ARIETTA ENERGY SYSTEMS LIBRARIES VA Tk MOT TRANSFER TO ANOTHER PERSON < g 4 i niame with document You wish some Hse § this decuf o weone aise fo see s document, e PP E o Mol and the tbeary will This report was prepared as on account of Government spensared wo_l:k; Neither the United States, _nor the Commission, ner any person acting on behalf of the Commission:. L . A, Mgkes any ‘warranty -of representation, express or implied, with respect to_ the :‘:o’crcg;acy‘l,_ completeness, or usefulness of the information contained in this report, or that the use-of any information, apparatus, method, or process disclosed in_ this report moy not infringe .prlvcrr.el.y owned rights; or A 7 . . Assumes any liabilities with respect to the use of, or for.,rjumuges_‘resulf‘ g from any”informqhon, appeoratus, method, or process disclosed in this report, ‘ o _ As used in the_above, *‘persan acting on behalf of the Commission’’ includes any employee or contractor of the Commission to the extent that such employee or contractor prepares, hondles or distributes, or provides access to, any information . pursuant to his .employment or contract with the Commission, . i et T - A ; This decument contains e I% * o S P : OAK RIDGE NATIONAL LABORATORY S o R 1N ' B . Operated by - - | \J ' ORNL-2157, Parts 1-5 ’ o o . . -84 Reactors-—Specwl Features of Aircraft Reactors < _‘" . o _ This document consists of 330 pages. _ " ' ‘ : | Copy Ql/of 237 copies. Series A, t ) . = Contract No, W=7405-eng-26 i . | AiééR}AF'f NUCLEAR-PRdPULSl-ON PROJECT - | QUARTERLY PROGRESS REPORT " ) | For Period Ending September 10, 1956 | o W. H. Jordan, Director - o ‘ .' o _‘-f | L S. J. Cromer, Co-Director A. J. Miller, Assistant Director l.-’ - - DATE ISSUED DEC 13 1956 o ’ - T UNION CARBIDE NUCLEAR COMPANY , - A Divlslon of Union Carbide dnd Carbon Corporation . o _ Post Office Box X~ 5 o , -~ Oak Ridge, Tnnnessae g [ED DATA Y .defined in the Atomic Energy Act of 1954, lts transmimel B ure of its contents in any manner to an unauthorized pers® oL | l'edc Reports previously issued in this series are as follows: - Period Er_l.d_i'n_/g November 30, 1949 o Period Ending February 28,1950 . ORNL-528 ORNL-768 ORNL-858 ORNL-919 ANP-60 ANP-65 ORNL-1154 ORNL-1170 ORNL-1227 ORNL-~1294 ORNL-1375 ORNL-1439 ORNL-1515 ORNL-1556 ORNL-1609 ORNL-1649 ORNL-1692 ORNL-1729 ORNL-1771 ~ ORNL-629 v ORNL-1816” ORNL-1864" ORNL-1896V ORNL-1947 ORNL.-2012 - ORNL-2061 ORNL.-2106 Vv’ Period Ending May 31, 1950 Period Ending August 31, 1950 ~ Period Ending December 10, 1950 Period Ending March 10, 1951 Period Ending June 10, 1951 Period Ending September 10, 1951 Period Ending December 10, 1951 Period Ending March 10, 1952 Period Ending June 10, 1952 Period Ending September 10, 1952 Period Ending December 10, 1952 Period Ending March 10, 1953 Period Ending June 10, 1953 Period Ending September 10, 1953 Period Ending December 10, 1953 Period Ending March 10, 1954 Period Ending June 10, 1954 Period Ending Septémber 10, 1954 Period Ending December 10, 1954 Period Ending March 10, 1955 Period Ending June 10, 1955 Period Ending September 10, 1955 Period Ending December 10, 1955 Period Ending March 10, 1956 | Period Ending June 10, 1956. 43 ¥) ORNL-2157, Parts 1-5 C-84 - Reactors=Special Features of Airéruft Reactors k® INTERNAL DISTRIBUTION oy, . B. Li-ndcuer 1. R. G. Affel 50. R 2. C. J. Barton 51. R. S. Livingston 3. M. Bender o 52. R. N. Lyon 4. D. S. Billington 53. F. C. Maienschein 5. F. F. Blankenship 54, W.D. Manly 6. E. P. Blizard | ' 55. E. R. Mann 7. C. J. Borkowski 56. L. A. Mann 8. W. F. Boudreau 57. W. B. McDonald 9. G. E. Boyd - 58, F.R. McQuilkin 10. M. A, Bredig _ 59. R. V. Meghreblian 11. W. E. Browning ' 60. R. P. Milford 12. F. R. Bruce. , 61. A. J. Miller 13. A. D. Callihan : 62. R. E. Moore 14. D. W. Cardwell ‘ - 63, J. G. Morgan 15. C. E. Center (K-25) ' : 64. K. Z. Morgan 16. R. A. Charpie . 65. E. J. Murphy 17. C. E. Clifford 66. J. P. Murray (Y-12) , 18. J. H. Coobs -67. M. L. Nelson L o= 19. W. B. Cottrell . : 68. G. J. Nessle LT 20. D. D. Cowen » 69. R. B. Oliver 21, S. Cromer - -~ 70. L. G. Overholser . 22. R. S. Crouse 71. P. Patriarca 23. F. L. Culler 72. R. W. Peelle 24, J. H. DeVan 73. A. M, Perry 25. L. M, Doney 74. J. C. Pigg fi 26. D. A. Douglas 75. H. F. Poppendiek 27. E. R. Dytko | 76. P. M. Reyling * 28. W. K. Eister ’ 77. A. E. Richt 29. L. B. Emlet (K-25) 78. M. T. Robinson e 30. D. E. Ferguson 79. H. W. Savage ‘ 31. A. P. Fraas 80. A. W. Savolainen ! 32. J.H. Frye ' . ' : 81. R. D. Schultheiss 33. W. T. Furgerson 82. E. D. Shipley 34. H.C. Gray | 83. A. Simon 35, W.R. Grimes : 84. O. Sisman ‘ 36. E. Guth- | - - 85. J. Sites ; 37. E. E. Hoffman ' - : ‘ 86. M. J. Skinner 38. H. W. Hoffman . - o 87. G. P. Smith 39. A. Hollaender =~ 88. A. H. Snell I 40. A.S.Householder - - 89, C.D. Susano | ) 41, J. T. Howe : / 90. J. A. Swartout 42, ‘W, H. Jordan R | - 91. E.H. Tayler o 43, G. W. Keilholtz | 92. R. E. Thoma 44, C. P, Keim _ 93. D. B. Trauger ‘ 45. M. T. Kelley B 94, E. R. Van Artsdalen . 46, F. Kertesz : 95. G. M, Watson -/ ' 47. E. M. King | 96, A. M. Weinberg 48-49, J. A. Lane 97. J. C. White iv 98. G.D. 9. E.P 100. G.C. 101. J.C 102, C. E Whitman igner (con sultanf) ilson Wi Willioms W Winters 124. 125. 126. 127-129. 130-131. 132. 133. 134. 135. 136. 137-139. 140, 141, 142, 143, 144, 145-150. 151, 152-153, 154, 155, 156. 157. 158. 159. 160. 161. 162-165. 166. 167. 168. 169. 170. 171. 172, 173. 174, 175. 176. 177. 178. 179. 180. =—==8FORET= L 103-112. ORNL - Y-12 Technical Library Document Reference Section 113-119. Laboratory Records Department 120. Laboratory Records, ORNL R.C. 121-123. Central Research Library EXTERNAL DISTRIBUTION AF Plant Representative, Baltimore AF Plant Representative, Burbank AF Plant Representative, Marietta AF Plant Representative, Santa Monica AF Plant Representative, Seattle AF Plant Representative, Wood-Rldge Air Materiel Area Air Research and Development Command (RDGN) Air Technical Intell:gence Center | | Allison Division ANP Project Office, Fort Worth Albuquerque Operations Office Argonne National Laboratory Armed Forces Special Weapons Project, Sandia Armed Forces Special Weapons Project, Washington Assistant Secretary of the Air Force, R&D "Atomic Energy Commission, Washington Battelle Memorial Institute Bettis Plant (WAPD) Bureau of Aerongutics Bureau of Aeronautics (Code 24) Bureau of Aeronautics General Representative Chicago Operations Office Chicago Patent Group Chief of Naval Research Convair-General Dynamics Corporation Engineer Research and Development Laboratories General Electric Company (ANPD) Hartford Area Office Headquarters, Air Force Special Weapons Center Idaho Operations Office Knolls Atomic Power Laboratory Lockheed Aircraft Corporation (Richard G. Rowe) Lockiand Area Office ‘Los Alamos Scientific Laboratory Mound Laboratory National Advisory Committee for Aeronautics, Cleveland National Advisory Committee for Aeronautics, Washington Naval Air Development and Material Center Naval Research Laboratory New York Operctions Office North American Aviation, Inc. (Aerophystcs Division) North American Avigation, Inc. (Canoga Park) e S iitrs=S wy s o 181. 182, 183, 184-187, 188, 189. 190. 191, 192, - 193210, © 2114235, 2360 237, Nuclear Development Corporatidn.of America Office of the Chief of Naval Operaflons (OP-361) 'Patent Branch, Washmgton Pratt & Whlmey Aircraft Dlwsmn (Fox Project) Sandia Corporation ‘Schoo! of Aviation Medicine Sylvania Electric Products, Inc. USAF Project RAND University of California Radiation Laboratory, Livermore ‘Wright Ait Development Center (WCOSI-3) Technlcal Infermation Service Extension, Ook thge “Division of Research ‘and Development, AEC, ORQ Technical Research Group, New York flw i iy wi L) FOREWORD This quarterly progress report of the Aircraft Nuclear Propulsion Project at ORNL records the technical progress of the research on circulating-fuel reactors and other ANP research ot the Laboratory under its Contract W-7405-eng-26.. The report is divided into five major parts: 1. Aircraft Reactor Engineering, 2. Chemistry, 3. Metallurgy, 4. Heat Transfer and Physical Properties, Radiation Damage, Fuel Recovery and Reprocessing, and Critical Experiments, and 5. Reactor Shielding. The ANP Project is comprised of about 550 technical and scientific personnel engaged in many phases of research directed toward the achievement of nuclear pro- pulsion of aircraft. A considerable portion of this research is performed in support of the work of other organizations participating in the national ANP effort. However, the bulk of the ANP research at ORNL is directed toward the development of o circulating- fuel type of reactor. The design, construction, and operation of the Aircraft Reactor Test (ART), with the cooperation of the Pratt & Whitney Aircraft Division, are the specific objectives of the project. The ART is to be a power plant system that will include a 60-Mw circu- lating-fuel reflector-moderator reactor and adequate means for heat disposal. Operation of the system will be for the purpose of determining feasibility and for studying the problems associated with the design, construction, and operation of a high-power circu- lating-fuel reflector-moderated aircraft reactor system. vii Ay wi CONTENTS FOREWORD ......ooocceeerreessesesessssscssesssessassssssesessessesesssssesseneesossnes SRR e vii < SUMMARY ...c.cccenrn e S, ] PART 1. AIRCRAFT REACTOR ENGINEERING 1.1. AIRCRAFT REACTOR TEST DESIGN .....covevrrmeiiieirrnirnnnenae eeeerereeeeesreeeaereertebtenaereseteanerneas 17 Status of ART Design....ccoeccerererreenincmniniinsnienan, ettt ettt a et et A s s bt eb e be A eras b es s re e sn s s saees 17 Reactor Assembly ......ccmvereevernienriinniicicsnsinineenaes e e areseaas eeeveseaenenetereeetansaaasntnes 17 Reflector-Moderator Cooling Circuit .. ivirmmsmmriincreerericnisssitsssissssssss s 17 Heat Exchangers .......ccccvcennniccricnnncinnsennisnnns eeutetebee st eaearte st seaRent s a et et s e et s e st e e eae w17 Fuel Pumps and Expansion Tank.........cccoiininiiiiitneseneseiese s s 17 Fuel Recovery Tank......ccoovvrrerrrenccencicninnencans eeestetetes et e bttt esensere R e bt eseanatares et 17 Applied Mechanics and Stress Analysis ............. e tai R AR AR 18 Reaetor SUPPOIt ...ttt ssree st mass s st st ne s a s st s s nesans e es e eaneenes 18 NaK Piping Inside Reactor Cell................... rerereenens eeteuereearebatesaeaereseserasstetebetasastasesaeaeatassstteearare 18 Fill-and-Drain Tank and SUPPOTT .......ccviireierenrictcnerreeecces st sss s snenssssssasssasssssases 21 Core-Shell Low-Frequency Thermal-Cycling Test................. eveseerensereettebatesatate s et assesensranesneater 23 Core Hydrodynamics ...t et b b s s sa st 26 - 1.2. ART PHYSICS ..ot eteeeeeeseattaeatestasssas et h A a e R e R e e e b e bR e s b ettt nb s s 29 Comparison of Basic Gamma-Ray Data with Bulk Shielding Reactor Gamma 5 Heating MeaSUIEMENTS .....ccoruiccrrrecresiecesircenseenseraseressesssssssassansses s s ssas s esb st ses s assssasssss st sesssnsueres 29 - Gamma-Ray Heating in ART ........... eevensrereneaenns errenenene ieereereeeebeeetesstesetateseeteteaner s et st eresnsasa bt b 31 GAMMA-Ray SoUrce SHENGHh ..ioviiecerrcererreecerseeneseserssessesiasesssssesssssssssssssresssssss s smsbessssssassasassasess 31 Gamma-Ray Heating ......ccovvcvnveimemminniininnn, enurpibine s ssatssnssesens reebeeeetebeertete e e asae e s ssanens 32 1.3. ART INSTRUMENTS AND CONTROLS ........ccccruveneen. ceeeteeetetiessasabenebete e e e At e s e aeanre e asesneneneater st bebetn 35 Reflector-Moderator Temperature Control Simulation........ocecumrecmmiveimiiisnniinssiiesns e 35 Instrument Development ......ccccevvcrreinrnciniesscnerineanen. ereeeeeatetesatasasararaserarasababesetassataeaebeteae et e nsaseasaneen 35 Fuel-Expansion-Tank Level |ndlcafor ............................................................................................ 35 Systems for Testing Liquid-Level-Sensing Devnces in Flowmg Liquids..ccoecnrnrncneenccenencnnas 36 High-Temperature Turbine-Type Flowmeter ... nsisssesssses 36 High-Temperature Pressure Transmitter ....cooveeencicccnnninnannenecs evererearaeserbasastetar s easnarasasn 37 Tests of Heliarc-Welded Inconel-Sheathed Thermocouples in Sodium ......covuviiniiiiniinininiinnnn, . 37 Thermocouple Data Reductlon Coiesessarassasasaissseesseissssatesnestasstereseseeastsesrsssssnats et . 37 1.4. COMPONENT DEVELOPMENT AND TESTING.........._ ........................................................................ 39 Pump Development T eSS iorimrersimseeressesssesssssssssasessens pebset e e sae st s b a s et an s reen rerrrsensnaseserenes .39 Bearing and Seal Tests ........... eeeeerereresr et ata et e e s s ses s bR e s srnees revetereeesae e e et e sbearaeaten 39 .- Test of Gas Attenuation by Seals seiesibsireresseietiressntse s et R s et et be s e st R et bR SRR E b e R e s e P SRR SRS 00 39 ; Fuel Pump Development Tests....ccocininninnniiine. ettt es rrnersessessoseseasiiibababea s sr s s s ansens 40 ‘ Fuel Pump Endurance Tests .irennnnnnnessnisiessesesascnsannnssncnns sesssemssssessinessarsssnsssisasnenias 42 e ~ Sodium Pump Development Tests .c..rireimnrmsnisscsserssssssisssssnsssssssrsesssersens reensveesereststesesesesteneneas 43 | Primary and Auxiliary NaK Pump Development Tests covnnnrrrrenens et st 43 l Heat Exchanger and Rodiator Development TEStS ............ccuwrmmmmimsssssessesssssssssmssssssssssssssssssesessseees 43 N Intermediate Heat Exchanger Tests................. eerst bbb se st sba st reen et sensiaes 43 Small Heat Exchanger Tests....cooeevreninenininnicsieiiiesinnenens rereereerenns evreerrereetesetesaaaransens 47 Cold-Trap Evaluation in Heat Exchanger Test Loops ...t 51 1.5. 1.6. 1.7. 2.1. S, 5 chtt p . v . Water Flow Tests of Aluminum North-Head Mockup..... et st 52 Dump Valve Development Tests ....ooeovecncccenncnne Cemeeeeneseedaess s iviseeeeenesieessirenns 53 Outer Core Shell Thermal Stability Test ..... 55 Auxiliary Component Development ........ccocooneerccncrrcnnnnene rereterenetarneserer s re s ea sy tassesansenors reesesnssaessansen 56 High-Frequency Thermal-Cycling Apparatus ......ccouiremvinermrssinssisinssississiistssetsssesesssessseseseser - 56 Cold Traps and Plugging Indicators ......ccvvccnverecnnrisninensennen: cnsnsnsanes reerneesearieraras rereresnes - 57 Liquid-Metal-Vapor Condensers ..........ccceemeercennnersnnesennieescsiseessseresesesssssssssssesssssasesasaesssssnssassenss 60 Zirconium Fluoride Vapor Trap ....cccocccinvmmnnncnininnsnenincnneressssnsssencsenes resreniossarsasensesasninsssrseare 61 PROCUREMENT AND CONSTRUCTION .....oouvireereisassesssesiessms s sasessssssssressssesssseressnnes reeverensee S 64 ART FEITIY wecverresrrseneeseessssssessssssesssssassasssssesessassssssssssssssseassssssesisresssesessassssssssssssescsseneers v 64 ART-ETU Reoctor Construcflon reesratsnsaeae st gttt re s s st etseresnaresanaseeres prerruenansnanes ereresrereensennnessesesesesss 18 ETU Facility oot crcrssscesiessss e snessssesnssesssesens eeseenereranannees veresersnsenensrnitearsassnons 78 ART-ETU Reactor Fabncahon and Assembly ...... ebensamsarersaeasasptste et sbssrantitsrsasnassrerenansne resesseennene 18 “ART Cell COMPONENES ..cuecvireenecerversssesiisesensaeressassssssesesssnssossssesssassons rreerersesseenes 79 ART-ETU Reactor Component Procurement .........ouo.vccciveesecnorssnnaens eeiereretsae sttt r s s sen e sasbene 79 ' Fuel-to-NaK Heat Exchangers ..........cuerenncninrcsteeneernsssseescsessssssesensens ereretensnenesbeseserenns 79 Sodium-to-NaK Heat Exchangers ...t reversensaressasesrsasans 79 Core Shells ..ottt st sassseresasseneesesassessonsrenasserasnens treeaneateneemaeneanaanastenasresens 79 Beryllium Reflector-Moderator ...........ciinmicnincncnemssmsssssessess 80 Boron Layers.......ooinieiecnnicienienisiaons reemsaseeesanresereasessass eeernerse et e rmensanssrreas rene 80 Pressure Shell... e sssesesesessssassnesees eereseesasasneneaes aerneseanisaenen eenenens 80 FRCONEL ettt et se e e ra s s s s s e n e nsase e SRRt s et b s b SRR bR 80 IN-PILE LOOP DEVELOPMENT AND TESTS.........ccooosmmmsercesomssrsssmsssssssssssssssssasessnes 1 Operation of In-Pile Loop No. 5....coeuereneneneen. s a1 R4 e s RS 81 In-Pile Loop No. 6 e ceeeeenne reteeeereneennanens eeeseeesesarerteresnerrrensettenresenesaasereastsbteraanes 81 Horizontal-Shaft Sump Pump for In-Pile Loops .......cccovreeerereeercrnrenennnnnens st snsisnsnenesesgassons 81 ADVANCED REACTOR DESIGN .......covceeeireiricrrssressessesaeseennes . ....... st snses veerenenes 82 Graphite and Beryllium Oxide Moderated Circulating-Fluoride-Fuel Reactor ...........cuiveuiureccernee .. 82 Hydrlde-Moderated Circulating-Fuel Reactor - 84 PART 2. CHEMISTRY | PHASE EQUILIBRIUM STUDIES ....cvroooovvornecmecnssecnscssssesnmessesns s eeeesrsrersssneesmssessssnseneess 8T The System LiF-RbF ..o veeneenens eeenenesene e sneanes tmserarannasonssaggresgnessienin 87 The System LiF-UF, ..o s emee e esesern e ren s eneemrasaeee peemunsaesaonsasin 87 The System RbF-UF, ............... et erdsbe s anrn s bbb e siee e siens 88 The System NaF-RbF-UF, .........cccrmmrsmreercsreersene eesensn e snsisn st 90 The System NaF-RbF-ZeF [ -UF ....ccccoeeorresnncesessnsessressssss s ssmsssssssssssssssssssssssssssissssssses — 91 The System RbF-CaF, o, rrerrenesreanareeraeetisaeses dersrsenses reeterteesrnneserarsarenaasans 91 The Alkali Fluoride~Cerous Flucride Systems ........ccocovccreviunnecirnnnnne ,f’;’{ ........................................... 92 The System CSF-BeF .......cooommmmmmererrcessmmmasnermeessemsnsnesssescecs ersurssn st RRSRS R ORARSO 92 The System NaF-LiF-BeF , c.c ettt sss s st s s 93 The System NaF-RBFBEF 5 oottt - 94 The System KCI-ZrCl, .....ovmrmmersssssssmsnssisssssomsssssssssississsnsssssmssssssmsssssmssssosssssssssssssmssosassssirs 9 ~—EORET — L n £) <\ o) 2.2 2.3. 2.4. 2.5. 2.6. CHEMICAL REACTIONS IN MOLTEN SALTS ....... et bRt er bR n R A bbb aee e s eas 96 Equilibrium Reduction of N|F by H, in NoF ZIF firermssssrmossssscssssesmonssrasstesmsasssnenassssssssecs weevrersaers 96 Stability of N{F2 in NaF-ZrF4 Mixtures ......ccocevevvrnane. veveerrrenens ........ 97 Activity of Chromium in Chromium-Nickel Alloys ... 100 Solubility and Stability of Structural Metal Flucrides in Vanous Fluoride Mixtures ......co......... 100 Reduction of UF , by Structural Metals ...ttt 103 Solubility Defermlnatlons by Measurement of Electromotive Forces of Concentration Cells .................. eeebsteasuenireanrsstetebontaeataeaasesstenenasbeberteastiseranansenseseseesasnsasans reerorsnenesnsanne 106 Activities in the RbF-ZsF | System e 108 Activities in the KF-Z:-F4 SYSEEIM ettt r e et e st ene e sbe e s ae et e st s ar et e aa b e e aeten 109 Solubility of LaF ; and of CeF ; in Molten Fluoride Mixtures ..o, 110 Solubility of Xenon in Fused Salts ...ttt se st sse s sesn st e e snesane 113 Molten Salt Polarography .......c.ccveiniecrncieecerrcnenannns e etreeeeterebeteieattereretetehes sttt bebaresesear st sebetetn e 113 Activities from EMF Measurements on Solid Solutions of Salts....c.cccoooiiiieicenniciriccceeenis 116 PHYSICAL PROPERTIES OF MOLTEN MATERIALS ...t s s 118 Surface Tension and Density of NaF-ZrF ; Mixtures ...ccoovrierrrireinnieennen e evenreneenererereas 118 Surface Tensions of Molten SAlts .t es 118 Fluorozirconate MIXBUTES .......coccoiiiereeieee e es st sttt sare e s e se s s ses e s et 118 Uranium Tetrafluoride .......coveeeeeerccrerreecrceereene, e eeterteetenrerearteintteare e e et et ereesteretensan et e e reneereereans 119 "PRODUCTION OF PURIFIED FLUORIDE MIXTURES ....ocmmeeseicirecteeeseesras s eeenesenssestasenseensseans 120 Laboratory-Scale Purification Operations .........cceereiorinieseneesnssessensascessssssscssesssssmmessesesssssssanss 120 Quality Control of Raw Materials and Products ..........coomeeeeeerererssesnersssessmmsensesssssmsssesssssseesssssssanns 120 Pilot-Scale Purification Operations ... iesiossvesessesssssssssssssssssssessasssessssssssssessssnses 120 Production-Scale Operations .......cccceeeveeemiresieinneesseesstessssssesssessrsssssssassssemssssssssnsassssessssesessansasesnsasas 120 " Batching and Dispensing Operations .................................................................................................... 121 Preparation of ZrF from ZrCI ................................................................ 121 SPECiGl SEIVICES uueurenirirerrrrnisereiereeesenesessnsisssessesssenes eeeeeeeresseseesemesesesa e sseesssseeroen rreeteesbsasasaen 121 Filling, Draining, and Somplmg Opefuhons eteeieetereesnesreseeraebeveeae it s et ebea e st a e e b et e s e te s ea s s g eraaes 121 Special Enriched Fuel for In-Pile Loops .. ssa s sessaas s s 122 Shield Mockup Core Materials ..o rieieeiecrreraeisetesiesesenaesivesssssisssssessesassssessssssssssssssssnssnsans e 122 Experimental Preparation of Various Fluorides 122 COMPATIBILITY OF MATERIALS AT HIGH TEMF’ERATURES ceveressresas s anbeneaas — 123 . Penetration of Graphite by Molten Fluarldes ...... etvre e e s e e e bt s srasasae spess ettt reneranes .. 123 Hydrogen Pressure of the NGOH-NI Reactlon et seeeaane rersieiaeeensasntsasasisanserne . wieeirensissentsessennsenenes 123 'Physical Properties of Elastomers Exposed to Attack by Liquid Mefals ...... eenssmsssnsssresssnerenensene 124 ‘Determlnahon of Oxygen in NuK .............. Laressssaninssisd s anssssanass sesssssmmenssssanamaen - eresstseiss e sasntrass Y ANALYTICAL CHEMISTRY .o covooesressoesssssssasssossssssssssssssss st s ssssssasessssssssississssessses 125 Determination of Rare-Earth Elemenfs in Fluonde Fuels ................................. et 125 Spectrophotometric Determination of Cerium .........oovueveeeeerercrerereresnesessisestiessscnensenens reerreenenes 125 Spectrophotometric Determination of Lanthanum ... 125 xi 3.1 3.2, 3.3, Nickel-Molybdenum Alicy Development eeeteeeessessees e e et se s easesasseresaaes et eaesresbenens eeeraresnssreressesnans xii Determination of Tantalum in Fluoride SIS ... iiiiieciecittinnnitscsssiesesssssesrssssssseissssssnsaesansersans Determination of Oxygen in NaK ... ssesssssesssesnes reveneeneenes eree - Sampler for Alkali Metals ... res s eseos Determination of Oxygen, Nitrogen, and Carbon in Metallic Lithium ......ccccccvvvnncennncee. Determination of Oxygen ..o ceiiccenrenienenessenisnsereresesaesssasassnsssssosssessassenes retenraensaensnras Determination of Nitrogen and Carbon.........cccvviccniinnniisnninnsisnisecerserenenens vrenssensassases S Detection of Traces of NaK in Air ........ccu.......... et e R s RS cresrrrnssesaas Compatibility of Fluoride Salts and Alkali Metals with Pump Lubricants ...ccconiivcieinirciennne. Determination of Chromium in Trioctylphosphine Oxide Extracts with DiphenYI CGrbaZide F e R e E e e R e ettt r e et e ra R et blasLuRiT IRl adtisRtIRsaIIa .--oocooaun---hl----n---; -------------------- ] ANP Service Laboratory .....ccooeerreecencvesneeenenencenenes reeereeere e e arasenes eevnsnassasesasssasns reeseerranenes ereeranane DYNAMIC CORROSION ..covervtrrrvsmerssesssensssisesnsnisesnes e Forced-Circulation Loop Tests .....eciecrnininierenieccnsessinisassssssessseresesessensnes cevesusesreasseasanessasrannans Fuel Mixtures in Inconel and Hastelloy B ........cccoviiiincicnnnnnnns reeeereeaas eresereserrrerasersneasssrassanens ' Sedium in Inconel and Stainless Steel...........ocvveriinnnnnnnn. aseereentn s rn s stessassnast e sa s rus b S RA S st 008 Sodium-Beryllium-Inconel Compatibility ....c.ccccoeimveverirninirinnenininnns ceeeesstetes st sreranenans R Thermal-Convection Loop Tests .c....cciiieeecicerecncresncsnenns ereerrerereanaeeeanesineesaseses ceeesseisenaerainine Sodium-Beryllium-Hastelloy B Compatibility ............... vernrrenns ceereteresesesesatentanesserenenasenesierasaesrens ‘ Sodium-Beryllium-Inconel Compatibility ......cccriniiveimsinniinniinccisn s s Fuel Mixtures in Hastelloy and Special Nickel-Molybdenum Alloy Loops............. rersnsinnensoserae Special NaF-Z¢F ,-UF Mixtures in Inconel Loops ... irrrnnc st snenssesesssssssssesssees Screening Tests of Special Fuel Mixtures ... e INCOnE] SHrAinN-Cycling TestS.vneinriirenerereeeseseissesssstsisassssstsnssesssassessssasssassesessssesisabesesntseessssensassasnens GENERAL CORROSION STUDIES ........covetirenrcerennnereresesesessssasssssssssasssenes _ Tests of Inconel Tube-to-Header Joints with Recrystallized Welds in NaK and IN FIUOride FUuel ...t v st s s sase e s e b s b a s sm e st et Tests of Coast Metals Brazing Alloy No. 53 in NaK and in Fluoride Fuel .cccovvvrennnee Lovrssreraseass Tests of Haynes Brazing Alloy No. 40 in NaK and in Flucride Fuel ........cccoveciicnnnnennnenesennncn. Niobium in Sodium and in Lithium ..coee........ rveeesertet et tetaanbebens s s aneas _ Vapor-Zone Attack in Inconel-Fluoride Fuel Systems ......oomicieieiinsivciniinninsnsinsniesnssnsssennenes Carburization of Various Alloys by Molten Sodium ..ot srores Sodium-Beryllium-Inconel Compatibility in a Static System (Test No. 2) ..cccocvrmrernccsinrerecenenne. | Effect of Diffusion Cold Traps on Mass Transfer in Inconel-Sedium ‘ Thermal-Convection Loops .....uiircicnireneninnssessssesess s sns st cssssensesssesesssassssessnasasarens seesssnnes Thermal-Convection Loop Tests of Various Structural Materials and Lithium crasssensesrrassssasastrsssens Tests of Intermetallic Compounds in Sedium and in Fluoride Fuel cerveeeesieseeseeress s e asene e st eaneeteens FABRICATION RESEARCH ..ccovieserseereensssepsssesissesesesssssessenessse st s sesssssssssssssnss s Battelle Memorial Institute ALIOYs ........cccceeeeenreserernreneerensnsesisiisesesstssnensssssesesesssasssssasnnsssssssssssns 152 153 1} iy T wr) 3.4. 3.5. - 3.6. 3.7. International Nickel Company AlloySs ..c.ccooviveerioicrnneiecct s e reetrennine . 176 ORINL AHOYS .ottt ee s st atenesnsssessrasanessesesasssasasassestessonsn eteeereenesnnresnssessensisersases 170 Hastelloy W Seamless Tubing.......... ettreeteeabiarebereeete ey e re ke ehe b e s b et e be e e e neetRe et s s e bt en e e sate e bR s EsRe e 177 Consumable-Electrode Experiments.........ccovcmicnnninnissnssss 177 Shield Plug for ART PUMPS c.ccoveserrecsereecmsressneesssssesssssesisssssssssssssssssssssssssssssssssssssesssss s s w177 Gamma-Ray Shielding Matenal .................................................... trvevesranererrenensanes eserssenrerssmansaaraents 177 THEIMAl SHIEIA .ooveeeeecre ettt e ersas b e ss s s an st s e eas seasrae st bsbsseneasane 180 Neutron Shield Material for the ART ..ot eeerraereeaebetere e e enseannns 180 Ceramic B, C Tiles oot reveestessasesseressesereseiateteeeaaasaeaetere et st sRsebesnebase s 180 Clad Copper-B4C COIMELS ..c.cevereeericrs e snecoressre e b s bbb srm R bbb bbbt b 180 Lithium-Magnesium ATLOYS 1ooveoresees s iees st v sssss s sses s ssse s es e bes e s aR SRR 180 Tubular Control Rods ......cccverunen. esessssansssssrmssesnees rerietebetastasbet st e b sa st e R s be et ere e b st R bR R 181 Seamless Tubular Fuel Elements ceereeeeneeenee e ieerb b st e e R e ba A b R b 182 Niobium Fabrlcahon Studies ..ccerrrneenene reerueseessteseereetesbeteaseaaaeanseatraeaaaeet e e eRaee et e s e e naseesanenbarbe b sorsnaeas 183 Evaluation of Arc-Melted Niobium .....c.ooiviiriicrscinceneeese et st sasees 183 Nb-UO, Compacts.......crerrsmreernnens eeteesveseeeeneesaaettsasienerent e et b st e oReR st e e e A eR TR RO A ae e ehaas et e eseREatese et saeranen 183 Nb-U Alloys ........................................................................................................ 183 Fobrlcahon of Hydrides ..o 183 WELDING AND BRAZING INVESTIGATIONS ...oooooos oo esecmemenessssessssssmseesesssssssmssesssnssess s 184 Fabrication of Primary NaK Pump VOIUBES wvvvvvvereeseeesceesesesessseeeeeeesssesesesesesesasesssessssssossssmssesnesenssess 184 Fabrication of Primary NaK Pump Impellers ... 192 Shrinkage of Inconel Core Shell Welds .........occcveciiniiicicniice e 192 Examinations of NaK-to-Air Radiators After Service ... 195 Fabrication of Cermet Valve Components .......ccoiicrinniiiniininccinine st sesersssnssssesssssssessns 197 Continuous Production of Sintered Brazing Alloy Rings ..oveeeeceicciicee eeraeeeanas 200 Fabrication of Heat Exchangers and Radigtors ... 201 Studies of the Aging Characteristics of Hastelloy B, 202 MECHANICAL PROPERTIES STUDIES .....oocovv R e e 205 Relaxation Tésting of Inconel.......... eeeeseeee it taen et r s s sttt R et e R R st enees reeneesestetennanssasasanane 205 Effects of Fused-Salt Corrosion on the Siress-Rupture Properties of Hastelloy X .....ccccovvnnnneee. 206 | Creep-Rupture Testing of an 80% Mg-—20% Li Alloy .......................................................................... 208 CERAMIC RESEARCH ............. rerereesenas eeeesesesssiereniensasains rrrerreeneanes rerenrens pevseseaessrenaden s sanesnaces vereeeennieee 211 Rare-Carth Mutenals ................ reeeeeeene 211 Fuel and Moderahng Matermls ettt eninn s snsaaes . ..... eesaeeensenetaanens 21 Component Fabrication............. eeesennenaes ~.....;._.’ ...... rereseesieessenestesananens s reveresessssvenesosssinssenns 211 High-Temperature X-Ray lefructlon Vesresiasesioieans Vivassasemsnsissasaseenitieattisas s st e et b e o sRsiabara bt an s 211 Flyoride Fuel |nvest|gcmons...'.r_......_...'..'.'.'........' ..... s ' ............ 211 NONDESTRUCTIVE TESTING STUDIES ....cooemrrrenrerennnisnsssessesssneseennee eeesenes s reevemmeeeessaens 212 Eddy-Current Inspection of Small-Diameter Tublng SO UT 212 Inspection of Tubing by the Ultrasonic Method ....... eveerestetetebererats et sersRoReanbed s At st et st et s en s s e se e s ben 212 xiii 3.8. 4.1. 4.2. xXiv VLUHLE -~ 218 INSPECHON OF PiPE .rveruriierercerieseensesesenessssnssessssessasmesssemsssnssasssssesss s e ssons besassastassssassans shemssassssssessases Inspection of Thin Sheet .....cccvvninieessininennncrcssnneesenseresierssssesensas peesmeresressaseerserensenens revirane yereesenannas INSPECTION OF MATERIALS AND COMPONENTS.......cccccvieeenvnernnens ........ ereenneieenennesnes | Weld Inspections ..ceeeuercemrreetnrscncn s sissascnanes ceeeeseeeensenenneeretase s ss s R benaes Material INSPECHONS ...ttt s s b sasn s s san et s s eeeresssesessenesesentans InSpection of COMPONENES .....ccerueerinessnsesessssacsssssessssssesssssssiess erereseesnetenaereinertesasatenaseaneres veereereens - Fluorescent-Penetrant Inspection of Tubmg..,...; ...... .......... ersss s e e st R0 Welder Qualification Program st ersesersssees eeeereesiseraiin st rsssmsmiernas e egedisens PART 4. HEAT TRANSFER AND PHYSICAL PROPERTlES RADIATION DAMAGE S FUEL RECOVERY AND REPROCESS\NG CRITICAL EXPERIMENTS o HEAT TRANSFER AND PHYSICAL PROPERTIES ......cconnrenececenranrenne rveereraeestsrsaeneasrerensies ART Fuel-to-NaK Heat Exchanger e eatesesesssitesstesriaseRsRe e SeTT TS LA S RS A s s e bR R Rt - ART Hydrodynamics - Core Hydrodynamics......... ©uomesesesessreberthebs et ettt A A SRS R SRS A SR RS eE PR SRR SRR RO RO SHSS e ASE SR E SR SRRSO RS RO S RS Sodium Flow in Reflector Cooling Sysfern Ceeeeeteesessesstssesssstsesstssesteesteesessstesrerararasantsassassenenasnetes Fluid Flow Visuglization ........cceecincreiinnniinne et s evereeresreeneveienenssnes ART Core Heat Transfer STUAY ........occcoiereeeieenennieennrincssssessossssssesssnsessisisisesmsssssssssassssssssasssssnsasnssasens Thermal-C'ycling Experiment .....cccccveeveiicenas Heeresesenesseseheserasaetatat et at Rt e s e s e et ensaenaseapnase L RO RS e SaR RO RESS Shield MOCKUP COre STUGY «.c..vvuemmeeemsssssssremenssssssssssssssssssssssasessssas sesssssssssssenssssssssssasasssssasassesssssssssessesses Heat Capaity ...occceenecicnecerecrcsnecrsnes e oot sesesnsosssssastshssssss sanssasessesisssnssossanasssssasssnsesns eeeveenensarasns Viscosity and Density ... rcrmnrenerecemesecesssisssisstiss s e _ Thermal CondUctivity ....ccccveeenirienii et b s s e e s s sbr s ab e e st s s me s s b s as s e e s sanenansasessene RADIATION DAMAGE ... scverseseessassse e s s s ssasstssesass st ssssasesessassasasasesssbesasssrsssnosssasssssss Examination of Disassembled MTR In-Pile Loops Nos. 3, 4, and 5 eeseeses e seesassssmaneeren Investigation of Materials Removed from MTR In-Pile Loops Nos. 3 and 5. Creep and Stress-Corrosion Tests of Inconel ... MTR Stress-Corrosion ApParatus ......co..ccerernsssemmissssssssorssrersstssssssssssssssenssessssns reeeeensasrtessnseresassaasens Effect of Radiation on Static Corrosion of Structural Materials by Fused | : - SAIt FUBIS ..o reecverenertree e e reesarsee e sessassessasaessesesobabes a0 beonisbes e Re s R R P e Sa SRS SRS S e O R Ra ST RS S s an S e e s R e84 Holdup of Fission Gases by Charcoal Traps ....c.c.wiciennineeincnsesessissessessasesssesssssssssssnsess ART Off-Gas System lnstrumentation AnGlySis ..ot sesssssssnssasessenss LITR Vertical In-Pile Loop.............. reveeressseeneesesertirnretsenesaetasrabant s et A st e e nboabasansienebenaensasnas yoeresreasssensins Design Calculations for LITR Vertical In-Pile Loop eereeseseseressasbe b s R b se st R e et s asrsri e e e et s s e snnne ART Reactivity Transients Associated with Fluctuations in Xenon-Removal 7 EFFICIency ettt ettt as st ne resentraseenesanenannsnes resrenreneesense Use of Zr?5 as a Fission Monitor .....ommeeccersnmmissasssessecsesssecesnssses reremerans rerenerensasares rsassssasaraes " Fast-Neutron Detectors .......ccccuueen.n. Effects of Radiation on Electronic Components .....cvuvceenneininsinrenensinsnsissrsssisessseraesens emesaaeeresensnes Xy " -l 4.3. 4.4. 5.1, 5.2. 5.3. 5.4. Irradiation Effects on Thermal-Neutron Shield Materials .........memcrssmimmmssismsssnssrsnssssssss 256 . Irradiations of Stressed Shielding Materials .......ccccvuvicirrerereccciiniic s 258 FUEL RECOVERY AND REPROCESSING ...t creensstseneraesssssisssssassssssssssssssnsasassanesens 261 Volatility Pilot Plant Design and Construction ... sesssssessnsens 261 Engineering Development ...t sss st sesi s s st sssesa s 261 Chemical Development .........ccocoevviiincconninnnn s 262 CRITICAL EXPERIMENTS ..ottt cnininenns e ereteesserr b et er e bt ren et e esasasarenaaan 264 Reflector-Moderated Reactor EXPeriments ......c.cooeeiriesieieneicierecentisinnncsseesisisisessssssessssesssssssssssassnas 264 Experiments at Room Temperature ........coccewreiiimemiiisinisissisesssssssssassssssssssssssnssssananssssssassscsenins 264 Experiments at Elevated Temperature ...t 264 PART 5. REACTOR SHIELDING SHIELDING THEORY ...ooiireecemrcreteimisesseseneseseresssssasssaesasesassssss sessssenssstsssassesesessesssstsssssssss sresasasesssasasesenes 271 A Monte Carlo Study of the Gamma-Ray Energy Flux, Dose Rate, and Buildup Factors in a Lead-Water Slab Shield of Finite Thickness ..o 271 LID TANK SHIELDING FACILITY ottt resensnsestsesssan s ssssessasasssstes st sssssannssesens 278 Study of Advanced Shielding Materigls ...t 278 BULK SHIELDING FACILEITY oot srerescnesissesssssastesssssnsssssssssasasssssssesssnsssanessasesasasessessans 291 The Fission-Product Gamma-Ray Energy Spectrum ...t 291 Gamma-Ray Streaming Through the NaK Pipes that Penetrate the ART Shield......coccoveevnnenenne 296 SHIELD MOCKUP CORE ...ttt eeestsess s e sessess st ssssensssasssnassnssbsrssssnssasssnsassasans sassssss 299 Selection of Caleulational Models ... s s snens 299 Caleulational Procedure ... i st s s bt sesesessssssssststssasssasassss s sa s asas e s sasessbesesas 301 Calculations for SMC Configurations Using UO,-Stainless Steel Fuel Elements ........ccccoinennennec 301 Caleulations for SMC Configurations Using Uranium-Aluminum Fuel Elements .......ccoovviinnenee 305 Core GammA ROYS c.ocvveveeeeemrerireiensieencsemsine s sssressasass s s ssss st ess s ss s st sas s s ssssssssess saessassebasants 309 Neutron-Capture Gamma Rays from the Core Shells rerame s ae RS e s 310 Neutron-Capture Gamma Rays from the Reflector .........umoreereessmnmerserissssnssssssssrsesssssssins 310 Gamma Rays from the Heat Exchanger Region ..., 310 Neutron-Ccpture Gomma Rays from the Pressure Shell and Shield woeeeeecicreeeenenrereereeeesereeneas . 310 Neutron-Capture Gamma Rays from the Air and Crew Compartment ......c..coceeeemsiemsreresisssssees 310 Design STatUs ccvcereireeececreeenssnscssnscscssnessssnsssesessens reesmensesirelassesetensarsestssabene et seane s et tes s s st shsasenpereaee 310 Xv .«3} SEC ANP PROJECT QUARTERLY PROGRESS REPORT | SUMMARY PART 1. AIRCRAFT REACTOR ENGINEERING 1.1.i Aircraft Reactor Test Most of the detail drawings of the reactor-core, heat-exchanger, pump, and pressure-shell assembly have been completed, and design work on the re- actor shield and the interior of the reactor cell is proceeding, A review of the detail drawings indi- cated that the pressure drop through the reflector- moderator cooling circuit would probably be higher than originally estimated, and therefore larger and higher-speed motors are being ordered for the sodiumepump drives. A re-examination is being made of the tolerances and clecrances specified for the fuel-to-NaK heat exchangers in an effort to ease the fabrication problems, Tests of a fuel pump in a high-temperature test rig have indicated that the cavitation limit is somewhat lower than anticipated.. However, the increased cavitation suppression head required for satisfactory operation can be accommodated by increasing the fuel expansion tank pressure by about 10 psi. In tests of a fuel pump and expansion tank assembly no objectionable aeration or pressure fluctuations were observed in the system over the entire speed range when the fuel levetl in the fuel expansion tank was varied from 1/2 to 3in. The preliminary design of the fuel-recovery tank was completed, The tank has been designed so that no electrical, cooling, or other lines will need - to be connected to it at the time it is removed from _ the cell. Lo Studles of the reactor supporf structure were come pleted The complete reactor and shield assembly is to be suspended from an overhead bridge-like structure. The weight of the reactor will be frans- mitted to the bridge structure through the four pump barrels. To cllow for displacement of the reactor as a result of thermal expansion of the NaK lines, - the lines are to be precut so that their exponsion’ will move the reactor to the desired operating posis tion. Several bends have been added to the NaK lines to allow for differential expansion between the lines. The stresses to be expected in the fill-ond-drain “tank were also evaluated. The support of the fill- and-drain tank assembly is to be accomplished by means of a nitrogen cylinder located beneath the tank. A series of low-frequency thermal-cycling tests of the core shells is under woy., The thermal cycle being used in the tests simulates that to be ‘expected in the reactor. In the initial test the model was subjected to 300 cycles, a factor of 10 more than expected in operation of the ART, Visual and dimensional inspection revealed no gross failures. Metallurgical examinations are being made. Additional core flow studies were made on the full-scale model of the ART core. Heat transfer coefficients were calculated for the 0,010-in.-thick fuel layer odjacent to the outer and inner core walls for the core flow pattern established by using a swirl-type header with and without guide vanes, 1.2, ART Physics The results of experiments performed to determine the rate of gamma heating in vorious target ma- terials near the Bulk Shielding Facility (BSF) reactor were compared with results calculated by the method used for determining similar ART gamma heating rates. The good agreement between the experimental ond the calculated heat generation ~ values for samples near the BSF indicates that the “basic gamma-ray data and the Prott & Whitney gamma-ray deposition code used for the ART cal- culations should give representative results for the gomma-ray heat- deposihon in various parts of the ART, - Two-dimensional gamma-ray hecfmg calculations ' for the ART have been performed by using a code developed for the IBM-704 computer by Kneidler and Wenzel ot Pratt & Whitney. Source strengths ~ for - the calculations were obtained -from two- dimensional multigroup neutron calculations and from critical experiments. The total gamma-ray deposition (excluding that caused by sources in the heat exchanger region) in the reactor was found to ANP PROJECT PROGRESS REPORT be 3.86 Mw, which amounts to approximately 84% of the total source. This means that about 16% of the gamma-ray source energy escapes. 1.3, ART Instruments and Controls The ART simulator was used for a study of tem- perature control of the reflector-moderator and the temperature responses of the reflector-moderator cooling system. A control system was evolved which will hold the mean reflector-moderator tem- perature constant to within +15°F for step changes between zero, 50, and 100% design power and be- tween 1200 and 1425°F mecan fuel temperature (of 50% and zero power). Tests were continved on a helwm-bubbler type of level indicator for the fuel expansion tank, and three test systems are being constructed for study- ing other liquid-level-sensing devices under con- trolled conditions in flowing liquids. Additional tests of a high-temperature turbine type of flow- - meter have indicated that these units can be ex- pected to operate satisfactorily at 1500°F for 3000 hr. A unit with a 3‘/2-in.-dia housing for oper- ation in o system with flow rates up to 1400 gpm is being fabricated, Evaluation tests of several types of high- temperature pressure transmitters were continuved. The effects of out-of-pile aging on the calibration of thermocouples are being studied, and plans are being made for in-pile tests, Life tests of two synchronized high-speed mercury-jet switches for scanning thermocouple signals are under way. 1.4, Component Development and T_esiifig Evaluation studies of lubricating and cooling fluids for pumps were continued. Tests are being made in order to determine the compatibility of various lubricants and reactor process fluids and to evaluate elastomers for seal application. Preliminary test data were obtained which indi- cated that fission-gas leakage from the sumps of the ART fuel pumps would be attenuated by a factor of at least 104 from the expansion tank to the shaft seal and by another factor of at least 104 between the seal and the iubricating oil reservoir, . These data are being analyzed to de- termine the radiation dose to be expected at the oil reservoirs, Developmental test worl: on the. fuel pump im- peller continued. Modifications were made to the slinger impeller and to the seal plate over the centrifuge cup to improve performance. An en- durance test of a fuel pump was terminated after approximately 2000 hr of satisfactory operation. The pump was disassembled, examined, and re- assembled for further high-temperature testing. More than 400 additional hours of satisfactory operation had been accumulated by the end. of the report period. Water testing of on ART sodlum pump continued, Various configurations of the flow passages above the centrifuge were tested, and a configuration was found which gave stable operation of the pump in a performance-acceptance loop. Water tests were also performed .on lnconel ports for the NaK pumps. Data for the Inconel parts were found to be satisfactory when compared with data obtained previously with a brass impeller. York NaK-to-air radiator No. 9, the first 500-kw radiator of the revised design to be tested, was found to be in satisfactory condition after comple- tion of a program of 30 thermal cycles.. Endurance testing of a heat exchanger at ART fuel-to-NaK heat exchanger design temperature and flow condi- tions with a heat load of 400 kw was initiated, Two 4-in.~dia circulating cold traps equipped with stainless steel cooling coils were placed in opera- tion with no difficulties from oxide plugging. A procedure for startup of the ART cold traps was developed. Water tests of the aluminum mockup of the ART fuel system components of the north head were continved. Fabrication of a second aluminum north head is under way in which the sodium sys- tem components will be mocked up. Four ART prototype dump valves received from vendors were fested, but none met performance specifications for ART operation. Assembly and and welding procedures ore being analyzed, and seat materials are being tested in fused salts. Apparatus and techniques are being developed for investigating the effect of high-frequency thermal oscillations and the resultant thermal fatigue stresses on the ART core. A high-frequency pulse pump has been demgned for dynamic mixing of hot and cold fluids to provide a thermal-oscilla- tion frequency range of 1 to 10 cps and a surface amplitude of about 100°F. Various types of cold traps and pluggmg mdn- cators are being evaluated in test stands. The test stands are also equipped with Argonne samplers and Mine Safety Appliances samplers -so that -} L} ) chemical analyses of the NaK being circulated can be used to calibrate the plugging indicators, Liquid-metal-vapor condensers are being - de- veloped for use in helium purge systems in the ART sodium and NaK circuits and the NaK dump - tanks, High~temperature chemical absorbents and thermal-precipitation traps arte being investigated in the development of a zirconium fluoride vapor . trap. - The most promising system consists of a . bed of hot alumina which reacts with gaseous - ZrF to form solid ZrO, and AlF,, 1.5, Procurement and Constructren Work on the building aodditions, building altera- tions, and cell instaliation for the ART has been - completed, except for installation of six circvit breakers and the completion of modifications to the ~ cell floor structure. A rigid leak test of the re- actor cell was made; a zero leakage rate was ob- tained with a detector sensitive to 5 micron 3 /hr., Since the contract specification allowed 32 micron " #3/hr, the wvessel satisfactorily met the leak tightness requirement. The installation of aquxiliary services piping was completed, except for re- placement of 15 incorrect valves in- the nitrogen system, The electrical control centers and the spectrometer room electrical and air-conditioning equipment were completed. Work on the diesel generators was delayed because one of the gen- erators and three control panels were damaged in shipment. - Two ‘of the four main blowers were installed, - Program and design planning for d:sossembly of the ART were initiated. Procedures and tools necessary for removing the reactor are’ belng de- veloped, and. a new, Iarge hot-cell facnllty is bemg ~planned.. - s Constructlon work started on the Engineermg Test Unit (ETU) facility, Details of the main "NaK piping were determined and the design of the air duct for the NaK-ts-air radiators was prepared.’ ~The. study of the reactor assembly problem was continued, -and genera! assembly procedures were ,prepured. ‘Nearly .all the components and spare parts . needed for three reactors have been ordered. ) '_Fabncotion of the north head for the ETU reactor ‘is 40% complete._ The lower half of the beryllium reflector-moderator has been contouted, und dnllmg' of the coolant holes was started. The York Corp, ond Black, Sivalls & Bryson, Inc., are developing techniques for fabricating PERIOD ENDING SEPTEMBER 10, 1956 the fuel-to-NaK heat exchangers, The sodium-to- NaK ' heat exchangers are being fabricated and should be available soon. -The Hydrospin machine has been enlarged to produce the various shells required for the reactors, and the outer core shells for the ETU are being fabricated. ' Acceptable samples of finished boron carbide tiles were received from the Norfon Company. An ‘order has been placed for forgings for the reactor pressure shell. 1.6. In-Pile Loop Development and Tests Examination of in-pile loop No. 5, which was inserted in the MTR but could not be filled with fuel, revealed that a plug had formed in the fill line that was a mixture of fuel and zirconium oxide. ~ Inepile loop No. 6, which is to be inserted in the MTR soon, is to operate at a maximum loop tem- perature of 1600°F with the fuel mixture NaF-ZrF ,- (53.5-40-6.5 mole %), o 300°F temperature differential across the nose coil, and a power density of about 0.75 kw/cm3, A horizontal-shaft sump pump, identical to that in in-pile loop No. 6, is operating satisfactorily in o test loop. It will have accumulated 1150 hr of operation by the time in-pile loop No. 6 is inserted in the MTR, 1.7. Advanced Reactor Design " Reactor designs are being investigated in which the fuel cools the moderator, in order to eliminate the sodium circuit used for cooling the moderator in reactors of the ART type. One of these designs depends on moderation by beryllium oxide in the reflector and some additional moderation by graphite in the core.’ Thermal stresses dictate the use of the - graphite in the core. The relatively poor moderating properties of graphite are not too ob- ~ jectionable ‘becouse of the relatively large core size (24 in. in digmeter) required by the total power and the allowable power densities. On the other hand, the use of a very good moderatfor, such as zirconium or yHrium hydride, in the core would make the reactor independent of moderation by the reflector, and metals could be used in the reflector. Metals require few cooling holes and are good gammastay and fast-neutron shields, With presently assumed materials properties, the hydride-moderated reactors would be better than the graphite~beryllium oxide moderated reactors; on the other hand, the properties of graphite and beryllium oxide are somewhat better known. ANP PROJECT PROGRESS REPORT PART 2. CHEMISTRY 2.1, Phase Equilibrium Studies Phase equilibrium studies of fuel system com= ponents were continved. A re-examination of the LiF-RbF system has revealed the compound LiF-RbF, and a revised equilibrium diagram has been prepared. Further examination of the LiF-UF, system has confirmed the identity of the previously arbitrarily designated compound 3LiF:UF,. There is evidence that the compound is metastable. Studies of the RbF-UF 4 System progressed suf- ficiently for a tentative equilibrium diagram to be prepared. The boundary curves, compatibility triangles, peritectic temperatures, and eutectic temperatures for the NaF-RbF-UF , ternary system are now well-established. Detailed examination of the four-component system NaF-RbF-ZrF 4-UF 4 has established several mixtures which are of possnble interest as reactor fuels. ' In studies of the RbF-CaF, system it has not been possible to get reproducible data for compo- sitions . containing more than 15 mole % CaF.. There is evidence of a compound with a cubic structure, which is presumed to be RbF-Con Examinations of the systems LiF-CeF,, NaF-CeF,, and RbF.CeF 3 were continued, but, as yet, 1here is insufficient data for the preparation of equilibrium diagrams. The NaF-CeF, system was shown to have a binary eutectic, whlch melts at 775 1 10°C, A study of the CsF-BeF, system was started in order to complete the study of the alkali fluoride— beryllium fluoride systems. This study is incom- plete, but, as in other alkali fluoride~beryllium fluoride systems, liquidus temperatures diminish rapidly in this system with increasing BeF, con- tent in the 33'6 to 50 mole % BeF , region. Study of the NaF-LiF-BeF, system progressed sufficiently for a plot of the boundary curves, com- patibility triangles, and pertinent temperatures to be prepared, although the composition of one of the ternary compounds is not completely established. The ternary system NaF-RbF-BeF, requires addi- tional study before the analogous diagram con be constructed. Lowemelting chloride systems, which are poten- t_nolly useful as heat transfer mediums, are being investigated. The present investigation of the KCI-ZrF, system covers the composition range of 10 to 75 mole % ZrF,. It will be of interest to compare phase equnl:brwm behavior in these MCI-ZrCl, systems with behavior in the analogous MF-ZrF systems, which have been thoroughly stuched ot ORNL. : 22. Chemical Reocflons in Molien Salts - : lnveshgatlon of the equnllbrlum reduction of NlF by H, in NaF-ZrF , was continued. A related mveshgohon of the stobzllty of NiF, in the solvent was made that indicated the essential validity of the equilibration data which have been obtained at 500, 575, and 625°C. The data obtained to _dofe indicate that the activity coefficient for NiF, NaF-ZrF, (53-47 mole %) must be about 2750 ot 600°C, The _corresponding value for FeF, in this solvent was previously found to be 3_28 In the investigation of the stability of NiF, in the solvent, a number of peculiar effects were observed that appear to be important in mechanisms of cotrosion of nickel by molten salts. Additional study of these phenomena is to be undertaken, The study of the activity of chromium in chromium- nickel alloys was completed, and the values ob- tained at 750 and 965°C are presented. The studies of the solubility and stability of the structural metal fluorides in fluoride mixtures were extended to include CrF » FeF,, and NiF, the solvents LiF-ZrF, (52-48 mole %), NoF-ZrF (59-41 mole %), and KF—ZrF (52-48 mole %). In addition, the solubility of CrF in the NeF-LiF-KF eutectic was determined. The solubility of CrF, in these solvents voried markedly with the solvent used. In all tests with FeF, at 600°C the solu- bility increased with excess Eer present and the zirconium-to-sodium ratios tended to decrease with increasing excess FeF,, ‘The solubility of NiF, in NaF-ZrF, and KF-ZrF, was found to be the same and nearly independent of the amount of NiF added. However, the solubility in LiF-ZrF, is a function of the amount of NiF, added at both 600 and 800°C, : Studies of the reducflon of UF by structural metals were extended to include the reaction media RbF-ZrF, (60-40 mole %) and L|FQBeF (48-52 mole %). The equilibrium chromium concentrahons in the RbF-ZrF, mixture at 800°C were found to be lower than those in any other alkali fluoride— ZrF, mixture studied, The equilibrium iron con- centrohons were found to be virtually independent ‘of the reaction medium and in cll cases were higher at 600 than ot 800°C. This behavior is in marked contrast to that observed for the Cr -UF b ¥ system, which exhibits strong dependence on the reaction medium, The solubilities of the structural metal fluor:des NiF,, FeF,, and CGrF, in molten NaF-ZrF, (53-47 mole %) that were determined by using concentra- tion cells and an emf method are presented. The results fall within the range of the results obtamed by filtration, - Vapor pressure data were used fo calculute the | effects of concentration and temperature on activi- ties in the RbF-ZrF and KF-ZrF, systems. Activities of KF were very slightly hlgher than the activities of RbF ot correspondmg compositions and temperatures, since the larger rubidium ion has less affinity for the fluoride and RbF is g better complexmg ‘agent than KF., The ZeF, activity was found to decrease with mcreasmg ~ ionic radius of the alkali cation. A systematic invesfigation of the solubilities of rare-earth fluorides in ART-type fuels was initiated, In the initial studies, LaF, and CeF, were used in the solvents NaF-KF-LiF (11.5-42-46.5 mole %)} and NaF-ZrF (50-50 mole %). The results obtained thus for show the solubilities to be comfortably high from the standpoint of reactor operation. . Apparatus is being assembled for the determination of the solubility of xenon in fused salts, - Polarography was utilized for quantitative analy- ses of structural metal jons in the NaF-KF UF ;-xNaF + 0.5F, Data obtained on the temperature dependence of ~ the reaction were successfully fitted by the ex- pression _ log r = 6.09 — (5.22 x 10%/T) , where r is the decomposition rate and T the abso- lute temperature. The activation energy for the decomposition reaction was calculated to be +23.9-'k¢c|/mp|e_ of UF ;.3NaF. 4.4, _Cri_ticdl_ Experiments N ‘Additional robmétempe'roture experiments ‘were " made on a reconstructed critical assembly that represents -the cnrculahng-fuel reflector-moderated ‘teactor. . The fuel" ‘region consists of alternate “Tamince of Teflon and enriched uranium foil, It -'wus found in experiments in which the outermost loyer of _the beryllium reflector was replaced with - stainless steel that beryllium is 2.75 times more “effective than stainless steel in this region. The reactivity coefficients of several materials of engineering interest were evaluated at various points ulong the radrus ‘at the mid-plane of the reactor. The results are presented as the change in reactivity introduced by filling @ void with the material. In the assembly being studied one end A suitable stub for. the ‘junction of | electrical cable to self-resistance-heated piping PERIOD ENDING SEPTEMBER 10, 1956 duct is thicker than the other so that the effect of end-duct thickness can be measured, : Equipment is being fabricated for an elevated- temperature critical experiment for investigating the design features and nucleor characteristics of the circulating-fuel reflector-moderated reactor being designed by Pratt & Whitney Aircraft, The design of the assembly is quite similar to that of the ART high-temperature critical assembly which was tested previously. . PART 5. REACTQ‘R SHIELD‘NG 5.1, Shielding Theory The results of calculations of gamma-ray energy flux, dose rate, and buildup factors in a lead-water shield of finite thickness are presented. A Monte Carlo method was used for the calculations, which included 1., 3+, and 6-Mev photons incident along a normal ond at on angle of 60 deg. Comparisons are made with data from the results of the moments- method solution and from an earlier Monte Carlo calculation for 3-Mev photons normally incident upon a one-region finite lead shield. The results of the calculations for radiation incident at 60 deg indicate that the practice of using only normal incidence data for shield designs can lead to a poor approximation. This problem becomes most acute when the number of mean free paths across the shield is small or when the angular distribution is such that a large portion of the radiation is not normal or nearly normal to the slab. _ 5.2. Lid Tank Shielding Focility The ‘,;tudies of udvcmced shielding materials were continued with mockups consisting of a beryl- - livm moderator region, a lead or depleted~uranium ‘gamma-ay shield, and a lithium hydride and oil ~ neutron shield. In some mockups a boral sheet was inserted outside the beryllium layer to prevent thermal neutrens from ‘entering - the gamma-ray shield material. : Data obtained thus for indicate a strong influence of the placement of the various materials on.the gamma-ray dose rate, The thermal-neutron traverses for the various configurations, however, show the flux to be independent of the order of the lithium hydride and lead, except for the difference at the beginning of each traverse caused by variations in the amount of oil trapped between the various 13 slabs. For a similar crrangement containing de- depleted uranium, moving the uranium out of the. intense neutron field redyced the thermal-neutron flux. The results of fast-neutron dose rate traverses are being analyzed, and studies of similar cone figurations are being continued, 53, Bulk Shielding Facility The gamma-ray energy spectrum and time decay characteristics of the fission products of U235 as related to circulating-fuel reactors are being investigated, Based on the preliminary energy- spectrum dota obtained thus far, an integration of the spectra between 0.28 and 5.0 Mev and between 1.25 ond 1600 sec gave a total of 2.81 photons emitted per fission with a total energy of 3.22 Mev . . per fission, These values are to be compured with data obtained from experiments on time decay characteristics which gave 2,92 photons per fission and 3.23 Mev per fission, All these values carry an estimated error of about $25%. Experiments are under way in an investigation of the effect on the dose rate outside the ART lead 14 shield of gamma-ray streaming through the NaK- filled pipes. Measurements made beyond a duct placed through a mockup of the ART shield af an angle of 51 deg 30 min are reported, Some measure- ments made beyond straight-through penetrations mocking up portions of the south-head ducts are being studied. Measurements are now being made in order to determine the effects of various com- ponents in the mockups and of adding patches to the shietding, o L 5.4, Shield Mockup Core Nuciear calculations were made in order to pro- vide information for detailed design of the reactor core-reflector-island region of the Shield Mockup Core (SMC). The SMC, c 5-Mw fixed-fuel reactor, is being designed for shielding studies of the circulating-fuel reflector-moderated ‘reactor, - The leakage fluxes of the SMC, with the exception of the fission-product radiation from the heat ex- changer, are to be the same as those from the reactor now being designed by Pratt & Whitney Aircraft and designated as the PWAR-1. = ta ) Part 1 AIRCRAFT REACTOR ENGINEERING S. J. Cromer ™ A LU ‘ molntenonce 1.1. AIRCRAFT REACTOR TEST DESIGN A. P. Fraas STATUS OF ART DESIGN Reactor Assembly Most of the detail drawings of the reactor-core, heat-exchanger, pump, and pressure-shell assembly - have been completed, and the drawings have been ‘approved for procurement. Design work on the reactor shield and the cell interior is proceeding concurrenfly with the construction of a one-half- scale model of the reactor- pressure-shell, NaK- ~ manifold, and lead-shield assembly, and a one- sixth-scale model of the reactor and cell, These models are proving most helpful in visualizing interference problems in both assembly and in cedures for assembly, operation, and servicing is - continuing, with major attention being given to vnusual contingencies. Reflector-Moderator Cooling Circyit A careful review of the detail drawings indicated that the pressure drop through the reflector-modera- tor cooling, circuit was likely to be substantially higher than that originally estimated. In order to provide for this probable increase in pressure drop, larger and higher-speed motors are being ordered for the sodium-pump drives. Equipment is being assembled for water flow tests on the crucial ele- ments of the sodium circuit, in porticular, the beryllium cooling holes and annuli and the mani- fold ‘at the inlet. to the beryllium. - ‘Preliminary . ‘calculations indicate' that the increased pressure drop -will ‘not present any serious problems’ from - ~the stress sfondpomt but the effects are bemg ,checked : S : Heat Exchangers i A greot deal of work and experlmentohon by 1he_",- D heat exchanger vendors has revealed that it will. be extremely cosfly in ‘time and money ‘to’ meet S "the close dimensional . folerances - specrfled A review’ of the design indicates that ‘certain’ toier- - ances may be- reloxed with only minor decreases in performance vendors it is believed that compromises may be reached that will permit production of heat ex- changers that will give acceptable performance} The preparation of written pro-. By working closeiy ‘with -the Fuel Pumps ond Expansion Tank Further test work on the fuel pump prepared for the high-temperature test performance rigs has indicated that the cavitation limit is somewhat lower than had been anticipated from the early water tests. This condition could be improved substantially, as indicated by the early water tests, through the use of a modified impeller vane, which would be somewhat more, but not unreason- ably, difficult to fabricate. Indications are, how- ever, that the increased cavitation suppression head required to obtain satisfactory operation with the present impellers is not excessive and can be accommodated readily by increasing the fuel ex- pansion tank pressuwe by about 10 psi. This problem is still being investigated (see Chap. 1.4, **Component Development and Testing’’). Tests on the north-head fuel pump and expansion tank assembly have shown that satisfactory op- eration can be obtained with fuel levels in the fuel expansion tank running from ]/2 to 3 in. with no objectionable aeration or pressure fluctuations in the system over the entire speed range. Test work is being continued with this setup in an effort to reduce the pressure difference between the fuel pump inlet and the expansion tank, and to improve its performance with one pump out or with one pump at a substantially different speed from the other. Reduction of the pressure differential be- tween the pump inlet and the expansion tank is important - because this pressure differential will largely determine the stresses in the upper and lower decks of the north head. It is believed that rather- snioll modifications in the details of the bleed passages coupling the pump inlet with the " exponslon tank will permit a very marked reduction in this _pressure dlfferentral Fuel Recovery Tonk The prehmmory desrgn of the fuel recovery tonk " has been completed. The ‘tank has been designed “on the premise that no electrical, cooling, or other lines need be connected to ithe tonk at the time ~that it is removed from the cell. . This means that " the - afterheat being generated in the fuel must balance the heat losses to convection and radi- ation to hold the temperature in the fuel between 17 ANP PROJECT PROGRESS REPORT its melting point and 1600°F. Because of the sharp increase in radiation heat losses with tem- perature this provides a fairly wide latitude. In addition, the tank has been designed so that the amount of air circulating up between the four 8-in.- dia fuel tanks and the lead shield will be con- trollable, and hence the air flow through this stack-type of thermal-convection cooling system can be varied. It is believed that in this way it ~.will be possible to hold the fuel temperature in the tank within the desired limits throughout a period of approximately 30 days, starting about 10 days after shutdown. Removal of the fuel from the cell to the reprocessing facility within 10 days of time ‘of shutdown has been considered to be very im- portant as a demonstration of the practicality of such an operation and, hence, the possibilities for operating with a low fuel inventory in a complete operational system. APPLIED MECHANICS AND STRESS ANALYSIS R. V. Meghreblian Reactor Support The complete reactor and shield assembly is to be suspended from an overhead bridgelike struc- ture, as shown in Fig. 1.1.1. Although the shield components will be in close contact with the outer shell of the reactor, the weight of these members will not be carried by the reactor. A separate system of tension members is to be provided for this purpose, and these members will transmit the shield loads directly to the bridge structure. The weight of the reactor will be transmitted to the bridge structure through the four pump barrels. The attachment of the individual barrels to the bridge will allow horizontal motion of the barrels in order to accommodate the relative thermal growth between the reactor {at operating tempera- ture) and the bridge (ot room temperature). Verti- cal motlon of the barrels will be restrained through Fabreeka! pads, which are sufficiently resilient to distribute the load fairly uniformly between the four barrels. The bridge will be fixed ot each end Irabreeka is o material composed of layers of tightly twisted, closely woven, lightweight cotton duck Ihoroughl impregnated with a special rubber com- pound. This material is manufactured by the Fabreeka Products Company, Inc., Boston, Mass, 18 to a flexible column consisting of 1.0-in.-thick steel plates, 28 in. wide and 123 in. long. The load carried by each column will be approximately 45,000 Ib. This value is somewhere between one- fourth and one-half the load required to cripple the column. A precise figure cannot be given since the "‘end condition’ of the top of the column is not well-defined. The principal function of the flemble co!umns is to allow the NaK lines, which remove the heat from the reactor, complete freedom when expanding from room temperature to the operating temperature - of the reactor (Fig. 1.1.2). During full-power op- eration, the upper row of NaK lines will be at 1070°F and the lower row at 1500°F. This will result in horizontal growth of 0.75 in. in the upper lines and 1.125 in. in the lower. If the reactor were mounted in the cold condition precisely over the center line of the column bases, these expan- sions would translate and rotate the reactor out of the neutral position and thereby cause bending stresses in the columns (Fig. 1.1.22). The present plan is to precut the NoK lines short by the amounts mentioned above so that at room tempera- twe the reactor will be located 1.0 in. off the neutral position. This displacement will be toward the cell wall through which the NaK lines enter (Fig. 1.1.2¢c). As the reactor heats up, the NaK tines will expand and move the reactor back to the neutral position and thus remove the bending loads on the columns and the axial loads in the pipes (Fig. 1.1.25). NaK Piping Inside Reactor Cell The flexible columns described above will allow for gross expansions of the NaK lines, but they do not provide for differential expansion between the lines of any one row. In order to provide some margin for operational incidents and accidents? and greater freedom in controlling NaK tempera- tures, some additional flexibility has been intro- duced into the piping inside the cell. This has been accomplished by the addition of several bends in each line. The bends will accommodate 300 to 400°F temperature differences between adjacent lines. A horizontal view of the NaK 2For example, the failure of the pump in any given NaK circuit would cause o change in the temperature of the pipes, and this in turn would produce thermal deformations in the piping relative to the plpmg in the other NaK circuits. 'y CONSDEMTHAN ORNL—LR—-DWG 16123 REACTOR CELL™| - \—Nal( LINES ‘.’ 121t 8%/, in. - b 5 ft 9% in.—— H._ " \— L WATER SHIELD CONTAINER Bemmmgeee] N FLEXIBLE COLUMN 5ft 05 in, l Y Fig. 1.1.1. ART Support Structure, 61 9661 ‘0l ¥IGWILJIS ONIANI QOI¥3d ANP PROJECT PROGRESS REPORT ~NoK LINE N TTIhNH AT -TEMPERATURE, WITHOUT PRECUTTING OF NaK LINES (o} \cowmas © SONPIDENTA. . .~ ORNL-LR-DWG 16124 PUMP BARRELS REACTOR 77 77777 N ' % AT TEMPERATURE, WITH PRECUTTING OF NaK LINES (&) POSITION OF REACTOR AT ROOM TEMPERATURE, WITH NoK LINES PRECUT {c) Fig. 1.1.2, Movement of Renctor‘ Cau#ed by Expansion of NaK Lines, piping inside the test cell is presented in Fig. 1.1.3. This layout was selected on the basis of a hand calculation for which the method of Spielvogel3 was used. A more exacting analysis, which in- cludes a careful study of the manifolding system, is now under way at the M. W. Kellogg Co. These calculations are being carried out with the aid of fast computing machines. A general schematic diagram of a complete NaK circuit is shown in Fig. 1.1.4. The essential feature of this system is that the. NaK lines are rigidly attached at only one point, the reactor cell wall. The ends of each circuit, which terminate at the reactor and at the pump-radiator assembly, have considerable freedom to expand. The only constraints imposed on individual lines of a given - circuit arise from the differential thermal growth between the various lines as a result of differences in length and irregularities in temperature control. The NaK piping external to the reactor cell was - described previously.4 ‘ 35. W. Spielvogel, Piping Stress Calculations Simpli- fied, 4th ed., Loke Success, New York, 1951. 4R, V. Meghreblian, ANP Quar. Prog. Rep. March 10, 1956, ORNL-2061, p 22, esp Fig. 1.7. 20 The principal stresses in the NaK lines will be due to the therma!l effects described above. Each time the reactor system passes from one operating condition to another, there will be a redistribution of the thermal stresses in the various lines. The stresses produced by relative thermal growths do not persist for all time, however. Relaxation tests (see Chap. 3.5, ‘“Mechanical Properties Studies'’) indicate that stresses due to a fixed strain, such as that present here, decay rapidly. This decay takes place as more and more of the initial strain on the member is transformed to plastic deformo- tion. The decay is sufficiently rapid that the plastic strain developed in a matter of minutes after loading is a significant fraction of that which would be developed in several hours. Thus with each change in operating conditions, the thermally loaded members will be subjected to some plastic deformation. An important criterion for design, then, is the strain-cycle life to which a given member can be subjected. The strain-cycle criterion was used as a basis’ for the design of the NaK lines. The objective of various analyses now under way, therefore, is the determination of the plastic strain that will be o developed with each chonge in operating condi- tions. The results will then be compared with strain-cycling data by using a Coffin® type of cor- relation. Some data for Inconel are presently available,® and additional information is now being collected ot the University of Alabama and at ORNL. L. F. Coffin, Jr., ""A Study of the Effects of Cyclic Thermal Stresses on a Ductile Metal," Trans. Am Soc. Mech. Engrs. 76, 931 (1954). 6pratt & Whitney Aircraft, Nuclear Propulsion Program Engineering Progress Report No. 18, PWAC-554, Oct, 1, 1955-Dec. 31, 1955, p 63, Fig. 21. PERIOD ENDING SEPTEMBER 10, 1956 Fill-and-Drain Tank and Support - On the basis of the present design and the op- erational philosophy of the ART system, there is some reason to believe that the safest position for the fuel during the normal progress of the experi- ment will be in the reactor proper, because in the event of trouble it will be easier to dump the fuel into the fill-and-drain tank than to pressurize the fuel back into the reactor. It has been suggested that this may also be the safest place for the fuel during certain off-design situations, such as the one-pump-out condition, partly for the same reason SESREP ORNL-LR-DWG 16022 CELL WALL 8-in. PIPE ‘ . _ i Fig. 1.1.3. NaK Piping in the Reactor Cell, 21 - ANP PROJECT PROGRESS REPORT - 1 |_—THERMAL SLEEVE ~p————— {070° F NoK . ORNL-LR=-DWG 16125 _CONSTANT LOAD , SPRING HANGERS ) MOTION ALLOWED REACTOR — FLEXIBLE COLUMNS ————— 1500 °F NaK - : RADIATOR AND PUMP ASSEMBLY A MOTION ALLOWED BY FLEXIBLE COLUMNS. B MOTION ALLOWED BY RADIATOR-PUMP ASSEMBLY. C FIXED POINT IN NaK PIPING. Fig. 1.1.4. Over-cll Suspension and Constraints on NaK Piping System, and partly because of the increased probability of a failure duwring the switching operations associ- ated with an emergency condition. A study is now under way to determine the validity of these premises. In addition, there are several off-design conditions and accidents in which it will be clearly unsafe to leave the fuel in the reactor; for example, if there is an internal leak between the fuel and sodium (or NaK) systems, a chemical reaction might occur which could produce metallic vranium.” Under certain operating conditions, the vranivm could be deposited throughout the hot portions of the fuel circuit. In order to control these reactions and to minimize their effects, it would be necessary to remove the fuel from the reactor. ' A fuel fill-and-drain tank for receiving the fuel is incorporated in the ART system design (Fig. 1.1.5). The tank will be an Inconel cylinder 38 in. in diameter and 38 in. in length, which will be kept at a steady-state temperature of 1100°F by means of a continuous flow of NaK from two in- dependent external circuits. These circuits have 7W. B. Cottrell et al,, Airc.raft Reactor Test Hazards Summary Report, p 38, ORNL-1835 (Jan. 19, 1955). 22 SURPIBENTA ORNL~LR-DWG 16126 NaK SUPPLY—1 j, (1100°F) FUEL DRAIN '/ (1200°F) ~~— INNER CYLINDER {He BLANKET AT 45 psi ) BAFFLES {FOR SUPPORT AND FLOW DIVISION) SUPPORT CYLINDER z ] (5000 1b UPLOAD) . Fig. 1.1.5. Fuel Fill-and-Drain Tank. U A i ot b e w) several functions. First, they will remove the decay afterheat from the fuel following a dump; second, they will keep the fuel above its freezing point at long times after the dump has occurred and the decay heat has been dissipated; and, third, they will maintain the tank at a sufficiently high temperature to minimize thermal shock effects when a dump takes place. The design heat load for the tank is ]75 Mw even though it is estimated that the decay heat at the instant of shutdown after operation at 60 ~ Mw for one month would be 3.6 Mw. This reduced operational requirement is based. on the premise " that the fuel will be cooled,in th:e‘ Vreoctor"to 1200°F before a dump and that @ minimum of 8 sec - ‘will elapse after the coolant rod ‘has been fully inserted and before the dump valves are opened. ~ Each of the two NaK cooling circuits will be capable of extracting the 1.75 Mw of decay heat under these conditions; however, it is expected that under one-circuit operation, the fuel and metal temperatures at certain locations within the tank . may be as high-as 2000°F. Such short-time effects will ‘not be encountered during o normal dump. During normal operation, both circuits will be in use, and the maximum metal temperature is not ~expected to exceed 1600°F. The primary structural loads within the tonk will be due to the NaK pumping pressure of 40 psi. - Since the tank is to serve as an ever-safe deposi- tory for the fuel, it is necessary that it be capable . of surviving some 2000 hr of continuous operation at temperature and possibly several fast dumps. The principal structural requirements of the tank, then, - are based on. creep-rupture ond thermol-:_____}; shock cons lderotrons Analyses show thot the most sensmve areas in - this design are at the - joints’ between the tube_,f- header sheets ond the mner cyllnder The dis-" continuity stresses in the'se areas are of the order_ o of 6000 psi. This is a relotlvely hlgh stress level, but it is- expected thot these stresses er be_‘f,:: _ _'_'Q_fi'tlon (1200°F tsothermol) to- full .power: some 30 " times.” One cycle consists of 8 hr ot idle ‘and’ 16 hr at tuil power. In- possmg from the idle condi- - tion_to full power - the temperature dnstributlon in markedly reduced os the ‘metal-in_ that region de- - forms in- creep. Slnce “the extent to which" dlS-_‘f.'»_ conhnmty stresses can relax is not known, a test __program’is under way to study this effect by means .. ~of tube-burst tests. Fmolly, because -of ‘the im-~ * 'portunce of thls component to the over—o“ sofety, of the experiment, it'is planned to test the entire tank assembly by using one of the NaK circuits. The test should demonstrate the adequacy of the PERIOD ENDING SEPTEMBER 10, 1956 design in withstanding the creep and thermal shock effects mentioned above. The support of the fill-and-drain tank assembly will be accomplished by means of a nitrogen cylinder located beneath the tank (Fig. 1.1.6). Although the tank will be attached rigidly to the reactor pressure shell through the fuel drain lines, the major portion of the tank weight will not be allowed to bear on the shell. The total weight of ~the tank, including the fuel, will be 6000 Ib, of which 5000 b will be carried by the nitrogen cylinder. The weight of the tank without the fuel " will be 4000 Ib. This support arrangement intro- duces some complication in the stability aspects ~ of the system, but calculations show that the proposed distribution of lever arms, structural stiffness of the drain lines, and the weights are ‘well within the stability limits of the system. In the event of the failure of the nitrogen cylin- der, all the tank weight will be exerted upon the drain line and reactor pressure shell, This situa- tion will produce high local stresses in the shell at the points where the two drain lines will be attached and will add an additional 1000 Ib to the load in each pump barrel. Although this will re- duce the design safety margin in these com- ponents, the system can support the tank in this manner. The purpose of the nitrogen cylinder, - then, is to reduce the operating loads in the barrels and in the pressure shell. Core-Shell Low-Frequency Thermal-Cycling Test A one-fourth-scale model of the core shells is . presently being subjected to the long-time tempera- ture variations expected in the reactor as a result - “of chonges in power level (see Chap. 1.4, ‘‘Com- "f'_-ponent Development and Testing’’). Such tempera- ~ ture ‘variations: produce thermal . distortions in the: shells, “and - these effects. are among the most severe condntlons ‘imposed on the shells.” The "'present operating :plan for “the ART anticipates that ‘the reacfor will be Cycled from the idle condi- the “core shells will change from a unn‘orm prof:le‘- " to a linear drstnbutlon, ‘with a maximum grodlent_ ' of about 300°F through the wall. The cyclic conditions described above have been simulated in the one-fourth-scale model of the 23 ANP PROJECT PROGRESS REPORT SUPPORT BRIDGE DUMP VALVE (( | N aK SUPPLY ATTACHMENT ————— o=~ TO FLOOR SES Fig. 1.1.6. Fuel Fill-and-Drain Tank Support, 24 : CONFIDENTIAL ) ORNL- LR~DWG 16127 C‘ outer core shell by means of a dual NaK circuit which heats the. inner surface of the shell and cools the outer surface. The test cycle selected was | hr at power and 1 hr |sothermal at 1200°F. At the power condition the shell is exposed to a maximum temperature gradient of--300°F through " the wall. The stress buildup and relaxation with each change in the operating conditions is illus- trated in Fig. 1.1.7. The solid-line curve gives the stress-time variation expected in the actual shells during the operation of the reactor. The broken-line curve gives the stress hme hlstory for the ‘model. seerrT ORNL ~LR-DWG 16128 T 17 1T b 1 1 | l"""" f\‘ MODEL NS REACTOR o [ £ & 0 —1 A 7 \/ | V y b 0 2 4 6 8 10 TIME {hr) Flg. 1.1.7. Stress.Time History of Model Used- for Core-Shell Thermal-Cycling Tests in Com- ith Stress- tory E ted in the - = panson Al "es T_We .I'_irs ory :_cpec”e rn € an upprecnable fraction of the applied deformation owill have been converted ‘to_plastic’ strain. On ”__r_)thas basis it was urgued that the cycle times for _ ¢ ‘relax ' it the- ‘model test’ could be reduced to.the 1-hr inter- was ¥ound fhut because of the very ropad decoy of ©the stress it was not.necessary to design the core- S "'she“ test for fhe 8- to 16-hr Cycle time: pianned_'}-i_-'f " for the reactor. - “The 2hr- Cycle time was:found to 7 be suffrcrem‘ to. develop a substan‘hal fraction of - < the plastic “deformation. expected in the- octual"‘ - shells, A" typlcol set of: relaxoflon ‘curves’ for ~ “Inconel,. "as. obtained" from various- sources, iso '_';,presenteds,ln Flg 11,8 The curves lobeled;:‘-f 'l}f"ORNL" were obtained from ‘measurements -made - " by the ‘Metallurgy DIVISIOI‘I of ORNL; the curves labeled *‘University of Michigan’’ are based on measurements made at the Engineering Research - PERIOD ENDING SEPTEMBER 10, 1956 SONEDENTIET" . ORNL~LR~-DWG (6329 25,000 E | ! INITIAL STRESS (24,000 psi)| 20,000 \ L —UNIVERSITY OF MICHIGAN 15,000 ~— o ‘-——L\-k INITIAL STRESS (12,000 psi) e N \v FROM CREEP DATA B W\ —~— 10,000 Rt RNL e = ORNL S SSemea [ —_—i _-:_:-___.:_::___-_.-. . —— "'z"“—‘-'r'-':-‘:‘.-;_-:;:;::“:::_ L —— ] UNIVERSITY OF MICHIGAN— 5000 0 0 04 02 03 04 05 06 07 TIME (hr) Fig. 1.1.8. Relaxation Data for Inconel at 1300°F, Institute of the University of Michigan; and the curves labeled **From Creep Data’’ were synthe- sized from creep curves. It is evident from these curves that within a ‘matter - of minutes after the start of the test, be- cause of the extremely rapid decay of the stress, vals bemg used, since opproxrmotely 70 to 90% of the - total strain’ deveioped in 810 16 hr could be _ioch|eved in L hr. . In this way it was possible to " reduce - the totol model ?est time - by a chtor of Cabow 0 " The flrst of rhese Iow frequency thermul-cycllng tests has been completed “The ‘model was sub- |ected to. 300 full cycles a factor of 10 more ‘than ) expected in:operation of the ART, and appeors to have survived without gross fcnlure It is believed “that the factor of 10 increase in the number of test cycles, in comparison with the cycles ex- " pected in the ART, is not an overly conservative 25 ANP PROJECT PROGRESS REPORT margin for tests of so vital a structural component. A complete metallurgical examination of the shell -has not yet been completed. [t should be men- tioned that the test program was interrupted after 57 cycles because of a leak in the external cir- cuits and the specimen was allowed to cool to room temperature. At the completion of repairs, the model was brought to temperature again, and the test was resumed. CORE HYDRODYNAMICS W. J. Stelzman W. T. Furgerson Further core flow studies were made on the full.- scale model of the ART core by using the tech- niques described previously.® The configuration .sG. D. Whitman, W. J. Stelzman, and W. T. Furgerson, ANP Quar. Prog. Rep. Dec. 10, 1955, ORNL-2012, p 23. EXPANSION JOINT FUEL PUMP VOLUTE FUEL FLOW ' l TN ' tested consisted of swirl-type header No. 2 with « simulated conical island expansion joint (Fig. 1.1.9). Tests were run with and without inlet guide vanes.. The inlet-guide-vane system tested consisted of a set of 14 paired vanes designed by G. F. Wislicenus? to induce radial mixing in the fuel annulus. The direction of rotation of the vortexes was to be alternately clockwise and counterclockwise around the annulus. = - The flow pattern produced by the blades agreed with the design pattern in approximately the upper “third of the core annulus, but the individual vortex cores rapidly degenerated into a region of low- velocity random motion at the equator. As the flow accelerated in the portion of the core belew the 9Consulfant. A ORNL~LR-DWG 16130 FUEL PUMP VOLUTE - OUTER CORE SHELL ISLAND WALL Fig. 1.1.9. Swirl-Type Core Header No. 2. 26 xp equator it plcked up a pattern of rotation centered around the core axis and was, in general, typical of the flow obtained in this region with other inlet configurations. A lack of circumferential symmetry was noted in the flow pattern produced by these guide vanes. Previous configurations have not been perfectly - symmetrical in this respect, but this system ap- pears to be worse than the others. It is possible that the vanes themselves were not symmetrical. They are complex in shape and have no suitable reference surface to use in effecting their align- ment. This analysis of the flow is based on observations of injected dye, and there are no quantitative data on this system. The decidedly three-dimensional character of the flow made the use of claw-probe traverses impractical. Complete probe traverses were made of the same core-header combination without inlet guide vanes. Circumferential asymmetry of the axial velocity profiles, as indicated by the maximum deviation of the reading of either. of two probes at each station from the average, was as follows: Maximum Deviation ‘ Station h . (%) 8 10 7 10 6 20 5 20 PERIOD ENDING SEPTEMBER 10, 1956 22 - A W M =] - The profiles given by the two probes at each sta- ‘tion were, in general, similar in shape. In an effort to obtain information on the heat conduction to be expected through the core shells, local heat transfer coefficients were calculated by Reynolds number analogy from the probe survey data extrapolated to the walls by means of uni- versal velocity distribution equations.'® The local coefficients thus obtained were then inte- grated from the wall out into the stream to obtain the film conductivity. The integration was carried out for an arbitrary distance of 0.010 in. from the wall. This dimension was chosen because further depth did not result in an appreciable change in the film conductivity. The fuel assumed for these caleculations was the mixture (No. 30) NaF- ZrF - UF (50-46-4 mole %). The heat transfer coefhcu- enfs for the 0.010-in.-thick fluid layer adjacent to the walls, as obtained for the swirl-type header No. 2 without inlet guide vones and with GS-2 guide wvanes (Fig. 1.1.10), are presented in ~ Table 1.1.1. ‘°w H. McAdams, Heat Transfer, 3rd ed., p 209, McGraw-Hill, N. Y., 1954, TABLE 1.1.1. HEAT TRANSFER COEFFICIENTS FOR 0.010-in.-THICK FUEL LAYER ADJACENT TO OUTER ' AND INNER CORE WALLS OBTAINED FOR SWIRL-TYPE HEADER NO. 2 WITH AND WITHOUT INLET GUIDE VANES Fuel mixture: NcF-ZrF4-UF4 (50-46-4 mole %) Heat Transfer Coefficients (Btu/hr-ft 20 F) GS-2 Guide Vanes Sfufiofi ' ' No Guide Vanes - ' | Outer Wall '~ Inner Wall Outer Wall - lnper Wall 8 4986 -2 5411 , 4913 7 4631 2469 2418 2066 6 3800 3182 2651 2526 5 3394 3243 2601 : 22 4 3032 4265 2559 2348 3 4959 : 4865 2494 2880 2 5678 5310 5794 3706 1 6193 5660 Not available - Not available Fig. 1.1.10. Guide Vane of ART Core. 28 -ANP PROJECT PROGRESS REPORT ‘UNCLASSIFIED. _ PHOTO'25581 and Baffle-Plate Arrangement (GS-2) for Distribution of Flow in Plastic Model PERIOD ENDING SEPTEMBER 10, 1956 1.2, ART PHYSICS A. M. Perry COMPARISON OF BASIC GAMMA-RAY DATA WITH BULK SHIELDING REACTOR GAMMA HEATING MEASUREMENTS H. W. Bertini C. M. Copenhaver The series of experiments performed to determine the rate of gamma heating in various target mate- rials near the Bulk Shielding Facility (BSF) re- actor ! afforded an opportunity for a direct verifica- tion of the basic gamma-ray data and the method used for the ART heat-deposition calculations.? Such a comparison is of considerable interest, since the agreement between the calculated results and the results for the BSF experiments tends to confirm the methods and data used in the ART calculations, : Experimental heat-generahon rates were measured in aluminum, iron, ‘and lead samples at several distances along the center line of the reactor, as illustrated by Fig. 1.2.1. In the experiments, measurements were made of the transient tempera- tures of the samples during heating in the gamma- ray field and during cooling after removal from the gamma-ray field, The calculated heat-generation rates indicate that the energy deposition in the target metals is due to the following principal radiation components: 1. core gamma rays, 2. reflector gamma rays (water capture gamma rays), 3. target gamma and beta rays. Since on Oracle code for the calcuiahon of gomma heating in reactors of rectangular geom- etry3+4 was available, the heat generation rate in the target samples caused by gamma rays emanating from the core was determined. The heat generation 1k, T. Binford, E, S, Bettis, and J, T. Howe, Gamma Heating Measurements in_the Bulk Sb:eld’mg Reactor. ORNL CF~56-3-72 (March 7, 1956). : 2I-i. V. Bertini et al,, Basic Gamma-Ray Data for ART Heat - Depos:tzon, Calculauons. - ORNL-2113 (Sepi'. 17, 1956). = = 3. C. Claiborne and T. B. Fowler, Tbe Calculatzon of Gamma Heating in Reactors of Rectangulozd Geometry. ORNL CF-56-7-97 (July 20, 1956) - ' 41. B. Fowler and H. C. Clcuborne, Operating Manualx for Oracle Code No. 243: The Calculation of Gamma Heating in Reactors of Rectangular Geometry, ORNL CF-56-8-126 (Aug. 20, 1956). - HEAT GENERATION (w/g OF MATERIAL, NORMALIZED TO A REACTOR POWER OF 1 Mw) in the target caused by gamma rays from the ith energy group, H(V), for one volume element of the core is K N/{V) E, P Hz(V) = . e-x Ba — s 4 [r (V) + (V)12 P UNCLASSIFIED ORNL-LR-DWG {6282 =0 EXPERIMENTIAL (REF {) -==® CALCULATED 05 | 0.2 Od ALUMINUM 0o O o o O r 0.014 0 2 4 6 8 10 DISTANCE FROM CORE FACE (in.) Fige 1.2.1. Gomma-Ray Hecting in Metals at Various Distonces from the Core Face of the BSF Reactor, 29 ANP PROJECT PROGRESS REPORT where a constant which normalizes the thermal flux to 1 Mw and converts Mev/sec to watts, . N,(V) = number of gamma rays per second in the ith energy group for the particular vol- ume element, V, E. = average energy of the ith energy group, = radii of core and reflector, respectively, x = pr(V} + pr(V), where p_r (V) and p,r (V) are the number of mean free paths in a volume element, V, of the core and reflector, respectively, B_ = A(e™ ~ 1) + 1, the energy absorption buildup factor (the constants A and a were evaluated for the materials of in- terest in the report ORNL-2113),2 ‘pg = energy absorption cross section for gamma rays for material being irradi- ated, ' p = density of material being irradiated. ~ It -~ 0 - ~ - { Appropriate B, data were then calculated by using an equivalent atomic number and interpo- lating the available buildup factor datas for point isotropic sources in homogeneous mixtures, Since the effective Z of the core is 9.9 and the effective Z of the reflector is 7.5, the homogeneous mixture approximation is valid, ond an equivalent Z of 9 was used. The BSF reactor is a thermal reactor, and there- fore the significant gamma radiation results from thermal neutron fission or capture. Thus, the spectrum of prompt gamma rays plus U235 capture gamma rays is 8.8 e~ 1-01E photons/Mev-fission.2 For 5500 kw/hr of reactor operation per day,$ the average ~daily fission-product decay gamma-ray energy would be about 85% of the saturation value and the resulting decay gamma-ray spectrum is 8.92 ¢~1.33F photons/Mev.fission. The gamma- ray energies per thermal neutron capture in alumi- num for the various intervals from 3 to 10 Mev were obtained by summing up the individual contributions for each interval from a paper by Kinsey, Bar- tholomew, and Walker.” An additional energy of 5H. Goldstein and J. E. Wilkins, Jr., Calculation of the Penetration of Gamma Rays, Final Report, NYO- 3075 (June 30, 1954). Se., B. Johnson, private communication (1956) to the authors, 7B, B, Kinsey, G. .A. Bartholomew and W, H. Walker, Phys. Rev, 83, 519 (1951). 30 1.80 Mev was added to the third group (1.5 to 3.0 Mev) to account for the decay gamma-ray energy resulting from Al28 (half life, 2.3 min). Thermal- neutron capture by water gives rise to a single gamma ray, 2,23 Mev, per capture by hydrogen, The heating in the target sample by capture gamma rays emanating from the water reflector was obtained by using disk sources 2 in. thick and 40 ¢m in radius. An average flux in the disk was obtained from the qverage flux along the disk center line, determined experimentally, ® as a func- tion of distance from the reactor north fece, multi- plied by 0.8 to account for the decrease in flux in the disk radially. - In computing the heating in the target sample from self-absorption of gamma and beta rays re- sulting from thermal-neutron capture in the targets, local perturbations in flux caused by the presence of the target sample were not considered. The only significant beta-ray source in the target samples was the 2,87-Mev beta ray resulting from the decay of Al?8, This beta-ray source was assumed to be entirely obsorbed in the aluminum somple. The fraction of gamma rays absorbed in the target sample was obtained from an expression by Storm et al.® based on the straight-ahead scattering ap- proximation for an infinite cylinder containing a vniform source distribution. Since the expression neglects multiple scattering effects, the energy absorbed in each cylinder was taken as the mean of the values calculated by using the energy- absorption and the total-absorption cross sections. A comparison of the experimental and calculated results is shown in Fig. 1.2.1. The rounding of the calculated curves close to the reactor surface results from the peaking of the thermal fiux curve in the reflector region. The magnitudes of the ex- perimental and calculated values agree quite well where the Z of the target metal is close to the Z of the surrounding mixture, Z = 9, but the agreement becomes worse as the difference in the atomic numbers increases. For materials having similar Z's, the equivalent Z method should yield good results, as confirmed by the aluminum sample. However, for a heavy Z material following a light Z material, the large, low-energy spectrum built up in the light Z material is highly absorbed in the first mean free path of the heavy Z material, ond o Sm. L. Storm, H. Hurwitz, Jr., and G. M. Roe, Gamma- Ray Absorption Distributions, for Plane, Spherical, and Cylindrical Geometrics, KAPL-783 (July 24, 1952). buildup factor based upon @ homogeneous mixture will yield a low result that becomes less and less accurate as the depth of penetration in the light material increases. This light-heavy sequence phenomenon has also been indicated by Monte Carlo calculations of gomma-ray energy deposition for water-lead slabs. ? In view of the good agreement between the exper- imental ond calculated heat-generation values for - samples near the BSF reactor, it seems reasonable to conclude that the basic gamma-ray data given elsewhere,? ysed in conjunction with the Pratt & Whitney gamma-ray deposition code, should give representative results for the gamma-ray heat de- _position in various parts of the ART. GAMMA-RAY HEATING IN ART R. B. Stevenson Gamma-Ray Source Strength The gamma-ray source strengths of the island, core shells, fuel, and reflector-moderator of the ART were determined so that a calculation of the heating due to these gamma rays could be carried out. The prompt, decay, and nonfission capture gamma rays in the fuel, the gamma rays caused by inelastic neutron scattering in the fuel, and the caopture gamma rays in the Inconel core shells, beryllium, and the sodium coolant were taken into account in the calculation of the gamma-ray source strengths., | The capture gamma rays and the gamma rays caused by inelastic scattering were computed by using neutron fluxes determined by a Curtiss-Wright two-dimensional neutron calculation for the ART. The gomma rays resulting from inelastic scattering were assumed to be emitted with an energy of 1 Mev. The distribution of the prompt and decay gamma rays in the fuel was calculated from the fission power distribution determined from. the Curtiss-Wright two-dimensional neutron calculation - and from an analysis of the hlgh- emperature criti- cal experiment, 10 ' In both the Curflss-erght two-dnmens:onol cal- culation and the determination of the gamma-ray source strength, the ART was divided intc a large number - of regions, the materials in each region being conmdered to be homogeneous for ease of 9s. Auslender, ANP Quar. Prog. Rep. Marcb 10, 1956. ORNL-2061, p 223. wA. M. Perty, Fission Power Distribution in the ART, ORNL CF-56-1-172 (Jon. 25, 1956). PERIOD ENDING SEPTEMBER 10, 1956 computation. The basic data for the calculation of the gamma-ray sources were taken from a report by H. W. Bertini et al. 11 The results of these calculations are shown in Figs. 1.2.2, 1.2.3, and 1.2.4. The distribution of the gamma-ray sources in the island, fuel, and reflector-moderator is shown in Fig. 1.2.2. The values given, in w/em3, are the sums over the seven energy groups actually used. The distribu- tion of the gamma-ray source strengthin the Inconel core shells as a function of the distance from the equatorial plane is given in Fig. 1.2.3, while Fig. 1.2.4 gives the source strength in the region adjacent to the control rod (a combination of sedium and Inconel) as a function of distance from the equatorial plane. The values given in these plots are felt to be fairly accurate, although there may be some discrepancies near the Inconel core shells due to the approximations in geometry used in the neutron flux calculations. There is some ambiguity in the source strength values assigned to the Inconel core shells, and the values may be in error by as much as 20%. This ambiguity arises from the fact that use was made of two different two-dimensional neutron calcula- tions in determining the core shell neutron absorp- tions. One neutron calculation was for the high- temperature critical experiment, which contained no sodium, while the second was for the ART with four times more sodium than will be present, The resulting neutron absorptions in the core shells are quite different for these two cases, so an appropri- ate average was used in calculating the values given in Fig. 1.2.3. The gamma-ray source strength in the core shells represents the absorption of approximately. 3.1% of _the total number of neutrons. This is substantially lower than the figure of 7.5% which was obtained for o spherical mockup of the ART, However, re- cent calculations!? of the Pratt & Whitney reactor, with both spherical and cylindrical geometry, have shown that for a cylindrical mockup the percentage of neutrons absorbed in the core shells is less than that for a spherical mockup by about a factor of 2. Thus the figure of 3.1% obtained from the two-dimensional neutron calculations does not My, W. Bertini et al,, Basic Gamma-Ra Data for ART Heat Deposition Calculations, RNL-2113 (SGPf- ‘7' 1956)0 12¢, Wagner, Pratt & Whitney Aircraft, private communication to R. B, Stevenson. 31 -ANP PROJECT PROGRESS REPORT ShSRES ORNL-LR-DWG t6288 0 !5 0.05 25 "' ¢ ¢ P 0.25% “‘ 0.20 $ 0 .'A / ) oy S e e X U)\ I > Z o 45 i \ 0.30 0.35 0.40 0.45 . 0.50 1 : 0.55 060 .65 CONTROL ROD ALL VALUES IN wattsfem® REFLECTOR — MODERATOR al 1 Fig. 1.2.2, Gamma-Rey Soutce Strength in ART, seem to be inconsistent with figures obtained for a spherical mockup of the ART. The total gamma-ray source strength in the ART was found to be 4.6 Mw, of which approximately 4.1 Mw comes from the core, and 0.28 Mw comes from the Inconel core shells. The remainder comes from the copture gamma rays in the island and the reflector-moderator. Gamma-Ray Heating The heating resulting from the gamma-ray sources described above has been obtained by using a two- dimensional gamma-ray heating calculation. The calculation, which was carried out on an IBM-704 computer, utilized a program developed by Kniedler and Wenzel of Pratt & Whitney Aircraft. The program uses a buildup factor approach to account 32 for the scattered gamma-ray photons. In the case described here, the buildup factor for water was used throughout the reactor. - The sources were computed at a number of points in the reactor, the volume associated with each source point being approximately a cube 4 ¢cm on @ side, Each source was divided into seven energy groups. The heating was determined at @ large number of points for each energy group for _all source points within a specified distance from the heating point. This distance was taken as 40 cm. Thus, a point at the outside of the reflector on the equatorial plane ‘‘sees’’ the entire fuel system source on the equator. | ' The results are shown in Figs. 1.2,5, 1.2.6, and 1.2.7. The heating in the island, core, and re- flector-moderator are shown in Fig. 1.2,5. The SESRET ORNL—~LR—DWG 16283 S 0 O, N @ o O O O o CORE SHELL W o SOURCE STRENGTH {(w/em®) n o o Q 0 10 20 30 40 50 DISTANCE ABOVE EQUATORIAL PLANE (cm) ~ Fig. 1.2.3. | Gcmma-Rfiy- Source Strength in Inconel Core Shells as o Function of Distance Above Equatorial Plane, | PERIOD ENDING SEPTEMBER 10, 1956 RS ORNL-LR-DWG 16284 04 SOURCE STRENGTH {w/cm3) T o o N o © o 0 ) 20 30 40 DISTANCE ABOVE EQUATORIAL PLANE (cm) Fig, 1.2.4, Gaomma-Ray Source Strength in Region Adjacent to Control Rod as o Function of Distance Above Equatorial Plane, SECRET ORNL- LR~ DWG 16267 . * CONTROL. ROD T T 7T LT T LTl T I T I I I T T 7T T ITIT T 77 oI o727 T 2ol e o222 - REFLECTOR —MODERATOR ALL VALUES IN watts /cm3 Figs 1.2,5. Gamma-Ray Heating in the ART, 33 ANP PROJECT PROGRESS REPORT values given, in w/cm3, are the sums over the seven groups used. The heating in the Inconel core shells as a function of distance from the equatorial plane is given in Fig. 1.2,6. The heating in the region adjacent to the control rod is shown in ‘Fig. 1.2.7. The heating caused by the gamma rays from the control rod is not included in these results. The heating from the heat exchanger region was reported previously!3 and is included in the outer regions of the reflector. 13H. W, IBertini, ANP Quar. Prog. Rep, June 10, 1956, ORNL-2106, p 28. TENGT ORNL-LR-DWG 16285 o o -] o INNER CORE SHELL 5 QUTER CORE SHELL GAMMA=RAY HEATING {w/em®) o o 0 10 20 30 40 50 DISTANCE ABOVE EQUATORIAL PLANE {cm) Fige 1.2.6. Gamma-Ray Heating in Inconel Core Shells as a Function of Distance Above Equatorial . Plane, . ' 34 The uncertainty in the source strength of the Inconel core shells does not have much of an effect on the heating values, since the source of gamma rays in the core is so much greater than that in the core shells, A test calculation has shown that a 50% change in the source strength of the core shells will change the heating values at a point close to the core shells by less than 5%. The total gamma-ray energy deposition (excluding that caused by sources in- the heat exchanger region) in the reactor was found to be 3.86 Mw, which amounts to approximotely 84% of the total “source. This means that about 16% of the gamma- ray source energy escapes, PEeREF ORNL-LR-DWG 16286 o ] \ T~ N o N 7 GAMMA-RAY HEATING (w/em®) o 10 20 30 40 50 DISTANCE ABOVE EQUATORIAL PLANE (cm) o Fig. 1,2,7, Gamma-Ray Heating in Area Adjacent to Control Rod as a Function of Distance Above Equatorial Plane, PERIOD ENDING SEPTEMBER 10, 1956 1.3. ART INSTRUMENTS AND CONTROLS E. R. Monn REFLECTOR-MODERATOR TEMPERATURE CONTROL SIMULATION J. M. Eastman! F. P. Green E. R. Mann The ART simulator was used for a study of a _reflectorfmoderator temperature control. system and the temperature responses of the cooling system ~ which will circulate sodium through the reflector- moderator and through sodium-to-NaK heat ex- -changers. The NaK is cooled in external NaK-to- . air radiators. With the reactor held at design-point conditions, abrupt closure of the louvers of one of the two NaK-to-air radiators caused an increase of approximately 100°F in the maximum sodium tem- perature (ot the core outlet) and in the mean beryllium temperature. The response was essen- tially first order, with a time constant of approxi- mately 2‘/3 min. A 50% reduction in power from design-point conditions caused o decrease of about 150°F in the maximum sodium temperature ond in the mean beryllium temperature, with a time constant of approximately 3 min. For operation at 50% power, an abrupt change of the mean fuel temperature from 1425 to 1250°F caused a de- crease of about 160°F in the maximum sodium and the mean beryllium temperatures, with, again, about @ 3-min time constant. The mean reflector-moderator temperature was computed in the process of simulation and was used as the main criterion for evaluating the tem- perature control system. ‘A control system was evolved which performed satisfactorily and will be used for the ART. The control method consisted in (1) holding constant a signal composed of the ‘maximum sodium temperature biased with the temperature differential in the NaK across the radiators and (2) causing the louvers to open when ‘this signal exceeded its set value by more than 2°F and to close when the signal became less than the set value by more than 2°F. This ar- rangement provided a proportionality characteristic adjusted to cause the maximum sodium tempera- ture to be held at 1200°F for zero reactor power and 1250°F for design-point power. These are 10n assignment from Bendix Products Division. C. S. Walker the values calculated (ond built into the simula- tor) for holding the mean beryllium temperature constant at 1200°F. The louver actuating speed was adjusted to give full stroke in 30 sec. This control system was operated on the simu- lator with.step changes between 0, 50, and 100% design power and between 1200 and 1425°F mean fuel temperature (at 50% and at zero power). For these disturbances the mean moderator tempera- ture was held constant to within £15°F. The system was not stable in the sense that the louvers remained fixed for steady-state conditions. In order to maintain the desired temperatures, the control system caused the louvers to shift posi- tions at 30- to 60-sec intervals. When the tem- perature control system hardware has been as- sembled, it will be checked on the simulator, INSTRUMENT DEVELOPMENT R. G. Affel Fuel-Expansion-Tank Level Indicator R. F. Hyland Tests were continued on a helium-bubbler type of level indicator, described previously,? for the ART fuel expansion tank. In these preliminary tests made to check the operation of the bubbler, two bubbler tubes are used and the fuel is static. One test rig was shut down after 3000 hr of opera- tion at a fuel temperature of 1150°F; neither bubbler tube had clogged. Another rig has com- pleted 2686 hr at 1500°F, and another has com- pleted 2352 hr at 1500°F. In both these rigs one of the two bubbler tubes has become clogged. Examination of a bubbler tube from the rig that was terminated showed that o thin film had de- posited on the inner surface. A chemical analysis of a 10-mg sample of the film showed it to contain 7 mg ZrO, and about 1.8 mg O,. - Since the helium supply of the building contains an average of 1.5. ppm O,, sufficient oxygen was present in 1000 liters of helium (the approximate flow for a 75-hr period) to account for the entire 1.8 mg of O,, if complete reaction is assumed. Analyses 2 . R. F. Hyland, ANP Quar. Prog: Rep. June 10, 1956, ORNL-2106, p 43. 35 ANP PROJECT PROGRESS REPORT of the contents of two bubbler tubes that plugged in a previous test at 1500°F, during which the system was accidentally contaminated with oxy- gen, also showed high ZrO2 contents. It is ap- parent therefore that the cause of bubbler-tube clogging is oxygen contamination and the subse- quent formation of Z0,, a high-melting-point (3000°C) solid, rather than ZrF ,-vapor deposits or fuel, as anticipated. It is planned to use either a NaK ‘scrubber or the incandescent-titanium- sponge method to purify the helium used in further tests. Systems for Testing Liquid-Level-Sensing Devices in Flowing Liquids “H. J. Metz Three test systems are being constructed for studying the operational characteristics of various liquid-level-sensing devices under controlled con- ditions in flowing liquids. The test system con- sists of pressure vessels at both ends of a motor- driven teeter-totter. Liquid will flow from the rising vessel to the vessel going down and the liquid levels in the vessels will remain constant. A gear-motor drive is provided which can rock the vessels at a rate of 0 to 10 cpm. A helium atmos- phere at a pressure of about 1 to 3 psig is pro- vided in a closed circuit over the vessels to pre- vent contamination of the fluid., The desired temperatures are obtained and maintained with heaters and variable autotransformers, Each vessel has six spark-plug probes con- nected to level-indicating lamps, and the liquid levels in the vessels can be adjusted from 1 to 10 in. Two level-sensing devices can be tested in each vessel at various levels. The tests will be of long duration to assure that the level-sensing devices will work for 2000 to 3000 hr with the surface of the fluid in constant motion. High-Temperature Turbine-Type Flowmeter G. H. Burger The first turbine-type flowmeter was designed and built in October 1955 for measuring flow in l-in.-ID tubing. Since then numerous tests have been made in order to determine the operating characteristics and the life of the units. The initial design has been modified as test data have indicated the need for changes and three succes- sive models have been built.? The test results obtained thus far have been disappointing, particularly the life test results. As stated above, four flowmeters have been tested, but none has even approached the required 3000 hr of operation. The brevity of the test, however, in all cases except the first, has been due to system failures, such as pipe leaks and pump difficulties. Performance tests of the ‘vnits have been quite satisfactory. The units have ‘been tested in both NaK and fused-salt fuel mixtures with excellent results. The accuracy of the measurements in the NaK system have been well within 1%, as checked against a magnetic flowmeter and venturi. . The accuracy of the units in the fused-salt mixtures could not be precisely determined because of the absence of a venturi in the system. The salt sys- tem also had a very low maximum flow rate which was out of the design range of the flowmeter. The bearing material and the bearing design of the unit tested in the fused salt appeared to be satis- factory, since no evidence of bearing-sticking or self-welding was evident from the operation of the units or from metallurgical examination of the units after removal from the system. The turbine blades and the turbine bedy appeared to be un- marked by the fuel or the NaK, Even though the tests in fuel and in NaK have been.relatively short, in the neighborhood of 300 hr, the units have operated at temperatures up to 1500°F, and the indications are that the units can be expected to operate satisfactorily for 3000 hr. - A fifth model of the flowmeter has been fabri- cated which incorporates improved bearings. - The water test and calibration of the unit have been completed, and it appears to be quite satisfactory. A similar turbine-type flowmeter with a 3}§-in.-dic housing has been designed for operation in a sys- tem with flow rates up to 1400 gpm and is now being fabricated. This unit will be evaluated in a loop for testing NaK pumps. It will be cali- brated against a venturi installed in the system. 3 G. H. Burger, ANP Quar. Prog. Rep. ]une 10, 1956, ORNL-2106, p 43. High-Temperature Pressure Transmitter W. R. Miller Pressure transmitters of several types obtained from five different vendors have been tested at temperatures between room temperature and 1400°F. Four modefs were of the pneumatic force- balance type which employs a bellows or dia- phragm as the sensing element. The all-welded “bellows appears to be the most sotusfactory Six of the transmitters tested produced an average accuracy of +0.37% full range if heid at a constant temperature, but zero shifts as large as $3% full scale were observed between room temperature and 1400°F. S - _Five .units were tested whlch utnllzed a dio- ,'Vphrogm-lsolated NaK fllled tube system in which ‘NeK hydrostottcolly ‘transmits the dlaphragm -sensor motion to a 3 to 15psi transmltter-mdlco- - tor.. These units are available in ronges from 0 to - 50 ‘and 0 o 200 psi w:th tube ‘system CQpl”Gry |engths ‘of up to 50 ft.. The units tested were random selectlons from a lot of 25 and were found . to have an average accuracy of iO 43% full scale, with- ‘zero shifts between room temperature and 1400°F which averaged 11.7% full scale. " Tests have been performed on three units which employ the tube-system type of sensor but termi- nate in a four-legged unbonded stram-gage bridge. i _ _modification of its input-damping network. Results to date show .an average -accuracy of 10.25 full scale at a constant temperature, w:th»-"':__ repeatedly (after intervals ranging from one to zero shifts of +1.7% between room temperature and 1400°F. Although these units are the most ac- cwate and promising to date, span- shifts of £2% full scale when the tempero-'_ ture of the strain-gage housing is elevated o’ 150°F. One more unit of this type remains to be tested as well as two units whuch uttllze a two-- 3 !egged strom-gage brudge. ) | o | S5 J-\fl“‘\"‘ifi*&h“wfliflt ‘!Fr-- ¥ i =,“ g‘;;.,,m w\-’h s_‘“&'h (&2 Tests of Heliurc-Welded lnconel Sheothed Thermocouples in Sodwm ' . T DeLorenzo o Tests have been made in‘an uttempt to determlne S the effect of @ hellorc-welded !nconel sheath on the calibration of a thermocouple. The data indi- f{ 4J T. D,Lo,emo and W. R. Miller, ANP Quar. Pro f Rep. June 10, 1956, ORNL-2106, p 42, Fig. 1:3.1. F N [). cate that the weldment produces no noticeable | o , . tests show that | temperature variations at the - strain gage - produce__. ) i, Mg B g AL ST 2l n,fufi‘ ] ~ purged switch showed indications of plugging. - The marked improvement produced by the nitrogen .~ purge is illustrated in Fig. 1.3.1. : ‘*:teSting"of‘the_syst_em ‘both switches will be purged. } PERIOD ENDING SEPTEMBER 10, 1956 deviation of the cahbratlon from the normal changes induced by aging effects over the interval of the tests. Equipment is being fabricated and assembled for aging tests of such welds in static sodium at temperatures above 1000°F for 3000 hr. A preliminary investigation of possible in-pile calibration tests is under way. For these tests sodium vapor pressure would be used as a tem- perature standard. The instrumentation require- ments and the system design are being studied. Thermocouple Data Reduction J. T. DeLorenzo | Life tests were begun on two 5ynchron|zed h:gh- speed mercury-jet switches, described prev;ously, - which operate at 1800 rpm and scan 80 thermo- ' couple signals per revolution. The input signal to the transmitter switch consists of a voltage- 'dlwder network with 79 one-millivolt increments i and one 0.5-v tap, which triggers the sweep of a 17-in. oscilloscope. The ‘oscilloscope monitors the common line between the switches. The input signal produces o stair-stepped trace that ranges on the oscilloscope from 1 to 79 mv in 79 steps. It was found that any ten consecutive output points ~on the receiver switch could be recorded reliably on a standard . 12-point recorder after a slight Initially, the life tests had to be terminated three weeks) because of plugging in the jet nozzles. After each case of plugging, the switches were disassembled and examined. -heavy black scum was found on the surface of the An each case, a " mercwy ond adhering to the various surfaces of “the mercury pool. A chemical ahclysistevealed that the scum was primarily mercurous oxide. ~ in an ottempt to eliminate the formation of this - oxide, one switch was continuously purged with “dry nitrogen. After approximately 20 days the un- In further life ey, . - & 37 nt_g-\vgew e Rt ik ek 4., wm&wwf* ANP PROJECT PROGRESS REPORT "UNCLASSIFIED Fig. 1.3.1. Nozzles of Two Mercury-Jet Switches Showing Clean Condition of Switch Thot Was Continvously Purged with Dry Nitrogen in Comparison with Mercurous Oxide Scum Formation on Switch That Was Not Purged. 38 o ( PERIOD ENDING SEPTEMBER 10, 1956 1.4. COMPONENT DEVELOPMENT AND TESTING | H. W. Savage - PUMP DEVELOPMENT TESTS E. R. Dytko! A. G. Grindell . Bearing and Seal Tests W. L. Snapp! W. K. Stair? A petroleum !ubncont for the ART reoctor pumps® has been specified on the basis of present infor- mation, but the search for better lubricating and - cooling fluids for pumps has’ continued. . program for determmmg the compatibility of vari- . ous lubricating fluids with the reactor process o fluids. (sodlum ond fuel) has been ‘undertaken. The" first tests. were conducted with Dowtherm A ond Cellulube 150 heated to 200°F under -a helium - atmosphere, The heated Iubrlcont was odmtfled to - g container of sodium at a temperature of 1100°F, which was alsé hefium blanketed. The reaction of the Cellulube 150 was vigorous, enoUgh to be near : exploswe proportions, while there was’ very little reaction between the molten sodium and Dowtherm - A. These: results ‘were- ‘sufficient to preclude -~ further Cellulube teshng. ‘A loaded journal-bearing test was conducted with . Dowtherm A as the lubricant and with an applied bearing load of 250 Ib. This test was terminated after 659 hr of trouble-free opercmon " The total measured seal Iéokdge was 1 cm® at the upper seal and 12 cm? at the lower seol The process sides of both seals were found to have the same type of gummy carbon deposit as that found previ- ously in tests in which Dowtherm A was used as the lubricant.® There was ‘some visible evndencefl - of bearmg ‘rubbing, which was opporenfly mmor* since no deleterious effects could be found in either. the beonng operoflon or.in- the drlve-mofor* : power trace. ‘ : “An mveshgohon of the possnble effects of in- creasing the present reoctor-pump (fuel ond sodium) - rotary-element _operating speed range "is to be - made. _operating speeds ‘and increased bearing loads. * Similar fesfs are plonned for the NaK pump rotory 'element On osugnment from Prufi & Whnney Alrcroff E _ Consultonf from the Umversny of Tennessee. - 3W L. Snapp and W. K, Stair, ANP Quar Prog Rep June 10, 1956, ORNL-2106, p46 A test ~ material. Tests will be storted soon with’ hlgher - Excessive axial motion was found in the original NaK pump thrust-bearing design. A change to a duplex pair of angular-contact ball bearings mounted face-to-face has provided the desired axial rigidity without sacrificing transverse flexi- bility. A seal design proposed for the NaK pump by a vendor is now being tested. In general, it has been found that the vendor’s seal operates with modest leakage rates during the early part of a ~.test but that subsequent operation always results _in much higher leakage rates.’ ‘seal, similar to that to be used in the reactor A bellows type of has been tested for NaK pump applica- Further tests pumps, 4 tion, with inconclusive results. ‘should result in a modification of this type of sea! which will give acceptable performance. Modified D_urometollic seals® have been received and are to be tested soon. Testing of elastomers for seal application has been impeded by delays encountered in obtaining test samples of recommended materials. A test was conducted, however, in which Cellulube 150 and Buna-N base elastomers were used. Since Cellulube 150 requires the use of butyl rubber elastomers, the results, as expected, were poor with respect to strength and hardness of the seal The Buna-N O-rings did, however, main- tain a satisfactory seal throughout the 800 hr of test operation at 220°F. It is evident that more ~ compatibility data are necessary for a realistic evaluation of elostomers, since the unsatisfactory performance observed thus far in tests at 220°F is " in contrast to' satisfactory results reporfed for " tests of elastomers at 150°F in other fluids.5 Tesfs of Gos Aflenuotlon by Sea!s S M. DeComp, Jr. V. K. Staie Tests have been concluded that were designed “to determine whether there would be fission-gas " leakage - from the sumps of the ART fuel pumps ~into the 7;'Iubricat_ing oil catch basins and thence ) ‘W L. Snopp end W. K, Stolr, ANP Quar Prog Rep. March 10, 1956, ORNL-2061, p 43. SUCON Fluids and Lubricants. Carbide and Carbon Chemicals Co., F-6500D, 1955. 39 ANP PROJECT PROGRESS REPORT into the lubricant reservoirs external to the reac- - tor cell. Attenuation of the fission-gas concentra-.. tion by a factor of 10'® from the fuel expansion tank to the lubricant reservoir was considered to be necessary to eliminate biological shielding of the lubricant systems. Preliminary test data in- dicate attenuation by a factor of at least 104 from the expansion tank to the pump shaft seal and by another factor of at least 104 between the seal and the reservoir. These data are being analyzed to determine the | upper limit to be expected for the radiation dose from the lubrlcahng-orl tank. Fuel Pump Development Tests S AW Simon ! M. E. Lackey "~ G. Samuels Developmental test work on the fuel pump im- peller’ continued, with water belng used as the pumped fluid. Efforts are under way to reduce the maognitude of the system pressure fluctuations, to produce the desired gas pressure gradient down the shaft annulus in the direction of the purge gas fiow, to determine the fuel flow through the centri- fuge, to reduce the pressure level of the system to meet the stress limitations of the north-head deck plate, and to obtain cavitation data. Some of these problems are, as yet, only partly solved. Part of the work on the reduction of the magni- tude of the pressure fluctuations has been carried on in conjunction with the efforts to obtain the desired gas pressure gradient down the pump shaft into the expansion tank and out into the off-gas line. These two problems are interrelated -be- cause both are affected by the upper seal face of the slinger impeller, area 1 in Fig. 1.4.1. The increase in the pressure fluctuations that was experienced duwing operation with high liquid levels in the expansion tank can be attributed to the priming and unpriming of the slinger impeller or perhaps to a hunting of the liquid level on the face of this impeller. As changes were made to the slinger impeller in attempts to remove the pressure fluctuations, the changes .in the gas pres- ‘sure gradient were noted. With the original slinger impeller design, complete with 12 blades on the upper. surface, both the fluctuations and the re- verse pressure gradient were at their maximums. The reverse pressure gradient indicated that the upper seal face of the slinger impeller could act as an effective gas pump and cause gas pressure in the expansion tank to build up to greater than 40 ~ the gas sipply pressure, with the gas pressure in the oil-leakage catch basin in the rotary element being the lowest pressure in the system. This condition would be especially undesirable if the pump stopped abruptly or even if there were a sudden reduction in the pump speed, beccuse the gas pressures would be such that fuel might be forced into the seal region of the rotary element. The upper seal face of the slinger impelier has been modified so that it is a smooth disk, and the axial clearance, as well as the radlal clearance, of the impeller is kept as small as possrble This configuration keeps the amount of liquid in and around the impeller face to a minimum, regardless of the liquid level in the expansion tank. Opero- tion with this modlfied impeller produced a normal gas pressure drop in the direction of flow, with the oil-catch-basin gas pressure being ot an lnter- mediate level and acbove the expansion tank gas pressure. Further modlflcahons of this region are being tested in an attempt to provide a pressure equalizing volume that may be necessary for alieviating any possible unequal liquid level dis- tribution around the pump shaft. These tests will provide a basic shape for further development tests in the twin-pump aluminum-north-head water tests, where the effects of unequal pump speeds can be noted. ‘ A ‘second trouble area with respect to pressure fluctuations was found to exist at the lower face of the slinger, that is, the degassing. face of the slinger impeller, which is area 2 of Fig. 1.4.1. The first mocllficahon for correchng thls situation consisted in simply rémoving the vanes, or blades, from the impeller face. The resulting large clear- ance rendered the degassing feature of the centri- fuge ineffective. The shape that appears to function acceptably consists of a smooth dlsk that maintains the previous clearances. ‘ Mcdifications were also made to the seal plate over the centrifuge cup, area 3 of Fig. 1.4.1, in an effort to reduce the centrifuge leakage flow. A lip was added that extended vertically downward from the inner surface of the seal plate to the top of the centrifuge cup. This resulted in operation that was, in general, poorer than before. The pressure fluctuations were greater and the degas- sing time was longer ‘because the greatly in- creased velocity of the liquid discharging into the already restricted flow area only further restricted the centrifuge cup inlet flow. When this. lip was dp PERIOD ENDING SEPTEMBER 10, 1956 SECRTT - ORNL-LR-DWG 16015 _— _— . - EXPANSION VOLUME ~ % — < =S EXPANSION VOLUME | ®/ , ' L— SEAL PLATE fi[ — . : r_- -‘]/——— 3 i 3 Y — SLINGER IMPELLER —MIXING CHAMBER N R SRR A L1 h CENTRIFUGE SEAL VANE CENTRIFUGE BLADE — ‘ CENTRIFUGE BAFFLE —1 | ' : ~~ CENTRIFUGE HOLE .PUMP | ' ‘ ' ' - " piscHARGe | FUEL PUMP 1 , ; / N = v - ' : / . PUMP SUCTION - ' \ I Fig. 1.4.1. 'Dfugrmfi ;f ..'Iu_npeller Region of ARf Fuel Pdmb. o | 41 ANP PROJECT PROGRESS REPORT removed and a similar one placed on the outer surface of the seal plate, the operation of the pump was improved. In locating this lip on the outer surface, the axial clearance from the lower edge of the lip to the top of the centrifuge cup was made identical to that existing between the top of the centrifuge seal vanes and the lower side of the seal plate. While this modification increased the resistance of the flow passage from the centrifuge discharge holes, area 4 of Fig. 1.4.1, to the centrifuge inlet, it did not reduce the available head of the seal vanes on the top of the centrifuge cup and thus did not affect the discharge pressure of the centrifuge in any way. This reduced flow has less effect upon the character of the flow at the centrifuge inlet, and pressure fluctuations are therefore less. Another series of tests was performed to deter- mine the amount of flow that passes through the shaft feed holes, area 5 of Fig. 1.4.1, the ex- pansion tank, and the centrifuge. An indirect method was used to measure the flow, since it is virtually impossible to measure it directly. The internal xenon-removal feed flow up the shaft was blocked, and data were token while known xenon- removal feed flows were added to the expansion tank through on external circuit. The pressure difference between the expansion tank and the pump suction was plotted vs the external circuit flow for a constant main loop flow and a constant pump speed. This pressure difference was equated to the difference determined with the unblocked shaft. The shaft flow was computed from this pressure difference and was found to be approxi- mately 11.5 gpm. The reduction of the fuel system pressure level will be attempted by two methods. The first will involve reducing the size of the centrifuge dis- charge holes in an attempt to reduce the head de- veloped by the centrifuge. The second will in- volve increasing the internal diameter of the centrifuge cup to reduce the centrifuge head. Results of incomplete tests performed on each of these modifications indicated that each will allow a reduction of the pressure level to obtain an acceptable stress in the north-head upper and lower decks. The final selection of one of these modifications will be determined by future tests concerning the stability of the pressure level as a function of the liquid level in the expansion tank. 42 The results from some initial test runs indicated that a potential problem of cavitation exists with the present impeller blade desngn Values of Thoma's cavitation parameter,® o, are about 0.71 for 2400 rpm and 0.51 for 2700 rpm. An impeller with new blades having an improved entrance angle was tested for cavitation, and o was found to be about 0.54 for 2400 rpm and 0.41 for 2700 rpm. Investigations of the cavitation characteris- tics of the present impeller design will be con- tinved. Fuel Pump Endurance Tests S. M. DeCamp, Jr. An endurance test of an ART fuel (MF-2) pump' was started’ on April 10, 1956, and was terminated on July 2, 1956, after 1964 hr of operation. Op- eration during this period was satisfactory. Troubles experienced earlier® with gas flow down the pump shaft and with removal of oil from the lower catch basin drain did not recur. Removal of the pump from the pump barrel was relatively easy, and disassembly of the pump was accom- plished by heating the pump impeller and as- sociated equipment and removing the parts while they were still hot. The test was terminated when the operating sounds of the pump changed and the source of the sounds could not be definitely located. Examina- tion of the components after disassemb|y gave no indication of trouble in the pump. It was reas- sembled and installed in the cold mechanical shakedown stand, where it was run for 100 hr. At the end of this period the pump was operating “satisfactorily, with no leakage detectable at the lower seal. The pump and a new hydraulic motor were then reinstalled in the high-temperature endurance stand for further testing. At the end of 400 hr, the pump was still operating satisfactorily, with no sign of lower seal leakage. Operation of this pump'is continuing. A, H. Church, Centrifugal Pumps and Blowers, p 82, Wiley, New York 1944, s, M. DeCamp, ANP Quar. Prog. Rep. ]une 10, 1956 ORNL-2106, p 48. 8. M. DeCamp, ANP Quar. Prog. Rep. March 10, 1956, ORNL-2061, p 48. i - C Sodium Pump Development Tests S. M. DeCamp, Jr. Water testing of an ART sodium pump in a per- formance-acceptance loop, initiated previously,? was continved. The tests conducted previously were aimed at solving the problems of bypass flow around the main impeller and ingassing of the main fluid circuit. The bypass flow was from pump discharge through a pressure breakdown labyrinth to the expansion tank, and it was returned to pump suction. Bypass flows as high as 10 gpm were possible without ingassing with this bypass flow circuit. Examination of the reactor north-head geometry and thermal stress conditions, however, indicated that bypass flow of this type was un- desirable because of the low sodium temperature (1050°F) in the expansion tank. A preferred solu- tion is to bring hot sodium (1250°F) from directly below the lower deck of the sodium expansion tank to cool the top lid of the sodium tank and then to return this flow to pump discharge by means of a centrifuge. , Various conflgurcflons of the flow passages above the centrifuge have been tested. The most recently tested configuration is shown in Fig. 1.4.2, and performance curves obtained for the configuration are presented in Fig. 1.4.3. It became apparent in earlier tests that the size of the slot in the side of the pump barrel was in- adequate to handle the return flow of liquid to the centrifuge. In order to handle the required flow, a second slot was cut into the barrel below the original slot. The addition of this slot increased the maximum obtainable flow into the centrifuge tion of the loop anury and Auxdlary NuK Pump Developmeni Testsm H C Young J.G. Teague 7 The flrst Inconel stahonary assembly - for an ART primary NaK- pump (PK-P), consisting of volute and pump tank, was received. This as- . - serrbly was welded into a hlgh-temperature test-fif" ‘95, M. DeCamp, ANP Quar ng Rep ]une 10, 1956 -' ORNL-2106, p 50. mThe pr‘imary NaK pumps are to be used to clrculo'fe. : NaK through the ART fuel-to-NaK heat exchanger sys- tem; the auxiliary pumps are to be used to circu- late NaK through the ART sodium-to-NaK heat exchanger system. PERIOD ENDING SEPTEMBER 10, 1956 loop made up of approximately 60 ft of 3}’2- and 4-in. IPS Inconel pipe. A heat exchanger and the necessary instrumentation were added to the loop, and water tests were conducted. Initial tests indicated that the pump would not prime because of excessive quantities of gas having been trapped in the discharge pipe during the loop filling operation. A ¥ .-in.-dia hole was therefore drilled through the top of the discharge pipe inside the pump tank to permit the trapped gas to be vented during filling. After the pump was primed, a continuous leckage of 2 to 3 gpm flowed through this vent to the pump tank. A baffle and a deflector were developed and in- stalled to control the splashing caused by this high-velocity leakage. The pump primed satis- factorily after the vent was installed, and such a vent will be mcorporuted into all future PK-P pumps. . - The water test performance data obtained on these Inconel pump parts agree very well with “initial data obtained with the original brass im- peller and aluminum volute;!! however, a slightly higher efficiency was obtained in these tests. The cavitation performance of the Inconel impeller was not so good as that of the brass impeller; however, the cavitation characteristics are still considered to be satisfactory. Weld shrinkage tests of pump volute halves have shown that the shrinkage can be controlled to - within oacceptable tolerances (see Chap. 3.4, *Welding and Brazing Investigations’). Several pairs of Inconel volute halves, with the volute passage profile mochmed hcve been received from the- vendor ond, at the same time, caused more stable opera- - HEAT EXCHANGER AND RADIATOR DEVELOPMENT TESTS E. R. Dytko R. E.'rMacP'l-ners'on' J. C. Amos | -lnt_ermedidik'erHecit,Eié_hanger Tests _ J.w Cooke ! H. C. Hopkins ! A summary of intermediate heat exchanger test - stand operations ‘during this quarter is presented - in Table 1.4.1. York radiator No. 9 completed a " - program of 30 thermal cycles and was removed from " test.stand A for metallographic examination. The My, c. Young, J. G. Teague, and M. E. Lackey, ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 51. 43 ANP PROJECT PROGRESS REPORT SEORER ORNL-LR-DWG 18132 SURGE TANK LIQUID LEVEL = N ©—— (B —>= ——= HELIUM PURGE GAS """" — -~ GAS SEPARATED FROM LIQUID (©)—-—= —-—= LIQUID FLOW THROUGH CENTRIFUGE (©)—> — MAINFLOW @——-— et C+ B Fig. 1.4.2. Diagram of Impeller and Centrifuge Regions of ART Sodium Pump. 44 PERIOD ENDING SEPTEMBER 10, 1956 SECRTY ORNL-LR-DWG 16133 140 120 |— : 3400 rpm wh—g : o ) — 3200 rpm =g 0~ 100 . . S —. | o _ - 3000 rpm deg | -5 g 80 . w 2700 rpm =—game_| ; .-‘."“o- 2 60 7 2400 rpm e - o= .-..__ 40 : 2100 7PM —gr— ~e=6 z | S - - u- 1800 rpm -o 8 —egg 3 150C rpm g4 0 20 T eete- & 0 : : 0 50 100 {50 200 250 300 350 400 450 500 550 600 : : FL.OW (gpm) Fig. 1.4.3. Water Test Performance Characteristics of ART Sodium Pump. TABLE 1.4,1. SUMMARY OF INTERMEDIATE HEAT EXCHANGER TEST STAND OPERATION Hours of Test Unit* Nonisothermal Total Hours of Number of Reason for Operatién Operation Thermal Cycles Termination Test Stand A York radiator No. 9 {revised ' 695 1283 30 Test completed "~ design) ' ' : ' Circulating cold trap Ne, 2 B , 1430 ! ' Test completed (4 in. in diameter) NaK screen filter o . ' 264 ' Test completed Test Sfund B Biock,_Sivalls & Bryson heat 1008 - 1398 ' 11 : Test completed exchangers Nos. 1 and 2 {(type 1HE-3) _ Cumbriage radiators Nos. 1 and 2 1466 , 2045 . 14 Test completed {modification 3) ' ' o Circulofing cold trap No. 5 o 1260 o Test completed (4 in. in diometer) *Includes only units tested during this quarter, 45 ANP PROJECT PROGRESS REPORT radiator performance data and details of the thermal-cycling program were reported previ- ously. 2 This radiator was the first 500-kw radia- tor of the revised design!3 to be tested. Post- operational examination showed no evidence of cracks in the radiator tubes. The available infor- mation on maximum mass transfer and total in- crease in NaK pressure drop for all radiators tested, including York No. 9, is summarized in Table 1.4.2, At the completion of York radiator No. 9 testing, test stand A was dismantled to provide floor space for the Engineering Test Unit (ETU). Analysis of the contents of the NaK screen filter tested in this stand revealed no significant amount of metal particles or foreign material. | Testing of Cambridge radiators Nos. 1 and 2 and Black, Sivalls & Bryson heat exchangers Nos. 1 and 2, type IHE-3, described previously, '2 was completed in test stand B. These test units are currently undergoing metallographic examination. The data on NaK pressure-drop increases and maximum mass-transfer deposits in the Cambridge radiators are given in Table 1.4.2. These radia- tors were of the original 500-kw design modified by removing the side plates and slitting the sup- 12 w. Cooke and H. C. Hopkins, ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 56. 13, W. Cooke, H. C. Hopkins, and L. R. Enstice, ANP Quar. Prog. Rep. March 10, 1956, ORNL-2061, p 52. - port plates and base plates. Me'tdllbg'rfq'phié__”ex» amination revealed cracks in the radiator tubes at the rigid base plate and support plates. With the exception of the degree of increase in NaK pres- suwre drop, the performance data obtained on these radiators and heat exchangers were in substantial agreement with data obtained from units previously tested. The NaK pressure drop in Black, Sivalls & Bryson heat exchanger No. 2, where heat was transferred from fuel to NaK, did not increase ‘throughout the test, whereas, in heat exchanger No. 1, where the heat was transferred from NaK to fuel, there was a total NaK pressure-drop increase of approximately 240%. This pressure-drop in- crease is attributed to mass-transfer deposits and has been observed as a function of time in all test units in which NaK has been cooled. ~ York radiators Nos. 11 and 12 (revised design) have been installed in test stand B, and tests will start as soon as Black, Sivalls & Bryson heat exchangers Nos. 1 and 2, type IHE-8, previously described, 14 have been received and installed in the stand. Construction of test stand C is essen- tially complete, with the exception of the installa- tion of the ART test radiator being supplied by York Corp. The delivery of this radiator has been delayed by fabricational difficulties. - e 4R, D. Peak e al., ANP Quar. Prog. Rep. Dec. 10, 1955, ORNL-2012, p 42, ' TABLE 1.4.2. SUMMARY OF DATA OBTAINED ON NoK PRESSURE DROP INCREASES AND MASS TRANSFER BUILDUP IN TEST RADIATORS Maximum Pressure Test Total Hours of Maximum NaK Temperature Drop Maximum Thickness Test Unit Stand Nonisothermal Temperature Differential Increase - ©f Mass Transfer Operation (°F) ©F) (%) Deposit {in.)* PWA-1 and 2 IHE-B 585 1500 500 50 0.004 ORNL-3 SHE-B 295 1500 230 30 €.0035 York-4 SHE-B 748 1500 265 30 0.003 York-9 IHE-A 695 1500 400 118 0.006 Cambridge 1 and 2 |HE-B 1466 1450 210 140 0.005 *Maximum mass ftransfer occurred in air inlet side (cold side) of the radiator in each instance. The Jensify of the mass transfer and the length of radiator tube over which mass transfer was found were greater in the units which ex- hibited the greater pressure drop increases. 46 Small Heat Exchanger Tests L. H. Devlin! ~ J. G. Turner! A summary of small heat exchanger test stand operation during this. quarter is presented in Table 1.4.3. Testing of York radiator No. 7 and Process Engineering Corp. heat exchanger No. 1, type SHE-2, is continuing in test stand B. The radiator NaK pressure drop increased approxi- mately 100% during the first 970 hr of nonisother- mal operation and remained essentially unchanged thereafter. This latter period of operation has been at a maximum NaK temperature of 1500°F, - with a radiator NaK temperature drop of 100°F for 216 hr and 430°F for 144 hr. The radiator was initially placed on power op- eration in a series of steps designed to simulate temperatwe conditions which will be experienced by the ART radiators as the reactor is taken to PERIOD ENDING SEPTEMBER 10, 1956 full power. Radiator NaK temperatures, operating times, plugging-indicator maximum break tempera- ture, and increase in radiator NaK pressure drop for these steps are given in Table 1.4.4, During the test, cold-trap flow was maintained at 0.25 gpm (equivalent to rate of ART cold-trap system flow) and an outlet temperature of 300°F. ‘Testing of the ‘Process Engineering Corp. small heat exchanger No. 1, type SHE-7, in test stand C " was interrupted by the failure of York radiator No. 5, shown in Fig. 1.4.4. This was the last ‘medified radiator of the original 500-kw design to be tested. The shift in elevation of the l/!é-in. support plates graphically illustrates the necessity of eliminating the rigid support plates and base plates from the radiator fin matrix in order to minimize thermal stresses. This radiator was replaced by York radiator No. 8 (revised design), and test operations were resumed. TABLE 1.4.3. SUMMARY OF SMALL HEAT EXCHANGER TEST STAND OPERATION - (4 in_.ri'n diameter, modifif:ation-l) 7 : Hours of Number of Test Unit* Nonisothermal Total Ho.urs** Thgrmol Status of Test ‘ Operation of Operation Cycles Test Stand B Process Engineering Corp. heat 1330 2165 13!’2 Test continving exchanger No. 1 (type SHE-2) * : York radiator No. 7 (revised - 1330 2165 - 133’2 Test continuing design) Circulating cold trap No. 6 (4 in. 2165 Test continuing in diameter, modification 1) ) Test Stand C Process Engineering Corp. heat ' 356 598 12 Test continving 'exchung'érrrlflor. 1 (type SHE-7) - ' Yor:k f_adi'afor No. 5 . 223 1265 22}'2 Terminated because of (modification 2) - ' _ radiator failure York ri_n.iki‘c'né_rr No. 8 {revised - 239 300 7}'2 Test continuing . design) _ . ' Circfildfing céid_ trap No. 4 298 : Replaced by cold frc.:lp (4 in. in diameter) - with stainless steel - _ cooling coil Circulating cold trap No. 7 300 , Test continuing *Includes only units tested during this quarter. **For tests in progress the operating hours are shown as of August 15, 1956. 47 ANP PROJECT PROGRESS REPORT The type SHE-7 heat exchanger, shown in Fig. 1.4.6. Endurance testing of thls heat exchanger is 1.4.5, was designed to operate at ART intermedi- currently under way. ate heat exchanger design temperature and flow A summary of the operatmg condmons and the conditions with a heat load of 400 kw. The actual corrosion of the fuel side of the heat'exchangers operating conditions obtained are shown in Fig. tested thus far is presented in Table 1.4.5. TABLE 1.4.4. STEP INCREASES IN OPERATING CONDITIONS FOR YORK RADIATOR NO. 7 Plugging indicator maximum break temperature: 300°F Radiator Inlef Radiator Outlet Hours at Increase in : Temperature Temperature Temperature Pressure Drop . Increase cF F) Indicated (%) (%/he) 1195 1125 20 0 0 1230 1160 25 0 0 1250 1070 17 0 - ' 0 1290 1070 72 14 0,192 1355 . 1070 30 3 0.100 1430 1070 49 14 0.285 1500 1070 98 35 0.365 TABLE 1.4.5. SUMMARY OF OPERATING CONDITIONS AND CORROSION FOUND ON FUEL SIDE OF TEST HEAT EXCHANGERS Maximum Test Total Hours of Operation at Yarious Maximum Depth of Unit Stand Hours of Fuel Temperatures Corrosion Remarks Operation 400_1500°F 1500-1400°F 1400-1200°F {mils) ORNL-1, SHE-A 1557 928 606 23 7 type SHE-1 ORNL-1, IHE-B 1129 965 164 12 _ Area where failure type IHE-3* occurred excluded in depth of cor- rosion determina- tion - ORNL.-2, IHE-B 1129 965 164 4 Examination of end type IHE-3 _ sections is not yet completed ORNL-1, - SHE-B 2071 214 412 1445 6 type SHE-2 *This heat exchanger transferred heat from NaK to fuel; therefore the NaK temperatures and the tube wall tempera- tures were higher than the fuel temperatures, This will not be the case in the ART intermediate heat exchanger. 48 PERIOD ENDING SEPTEMBER 10, 1956 49 &8 wl 3 < 45 gz 5o 5 After Failure in Test Operation, York Radiator No, 1.4.4, ige F 0S TCONEGELLLM. ORNL-LR-DWG 15289 MATERIAL: INCONEL ’ i : 25 TUBES {0.1875-in. OD, 0.025-in. WALLS) NaK OUTLET NaK INLET SECTION A-A NoX HEADER {0 in. {REF} HEAT EXCHANGER TUBES FUEL HEADER HEAT EXCHANGER SHELL FUEL OUTLET SETS COMB SPACERS INLET SPACER WIRE Q.031 x 0055 in 66in. Fig. 1.4.5. Small Fuel-to-NoK Heat Exchanger Type SHE-7. LAODIY SSIII0Vd LI23r0dd NV R ORNL-LR~DWG 46134 442 kw - 1471°F Nak-TO-AIR | 1046°F 328 psi| RADIATOR [35 pgi APy = 7.8 psi , Y ' 25 psi ) . ; NaK J 354.9Pm| oime ReNCIK = 450,000 i ] 874 psi APNOK = 54.6 psi ) . AT = 425°F o Y 1471°F 1046°F 32.8 psi FUEL—-TO—-NgK . 874 psi- 1594°F HEAT EXCHANGER 1247°F 56.6 psi - 25 psi | APFUEL== 31.6 psi . 25 psi Repyg = 3580 . FUEL ] 975 gom| oive _ 442 kw Y 1594°F | REsISTANCE | 1247°F 56.6 psi HEATER Fig. 1.4.6, Operating Conditions for Tests of Process Engineering Corp, Small Heat Exchanger No. 1, Type SHE-7. Conditions given are approxi- mately the design conditions for the ART, Cold-Trap Evaluation in Heat Exchanger "~ Test Loops F. A. Anderson 13 J. C. Amos Two 4-in.-dia circulating cold traps equipped with stainless steel cooling coils have been placed in operation with no difficulties from oxide ‘plugging. The procedure listed below, which has been proposed for - ART cold—trap startup, was followed in each case: - ‘1. Operate with maximum cold-trap Nak flow und no cooling until main NaK system reaches 1200°F. 2. Turn qir cooling to maximum and maintain the cold-trap inlet temperature above 1he pluggmg in- dicator break temperature. 3. Reduce the cold-trap outlet temperature by' reducing the cold-trap NaK flow as the break tem- perature drops. 4. When cold-trap outlet temperature reaches ' 300°F or less, transfer to water cooling and in- YSConsultant from the University of Mississippi. PERIOD ENDING SEPTEMBER 10, 1956 crease the cold-trap NaK flow to design rate. Ad- just water flow ‘to obtain desired cold-trap tem- peratures. During ART operation it may be necessary, for maintenance purposes, to cool the reactor to ap- - proximately 300°F and subsequently reheat it. A test was therefore run on the NaK loop of SHE test stand C to determine whether such operation would seriously upset the oxide balance or other- wise cause cold-trap operating difficulties. With a system temperature of 1100°F and the cold trap operating at approximately 0.75 gpm at a NaK out- let temperature of 300°F and an oxide saturation temperature of 270°F, the main loop was cooled to 300°F and reheated without disturbing the cold- trap NaK or coolant flows. No adverse effect on the cold trap was noted, and at no time during the course of the test did the oxide saturation tem- perature rise above its initial level. The results of this test are plotted in Fig. 1.4.7. In order to obtain a comparison of actual and predicted over-all heat transfer coefficients for the cold-trap circuit, a series of tests were run on the 40-in. economizer and on the 4-in. circulating cold trap on test stand C with water as the coolant passing through the slightly flattened, copper cooling coil wound around the cold trap. Experi- mental over-all heat transfer coefficients were ob- tained that ranged from 9.9 to 20.9, in comparison UNCLASSIFIED ORNL-LR-DWG 16135 1200 1000 _ 8OO o MAIN SYSTEM NaK o TEMPERATURE @ 2 600 x Y COLD TRAP NoK OUTLET 2 a00 TEMPERATURE L= - 200 ' PLUGGING INDICATOR = . BREAK 12 & = 0 o o TRAP FLOW 5 : 08 8 z o 06 g 0 10 20 30 a0 50 60 OPERATING TIME {hr} - Fig. l..4.7. Cold'Trdp. Operation During Cooling and Reheating of Main NaK System in SHE Test Stand C, 51 ANP PROJECT PROGRESS REPORT with predicted values of ~20 Btu/hr-ft2.°F. The coefficients increased significantly with increased NaK flow (range, 0.45 to 1.5 gpm), but they were essentially unaffected by a 30% variation in the water flow rate. A plot of the over-all coefficient, U, vs the logarithm of the NaK flow rate, x (in gpm), yields a straight line, which can be repre- sented by the following equation: log x = 0.049U - 0.855 . The over-all coefficients obtained with air used as the coolant ranged from 7.2 to 10.7, in compari- son to a predicted value of ~ 10 Btu/hr-ft2.°F, and appeared to be, within experimental error, almost independent of NaK flow rates in the range from 0.50 to 1.0 gpm. For both air and water cooling, the over-all coefficients were based on the inside circumferential area of the packed section of the cold trap. In the economizer unit the over-all coefficients (based on the outside area of the inner pipe) ranged from 458 to 664, in comparison with an average calculated valve of 656 Btu/hr-ft2.°F, and increased as the NaK flow rate increased from 0.50 to 1.5 gpm. Since the ART cold traps will be equipped with %-in. stainless steel cooling coils, the tests re- ported above were repeated for an identical 4-in. cold trap wound with the ART type of coil. With water as the coolant, the over-all coefficients ranged from 17.9 to 29.0 Btu/hr-f12.°F as the NaK flow rate was increased from 0.43 to 1.25 gpm. These values for the coefficient are approximately 50% greater than the values obtained for the trap wound with the copper coil. |t is thought that the higher values are a result of better contact of the stainless steel cooling coil with the cold-trap shell. With air as the coolant, the effect of the better contact was again noticeable, the over-all coefficients for the new trap being somewhat better than those obtained for the original copper coil trap. WATER FLOW TESTS OF ALUMINUM NORTH-HEAD MOCKUP E. R. Dytko R. E. MacPherson R. Curry! ~D. R. Ward It was originally planned that the ART fuel sys- tem components of the north head would be fabri- 52 cated and water-tested and then that the sodium system components would be added for testing. The desire to extend the fuel circuit testing and the urgent need for starting the sodium circuit testing have led, however, to the decision to build a second oluminum north-head mockup for water tests of the sodium system components. The aluminum parts for the sodium-circuit water tests. are about 75% completed. Water tests that simulated reactor .fuel system filling revealed that the addition of water into the bottom of the system at a rate of 1 gpm with both pumps running at 200 rpm or less caused only minor ingassing. Most of the gas in the system apparently vented upward into the surge tank during filling. At pump speeds of 400 rpm, how- -ever, the trapped gas was mixed thoroughly into the liquid being pumped, and extensive ingassing was observed. In another set of experiments, water was drained from the simulated fuel system while both pumps were running at various matched speeds from 800 to 3000 rpm. It was found, in all cases, that the liquid level dropped to the floor of the fuel surge tank before ingassing could be observed. During this drop in liquid level the pump discharge and suction pressures dropped in unison, and the pumping heads remained essentially constant, When the ingassing occurred, the pump suction pressure was within 1 psi of the surge tank pres- sure. Final confirmation of satisfactory perform- ance, with water, of the twin fuel pumps in the north-head configuration will be sought when a final impeller design has been established, based on single pump tests. Work has progressed on solving the problem of pressure and flow surging in the test system. As reported previously, 16 these pressure fluctuations were believed to be due partly to the hydraulic characteristics of the pump centrifuge section and partly to the unfavorable pump inlet configuration. In an attempt to separate these two effects, the external circuit was rebuilt to provide a 10-dia straight inlet to the pump suction. To further idealize the pump inlet, straightening vanes were provided in the inlet section. Subsequent opera- tions demonstrated a 75% reduction in pressure and flow fluctuations, and it is felt that some ME. R, Dytko et al., ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 22. : portion of the remaining instability will be re- lieved by the final impeller design. The sensitivity of the loop to pump inlet geome- try has demonstrated the desirability of mocking up, as a later step in the fuel circuit mockup test program, the actual reactor fuel pump inlet region. As a further refinement to the test unit, provisions are being made for the later addition of a full- scale aluminum core mockup that is to be tested in conjunction with the mockup of the reactqr pump inlet region. T DUMP VALVYE DEVELOPMENT TESTS E. R. Dytko R. E. MacPherson L. P. Carpenter M. H. Cooper! Four ART prototype dump valves have been re- ceived from vendors, but none have met the per- formance specifications for ART operation. . _In all cases the seat leakage rate was excessive, 17 Other difficulties encountered have included oxida- tion of the flame-plated stem in air and fcnlure of the seat-ring braze joint. The cause of the braze joint failure has been corrected, but a satisfcéfery oxidation-resistant flame-plated stem has nqt yet been tested. 18 71, P. Carpenter, J. W. Kingsley, and J. J. Milich, ANP Quar. Prog. Rep. March 10, 1956, ORNL-2061, p 60. 18 p, Carpenter, ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 62. PERIOD ENDING SEPTEMBER 10, 1956 The seat materials of the valves tested thus far were Kennametals 152B, 151A, and 162B (see Chap. 3.4, ‘'Welding and Brazing Investigations,"’ for compositions of these cermets). These Kenna- metals showed the greatest- resistance to solid- phase bonding in screening tests. ' The exterior of the valve body ‘will operate in air at a high tempe_rafure for an-extended period of time. Therefore a ‘stem plohng which will not oxidize is required. A Stellite:6 weld overlay on the stem at the stem guides and an aluminum oxide flame plate. are ‘to-be’ performance fested. The excess;ve seat Ieakoge of the first four prototype duinp’ ‘valves was caused by welding distortion and. mlsahgnmen'r of the assembly prior to final welding. - The assembly and welding pro- cedures have’ therefore been altered in an attempt to attain satlsfactory alignment. The seat ma- terials were satisfactorily tested for leakage, with water and air, ‘prior to assembly in the valves. In each case: “the, -leakage with water and air in- creased after the ‘valves had: been assembled and welded. The test results on the four prototype dump valves are-summarized in Table 1.4.6. Several - rngs are being constructed for testing . valve seat. motermls in contact with fused salts. In addition ‘to the Kennametals that are being tested, tests will be made on"molybdenum, copper, tungsten, and titanium carbide. For these tests the seat ring is brazed to fhe bottom of the valve TABLE 1.4.6. SUMMARY OF PROTOTYPE ‘DUMP YALVE TESTS IN THE FUEL MIXTURE (No. 30) NaF-ZrF - UF4 '(50-46-4 mole %) AT 1200°F : - _Pressure 7 S Profofyp_e ' _ Seat Leokage - 'Dif_ferenfial_ Total _ Valve No, - . (cm /hr) o Aeross Seat Stem Thrust o Reme;ks - L s () T _]V :‘."7'-'-‘_ 0.59 C 6 - 1200 _ Leak rate _E;;teosed with time . - 0.9 90 1200 : Removed,tadetermine c&iuse for sticking of stem 2 - , .27.1_3 6 1200 Valve remc;{red when sedt ring pulled loose 3 055 6 700 Valve would not open with 225015 stem pull L ' and would not fully close 4 R P - -6 - 1050 Spring-loaded valve operator used 35 ' 50 1050 After being closed for two weeks, seat leakage increased ropidly on opening and reseating 53 ANP PROJECT PROGRESS REPORT barrel, and the plug is rigidly mounted on the stem, which passes through an O-ring seal to a hydraulic operator. The seat ring and plug are shown in Fig. 1.4.8. Salt from a sump is forced through a dip line and against the seat and plug by helium pressure. The leakage collects in the valve body and drains through an overflow line into a tared receiver. To ensure the removal of gas pockets from the dip line, the plug is not seated in the ring until after the salt flows through the over- flow line. The stem thrust, the pressure dif- ferential across the seat, and the leakage are accurately measured. , The results of seat materials test No. 1, in which the seat ring was Kennametal 162B and the plug was Kennametal 152B, are summarized in Fig. 1.4.9. The valve had been cycled 15 times as of August 15, with the maximum leakage being 2.0 cm3/hr. The stem force is 750 Ib; the seat pressure differential, 80 psi; and the test tempera- ‘ture, 1200°F. After 1000 hr in test, the valve will be left closed for 500 hr to test for long-range self-welding. , , Seat materials test No. 2 was terminated 24 hr after its start because of excessive leakage.” The seat ring material was Kennametal 152B and the plug material was Kennametal 151A. The leak rate was 38 em3/hr at 25-psi differential pressure across the seat and a 750-Ib stem thrust. The high leakage was caused by misalignment of the stem. ' UNCLASSIFIED PHOTO 28937 IT—II_'._l_IIITT“II]I]I‘l‘lj-l‘ [ o 1 2 3 INCHES Fig. 1.4.8. Kennametal Seat Ring and Plug. PERIOD ENDING SEPTEMBER 10, 1956 GECREA- ORNL—LR-DWG 16t36 VALVE OPENED AND CLOSED; NUMBER INDICATES FORCE REQUIRED TO OPEN (Ib) o oo 0 Soowl 28 3 23 ® oRRB Qo vy v ¥ Y ¥ ¥ Y. ¥ 24 _ 28 RING MATERIAL: KENNAMETAL 162 B 2.0 PLUG MATERIAL: KENNAMETAL 1528 TEST TEMPERATURE : 1200°F LEAK RATE— i.6 1 R 1C0 - DIFFERENTIAL PRESSURE n{ ACROSS SEAT £ L 7 - S S DA Dt BoveBoeflioe e bractioelom Lol Lo dios e ol ke Ben e Ben iy et w12 &g bt t 7 3 <« Pl : K E @ i § Z ¥ . : T L < o w w . /\ i 1000 0.8 50 5§ — i = o = \1 \/ 2 E m > o T 750 0.4 —e e—8—g® *—0—0—=8 e—o—p—8— 25 0 - \ 5 = ISTEM THRUST ud W w l_ w 500 0 0 25 30 35 40 OPERATING PERIOD (days) Fig. 1.4.9. Results of Valve Seat Materials Test (No. 1) in the Fuel Mixture (No. 30) NoF-Z¢F-UF (50-46-4 mole %). OUTER CORE SHELL THERMAL STABILITY TEST E. R. Dytko R. E. MacPherson J. C. Amos L. H. Devlin The 300 thermal cycles scheduled for the outer core shell thermal stability test were completed. The model was disassembled, and the core shell was removed for examination. Dye-penetrant and x-ray inspections revealed no surface cracks or internal flaws. ' Measurements were carefully made of the inside surface of the shell by using the same procedure as that used for obtaining measurements after 57 cycles.!? The previous measurements showed the shell to be slightly elliptical, and after com- pletion of 300 cycles the elliptical deformation was even more pronounced. 5. Curry and A. M, Smith, ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 65. The core shell was subjected to sodium tempera- tures of 1200°F or higher for 1050 hr. A thermal cycle comprised 1 hr with a temperature dif- ferential and 1 hr of isothermal operation at 1250°F. The transient time was 10 min. Bursting pressures of 24 and 16 psi existed across the shell at the bottom and top, respectively. The temperatures of the sodium flowing inside and outside the shell were measured by thermo- couples located on the pipe walls at their respec- tive entrances and exits to the model assembly. During the temperature differential phase of a ‘cycle these temperatures were approximately as follows: Inner sodium entrance 1600°F Inner sodium exit 1235°F Outer sodium exit ' 1085°F Outer sodium entrance 980°F The temperature differential fromr the inner stream to the outer stream, which were flowing counter- 55 ANP PROJECT PROGRESS REPORT current to one another, was 515°F at the bottom and 255°F at the top. The core shell will now be subjected to a creep- buckling test at o temperature of 1500°F ond an external pressure of 52 psi. The time to buckle and the type of failure will be observed. A new shell will be installed in the thermal stability test apparatus, and a second test program, which will be identical to the one just completed, will be initiated. ' AUXILIARY COMPONENT DEVELOPMENT J. J. Keyes High-Frequency Thermal-Cycling Apparatus W. J. Stelzman J. M Trum_mel20 Apparatus and techniques are being developed for investigating the effect of high-frequency ther- mal oscillations and the resultant thermal fatigue stresses on the ART core. Based on volume heat 2°Consu|fant from the University of lowa. HOT FLUID COOL FLUID FLOW, q\ =3 N _ " VELOCITY, v, \ \-. HOT FLUID TEMPERATURE ]_-' AT (AT gax COOL FLUID TEMPERATURE TEMPERATURE TIME ~————— sowce data,?! a frequency range of from 1 to ‘ ‘ g 10 cps and a surface temperature amplitude of about 100°F have been suggested for simulating possible ART conditions. Dynamic mixing of hot and cold fluids appears to be a feasible method for achieving this frequency range and amplitude. A schematic diagram of e high-frequency pulse pump cutrently being developed is shown in Fig. 1.4.10. Alternate slugs of hot and cold fluid are dynamically mixed at the ‘T’ joining the legs of the system by pneumatically pulsing the level in two static pots joined to the hot and cold legs and connected to a pair of pistons operating in phase opposition. This design produces nearly sinus- oidal temperature variations in the common line, as well as steady flow, for convenience from the “point of view of data analysis. The amplitude of the temperature oscillations generated depends on many factors, but it can be MEAN GAS VOLUME, @, MEAN GAS PRESSURE, £ 21N, D. Greene et al., ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 222. : . GOMNFITTETHRLLL ORNL-LR-DWG 16137 EQUIVALENT DAMPING LENGTH = /o PIPE FRICTION FACTOR = ' GAS COMPRESSION EXPONENT =4, P-Q¥=C ANGULAR ROTATION OF CRANK SHAFT = w FREQUENCY OF PUMP PULSES = F Fig. 1.4.10. Schematic Diagram of High-Frequency Pulse Pump and Outlet Flvid Temperature Variation. 56 " made to approach, as a theoretical limit, A - (THot Leg — TCold Leg) = ATmax max _ 2 | . 2 * From energy-balance - consideration, it can be shown that the efficiency of the pulse pump, de- fined as the ratio of output amplitude to the maxi- mum amplitude, is given by the equation PERIOD ENDING SEPTEMBER 10, 1956 tially maximum efficiency up to 7.5 cps, with only a slight drop in efficiency at 10 cps. CoLD TVR_APS AND PLUGGING INDICATORS R. D. Peak? The first cold trap evaluation test stand, de- " scribed previously, 22 has been shut down and dis- 2 2\ ~1/2 l 4y 2 2 4 2 — [pyy | wp [L + I[— A AT wV 2 dq 2 dy A AT "2 R U | ‘ max max 7 P°7Td$ gckpofld% 40, 490, angular rotation of crank shaft, piston displacement, flow rate of fluid, equivalent damping length, = pipe friction factor, density of fluid, velocity of fluid, pulse chamber diameter, pipe diameter, mean gas pressure, mean gas volume, mean height of fluid in pulse chamber, length of pipe from pulse chamber to T, = gravitational correction factor, = gas compressron exponent (P-0% = a con- stant) : : : un u I o a0 . PN NG ~e g X 1/ TEMPERATURE RATIO, AT/A T yax, PER UNIT VOLUME OF PISTON DISPLACEMENT, V (ft3) ISOTHERMAL COMPRESSION {4 =1) Vi 0 l 0 2 4 6 8 10 12 FREQUENCY (cycles /sec) Fig. 1.4.11. Response of Pulse Pump as a Function of Frequency. 615°F (equivalent to 135 ppm O,), while chemical analysis of the NaK gave only 82 ppm O,. This type of plugging indicator was considered un- satisfactory, because the characteristic break in NaK flow was not obtained for oxide saturation temperctures below 400°F. This made interpreta- tion of low oxide contents virtually impossible. The second type. of plugging indicator used o plugging disk with 18 holes 0.31 in. in diameter and one hole 0.049 in. in diameter. The stand 58 operated for 25 days, during which time the plug- ging indicator was run 68 times and 47 samples were taken for chemical analysis with two Argonne samplers. Averaged data for a 24-hr period under constant conditions showed that, with the cold- trap outlet temperatwre at 900°F (equivalent to 470 ppm 0,), the plugging indicator breck. tem- perature was 640°F (equivalent to 160 ppm O,), while the chemical analysis of the NaK showed 170 ppm O,. Performance of this plugging indica- tor was, in general, satisfactory. S 65 ADONPIDENEL ORNL-LR-DWG 16139 THERMOGOUPLE 0.034~in~DIA HOLE 0.050-in—DIA HOLE 3jg=in. SCH 40 INCONEL PIPE Fig. 1.4.12, ART Plugging Indicator, 9561 ‘0l ¥3dWILdIS ONIGNZ QOld3d ANP PROJECT PROGRESS REPORT Y% —in-DIA. HOLES STAINLESS STEEL TUBING Va~in. INCONEL PLATE, ROLL TO 8%-in. OD YORK DEMISTER PACKING AIR OR WATER IN \{ NaK OUT —\ © BENEHBENT I ORNL—-LR-DWG 16140 THERMOGOUPLE WELL Fig. 1.4.13. ART Circulating Cold Trap for 70-gal NaK System, LIQUID-METAL-VAPOR CONDENSERS M. H. Cooper The specification of bleed gas purges through ART sodium and NaK pumps requires that con- densers be provided for removing liquid metal vapor from the exit helium streams. in addition, the NaK dump tanks must be provided with similar condensers capable of removing the NaK vapor from the helium exhausted by an emergency NaK dump. Development work on such condensers is under way, and the dimensions and design parameters of the units currently being investigated are given below: For Na For NaK Vapor Vapor Condenser diameter %-in.-IPS pipe l-in. tubing Condenser length 60 in. 36 in. Helium flow rate 500 liters/day 1.7 cfm Inlet temperature 1200°F 1200°F Outlet temperature 200°F 600°F 60 Schematic test layouts are shown in Figs. 1.4.14 and 1.4.15. For the NaoK test, helium at 1200°F, saturated with NaK vapor, was periodically vented from the 15-ft3 sump at a flow of 1.7 c¢fm and exhausted to the atmosphere through the con- denser. Two condensers were tested, the second of which was refined by adding external fins (2}’2-in.-dia Inconel, eight per inch) and internal Demister packing. - The first test, for which the conditions were those tabulated above except that the outlet tem- perature was 300°F, was terminated after the second cycle because of entrained NaK in the control panel. The second test, for which fins were added to the condenser and the outlet tem- perature was 100°F, was terminated after 70 cycles because of entrained NaK in the control panel. A larger trap downstream of the condenser was used in the second test and may have been responsible for the improved performance. The test results obtained to date indicate that the NaK condenser design is inadequate for ART g e He {4.7 cfm} — W =t — NaK {(1200°F) He (500 l'ifers/doy)—-- ‘3/ in. PIPE . B : CONDENSER 1-in. TUBING symMP o o TRAP. PERIOD ENDING SEPTEMBER 10, 1956 UNGLASSIFIED ORNL=-LR~DWG 16141 THROTTLE VALVE EXHAUST TO ATMOSPHERE ROTAMETER Fig. 1.4._14.i.|l.nlyout of Nc_l(i-Vupor-Condensgf'.l_'e;t Apparatus, UNCLASSIFIED ORNL—LR—DWG 16142 )__,,—;—60 in-—"""”-l EXHAUST . _ AL WET-TEST GAS METER TRAP Fig. 1.4.15, L‘.cl:you_t of Sod ium-Vupor-Condenser Test Apparatus, application. A reliable ‘gas‘-'lic';uid:' separ'a'tér_ must’ be used to remove condensed NaK entrained in the helium stream. - e Lo B For fhé 56dium"'fe§f; helibm' at a flow of 500 liters/day was bubbled through a small sump con- taining sodium ot 1200°F and vented to the atmos- phere through the sodium condenser. The sodium condenser plugged 10 hr after flow commenced. A sodium oxide plug had formed at the Swagelok joint between the sump and the condenser inlet _tube. Therefore the test equipment is being re- ‘designed to eliminate the mechanical joint. ZIRCONIUM FLUORIDE VAPOR TRAP F. A. Anderson M. H. Cooper High-temperature chemical absorbents and ther- mal-precipitation traps are being investigated in the program for the development of a zirconium fluoride vapor trap. The most promising system consists of a bed of hot aluming which reacts with gaseous ZrF , to form solid ZrO, and AIF,. All 61 ANP PROJECT PROGRESS REPORT the thermal-precipitation traps tested have formed ZrF , plugs at the initial cold section. The second test with an alumina bed, initicted previously,24 was terminated after 450 hr of opero- tion because of flow stoppage. X-ray diffraction analysis of the packing showed that the Al,O, had reacted with the ZrF , according to the follow- ing equation: 3ZeF, + 2A1,0, —> 3210, + 4AIF, Wet analysis of the packing showed that the reac- tion was about 90% complete. The wet analysis also showed that the plug was caused by thermal precipitation of ZrF ,, in the cooled outlet section of the test piece, ofter the Al,0, had been depleted. A shell-and-tube prototype trap2® was packed with 4- to 8-mesh Al,O, pellets. The schematic flowsheet for the test is shown in Fig. 1.4.16. - 24 M. H. Cooper, ANP r. Prog. Rep. June 10, 1956, ORNL-2106, p 63. Qua & Reb 25), J. Milich and J. W. Kingsley, ANP Quar. Prog. Rep. Dec. 10, 1956, QRNL-20'I2, p 60. Helium was bubbled through the sump containing the fuel mixture (No. 30) NaF-ZrF -UF , (50-46-4 mole %) at 1500°F at o rate of 5000 liters/day and exhausted through the trap, which was heated to 1350°F. This test, in which the fuel temperature was 100°F higher than the temperature anticipated in the ART, operated for 690 hr before fiig pres- sure required to maintain the ART design flow rate started to increase. Examination of the pack- ing after termination of the test indicated that the flow stoppage had been caused by thermal precipitation of ZrF, after the Al O, had been consumed. The test of UF, pellets as a high-temperature absorbent for ZrF, was unsuccessful. The ZrF, was. not removed from the helium, and plugs of thermally precipitated ZrF, were formed in the outlet of the test piece. | A thermal trap was modified by the addition of four water-cooled baffles, as shown in Fig. 1.4.17. Helium, ot the ART design flow rate of 5000 liters/day, was bubbled through the fuel (No. 30) at 1400°F and passed through the trap. This test was terminated after 231 hr as a resvit of a ZrF plug, shown in Fig. 1.4.18, at the first baffle. VEORCY ORNL-LR~DWG 16143 TRAP PACKED WITH Al;,0, RISER He {5000 liters /day) (1350 —1400°F} i —W_—-_\ ‘ e 40 in, ———fa— {2 in. — 1350-1300°F 1300 -80C°F Iz — -3 = _ .O —_— . = - —_—— - =__ - — - _ P WET-TEST | _ - = GAS METER FUEL (1500°F) . WATER TRAPS Figs 1.4.16. Schematic FI_&wsheet for Tests of Al, 0, Trap for ZrF, chér. ' 62 { i ~ PERIOD ENDING SEPTEMBER 10, 1956 SEGRET ORNL-LR-DWG 46144 COOLING WATER IN f WATER COOLING coiL THERMAL-TRAP TUBE BUNDLE BAFFLES ' - COOLING ‘ WATER OUT 5 in. 5 in. 5in. 5 in. 5in, I 50 in. =j £ i Fige 1.4.17. Modified Thermal Trap for ZrF ¢ Yapor, Fig. 1.4.18. Pfug of ZtF, at First Baffle of Modified Thermal Trap. Deposit broken through to show thickness. (Secret with caption) 63 ANP PROJECT PROGRESS REPORT 1.5. PROCUREMENT AND CONSTRUCTION W. F. Boudreau ART FACILITY F. R. McQuilkin Construction work on the contract portion of the ART facility in Building 7503 is in the final stages. Work on the building edditions, building alterations, and cell installation (package 1) has been completed, with the exception of installation of the last six circuit breakers (which were delayed by the Westinghouse strike) and completion of two modifications to the cell floor structure (which were ordered late in the contract period). Work performed during the quarter included com- pletion and testing of the cell; installation of mechanical items, such as the penthouse coolers, a 3-ton hoist, the air duct liner and insulation; electrical testing; painting; grading and paving; and miscellaneous cleanup jobs. The contractor vacated his site office on July 20, 1936. In order to ascertain the leck tightness of the reactor cell, the 24-ft-dia tank was completely assembled (with temporary test plates over all nozzles) and subjected to a rigid leak test. Fol- lowing a satisfactory pressure rise test of 50-hr duration, the annulus between the two vessels was flooded with 21,000 13 of circulating helium for the leak-rate measurement, By using standerd leaks of 14.7 and 28.24 micron ft3/hr on the sys- tem, it was established that the two Model 24-102 Consolidated Engineering Company leak detectors could detect a system leakage rate of 5 micren f13/hr or larger. Inasmuch as a zero leak rate was measured and the contract specification allowed 32 micron f3/hr, the vessel satisfoctorily met the leak tighiness requirements. Work on the installation of auxiliary services piping (package A) has been completed, with the exception of replacement of 15 valves in the nitro- gen system. Erroneous vendor information led to the installation of incorrect vaives for the service. They were rejected when pressure and leck testing - revealed the error and their inadequacy. Other- wise, work performed during the quarter included completion and testing of the contractorsestablished portions of the lube oil, hydraulic drive oif, cooling water, gir, nitrogen, helium, ond vent piping systems, The work which includes the installation of the diesel generators and facility, electrical control 64 centers, and spectrometer room electrical air con- ditioning equipment (package 2) was at the 91.8% completion point on September 1, 1956. All work has been completed on this contract, with the ex- ception of the installation and testing of the diesel generators. This work has been delayed because one of the deisel-generator units and three control panels were damaged in shipment, The damaged components were returned to the manufacturer for repair and are scheduled for reshipment early in October 1956, Special vibration testing of the units will be performed during the field tests to ascertain that all damage has been accounted for. Two other items of work were negotiated with the package 2 contractor for execution while he com- pletes his other work. The contractor was given an order to remove the shield blocks now occupying three sides of the ARE pit to nearby outside storage so that this building space can be used. To provide compliete access into the high bay through the north door and to provide greater flexi- bility in the use of the building cranes in unloading, the contractor was also given an order to remove the parapet wall above the floor level at an-eleva- tion of 852 ft around the ARE pits. The roof plugs for the pits will provide the floor surface over them, Design work continued and installation work was started on package 3, which concerns the installa- tion of process piping, process equipment, etc. Completion of the design is currently scheduled for October 1, 1956. During the quorter two of the four main blowers were installed. Some of the construction work may be seen in Figs. 1.5.1 through 1.5.13. Program and design planning for disassembly of the ART were initiated. The problem is being handled in two basic parts; the first being that of removal of the reactor from the cell, and the second that of transporting and disassembling it in hot- cell facilities. Procedures and tools necessary for removing the reactor are being developed. A new, large hot-cell facility is being planned. It is proposed that the new facility be constructed in conjunction with the construction of a group of smaller examination cells. The tentative location of these cells is on the north side of the ORNL Area adjacent to the Bethel Valley Road. Transfer o w s —— A ' UNCLASSIFIED PHOTO 177233 Fig. 1.5.1. View Looking South Across the 7503 Cell. In the foreground is the top of the cell inner vessel. At the top right is the concrete penthouse structure which will house the primary coolant pumps and their drive motors. To the left of the penthouse are the roof plugs which will cover the special equipment room immediately below. 9561 ‘01 ¥39WILd3S ONIANI 4oly¥3d AT UNCLASSIFIED PHOTOC 17960 o . . , % 3 b . - 4 Fig. 1.5.2. View Looking South from the North End of the Original 7503 Building. This view, showing the fop of the call water tank in place as it will be located during ART operation, was photographed during the vacuum-testing operction on the inner vessel, which is enclosed by the water tank. | ) __ - S LY¥0dIY SSITY20V¥d LI3r0¥d ANV O~ ~I UNCLASSIFIED PHOTO 17961 "."'i“g.'l_"‘l.5.3., Vievlv_: Lodkihg. N.orf'hl, Acfoss the Cell While the Water Tank Top Is in Place. At the lower left are the penthouse structure and the area in which the main coolant pumps and motors will be located. 9561 ‘0l ¥33IWILd3IS ONIANT QOId3d ¥ ¥ e ??‘i”?}a Fig. 1.5.4. View Looking Norfh Across the Opened 7503 Cell Tunk. The rectangular plafes shown on ihe floor sfructure of the inner. vessel will support the reucfor. The circular plates shown will support the shielded fuel fill-and-drain tank and fuel removal tank. The 24-m.-d|o nozzles through the sude wall of the vessel will provide for the penetration of the insfrumentaflon and control lines, lube oil lines, hydrouhc drive imes, cooling water lines, gas lines, and heater leads to the cell equipment. LA0dIY $SSIVO0V¥d LDIFr0dd dNY i 69 UNCLASSIFIED PHOTO 18080 Fig. 1.5.5. View Looking North Across the Completed and Painted 7503 Cell Tanks. The floor 'structure was removed from the inner vessel at the time this photograph was taken and therefore the fluid distribution weirs located in the bottom of the vessel may be seen. The scalloped ring shown outside of the weirs is the support for the floor structure. At the exireme bottom of the photo- graph are the NaK piping sleeves with their expansion joints which penetrate the two cell tanks. To the right of these sleeves are portions of three of the 24-in.-dia spectrometer tubes. 9661 ‘0l ¥IIGWILJIS ONIANI QOId3d 0L ' UNCLASSIFIED ' | .-, PHOTO 18032 Fig. 1.5.6. Vnew Lookmg Soufhwest Towutd the Modified Bui|ding. In th_e_'fkbr‘egr‘oondl cré _tHe two top héad's.fo'r ‘:the‘ cell tanks whnch hove been removed to storoge outsude the facility. o : ‘ . ‘ LY0d3IY SSIAO0Vd LIIr0dd dNV 1L UNCLASSIFIED PHOTO 18033 Fig. 1.5.7. Vie\.n'Looking. Southwest Toward the Diesel-Génerafir House. In the lower left is the substation for the 13.8-kv purchased-power supply from the TVA system. The gas cylinders in the lower center of the photograph will be used for storage of nitrogen during ART operation. 9561 ‘0L ¥39WILJIS ONIANI GoiId3d ZL - Fig. 1.5.8. View Looking West Inside the Diesel-Generator House Which Shows a Stage of the Insfallofion Work on the. Five Diesel-Generator Units for the Auxiliary Power System, 13043y SSIFH00dd LI23r0odd dNY UNCLASSIFIED PHOTO 12966 Flg. 'I 5 9, View 'Tuken in the Switchhouse Which Shows the Primory Switchgear and Instruments Associated With the Receipt and Dlstribuflon of the: Two Power Supplies to the Facility. The switchgear and instruments on the left will serve the i incoming purchused-power from TVA. Those on the right are for the diesel-generator power. ~4 W 9561 ‘0L ¥IIWILJIS ONIANI AOI¥3d vL T TR UNCLASSIFIED “PHOTO 18233 - Fig, 1.5.10. View Taken Inside the Blower House Which Shows the Supply End of the Main Cooling Air Duct and Installation Work on Two of the Four 82,000-cfm Blowers. On both sides of the main duct are the 10,000-cfm blowers which will supply air into the annulus between the building concrete and the insulated steel duct for the purpose of preventing overheating of the building structure, At the lower left is the opening for the ramp which leads down to the radiator pit beneath the cooling air duct. - 130d3Y SSIIO0Vd LI3rQ0dd dNV . . ¢ a roa ¢ " » L . e o 74 R e e et UNCLASSIFIED ~ PHOTO 18152 Fi.g;'Jll.S..i.l.\ ViewLooking -Wésf and from Inside the Cooling Air Duct. This view shows the 16-gage stainless steel facing over the insulated steel liner of the duct. The supply end of the duct is the opening shown in the center of the photograph. 9561 ‘01 ¥3GW3ILJ3S INIANI QOI¥3d 9L UNCLASSIFIED PHOTO 18153 Fig. 1.5.12. View Taken Inside the Main Air Duct from the Base of the Discharge Stack. This photograph shows the portion of the duct in which the NaK-to-air radiator banks will hang. At the right center are the exterior facing of the cell water tank ond the sleeves through which the NaK pipe lines will pass. Above the duct, fhrough the opening shown, is the penthouse. Below the duct, through the opening in the floor, is the radiator pit. LI0dTY $53490dd LI3r0dd ANV LL UNCLASSIFIED PHOTO 17734 e e Fig. 1.5 '|3. View Showing the Southwest Corner of the Auxilmry Equipment Room. This room, which was used during ARE operation as fhe heat exchanger pit, is being modified to serve as the auxiliary equipment room ond will house the lubricating oil pumping system ond the hydraulic drive ‘equipment. 9661 ‘01 ¥IGWILJIS ONIANT aold3d ANP PROJECT PROGRESS REPORT of the rodioactivé reactor from Building 7503 to the disassembly cell would be by a shielded *‘low boy’’ carrier and from the larger cell to the smaller cells by means of a conveyor in a shielded transfer - ond storage facility. ART-ETU REACTOR CONSTRUCT'ON M. Bender - G.D, Whltman . ETU Facility | To J. BO”esl » V. Jo Ke“eghun P,A.Gnadt = A, M. Smith Construction work has started on the ETU facility. Structural steel is being fobricated, and erection of the stand from which the NaK pumps and piping are to be supported has begun. The H. K. Furgerson Company is procuring steel which will be installed under the track floor to support the weight of the stand. Designs for this work are complete. Design of the control room enclosure is complete except for minor details. - The ‘details of the main NaK piping have been determined, and piping layouts are being reviewed. The economizers and the plug indicators for the NaK cold-trap loops are being fabricated. The design of the ait duct for the NaK-to-air radiators is being reviewed. The air duct is to be supported from the stand which supports the NaK system. A variable-speed blower will send air through the duct and across the radiators. The air will be discharged outside the building. A second layout of the piping for auxiliary services has been completed. The auxiliary services include instrument air, plant air, water, . helium, lube oil, and hydraulic oil. Changes in the system requirements have caused delays in the start of construction of these systems. Neces- sary changes have been made in the existing gas pipingto the furnaces which will be used to provide the heat load for the ETU. The basic electrical system has been de.fugned | A normal-supply substation of 1500-kva capacity will be used, This power will be distributed by means - of a 450-v switchgear unit, Individual circuit breakers in this switchgear unit will supply _ power to six wound-rotor NaK motors; to a process- air-blower motor; to a motor control center for the ‘reactor pump drive motors; fo motor control centers for "the lubrication pump drive motors; fo trans- " Ton assignment from Pratt & Whitney Aircraft. 78 former banks which will, in turn, supply power for controls and instrumentation, valve actuators, and control - motors; and to two sets of transformer banks which will, in turn, supply power to fhe heating circuits. An emergency electrical supply will also be available which will consist of @ 300-kw capacity, ' 480-v, 3-phase, diesel-generator - unit. Upon failure of the normal supply this emergency unit ~will supply power through automatic switching devices for an orderly shutdown of the equipment. The spare lube oil systems drive motors, instru- ments and controls, valve actuators, and some ~ necessary heaters in the system will continue to . function until it is felt that the experiment can be safely stopped or restarted. The drives for the fuel, sodium, and NaK pumps are not connected to -this emergency supply because of its limited - capacity. - Preheat and additional auxiliary heat required during the course of the experiment will be supplied by means of ceramic clamshell and tubular heaters designed for high-temperature applications. The 1500-kva substation, 460-v switchgear unit, and the emergency diesel gener- ctor have been installed, ART-ETU Reactor Fabrication and Assembly R. Cordova C.K. McGlothlan The study of the reactor assembly problem has continued, and general assembly procedures have been prepared and reviewed. A sequence of opera- tions has been outlined which covers the assembly of the reflector-moderator, nérth head, heat ex- changers, island, and the various shells into a complete reactor. Detailed instruction sheets are being prepared from this general outline,. which will serve as the assembly manual for the craftsmen who will assemble the reactor. - 7 The more detailed review of the reactor assembly problem has indicated the need for many jigs and assembly fixtures, and design of such tools is proceeding. A device for checking the concen- tricity of the reflector-moderator components about the polar axis during assembly has been designed. “Weld ‘shrinkage tests on geometries pecuhor to the reactor design are continuing (see Chap. 3.4, “Welding ond Brazing - Investigations® ). It is ‘necessary that the weld shrinkage be predictable to within a few thousandths of an inch if the close tolerances designed into the reactor are to be held. " This is porhcuiarly true in the fitup of the thin _lnconel shells surroundlng the reflector-moderator.- - Nearly all the component parts and spare parts -for three reactors have been ordered. Fabrication of the north - head for the ETU reactor has con- tinved. wflhout dlfflculty and is approxumately 40% | complete. - A gas-drying facllxty is. being desngned for ine stallation in the Y-12 foundry, This equipment “will be used to supply the controlled atmospheres - necessary for proper heat treutment of the north head and snmllar weldments, ‘ The lower half of the berylhum reflector-moderator has been contoured, and drilling of the coolant holes has been started. A rough dimensional check of the contoured piece of beryllium mdlcated that it was acceptable, The water-flow tests to be run on the first re- - actor assembled have been defined, and design work is practically complete on the equipment for the first of these tests, which is to be run on the beryllium reflector-moderator, Four water-flow tests are scheduled for the critical sodium circuits to determine flow distribution and pressure drops in the reflector-moderator and island sodium passages. Dummy ‘parts will be used in some cases in order to expedlte these tests. The construction work on the reactor assembly and inspection area was completed. This area has been air conditioned and a 20-ton bridge crane has been reinstalled over the building main floor for lifting-access to the assembly area through a hatch equipped with removable covers. Space has also been made available for a degreaser, a vacuum- drying chullty, and a storage area. for reactor parts. ART Cell Components ‘ A M Smlth - Desugn work on various ART ceII components', '.__.':cnd mstallatlon loyouts is contlnumg. Loyout _ . .drawings of the lead shield, the top support plate - form, the fuel flll-and-dram tank, piping manifolds, - ~_auxiliary piping, and the junction panel and junc- tion panel expansion joint have been completed, - and- fabrication and procurement are under way. " These. items constitute approximately 5% of the - * ‘equipment ‘needed - for the cell, -hydraulic piping layout ms|de the cell is approxi- .mately 90% complete, and the lube oil plplng layout is about 40% complete. Design of ‘the PERIOD ENDING SEPTEMBER 10, 1958 ART-ETU REACTOR COMPONENT ~ PROCUREMENT ) W. R. Osborn. J. Zasler - The fabrication of the fuel-to-NaK and the sodium- to-NaK heat exchangers still appears to be one of the most difficult procurement problems. Al- though considerable progress has been made, much remains to be done in refining the design ond in developing machining, tube bending, welding, and . brazing techniques for both these heat exchangers. Fuel-to-NoK Heut Exchangers The York Corp. has ordered all the toolmg necessary for production of the fuel-to-NaK heat exchangers based on the use of siretch-formed tubes and of channels made from premachined parts, which will be welded and stress-relieved -as afinal operation, Black, Sivalls & Brysen, Inc., feel that the channels must be machined after ‘welding to hold the required tolerances, and they are providing special tooling for their large boring ~mill for this purpose. They are also developing at least two different means of forming tubes to avoid some of the uncertainties of stretch-forming, Work is being carried out at ORNL, as well as at - the plants of the two vendors, on the various welding and brazing problems, and some design changes are being studied for overcoming existing difficulties with close tolerances, accessibility for welding, and distortion problems. - Sodium-to;Na'K Heat Exchangers Minor redesign of the sodium-to-NaK heat ex- changers and major improvement of the vendors’ - tooling have -been found necessary to permit fabris ~ cation of this unit ‘within the required tolerances. “This work is now in progress, and should result in’ ' the dellvery of the sodium-to-NaK heat exchongers "~ for the ETU durmg the next quarter, Core Shells The modlflcatlon of the Hydrospin machme to - increase the size of the shell_s that- can be pro- " duced has been completed, and a number of reason- ‘ably successful efforts have been made to spin the |s|und core . ‘shell. The mandrels for the outer _core . shell are ‘essentially complete, and both these shells should be available for shipment to ORNL - in September. The mandrels for all the remaining shells are in various stages of manu- facture and will be completed. during the next quarter, 79 ANP PROJECT PROGRESS REPORT Beryllium ReflectorsModerator - The Brush Beryllium Co. is now machining both * the north beryllium hemisphere for the ETU (to be shipped about October 5, 1956) and o hemisphere for the ART., The major difficulties in the fabri- cation of these units seem to have been overcome, Boron Layers Acceptable samples of finished boren carblde tiles have been received from the Nerton Company, and no major difficulties are expected with pro- duction of these components. Sample sheets of the boronecopper cermet have been received from Allegheny-Ludlum which are acceptable except for surface finish. Some difficulties are antici- pated in forming and finishing this material. 80 Pressure Shell An order has been placed for forgings for the reactor pressure shell with the Ladish Co. These forgings will be muchmed by fhe Allls-Cholmers Mfg. Co. |ncane| The availability of adequate supplies of Inconel in the proper forms that are acceptable according toe ART processing and inspection standards cons tinves to be a major problem and a limitation on the speed of fabrication for many items. A great deal of work is being done with the International Nickel Company and the Superior Tube Co. toward the solution of this problem. W " PERIOD ENDING SEPTEMBER 10, 1956 1.6. IN-PILE LOOP DEVELOPMENT AND TESTS H. W. Savage OPERATION OF IN-PILE LOOP NO. 5 VVC. C. Bolte! B. L. Greenstreet J. A. Conlin D. M. Haines! R. A. Dreisbach! M. M. Yarosh ‘In-pile loop No. 5 was inserted in the MTR on June 11, 1956, but it could not be filled. It was therefore returned to ORNL for disassembly and inspection. The fill system consisted of a fill tank, a vent line, and a fill line to the pump sump. In this system the fluid flows by gravity into the preheated loop when a plug of frozen fuel mixture is melted out of the fill line. Upon disassembly ‘and analysis a section of the fill line was found to have ¢ plug formed of a 50-50 mixture of fuel and zirconium oxide. Subsequent examination of the initial batch of fuel showed it to contain 0.5% zirconium oxide. It was also determined that repeated melting and freezing of the fuel salt could cause the zirconium oxide to separate out and settle in the bottom of the fill tank; however, it has not been possible to positively establish that this was the mechanism of formation of the oxide plug in the fill lines. Another consideration was the possibility of accidental exposure of the molten salt to moisture. However, the evacuation and purging of the system with inert gas, which were done prior to filling the tank, plus the evacu- ation before and purging during the attempted melt-out preclude the presence of. air-or moisture . in the system. - Close control was held on loop cleaning during assembly. Future. |oops will be filled ‘with salt” produced under stricter ‘control, -and every precautlon will be. conflnued to prevent “air and moisture from - entermg the - loop ‘prior ‘to - and after. charging of - the fill tank. - Increased - o efforts will be -made to-minimize - thermal cyclmg ‘durmg mmol f;llmg und melt-out of the flll system.' 10“ dé?ignmefif from Pran & 'w”"i'l;ne'y"Aircrdft, e e D. B. Trauger IN-PILE LOOP NO. 6 In-pile loop No. 6 is nearing completion and is scheduled for insertion in the MTR on September 3. This loop is similar to previous loops, with the principal variations being the mstallahoh of a modified (Mark 11) horizontal-shaft sump pump, lmproved electrical hermetic seals between the " nose and intermediate, or bearing housing, region, and the use of fuel in which the zirconium oxide content has been minimized. The planned oper- ating conditions for this loop are a 1600°F maximum loop temperature with the fuel mixture (No. 44) NaF-ZrF -UF, (53.5-40-6.5 mole %), a 300°F temperature daf?erenhal across the nose coil, and a power density of about 0.75 kw/em3. These conditions are to be maintained over two MTR operating cycles to give approximately 700 hr of operation with the reactor at full power. HORIZONTAL-SHAFT SUMP PUMP FOR IN-PILE LOOPS J.A. Cpnfin 'D. M. Haines - W.S. Karn! A prototype (Mark 11) horizontal-shaft sump pump for in-pile loops was modified to increase the radial clearance between the pump housing and shaft slinger from 6 to 60 mils in an effort to prevent shaft seizure such as that which occurred previously because of ZrF, vapor deposits.2 This _-pump, which is identical to that installed in loop ‘No. :6,-is now being tested and should have ac- ,cumulated 1150 hr of operation by the MTR in- : semon date for foop Ne. 6. The’ pump temperature in_toop-No. 6-will be hlgher than in previous tests, - buf it will be 50 to 75°F lower than the _temperature of the test pump now operating. ~It should therefore’ ~have @ lower rate . of zircomum fluoride ac- i c’Urfiuldtion than the test pump wnll have. : 2w. S, Kam, ANP. Quar. Prog. Rep. ]une zo, 1956. ORNL-2106; P 76. 81 ANP PROJECT PROGRESS REPORT 1.7. ADVANCED REACTOR DESIGN W. K. Ergen GRAPHITE AND BERYLLIUM OXIDE MODERATED CIRCULATING-FLUORIDE-FUEL REACTOR W. K. Ergen The complexities involved in the use of sodium to cool the beryllium reflector-moderator of the ART have directed thinking toward a reactor design in which the moderator is cooled by the circulating fuel. A new design now being studied is based ~ on a multitubular core, similar to that used for the ARE,. in which fuel passages would penetrate o moderator block or, altematively, moderator rods ‘would be immersed in the circulating fuel. If the fuel is to accept heat from the moderator, ~ the moderator must be at o higher temperature than the fuel. This eliminates beryllium metal, as used in the ART, from consideration! because of its poor mechanical properties at temperatures above the fuel temperature of interest, that is, 1600°F. Therefore graphite, beryllium oxide, and the hydrides of zirconium and yttrium are being considered as possible moderators. The required volume of the multitubular reactor core will, of course, depend largely on the desired total reactor power and the permissible power density in the fuel. A 24-in.-dia sphere has a volume of 1.2 x 105 em3, and, if it is assumed, optimistically, that 80% of this volume could be occupied by fuel, about 105 cm? of fuel volume would be available, Thus, in a 24-in.-dia multi- tubular core, 100 Mw of power could be generated with a power density of 1 kw/cm3, which is approximately twice the average power density of the ART, and 300 Mw would require 3 kw/cm3. It is clear that in many cases of interest, therefore, larger, rather than smaller, core volumes may be necessary. Thus far, this study has been limited to an evaluation of 24-in.-dia spherical cores, but the general conclusions are applicable to larger cores. The rate of heat generation in the moderator and the resulting thermal stresses would be of prime Vit would, conceivably, be possible to use beryllium in some reglons of the reactor, such as the cutside of the reflector, and to cool these regions with the rela- tively cold fuel coming from the heat exchanger. 82 A. M. Perry importance. The gamma-ray heat sources? have been shown to be prompt gamma rays plus U235 capture gamma rays (9 Mev), fission-frogment-decay gamma rays (1 Mev, based on 20% of the fuel being inside the core), gamma rays resulting from cap- tures in material other than U235 (5 Mev, depending strongly on the *‘thermal utilization” of the reactor), and gamma rays resulting from inelastic scatfering (1 Mev). This adds up to about 8% of the fission energy. The density of the fuel was considered to be 3.2 g/cm3, which is approximately the density of the fuel mixture (No. 30) NaF- ZrF -UF, (50-46-4 mole %), Hence, the radius ussumed for the core would correspond to 2 or 3 reloxation lengths of the gamma rays, even if aliowance were made for the moderator, which would occupy some of the core and would have a somewhat lower density than that of the fuel ‘Most of the gamma rays generated at the center of the core would be absorbed in the core. At points closer to the surface, the power generotion by gamma-ray absorption would be less. The gamma-ray absorption would be distributed between the fuel and the moderator approximately according to the densities. In a volume filled to 80% by fuel of density 3.2 and to 20% by graphite of density 1.8, the graphite would contribute about '/ to the density. Beryllium oxide, of density 2 8, if it occupied the same volume fraction in the saome fuel, would contribute about ¥ to the density. With a heat generation rate in the fuel of 1 kw/cm3 and a fuel volume fraction of 80%, 0.8 kw/cm® would be generated, on the average, over the core; 8% of this, or 64 w/cm3, would be gamma-ray energy, of which ‘é or 1’1'5 would be absorbed in the moderator, depending on whether the moderator is graphite or beryllium oxide. Since the moderator would occupy Y% of the volume, the power density in the graphite would be 40 w/cm?, and, in the bery!lium oxide, 60 w/ecm3. 2H. W, Bertini et al., Basic Gammé-Ray Data for ART Heat Deposition Calculanons, ORNL-2113 (Sept. 17, 1956). 3ANP Physical Properties Group, Physical Properties Charts for Some Reactor Fuels, Coolants, and Miscel- laneous Materials, 4th ed., ORNL CF-54-6-188 (June 21, 1954). at - e i e e M e gk e AL et n would shatter,” The heating as a result of neutron moderafion should, of course, be added to the gamma-ray heating. However, the neutron heating in the core would depend strongly on the design of the reactor, and, therefore, for the purpose of this discussion, it has been assumed. that it is compensated by the loss of gamma rays from the core. _ The permissible stress for graphite has been assumed to -be 2000 psi, and, for beryllium oxide,4 1000 psi. The stress data® and the power density data given above were used to obtain the following maximum dimensions for berylllum oxide and graphite bodtes in the proposed mulhtubular core: Flat Plates, Rods, ‘Thickness ~ Diameter 7 {in) {in.) Beryilium oxide 0.16 | 0.26_ ' Graphite - 5.6 2.1 As may be seen, if beryllium oxide were used, the structures in the core would have to be rather . delicate. - Graphlte, on the other hand, would allow rather rugged construction, in spite -of nuclear considerations, such as self-shielding, which would require o somewhat finer structure than that indicated by the stress limitations. -Furthermore, the cladding that would be required around be- ryllium oxide bodies would add appreciably to the ~ poisoning of the core, whereas the cladding around graphite, with its small surface-area-to-volume " ratio, would odd negligible poisoning. With the advantages of graphite in mind, ‘a three- - group reactor calculation based on the mulmubular: . core was performed on the Oracle. Accordmg to ~ this. CG|CU|Gfl°n.,G 24-in. -dia .'.phere Cm*fllfllflg' :‘fluorade-fuel reactor with graphite as a moderator - 80 vol %, -of the fuel mixture - (No, 30)- NaF-ZrF . : _"'UF (50-46-4 mole %) ‘and 20 vol % of: graphne,?' ' "W";'would be somewhat supetcritical. ~lation poison was consndered to be ‘present in the _'_form of a motertal havmg a macroscoplc obsorpflon Af substanha"y higher sh’esses, bery!llum cxide - . The consequences of such shattering. - - are, however, not quite clear and could be determined Concelvably such experiments "-‘of fast neutrons, whlch ‘on one hand mlght flnd only by experiments, - '-could make berylHum oxide appear more am'acfive fhun it is now considered to be, Sk, A Field, Temperature Gradient and Thermal Stresses in Heat Generatmg Bodies, ORNL, CF-54-5-196 (May 21, 1954). assumed in the Oracle calculation. In this calcu- " outer part of the reflector. PERIOD ENDING SEPTEMBER 10, 1956 cross section of 10% of the macroscopic absorption cross section of the fuel, both absorption cross sections being averaged over the core. This amount of poisoning is far in excess of that which could be caused by the cladding of the graphite. The results of the calculation indicated that there would be no need from the criticality viewpoint to use beryllium oxide in the core and that graphite would be an acceptable moderator. Since the reactor power and the permissible power density prescribe a relatively large core volume, the ~ somewhat poorer moderating properties of the graphite appear to be adequate. ~An_essentially infinite graphite reflector was Because of the relatively poor moderating properties of the ‘graphite, such a reflector would have to be rather thick. Also, graphite has a relatively small macro- ‘scopic removal cross section for fast neutrons, that is, 0.073 cm=1 compared with 0.14 cm=! for beryllium oxide, 0.17 cm=1! for nicke! and copper, ~ and 0.096 cm~ for zirconium metal.® The small removal cross section would result in large neutron leakage and intense sodium activation in the heat exchanger. On the other hand, the power density in the reflector would be smaller than that in the core, and there would be no need to use graphite. An *‘essentially infinite'® beryllium oxide reflector would be thinner than an ‘‘essentially infinite"’ graphite reflector, because of the lower age in beryllium oxide, and in any geometry, except plane geometry, the beryllium oxide reflector would give greater reactivity because of the larger solid angle ~which the core would subtend, as viewed from the “location of @ neutron ‘which had slowed down to thermal or any other given Iethargy. ‘Thus, there emerges a picture of a circulating- on the msnde, ‘where' the heat generation is large, and beryllitm oxide as @ moderator in ot least the ‘The “design of the interface between -the beryllium - ox;de and the " graphite remains to be established. “Even at the outsnde of the reflector fl'lere would ' ;be a certain amount of heof genercmon, and, hence, - fuel passages would “be required for ‘cooling. - Fissions in these passages would act as sources ‘ 6Ct:il'npufecl from data given by G. T, Chapman and C. L. Sterrs, Effective Neutron Removal Cross Sections for Shielding, ORNL -1843, p 22 ond 26 (Aug. 31, 1955). 83 their way back to the core and increase the re- activity, but which, on the other hand, would also leak out and increase the shielding requirements and the sodium activation. Therefore it would probably be advantageous to poison some of the outer fuel passages, but the radius at which the poisoning should start also has not yet been - established. - Both graphite and beryllium oxide are poor gamma-ray shields. 1t may therefore be.advan- tageous to introduce some heavy material into the moderator, at the expense of reducing the moder- ating effectiveness. The heavy material would protect the outer layer from gamma-ray heating and would make it possible to use beryllium oxide rather than graphite farther into the core and thus gain back the lost moderation. HYDRIDE-MODERATED CIRCULATING-FUEL ' "REACTOR ' A. M, Perry An alternative design now being studied includes the use of a somewhat more effective moderator in the core, such as the hydrides of zirconium or ytirium, than the graphite and beryllium oxide moderators discussed above. The hydride moder- ators would reduce the dependence of the reactor on its reflector for moderation and would introduce the possibility of using a heavy material, such as nickel or .copper, for the reflector. These materials have unusually large removal cross sections for fast neutrons, as given above, and are far more effective as gamma-ray absorbers than the moderating materials usually used as reflectors. Thus a significant portion of the gamma-ray shield might be placed immediately around the core, where its volume would be greatly reduced. While the density of radiation heating’ in a metal reflector would be greater than in a beryllium oxide reflector, the allowable stress would be much greater also, so that fewer cooling holes would be required for a nickel than for a beryllium oxide reflector, In the case of copper, the conductivity is so high that cooling might be required only at the inner and outer surfaces of an 8-in.-thick reflector. Preliminary calculations indicate that a few thousand pounds in shield weight might be saved by using a dense metal reflector. The effect of capture gamma rays in the reflector, as well as the activation of NaK in the heat exchanger, remain to be determined. Further calculations of fuel concentrations in reactors with metal re- flectors are also required. sl e reim e . o m————— Part 2 - * CHEMISTRY w.‘ R. Grimes * » 2.1. PHASE EQUILIBRIUM STUDIES! C. J. Barton R. E. Moore R. E. Thoma H. Insley, Consultant Phase equilibrium studies with a variety of binary, ternary, and quaternary systems have been continved. Each of the several methods described in previous reports of this series has been applied to several of the systems, Although earlier studies in this laboratory and elsewhere did not disclose its existence, careful examination of the LiF-RbF system has revealed the compound LiF:RbF. Careful re-examination of the LiF-UF , system has yielded evidence which strongly indicates that the compound 3LiF:UF, is metastable. Study of the RbF-UF , system has progressed to a point such that o tentative diagram for this complex system can be presented. The boundary curves, compatibility triangles, peritectic temper- atures, and eutectic temperatures for the NaF-RbF- UF, ternary system are now well established. The four-component system NaF-RbF-ZrF +UF, has been examined in some detail. Several mixtures which are of possible interest for reactor fuels have been established. Study of the NaF-LiF-BeF , system has progressed sufficiently for a plot of the boundary curves, compatibility triangles, and pertinent temperatures to be presented, although the composition of one of the ternary compounds is not completely es- tablished. The ternary system NaF-RbF-BeF, requires additional study before the analogous diagram can be constructed., THE SYSTEM LiF-RbF L. M. Bratcher A recently published? equilibrium diagram for the system LiF-CsF showed the existence of an incongruently melting compound LiF.CsF that was analogous to the binary compound previously observed3 on petrographic examination of mixtures containing LiF and RbF. Earlier thermal analysis 1 The petrographic examinations reported here were performed - by G. D. White, Metallurgy Division, and T. No McVay and H. insley, Consultants. The xway examinations were performed by R. E. Thoma and B. A. Soderberg, Chemlistry Division, : 2|, M. Bratcher, ANP Quar. Prog. Rep. June 10, 1956, 0RNL'2]06; P 90, Figo 2:147, 3L. M. Bratcher et als, ANP Quar. Prog, Rep. June 10, 1954, ORNL-1729, p 44. studies43 failed to reveal the existence of the LiF-RbF compound. The system was re-examined recently by the thermal analysis technique, and selected compositions were examined petro- graphically and by x-ray diffraction. A revised equilibrium diagram for the system, based on the results of these examinations, is shown in ‘Fig. 2.1.1. The shape of the liquidus curves gives evidence of complex formation. However, the difference between the melting point of the LiF- LiF.RbF eutectic and the incongruent melting point of the LiF-RbF compound (about 5°C) is less than the uncertainty of the temperature measurements. Petrographic examination of a mixture containing 50 mole % LiF showed that the birefringent complex was the predominant phase present; excess LiF and the complex were found in a mixture containing 70 mole % LiF. THE SYSTEM LiF-UF, B. A. Soderberg R. E. Moore H. Davis® The discussion of the system LiF-UF which was given previously,”’ mentioned a phase con- taining 20 to 25 mole % UF, whose formula and stability characteristics were unknown., Recently, thermal gradient quenching studies were made in an attempt to completely characterize this com- pound, The evidence from these and other quenched samples from this system indicates that there is a metastable compound in the LiF-UF, system whose formula is 3LiF-UF,. In the quenched samples the compound has the appearance of quench growth. The crystals are sufficiently well formed to give a good uniaxial-negative inter- ference figure and a refractive index of approxi- mately 1.514, but they have the characteristic mottled texture and wavy extinction of quench 4J. P. Blakely, L. M. Bratcher, and Cs J+ Barton, ANP Quar. Prog. Rep. Dec. 10, 1951, ORNL-1170, p 85. 5E. P, Dergunov, Doklady Akad. Nauk 5.5.5.R. 58, 1369 (1947)« ' 50n assignment from Pratt & Whitney Aircraft, 7R, E. Moore, ANP Quar. Prog. Rep. Dec. 10, 1955, ORNL-2012, p 77, 87 ANP PROJECT PROGRESS REPORT SOMRB Nt ORNL-LR-DWG 16145 S00 4 — -— 800 T ) ‘\ . . ‘f(— \ -~ T0C § i w / a 2 600 - h < o . . Wl o - - . 500 , N / _—“-____’_' . s L ———-hé-—u ————— 400 ' s [ L - 300 ROF 10 20 30 40 50 60 70 80 90 LiF LiF (mole %) ‘ ' Fig. 2.1.1. The Sysfem LiF-RbF. growths. X-ray diffraction patterns of these Since the quenched samples of 25 mole % UF ,-75 samples ore distinct and unique in the LiF-UF " system. Large quenched samples (~10 mg) contained 3LiF-UF 4 95 quench growth. Medium-sized (~5mg)or very small samples (2 to 3 mg) quenched in platinum tubes allowed the liquid to freeze to material with a glassy appearance and isotropic optical character, but even in these samples x-ray analysis showed the glassy material to be largely 3LiF.UF «+ There is no observable difference, optically or by x-ray pattern, in the 3LiF.UF 4 which occurs below the liquidus and that from quenches which were attempted well above the accepted liquidus. Under no conditions has 3LiF:UF, been observed as homogeneous, well- crystallized moterial free of quench growth texture. The inconsistency of the appearance, petro- graphically, of the 3LiF.UF , which occurs as a glass (isotropic, no optic figures) and that which occurs as quench growth, with texture gradation, is evidence of the metastability of the compound. 88 mole % LiF are more consistently single-phase than those of any other composition, the formula of the compound is considered to be 3LiF-UF‘. THE SYSTEM RbF-UF4 H. A. Friedman R. E. Moore Quenching studies on the system RbF-UF, are not complete, but considerable progress has made. A tentative phase diagram, based on thermal analysis studies and petrographic and x-ray dif- fraction examinations of quenched and slowly cooled samples, is presented in Fig. 2.1.2. The formulas of RbF-UF4,' 2RbF-3UF‘, RbF-3UF4, and RbF.6UF, were established on the basis of nearly single-phase material that was found in quenched samples of the corresponding compo- sitions which had been equilibrated just below the solidus temperatures. The recently obtained thermal analysis data are given in Table 2.1.1, and the results of quenching studies are presented in Table 2.1.2. een 600 TEMPERATURE (°C) PERIOD ENDING SEPTEMBER 10, 1956 fecRes ORNL—LR-~OWG 16146 1100 ] 1000 g 200 800 700 500 3RbF - UF, 2RbF- UF, 7RbF-BUF,4 ! . RbF - UF4 2RbF -3UF, RbF - 3UF, RbF- 6UF, 400 RbF . 10 20 TABLE 2.1.1. THERMAL ANALYSIS DATA FOR RE:I_'-'-_IV.IF4 MIXTURES . 30 . 40 50 60 UF, {mole 7o) Figs 2,1.2, The System RbF-UF‘l. 70 80 90 UF, CLos0 UF, . Ceontent L oo (mole %) - Casa s osg - 57 R | Liguidus '{ém_berhf“’e. - o ' ceetey 90 g . Solidus Tempérgturp- e e[ cess | ~UF, 60 .75 80 -»-;_1.~C6r_lfent : . /:':_‘r,‘(m'o'le' %y - 625 66.7 . . :85;8': :;-.; - , Liqui&us -+ . Temperature o) = 740 740 802 825 937 Solidus Temperature €O /1 745 - -743 o ANP PROJECT PROGRESS REPORT TABLE 2.1.2. RESULTS OF QUENCHING STUDIES OF RbF-UF, MIXTURES | Ct:an:nf TeLl::’:iil::fe Primary TZ':.:::;:, Phases™ Below T:;::m | #h“‘e"_ Below (mole %) ‘ (°c) Phase* (°C) Teonsition °C) Solidus 333 o <572 2RbF-UF, >745 2RbF-UF, | 46.2 712 RbF-UF, - | | 683 7RbF-6UF4,'R'bF‘-UF4_ | 50 727 ..RbF-UFa 727 RbF-UF . B 51 736 RbF-UF, 706 RbF-UF,, [u.m.] 53 rp3 RbF-UF, 709 RbF-UF 4, [usm.] 53 725 RbF-UF, 710 RbF-UF ,, 2RbF-3UF 53 721 RbF-UF 715 RbF-UF ,, 2RbF-3UF 57 737 RbF-6UF, 724 2RbF-3UF ,, Q.G. 714 2RbF-UF ,, RbF-UF - 60 2772 (RbF-6UF ) 60 >770 (RbF+6UF ) 731 RbF-3UF , Q.G. 720 2RbF3UF ,, [ vem.] 62.5 723 2RbF-3UF,, RbF-3UF, [Q.G.) 6647 >752 (RbF-6UF ) 733 RbF:3UF , Q.G. 709 RbF3UF , 2RbF-3UF4 6647 >757 (RbF6UF ) 725 RbF:3UF,, Q.G. 717 RbF-3UF ,, 2RbF-3UF 66:7 >758 (RbF-6UF ) 745 RbF:3UF ,, Q:G. 714 2RbF-3UF ;, RbF-3UF 75 >755 (RbF-6UF ) 714 RbF-3UF 75 >753 (RbF-6UF ) 721 RbF3UF 80 >760 (RbF+6UF ) 724 RbF-6UF ,, RbF-3UF 85.8 >824 RbF-6UF *The phase present in greatest quantity is given first, Phases found in very small or trace amounts are in brackets, Where a phase is enclosed in parentheses it is the phase found below the indicated temperature, but it is not neces~ sarily the primary phase. Quench growth is indicated by Q-G., vnidentified material by vem. THE SYSTEM NoF-RbF-UF, B. A. Soderberg V. Frechette$ H. A. Friedman H. Davis$ An intensive study of the phase-equilibrium characteristics of the system NaF-RbF-UF has been made. Equilibrium dota were derived from petrographic and x-ray diffraction examination of quenched ond slowly cooled melts. The boundary | curves, compatibility triangles, and the peritectic 5 and eutectic temperatures characteristic of the 90 system are shown in Figs. 2.1.3 and 2.1 4. It is noteworthy thot there is extensive solid so- lution along the join 7NaF. 6UF4-7RbF-6UF The extent of the miscibility gap is not yet de- termined. This solid solution effecr is general among the compounds 7MF-6UF ,, where MF is an alkali fluoride. There is one ternary compound in the system, with the formula NaFRbF-. UF ,. The compound has a subsolidus existence at its compo- sition below 530°C. The system contains the following five ternary eutectics: -~ UF,, - Composition (mole %) 7 ' Melting Point ) NaoF RbF UF, (°c) 25 25 50 630 26 27 47 .. 620 . 33 30 a7 535 47 33 20 | 470 wooB 9 e o o ORNL-LR-DWG 16147 - UFg | {wwe-SOLID SOLUTION) NaF - 2UF, e e i i o - —fi__‘ - - BNaF -3UF, 2NoF - UF,, i .t —— —— -"--.__ - - - P - - - - " - - ——— - - - - - Fig; 2.1.3. B'o'ur:llaary’ Curves and Compatibility Triangles in the System NaF-RbF-UF". PRI * ORANL-LR-DWS 16148 UF, =\ RbF - 6UF, and Eutectic Temperature in the System NaF-RbF- ‘Fig. 2.1.4. Primary Phase Fields and Peritectic PERIOD ENDING SEPTEMBER 10, 1956 The * boundary curve between the 2RbF-UF, and NaF-.RbF.UF, primary phase fields has its highest liquidus temperature at the intersection of the boundary curve and the extension of the 7RbF-6'UF“-?\lc:F-RbF-UF4 Alkemade line. (“‘Alkemade Line: In a ternary phase diagram a straight line connecting the composition points of two primary phases whose areas are adjacent and the intersection of which forms a boundary curve, "B) - The subsohdus compound NaF. UF has no primary phase field in the equilibrium diagram, At liquidus temperatures. which will permit an NaF.2UF primary phase field, RbF-UF , compounds . are the primary phases. THE SYSTEM NdF-RbF-ZrF4-UF4 H. A. Friedman B. A. Soderberg I H. Davis The system NaF-RbF-ZrF +UF, offers several mixtures which may be sohsfcctory fuels for a cuculohng-fuel reactor.? Quenching studies of this system were continued in order to establish the primary phase fields of the uranium-containing compounds. Primary phase determinations were also made on slowly cooled melts. These studies provide a basis for indicating the potential fuel mixtures from which the high-uranium-content compounds might segregate. The results shown in Figs. 2.1.5 and 2.1.6 are for the ternary compositions NaF-RbF-ZrF, re- calculated from the quaternary data. Where both the primary and secondary phases contain uranium, ‘the secondary phase is shown as the second ~ listing, THE SYSTEM RbF-CuF L. M Bratcher : Apphcotlon of fhermcl anolysns to mlxtures of R ff_"RbF and Cc:F2 has “consistently failed to yield "7 reproducible data for compositions “containing more than 15 mole % CaF .. . than ‘normal - cooling rates nor the subsmuhon of Covery thin thermocouple ‘shields served to improve . the _data ,obtomed. , Nelther the use of slower The fmlure of. fhermocouple 8E. Me Levin, He F- McMurd;e, and’ F. P. Hall Phase Diagrams ~ for Ceramists, American Ceramic Socaefy, Columbus, Ohio, 1956, ;9H- A. Friedman and H. Davis, ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 86. N ANP PROJECT PROGRESS REPORT 2rF, T ORNL-LR-DWS 18149 A= TNaF - 6Zr{U)F, 6= RbF-3UF, € = RbF-UF, D« SRYF- 42¢ (UIF, £ = RbF - 6UF, F=7RLF-6UF, - 75 LIQUIOUS TEMPERATURES 70 PLOTTED IN°C mp7 A.‘.b.‘ Fig. 2.1.5. Primary end Secondary Uranium- Containing Phases in the System NaF-RbF-Z¢F - UF, wnh 4 mole % UF,. shlelds after a few experiments suggests that the mixture is extremely corrosive; this may be due to the presence of some sulfate in the CaF .. ' Examinations of slowly cooled melts by x-ray diffraction and by the petrographic microscope reveal a compound with a cubic structure, which is presumably RbF.CaF, (ref 10). Since free CaF, has been observed in mixtures containing 40 to 45 mole % CaF,, it seems likely that RbF -CoF, melts incongruently. THE ALKALI FLUORIDE-CEROUS FLUORIDE SYSTEMS L. M. Bratcher Examinations of the systems LiF-CeF,, NaF- CeF,, and RbF-CeF , were continued; however, the data do not yet perrmt construction of satisfactory equilibrium diagrams for these systems. Com- parisons of data obtained with CeF; prepared at ORNL with dats obtained - with commercially prepared CeF , strongly suggest that the commercial material contums a small amount of water. It appears that treatment of the MF-CeF , mixture . with NH HF, will suffice to eliminate difficulty from this source. The NaF-CeF, system has been shown to have a binary eutectic, which melts at 775 £ 10°C, between CeF, and a binary compound believed to have the composition. NaF-CeF_s. 10y, L. W. Ludekens and A J. E. Welch, Acta Cryst. 5, 841 (1952). 92 A'w TNGF: 6Zr () F " 8 = RbF-3UF, ‘ Zrhy € = RbF - UFy | O =5ROF-4Zr{UIF, E = AbF-6UFy LIQUIDUS . TEMPERATURES PLOTTED IN °C Fig. 2.1.6. Primary and Secondary Uranium- Containing Phases in the System NaF-RbF-ZrF,. UF, with 7 mole % UF,. THE SYSTEM'CsF-Bch L. M. Bratcher Study of the system CsF-BeF, was started . recently in order to complete the study of the alkali fluoride—beryllium fluoride systems and to obtain additional data on the effect of varying ion size on phase relationships. Thermal analysis data have been obtained with mixtures at § mole % intervals in the range 5 to 50 mole % BeF,, but study of the slowly cooled melts is mcornpfefe. There is a eutectic, presumably between CsF and 3CstBeF2, which melts at 605 + 5°C and con- tains about 14 mole % Ber. Melting point re- lationships of the compound 3CsF:BeF ., have not been definitely established. The compound 2CsF-BeF , undoubtedly melts congruently, proba- bly at a femperoture befow that of the analogous compound 2RbF. BeF (800°C). As in the other “alkali fluorlde-berylhum fluoride systems, liquidus temperatures diminish rapidly in this system with increasing BeF, content in the 33',5 to 50 mole % BeF, region. The solidus temperature for mixtures containing 35 to 45 mole % BeF, appears to be about 450°C, which is very close to the repc;rfet:!I P mcongruent meltmg point of RbF.BeF,,. HL. M. Bratcher, R« E. Meadows, and Re Jo Sheil, ANP Quar. Prog. Rep. March 10, 1956, ORNL+2061, p 74. ORNL-LR-DWG 16180 RLF THE SYSTEM NaF-LiF-BeF, R. E. Meadows A dmgrom of the half of the system NaF-LiF- BeF2 represented by the traungle NaF-LiF- NaF.BeF_ was published recently.12 The quench- ing study of the LiF-NaF.BeF,-BeF, triangle is now sufficiently complete to permtt construchon of a tentative diagram for the entire system, as shown in Fig. 2.1.7. Binary eutectics, ternary eutectics, and -peritectic points are indicated in the diagram, along with the corresponding temperatures. Sub- solidus compounds are shown in parentheses. The temperatures following all the compound formulas are the liquidus temperatures for congruently 'melhng compounds or the upper sfob|||ty limits for other compounds. : Difficulties in the petrographic identification of the phases in the system, as mentioned previously, are also present in the work on the LiF-NaF.BeF .- BeF, portion of the system. Consequently it has been necessary, in some cases, to prepare and examine petrographically and by x-ray diffraction o great many samples in order to obtain a single result, _ Examination of slowly cooled samples, as well as quenched ternary samples, shows that a ternary compound containing more than 50 mole % BeF exists in this system. Its identity has not been completely established, but it is indicated on the 12n. E. Meadows, ANP Quar. Prog. Rep. June 10, 1956, ORNL'2]06J p 88, Fig' 2616 TS ORNL-LA-0WS 16451 £= EUTECTIC ’ S T= TERNARY EUTECTIC Befy TC= TEANARY COMPOUND ) oo Bez P= PERITECTIC : : SUBSOLIDUS COMPOUNDS SHOWN IN PARENTHESES . LIQUIDUS TEMPERATURES ARE IN °C LiF i 844 _ £649 Fig. 2.1.7.. The System NaF-LiF-BeF,, PERIOD ENDING SEPTEMBER 10, 1956 diagram by the symbol ““TC"" at 60 mole % BeF ,-20 mole % LiF-20 mole % NaF. The possibility that it may be an unreported binary compound has been eliminated by x-ray diffraction examination of equilibrated and quenched samples containing 66.7 mole % BeF,-33.3 mole % NaF and 66.7 mole % BeF2—33 3 mole % LiF. The join 2LiF.-BeF o"2NaF.LiF.2BeF, was established as a quasu-bmary system f:y the examination of equilibrated and quenched samples at 38 mole % BeF2—35 mole % LiF-27 mole % NaF. The liquidus is 334°C and the primary phase is 2LiF.BeF,. The phasesbelow the solidus (315°C) are 2LiF- BeF ‘and 2NaF.LiF. 2BeF2 An extrapolation places t 2he binary eutectic at 38.5 ‘mole % BeF 2~30.5 mole % LiF-31 mole % NaF. The join is not a true binary system because the primary phase field of LiF overhangs the compound 2LiF-BeF, very slightly.13 The triangle LiF- 2LiF.BeF -2NaF «L.iF:2BeF, is a compatibility triangle; no phases other than the three mentioned have been found in samples within the triangle. . Examinations of quenches of a composition con- taining 50 mole % BeF2-38 mole % LiF-12 mole % NaF, which is on the join 2LiF.BeF,-TC, show that the liquidus is above 354°C ond that the primary phase is 2LiF.BeF The solidus is 280°C, and the secondary ilase is the ternary compound TC. The absence of a third phase indicates that the composition is on or near the binary system 2LiF.BeF -TC, At the composition 50 mole % BeF ,~12 mofe % LiF~38 mole % NaF the liquidus is 344°C, and the primary phase is NaF.BeF,. Below the solidus (280°C) the three phases NaF-Ber,'ZLiF-Ber, and the ternary phase TC coexist. At the composition 50 mole % BeF ,~25 mole % LiF-25 mole % NaF the liquidus- was found to be 276°C, with 2LiF. BeF, as the primary phase.” The same three phases, NaF BeF2 2LiF.BeF,, and the ternary phase TC, were found to be present below the solidus (271°C). The results of these experiments show that _2LiF-_BeF2,' NaF.BeF,, and the ternary phase TC form a com- patibility triangle. However, the join 2LiF-BeF2- " NaF.BeF is not a true binary system, even if the ~slight overhang of 2LiF.BeF, by the primary phase field of LiF is disregarded. Quenching experiments with the composition 45 mole % BeF2-20,mole % LiF~35 mole % NaF, which is on the join, show that, down fo 300°C, 2LiF-BeF, ]3J- L+ Speirs, Thesis, Michigan State College, 1952, 93 ANP PROJECT PROGRESS REPORT and 2NaoF.LiF 2BeF, are present. Thus the primary phase fueld of 2NaF-.LiF 2Bef, cuts across the join, There must be a pentecflc point below 300°C in the triangle 2LiF. Ber-NaF Ber- TC, where 2LiF. BeF,, 2NaF.LiF. 2Be|:2 NaF.BeF, are in equilibrium, Locahons of the phase bounda- ries, the peritectic point, and the ternary eutectic were deduced from the evidence given above and _are considered to be tentative. THE SYSTEM No F-RbF-Ber L. M. Bratcher Study of the ternory system NaF-RbF- BeF, by thermal analysis supplemented by petrographic and x-ray examination of the slowly cooled melts was continued. Cooling curves obtained with compo- sitions within the triongle NaF-RbF-3RbF. BeF,, prevuously reported to comprise a compahblllty trlangle, indicated the existence of a ternary eutectic having a melting point of approximately 625°C. Additional data will be required in order to determine the eutectic composition, but the available data show that it will be close to the NaF-RbF eutectic (74 mole % RbF, melting point 675°C). An unidentified crystalline phase, believed to be a ternary compound, has been observed in a number of siowly cooled melts containing 45 to 60 mole % BeF,; this phase is usually associated with an isotropic phase believed to be glass. The cooling curves for the mixture NaF-RbF-BeF (35-15-50 mole %)} showed only one thermal effect ‘at about 240°C, and the cooled melt was found to be predominantly glass. The occurrence of glass in ternary mixtures having BeF, concentrations that produce only crystalline compounds in the corresponding binary alkali fluoride—beryllium fluoride systems was noted previously in the NaF-KF-BeF, system.'5 This interesting phe- nomenon may be due either to the difference in ion sizes of the alkali ions involved or to the lowering of solidus temperatures in the ternary system to the point where the high viscosity of the melt inhibits crystallization. Recent petrographic and x-ray diffraction studies of slowly cooled melts have confirmed the existence of the ternary compound 2NaF-RbF-_BeF2; it also appears that "L. M. Bratcher, ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 89 5., M. Brotcher et al, ANP Quar, Prog. Rep. Dec. 10, 1955, ORNL-2012, p 84. 94 the compounds 2RbF. BeF,, 2NaF-RbF BeF o and NaF.BeF, comprise a compahbnhty trmngle. The ~Nof.BeF -2NaF-RbF-BeF, side of this triangle passes through the minimum melting-point mixture (480 5°C) on the 2NaF BeF ,-2RbF.BeF, |°|no THE SYSTEM ‘KC'I-ZrCI‘ R. J. Sheil Interest in zirconium chloride—~alkali chloride systems stems from several sources. Low-melting- point chloride systems are potentially useful as heat transfer media. It is possible that fused- salt systems containing Z:Cl, will be of interest in the ZrCl, volatility method proposed for processing o? metallic zirconium-uranium fuel elements.'® [n addition it is of interest to com- pare phase equilibrium behavior in these MCl-ZrCl systems with behavior in the analogous MF-ZrF systems, which have been thoroughly siudied at ORNL. The alkali chloride—zirconium chloride systems have received relatively little attention. A Russian investigation of the NaCl-ZrCl, system is on record.'? Preliminary data obtained at ORNL indicate serious errors in the published data; the study being made here is not, however, sufficiently complete to permit a diagram to be drawn. Some data on the KCI-ZrCl, and NaCl- ZCl, systems in the ZrC|4-r|ch parts of the systems were obtained by investigators at Columbia Uni- versity.18 The present investigation of the KCl- ZrCl, system covers the composition range 10 to 75 mole % ZrCl4. The data obtained are in reasonably good agreement with the Columbia data in the limited area where the two investigations overlap. The thermal analysis data obtained from cooling curves with mixtures contained in sealed glass, quartz, or nickel tubes fitted with thermocouple wells ore shown in Fig. 2.1.8. In addition to the use of the thermal analysis tech- nique, visual observation and quenching techniques were applied to the study of several mixtures in the V6Chem. Tech., Semiann, Prog, Rep. March 31, 1956, 17N, A, Bolozersky and O. A, Kucherenko, Zbur, Prikad. Khim, 13, 1551 (1940). 184, H. Kellogg, L. J. Howell, and R, C. Sommer, NaCl-ZrC|4, KCl-CrC|4, and NaCI-KCl-ZrCl4. Summary Report, NYO-3108 (April 7, 1955). " e et et — e o PERIOD ENDING SEPTEMBER 10, 1956 UNCLASSIFIED CRNL-LR-DWG 16152 800 - r . 600 - - 3 —-——‘ 5 < \ s 5 500 w = o S | E el < i 3 ——— & 400 gL \ = = = /" Lt = - @ W I Q T S ? | / 300 = ~ g © o — O G | J ~ - i 200 -.”— '-—.-.- v o' —V — - ——-.——-}——-—- _____ 100 ' KCI 10 20 30 40 50 60 70 80 20 2rCl, ZrGly (mole %) Fig. 2.1.8. Tentative Diagram of the System KCI-Z+Cl . system, and most of the slowly cooled melts used for thermal analysis were examined petrographi- cally. As shown by Fig. 2.1.8, there is a eutectic between KC! and the compound 2KCl.Z:Cl & which melts congruently ot 790 + 5°C. The eutechc melts at 600 1 5°C and contains about 23 mole % ZrCl,. There is also another compound, tentatively ass:gned the formula 7KCl.6ZrCl 4+ Which appears to melt incongruently to give 2KCI.ZrCl . and liquid at 565 + 5°C. The choice of the compound formula was based partly upon thermal analysis dota, which indicate thot the compound contains less than 50 mole % ZrCl 4+ and partly upon the “results of petrographic excmmahon of slowly cooled and quenched samples in this region. More careful study, particularly of quenched samples, will be required to eliminate uncertainty in regard to the compound composition. Thermal effects on cooling curves at temperatures ranging from about 218 to 225°C were noted with compo- sitions containing more than 33'6 mole % ZrCl,. Since this is reasonably close to the reportedi8 “melting point of the eutectic at 65 mole % ZrCl (234.5°C), the appearance of the thermal effect ot a temperature close to this value with compo- sitions containing less than 50 mole % ZrCl ‘might be taken as an indication of incomplete reaction of liquid with solid 2KC|-ZrC|4 at the peritectic point. However, the magnitude of the thermal effect on cooling curves, coupled with the fact that no liquid was visible in mixtures containing less than 50 mole % ZrCl, ot temper- atures below 250°C, indicates the likelihood that a solid phase transition occurs at a temperature so close to the solidus temperature that the two effects cannot be differentiated either by thermal analysis or differential thermal anclysis. Petro- graphic examination of a quenched sample con- taining 55 mole % ZrCl, showed that, above 237°C, well-crystallized phoses were present but that at 219°C and lower temperatures the samples con- tained only microcrystalline aggregates. 95 ANP PROJECT PROGRESS REPORT 2.2, CHEMICAL REACTIONS IN MOLTEN SALTS F. F. Blankenship R. F. Newton EQUILIBRIUM REDUCTION OF NIF, BY H, IN NaF-ZrF, C. M. Blood Investigation of the equilibrium ~ NiF,(d) + H,(g) == Ni(s) + 2HF(g) with NaF-ZrF, (53-47 mole %) as the reaction medium was continued. Since the system was observed to behave rather peculiarly at high temperatures, this study was interrupted while reloted investigations of the stability of NiF, L. G. Overholser G. M. Watson in this solvent were conducted. These experiments have served to indicate the essential validity -of the equilibration data previously obtained! at 625°C and of the data obtained at 500 and 575°C, as reported below. The experimental data obtained by using sintered-copper filters and the general technique described previously are shown in Table 2.2.1. The mean of 18 values previously . obtained for the equilibrium constant at 625°C was 3.0 x 10=4 atmosphere. - ]C. M. Blood, ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 100. TABLE 2.2.1. EQUILIBRIUM RATIOS FOR THE REACTION NiF,(d) + H,(g) = Ni(s) + 2HF(g) Ni in Melt Pressure of H,* Pressure of HF K ** % 10-4 (ppm) (atm) {atm) x Measurements Made at 575°C 315 0.0266 0.434 1432 345 0.0257 0.453 1.25 365 : 0.0248 0.472 1.45 385 0.0245 0.477 1.42 420 0.0243 0.482 1434 435 040241 0486 1.32 180 0.0301 0.358 1.38 165 0.0297 0.368 1.62 185 0.0293 04375 1.52 220 0.0290 ' 0.382 1.35 150 0.0309 0,342 ‘ 1.49 Av 1.40 0.09 Measurements Made at 550°C . 175 0.0318 205 0,0309 205 ~ 0.0306 200 | 040306 0.321 1.09 0.341 1.08 0.347 112 0.347 1.15 *Initial H2-H_e mixture was 4.69% H2. "'*Kx = PE!F/XNIszfiz' where x is mole fraction and P is pressure in atmospheres, 96 + o o -"obtamed for NlF From the data obtained to date it appears that the activity coefficient for NiF, must be about 2750 at 600°C, with NiF ,(s) taken as the standard in this - solvent was 3.28 (ref 2). The very large apparent is considered to be state. The correspondmg figure for Fer activity coefficient for NiF, . quite surprising. ' STABlLlTY OF NiF J G. Eversole lN NGF-ZrF MIXTURES C M Bload The very hlgh apparenf achvnty caeff:cnents in molten NaF-ZrF, prompted | " .a careful exammahon of the behavior of such " solutions under a variety of conditions. " - this ‘examination -has served to substantiate the’ While validity of the equilibration data obtained, a number - of rather peculiar effects have been observed. ‘ - An initial mveshgahon was performed to dls-. close whether any yncerfainties in the chemical - X analyses were obscuring the real picture. A total of ten standard samples prepared by adding , weighed quantities of NiF, with or without added 'CuF to powdered NaF-Zer mixture were prepared ‘and submmed for analys:s. Analytical results for these samples’ covered the range from 125 to 2000 ppm nickel and were, with a single exception, . within 5% of the theoretical value; the single exception was the most dilute sample, for which ‘the analytical result differed from the *‘true” - value by 22%. In addition the reproducibility of the chemical analyses was tested by submitting, in duplicate, finely ground samples of filtrates from solutions of NiF,in NaF-ZrF ., The duplicate samples were submmed under glfferent sample " numbers, with one to two weeks being allowed “between the time - the. sample and its duplicate ~were submitted. The data presented belaw show i excel!ent agreement in every case' ' O Nlckel Found (ppm) Inltlal Sumple s Dupllcafe'Sampl_é s Cess s 1040 1040 oems. o s 005 1015 1030 - -7 1025 40 40 ZC. Me Blood and G. M. Wotson, ANP Quar. Prog. Rep. March 10, 1956, 0RNL-206'|, P 89. 50 hr of sparging with helium at 625°C. ~cant quantity of NiF, PERIOD ENDING SEPTEMBER 10, 1956 40 35 40 35 260 260 1165 . ms - 1070 1050 1030 1010 1030 1030 It seems clear that uncertainties in analyses for -NiF, in the samples submitted is not a contributor . to the surprising observations. It has been observed thot NiF, dissolved in NaF-ZrF, mixtures is stable in containers of A-nicke! under continuous sparging with helium - up to temperatures of 625°C. At temperatures of 700°C ond above, .however, the concentration - of dissolved nickel decreases continvously. The . results of a typlcal experiment are presented in Table 2.2.2. concentration . It may be observed .that the N|F2 remained - constant during about During equilibration experiments it is necessary to sparge the system with helium for about 30 min when ~ samples are taken. In view of this stability under long sparging, however, it appears that no signifi- is lost during the short treatment required in equilibration experiments. The disappearance of NiF, from solution at the ‘higher temperatures is not well understood. The presence of significant concentrations of reducing impurities, such as hydrogen, methane, or oil vapor on the helium, can apparently be eliminated from consideration. The influent helium stream was passed over a bed of copper oxide at 500°C, - a magnesium perchlorate drying tower, and a liquid ‘nitrogen cold trap without affecting the experi- ~ mental results. - Moreover, the effluent helium was ‘carefully ‘monitored by . passage through KOH solution. The amount of HF recovered was only a few per cent of that recoverable fram the NiF lost from solution. ‘And fmally, even though the . NiF, is apparently lost from the soluhon, passage of - fiydrogen gas through the system hberatas 'essenhally the stoichiometric quantity of HF, _ Thermodynamlc_ considerations indicate that the : ethbrwm pressure for decomposition -of NiF, ‘into Ni® and F,is neglxglbly small, It is possible, however, that a volatile subfluoride of nickel may be responsible for this peculiar behavior. [f the reaction NiF ,(d) + Ni{s)== 2NiF(g) 97 ANP PROJECT PROGRESS REPORT TABLE 2.2.2. STABILITY OF NiF, IN NaF-ZrF ; (53-47 mole %)* UNDER CONTINUOUS SPARGING WITH HELIUM ‘ - Filter . - - ' Found in Filtrate (ppm) Temperature Time at Temperature _(°C) ' _ (hr) Material - - Ni Fe " Cr 625 . 24 Ni IE 225 ' 25 48-50 NP 1050 255 25 Ni 1030 240 25 ‘cw - 1000 225 - 30 Ni 1025 250 30 Cu. 015 245 .25 Ni 1040 - 250 20 NN . . 35 145 - - 20 80 . 65 *Initial charge: 1 kg of NaF-ZrF , containing 260 ppm Ni and 220 ppm Fe. Additive: 1000 ppm Ni as NiF,, can occur, it may be possible to explain the phenomenon observed. No concrete evidence for existence of this compound has been found.. A large number of equilibration measurements have been ‘performed by using copper filtering tubes. During the course of the sampling procedure the copper filter tube becomes plated with nickel. It has, accordingly, been necessary to determine whether the concentration of nickel in solution is changed by the introduction of the copper tube. The results listed in Table 2.2.2, for samples taken in the interval of 48 to 50 hr indicate that nickel plating of the copper tubes does not change the nickel concentration in solution. In addition, a copper tube was immersed in the solution for 24 hr lmmedlately following the 48- to 50-hr period, and a massive nickel deposit, which contained large dendritic nickel crystals, was obtained (Fig. '2.2.1). The weight of the deposit was determined to be approximately 15 g; the weight of nickel in solution just before immersion of the copper tube was only about 0.4 g. The nickel concentration of the solution immediately after the heavy nickel plating occurred was determined to be 960 ppm, as compared with a value of 1030 ppm obtained by averaging the results for all the 'samples taken during the 48- to 50-hr period. The formation of a nickel plate on the copper | tubing could be explained by simple concentration cell action; however, such a mechanism would 98 cease to function after a very thin film of nickel . was laid down unless a copper-nickel alloy were formed. Careful examination of the massive deposit- shown in Fig. 2.2.1 revealed essentially no copper.on the outer surface. force for this plating process were the temperature difference between the nickel reactor wall and copper tube, which was cooled by influent helium gas, similar deposits of nickel should have ap- peared on the nickel sampling tubes, but no such deposits have ever been observed. In all these experiments the copper tubing was inserted through a stainless steel Swagelok fitting at the top of the reactor, where the temperature was about 200°C. Accordingly a low-temperature copper- steel-nickel junction was in electrical contact in series with a high-temperature copper-nickel junction through the molten salt. It appears likely that the resulting thermoelectric potential may have been the driving force for the plating process. Although experiments are planned to supply explanations for some of these phenomena, it can be concluded that the use of copper filters introduces no significant error into the results presented. In an effort to obtain concrete evidence for the formation of ‘a volatile nickel subfluoride, two platinum crucibles were loaded with known weights of nickel fluoride and three nickel crucibles were charged with mixtuwes of metallic nickel powder if the driving - - L A Al Sy PERIOD ENDING SEPTEMBER 10, 1956 SXT UNCLASSIFIED R PHOTC 26802 Fig.2.2.1. Massiie Nickel DepoSit with Lafge Dendritic Crystals on Copper Tube That Was Immersed for 24 _hr in an NaF-Z¢F, (53-47 mole %) Solution Containing NiF,. -und mckel fluornde. The five crucibles were. enclosed ‘in a nickel ‘vessel fitted with ‘a copper spiral coil to serve as a condenser. Cool helium - gas was passed "through ‘the copper -coil cone ~f|nuous|y at high rates. A flowing" atmosphere -of “helium gas was also: conhnuousiy passed over the samples _as ‘the ‘temperature of the “system was - brought to 800°C and held constant for a period of 3 hr. - The samples were allowed to cool and were reweighed. The material condensed on the copper - coil was curefu“y collected and " submitted for _petrogrophlc -and . xeray : d:ffrachon ‘examination, After having been weighed, the Ioaded crucibles - were heated again in the same manner for. an additional 19 hr. The results of the experlment are summarized in Table 2.2.3. At the end of 3 hr the N:F n- the platinum crucibles was discolored - Ond a considerable amount of magnetic material, later identified as Ni% was present. . No apprecioble change in welght of the platinum occurred. Slight deposits - on the copper. condenser tube were tentatively | identified as NiF,, Ni, and CuF2 - The data may be mterpreted as. p0551bly sup- ‘ porhng the reversible reaction: NiF (s) + N;(s)-—->2N|F(g) ‘ The magnitude of the effect noted ‘was, however, so small that there is no conclusive proof of such a mechanism. No explanation can be offered for the partial reduction of NiF, in the platinum crucibles, 99 » ANP PROJECT PROGRESS REPORT TABLE 2.2.3, WEIGHT LOSSES OF NiF ,(s) ON HEATING TO 800°C " Crucible Weight Loss of Initial Nfl:2 (%) Material ‘After 3 he After 22 hr Total Platinum 1.78 0.18 1.96 Platinum 234 0.38 7 272 Nickel 3.58 0,93 4,51 Nickel 3.06 0,78 3.84 Nickel 3411 0.86 3.97 All these phenomena will, presumably, be important in mechanisms of corrosion of nickel by molten salts. It is anticipated that considerable additional study of these problems will be under- taken. ACTIVITY OF CHROMIUM ‘IN CHROMIUM-NICKEL ALLOYS M. B. Panish The determlnatnon of the activity of chromium in chromium-nickel alloys has been completed with the use of the electrode concentration cells and the techniques previously described.34 The values obtained for the activity of chromium in various chromium-nickel alloys at 750 and 965°C during this investigation are shown in Fig., 2.2.2. The curve obtained by Grube and Flad> at 1100°C is shown for comparison. From these data it appears that at concentrations below about 25 at. % the activity of chromium is somewhat lower than its atomic fraction in the alloy. The curve obtained at 750°C in this investigation is consistent with the expected behavior of a system which has an immiscibility gap. However, both the Grube and Flad curve and the data ob- tained at 965°C in this investigation seem to yield implausibly fow chromium activities for the chromium-rich solid sclutions. M. B. Panish, ANP Quar, Prog. Rep. March 10, 1956, ORNL.-2061, p 92. ‘M- B. Panish, ANP Quar, Prog. Rep. June 10, 1956, ORNL-2106, p 93, 5G, Grube and M. Flad, Z. Elektrochem. 48, 377 (1942). 100 UNGLASSIFIED . . ORNL=LR—-DWG 16153 100 o TWO-PHASE REGION AT 750°¢" | / . 7 lo 075 TWO-PHASE REGION — p—t—ar 965°C e ¥ i- ‘L/ - > C ': 7. = 050 ‘ / - P [ & /'_ ; < / TWO-PHASE REGION ) |=t— aT t100°¢ ——| 0.25 S o 750°C / A965°C i /7 e 1100°C (GRUBE AND FLAD) o o Ni 25 50 75 - Cr CHROMIUM (ot. %) Fig. 2.2.2, Activity of Chromiufi in :Nilckel- Chromium Alloys, SOLUBILITY AND STABILITY OF STRUCTURAL , METAL FLUORIDES IN VARIOUS FLUORIDE MIXTURES J. D. Redman Data gathered from studies on the stability and solubility of CrF,, FeF,, and NiF, in NoF-ZrF (53-47 mole %) at 600 and 800°C were reported previously.® The results obtained at 600°C demonstrated that the solubility of the structural metal fluoride was a function of the excess of structural metal fluoride present. It is believed that the change in solubility is caused by the presence of a solid phase, probably MF2-ZrF4, and the resulting change of the NaF concentration in the melt. The studies have been extended to include the solvents LlF-ZrF (52-48 mole %), NaF-ZrF (59-41 mole %), ond KF-ZrF (52-48 mole %), anci additional studies have been mcde on the solubility of C!'l"'3 in the NaF-LiF* DIFFUSION CURRENT‘/ - g 50 —— ‘ o ’ . 5 ) . 8 40 W a 30 ' / 0.19 wt % ) / | / 20 / /I 4/' ‘ 10 /' i+* oiFFusion 7 CURRENT -4 ' . N 0 o — 0 10 20 30 40 50 60 70 FeF, ADDED (mg) Fig. 2.2.10. Diffusion Currents Obhinéd from Adding FeF3 to 26,6 g of NaF- KF-LIF (11.5-42- 4605 mole %) at SSOOCO ACTIVITIES FROM EMF MEASUREMENTS - ON SOLID SOLUTIONS OF SALTS M. B. Panish Measurements of activities in the system AgCl- NaCl by an emf method have been of interest both as a trial study with a solid elecfrolyte of o technique which may be applicable ‘to fluoride systems and also as o demonstration of some metastable solid solution behavior. The cell being investigated was comprised of ‘Agf% and CI'J electrodes, with AgCi-NaCl as the elecfrolyte, so that the cell reaction was the formation of AgCl from the elements. The activity of AgCl was determined from the emf by comparison with pure AgCl. The eiectrolyte concentration was varied from 3 mole % to pure AgCl, and the temper- ature range was extended from the liquid into the L I. solid region, that is, to temperatures as low as 250°C. - The feasibility of makmg equulsbrlum mecsure--_' ments with a solid electrolyte, as reported by . earlier workers,’? 'was demonstrated. This is probably easier’ with the Ag* ion.in a chloride system, but, with a_ svitable amplifier, measure- ments on solid fluorides might also be feasible. If so;, a ceramic container’ would no longer be.a problem for cells containing fluorides, measure- ments could be made on solids by using cells - designed to cvord ‘transference potentials, and - compositions - with™ mconvenlently hlgh meltmg pomts could be studied.” “In the liquid state, the AgCI-NcCI system ‘showed a slight posifive deviafion from’ ideality. From the phase dragrum, a continuous series of solid so- Jutions was expected below the solidus. However, the activity of AgCl in the solid so!uhon was not a monotonic function of concentration, The AgCl- rlch soluhons showed marked posmve devrahon .lqu Wachter, J. Am. -Cbem. Soc. 54,'919 (1932). PERIOD ENDING SEPTEMBER 10, 1956 ‘from Raoult's law and gave an activity vs concen- -tration curve which rose well above, and pre- sumably did not connect with, the curve for NaCl- rich' compositions. The NaClerich compositions showed - a.slight pegative deviation, 'if any, from Raoult’s law. There was apparently a region in the middle where the two curves overlapped without _ ‘cbrmecfing, however, the ‘experimental uncertainty _in this region was large. Obviously the solution wrfh the higher AgCl activity was unstable with respect to the lower one, but no transition from ‘the higher curve to the lower was noted. Also an achwty-concenfraflon relcmon of thls type requires ‘that * the 'stable condition is separation into ‘c.oniugate solutions; no evidence of such phase separation was found by petrographic or x-ray fe_xaminotions.' The \m‘etostable_' condition was - praduced by quenching from the liquid and was maintained, in spite ‘of annealing, because of the slowness of diffusion in the solid. These findings “agree qualitatively with those found for the same solid solution by Wachter,1? who worked at a lower temperature and used different electrodes. 117 ANP PROJECT PROGRESS REPORT 2.3, PHYSICAL PROPERTIES OF MOLTEN MATERIALS F. F. Blankensh i-p SURFACE TENSION AND DENSITY OF NaF-ZrF, MIXTURES. F. W. Miles Attempts to apply the maximum bubble-pressure method! to determinations of surface tensions of NaF-ZrF mixtures have been accompanied by dlfficulfles caused by partial plugging of the capillary tips. Therefore the values that have been obtained are not considered to be sufficiently accurate to report. Additional equipment is being added to the gas-purification train, and the bubble- pressure apparutus is being simplified in an effort fo minimize contamination of the system and consequent pluggmg of the tips. The results of several densuty determmutlons on NaF-ZrF (53-47 mole %) are now available. The density data (in g/ml) are summarized below: At 600°C At700°Cc At 800°C 3.049 2.916 2.805 3.014 2,948 2.813 3.051 2.895 2.821 - 3.039 2.946 2.833 3.003 3.020 o T Av 3,03 % 0.02 2.93 T 0.02 2.82 t 0.01 The average values may be represented by the equation d = 3651 - (1.4 x 10~3) T , where T is temperature in °C. The data were obtained on several different batches of NaF-ZrF mixture of the same compo- sition. |t is believed, however, that the density values are probably correct to within 2%. The calculated densities? corresponding to this compo- sition are higher by approximately 5% than the VE. W. Miles, ANP Quar. Prog. Rep. March 10, 1956, ORNL-2061, p 106. 25, |, Cohen and T. N, Jones, A Summary of Density Measurements on Molten Fluoride Mixtures and a Corre- lation Useful for Predicting Densities of Fluoride Mgu:tures of Known Compositions, ORNL-1702 (May 14, 1954). 118 G. M. Watson average values listed, Efforts are being made to improve the precision of the density ond surfuce tension measurements. : SURFACE TENSIONS OF MOLTEN SALTS S. Langer Flucrozirconate Mixtures Additional measurements of the surface tensions of NaF-ZrF, (53-47 mole %) mixtures have been made by USmg the sessile-drop technique, pre- viously described.3+4 Recent experiments at 600°C tend to substantiate the results reported previously. Some data have been obtained.at 700°C, and o preliminary experiment at 800°C has been run. The data accumulated to date are summarized in Table 2.3.1. Data from the previous report4 are included and corrected by the use of more precise densities calculated from the equation d = 3.651 - (1.04 x 10-3) T(°C) , reported above. The “‘best’’ values for the surface tension of these mixtures are 127 * 6 dynes/cm at 600°C and 119 t 4 dynes/cm at 700°C, where the error range is the root-mean-square deviation. It should be noted that there appears to be a small difference (within experimental error) between the data obtained in helium and in hydrogen atmospheres at 600°C. The average surface tension in helium is 128 dynes/cm, while the average in hydrogen is 122 dynes/cm. More experimental work will be needed to demonstrate the effect of the atmosphere. The first evidence of a large change of surface tension with composition is shown by the data at 800°C. The large loss of weight due to vapori- zation of ZrF‘ from the sessile drop and the concomitant change of composition probably ex- plain the increase in the surface tension of the melt. Experiments designed to study the effects of change of composition as a function of temper- ature, pressure of the confining gas, and time of equilibration are under way. The results will 3S. Langer, ANP Quar., Prog. Rep. March 10, 1956, ORNL-2061, p 105. 45, Langer, ANP Quar. Prog. Rep. June 10, 1956, ORNL<2106, p 118- v U D et i My e St Pt L S ek L] AW o b PERIOD ENDING SEPTEMBER 10, 1956 TABLE 2.3.1. SURFACE TENSIONS OF NuF-ZrF (53-47 mole %) MIXTURES CBTAINED . BY THE SESSILE-DROP TECHNIQUE Temperature Sompie Weight Pressure Weight Loss b Compositidn Nurfiber of -Surface Tension (°C) )] ~ {mm Hg) {%) - (mole % Z?F4)"' .Photographs Taken (dynes/em) 600 04502 240 3 46,0 6 132 £ 6 0307 250 3.7 458 6 125 £ 4.3 05978 . 410 ) 2 126 £ 4.6 03154 . 350 43 45.6 6 122 1 2,6° 700 07758 260 173 | 40.8 5 118 1 0.8 04632 . 51s 129 400 4 16145 - 04487 515 9.0 . 414 4 122t 1.4 S sIs 0 1S 4l 3 123 1 0.6 800 - 03513 530 - 354 29.8 s 138 1 7 ) aCalculated from the weight loss of the scmple by assummg thot only ZrF4 is volatilized. _ ‘ation, 'bSumples meusured on. graphite p!oques under a hellum otmosphere. The error range is the root-mean-square devi- cThls #cmpie was meusured under hydrogen ruther tl'mn helium. permlt the estlmahon of -the surface fensions over ‘@ range of composmons from | smgle sample. : o " In the prevuous reports the penetrahon of APC. - graphite by NaF-ZrF (53-47 mole %) mixtures was described. This behavtor is surprising in view of - the nonwemng properties -of this mixture on C-18. . ‘;grqphlte, as evidenced by the surface-tensson : meosurements reported here, Several. prellmmary ‘ "expenments in which APC grcphlte was used as the -supporting plaque have therefore been carried - out,. While the’ experimental conditions are some- . what - different : from those of the penetruhon'i - . “studies, the sessile drops do not appear to have " either wet ‘or penetrated the plaque, - The reason - “for the discrepancy between these data and the _prev:ous observahons is not mmedmtely upparent. ' SH. J. Botram and G, F. Schenck, ANP Quar. Prog. ' Rep. ]une 10, 1956. 0RNL~2106, P 'I25o L ' Uraniurh Téfraflucride Several preliminary experiments have been . carried out to study the wetting properties and ‘surface tensions of UF, and of UF, with small . additions of vo,. - viously® indicated that melts containing 4%, or - _more, UO would penetrate graphite. This obser- * vation wus borne out by preliminary observations _of the surface tension. No values for the surface ~ tensions are available as yet, but the preliminary 'expen_ments have shown that, even though UF “with about 1% UO, did not immediately wet C-18 Visual observation had pre- or APC. graphlte, after o period of time, during - .which small amounts of air leaked into the system, " the contact gngle receded and the UF 4-UO, melt started. t6 wet the plaques. These studies will - be contlnued 6R, J. Sheil, ANP Quar. Prog. Rep. June 10, 1956, 'ORNL-ztoa. P 91. - 119 ANP PROJECT PROGRESS REPORT 2.4, PRODUCTION OF PURiFIEDM FLUORIDE MIXTURES G. J. Nessle LABORATORY-SCALE PURIFICATION OPERATIONS W. T. Ward The standard hydroflucrination-hydrogenation process was used, with appropriate meodifications to laboratory-scale equipment, to prepare a number of especially pure flucride mixtures requested for various' reseurch programs, Several of these compositions were dispensed directly into con- tainers furnished by the requester to minimize atmospheric-contamination and handling. QUALITY CONTROL OF RAW MATERIALS AND PRODUCTS W. T. Ward . The processing characteristics of an NaF-ZrF mixture supplied by a commercial vendor were determined, and the product was found to meet specifications. Samples of all production batches were obtained for petrographic and x-ray diffraction analysis to assure quality control. PILOT-SCALE PURIFICATION OPERATIONS C. R. Croft J. Truitt J. P. Blakely The pllot-scoie purification facility processed 53 batches totaling 835 Ib of various fluoride compositions for use in smallsscale corrosion testing, phase equilibrivm studies, and physical property studies. The demand for special compo- sitions continved at a fairly constant level, and thus efficient operation and maintenance of these facilities at about 30% capacity was possuble without accumulatmg a backloeg. The use of copper-lined stainless steel reactor vessels in these small size units has proved successful in all but one case. Mixtures con- taining NaF, LiF, KF, and UF cannot be frozen in the copper liners and then remelted In three attempts to process NaF-KF-LiF- UF (11.2-41- 45.32.5 mole %) the copper liners ruptured when the melt was frozen and then remelted in the reactor vessel. This means that any full-scale operation of the 250-1b facility for the preparation of such mixtures would necessitate a 24-hr-day 7-day-week basis of operation so that the mixture could be maintained molten at all times. 120 G. M. Watson PRODUCTION-SCALE OPERATIONS J. E. Eorgan J. P. Blakely The production-scale facility processed 19 batches totaling about 4550 1b of purified fluoride melts. The use of copper-lined stainless steel reactor vessels continued to prove very satis- factory in this facility. One reactor vessel has now produced sixteen 250-Ib batches, and it. shows " no signs of deterioration. Product quality has remained excellent throughout the equipment life; however, some difficulties have been encountered because of the high oxygen and water content of the poor-grade hydrogen now being supplied in this area. This difficulty has apparently been overcome by installation of catalyst trains and - cold traps of higher capacity than previously used on the hydrogen purification system, The production facility was shut down during July because the relatively low consumption rate resulted in a shortage of receivers for storage of purified product. The present disposition of sixty- four 250-Ib containers is the following: - Held by Pratt & Whitney 23 | To be shipped to Pratt & Whitney (August) - 8 ART High-Temperature Critical Expenmenf _' 3 storage } Stock inventory (full) ' 20 Ernpt} containers on hand _ 10 Total . : &4 Forty new 250-1b storage containers are presently on order. Delivery of these containers is expected to begin in September, and reactivation of the production facility is scheduled accordingly. It is extremely unlikely, however, that the shortage of such containers will be alleviated before January. Accordingly production of purified ma- terials will probably be 8,000 to 10,000 Ib behind presently scheduled demands by the end of this calendar year. An order for 30,000 Ib of NeZrF, has been placed with a commercial vendor. It is estimated that this amount should be sufficient to maintain operation for FY 1957. Delivery of 4000 lb of this material per month is expected to begin in . w ne August 1956. Each shipment will be analyzed thoroughly to ascertain that specifications of purity have been met. : ' BATCHING AND DISPENSING OPERATIONS F. A. Doss - J. P. Blakely The batching and dispensing facility dispensed 128 batches totaling approximately 8025 Ib of processed fluorides in batch sizes ranging from -1 to 250 Ib.. This is a sizeable increase over the amount normally handled because of large ship- ~ments to Pratt & Whitney. As a result, the stock inventory of processed fluorides decreased ap- preciably. A material balonce for the quarter is given in Table 2.4,1. The main consumers of the processed fluoride mixtures and their allotments during the quarter are given below: - - ORNL-ANP groups For chemical and physical properties - 258 1b - studies : For experimental engineering tests _ 2680 For metallurgical studies and fuel 593 reprocessing development Pratt & Whitney Aircraft 3207 Other cofiirccfors, including BMI, NRL, 200 and WADC Salvage and reprocessing 1087 Total _ 8025 b PERIOD ENDING SEPTEMBER 10, 1956 PREPARATION OF ZrF‘ FROM Zl'(:l4 J. E. Eorgan J. P. Blakely A dust filter for the ZrCl -conversion unit is presently being installed. tonversion of some 2000 Ib of low-hafnium ZrCI4 to ZrF, will begin as soon as installation of the filter is complete. It is hoped that the new dust filter will prevent large-scale loss of the product during the con- version and permit efficient operation. The con- version unit will be operated on a 24-hour-day five-day-week basis until the present supply of ~ZrCl, is all converted to ZrF,. The recent demand for 1200 Ib of low-hafium NaF-ZrF, (50-50 mole %) for the Pratt & Whitney high- temperature critical test has made it imperative that this unit be operated as soon as possible. In order to meet known demands in FY 1957 for fluoride compositions requiring low-hafnium ZrF , an order for 7000 Ib of low-hafnium ZrCI4 has recently been initiated. At least half this order has been requested for delivery before December 1956, and the remainder is to be delivered some- time in the last half of FY 1957. SPECIAL SERVICES F. A. Doss J. P. Blakely J. Truitt N. V. Smith Filling, Draining, and Sampling Operations The filling, draining, and sampling operations were at a normal level during the quarter. Ap- proximately 2700 |b of processed fluorides and 1000 b of liquid metals were used to charge TABLE 2.4.1. MATERIAL BALANCE FOR FLUORIDE MIXTURE PRODUCTION AND USE Material (Ib) - . " - Mixture No. 30, Mixture No. 31, Mixture No. 108, . Mixture No. 71, Special " NoF-ZF (UF, - NaFZiF, NaF-ZrF,UF, - NaF-ZrF pPecial ol (50464 mole %) (50-50 mole'%) (56-37.56.5 mole %) (54.145.9 male %) Onhand at beginning of 595 2,073 | 1,302 9,325 quarter "~ T A | | Produced dur-ing:iqq‘m_e'r' | 655 0 2673 1232° 835 5395 Tewl - eg0s . 2,07 e wm o 20w 4o Dispensed duringquarter - 4,161~ 833 1679 242 - 1,10 8,025 Onhand of end of quarter 2,444 1,240 994 990 1,027 6,69 12] ANP PROJECT PROGRESS REPORT engineering tests in charge sizes ranging from 1 to 500 Ib. Electric power has been installed at the liquid metals disposal quarry to facilitate the disposal of bulk quantities of NaK and Na. Special Enriched Fuel for In-Pile Loops Inpile loop No. 5 failed to operate at the MTR because of inability to move the fuel from the sump tank to the pump (see Chap. 1.6, ‘““‘In-Pile Loop Development and Tests"). Later investigations showed that the fill line between the pump and the sump tank was plugged. This plug was found to be nearly 50% ZrO s and subsequent analysis of the original batch showed evidence of oxides and oxyfluorides. Examination of two other unused batches showed oxides and oxyfluorides to be present in undesirable quantities. As a result - of these findings, a new batch of the fuel mixture (No. 44) NoF-ZtF, -UF, (53.5-40-6.5 mole %) was processed with extreme care for use in in-pile loop No. 6. The batren carrier was prepared and processed first to determine whether oxides of zirconium were persisting in the batch. Since the carrier was found to be oxide-free, the U235. enriched UF, was added to the batch, and the processing was repeated. When the batch was shown fo be oxide-free, a fuel charge was trans- ferred from the batch into a loading can. A sample was token during this operation to reverify the oxide purity of the charge. The charge was then loaded into the loop, with a sample being taken as the charge moved from the loading can into the loop. This sample was also verified as being oxide-free. Therefore, it can be assumed that the charge that went into the sump tank of in-pile loop No. 6 was in satisfactory condition. Spectrographic anclysis of the plug in the fill line of in-pile loop No. 5 revealed sufficient amounts of foreign impurities, such as aluminum, silicon, calcium, and zinc, for there to be reasonable doubt as to whether the oxide content of the original charge was the actual cause of the plug formation. However, to remove at least one uncertainty in future loop operation, extreme care will be taken to assure the delivery of pure material into the loop sump tanks. Shield Mockup Core Materials Fabrication of two *‘orange slices’ which simu- late sections of the SMC has been completed (see Chap. 5.4, “*Shield Mockup Core’). Attempts will 122 be made to pour into these sections the molten salt NaF-ZrF -UF ,-KF (61.7-16.4-1.4-20.5 mole %), which wiil be used to simulate the fuel and NaK coolant in the heat exchanger region. Critical measurements of the ‘‘orange slices’ have been made and any changes in these dimensions as a result of the pouring operation will be noted. One section is fabricated from l/‘-ln type 310 stainless steel plate and the other from é-m type 310 stainless steel plate. After pouring and cooling, the **orange slices’ will be sectioned thoroughly to determine the possibility of void formation in the frozen salt and whether any segregation of the uranium occurred in the freezing process, EXPERIMENTAL PREPARATION OF VARIOUS FLUORIDES B. J. Sturm L. G. Overholser Additional quantities of the several structural metal fluorides have been prepared by the methods employed previously. Increasing interest in the use of rare-earth fluorides for various experimental purposes required the preparation of several such fluorides in substantial quantities. Continuved use has been made of chemical, x-ray, and petrographic examinations to establish the identity and purity of the products. Several batches of CrF, were prepared by hydrogen reduction of (NH4)3CrF6, and additional N:F was obtained by hydroflucrination of NiCl,. Three pounds of K,FeF, was synthesized from aqueous solutions of FeCf and KF. Four pounds of CeF; was prepared by the interaction of an aqueous HF solution with Ce,(CO,),, followed by washing and drying. A small batch of YF, was prepared in a similar manner, with YCI, bemg used as the starting material. Approximately ‘é Ib of NdF, was synthesized from Nd,(CO,) by using an aqueous HF solution. Also a smafl' batch of PrF_ was prepared by the same general method by starting with PrCl,. A small quantity of K,TeF, was synthesized by the addition of aqueous K% to an aqueous mixture of TeO, and HF. A sample of CaF, was treated wifl'\'NH‘F HF and ignited at 1125°C. A somple of K, ZrF was freed of oxide by hydrofluorination at 780°C Recently, work was started on the preparation of molybdenum fluorides by using the reduction of MoF and separation of the reduction products by condensatlon at various temperatures. 'y IR 1 a4 . -+ s S bt e . sl g i o " PERIOD ENDING SEPTEMBER 10, 1956 2.5, COMPATIBILITY OF MATERIALS AT HIGH TEMPERATURE F. Kertesz PENETRATION OF GRAPHITE BY MOLTEN FLUORIDES H. J. Buttram - G. F. Schenck! The study of the behavior of graphite exposed to molten fluoride fuel mixtures was continved. Pre- viously reported preliminary results? showed that APC graphite was completely penetrated by NaF- ZrF, (53-47 mole %) in 1 hr ot 600°C, while NaF-ZrF 4-UF, (53.5-40-6.5 mole %) had not de- tectably penetrated the .graphite after 10 hr. In an attempt to investigate this behavior more thoroughly, an alpha counter was employed to follow the dif- fusion of the urdnium-fluoride-confaining mixture into the graphite. After exposure of 1’ X / x ¥ in, graphite specimens to the fluoride mlxture at 606°C counts were taken on various samples obtained by filing and scraping the surfaces in @ .uniform manner, ' : R The petrographic microscope was used for the examination of samples, as in the previously described experiment,? and it was found that after 10 hr of exposure to NaF-ZrF .UF, (53.5-40-6.5 mole %) there was no evidence of penetrahon into the graphite specimens. After 50 and 250 hr, only slight increases in alpha count were found, while no penetration could be detected by microscopic examination, A graphite sample was then exposed - to the NaF-ZrF -UF, mixture at 800°C and it was observed that after 10 hr there was a sufficient increase in the alpha count to indicate the presence of the su!f msude the tesf specimen. o _ HYDROGEN PRESSURE OF THE NaOH-NI REACTION o H J. Bufl’ram 5"' F. A. Knox | The equ:llbrwm pressure of hydrogen from the 'reachon of NaOH and mckel was. studled earher in - Can: apporatus in which a manometer . was connected : to a quartz. envelope contommg the NaOH ‘in‘a metal - crucible,-- More recently an: opparatus was developed ‘which. ‘permits - measurements of the f.pressure inside a -sealed quartz ‘envelope - con-_' _tummg the NaOH-Nl system. The measurements lOn oks.sierrm_lent /fro-m' Pratt & Whitne} Aircfuh. 2H. Buttram and G. F. Schenck, ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 125 can be made after the apparatus has cooled to room temperature. In a series of experlments, 10-g somples of NaOH were sealed in l/.‘,,x 3 in, tubes of nickel, and each tube was sealed in a quartz jacket. Three sizes of quartz jackets, approximately 25, 42, and 73 ml, respectively, were used. The specimens were soaked at 800°C for periods of 3 to 729 hr, and after cooling to room temperature the pressure of H, in each was determined by breaking the quartz jacket inside a calibrated volume in which the pressure could be ascertained. The quantities of hydrogen recovered from the capsules are shown in Table 2.5.1, and the pres- sure (extrapolated to 800°C) of hydrogen in each capsule is plotted in Fig. 2.5.1. It is apparent ~ TABLE 2.5.1. HYDROGEN COLLECTED FROM REACTION OF NaOH WITH Ni® AT 800°C Hydrogen Collected {cm3+atm) Exposure Time From 25ml From 42-ml From 73-ml (hr) Capsule Capsule Capsule 3 11.3 16.2 21.8 27 37.2 42,0 50.6 81 47.5 54,5 56.0 243 23.0 32.4 47.4 79 0.61 'I.l_ 3.8 . . chmsétfleo .- ORNL-LR~DWE 16162 O o 2 1600 5 ~B 1200 : £ o0 ] .. 5 &% 400 ' & 0 : = _ S 10 C 100 ' 1000 TIME (hr} " Fig. 2.5.1. -»Hicliregen".Pres;ures Developed in Quartz-Jacketed Nickel Capsules Containing NaOH After Various Periods of Exposure at 800°C, 123 ANP PROJECT PROGRESS REPORT from the data that in these experiments the hydrogen pressure, as calculated, ond the amount of hydrogen recovered passes through a relatively sharp maxi- mum at cbout 100 hr and drops to a negligible value. The amounts of hydrogen recovered seem to be somewhat dependent on the volume of the system. ‘ : . These results must be considered somewhat sur- prising in that they indicate that after about 100 hr the rate of evolution of hydrogen from the system become slow with respect to the rate of loss of hydrogen by diffusion from the quartz envelope. Whether this phenomenon is due to @ much reduced rate of reaction or whether the character of the reaction changes abruptly is not known at present, Further studies of this situation will be conducted. - PHYSICAL PROPERTIES OF ELASTOMERS EXPOSED TO ATTACK BY LIQUID METALS D. Zucker Equipment has been assembled for the testing of elastomers exposed to NaK. In the tests, fresh samples of various elastomer materials are heated slowly under tension to 185 to 190°C in NoK in a dry box coentaining a high-purity helium atmos- phere. Two samples of G-E silicone 81576 failed below 165°C, both samples of silicone 81577 failed shortly after reaching 180 to 185°C, ond one sample of silicone 81578 failed cfter a few hours at 185 to 190°C. A second specimen of silicone 81578 was still intact after 25 days of exposure. Two of three Du Pont materials showed promise; sample SR-5550 failed at 150 to 160°C, but SR-5570 remained intact for nine days at 185 to 190°C, Sample 5806, a silicone-resin-treated fabric, was still intact after 25 days. 124 DETERMINATION OF OXY GEN IN NaK Eo Eo KetChen Go Fo SChean A source of NaK of known oxygen content is needed in connection with studies on the solu- bility of structural metals in NoK, and thus « reliable and practical method for analyzing for oxygen is required. A study has been made of oxygen determinations by the amalgam and the butyl bromide methods in an attempt to evaluate the methods and establish whether or not any corre- lation exists between the results obtained. The amalgam method, in which NaK is removed from the oxides by mercury, appeared to be the most adapt- able, since it could be carried cut in the vacuum drybox at the time samples of NaK were loaded into the capsules used in the solubility studies. In this method, samples of NaK which had been passed through a 10-p sintered-glass filter, as well as samples from the same lot that had not been filtered, were analyzed. The results showed about 10 and 25 ppm of oxygen for the filtered and un- filtered samples, respectively, The precision was: good in all cases. To test the butyl bromide method, samples of the filtered NaK were removed from the drybox and reacted with dry, purified n-butyl bromide in a glass vacuum system. When the water extracts were titrated with 0.1 N HCI to a pH of 6.5, values of 80 and 130 ppm of oxygen were obtained. Thus the values obtained by the butyl bromide method are completely out of line with those obtained by the amalgam method.” It was observed, however, that the aqueous extracts behaved as if they were buffered, and consequently these results cannot be considered at all reliable. The cause of this buffering action is being studied (see Chap. 2.6, *‘Analytical Chemistry’’), * W “aw PERrOD'ENDlNG SEPTEMBER 10, 1956 2.6, ANALYTICAL CHEMISTRY J. C White DETERMINATION OF RARE-EARTH ELEMENTS:IN FLUORIDE FUELS A. S. Meyer, Jr. B. L. McDowell - The procedure for the determlnoflon of lanthanum in samples of NoF-ZrF +UF, by precipitation as the oxalate ! has been opphed with good precision ~.dnd accuracy to the determination’ of lanthanum-in virtually - pure LoFa. “The same . procedure, how- ‘ever, did not yield very precise or accurate resylts- - for the determination of cerium in samples of CeF,, - or ‘in NaF-ZrF; -UF ~that ‘contained CeF,, Ac- cordingly, spectrophotometnc methods for the de- termination of cerium are being investigated. The “procedure for the determination of fanthanum is, as - reported earlier, } quite lengthy, and therefore spec-- ~trophotometric methods for the defermmoflon of : lanthanum are also. being studied. ' Speetrophotometric Deiermlnoflon of Cerium - A spectrophofomefnc merhod2 for the determi- nation of cerium with Tiron (dlsodium-'l 2-drhy-‘ . droxybenzene-3, 5-d|sulfonote) was opphed o the - - determination of ¢erium in samples of NoF-LlF-KF ' ~and ‘NaF-ZrF, ‘4 ‘which contained CeF " The re-. - action. of cerium with Tiron yields a colored com- plex which exhibits an -~ absorption - moxlmum at 500 mp: This reaction is selective for cerium in - that colored reaction products are. not formed with- ~any ‘other lanthanide element,2- ‘The .color is de- - veloped in solutions.of high ionic strength at a pH ~_greater. than 8. ‘Although the .color. developmenr is " instantanecus with cerium(lV), ot least 6 hr is: - required __for complete " color - development with. - - cerium(lIl), " For - soluhons of cerium(1il) su!fote,; o - the coefficient of variation of this method is less. ,;_._ferences. . - than 2% for the ronge of cerium concentrohon of - ' . 2to 50 pg/ml. - I, : 1 " Several . elements thot are normoily present in '_"ffiuorlde foels” mterfere with this determmation.' -~ lron"and uranium react with Tiren to form colored* 'f,_;products that obsorb in this region.- erconium o . reacts to" form a colorless comp!ex ond ‘thereby ";r'consumes reogent, whale fluorlde destroys the 5 TA. 5. Meyer; Jr., and B. L. McDowell, ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 126. 2B, Sarma, J. Sci. Ind. Research 14B, 538 (1955), cerlum(lll) Tlron product. Fluoride is normally removed by volatilization during the dissolution of the somple with sulfuric acid. The metallic inter- ferences can be removed by extraction with tri- octylphosphine oxide (TOPO) (see subsequent sec- tion, ‘‘Extraction of ART Fuel Components with Trioetytlphosphirie Oxide,’’ this chapter). - Spectrophotometric Determination of Lenthanum The method of Rinehart3 for the spectrophoto- B 'ihetric:deferminofion of the rare earths and yttrium ‘with sodium glizarin sulfonate (alizarin red-S) is being investigated. The complex which is formed by the reaction of alizarin red-S and lanthanum in an acetate-buffered solution exhibits an absorption -maximum “at 535 mu. This system conforms to “Beer's law for concentrations of lanthanum from 21020 pg/ml. For solutions of lanthanum chloride " the coefficient of variation of the method is 2%. " All the rare-earth elements form complexes which - possess similar molar absorbance indices. There- fore, samples which contain both lanthanum and cerium can be analyzed by determining the total moler concentration of rare-earth elements by the alizarin red-S method, determining the cerium by the Tiron method, ond cnlculatmg the lanthanum by _ dlfference.‘ o Most ions which would normally occur in sulfate " solutions of the samples of NaF-ZrF ,-UF, and rare-earth elements interfere serlously wrth this ‘method. Some of these interfering ions are UOz'H Fe***, AI***, Z:0**, F~, and 50,=~. The ex- troctlon of the solutlon ‘with TOPO is being tested ":cs a method for fhe ehmmoflon of these inter- - _‘-oErékulunlofl_oF TANTALUM IN FLUORIDE SALTS JP. Young J. R, French - The determination ‘of tantalum in NoF-ZrF ‘-UF was studied further in order to ascertain the mini- “mum concentration of tantalum that can be found by - the present method, The determination of tantalum in the range 900 to 1500 ppm in fluoride solts was 3R. W. Rinehart, Anal. Chem. 26, 1820 (1954). 125 ANP PROJECT PROGRESS REPORT discussed previously.4¢5 The sample of approxi- . mately 1 g of salt is carefully digested with dilute H,SO, in order to hydrolyze any TaF to Ta 05, then the solution is evaporated to. dryness. The residue is fused with potassium pyrosulfate, fol- lowing which the melt is dissolved in a solution of ammonium oxalate. If the sample consists of alkali fluorides only, the tantalum is determined.colori- metrically in - the solution of ammonium oxalate with pyrogallol.8 The absorbance of the complex is measured ot 330 mu. If the sample contains either ZtF, or UF,, the tantalum is precipitated with tcmnm at a pH of 5 in a hot solution,. The precipitate is filtered, washed, and then ignited to Ta,04. This oxide is fused with potassium pyro- sulfate, the melt is dissolved in ammonium oxalate, and the tantalum is determined with pyrogallol. An investigation was made of the minimum con- centration of tantalum that could be recovered from synthetic samples by means of the tannin precipi- tation. It was found that approximately 100 ug of tantalum was lost in the precipitation step. This amount of tantalum appeared to be fairly constant for the separation of tantalum from vranium, zirco- nium, or a mixture of these two elements., The separation of tantalum from uranium or zirconium seems to be practical only for amounts of tantalum greater than 1000 ug. DETERMINATION OF OXYGEN IN NaK A. S. Meyer, Jr. T. W. Gilbert, Jr. A comparison of the methods for the determina- tion of oxygen in highly purified NaK has shown that significantly higher oxide concentrations are indicated when the butyl bromide method? is used ‘than when the amalgamation procedure® is used (see Chap. 2.5, *“Compatibility of ngh—Temperuture ‘Materials®’). It was noted that the usual strong- base—strong-acid titration curves were obtained when the oxide residues from the amalgamations 4. c White, - Determination of Small Amounts of Tantalum in NaF-LiF & FITTING ' VESSEL ' Py A 7 , I ] .—RrEACTION 3g-in, SWAGELOK : FITTING MICROMETALLIC FILTER Figs 2.6.1. Reactor for the Determination of Oxygen in Lithium by an Amalgamation Metheod, is dissolved much more rapidly in the presence of ether, in a manner analogous to the Grignard re- action. The most rapid dissolution was accom- plished with solutions of n-butyl iodide in diethyl ether, lodine is added to the solution to destroy organolithium compounds. Samples of lithium of approximately 2 g will dissolve in '/2 hr. The following reactions are postulated for the dissolu- tion of lithium: 2Li +2C Hgl —> 2Lil + CH,, 2Li + CHyl +(GH;),0 —> LiCH,-(C,Hg),0 + Lil LiC Hg(C,H,),0 + 1, —> Lil + C Hyl + (C,Hy),0 - The halogenation procedure may offer additional advantages in that the free iodine is postulated to 128 oxidize lithium nitride and carbide and thus elimi- nate their contribution to the alkalimetric titration. Thus the oxide is determined by direct titration, and the uncertainties inherent in the determination by difference are eliminated. These postulated oxidation reactions, which are thermodynamically feasible, have not yet been demonstrated with pure Li,N and Li,C,, but they are supported 'by an- alyses of lithium samples in which the oxide titration is significantly less than that required for the titration of the nitrides and carbides. In the procedure that is now being tested, a 2-g sample of metallic lithium is added to a solution that contains 100 ml of n-butyl iodide, 200 ml of diethyl ether, and 15 g of iodine. After the dis- solution of the sample is completed, the ether is volatilized and the organic solvent is extracted with water, The aqueous phase is titrated poten- tiometrically with a 0.01 N solution of hydrochloric - acid to a pH of 7. The iodine is removed from an' aliquot of the titrated solution and the lithium iodide in the aliquot is titrated with a solution of-. silver nitrate to determine the sample weight. Concentrations of oxygen in lithium as low as 200 ppm have been determined by this procedure. Determination of Nitrogen and Carbon The method proposed by Nuclear Development Associates 15 for the determination of nitrogen and carbon in metallic lithium is being evaluated. In this procedure the sample is dissolved in water to ‘convert the nitrides and carbides into ammonia and acetylene, respectively. Ammonia is absorbed from the exit gases by a boric acid scrubber and is determined spectrophotometrically. After the mois- ture is removed, the acetylene is passed over heated copper oxide to convert it to carbon dioxide, which is absorbed on ascarite and weighed. A modification of this procedure in which the acety- lene is absorbed in a solution of AgNO, is being tested as a possible means for improving the sensitivity of the determination of carbide. The acetylene is absorbed in a 35% solution of AgNO,,. When the solution is diluted twentyfold, the acety- lene is precipitated and weighed as the compound - Ag,C,-AgNO;. This procedure offers a fivefold increase in the sensitivity of the determination and is selective for acetylene. . L tion of a radiator leak. DETECTION OF TRACES OF NcK IN AIR A. S. Meyer, Jr. J. P. Young Two methods are being studied for the detection of traces of NaK in air. Functional descriptions of the methods and the operational requirements of the necessary apparatus were presented previ- ously. 16 The electronic and optical components of the instrument designed for the photometric de- tection of microgram quantities of NaK in air have been fabricated. The detailed engineering drawings of the instrument for the detection of submicrogram quantities of NaK in air by observation -of the - sodium resonance radiation have been prepared, Experiments were' continued in an effort to find a means for introducing reproducible concentrations of sodium into air.'7 When helium which contained 20 to 100 pug of sodium vapor per liter was mixed with 10 times its volume of air, at least 95% of the sodium was removed from the air stream before it had traversed a 3-in. length ‘of transfer line. The stability of such suspensions of sodium oxide was not improved by varying the tempercture of the transfer line from 150 to 900°C or by altering the design of the mixing chamber to increase the velocities of the streams of helium and air. | Since it was not possible to prepare mixtures of sodium oxide in gir that could be transferred by a jet of helium, @ small test facility will be fabri- cated which will more closely simulate the condi- constructed for the ejection of minute quantities of NaK at a metered rate. This jet of NaK will be injected directly into an air duct, The air will be passed through the duct at velocities comparable | . to those required for: the cooling of NaK radiators in the ART, and the air will be’ heated electflcolly, This test = ‘facility should prov;de samples for definitive tests ' by usmg the pot furnace; g heating mantle is used _' of the opplscoblhty of the proposed Ieok detectors. e to temperatures as high .as 1400°F. COMPATIB!LITY OF FLUOR!DE SALTS AND -. ALKAL! METALS WiTH PUMP LUBR!CANTS ' A S Meyer, Jro - G Goldberg R h.fieyer.‘ ’Jr.;. et Ihl “ANP -Q‘fiar‘.-"'Prorg. 'Rep._ Maer 10. 19561 0RNL'206‘, P 207 VA S, Me yer, Jr., and J. P. Young, ANP Quar Prog Rep. June 10. 1956, ORNL-2106, p 128. An apparatus ‘is being - PERIOD ENDING SEPTEMBER 10, 1956 and molten fluoride salts with the lubricants which are proposed for use in ART pumps. These ex- periments dre designed to simulate the conditions that would exist if leaks developed in the seals of the fuel or coolant pumps. It is necessary to determine whether a hazardous condition would result from an exothermic reaction between the molten materials and the lubricants. The reactions between samples of Dowtherm-A (diphenyl oxide) and of Cellulube-150 (tricresyl - phosphate) -at 200°F with sodium maintained at " 1100°F were carried out in the apparatus 18 which was developed for the determination of oxygen in - sodium by the addition of molten sodium to butyl bromide. Approximately 100 ml of each of these lubricants was placed in the glass reaction vessel and heated to the desired temperature with o heating mantle. The temperature of the oil was measured by placing a thermometer within the re- action vessel, After the transfer line was flushed with sodium at 1100°F, approximately 5 g of the ‘metal was allowed to drop into the lubricant. The reaction with the Cellulube was quite vigorous and exothermic, while there was very little reaction between the molten sodium and the Dowtherm-A sample. In order to provide a versatile apparatus for testing the compatibility of additional lubricants with both alkali metals and fused fluoride salts, the test apparatus shown in Fig. 2.6.2 was de- signed and constructed, In operation, the melt pot is half-filled with either alkali metal or fluoride salt. After the oddition of the oil sample to the reaction vessel, the apparatus is assembled as ~ pictured.. The apparatus is then evacuated and the “system s pressunzed with helium at 1 to 2 psi. _The material in the melt pot is heated to 1100°F to raise the temperature of the oil to 200°F. ‘When the desired temperatures are reached, the ~ helium outlet valve is opened, and the helium inlet ~ regulator is set ot 1 to 2 psi to maintain a helium .. flush of the opporatus. ~ provided to relieve possible pressure surges after » Tests are bemg conducted at elevated tempero-'- _'itures to oscertam the compot:blhty of olkoll metols (The helium outlet is. the addition of the molten material to the oil.) By means of the manual gear drive the hollow plunger " is lowered to displace the melt which flows through mA. L Meyer, Jr., G. Goldberg, and W. J. Ross, ANP Quar. Prog Rep. Dec. 10, 1955, ORNL-2012, p 188, Fig. 9.1. 129 ANP PROJECT PROGRESS REFORT UNCLASSIFIED ORNL— LR—DWG 16164 GEAR DRIVE - THERMOCOUPLE WILSON SEAL TO VARIAC THERMOCOUPLE—_ - BT . | | N - TO VALVE | TO VALVE - HELIUM OUTLET - = Gl HELIUM INLET VACUUM QUTLET FLANGE PLATE i . . . TEFLON GASKET " 1) 7, i) It Hal| I PLUNGER h b 1 \ ”= FLANGE PLATE | || |=—— 6-in. GLASS PIPE , : MELT POT OVERFLOW PIPE S J FURNACE SUPPORT ill‘“ INSULATED POT FURNACE 3-in. GLASS PIPE OIL UNDER TEST Fi-g:.tv-2.6.2. Apparatus for Testing the Compatibility of Alkali Metals and Molten Flruoi-id'e‘SaAlts with | Pump Lubricants. 130 - the overflow tube and drops into the heated lubri- cant, Any reaction that occurs is-observed visually through a }-m.-thlck plastic pipe (not shown), which encloses the entire apparatus. Temperature changes are recorded on a fast-chart-drive, one- point recorder. EXTR-ACTION OF ART FUEL COMPONENTS WITH TRIOCTYLPHOSPHINE OXIDE A, S. Meyer, Jr. W, J. Ross “An investigation has been initiated of the ex- traction of ART fuel components, corrosion prod- ucts, and fission products from acidic solutions into solutions of trialkylphosphine oxides in an organic solvent. Trioctylphosphine oxide (TOPO), which has been recommended for the extraction of vranium from leach liquor,'® has been found to offer two methods for supplementing the existing analytical methods. Traces of structural metals, iron, and chromium can be readily concentrated by extraction from solutions of the fluoride fuels or alkali metals, and thus the sensitivity of the meas- urement of the accumulation of corrosion products can be improved. The extraction may also be used toisolate certain elements, particularly rare earths, so that sensitive and convenient methods may be applied to their determination. , Initial results show that TOPO extracts metal ions from acidic solutions into organic solvents, such as cyclohexane or Varsol, through the forme- ~ tion of definite complexes. The extent to which the extraction of Cré*, Fe2*, and Zr4* occurs can . be made quantitative, or the extraction of the ele- ments can be completely prevented by varying such experimental conditions as type and .concentration of the acid that is used as a solvent for the fluo- The results are practically - ride salt mixture, independent of equnl:braflon periods longer than ‘5 min and of ratios of volumes of the aqueous and - organic. phases. A rapid method of isolating and concentrating the components is fherefore avml- “able. . Experimental resulis have shown fhat 0.5 mmoles of TOPO in cyclohexane extracts quunhtatwely, in “a single equilibration, (1) 10 mg of Cré* from 1 to -7 MHCland 1 to 7 M H,S0,, approximately 9 mg ~ from 8 fo 12 M H PO cmd 'l M HN03, cnd lesser "C A. Bloke, K. B. Brown, and C. F. Coleman, Solvent Extractions of Uranium (and Vanadium) from Acid Liquors with Trialkylphosphine Oxides, ORNL- 1964 (Nov. 4, 1955). and S PERIOD ENDING SEPTEMBER 10, 1956 amounts from other concentrations of these acids; (2) 10 mg of Fe*** from 4 to 7 M HCl and lesser amounts from 1 M HCI, but none from HNO,, H 290, or H,PO,; (3) 10 mg of Zr*t from 1 to 10 M HCI , but none from H3PO4 or HNO.; 3i and (4). mllhgram amounts of U0, * from all four ocnds. Trioctylphosphine oxide does not extract Nit** Crt*t, La***, or the rare earths. Ferrous ion is pamally extracted from HCl because of slow oxi- dation to Fet**, Since the excess TOPO remains in the organic phase, the elements which are not extracted can be determined by the usual methods . Thus, the rare- earth elements can be separated from the fuel constituents which would interfere with their spec- " trophotometric determination. Preliminary studies indicate that other typical fission products, such as barium and rubidium, also remain in the aqueous solution, ‘While the extracted ions can be stripped from the organic phase for subsequent spectrophotometric determination in the aqueous solution, more con- venient determinations of iron and chromium can be carried out by direct measurement of the absorbance of the extract. The molar absorbance indices for the absorption maxima of the Fe*** complex that is extracted from 2 to 9 M HCI is 8000 ot 365 myu and 7500 ot 317 my. These values are comparable to the molar absorbance index of the iron ortho- phenanthroline complex (11,400 at 508 mp).20 Extracts of hexavalent chromium exhibit absorp- tion maxima at 366 mu and 285 my, with indices of 1300 and 1700, respectively. A more sensitive spectrophotometric determination of chromium may be obtained by developing the color with diphenyl carbazide in the organic phase. A survey of the extractability of other elements which may be present in experimental fuels is now being carried out, DETERMINATION OF CHROMIUM IN TRIOCTYL- ' PHOSPHINE OXIDE EXTRACTS WITH DIPHENYL CARBAZIDE _ C. K. Mann?! | The diphenyl carbazide method has been applied _to the spectrophotometric determination of hexa- - yalent chromium after its extraction into solutions 20G, F. Smith and F. P, Richter, Pbenatbroline and | Substituted Phenanthroline Indicators, Their Preparation, Properties, and Applications to Analysis, G. Frederick Smith Chemical Co., Columbus, Ohio, 1944, p 78. 21Research participant, University of Texas. 131 of trioctylphosphine oxide (TOPO) in organic sol- vents. When a 0.25% solution of diphenyl carbazide in ethano!l is added to solutions of hexavalent chromium which have been extracted from 2 M solutions of sulfuric or hydrochloric acid, ¢ violet color is developed which is similar to that formed in an aqueous solution. The organic solutions exhibit an absorbance maximum at the same wave- length as that observed in the aqueous solution (550 mp). When ethanol is used as a diluent, the molar absorbance index is identical to that ob- tained in aqueous solutions. Since the chromium may be concentrated at least twentyfold by ex- tracting the aqueous solutions with small volumes of TOPO solution, the method affords a significant ‘increase in sensitivity, The method may also offer a means for the elimination of the interference of vanadium, which is not extracted from solutions of sulfuric acid. It is proposed to apply this procedure to the de- termination of hexavalent chromium in a sulfuric acid solution of fluoride salts or in acid solutions of alkali metals in 2 M H.‘,SO4 or HCi. The solu- tion is equilibrated with 5 ml of a 0.1 M solution of TOPO in benzene. An aliquot of the organic phase is transferred immediately to a 25-ml volu- metric flask, Two mitliliters of a 0.25% solution of 132 diphenyl carbazide is odded, and the solution is diluted to volume with ethanol. “After 1 hr the absorbance of the solution is measured at 550 my. Tests are currently being made to evolucte this particular application. : ANP SERVICE LABORATORY W. F. Vaughan ‘ A total of 1427 samples was analyzed, which involved 5247 reported results, an average of 3.7 per sample. A breakdown of the work is given below: - - Number Number of of - Reported Samples -~ Results Reactor Chemistry 738 _ 2615 Experimental Engineering 626 2370 Metallurgy 28 30 WADC 27 189 Miscellaneous 8 43 1427 5247 The program of the analysis of special samples from the Wright Air Development Command (WADC) was completed. & ' Part 3 METALLURGY W. D. Manly -4 ” . 3.1. DYNAMIC CORROSION " JH.DeVan FORCED_'-CIR'C'UL'ATIONfLOQIP'T_EST'S | J. H. DeVan - Fuel Mlxtures in. Inconel und Hasfe"oy B. " Two Inconel forced-cnrculohon loops were' ex- ~ amined following-1000 hr of operation with the fuel . . mixture {No.’ 70) NaF- ZrF -UF, (56- 39-5 mole %). The purpose “of these tests was fo compare the corrosion properties of this fuel mixture with those of the similar, more commonly used, -fuel mixture (No. 30) NaF-ZrF ,-UF (50-46-4 mole %) : The lower ZrF4 content of maxture No. 70 gives a fuel . with a considerably lower vapor pressure at reac- tor operahng temperatures than that of muxture o - No. 30. The conditions of operahon for these two Ioops,' 7425-14 and 7425-15, are given in Table 3.1.1. Operation of the Ioops, including the cleaning procedue, followed the same pattern as that normally employed for tests with fuel mixture No. 30. Visual examination of loop 7425-14, which was operated with @ maximum fluid temperature of about 1500°F, showed the presence of few very small metallic crystuls in the top section of the cooling coil and ulong the fluid—inert gas inter- face in the pump. These crystals wer_e_ found by sééctermphic examination to be >5% Cr, >5% Zr, 2% Ni, and 0.2% Fe. Although the fuel undoubtedly icompnsed a large portion of the crystals, the - presence of chromium indicates possible mass transfer. However, no crystals could be found in loop 7425-]5 which operated with the same tem- . -perature gradient but a higher maximum fiuid tem- perature of about 1650°F, The .maximum hot-leg attack observed metallo- . graphically was to a depth of 7 mils and was ap- proximately the same for both loops. The attack occurred as light to moderate void formation, as ~ shown in Fig. 3.1.1.. The similarity in the attack in the two loops (7425-14 and -15);, which operated at different -maximum fluid temperatures, is thought to have limited significance. For loops operated with fuel No. 30 under similar conditions, the average - hot-leg attack is 4.5 mils at a maximum fluid temperature of 1500°F and 9 mils at a moximum fluid temperature of 1650°F. The value of 7 mils for the hot-leg attack at both temperatures in the tests with fuel No. 70 is felt to be within the range of possible experimental deviations and does not represent an abnormally large difference - from either of the values obtained for fuel No. 30. TABLE 311 CONDITIONS OF OPERATION OF THREE FORCED—CIRCULATION LOOPS WITH FUEL MIXTURES Operating time: ]000 hr Temperuture gradienf 200°F o L - Ratie of hot-leg’ ,s_u_rf_ace' to _loop volume: ‘2.08 in.” e Op"“f'flg iéd'na'ir'tibfi;ij? - —_— Loop Number o 77425_14 7425-15 7425-13 '._.:-VLoc;p mataria] Inconel i !ncbnel o ,'H‘o‘.strelrloY‘B ' "'-'::;:i.Fuel rmxiure 7 el B N°-70* - : NO 70* No. 107** {Maxsmum fuel miinxwre te-mpernfun.a‘,‘rm F 1525 - .71643 , | ‘-"505 o ‘:MQXImt;I'I:l fube woli temperuture, F T ]595 | ‘.‘712 ‘555 : - V'JL’-‘:,:-':‘Reynolds number of fuel ‘7 : 10,900 :' o 95_00.. - . 10,000 ;--;Q-'_V.,.;;:,Ve|ocfly of fuel fps L | 48 L S 26 4 | *Composition: ‘NaF-ZrF -UF , (56.39-5 mole %). **Composition: NaF-KF-LiF-UF (”;2-43-45.3-2.‘5 rfiole %). 4 135 ANP PROJECT PROGRESS REPORT Fig. 3.1.1. Hot-Leg Surface of Inconel Forced- Circulation Loop 7425-14 After Operation with the Fuel Mixture (No. 70) NaF The results of the first test indicated that a chromium plate on the Inconel surface in contact with the beryllium would reduce the extent of the alloying reaction which occurs when Inconel and beryllium are in direct contact in sodium., The thin chromium platings (1 and 2 mils) used in the first test were entirely consumed by alloying during the 1000-hr test period at 1300°F. Therefore o second test was conducted in which heavier platings were used. A beryllium oxide specimen was alsc included in this test to determine whether a reaction would occur between 3E. E, Hoffman, ANP Quar. Prog. Rep. June 10, 1956, 160 beryllium oxide and beryllium or beryllium oxide and Inconel. * The results of this test are given in Table 3.2.6. The presence of a chromium plate *‘diffusion barrier’’ on the Inconel specimen resulted in the formation of an alloy layer (principally Be,Cr) which was approximately one-third the thickness of the alloy layer which forms when Inconel is ‘placed in direct contact with beryllium under the conditions of this test. Analyses of the test resuhs indicate that a minimum of 5 mils of chromium plate will be neces- sary to ensute that all the chromium is not con- sumed by the alloying reaction with beryllium. The Be,Cr phase was identified by x-ray analysis. Thls phase has a Vickers hardness of 2440, as compared with hardnesses of 1495 for Be,Nig, the phase which forms when Inconel and beryllium are in direct contact, and 180 for Inconel (Figs. PERIOD ENDING SEPTEMBER 10, 1956 TABLE 3,2,6. RESULTS OF METALLOGRAPHIC EXAMINATION OF INTERFACES OF SPECIMENS FROM SODIUM. BERYLLlUM-INCONEL COMPATIBILITY TEST NO. 2 " Test duration: 1000 hr ~ Test temperature: 1300°F Contact pressure between specimens: 500 psi Irnterche_ Results of Metallographic Examination Inconel vs berylli‘urm, ‘ direct contact {standard) Inconel plus 4-mil chromium Alloy formation (Be2lNi5 and BeNi) 25 mils deep along interface; in carlier test 24 mils of alloy formed; 4 to 5 mils of Inconel consumed by alloying reaction Alloys formed between chromium plate and beryllium to a depth of 8 mils, - plate vs beryllium which consisted of 7 mils of Be,Cr and 1 mil of Be Cr(?); 0 to 2 mils of chromium plate remained after the test; 2 to 3 mils of Inconel consumed, probably by alloying with the chromium plate. ‘Inconel plus 6-mil chromium ‘plate vs beryllium Alloys formed between chromium plate and beI-yllium to a depth of 9 mils, which consisted of 8 mils of Be Cr and 1 mil of Be Cr(?), 3 mils of " chromium plate remained after the test; upproxlmfeiy 2 mils of Inconel consumed by alloying with the chromium plate Beryllium vs beryllium oxide Surface of beryllium oxide discolored slightly; specimens were easily separated and no bonding was evident Beryllium oxide vs Inconel Neither the beryllium oxide nor the Inconel surface affected by test 3.2.15 and 3.2.16). No reactions were anticipated between beryllium oxide and beryllium or beryllium oxide and Inconel, and none were found to have occurred during this test, EFFECT OF DIFFUSION COLD TRAPS ON MASS TRANSFER IN INCONEL-SODIUM : The purpose of opercflon of 1I1e two Inconel- = sodwm thermal-convechon Ioops dtscussed here;{,; “was _to “determine - ‘whether fhe presence of o dif-: - fusion - ‘cold’ trap wouId affect the extent of mass-.f_ o : r’,transfer observed in the cold legs. The Ioops were . .. loaded wnh sodlum ‘which had :been preirected by design, with no cold trap. : Although many " thermal-convection loop tests conducted on Inconel-sodium systems in .the past E. E Hof{man have failed to show any mass transfer, it has been demonstrated that mass transfer will occur in such systems if the hot-zone temperature is in excess of 1500°F and if a steep temperature gradient is induced in the cold leg by directing an air blast on a small area near the bottom of the cold leg. ~ These loops (Nos. 28 and 29) were therefore . o = “operated for 1000 hr at hot- and cold-zone tempera- . TH,ERMAL CONVEC“ON LOOPS L7 tures “of 1600 and 990°F, respectively, - The loops “were - carefully: sinpped of sodium after the tests, The mass-transfer deposits found in the cold zones of these |oops are shown iin. Fig. 3.2.17. Although ‘;only smull “amounts of mass-fransfer crystals were :found in these Ioops, it is apparent thot- slightly more - deposmon occurred in the loop. which had no _ ;dlffuswn cold trap than in'the loop ‘which mcluded ' i-"cold trapping -for several months to lower the oxy:- a cold trap. -The results of. specfrograph:c analy- 7 gen. ccntent., One loop was - prowdec! WIfl'I a dlf-,_’f-' ~fusion cold trap near the- boftom of the" cold:leg ln"‘"-_.]r-‘. . order - to reduce ‘the sodium - oxide «content during o operoflon, ‘while the other loop was of standard"ii - sis of the metallic crystals recovered from these loops ‘are presented in Table 3 2.7 -along with. the *onalysis’ of- the !nconel plpe prior to fest. More ~ tests will be reqmred 1o establtsh whether the " differences in-chromium and iron confents of the crystals from the two loops are significant. The mass-transfer crystals found in a thermal-convection 161 ANP PROJECT PROGRESS REFORT Fig. 3.2.15. Alloying Between Chromium-Plated Inconel and Beryllium During 1000-hr Compatibility Test in Static Sodium at 1300°F. Dark area between Be ,Cr and beryllium is due to separation of speci- mens after the test. Unetched. 300X. (Secrei-wfl-lfl:up-rren) Flg. 3.2.16. Enlargement of Area Shown in Fig. 3.2.15. Note extreme hardness of Be,Cr phase as compared with the hardness of the Inconel and the chromium plate. Unetched. 500X. (Seere!—wflh' ceptiom~ 162 LY ] T UNCLASSIFIED | PERIOD ENDING SEPTEMBER 10, 1956 S UNCLASSIFIED § Y-18622 Fig. 3.2.17. Mass-Transfer Deposits in Cold Zones of Inconel Thermal-Convection Loops Which Circulated Sodium at @ Hot-Zone Temperature of 1600°F and a Cold-Zone Temperature of 990°F. (a) Cold-zone section of loop which had no diffusion cold trap {loop No. 29). (&) Cold-zone section of loop which had a diffusion cold trap (loop No. 28). {Cerfidertiatwith~caption) TABLE 3.2,7. ANALYSES OF MASS-TRANSFER CRYSTALS FROM INCONEL-SODIUM THERMAL-CONVECTION LOOPS Operating period: 1000 hr Hot-zone temperature: 1600°F Cold-zone temperature: 990°F Material Analyzed Chemical Analysis (wt %) Ni Cr Fe Cu Mn Crystals from loop No. 28 which included a dif- 84.4 10.6 4.6 0.33 0.07 fusion cold trap Crystals from loop No. 29 which had no cold trap 77.4 13.3 9.1 0.06 0.08 Inconel pipe {as-received) 73.15 14.81 6.62 : 0.40 loop operated previously at a hot-zone temperature of 1500°F had a much lower iron concentration.* The results of metallographic examination of speci-_ mens from similar locations in the loops are given in Table 3.2.8. ‘Examinations of these results indicate that the loop with no cold trap (No. 29) had slightly heavier attack in the hot leg (Fig. 3. 2.18) than the loop with a cold trap (No. 28). 4E. E. Hoffman, ANP Quar, Prog. Rep. Sept. 10, 1955, ORNL-1947, p 113. The deposits found on.the cold-leg surfaces of these Ioops may be seen in Flg. 3. 2 19 THERMAL-CDNVECTION LOOP TESTS OF VARIOUS STRUCTURAL MATERJALS AND LITHIUM E. E. Hoffman There has been considerable interest in the possible application of lithium as a reactor coolant and, in particular, as a coolant for a solid-fuel- element aircraft reactor, The corrosion problems 163 ANP PROJECT PROGRESS REPORT TABLE 3.2,8. RESULTS OF METALLOGRAPHIC EXAMINATION OF VARIOUS SECTIONS FROM INCONEL-SODIUM THERMAL-CONVECT!ON LOOPS ' Operating Metallographic Reéu'lts, _ L Location of Temperature Loos No. 28 — Loop No. 20 - Specimen ('QF) oop MNo. - L pp - 29 - , . - {Diffusion cold trap) {No cold trap) pr'qf hot leg 1600 Grain boundary voids to a deptfi 1 to 2 mils of grain boundary 'aflack | _ 7 of less than 1 mil ' o Top of cold leg 1350 S!ighfly irregular surface Slightly irregular surface ‘Bofiom’of cold leg 990 Two-phase surface layer 0.5 mil Two=-phase surface |ayer O.b mlt s thick thick Fig. 3.2.18. Hot-Leg Surfaces of Inconel-Sodium Thermal-Convection Loops Operated for 1000 hr with a Hot-Zone Temperature of 1600°F and a Cold-Zone Temperature of 990°F. (a) Hot-leg specimen from loop No. 28 which included a diffusion cold trap. ' (b) Hot-leg specimen from loop No. 29 which had no cold trap. 500X. Reduced 4%. (Cenfidential-witheaption) encountered in attempting to contcin lithium in high-temperature systems are much more difficult than those found with liquid sodium at high tem- peratures. Analyses of o large number of static tests have indicated that nickel and nickel-base alloys are very heavily aftacked even by static, isothermal lithium, and thus many of the commer- cially available high~temperature alloys, are elimi- ncted from consideration. Pure iron has shown 164 good resistance to lithium, and therefore con- siderable effort has been directed towards tests on iron-base alloys, in particular, the stamless steels. The results of thermal-convection Ioop tests of various materials are summarized in Table 3.2.9. The important information contained in this table is whether or not the loop plugged, the weight of metal crystols recovered from the cold leg of the loop, and the depth of attack in the hot . PERIOD ENDING SEPTEMBER 10, 1956 Fig. 3.2.19, Cold-Leg Surfaces of Inconel-Sodium Thermal-Convection Loops Showing Mass-Transfer Deposits. The decreased width of the scratch ot the edge of the specimens indicates that this deposit is quite hard (possibly a carbide) as compared with Inconel or nickel plate. (a) Cold-leg specimen from loop No. 28 which included a diffusion cold trap. (&) Cold-leg specimen from loop No. 29 which had no cold trap. Etched with aqua regia. 1000X. Reduced 6%. {Cenfidentiatwithraprion— leg. All the loops were constructed of Y%-in, sched-40 pipe (0.84 in. OD, 0.622 in. ID). The {ow-sodium-content lithium used in these tests was received from the vendor packed in gas- tight helium-filled containers, The loops were foaded with lithium and. inert-gas arc-welded. Inconel (nommal ‘composition, in wt %: Nl, 77; Cr, 15; Fe, 7) showed extensive mass transfer and _hot-leg attack, even at hot-leg temperatures as low as.1300°F." This result is typical of nickel- base -alloys exposed to Ilfhlum under these con- dmons. S The six type 3]6 stainless steel (nommal com- . position, in wt %: Fe, 68; Cr, 17; Ni, 12; Mo, 2)" loops tested showed less mass transfer and less of a ‘tendency to plug in the three tests conducted with’ hot-leg temperatures of approximately 1500°F - than in the other tests conducted at hot-zone tem- peratures of approximately 1400°F and 1300°F (Figs. 3.2.20 aond 3.2.21). The loop (No. 25) operated with a hot-zone temperature of 1400°F had more mass-transfer crystals in the cold leg than did the loop (No. 16) operated with a hot-zone temperature of 1310°F; however, loop No. 25 operated over seven times longer before it plugged than did loop No. 16. The hot- and cold-leg surfaces of loop No. 25 after the test are shown in Fig. 3.2,22, and the hot- and cold-leg surfaces of loop No. 16, which plugged in 290 hr, are shown in Fig. 3.2.23. The formation of carbide crystals on the surfaces of lithium loops has been noted in almost all tests conducted with stainless steels, and it usually occurs in that section of the loop which is at a temperature of approximately 1300°F. The carbide deposits have been found in the hot legs of some loops and in the cold legs of others. ‘Type 321 stainless steel (nominal composition, in wt %: Fe, 70; Cr, 18; Ni, 10; Ti, 0.5) when tested at a hot-leg temperature of 1500°F com- pletely plugged with crystals in 204 hr, Type 347 stainless steel (nominal composition, in wt %: Fe, 70; Cr, 18; Ni, 10; Nb, 1) showed results similar to those for type 321 stainless steel, with 165 ANP PROJECT PROGRESS REPORT TABLE 3.2.9. RESULTS OF THERMAL-CONVECTION LOOP TESTS OF VARIOUS ALLOYS EXPOSED TO CIRCULATING LITHIUM Mefalldgrnphic Results System Temperatures (°F) Test . — ~ Material ~ Duration Hot-Leg _ R ' Hot Leg Cold Leg Differential ~ (hr) . Aftack ) Col_d Lg,g : S , o : {mils) o Inconel 1300 1200 100 " 1000 16 151 g of crystals . Stainless steel . Type 316 1500 1292 208 500 13 0.5 g of crystals | 1490 1220 270 000 L5 0.1 g of crystals 1472 . 1335 117 1000 . 3 . 0.1 g of erystals 1400 1130 270 2150* 23 4.7 g of crystals 1310 1058 252 290* 2 0.8 g of crystals 1292 1094 198 1000 15 0.25 g of crystals Type 321 1500 1220 280 204* 0.5 1.0 g of crystals 1310 980 330 1230* 3.0 0.7 g of crystals Type 347 1500 1112 388 280* 2 1.5 g of crystals 1310 1060 250 1000 3.0 1.3 g of crystals 1000 618 382 1000 1 Crystals 0.2 mil thick 1000 618 382 3000 1.5 Crystals 0.3 mil thick Type 430 1500 1220 280 1500 4.0 L0 g of crystals Type 446 1500 1166 334 864* 1.0 9 g of crystals (84 mils of attack) 1500 1200 300 700* 16 6.8 g of crystals (84 mils of attack) 1292 1058 234 1500 25 0.23 g of crystals {65 mils of attack) *Loop plugged with crystals. 166 UNCLASSIFIED Y-18640 (o) NS 'N;-CH__ (&) Fig. 3.2.20. Type 316 Stainless Steel Thermal- Convection Loop (No. 25) Which Circulated Lithium for 2150 hr at a Hot-Zone Temperature of 1400°F and a Cold-Zone Temperature of 1130°F, (a) Speci- men from hot zone, (&) Specimen from celd zone. plugging occurring in 280 hr when a hot-leg tem- perature of 1500°F was used. A 3000-hr test of a type 347 stainless steel loop with a hot-leg tem- perature of 1000°F indicated that this material would be satisfactory for operation in this tempera- ture range for very long periods of time if a small amount of attack and mass transfer could be tolerated. . A type 430 stainless steel (nominal composmon, in wt %: Fe, 83; Cr, 16; C, 0.10) loop operated at a hot-leg temperature of 1500°F (cold leg tem- perature, 1220°F) for 1500 hr without plugging, but it showed considerable mass transfer (Fig. 3.2.24). Two type 446 stainless steel (nominal composition, in wt %: Fe, 74; Cr, 25 C, 0.35):loops tested at a hot-leg temperature of 1500°F (cold leg tempera- ture, ]200°F) completely plugged in less ‘than = 900 hr and had 6.8 and 9 g of metal crystals in PERIOD ENDING SEPTEMBER 10, 1956 UNCLASSIFIED Y-18373 () Fig. 3.2.21. Type 316 Stainless Steel Thermal- Convection Loop (No. 16) Which Circulated Lithium for 290 hr at a Hot-Zone Temperature of 1310°F and a Cold-Zone Temperature of 1058°F. (a) Speci- men from hot zone, (b) Specimen from cold zone. (Eorfidentiai-with-eaption) metals are transferred in appreciable quantities. If the alloy contains nickel, the mass-transfer crystals are richer in nickel than is the container - alloy, . In iron-chromium alloy loops, the crystals jdepoSited ‘are -richer in iron than is the container the cold legs (Fig. 3.2.25). The wall-crystal inter- face on the coldsleg surface is shown in Fig. 3.2.26, The bulk of the crystals are ferrite (composition,: in 'wt %: Fe, 90; Cr, ), while those attached to the surface of the type 446 stainless steel wall have been identified by x-ray analysis as Cr,,C,. - The relative hardnesses of these two phases may ; be noted in the illustration. The “results of -chemical analysis of the mcss-:" transfer crystals from the nickel-iron-chromium and iron-chromium _ alloy loops show that all three L _alloy. Of the three metals - nickel, iron, and "__‘rchromlum ~ chromium has the least tendency to “mass’ transfer, but-it is still found to be present to the extent of 5 to 10% in the mass-transfer crystals found in stainless steel=lithium tests. . . The data obtained in these thermal-convection f’loop tests indicate clearly: that further work needs " to be ‘done to resolve the" apparent discrepancies -which exist between the results for type 316 stain- less steel and ‘other austenitic stainless steels which contain little or no molybdenum, Also, types 317 and 318 stainless steels, which contain 167 ANP PROJECT PROGRESS REPORT 227 VHN 25-g LOAD 14543 VHN '266 VHN Fig. 3.2.22. Hot- (2) ond Cold-Leg (b) Surfaces of Type 316 Stainless Steel Loop No. 25 Which Circulated Lithium (See Fig. 3.2.20). Hot-leg surface attacked along grain-boundary carbides. Two- phase mass-transfer crystal (chromium carbide, VHN, 1513) plus attached crystal containing iron, nickel, and chromium are shown on cold-leg surface. Etched with glyceria regia. 500X. Reduced 13%. (Gent dential-witirception Fig. 3.2.23. Hot- (2) and Cold-Leg (b) Surfaces of Type 316 Stainless Steel Loop No. 16 Which Circulated Lithium (See Fig. 3.2.21). Crystal on hot-leg surface is chromium carbide. Note mass-' transfer crystals on cold-leg surface. Etched with glyceria regia. (Genfidentielwith-coption) 168 L UNCLASSIFIED “Y-18087 lal_.. (b) Fig. 3.2.24. Type 430 Stainless Steel Thermal- Convection Loop (No. 14) which Circulated Lithium for 1500 hr at @ Hot-Zone Temperature of 1500°F and a Cold-Zone Temperature of 1220°F. () Speci- men from hot zone., () Specimen from cold zone, more molybdenum than ls'present in type 316 stain- less steel, will be tested to determine the effect -of molybdenum on the mass-transfer tendency of - stainless steels in contact ‘with molten lithium. The ferrmc stainless steel test results are not encouraging, especially those for type 446 stain- - less steel, which suffers very deep mtergranular. attack ‘because of. preferential- ‘attack’ of the grain- boundary carbides. Further tests are- planned for - type 430 stamless steel. “The superior corroslon',_t resistance of tfype 430 stainless - steel, in com- parison - wrth type 446 stainless steel, can be attributed to the lower carbon ‘content of the ‘type . . 430 stainless steel (0 10% C maximum). Type 446 stainless steel has a maximum of 0.35% C. 'The data obtained in static and thermal-conVectian, - loop tests ore summarized in Fig. 3,2.27, which shows that the materials tested thus far are un- PERIOD ENDING SEPTEMBER 10, 1956 satisfactory as containers for lithium at hot-leg "vtemperatures of ‘approximately 1500°F in flowing systems. Tests are presently belng conducted on ~niobium, - molybdenum, and zirconium in dynamic systems, and it is believed that these refractory metals, especnally molybdenum, will withstand the . cotrosive action of lithium under these conditions, - but, as yet, this has not been demonstrafed. ' The effect of various. amounts of nitride cantammahon on the amount of mass transfer will be studied further. Experiments will be conducted in order to ~ determine the maximum. temperature (in the range ~ from 1000 to 1300°F) ot which a stainless steel- 'dynamic lithium system may be operated, since these alloys at present seem to be satisfactory at 1000°F -and, in general, unsatisfactory at 1300°F. " The temperature limits indicated in Fig. 3.2, 27 are merely estimates arrived at by evaluating the data 4. 'obtained thus far, and they may. be in error by as | much as 100°F. TESTS OF lNTERMETALLIC COMPOUNDS IN SODIUM AND IN FLUORIDE FUEL W, H. Cook The mtermetall:c compounds NlAI NiAl + 5% NI, NiAl + 4% Zr, and MoAl have been screen tested for 100 hr in static sodium and in static NaF- ~ZrF - UF4 (53.5-40-6.5 mole %) at 1500°F. No attack was found on the specimens tested in the sodium. Those tested in the fused salt were severely attacked. . The test results, which were obtained by metallographlc comparisons of untested and tested specimens, are presented in Table 3.2.10. Chemical analyses of the test medlums are bemg made, . LAl the mtermetall:c compounds were found to have a common phase that may have been ‘an oxide added by the fabricator for strengthening.d - This phase constituted less than 5% of the su'rfac'e‘ of any -sectioned intermetallic ”co'mp'ouni:l it was -homageneously distributed in-all specimens’ with the exception of the N|A| + 5% Ni compound, in ~-which some ‘segregation was apparent. 'In addition to this common phase, all ‘the intermetallic com- ' ,pounds had one major phase, except the NiAl + 4% 'Zr - compound, which contamed two -additional ‘phases.' _ ',bon.c"le_d‘ globular particles, and the three minor The major phase was - present as well- G, Sterh,‘ Refractory Type Materials for High Teme .berature Application, NP-4527 (August 1953), p 139-140. 169 ANP PROJEC:I' PROGRESS REPORT B CRYSTALS e Fig. 3.2.25. B UNCLASSIFIED 8 g Y-14932 § FLL poT E Type 446 Stainless Steel Thermal-Convection Loop Which Circulated Lithium for 864 hr. (Eonfidentiot-withTaption 170 | _ - "PERIOD ENDING SEPTEMBER 10, 1956 ” Fig. 3.2.26. Cold-Leg Surface of Type 446 Stainless Steel Loop Which Circulated Lithium (See Fig. 3.2.25). Very hard phase (VHN, ~1100) is Cr,,C,, while soft crystal attached to it is a ferrite crystal containing 90% Fe and 6% Cr. Efched wnth glyceria regia. (Een{-rd'en+ru-|-m+'l-r-eu-pt-oon) ' TABLE 3.2.10. SUMMARY OF THE RESULTS OF EXPOSURE OF INTERMETALLIC COMPOUNDS FOR 100 hr IN STATIC NflF-Zl‘F4-UF4 (53.5—40-6.5 MOLE %) AT 1500°F 1 Hic o Attack (mils) ntermeta € — — — Metallographic Results 'CP""PQUHJ : Maximum Average Minimum : MoAl - L o C SR Quanfltaflve measurements could not be made on the ' ' ST T - Mo Al compound becnuse of its extreme brittleness, but visual examination indicated that n hud been : _ ~attacked by ihe fused snlt 7 L N:A! S 20 o 15 8 The 6tfack pro_duced alternate zones (bands) of ' L : 7 R - B R - various degrees of porosify"pqrallei‘ to the surfaces L e B L L S ~ of the specimen. : : NiAD+ 5% NTRE '7,-'32 e 3 o The cnd of the specimen where 1he maxamurn qflack o b i o - - was found did not appear to ‘be bonded as weli as * Ve m | L e - :th_e re_mgmder of the_sp:ec:men ' 1‘ | NiAl + 4% Zr 74 61 8 The attack was severe to the depths indicated, but it ) S ' was most severe along the edges fer an average depth of 6 mils 171 ANP PROJECT PROGRESS REPORT UNCLASSIFIED ORNL-LR-DWG {2080A IRON LOW-ALLOY STEELS FERRITIC (Fe —Cr) STAINLESS STEELS T FRRRRTRFERTTTETTR AUSTENITIC (Fe—Ni—Cr) STAINLESS STEELS NICKEL N R TITRER NICKEL -BASE ALLOYS T I imamarars™ (INCONEL) REFRACTORY METALS {Mo, Nb, Ta, Zr, Ti, W, Va) PRECIOUS METALS Y POOR RESISTANCE (Ag, Au, Pt) VER lOO EIS S (IJ PREDICTED NN STATIC SYSTEMS RESN DYNAMIC SYSTEMS Fig. 3.2.27. Corrosion Resistance of Various Metals and Alloys in Lithium. Bars indicate approxi- mate temperatures below which a system might be operated for 1000 hr with less than 0.005 in. of ettack or container surface removal. (Cenfidentiel-withcaption) phases were located in the few open spaces be- tween the particles of the major phase. The NiAl + 5% Ni specimens were the only ones in which potosity was found, and it was negligible in these, _For the NiAl-base intermetallic compounds the tused-salt attack was probably due to the aluminum. The attack began intergranularly and advanced into the grains. The differences in the extent of the fused salt attack on the NiAl-base intermetallic compounds was probably caused by differences in the accessibility of the NiAl to the fused salt, In the NiAl + 5% Ni specimen it was noted that the attack was most severe where the ‘‘oxide’’ phase was concentrated. It is assumed that oxide quickly 172 corroded away and allowed the fused salt to attack the intermetallic compound. The severe attack in the case of the NiAl + 4% Zr specimen can probably be explained by the removal of both the common phase and the zir- conium, These corrosion tests indicate that the NiAl compound in the NiAl-base intermetallic com- pound was attacked when it was exposed to NaF- ZrF UF,, (53.5-40-6.5 mole %) at 1500°F. The poor corrosion resistance of these inter- metallic compounds in the fused salt obviates further tests in this medium, However, the results of the sodium tests are encouraging enough to warrant more severe corrosion tests in sodium. X PERIOD ENDING SEPTEMBER 10, 1956 3.3. FABRICATION RESEARCH J. H. Coobs NICKEL-MOLYBDENUM ALLOY DEVELOPMENT T. K. Roche The search was continued for a superior nickel- molybdenum-base alloy with the corrosion re- sistance of Hastelloy B and elevated-temperature strength at least equal to that of Hastelloy X. Seamiess tubing has been fabricated from the various alloys under consideration to provide material for evaluation in thermal-convection loops. The extrusion techniques described previ- ously! have been successfully used to produce tube blanks of the variety of alloy compositions being studied. The nickel-molybdenum-base-alloy tube blanks submitted to the Superior Tube Co. for redrawing responded satisfactorily, in general, with the several exceptions being notably the high-carbon- content compositions. The processing procedure used for the as-extruded, 1.5-in.-OD, 0.25-in.-wall tube blanks is to (1) condition the tube blanks by machining (resultant average weight loss, 50%); (2) reduce to 0.875-in. OD and 0.095-in. wall tubing; (3) anneal at 2050°F in a dry hydrogen atmosphere and water quench; (4) rod draw and sink to final size, with intermediate anneals; and (5) finish by sand blasting and rotary polishing. An average yield of 70% is being obtained from the blanks after the initial conditioning operation. It has been found in the processing of these alloys that lighter cold reductions per pass must be taken than in the case of stainless steel or Inconel. A 20% reduction in area per pass is considered to be optimum. ' ' H. Inouye Battelle Memorial Institute Alloys Ten tube blanks of three different alloys were extruded and submitted to Superior Tube. Co. for redrawing in order to assist Battelle Memorial Institute “with nickel-molybdenum alloy develop- ment work - under. way there.! The results of processing these extrusions to 0.500-in.-0D, 0.035-in.-wall tubing are presented in Table 3.3.1. H. Inofiye and T. K. Roche, ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 163. The response of these compositions to cold tube- forming techniques was only moderately suc- cessful. The difficulties encountered are illus- trated in Fig. 3.3.1. Metallographic examination of the failure found in the tube blank of alloy B-2898 (20% Mo-1% Nb-2% Ti-0.80% Mn-0.12% C-bal Ni) revealed the stringering of second-phase particles (that is, carbides), as shown in Fig. 3.3.2, to be the probable source of the difficulty. All attempts to produce seamless tubing of alloy B-2899 (20% Mo-1% Nb-0.80% Mn-0.20% C-bal Ni) failed. Two of the four extrusions of this alloy developed longitudinal cracks along their lengths, as shown in Fig. 3.3.1, while on the draw bench. The other extrusions, also shown in Fig. 3.3.1, failed during the initial tube-reducing operation, Examination of the cracks again revealed that stringering of the numerous precipi- tates was the source of failure. The results indicate that carbon in amounts of 0.20% or greater in the nickel-molybdenum alloys being ~ investigated will make improbable the success of the tube-forming process and that 0.12% carbon is marginal in this respect. UNCLASSIFIED Y-19445 Heat B-2899, Extrusion 64 Alloy Composition: 20% Mo—1% Nb-0.80% Mn- 0.20%.C-Bal Ni Heat B- 2899, Extrusion 60 Aloy Composition: Same as Above Heat B-2898, Extrusion 107 Alloy Compaosition: 20% Mo-{% Nb-2% Ti~0.80% Mn-0.12%C - Bal Ni Fig. 3.3.1. Examples of Failures Which Occurred During Cold Reduction of Tube Blanks Fabricated from Battelle Memorial Institute Nickel-Molybdenum Alloys. 173 Vil TABLE 3.3.1. RESULTS OF REDRAWING OF BATTELLE MEMORIAL INSTITUTE; INTERNATIONAL NICKEL COMPANY, AND ORNL SPECIAL NICKEL MOLYBDENUM ALLOYS TO 0.500:in,-0D, 0,035-In.~WALL TUBING All Nominal C 1 Number of Total Length "d'*' _ bor of L \ oy ominal Composition Extruded Tube of Tubing - Yie . | N':’m Fr : l OOZ:* Remarks O (wt %) Blanks Received (%) Being Fabricate Battelie Memorial Institute Alloys B-2897 20 Mo=1 Nb-? Ti-0.80 Mn-0.12 C—-bal Ni 2 9f2in, 40 1 B-2898 20 Mo~1 Nb-2 Ti-0.80 Mn—0,12 C—bal Ni 4 19 ft 6 in. 2 One tube blank split on ’ ; ' tube reducer B-2899 20 Mo=1 Nb~0.80 Mn..0,20 C=bal Ni 4 0 0 0 " Two tube blanks split on tube reducer; two split on draw bench International Nickel Company Alloys T-23011 15 Mo=5 Cr-3 Nb-3 W-0,5 Al~0.02 C~bal Ni 2 11 #t 11 in. 43 1 T-23012 17 Mo-0.5 A1-0,02 C~bal Ni 1 12 ft 69 1 T-23013 15 Mo=3 Nb-3 W_0,5 A1-0,02 C~bal Ni 3 27 ft 6 in, 76 2 T-23014 15 Mo=1,5 Ti=1 Al—-0,02 C~bal Ni 1 7 ft4in, 80 0 Both blanks failed on tube reducer ORNL Alloys 30-1 17 Mo=3 Cr=0.06 C=bal Ni 3 30 ft 10 in, 76 2 37A-1 20 Mo-3 Cr-0.02 C~bal Ni 1 8 ft 2 in. - 69 1 30-2 17 Mo~5 Cr=0.06 C—bal Ni 3 36 # 3 in. 81 '3 43A-3 20 Mo~7 Cr—0.02 C~bal Ni 1 8 ft 4 in. 73 . 1 30-4 21 #1 6 in. 81 2 17 Mo=10 Cr-0.06 C-bal Ni 2 *Percentage yield of tubing from conditioned tube blank, **Approximately 10 ft of tubing is required to fabricate a thermal-convection loop, 1A0dIY SS3¥I0Ad LIO3r0dd dNV av The compositions of new alloys that have been received from Battelle Memorial Institute for evaluation are presented in Table 3.3.2. These alloys were received in the form of forged ex- trusion b:llets, swaged impact-tensile specimens, and rolled creep-rupture specimens, The impact- PERIOD ENDING SEPTEMBER 10, 1956 tensile specimens were sent to Rensselaer Polytechnic Institute for weldability studies, and the creep-rupture specimens will be tested at ORNL in the fuel mixture (No. 107) NaF- KF-LlF- UF, (l] .2-41-45.3-2.5 mole %). Extrusion experi- mem‘s were conducted at ORNL on two forged - Fig. 3.3.2. Longitudmal Section of Tube Blank of Alloy B.2898, Showing Crack Propugnhon Along Carb:de Sf;ingers Thai Occurred Durlng Tube Reducmg. : Etched wuh chrome regm. IOOX TAB LE'f_3'.V3'..'2. - COMPOSITION OF NEW -A'LLofs, RECEIVED FROM BATTELLE MEM'oR_l_Al: _ms‘r_rr_uTE FOR _EYALUATION c Nommul Composiflon {wt %) Alloy No.. Llmon ‘ , | . — D e T e W e e R W - B35 20 o 0-.1.2: . ".‘,'-0.'50":'.:.:}" e 7 . . Bal -6;32})6»" - 20 o '“'iz,z-’ .: _ :.'6{36' : 2 7 7 Bl [B-327'7T-’;f’ Coa00 a2 080 i 2 T | 2 7 . Bal B8-3278 20 0.12 - 0.80 175 ANP PROJECT PROGRESS REPORT billets of each composition to investigate the feasibility of producing seamless tubing of the alloys. The tube blanks were extruded at 2150°F at a ratio of 7:1. All compositions were fabricated successfully and sent to Superior Tube Co. for redrawing. International Nickel Company Alloys Tube blanks which were fabricated from air- melted nickel-molybdenum base alloys supplied by the International Nickel Company have been returned by Superior Tube Co. after having been processed to 0.500-in.-OD, 0.035-in.-wall tubing. The results of the processing are given in Table 3.3.1. ' For the most part the limited data indicate that these ' International Nickel Company alloys are amenable to tube-producing practice, with one exception. Alloy T-23015, which contained 0.25% C, could not be successfully reduced to tubing. Although the failures from these particular tube blanks were not examined metallographically, it is felt that carbide stringers, similar to those shown in Fig. 3.3.2, were again the cause of the difficulty., Thermal-convection loops are to be - constructed from the various lengths of tubing received. ORNL Alloys Extrusion experiments outlined previously! for producing seamless tubing of ternary alloys con- taining nickel, 15 to 20% molybdenum, and a third element were continved, with the intention of determining the amount of a particular strengthening element which can be tolerated without serious effect on the corrosion resistance of the base composition to fuel No. 107. The tube blanks which were fabricated during the quarter and submitted to Superior Tube Co. for redrawing are listed in Table 3.3.3. The extrusion of these TABLE 3.3.3. ORNL ALLOYS EXTRUDED FOR TUBING FABRICATION AND CORROSION EVALUATION Alloy No. Nominal Composition (wt %) Number of Extruded ' Tube Blanks 30-6 17 Mo—=0.06 C~7 Cr—bal Ni 2 30-7 17 Mo~0.06 C=2 Al-bal Ni 3 30-17 17 Mo—0.06 C—4 Al-bal Ni 2 30-8 17 Mo~0.06 C~2 Ti~bal Ni 3 30-18 17 Mo=0.06 C—4 Ti-bal Ni 3 30-9 17 Mo~0.06 C—2 W-bal Ni 3 30-19 17 M0~0.06 C~4 W=bal Ni 3 30-10 17 Mo=0.06 C=2 V~bal Ni 3 30-20 17 Mo—0,06 C—4 V—bal Ni 3 3011 17 Mo=0.06 C—4 Fe—bal Ni 3 30-12 17 Mo—0.06 C—3 Nb—ba! Ni 3 30-21 17 Mo=0.06 C—5 Nb—bal Ni . 30-13 (INOR-3) 16 Mo=1.5 Ti—1 Al-0.06 C—bal Ni 3 30-14 (INOR-4) 16 Mo~1.5 Ti=2 Al..0.06 C—-ba! Ni 3 30-15 (INOR-5) 15 Mo~2 Nb—2 W~0.06 C—bal Ni * 30-16 (INOR-6) 16 Mo—5 Cra1.5 Ti=1 Al-0.06 C~bal Ni 3 30-22 (INOR-7) 16 Mo—=6 Cr~1 Nb—1 Al-0.08 C~bal Ni * *Extrusion not yet made, 176 o alloys was carried out at 2150°F at a ratio of 7:1. The extrusion techniques described previously were used, and no difficulty was encountered with these ternary alloys, except the 79% Ni-17% Mo—4% Al alloy, which offered relatively high resistance to plastic deformation. This is an indication of the potency of aluminum in increasing the elevated-temperature strength of the base composition, In addition to the ternary alloys, 36-1b vacuum induction melts of the new alloys INOR-3, -4, and -6 were prepared and extruded to tube blanks. The compositions of these alloys and the results of the extrusions are given in Table 3.3.3. The purposes of this series of experiments were to determine the extrudability of these afloys and to gain a supply of tubing for corrosion testing, As indicated in Table 3.3.3, three billets of eoch alloy were successfully extruded. These ex- trusions were carried out at 2150°F at a ratio of 7:1. The quantities of tubing obtmned from extruded tube blanks previously submitted to Superior Tube Co. are listed in Table 3.3.1. In general the redrawing of the chromium-bearing alloys with intermediate carbon contents (0.06% or less) proved to be satisfactory. The percentage yields of 0.500-in,-0D, 0.035-in.-wali tubing from con- ditioned tube blanks, as well -as the total length of tubing of each composition, are also shown in Table 3.3.1. This tubing is to be fabricated into thermal-convection loops for corrosion testing. Hastelloy W Seamless Tubing Results of processing a tube blank of Hastelloy W, which was successfully extruded at ORNL, to 0.187-in.-0D, 0.025-in.swall tubing indicate the feasibility of producing small-diameter seamless - ~ tubing of Hastelloy-type alloys for heat exchanger application. Ayield of 25 ft of tubing was obtained from this extrusion. Longitudinal and transverse 7 sections of the tube wall are shown in F:g. 3.3.3. Consumable-Elechode Expenments Electrodes of mckel and four nickel-base ulloys,' ‘_ -~ 83% Ni-17% Mo, 76% N|—l7% o-7% Cr, Hastelloyv"_ ‘B, and Hostelloy W, which were submitted to Battelle Memorial Institute for consumable-elecs trode arc melting, have been melted and returned to ORNL for evaluation. Suitable test specimens will be fabricated from the ingots to determine PERIOD ENDING SEPTEMBER 10, 1956 whether arc-melting is mstrumental in |mprovmg the properhes of these alloys. SHIELD PLUG FOR ART PUMPS J. H. Coobs J. P. Page Gamma-Ray Shielding Material The thermal conductivities of the six tungsten carbide—constantan specimens described in the previous report? were determined, and the data are presented in Table 3.3.4. A plot of these data as functions of the volume percentages of tungsten carbide, constantan, and pores is presented in Fig. 3.3.4. Examination of this graph shows the thermal conductivity of this system to be primarily a function of the tungsten carbide content, with only minor effects resulting from variations in porosity. These specimens bracket quite completely the composition range at a bulk density of 12.0 g/cm?® ond indicate fairly conclusively that the specifications for the gamma- ray shielding materia! (density, 12.0 g/cm3 ~ mini- mum; conductivity, 0.10 cal/cm.sec:°C —~ maximum) cannot be met by the tungsten carbide—constontan material. A search for a lower conductivity nickel-base ~alloy for use as a binder resulted in the selection of Hastelloy C. This alloy has a reported conduc- tivity of 0.03 cal/em.sec-°C, as compared to 25, H. Coobs and J. P, Page, ANP Quar, Prog. Rep, ]une 10, 1956, ORNL-2106, p 173. TABLE 3.3.4. THERMAL CONDUCTIVITY OF EXPERIMENTAL COMPOSITES OF TUNGSTEN CARBIDE AND CONSTANTAN : Composition Densify o (fi' %) (g/em®y — Thermal Conductivity —_— e {eal .sec®C wC _Constanfun (cal/em see ) 153 57 4o N 7 29 012 0 85 15 012 | 1:1.1'.'55. w6 | 0.12 1296 79 21 0.13 1267 93 7 0.15 177 ANP PROJECT PROGRESS REPORT Fig. 3.3.3. Longitudinal (a) aond Transverse (b) Sections of 0.187-in.-OD, 0.025-in.-Wall Seamless Hastelloy W Tubing. 178 ge UNCLASSIFIED ORNL-LR-DWG 15034 ~ TUNGSTEN {0 20 30 40 CARBIDE CONSTANTAN {(vol %) - Fig. 3.3.4. Thermal Conductivity (cal/cm-sec.°C) of Hot-Pressed Tungsten Carbide~Constantan Gamma-Ray Shielding Material as a Function of Composition and Porosity. 0.06 cal/cm-sec:°C for constantan. Also, pre- alloyed powder is commercially available at a reasonable price (approximately $5/1b). Four tungsten carbide~Hastelloy C thermal- conductivity specimens were hot pressed, ma- chined, and tested. The results are shown in Table 3.3.5. These data are plotted in Fig. 3.3.5, again as functions of the three composition parameters: tungsten carbide, Hastelloy C, and porosity. It is evident that the specifications for the gamma-ray shielding material are satisfactorily fulfilled by this material. The 70 wt % tungsten carbide~30 wt % Hastelloy C specimen was considered to be the easiest to hot press ond was tentatively chosen as the material for the ART shield plug. Two models of the shield plug have been hot pressed from this 70 wt % tungsten carbide materlal that hove the followmg dimensions and densi :ty Model'l Model 2 : _ ' lnslde dwmeter, ine . 1495 ‘_727‘.3762_ _ Outside dlameter, in. 2.229 3-860 7 | | . ‘He-ghf..m,- o 1aas 1740 Densityygfen® S0 a5 These models are not true MII'IIO'I'UI'eS of the - plugs, but they have the same ratio of wall-surfoce-i_ area to- cross-secflonoi area as. the shle!d plugs,'._-' this ratio is on important variable in the hot- pressing operation. Neither of these models was PERIOD ENDING SEPTEMBER 10, 1956 TABLE 3.3.5. THERMAL CONDUCTIVITY OF EXPERIMENTAL COMPOSITES OF TUNGSTEN CARBIDE AND HASTELLOY C , Composition Density Thermal Conductivity 3 (wt %) 0 (g/em?y —M8M ———— (cal/cmesec+"C) WC Hoastelloy C 11.80 92 8 0.08 1.99 81 19 0.07 13.45 90 10 0.13 1190 70 30 0.07 UNCLASSIFIED ORNL-LR-DWG 45030 TUNGSTEN 10 20 30 40 R CARBIDE HASTELLOY G (vol %) Fig. 3.3.5. Thermal Conductivity (cal/cm.sec-°C) ~ of Hot-Pressed Tungsted Carbide~Hastelloy C Gamma-Ray Shielding Material as a Function of Composition and Poresity. _difficult to press, although an alumina coating -~ on the grcphlte die is required to prevent a - graphite-nickel reaction at the. pressing temper- oture. Unless severe temperature gradients are ~ ~encountered in the larger die to be used for the 'ART plug, it should be possnble to hot press the ~ plug as one piece. ‘A secondary specification is that the material " be readily brazed to Inconel. The 70 wt % tungsten cnrblde material sahsfactorlly meets this re- - quirement. - Copper readlly wets the’ composate, and the coefficients of expansion of the composite ~and of Inconel appear to be similar. Several small fest pieces. and the two models have been suc- cessfully copper-brazed to Inconel plate. 179 ANP PROJECT PROGRESS REPORT Thermal Shield Three specimens of ZrO, were fabricated at ORNL and sent to Bottelle KAemorial Institute for thermal conductivity determinations. Results for the first two specimens tested are presented in Teble 3.3.6. TABLE 3,3.6. THERMAL CONDUCTIVITY OF ZrO, SPECIMENS Thermal Conductivity Test (cal/cmesecs®C) Tem(p: (r:t;fure Specimen with Specimen with Density of Density of 3.52 g/cm’ 3.08 g/em® 20 0.00251 0.00163 100 0.00244 0.00158 200 0.00235 0.00153 300 0.00225 0.00146 400 0.00218 0.00141 500 0.00211 0.00134 600 0.00203 0.00129 700 0.00196 0.00122 800 0.00192 0.00115 NEUTRON SHIELD MATERIAL FOR THE ART J. H. Coocbs M. R. D’Amore3 Cercmic B4C Tiles The configuration of the neutron shield for the - ART was described previously.# The contract for fabrication of the ceramic B,C tiles for the ART was awarded to the Norton Company during the quarter. The tiles will be fabricated by hot- pressing high-purity B ,C powder to a minimum guaranteed density of 1.7 g of boron per cubic centimeter, Clad Copper-B «C Cermets The Allegheny Ludlum Steel Corp. is being considered as a potential supplier of the 0.100-in.- 30n assignment from Pratt & Whitney Aircraft, “M. R. D*Amore and J. H. Coobs, ANP Quar. Prog, Rep. March 10, 1956, ORNL-2061, p 151. 180 thick type 430 stainless steel—clad copper-B4C (6.6 wt % B C) neutron shield material. A plate of this material that was fabricated by Allegheny Ludlum Steel Corp. was evaluated, The plate was approximately 7% x 23 in. and 0.100 in. thick, The evaluation included examination of the- surface finish, radiography, bend fests, tensile tests, thickness uniformity measurements, core density measurements, microstructure examinations, and investigation of the integrity of the clad«to- core bonding. No pinholes were evident in the 10-mil-thick cladding stripped from the core material, although the surface finish on the cladding was rough. No.segregation of the B,C or cracking of the core was observed, and the clad was well-bonded to the core material. The density of the copper-B 4C cermet, as measured by the water-displacement method, was 97.9% of theoretical. Small specimens that had been cut parallel and transverse to the rolling direction were bent to a radius of curvature of %, and 27, in., respectively, before fracture occurred, The plate tapered along the length from a maximum thickness of 0.099 in. to a minimum thickness of 0.092 in. The dispersion of B C particles in the copper matrix was excellent. The room-temper- ature tensile strength data for the material are presented in Table 3.3.7. The tensile specimens were punched out with a blanking die and were 5 in. long with a 2-in, gage length. The as-punched tensile specimen showed no elongation and failed at the yield point of the material, Specimen 4 was prepared by stripping the cladding from the copper-B 4C core prior to punching the specimen. : LITHIUM-MAGNESIUM ALLOYS R. E. McDonald? Efforts to develop a 20% Li-80% Mg alloy for shielding application were continued, with the possibility of roll cladding the clloy being investi- gated further. As reported earlier,® 25 aluminum forms a brittle intermetallic compound with the alloy, and, in an effort to overcome this difficulty, nickel foil was used as a diffusion barrier; some bonding resulted. Work will be continved with barrier materials during the next quarter. 50n assignment from Pratt & Whitney Aircra.fl.— 6R. E. McDonald and C. F, Leitten, Jr., ANP Quar, Prog, Rep. June 10, 1956, ORNL-2106, p 173. ke < g PERIOD ENDING SEPTEMBER 10, 1956 TABLE 3.3.7. ROOM-TEMPERATURE STRENGTH OF TYPE 430 STAINLESS STEEL CLAD COPPER-B,C CERMETS Specimen ield t Vield Strength at o sile Strength Elongation No. Orientation Condition 0.2 % C')ffsef (psi) % in 2 in.) (psi) 1 Trunsvefse torolling As punched 24,500 24,500 0.0 direction 2 Transverse torolling Annealed 20 min at 20,000 31,200 5.6 direction 1650°F 3 Parallel to rolling Annealed 20 min at 19,000 34,000 7.5 direction 1650°F ' 4 Core material unclad; Annealed 20 min at 15,100 21,400 3.5 transverse to rolling 1650°F direction Creep and stress-rupture data obtained for this lithium-magnesium alloy are reported in Chap, 3.5, *‘Mechanical Properties Studies.”” Further work is being done on the powder-metallurgy approach to the preparation of this shield material. Lithium orthosilicate is being tested as o substitute for lithium oxide, since it is less reactive in air and water and easier to handle. However, it has only one-half the lithium content of lithium oxide. TUBULAR CONTROL RODS M. R. D’Amore The feasibility study of extrusion of tubular control rods, described previously,” was con- tinved, A simulated control rod billet consushng , of a type 316 stainless -steel can with a core composed of 34 vol % AI,‘,O3 (to simulate Lindsay “oxide) m nickel powder was exfruded at 2150°F - to a1} -m.-OD /-m.-lD “tube. " The ‘billet was - .deSIgne to determme the feasubtllty of extrudmg; o 'thm (obout 0. OSO-m.-thlck) inner and outer claddmgr’ on - control -rod cores ‘and of prepormg cores by - -fampmg the Ioose powder mixiure into the blllet{'r' L followmg ranges' I to 3 e 1to 200 3 ‘and 52 to 200 .. CGHS. : '_ ‘,*;,'7. Evaluation- of the extruded tube showed flmt the_ -' . 'cloddmg thlckness was . sotlsfoctory. ‘However, - 'tompmg of the 1oose powder does not appear to be' , L 7 T —,'I-kg somple of the - Lmdsay omde .was converted " into latger particles by pressing the as+eceived -7y, H. Coobs, R. E. McDonaH, and M. R. D'Amore, AI;JP3 Quar, Prog, Rep. March 10, 1956, ORNL-2061, p 163. promising as a core-preparation method., Trans- verse sections of the extruded tube showed the core cross section to be fairly uniform, except in one area. The nonuniform area can be atiributed to the sintering and concomitant shrinkage of the core during the preheating operation, which can cause longitudinal cracks in the core as a result of shrinkage around the inner cladding layer or the formation of void areas between the core and the outer cladding layer. The void areas can, in turn, ~ cause buckling of the outer cladding layer during extrusion, A Hastelloy X billet containing a hot-pressed core of 30 wt % Lindsay oxide=70 wt % nickel | ~has been prepared for extrusion, Tensile speci- ~.mens are being machined from plates of 30 wt % ~ Lindsay oxlde-—70 wt % nickel cermet clad with inconel, which were fabncated by hot rolling and 5‘_usmg the p:cture-frame techmque. - The specimens - will be used to investigate the effect of the 2 Lmdsuy oxide particle size on the elevated- - ‘temperature ‘tensile strength of the cermet. The oxide particle sizes Beung mvestlgofed are in the -As-received Lmdsay oxtde is @ mixture of very fine -particles that are about 1 to 3 g in size. A powder to a low-density compact and sintering at 2460°F in air. The sintered compacts were then 181 ANP PROJECT PROGRESS REPORT crushed ond ground in @ micropulverizer to attain the desired particle size. - Thermal conductivity data were obtained for a 30 wt % Lindsay oxide~70 wt % nickel specimen at temperatures between 165 and 554°C, and the results are shown in graph form in Fig. 3.3.6. -The slopes and intercepts of the two lines were calculated from test results by using the least- squares methed. " SEAMLESS TUBULAR FUEL ELEMENTS M. R. D’Amore Two three-ply blanks containing cores of 30 vel % Al,O; (to simulate UQ,) dispersed in type 302B stainless steel powder were redrawn from 1.0-in.-0D, 0.125-in.-wail tubing to 0.187-in.-OD, 0.015-in.-wall tubing at the Superior Tube Co. The inner and outer layers of the three-ply composites were type 316 stainless steel. The finished tubing and samples taken at various ~'stages in the reduction schedule have been re- ceived and evaluated. The first extruded tube, for which the core was prepared by hot pressing a mixture of <325 mesh particle size Al .05 and stainless powder mixture, exhibited tenslle g-actures in the core after 30% total reduction. The finished tubing had severe fractures in the core, The second extruded tube blank was prepared by tamping the powder mixture (=105 +325 mesh Al,O, in type 302B stainless steel) into the 0.40 billet can. Examination of the redrawn tubing did not reveal any tensile fractures in the core. How- ever, the relatively large particles of A|203 were fractured, and they tended to form stringers that finally resulted in longitudinal cracking of the core as the total reduction values increased. Radio- graphs of the finished tubing revealed the cere fractures in tube No. 1 but did not show the longitudinal cracks in the core of tube No. 2. It is evident from the examinations of the re- drawn’ tubing that the core containing the larger Alzo3 particles was stronger than the core with the fine Al o particle dispersion. Tamping of a loose powdzer mixture into a billet can therefore appears to be a satisfactory method of preparing cores of 30 vol % Al,O, dispersed in 70 vol % of type 302B stainless steel The study of flow patterns in three-ply extruded tubes containing cermet cores was continued with the extrusion of two three-ply billets to 1Y-in.-OD, 3{‘-in.'lD tubes at 2150°F. The billet can material was type 316 stainless steel, and the cores were fabricated by hot pressing @ powder mixture of type 304 stainless steel and 30 vol % Al,O,. A 29-deg taper was machined into the ends of the core of one billet in an effort to eliminate end defects in the extruded tube and thus to improve material recovery. This taper resulted in a slight reduction in the length of the defect. The second billet contained a core made in three sections to determine whether nonuniform areas NPT TR ORNL-LR-DWG 15029 0.09 / o 9 '-._\ [ X (cal/cm-sec-"C) o o o 8 + 0.03 0 50 100 150 200 250 300 350 400 450 500 550 €00 650 700 750 800 TEMPERATURE (°C) Fig. 3.3.6. Thermal Conductivity of a 70% Nickel-30% Lindsay Oxide Specimen. 182 * would appear at the interfaces between core sections, The extruded tube is being sectioned for examination, NIOBIUM FABRICATION STUDIES V. M. Kolba® 1. P. Page 7 Evaluation of Ar-c-Melli'e'd 'N‘i_éb'ium H. Inocuye The physical and mechanical properties of arc-melted niobium are being investigated in order to compare this material with niobium prepared by conventional powder-metallurgy techniques. Macro- etching of a section of an arc-melted, cast, and cold-worked ingot has revealed the large-grained, ‘severely distorted structure. that is usually ob- tained through such a fabrication process. An ottempt to cold roll a section of as-wreceived y.in. plate was terminated after two 3% passes when some end cracking was noticed. The cracked ends were cut off, and the re- maining material was annealed 1 hr ot 1100°C at @ vacuum of 2 x 10=5 mm Hg. This 0.485-in.-thick section was then successfully cold rolled to 10-mil strip, a reduction in thickness of about 98%. The hardness increased rapidly from 160 VHN (2.5-kg load) to 190 VHN in the first few passes, and then it increased very siowly to 206 at 98% reduction in thickness. Samples were taken at several stages in the reduction, split into thirds, and annealed for Y hr at 850, 1050, and 1150°C in high vacuum, 'fi'\ese samples will be examined metallographically to determine the amount of reduction necessary to produce fme- grained materloi after annealung. - Nb-UO Compocts e %on as’si'game'm from The Glenn L;- Mdrtih‘Co."_'_:‘ s PERIOD ENDING SEPTEMBER 10, 1958 somewhat contaminated, The analytical results are presented in Table 3.3.8. The Nb-3-C batch was chosen to make five 70 wt % Nb-30 wt % UO, compacts for sintering and reaction studies. The UO, used had previously been high-fired in a hydrogen atmosphere. The compacts were 82% dense, as compacted, and had good ‘‘green® strength. These compacts are being sent to GE-ANPD for sintering at 2000°C. Attempts to secure high-purity powder from commercial sources have been unsuccessful, Nb-U Alloys A finger melt of an 80% Nb~20% U alloy was made. Segregation noted in the initial casting was ~ reduced by subsequent remelting. The hardness of the as-cast sample was R 86. Attempts to cold roll the as-cast material proved to be unsuccessful. A sample of the alloy was then encapsulated and hot rolled successfully at 1050°C to a reduction of 85%. Metallographic examination of the hot- rolled structure revealed that the as-cast structure had not been broken up during the rolling process. The as«olled hardness was R 95. Vacuum-annealing tests are bemg conducted on - the hot-rolled alloy at 1100, 1200, and 1300°C to determine the recrystallization temperature of the clloy. Further rolling studies will then be made. FABRICATION OF HYDRIDES R. E. McDonald A survey of available literature on the fabri- cation of hydrides has been completed, and @ - hydriding furnace and purification system is being _ Two bc:tches of moblum powder made from : o wroughf sheet were ana lyzed and were found to be _ built ‘(see Chap. 3.6, ‘‘Ceramic Research), - Yttrium and zirconium and their alloys will be -hydnded and fabricoted, and the compatibility of various barrier materials will be studied. TABLE 3.3.8. IMPURITIES FOUND BY ANALYSES OF TWO SAMPLES OF NIOBIUM POWDER lmpurifies (wt %) . ""’S:dm-prorle:‘f L = - - Hordness .c 0, Ny - H, ~ (VHN) NB2-C 0042 029 0018 s2x 107t 292 Nb+3-C 0.103 0.18 0.019 5.8 x 1074 179 183 ANP PROJECT PROGRESS REPORT 3.4. WELDING AND BRAZING |NVESTIGATIONS ' P. Patriarca FABRICATION OF PRIMARY NaK PUMP VOLUTES P. Patriarca E. J_; Wilson . Preliminary experiments designed to determine the approximate weld shrinkage to be expected in the fabrication of primary NaK pump volutes were conducted, and the results were reported.! (The primary NaK pumps are the pumps for circulating NaK in the ART fuelsto-NaK heat exchange system,) The information obtained from these tests has now been used to successfully fabricate three Inconel volutes, one by the metallic-arc welding procedure and two_ by inert-arc weldmg procedures. The shrinkage of volute-No, 1, fabricated by ‘using the metallicearc, (coated-electrode) welding - procedure, compared favorably with that of simu- .lated test components, -However, radiographic inspection of the weld revealed porosity to an ‘unacceptable degree. The exclusive use of inert- arc welding was therefore specified for the ~ joining of future volute halves in order to ensure ~acceptable quality, 1p. Patriarca and G, M. Slaughter, ANP Quar, Prog, Rep. ]une 10, 1956, ORNL-2106, p 176. UNCLASSIFIED ORNL—LR—DWG 16166 ' PASS ELECTRODE ELECTRODE NUMBER ~ SIZE {in) - MATERIAL CURRENT (amp) ™ ¥ in. INCO 62 . 90 2 33 in. INCO 62 140 3 g in. INCO 62 170 4-8 g in. INCO 62 190 * ROOT PASS: 6 TACKS, EACH APPROXIMATELY Zin. IN LENGTH, FOLLOWED 8Y TIE-INS BETWEEN TACKS. Fig. 3.4.1, Procedure and Joint Design for the Inert-Arc Welding of Primary NaK Pump Volute Weld Shrinkage Test Pieces (No, 1). Test pieces rotated in horizontal position. 184 It was recognized that the magnitiude of the weld shrinkage would differ sugmflcantly for the two welding processes, and, since the existing data were not adequate for predlchng the dimensional changes, an additional shrmkage test was con- ducted, © Two test pieces were machined from 2-in. Incone! plate, as described previously,! and welded - in accordance with the procedure described in Fig. 3.4.1. Micrometer measurements were made at four radial sections, as described in Fig. 3.4.2, prior to and after each subsequent_ operation. The completed test piece is shown in Fig. 3.4.3. The results of the micrometer measure- ments are summarized in Table 3.4.1. These data provided a first approximation of the shrinkage to be expected when welding of volute Ne. 2 was undertaken. - A 0.125-in. shrinkage allowance was incorporated into the joint design and, after machining, the volute halves were scribed as shown in Fig. 3.4.4. . Micrometer UNCLASSIFIED ORNL - LR~ DWG 16167 Rl Fig. 3.4.2. Details of Micrometer Measurements on Welded Primary NaK Pump Veolute Weld Shrinkage Test Pieces, Measurements made at four radial sections (A through D) at 90-deg intervals at positions 1, 2, and 3 (see Table 3.4.1). o : ‘ S e PERIOD ENDING SEPTEMBER 10, 1956 UNCLASSIFIED PHOTO 17788 | F|g. 3-4.3.,‘. Completed Inert-Arc-Welded Primary NaK Pump Volute Weld Shrinkage Test Piece, measurements were made before and after each operation at each intersection shown, ‘The volute halves are shown dséembled for ‘welding in Fig, 3.4.5. The welding procedure used ‘was. essenflally fhe same ds that shown in : "except'f fl\at .as ‘many’ as six addrtionai passes were - requnred to compiete the weld at the - ‘volute . exit where ‘the section ‘was abouf 1 ing fhlck" The complefedvolufe is shown in Flg. 3 4.6 The results ~_:of fhe mlcrcmefer measurzments are summarlzed in Tnble 34 2. It may be noted that ‘the’ shrlnkuge was somewhat less than that cleslred ‘The velute ‘entrance dlmensmn ns consldered to be “the mosf“cnhcci dlmensmn in-the pump, and, after an analysus of the daia, ‘a shrmkoge ul- lowance of 0.115 in. was selected for incorporation into the joint design of volute No, 3. ' The volute halves for volute No, 3 were ma- -chined, assembled, welded and annealed, with the IWeidmg procedure agam being similar to that ~given on Fig, 3.4.1. Micrometer measurements ~were made before - ond after each important oper- aflon, und the data are summcnzed in Table 3.4.3. It may be noted that adjusting the shrinkage uflowunce to-0.115 in, achieved a close approach. ~to the dimensional requirements ~of the -volute entrance.- A further refinement 'is “planned for ~volute No. 4 in that a shnnkage allowance of 0.110'in, will be mcorporated into the joint design, Any shrmkuge of ‘the ‘volute entrance to less than 70,650 in, will be’ prevented by the use of Inconel spocers where needed, - 185 ANP PROJECT PROGRESS REPORT UNCLASSIFIED ORNL-LR—-DWG 16t68 180° 160° VOLUTE ENTRANCE, A RADIAL POSITION C J ‘ 60° EXTERNAL BEVEL EDGE, RADIAL POSITION A _ — VOLUTE CENTERLINE, RADIAL POSITION B 340° 0° l— T> i (D) I e 0 =i SECTION A-A Fig. 3.4.4, Details of Micrometer Measurement on Volute No. 2. 186 i i i i i i ! ! i i PERIOD ENDING SEPTEMBER 10, 1956 UNCLASSIFIED PHOTO 17583 A UNCLASSIFIED PHOTO 17879 Fig. 3.4.6. Yolute No, 2 After Inert-Arc Welding and Annealing, 187 i § i - ANP PROJECT PROGRESS REPORT TABLE 3.4.1. INERT-ARC WELDED PRIMARY NaK PUMP YOLUTE WELD SHRINKAGE DATA FOR TEST NO. 1 Dimensions (in,) Position Before After After - Net Welding -Welding Change Annealing** Change Alr 0.366 0.244 0.122 0.240 0.126 B1 0.382 0.254 0.128 0,248 0.134 C 0.388 0.264 0.124 0.259 0.129 D1 0.378 0.249 0.129 0.244 0.133 A 2% 0.377 0.259 0.118 0.254 0.122 B2 0.381 0.260 0.121 0.254 0.127 c2 0.388 0.268 0.120 0.264 0.124 D2 0.384 0.263 0.121 0.257 0.127 A3 0.468 0.349 0.119 0.346 0.122 B3 0.469 0.348 0.121 0.343 0.126 C3 0.469 0.349 0.120 0.345 0.124 D3 0.469 0.348 0.121 0.344 0.125 *Add 3.317-in, micrometer correction to readings shown at positions 1 and 2, **Annealed at 1950°F for 2 hr and furncce cooled at a rate of 500°F /hr. TABLE 3.4.2. RESULTS OF MICROMETER MEASUREMENTS OF WELDED YOLUTE NO, 2 Angular Position Dimensions (in.) 188 Before After After Total b (deg) Welding Welding Annealing? Shrinkage Deviation At Radial Position AC _ 0 0.754 0.646 0.648 0.106 -0.019 200 0.765 0.649 0.647 0.118 '-—o.o_oj 220 0.771 0.655 0.652 0.119 -0.006 240 0.758 0.641 0.639 0.119 ~0.006 260 0.763 0.642 0.641 0.122 -0.003 280 0.763 0.639 0.639 0.124 ~0.001 300 0.777 0.654 0.654 0.123 -0.002 320 0.778 0.653 0.654 0.124 -0.001 340 0.773 0.644 0.647 0.126 +0.001 PERIOD ENDING SEPTEMBER 10, 1956 TAB LE 304.2. (con'I flued) Dimensions (in.) Angular Position : l (deg) Before After After Tota b Welding Welding Annealing? Shrinkage Deviation T At Radial Position B¢ 0 0.804 0.690 0.689 0.115 ~0.010 20 0.771 0.665 0.665 0.106 ~0.019 40 0.802 0.696 0.694 0.108 ~0.017 60 0.787 0,675 0.674 0.113 -0.,012 80 0.778 0.664 0.662 0.016 -0.009 100 0.774 0,659 0.656 0.118 ~0.007 120 0,771 0,655 0.653 c.118 ~0.007 140 0.779 0.664 0.661 s -0.007 160 0.773 0.656 0.654 0.119 ~0.006 180 0.766 0,650 0.649 0.117 ~0.008 200 0.779 0.665 0663 0.116 ~0,009 220 0.777 . 0.668 0.663 0.111 -0.014 240 0.764 04649 0.648 0.116 ~0.009 260 077 0,654 0,652 0.119 ~0.006 280 0,776 04659 0.657 0.119 ~0.006 300 0.780 0.664 0.663 0.117 ~0.008 320 0.784 0,668 0.667 0.117 ~0.008 340 0.794 0.674 0,673 0.121 ~0.004 Af Radicl Position Cd‘ -0 - oo 0.780 - - 04675 EERE 0.674 0.106 +0.014 200 . 0781 . 0478 0677 . 0.104 +0.017 40 0784 0681 - 0680 . -0J04 +0.020 60 - 0785 . 0679 0674 ~eamn - +0.014 80 . o785 0678 . 0679 . 006 +0.019 100 - o784 0675 0475 - 0109 +0.015 120 . 0783 D674 . 0674 009 - +0.014 Mo - 0781 0673 0673 - 0108 . - +0.013 160 0780 0673 0472 Cod0s 0 +0.012 180 o781 0673 0673 0108 - +0.013 200 0781 . 0675 0673 - 0108 . +0.013 220 0782 0476 04674 . 0008 +0.014 240 0782 0476 0.674 0.108 +0.014 189 ANP PROJECT PROGRESS REPORT TABLE 3.4.2. (continued) Dimensions (in,) Angular Position (deg) Befo:'e ' Afte.r Afte'r . - T.otal Deviation? Welding Welding Annealing Shrinkage At Radial Position C? 260 0.782 0.675 0.674 - 0,108 7 - +0.014 280 0.782 0.674 0.674 0.108 +0.014 . 300 0.782 0.674 0.673 0.109 : -.1-70.013 320 0.781 0.673 0.673 0.108 +0.013 340 0.781 0.673 | 0.674 0.107 +0.014 2Annealed at 1950°F for 2 hr and furnace cooled at a rate of 500°F /hr. o bAt radiol positions A and B the deviation is given as the deviation from the ideal change of 0.125 in. At radial position C the deviation is given as the deviation from the desired dimension of 0.660 X 0.010 in. €Add 3,317-in. micrometer correction to readings shown, | Dimensions shown are absolute micrometer measurements., TABLE 3.4.3. RESULTS OF MICROMETER MEASUREMENTS OF WELDED YOLUTE NO. 3 Dimensions (in.) Angular Position (deg) Befo.re Afte'r Af‘l'e.f . T.ofol Dev iotionb Welding Welding Annealing Shrinkage At Radial Position AS 0 0.741 0.597 0.597 0.144 +0.029 180 0.723 0.604 0,602 0.121 +0,006 200 0.723 0.604 0.601 0.122 +0.007 220 0.719 0.598 0.596 0.123 +0.008 240 0.716 0.593 0.593 0.123 +0.008 260 0.719 0.594 0.593 0.126 +0,011 280 0.725 0.597 0.596 0.129 +0.014 300 0.737 0.604 0.606 0.131 +0.016 320 0,730 0.594 0.599 0.131 +0.016 340 0.733 0.594 0.596 0.137 +0,022. At Raodial Pesition B 0 0.741 0.616 0.608 0.133 +0.018 20 0.755 0.651 0.645 0.110 ~0.005 40 0.780 0.678 0.673 0.107 ~0.008 60 0.770 0.666 0662 0.108 ~0.007 80 0.738 0.629 0.625 0.113 —0.002 190 : L) TABL E 3-4. 3. (Cfln'i nued) PERIOD ENDING SEPTEMBER 10, 1956 Angular Position Dimensions (in,) (deg) Before After After Total Deviation® Welding Welding Annealing® Shrinkage At Radial Position B 100 0.738 0.626 0.624 0.114 ~0.001 120 0.734 0.619 0.618 0.116 - +0.001 140 0.728 0.614 0.610 0.118 +0.003 160 0.728 0.611 0.610 0118 +0.003 180 0.728 - 0.610 0.609 0.119 +0.004 200 0.723 0.606 0.603 0.123 +0.008 220 0.725 0.608 0,605 0.125 +0.010 240 0.728 0.609 04606 0.128 +0.013 260 0.728 0.608 0.606 0.128 +0.013 280 0.733 0.609 0.606 0.127 +0.012 300 0.735 0,609 0.607 0.128 +0.013 320 0.735 0.607 0.603 0.132 +0,017 340 0.735 0.606 0.603 0.132 +0.017 o At Radial .Position c? _ 0 0,773 0.663 0.656 0.117 ~0.004 20 0.774 0.670 0.662 0.112 +0.002 40 0.775 0.673 0.669 0.112 +0.002 60 0775 0.672 0.671 0.104 +0.009 80 - 0775 0.671 0.671 0.104 +0.011 100 To774 | 0.68 0,668 0.106 +0.008 120 S o774 osss 0.666 0108 +0.006 140" 0774 - 0.665 0,665 0109 - +0.005 16 0774 - 0.665 - 0.665 0.109 +0.005 10 0774 0665 . 0.663 0.11 +0.003 200 0774 0665 - 0.663 011 40,003 o200 oam o 0.665 0.662 L0.M2 . 40,002 '?,;jf;;r'?&‘l R °774 - ir,:'.:_'jé:'-éefviiji;” | 0.661 03 +0.001 Lo as0s o togm o oses 0.659 01s - -0.001 280 e 0774 0,663 o 058 0016 ~0.002 191 ~ ANP PROJECT PROGRESS REPORT TABLE 3.4.3. (continued) Dimensions (in.) Angular Position (deg) Before After After . Total De-viétio;b Welding Welding Annealing Shrinkage At Radial Position C% 300 \ 0.773 0.660 0.654 0.119 -0.006 320 0.773 0.659 0.653 0.120 ~0.007 340 0.773 0.660 0,653 0.120 ~0.007 “Annealed at 1950°F for 2 hr and furnace cooled at a rate of 500°F/hr, bay radial positions A and B the deviation is given as the deviation from the ideal change of 0,115 in. At radial position C the deviation is given as the deviation from the desired dimension of 0,660 £ 0.010 in. €Add 3.317-in, micrometer correction to readings shown, Dimensions shown are absolute micrometer measurements, FABRICATION OF PRIMARY NaK . PUMP IMPELLERS P. Patriarca G. M. Slaughter Four primary NaK pump impellers have been fabricated by furnace brozing with Coast Metals brazing alloy No. 52. Since slow heating and cooling rates are required to minimize distortion and braze cracking, each impeller brazing cycle entails o furnace time of 10 to 12 hr. As can be seen in Fig. 3.4.7, which shows a typical impeller, the vanes were inert-arc-welded to the housing, where accessibility permitted, to achieve ad- ditional reinforcement, SHRINKAGE OF INCONEL CORE SHELL WELDS P. Patriarca A. E. Goldman Preliminary transverse weld shrinkage tests of Inconel core shell welds were discussed in the previous report.2 Further experiments have now been conducted to study the shrinkage for other thicknesses of plate, The procedure for the shrinkage tests of welds of l/s-in.-_thick plates is described below, and the results are compared with the results obtained for other thicknesses of Inconel plate. , The initial phase of the test consisted of the inert-arc welding of two Y%-in. Inconel sheets, each 6 x 20 in. A 50-deg bevel! with a I'Sz'i“' 2p, Patriarce and A. E. Goldman, ANP Quar. Prog. Rep. June 10, 1956, ORNL<2106, p 184. 192 land was machined on one long edge of each sheet, The sheets were assembled as shown in Fig. 3.4.8 and held against a flat horizontal plate by means of C-clamps. The edges of the assembly were sealed with tape to prevent air leakage, since only the torch gas was used to supply Fig. 3.4.7. Impeller in Which the Vanes Were Both Inert-Arc- Welded and Furnace-Brazed, Completed Primary NoK Pump 40 49 UNCLASSIFIED ORNL-LR-0WG 6469 Y% x 6 x 20-in, INCONEL SHEETS HELIUM INLET TUBE Fig. 3.4.8. Joint Design for Shrinkage Test of | Inconel Plate Welds. backing gas and weld coverage for this test, A l,é--in.-long tack weld was placed near each end of the root, and two other z-in.-long tack welds were equally spaced along the root. The root pass was made, without filler metal, and, after wire brushing and cleaning the weld with acetone, a final pass was made for which 1,16--in. inco No. 62 filler wire was used. A second test was made that was similar to the first, except that backing gas was used to elimi- nate oxidation of the underside of the weld, Neo shrinkage measurements were taken on these plates, since the test plates were made solely to aid the welding operator in adapting his technique. Based upon the plate tests, two tests were made ~with Y%-in. Inconel sheets, each 6 x 33 ihr.,-th_af_ were bent and welded into - lO‘é-in.—dia ‘hoops. A 50-deg bevel with a 'z'z-in. land was machined. on one long edge of each sheet, The first hoop - was placed ‘on the flat, horizontal bed of the ~welding" positioner, ‘and the second hoop was . placed above it. The two hoops were aligned by - placing a large perforated steel band around the “root girth and tightening the band with C-clamps, ‘as shown- in Fig. 3.4.9. Transverse micrometer measurements were then taken at 4-in. intervals - around - the circumference, = Eleven _Z-ih;-l_bng ' tdcrkl_".jw'gids',_f,we'r_é', equally spaced around the circumference, about 3 in, apart., - Helium backing gas was applied through a copper cone, No dressing of the land or feathering of the weld beads was permitted prior to deposition of the root pass, ‘ PERIOD ENDING SEPTEMBER 10, 1956 After the tack welds were made, the steel band was removed, micrometer measurements were again taken, and the assembly was enclosed as shown in Fig. 3.4.10. The interior was purged UNCLASSIFIED ORNL.-LR-DWG {6470 "¢" CLAMPS TEST PIECE PERFORATED STEEL BAND O TEST PIECE Fig, 3.4.9. Method of Assembling Hoops for Welding. UNCLASSIFIED ORNt=LR-DWG 1671 SHEET METAL COVER ELIUM INLET TUBE TAPED IN PLACE TEST TACK ASSEMBLY WELDS NCONEL DISK < TO WELDING POSITIONER Fig, 3.4.10. Hoop Welding Test Assembly After Tack We’dil'lg. 193 e R P e et et e ANP PROJECT PROGRESS REPORT for 10 min, after which the root pass was made. After wire brushing and cleaning the weld with acetone, the final pass was made, with '4 -in, Inco No. 62 filler wire again being used. Féinal micrometer measurements were taken after the disks were removed from the ends of the assembly. Two additional tests were then carried out in an identical manner with Y-in, Inconel sheets, each 6 x 70 in., bent and werded into 22-in.~dia hoops. These hoops were aligned with C-—clamps, rather than with a steel band, The pertinent welding data for the four hoops are listed in Table 3.4.4, and the micrometer measurements are given in Table 3.4.5. The deposited welds were clean and had slight concavity on the inside surfaces. It was found that short tack welds enabled the welding operator to fuse the tack welds into the root pass more readily. For Y-in. Inconel sheets, inert-arc- welded under the conditions utilized in this test, the transverse shrinkage to be expected for loz-in. hoops is 0.050 £ 0.010 in. and, for 22«in, hoops, 0.041 £ 0.010 in. The circumferential shrinkage was less than measurable for both hoop sizes. A summary of the weld shrinkage results for Y . 1/8-, V., and ¥%-in. Inconel sheet is presented in Table 3.4.6. It should be noted that the data are pertinent only under the conditions utilized for these tests, Any change in the variables may cause significant variations in the actual weld shrinkage encountered. TABLE 3.4.4. WELDING DATA FOR HOOP WELD SHRINKAGE TESTS Current Time Torch Gas Backing Gas Filler Rod (amp) (min) (H3/hr) (3 /hr) (in.) Hoop Nc. 1 (10% in. dia) Tack weld 100 8 25 15 Fusion weld 100 13 25 10 Final weld 80 25 25 | 10 110 Hoop No, 2 (10% in. dia) Tack weld 75 7 25 10 Fusion weld 80 n 25 10 Final weld 80 26 25 10 115.5 Hoop No. 3 (22 in, dia) Tack weld 100 13 26 17 Fusion weld 90-95 18 26 17 Final weld 75--80 47 26 17 212.5 - Hoop Ne. 4 (22 in. dia) Tack weld 80 13 26 20 Fusion weld 100 15 26 20 Final weld 80 47 26 20 209.5 194 PERIOD ENDING SEPTEMBER 10, 1956 TABLE 3.4.5. MICROMETER MEASUREMENTS OF HOOP WELDS Tronsver#e Shrinkage (in.) . : Deviation, Average Maximum Minimuym ) . Station to Station Hoop No, 1 (10.5 in, dia) 0.051 0.056 0.044 0.007 Hoop No, 2 ‘ - (10,5 in. dia) 0.050 0.054 0.045 ' 0.009 Hoop No_;. 3 ‘ ‘ - (22 in, dia) 0.041 ' 0.048 0.036 0.010 Hoop No. 4 - " Neote: Longitudinal shrinkage was less than measurable because of the small diameters of the hoops. TABLE 3.4.6. SUMMARY OF HOOP WELD SHRINKAGE DATA Inconel . Transverse Shrinkage {in,) Recommended Sheet H.oog""-r Greatest Design Allowance Longitudinal . Diameter ) . .. Shrinkage Thickness o Average Maximum Minimum Deviation, for Transverse inu/in.) (in.) (l»n.) - Station to Station Shrinkage (in.) (in./in. 14 6 45 0.024 0.026 0.018 0.005 : 10.024 1 0.010 0.00045 % 105 0.051 0.056 0.044 0.009 0.051 £0.010 Not measurable % 22 0.041 0.045. 0.036 0.010 0,041 £0.010 Not measurable % 44 0.120 0.138 0.111 0.011 0.120 £ 0.015 0.002-0.003 ¥ 52 0de8 0184 0.155 0.014 0.168%0.020 0.002-0.003 *Vertical dimension of all hoops was 12 in, EXAMINATIONS OF NaK-TO-AIR RADIATORS AFTER SERYVICE o P. Patriarca A. E. Goldman ~ G. M. Slaughter A 500-kw high-conductivity-fin NdK-fo-efr-roJi- ator, designated York HCF Radiator No, 4, was’ removed from a test stand on March 3, 1956, after 1356 hr of service. This radiator was tested in the temperature range of 1000 to 1600°F, with a temperature differential imposed on the system during 748 hr of the test. This radiator differed from radiators tested previously in that the side plates were removed, the horizontal spacer plates were cut, and the base plate was sliced through the middle parallel " to the air flow. These modifications were among - those suggested in previous reports,3+4 3. 4. Gray and P. Patriarea, Metall&grapbic Ex~ amination of ORNL Radiator No. 1 and York Radiator No. 1 Failures, ORNL CF-55-10-129 (Oct. 31, 1955). Q. J. Gray and P, Patriarca, Metallographic Ex- amination of PWA HCF Radiator No, 2, ORNL CF-56-3- 47 (March 12, 1956). 195 ANP PROJECT PROGRESS REPORT Twenty-four specimens were cut from the radi- ator as shown in Fig, 3.4.11. Each specimen contained a portion of three tubes, with each tube joined to 15 or more fins, The specimens were cross sectioned to expose two opposed joint areas, mounted, and examined at 80X. A total of 2757 joint areas were examined, The percentages of fin-to-tube adherence and the degree of oxidation of the fin collars were noted. UNCLASSIFIED Y-18442 The results of this examination in comparison with the results of examinations of five radiators operated previously are presented in Table 3.4.7. Small cracks were found in several of the tube-to- support plate joints, and evidence of cracking was observed in several tube.‘ ‘deep, 0, 012 in.- wide, and 32 in. long,-and on the outside ‘of another’ tube there .was a small crack : upproxlmately 0.003 in; deep. Agam, trouble was encountered in precisely locating the defects, and there was some smearing durmg metcllogrcphlc polishing. 44 Over 1000 ft of 3/-ln.-OD 10.035-in.-wall, Hastel- on C. tubing .in rundom lengths. was inspected ultrasonlcaily. The rejection rate for this tubing was approxlmately 30%. Metallographic dectioning disclosed the presence of the ‘cracks and other discontinuities -illustrated in Flgs. 3.7.2 through - 3.7.5. A photomacrograph of the inside surface along the weld of weld-drawn Hastelloy C tubing ‘is_shown in Flg. 3.7.2. There are tiny tears along _the -weld to parent metal interface, and there is an -:-;FGpparenf lack of fusion. An internal crack which ~did not extend either to the inside or to the out- side surface of the tube. wall is shown in Fig. 3.7.3. The crack appears to be in the heat-affected zone adjacent to the weld. An unetched sample of Hastelloy C tubing that has two c¢racks is shown 213 0o< ANP PROJECT PROGRESS REPORT Fig. .7.2. lnside Surface of Weld-Druwn Hastel- loy C Tubing. _ in Fig. 3.7.42. One crack is approximately 0,015 in. deep, and the shallower one is about 0.005 in. deep. Etching of this sample, as shown in Fig. 3.7.4b, disclosed that the cracks were actually about 0.020 in. and 0.015 in. deep. Another crack is shown in Fig. 3.7.5¢ and & in the etched and unetched conditions: It may again be noted that the crack, although it is about 0.015 in. deep, does not extend to either surface. Again, difficulty was encountered in the preparation of metallographic sections because the soft matrix smeared over the defects during the polishing operation. However, a greater degree of success was achieved in locating the defects in this ma- terial than in the Inconel tubing, This can probably be attributed to the larger size of the defects that seem to regularly occur in Hastelloy tubing, ~ Fig. 3.7.3. lnternal Cruck in Heat-Affected Zone Adjacent to Weld in Weld- Drawn Hastelloy C Tubmg. Etched with chrome regia. 100X, 214 PERIOD ENDING SEPTEMBER 10, 1956 » " in Hastelloy C Tube. (a) Unetched. (b) Etched with chrome regia. Intergranular Cracks ig. 3.7.4. Reduced 13% F 100X 215 ANP PROJECT PROGRESS REPORT LAY 4 ith chrome regia. 100X, (6) Etched w () Unetched. Fig. 3.7.5. A Crack in Hastelloy C Tubing. Reduced 13.5%. Ces N " @ 216 e e Over 1000 ft of %-in.-OD, 0.035-in.-wali, CX-900 Inconel tubing in random lengths was examined ultrasonically, and a total of approximately 25 ft was rejected. The few defects detected seemed to be very small. In general, the quality of this tubing was quite high, as indicated by the rejection rate of less than 2.5%. Metallographic examing- tions also failed to disclose any defects of appre- ciable size. Of the 229 pieces of 0.229-in.-0D, 0.025-in.=walf, 78-in.-long, CX-900 Inconet tubing also examined ultrasonically, 31 pieces were re- jected, An attempt is being made to compare the de- tection of defects by ultrasound and by other in- spection methods, and, for the larger defects, correlations are normally possible. However, for many of the very small defects, particularly those on the inside of tubing, ultrasound seems to be the only feasible detection methed. It is hoped that, despite the inherent difficulties, metallo- graphic examinations will continue to provide additional information concerning the nature of the defects. INSPECTION OF PIPE Jo Ko Whi'e Ro Bo :Oliver The pipe inspection facility was installed in April, and about 6000 ft of pipe ranging in size from "3’ in. IPS to 7 in. OD has been inspected; most of this material has been Inconel. As a result of this experience only minor changes in the details of the inspection technique previously described? have been made. Because of the ex- cessive camber in large pipe and the end whip frequently experienced in small pipe, the present system of using the pipe as a transducer guide appears to be the only practicel method for main- taining alignment, In this -system a hand-held - search-tube positioner is used with a series of interchangeable side plates cut to fit each plpe size. The positioner is translated along the pipe by hand, and @ moter-driven string is used as @ speed guide (Fig. 3.7.6). The positioner (Fig. 3.7.%) con . be adjusted to change the transducer-to-pipe distance (delay), the rotation of the transducer in the plane perpendicular to the sound beam, and the angle of incidence of the beam upon the outside surface of the pipe. The angle of incidence de- 2), K. White, ANP Quar, Prog. Rep. June 10, 1956, ORNL-2106, p 216. 644 PERIOD ENDING SEPTEMBER 10, 1956 Fig. 3.7.6. Facility for the Inspection of Pipe by the Immersed Ultrasonic Method. termines the depth of inspection of the pipe wall., With materials of the same acoustic impedance as Inconel and approximately the same wall-thickness- to-diameter ratio as sched-40 pipe, the usaoble range of angles is from 10 to 30 deg. Incident angles of less than 10 deg cause interference - because of reverberations across the pipe wall. Angles greater than 30 deg set up Rayleigh (surface) _waves3 and occentuate scraiches selectively more than deeper defects. Within these limits the inci- dent angles used are chosen to balance the indi- _cations from 5% notches on the mslde and outslde surfaces. Internal defects show up strongly when angles of about 10 to 17 deg are used, External ,_defects show up strongest ‘when angles of about ,23 to 90 deg are used. " Rotation of the transducer in the plane perpen- dtcular to the 'sound beam is sometimes necessary 3E, G. Cook and H. E. Van Valkenburg, ASTM Bull 198, May 1954. 217 ¢oo ANP PROJECT PROGRESS REPORT UNCLASSIFIED PHOTO 17936 . TV - ‘ mfr"\:‘\"‘" :‘B“ }www.w} Fig. 3.7.7. Ultrasound Search Tube Mounted for the Inspection of Pipe. because the transducers presently available do not emit energy uniformly. Apparently, uniform crystal damping is difficult to achieve and poorly damped areas, or ‘‘hot spots,’”’ are common. The purchase of a square transducer is contemplated to minimize this “hot spot’” effect and to achieve higher coverage per revolution and hence greater scanning speeds. Heretofore, experience had been limited to in- spections made at a frequency of 5 Mc, but for thick-walled pipe, 2.25 Mc appears to be a promis- ing frequency, as evidenced by some recent success. More observations will be made ot this frequency in the near future, It has been observed that, in general, pipe quality has improved since the inspection program was initiated. Several entire pipe orders were rejected 218 ~ during the early stages of the inspection program. During the lost few months the special core exer- cised in preparing the pipe, crating it for protection during shipment, etc., have resulted in marked improvement in quality, as evidenced by numerous pipe orders inspected without detection of a single defect, INSPECTION OF THIN SHEET J. W, Allen R. W, McClung- R. B. Oliver o Sheet material has extensive application in power reactors, and, in many cases, formed sheet will function both as a container for fluids and as a heat transfer medium between two fluids. Since high heat fluxes and thermal stresses are involved, the presence of laminar defects is very undesirable. The problem of detecting such defects in sheet has, currently, only one solution, which, in turn, poses a very difficult mechanical problem. This situation has motivated an investigation to develop a similor and betfer inspection method. Liquid penetrant, radiographic, and eddy-current methods . cannot be used because of the unfavorable defect orientation. The ultrasonic resonance method is not an adequate approach, since it is not capable of resolvingsmall defects and it inherently requires slow contact scanning of the sheet. The conven- tional pulse-echo vitrasonic inspection, even with a pulse length as short as 1 usec, is not practical for sections thinner than 0,20 in. The transmission- attenuation technique is capable of detecting small laminations in thin sections, but, since it requires critical alignment of two transducers on opposite sides of the sheet, it is a very difficult method to apply to the inspection of large or nonplanar areas. A new ultrasonic method has been proposed, and preliminary experiments have given promising results, The new method requires a pulse of ultra- sound having o duration of 5 to 20 psec, with the sound tuned to such a frequency that the sheet thickness is an exact multiple of the half wave- length, With these conditions the ultrasound reverberates, or rings, between the two sheet sur- faces for a period of time that is several times greater than the pulse duration, When a lamination exists in the sheet the ringing is decreased, both in duration and amplitude, as a function of the area of the lamination relative to the transducer area. To test this method, flat-bottom holes with GCid 0o7 C several different areas were milled into one side of a sheet of Inconel to various depths. In experi- ments with existing equipment most of these reference defects were located. To properly in- strument this test method, the reflectoscope has been drastically altered. The pulse repetition was increased from 60 to 500 pulses per second, PERIOD ENDING SEPTEMBER 10, 1956 appropriate filters were added to the circuits, a variable-sweep delay circuit was added to permit immersed scanning, and external connections were provided for synchronization signals. This last chonge will permit the addition of gated alarm circuits and various data presentation and recording units in the near future, 219 644 008 ne PERIOD ENDING SEPTEMBER 10, 1956 material was not available, this material was re- worked by centerless grinding in order to remove most of the surface imperfections. A waiver was made on the porosity, which resulted in rejection of only large-size porosity defects, INSPECTION OF COMPONENTS R. L. Heest_ands = Sixty thermal-convection loops were received, and the welds were inspected by the dye-penetrant ‘method. Thirty-six of these loops were found to have welds which showed numerous indications of cracks, pinholes, and other defects. The remaining loops were accepted for use. The defective loops were repaired and reinspected prior to acceptance. Thirty-eight pieces of Inconel plate ‘were in- spected prior to shipment for fabrication into dished heads, and then they wete re-examined for manufacturing defects upon return. Several were found to have numerous pits that were apparently caused by a foreign material becoming embedded in the surfaces during the pressing operation. Four heads were found to have cracks running from the edge inward, which appeared to be as deep as Y in. These cracks are to be repaired, and the areas will be reinspected prior to use. Four small heat exchanger units were received and inspected, They were found to be acceptable for use in test operation, 10on assignment from Pratt & Whitney Aircraft, 644 009 .FLUORESCENT-PENETRANT INSPECTION OF TUBING G. Tolson The installation of the fluorescent-penetrant equipment to be used in the inspection of small- diameter, thin-walled tubing was completed, Tests are now being performed to compare the fluorescent- penetrant with the dye-penetrant type of inspection used heretofore. Indications are that it will be possible to ensure higher quality tubing with the new method because of its higher sensitivity., The fluorescent penetrant has revealed small pinholes “and tight laplike defects which were not dis- cernible by the dye-penetrant method. Exploration of some of these areas by polishing showed them to be as deep as 0.002 in. in some cases. How- ever, the greater sensitivity of the new method will present some problems, initially, because ex- perience will be required to distinguish indications of superficial imperfections from indications of true defects. WELDER QUALIFICATION PROGRAM A, E. Goldman - The number of welders qualified under ORNL welding specifications for ANP work is now 42, An estimate of the additional qualified welders " that will be required is now being prepared, and as soon as these requirements are known, qualifica- tion tests will be given. A qualification program is now under way at the Paducah installation, but, to date, no welders have qualified. 221 Part 4 HEAT TRANSFER AND PHYSICAI. PROPERTIES H. F. Poppendiek RADIATION DAMAGE G. W. Keitholtz 'FUEL RECOVERY AND REPROCESSING R. B. Lindaver CRITICAL EXPERIMENTS A. D. Callihan " " P R L LA e A e e P 0 4.1. HEAT TRANSFER AND PHYSICAL PROPERTIES H. F. Poppendiek ART FUEL-TO-NcK HEAT EXCHANGER J. L. Wantland - S, |, Cohen The fluid friction characteristics of the ART fuel-to-NaK heat exchanger were determined with six 60-deg staggered spacers and six 60-deg in- clined spacers alternately placed at 6-in. inter- vals.! These data are presented in Fig. 4.1.1. A full-scale model of the present ART fuel-to- NaK heat exchanger (which contains more tubes and @ somewhat different spacer configuration than the previous system) has been assembled, Pres- sure drop data will soon be obtained for it. The friction characteristics of this heat exchanger have been predicted by using the technique previously described.?2 The results indicate that the experi- mental data to be obtained should lie very near the curve in Fig. 4.1.1 for the staggered and inclined spacers, : 5. L. Wantland, Transverse Pressure Di ference Across Staggered and Inclined Spacers in the ART fguel;to-NaK Heat Excbanger, ORNL CF-56-6-'|43 {June 56 25, L. Wantland, A Method of Correlating Experimental Fluid Friction Data for Tube Bundles of Different ‘Size and Tube Bundle to Shell Wall Spacing, RNL CF-56-4-162 (April 5, 1956). woTREY ORNL—LR—DWG 16180 O * N~ l—r= 0.475 #3039 « R SL ,EJ0.0S' : - - H‘IW ~== - — : , w . *025 7\ —— | reomemon T | 2000 - 5000 - - -10,000 REYNOLDS NUMBER, My, 1" Fig. 4.'|.'|.. Fricfi'o'n 'Churucterisflcs of 60-Jeg Staggered Spacers and ‘60-deg Inclined Spacers in the ART Fuel-to-NaK Heat Exchanger, Isothermal friction measurements for the delta- array heat exchanger have nearly been completed. Heat transfer experimentation will commence in the near future. ART HYDRODYNAMICS C. M. Copenhaver F. E. Lynch Go Lo MU“er3 Core Hydrodynamics A new method for stabilizing the flow in the ART core was studied. |t consisted of placing a number of screens in the northern hemisphere of the core, as shown in Fig. 4.1.2. Qualitative velocity pro- files, which were determined by the phospho- rescent-particle flowevisualization technique, are shown in Fig. 4.1.3. The regions of flow sepa- ration on the outer core shell walls, which existed in a system without screens, apparently were eliminated because of the local turbulence gener- ated by the screens. The over-all friction loss in the core was deter- mined and is plotted in Fig. 4.1.4 in terms of a loss coefficient and the Reynolds number, The loss coefficient of this system was also estimated on the basis of screen friction data available in the literature. The two loss coefficients were in 30n assignment from Pratt & Whitney Aircraft. SELRET ORNL—L.R—DOWG {6181 STRAIGHT-THROUGH FLOW ~ ALL SCREENS ‘/ ARE 20 mesh/in. . 00108-in. WIRE DIA =5 b - 0.385 SOLIDITY 2.24in. ) 3 2.98in. 818in. Fig. 4.1.2. Schematic Diagram of Screen Arronge- ment in 10/44-Scale Model of ART Core. 225 ANP PROJECT_PROGRESS-REPORT | SEGRET ORNL-LR-DWG 416482 CORE ENTRANCE SCREEN POSITIONS THICK BOUNDARY LAYER OBSERVED AT A MIDPLANE REYNOLDS NUMBER OF 6000 CORE EXIT | Fig.r 4.1.3. Qualitative Velox":il'y Profiles for * Straight-Through Flow in Model Shown in Fig. 4.1.2. general ogréemént. In an actual reactor system the screens would probably be replaced with perforated plates having an equivalent pressure drop. A rough “estimate of the amount of Inconel that might be required for such plates in an ART core is about 7 kg. , Flow studies are to be started soon on a straight, anmular core with a diverging entrance and con- verging exit. Screens will be placed in the en- trance. The advantage of this arrangement is that " the screens will be in a relatively low neutron-flux region and thus will yield less reactor poisoning than they would in the system described above. 226 SeREFS ORNL-LR-DWG 16183 ED IN ART n ESTIMATED FROM TURE DATA | FRICTION LOSS COEFFICIENT, & n 104 2 5 40 REYNOLDS NUMBER AT MIDPLANE Fig. 4.1.4. Friction Loss Cocfficient in ART Core with Screens as a Function of the Mid-Plane Reynolds Number. Sodium Flow in Reflector Cooling System The pertinent data for the flow distribution in the reflector cooling system for concentric and ec- centric annulus conditions were presented in detail in the preceding report.4 The static pressure dis- tributions for the reflector and island annuli were not included, however, and they are presented here. The experimental studies indicate that the pressure losses through the annuli can be expressed as fL pV2 AP _(2t + nKs+Ka) 2% . where | { = friction factor in a smooth duct having an equivolent diameter of 27, t = annulus thickness, L = annulus length, n = number of spacer rows, K, = resistance coefficient per spacer row, K, = diffusion coefficient (1.0 for reflector an- nulus and 0.5 for island annulus), o P = fluid density, V = mean velocity in annulus, g = acceleration of gravity. 4C. M. Copenhaver, F. E. Lynch, and G. L. Muller, & o . o n The results of static pressure distribution calcule- tions are presented in Fig. 4.1.5. : Fluid Flow Visualizction Stu&y | A new attack on the problem of photographing phosphorescent particles used in-the flow-visuali- zation technique has been initiated. The following factors are being investigated: (1) -the matching of the phosphorescence emission spectrum and the ~ film sensitivity spectrum, (2) optimization of the ~ optical system of the camerg, -and (3) optimization ~of the film development process. - If satisfactory ‘photographs can be obtained, the qualitative phos- ~ phorescent-particle -flow-visualization method can become « quanhtatlve method. | 'ART CORE HEAT TRANSFER STUDY - N, D. Greene - H. F. Poppendiek ~G. L. Muller . G. M. Winn The results of an exPenmental and unolytlcul._ - study -of the temperature structure in an uncooled ‘ART core with a swirl entrance system were pre- ~sented in the prevsous report.5 - Experimental mean " ‘and transient temperature fields in the outer and ‘inner “core shell walls and within the circulating electrolyte were obfomed from the ART voiume- heat-source model, ' - Recently « study pfr the temperature'_Stru_cture in " the ART core With a vaned entrance was come pleted. The vane system, described previously,® ~ eliminated the flow separation region on the island in the northern hemisphere. Six complete power runs ‘were nidde for fh’e case of bofh pumps. in '=runged from below to above dessgn flow conditions. o = -Three power runs were ‘also made for the slmulcted - case of ‘‘one pump off,"’ Heat bolances were agcun L rwnhin +4% of being perfect, . The.inean, uncooled wall and flund temperature’ - profiles . obtained in'the ART core with- the vaned = entrance are presented iin Fig. 4.1,6. ‘The asym- "_r'f_.;fQ_metfles in-the outer core shell and island (inner) - ~core "shell ‘wall temperatures ‘can be explained on . the baosis of hydrodynamlc flow usyrnmetrles. Note . that the high island core shell wall temperatures "",?'—';found for the prevnous swarl-entrcmce case «are no - SN, D, Greene, et al,, ANP Quar Prog. Rep. june 10, 1956. 0RNL-2106 p 222. 6G. D. Whitman, W. J. Stelzman, ond W, Furgerson, ANP Quar. Prog. Rep. March 10, 1955, 0RNL-206|, p 24. PERIOD ENDING SEPTEMBER 10, 1956 BECRE T ORNL-LR-DWG 16184 N =~y = o o (=] REFLECTOR ANNULUS bH o ANNULUS STATIC PRESSURE (psi) -8 8 3 o‘ . .20 16 {2 8 4 0O -4 -8 -42 -6 -20 Z, AXIAL DISTANCE FROM CORE EQUATOR (in.} Fig. 4.1.5, Static Pressure Distributions for ART - Reflector and Island Annuli. ‘ “BEoRME ORNL-LR-DWG 16185 QUTER CORE SHELL IDEALIZED ART (THEORETICAL}) I.SLAND CORE SHELL . MIXED MEAN FLUID ~ HELICAL Re 134,000 (VANED ENTRANCE) W= W (UNIFORM) Pr=4.0 0. 2 4 8 B 10 12 14 6 8 ' AXIAL DISTANCE FROM INLET {in.) " Fig. 4.1.6. Mean, Uncooled Outer Core Shell and . "'Islund (Inner) Core Shell Wall. Temperature Meas- ~-urements for the Half-Scule ART Core Model with @ Umform Vo!ume Heut Source. . *|onger present because the vane system confumed . a deflector ring which prccflcully eliminated flow ~separation ‘on the island core ‘shell wall. “The wall ‘temperature solution for -an idealized ART (a par- 'allel-plutes Sysfem7) is also plotted in Flg. 4.1.6; 7H. F. Poppendiek und L. D Pulmer, Forced Convec- tion Heat Transfer Between Pargllel Plates and in Annuli with Volume Heat Sources Within the Fluids, ORNL-1701 (May 11, 1954). 227 ANP PROJECT PROGRESS REPORT this predicted uncooled-wall temperature profile lies between the island and outer core shell wall temperature measurements. Typical experimental transient wall and fluid temperature measurements are shown in Fig. 4.1.7. The results are expressed in terms of the total temperature fluctuation divided by the axial temperature rise of the fluid gomg_ through the core. SEOREY ORNL—LR—-DWG 16186 U 2 P~ A~ = 0.25 g Bty ISLAND SHELL AT EQUATOR be—{ §8C— T\ — At — = 0.8 Bim AV _r CORE SHELL AT EQUATOR 1A (AU NN | e MID-STREAM BEYOND EXIT 0 1.0 2.0 3.0 TIME (sec) HELICAL Re = 131,000 . - W = W {UNIFORM) - . Afs = TOTAL CHANGE IN WALL TEMPERATURE Aty = AXIAL FLUID TEMPERATURE RISE IN GOING THROUGH CORE Fig. 4.1.7. ‘Tfl:n.sient ART Surface and Fluid Temperatures (Uniform Yolume Heat Source). The mean, uncooled outer and island core shell wall temperatures (except the island core shell wall in the northern hemisphere) for the vaned system were greater than those found previously for the " swirl flow system; this results from the helical Reynolds number for the vaned system being signif- - icantly lower than that for the swirl system. The 228 temperature fluctuations of the island and outer core shell walls in the northern hemisphere were rela- tively lower for the vaned system than for the swirl system; conversely, the temperature fluctuations in the southern hemisphere were relatively larger for the vaned system _than' for- the swirl system. - The fluid temperature fluctuations beyond the core ‘exit were also larger for the vaned system than for the swirl system.. ~ The frequency spectrums of the temperature fluc- tugtions for the vaned system were very similar to the spectrums observed in the swirl system. The ‘peripheral asymmetries were greater in the vaned system than in the swirl system; this characteristic' was also observed in the correspondmg hydrody- ) nomlc Structure, "The uncooled wall temperature rises obove the mixed mean temperature in the actual ART system, where the radial volume heat source distribution will be approximately a hyperbolic cosine function, were previously shown to be greater then twice those for a uniform power density distribution such as that which existed in the volume heat source experiment. Consequently, temperatures as high as about 1850°F may occur in the vaned system if the core shell walls are not cooled properly, Some interpretations of the wall temperature fluctuations in the outer core shell, island core shell, and heat exchanger walls in the actual reactor system have been described previously. For example, in the event that a momentary flow stagnation exists adjacent to the core shell wall, the fuel-Inconel interface temperature fluctuation will be only about one-fourth as large as the fluctuation in the tem- perature of the fuel at some distance from the wall; this reduction in the temperature rise -occurs be- cause of the relatively good heat transfer for the Inconel when the fuel is momentarily stagnant, However, when a high-velocity turbulent eddy of lower or higher temperature level suddenly wipes the Inconel surface, calculations have shown that the relative heat transfer to or from the Inconel is then poor, and consequently the fuel-Inconel inter- face temperatures are nearly as large as those that would exist if the wall were insulated, THERMAL-CYCLING EXPERIMENT H. W. Hoffman D. P. Gregory® The experimental system designed to investigate the effect of thermal cycling at an intermediate 80n assignment from Pratt & Whitney Aircraft. 5% M,' frequency range ( 1/4 to 2 cps) on metals in the . presence of reactor fuel mixtures has been suc- ' cessfully operated, The data obtained at two .. temperature levels with NaF-ZrF +UF (50-46-4 mole %) flowing turbulently through 12 in,-0OD, : 0.035-in,-wall |nconel tubes are summarized in 5 Table 4.1.1. The corrosion attack on the metal in both the heater and test sections during these runs is shown in Figs. 4.1.8 through 4,1.12, PERIOD ENDING SEPTEMBER 10, 1956 In the ART core, the high differential temperature cycling will be experienced by the Inconel core shell surfaces and the heat exchanger tubes as a result of the hydrodynamic instabilities that will exist in the system. These temperature fluctua- tions have been simulated in a bench-scale experi- ment by subjecting the salt flowing through an electrically heated tube to cyclic heating. The experimental system is shown in Fig. 4.1.13. As TABLE 4.1.1. PRELIMINARY THERMAL-CYCLING DATA FOR INCONEL TUBES EXPOSED TO THE FUEL MIXTURE (NO. 30} NaF-ZrF ,-UF, (50-46~4 MOLE %) Heoter Section Test Section Mean Duration Inlet Fluid Inside Surface Average inside Surface Average Cause of Run (hr) Temperature . Temperature (0 F) Depth of Temperature ‘o F) Depth of Termination (°F) - Attack Attack of Test Average Fluctuation (mis) Average Fluctuation (mils) ET-A 16 1265 1455 tis0 1273 17 0.5 Oxidized electrode 4 1265 1546 +225 1287 112 . ET-B 240 1275 1415 *170 1330 116 1 Stopped to alter test conditions ET-C 4 1580 1755 189 1613 136 2 Melted electrode : ET-D 14 1580 1845 1240 1590 183 4 Tube break ET-E 5 1615 1905 1235 1635 143 0.5 Tube break duced 31%. (Secret-with-eeptiony—— . Fig. 4 1.8. Corrosuon Results of Thermal Cyclmg of Inconel Tubes Exposed to the. Fuel Mlxiure‘ (No. 30) NuF-ZrF ‘UF, (50-46-4 mole %). Run ET-A: mean fluid temperature, 1265°F; heater inside surface temperafure, 1500 1+ 200°F; test section msude surface temperature, 1280 + 10°F, 250X, Re- 229 ANP PROJECT PROGRESS REPORT the fluid moves downstream, the surface of the test section is exposed to periodic temperature fluctua- tions imposed on the fluid in the heater. Because of the poor thermal diffusivity of Inconel, the major effect of this thermal. cycling is confined to a region close to the metal-fuel interface. In this region the combination of the high coefficient of CTREN EEEPEBRET thermal expansion and the high modulus of elas- ticity of Inconel, with its poor thermal conductivity, may result in thermal stresses which are of suffi- cient magnitude to cause eventual damage to the metal by fatigue. These thermal-cycling experi- ments are being made in order to defermme the extent of fatigue damage. | Fig. 4.1.9. Corrosion Results of Thermal Cycling of Inconel Tubes Exposed to the Fuel Mixture (No. 30) NaF-ZrF -UF (50-46-4 mole %). Run ET-B: mean fluid temperature, 1275°F; heater inside surface temperature, 14]5 t 170°F; test section inside surface temperature, 1330+ 1°7F. 250X. Re- . duced 31%. {Secretwith-ceptiond INCHE! TEFEIAELE ¥ L Fig. 4.1.10. Corrosion Results of Thermal Cycling of Inconel Tubes Exposed to the Fuel Mixture (No. 30) NaF-Z(F ,-UF, (50-46-4 mole %). Run ET-C: mean fluid temperature, 1580°F; heater inside surface temperature, l755:|: 190°F; test section inside surface temperature, 16]3137°F 250X. Re- duced 31%. 4beTretwith-eaption) 230 PERIOD ENDING SEPTEMBER 10, 1956 The use of the ““floating’’ electrodes previously supported from above by thin, flexible rods. The described? has been temporarily abandoned be- power connections are made through copper braids cause of continuing failures at the electrodes. clamped to the electrodes. The bellows provided The modified electrodes (Fig. 4.1.13), fabricated to relieve the stresses of linear thermal expansion of Inconel, are welded or brazed to the tube and has been relocated so that the forces are applied - “along the axis of the bellows, 9H. W. Hoffman and D. P. Gregory, AfiP éwr. Prog. While the thermal cycling data presented here Rep. June 10, 1956, ORNL-2106, p 227. - are preliminary, the rate of corrosive attack on the -Fig. 4. l 'll Corros:on Results of Thermal Cycling of Inconel Tubes Exposed to the Fuel Mixture (No. 30) NaF-ZrF -UF, '(50-46-4 mole %). Run ET-D: mean fluid temperature, 1580°F; heater inside surface temperature, 1755 1 190°F; test section inside surface femperature, 1590 £ 83°F, 250X. Re- duced 30%. {Sem#-wn'frcupfrefi)—- Fug. 4.1 12 Corrokion Results of Thermal Cyclmg of |nconel Tubes Exposecf 1o fhe Fuel Mnxture (Ne. 30) NaF.IF -UF, (50-46-4 mole %). Run ET-E: mean fluid temperature, 1615°F; heater inside surface temperature, 3905:!:235" F; test section inside surface temperature, 1635 + 43°F, 250X. Re- duced 31%. «(Serret-withcoption)~ 231 Zee UNCLASSIFIED g PHOTO 26932 [ AR TN I SRR AT EERALE LA IAESREEANEREE) T wt LAOJIN §SFJ00¥d LI3IF0dd ANV inconel for the runs ET-C, ET-D, and ET-E is note- worthy. Whether this is to be attributed to the very large fuel volumes and the small areas of heated surface, to the high temperatures at the metal-fluid interface, or to thermal cycling of this interface has not yet been established. An experiment is now in progress to establish a base corrosion rate for this specific system under isothermal conditions with interface temperatures in the range of 1600 to 1800°F. It is to be noted that both runs ET-D and ET-E failed due to breaks in the Inconel tube at the outlet end of the heater, immediately adjacent to the electrode. The effect on the tube of the constraint imposed by the electrode is not known, Experiments are planned in which the electrode will contact only .one-half of the tube circumference at the electrode position. Alternative approaches to thermal cycling experiments that eliminate the need for electrodes are being investigated, SHIELD MOCKUP CORE STUDY Lo. Co Pdlmer The recently initiated study of the temperature structure in the Shield Mockup Core is continuing. In particular, the temperature distributions in the beryllium reflector and the various Inconel shells are being determined, An effort is being made to explore the possibility of eliminating the cooling system for the beryllium reflector by the reduction of certain high thermal resistances in the system. 105, |. Cohen, W. D. Powers, and N. D. Greene, A szstcal Property Summary for ANP Fluonde sztures, ORNL-2150 (Aug. 3, 1956) _ o PERIOD ENDING SEPTEMBER 10, 195§ HEAT CAPACITY W. D. Powers A study was conducted to determine the influence of fission products in a fluoride fuel mixture on the enthalpy and heat capacity in the liquid state, Quantities of RbF, BaF,, and LaF, were added to a zirconium-base fuel to sumu!ate an ANP fuel heavily laden with fission products resulting from 1000 hr of exposure to a neutron flux of 1015 neutrons/cm2.sec. The ‘results of the measure- ments, presented in Table 4.1.2, indicated that, within expenmental error, the change in heat copoclty as a result of the addition of simulated fission products was negligible. This result had previously been predicted on the basis of the established heat capacity correlation expression. The enthalpies of the fuel with the additives were about 10 cal/g less than those of the fuel alone. Earlier enthalpy and heat capacity measurements on the fuel without additives are also given in Table 4.1.2 to illustrate the reproducibility of the data. VISCOSITY AND DENSITY S. I. Cohen A summary of the density, viscosity, heat ca- pacity, thermal conductivity, electric conductivity, and surface tension measurements that have been made at ORNL during the past several years was prepared and issued.'® Viscosity measurements were made on LiF-BeF,-ThF -UF, (71-16-12-1 ~ mole %), which is a mlxture of possntle interest in TABLE 4.] 2. COMPAR!SON OF ENTHALPIES AND HEAT CAPACITIES OF FUEL MIXTURES ~ WITH AND WITHOUT SIMULATED FISSION-PROPUCT ADDITIVIES Entholpy, = Hypoc : Heat Cupdcity : {eal/g) : _ - {cal/g:°C) Material — T - A T A - At At At A -600°C 700°C 800°C 600°C 700°C 800°C _NaF-ZfF ~UF, (50-46-4 mole % o | | Data obtained in 1955 1710 197.2 222.6 0266 0.258 - 0.249 ' Dota obtained in 1956 - 1681 1952 - 221.3 0.276 0.266 0.256 NaF-ZrF -UF ,-RbF-BoF ,-LaF, 158.5 186.0 212.0 0.283 0.268 0.253 (47.6-43.7-3.8-1.1-1.0-2.8 mole %) 233 ~the fused-salt power reactor program. ANP PROJECT PROGRESS REPORT The vis- cosities were found to vary from about 13 centi- poises at 600°C to 4.8 centipoises at 800°C, THERMAL CONDUCTIVITY W. R. Gambill A tentative correlation has been developed for the prediction of the thermal conductivities of fused salts and salt mixtures at their melting points,!! For this correlation the total conductive heat trans- mission of the fused salt is divided into the two distinct contributions, One is the atomic or lattice portion, which arises from the short-range atomic or moleculor order present in liquids, in general; and the other is the ionic portion, which arises from a drift of ions, and subsequent energy transfer, between atoms. Ny, R, Gambill, Prediciion of the Thermal Conduo thty of Fused Salts, ORNL CF-56-8-61 (Aug. 10, 1956) The atomic portion of the total conductivity, (k,),, in Btu/hr-ft2(°F/ft), may be correlated by- the expression T!/z 4/5 (k,), = 1043 — . F9/5 T, = melting temperature or Ilqmdus tempera- ture (for mixtures), °K P, = density of fused salt ut T, g/em?, M = molecular weight for pure salts = Fx.M, for salt mixtures, x. = mole fraction of ith component, | M, = molecular weight of ith component. The ionic portion of the total conductivity is plotted in Fig. 4.1.14. With the exception of NaOH, Saener ORNL-LR-DWG 16187 NONELECTROLYTES 100 NaOH 90 = - " N=§ X%; N;, WHERE x; = MOLE FRACTION OF /™" COMPONENT - N; = NUMBER OF IONS PER MOLECULE 80 — A=100 — . WHERE &, = ATOMIC PORTION OF THERMAL CONDUCTIVITY T #,= TOTAL THERMAL CONDUCTIVITY \ 80 , 14 \ 30 50 35 \ a4 104 40 4 (%) SALT MIXTURE NO. COMPOSITION " NoF-KF~LIF (11,5-42-46.5 mole %) NoF=KF-LiF-UF (10.9-43.5-44.5-1.1mole %)} NaF—Zl;F"-UF"(SO—46-4 mole %)} NoF -~ BeF, (57-43 male %) NoF-ZrF - UF,{53.56-40~6.5 mole %) LiF ~RbF (43~ 57 mole %) ~N e N SALT MIXTURES > - HTS NO. 42 AND 104 ‘\‘( NO. 35 20 . , NO. {4 \\ NO. 44 NO. —g\ 10 30 Q o [ 2 3 4 5 6 7 8 N Fig. 4.1,14. Thermal Conductivities of Fused Salts at Their Melting Temperatures and Snlt Mlxtures at Their Liquidus Temperatures. 234 the data lie along @ smooth curve, The deviation of NaOH is believed to be caused by the intense hydrogen bonding (between OH radicals), which. effectively reduces the ionic component of the conductivity to a negligible value. PERIOD ENDING SEPTEMBER 10, 1956 ~ A relation for predicting the temperature de- pendence of the thermal conductivities of fused salts is currently being studied. [t can not satis- factorily be evaluated, however, until more tem- perature-dependent conductivity data are available, 235 ANP PROJECT PROGRESS REPORT 4.2. RADIATION DAMAGE G. W. Keilholtz EXAMINATION OF DISASSEMBLED MTR IN-PILE LOOPS NOS. 3, 4, AND 5 A, E. Richt c. E”is Ro No Ramsey E. J. Manthos E. D. Sims W. B. Parsley R. M. Wallace Examination of MTR in-pile loop No. 3, described previously,! was completed, except for metallo- graphic examination of the pump impeller, Corro- sion of the nose coil was found to vary from 1 mil near the inlet to a maximum of 3 mils of inter- granular attack near the outlet, Fig. 4.2.1. Com- parisons of the inner wall of the nose-coil tubing . P.LCarpenfer et al, ANP Quar. Prog. Rep, Dec, 10, 1955, ORNL-2012, p 27. UNCLASSIFIED ORNL~-LR~-DWG 16188 264 260 1455°F 1395°F 2 mils 261 263 . £ mil 262 1Y% mis 269 2 mils 273 270 1490°F 1435°F 3 mils 2' mils ' ' 271 {390-1490°F 2 mils 274 1415°F OUTLET g-amis “‘_Fig. 4.2.1. Diagram of Sectioned Nose Coil of MTR In-Pile Loop No. 3 Showing Locations from Which Metallographic Samples Were Taken, the 'Maximum Temperature Which Occurred at the Par- ticular Location During Operation, and the Depth of Corrosive Attack Found by Subsequent Metallo- graphic Examination of the Samples. 236 785 (01 on the compression side of the coil with the opposite inner wall on the tension side showed no differences ‘in depth of attéck, The depths of attack found on the straight sections of the fuel tube are compared in Fig. 4.2.2. Photomicrographs of samples from the nose coil and the straight sections are presented in Figs. 4.2.3 through 4.2.7. No mass-transferred crystals were found on any of the specimens, | Disassembly of MTR in-pile loop No. 4 has been started, but as yet no results are available. Opera- tion of this loop was described previously.?2 In-pile loop No. 5, which was inserted in the MTR but could not be filled, was shipped to ORNL for disassembly after the section behind the pump motor was removed at the MTR., During the attempt to fill the loop the fuel-circulating pump seemed to be chattering, some of the thermocouples were not functioning properly, leakage occurred throudh the pump bulkhead, and a fill-line Calrod heater failed after it had been raised to higher than normal operating temperatures. 2¢, C. Bolta et al., ANP Quar. Prog, Rep. June 10, 1956, ORNL-21064, p 75. CONFIDENTIAL ORNL-LR-DWG 16189 TO NOSE COIL DISTANCE FROM PUMP (in.) 35 30 25 21t 7 12 s T FUEL LINE - T =1 1 I SAMPLENO. 309 308 307 306 305 304 SLIGHT ATTACK FROM SAMPLE NO. 304 TO 309. VARIED FROM O TO 0.5 mil : 0 . = : =2 FROM NOSE COIL o DISTANCE FROM PUMP (in.) 42 37 3 26 22 a2 FUEL LINE ° ] | | 1 i 1 ! 1 I | | 1 i SAMPLE NO. 284 ' 287 291 295 296 303 DEPTH OF ATTACK (mils) 3 2 i § i i Fig. 4,2.2. Results of Metallographic Examina- tions of Straight Sections of Fuel Tubing from In- Pile LOOP Noo 3. | | ! j 1 ) i 5 } i ! i | i : Fig. 4.2.3. Sample No. 264, Taken Near Inlet End of Nose Coil of MTR In-Pile Loop No. 3, Showing Intergranular Corrosion to a Depth of 2 mils. 250X, {Cenfidemiai-with-ception u Fig. 4.2.4. Sample No. 273, ‘Taken Near Outlet End of Nose Coil of MTR In-Pile Loop No. 3, Showing Intergranular Corrosion to a Depth of 3 mils. 250X, “FSonfidentictwith option~ 237 e %35 002 ANP PROJECT PROGRESS REPORT Fig. 4.2.5. Somple No. 284 Taken from Straight Section of Fuel Tubing Near Outlet of Nose Coil - at a Point 42 in. from the Pump of MTR In-Pile Loop No. 3. Intergranular corrosion to a depth of 3 mils may be seen. 250X. {Confrdentivtwittrraption) Fig. 4.2.6. Sample No. 303 Taken from Straight Section of Fuel Tubing Between the Nose Coil Outlet ‘and the Pump at o Point 12 in. from the Pump of MTR In-Pile Loop No. 3. Intergranular corrosion to a depth of 1 mil moy be seen. 250X. {(Confidentielwith-caption)- 238 " PERIOD ENDING SEPTEMBER 10, 1956. L Fig. '4 2.7. Snmple No. 304 Taken from Straight Section of Fuel Tubing Between Pump and Nose Coil Inlet at a Point 12 in. from the Pump of MTR In-Pile Locp No. 3. Intergranular corrosion to a depth of 0.5 mil may be seen. 250X {Confidentiehwith-capton) The inlet fuel line to the nose coil was found to be full of fuel at a point 24.!/2 in, from the face of the pump, and a shallow layer of fuel was found in the outlet line from the nose coil at the same dis- tance from the pump. The fuel in the inlet and outlet lines at a point 2234 in. from the face of the pump is shown in Fig. 4.2.8. A radiograph of the fill tank, fill line, and vent line, Fig. 4.2.9, showed the fuel in the fill line to have frozen solid for the first 4/ in frorn the fill tank, and, from fhut pomf on, there were - cavities in the fuel. - The vent line appedred to be clear, except for a few particles and a deposit close to the pump. The fuel in the fill line at the point of maximum - fuel porosafy, 1.6 in. from the pump, - is’ shown in- Fig. 42,10, The plug probably - r_:occurred in the fill line ot this point, since the fuel found here was dlfferent in appearance from that. found closer to the fuel tank where no’ porosity .. was observed.,The fuel m the flll tonk is shown in - Fig. 42,11, The pump sump was' devoid of fuel. Thirteen- fuel samples ‘were obtcnned for_chemical analysis at various focations around the Ioop. The fill-line Calrod heater failure was located and is shown in Fig. 4.2.12. The pump chattering slight, if any, corrosion occurred. was probably caused by rubbing of a wiper ring behind the slingers on the slinger shaft, Fig. 4.2.13. All the thermocouples in the nose-coil region appeared to be in operating condition. Thermo- couple No. 28, which was at the rear of the fill line, could not be found. The cause of the leakage through the pump bulkhead could not be determined, since the damage of the glass seals may have ~ occurred during disassembly,3 'INVES‘TI-GA'HON OF MATERIALS REMOVYED "FROM MTR IN-PILE LOOPS NOS.3 AND 5 R. P, Shields Chemical analyses of fuel and other materiais . from MTR in-pile loop No. 3 have been made. A ~partially completed :study of the results has indi- cated that, on the basis of the chemical analysis “of the fuel for iron, chromium, and nickel, very The amber- colored material found in the pump region was * probably an oil-decomposition product, This con- - clusion is based on the fact that the material is 3¢, Ellis et al., Examination of ANP In - ; (e} - - - Flg _4.2.16.7 Modified Motor and Pump for Use in LITR Vertical In-Pile Loop. (a) Impeller and ‘shaft components. (b) Pump assemblies. (c) Motor parts. - : 246 s ¥ “j L ~3 ) o . ~ance tests were run with water in a plastic model . T 7/ - at the design point, so the pump is still operated at ‘about the same speed, _were converted to valves for fused salt fuel by ' ‘UF (63-25-12 mole %). A plot of the estlmdted PERIOD ENDING SEPTEMBER 10, 1956 increase caused- @ measurable but unimportant e : . S - ORNL—LR-DWG 16191 decrease in pump efficiency. The pump perform- 9000 - of the pump. The test results are presented in ' = 7 . . - | | / Fig. 4.2.17, which shows the pump pressure as a | function of water flow before and after the increase 5000 |— _ in clearance for various pump speeds. The resist- r'/ ance curve for the test loop is also shown. Both port than was formerly used. The results of the | /,/ 2000 sets of curves are for a pump having a larger exit two changes are approximately equal and opposite - REYNOLDS NUMBER “The pressure vs flow data obtamed with water - using recently obtained values for the viscosity 1000 L and density of the fuel mixture (No. 41) NaF-ZrF - .~ 1000 2000 - 5000 10,000 . PUMP SPEED (rpm) Reynol&s number for the fuel in the LITR vertical _ | | in-pile loop as a function of pump speed -is pre- Fig. 4.2.18. Estimated Reynolds Number of the sented in Fig. 4.2.18. L : Fuel Mixture (No. 41) NaoF-ZrF «UF, (63-25-12 A full-scale Inconel pump -has been ussembled mole %) in LITR Vertical In-Pile Loop as a Func- that is identical with the pumps to be used in the tion of Pump Speed UNCLASSIFIED ORNL-LR-DWG 16190 ———0.050TO 0.07Qin. IMPELLER CLEARANCE 0.010 TO 0020 in. IMPELLER CLEARANCE ‘if HEAD {psi) | - LOOP RESISTANCE ~ N - . B i ‘“ . —— o e TRl . 03 - 04 - 08 - 06 or o : '*VFLOW(gprn) C ‘ ) ' ' Fig.42 17. Performcnce Curves with Wcter for Modlfied Pump for LITR Vemcul In-Pile Loop Before and After |ncreasing Impeller Clearances. 247 "ANP PROJECT PROGRESS REPORT Inconel in-pile loops. It will be filled with de- pleted fuel and run in the laboratory under con- ditions as nearly like those which will exist in the reactor aos possible. The existing parts for in-pile loops are being modified to incorporate these recent changes. B DESIGN CALCULATIONS FOR LITR VERTICAL IN-PILE LOOP M. T. Robinson J. F. | Krause? An extensive series of design calculations for the LITR vertical in-pile loop was previously re- ported.'® [n an appendix to that report, a method was presented for making such calculations by _electronic analog simulation. ential equations describing the heat balance in the loop, followed by analog computation, supported by . some desk-machine computation. The correct solu- tions were then found :by'ir_lterpolation among a femily of analog computor results. A much- improved method of analog computation has now been devised which eliminates the interpolation - procedure and the desk-machine work, as well as substantially shortens the amount of analog com- putation required to obtain correct solutions. The new circuit has been used to make calcu- lations of the behavior of the Mark VIl Joop!® when circulating the fuel mixture (No. 41) NaF- ZrF‘-UF4 (63-25-12 mole %) in position C-46 of the LITR. The flux data previously reported!! for that position were used. The physical property 90n loan from Pratt & Whitney. 10M. T. Robinson and D. F. Weekes, Design Calcula- tions for a Miniature High-Temperature In-Pile Circulat- ing Fuel Loop, ORNL-1808 ( pt: 19, 1955) with **An Appendix on Analog Simulation’ by £ R. Mann, F. P. Greene, and R, S. Stone. Um, 1. Robinson, Solid State Semiann. Prog. Rep. Feb, 28, 1954, ORNL-1677, p 27. ~ tip at the mid-plane of the LITR. “assumed to be depressed 50% by the presence of _ The method was based on partial analytical solution of the differ- data used for air were the same as those used for the previous calculations.1? The physical property data for the fuel mixture are given below: | Viscosity Specific heat Thermal conductivity . Density 4,3 centipoises 1.0 joule/g°C 0.14 w/em°C 3.38 g/cm’ The loop was considered to be located with its The flux was the loop. The initial air temperature, To, was always taken as 30°C. A Reynolds number of 4000 was assumed for the fuel, ' Some results of the calculations are summarized in Tables 4.2.1 and 4.2.2 and in Figs. 4.2.19 and 4.2.20. = Typical profiles of the mixed-mean fuel temperature, T,, the fuel-metal interface tempera- ture, T,, and metal-air interface temperature, T, ‘are shown in Fig. 4.2.19, and profiles of the air ~ temperature, T, and of the heat transfer function, y, are shown in Fig, 4.2.20. The discontinuities that occur in the plots of T,, Ty T, end y result from the use of two separate cooling-air streams, Local heat flow would be expected to remove these discontinuities in the actual loop. The very dif- ferent shapes of T, and of T, and T, are notable. This feature is important in analyzing experimental data, since T, is the only quantity for which a direct measurement can be obtained. Data on the effect of the initial fuel temperature (T9) on other quantities are presented in Table 4,2,1. As was realized before, 19 AT, is very little affected by changes in T?. It should be noted, however, that " appreciable changes in cooling-air requirements are associated with changes in T9 Table 4.2.2 itlustrates the effect of fuel flow rate on other quantities, The important points to notice are the close similarity of the values AT, and AT, ex- cept at the lowest fuel Reynolds numbers; the TABLE 4.2.1. EFFECT OF INITIAL FUEL TEMPERATURE ON COOLING-AIR REQUIREMENTS . Air T}’ °c) Reynolds 73 (°0) T3 (°C) AT, (C) AT, (°Q) AT, °C) Number 750 34,500 659 641 87 81 83 800 31,500 708 690 88 83 85 850 29,750 756 733 89 85 8 248 w3 i w - PERIOD ENDING SEPTEMBER 10, 1956 TABLE 4.2.2. EFFECTS OF FUEL FLOW RATE ON COOLING-AIR REQUIREMENTS FOR AN - INITIAL FUEL TEMPERATURE, T?, OF 800°C ~ Fuel - Air - Reynolds Reyfio_l_ds . T‘,’(°c') | _ | Tg (°c) ATI (°C) AT, (°0 = A1, {°O 7 MNumber | Number ' 2,000 45000 477 450 163 134 139 13,000 .~ 34000 650 632 14 04 106 4000 31,500 708 690 88 83 85 6000 29,50 7:1 732 59 57 58 8000 - 29,000 764 74 M 43 45 10,000 29,000 . - 770 - 752 - 3 35 36 :860 - aeo " g20 | © 860 . “TEMPERATURE {*C) " 740 L re0 100 b SR ORNL-LR-DWG 15244 ' FUEL COMPOSITION: NGF-Zrf,-UF, (63-25-12 mole %) " FUEL REYNOLDS NUMBER: 4000 - ‘ : _ AIR REYNOLDS NUMBER: 31,500 ,. FLUX: 50% OF LITR FLUX LITR POSITION: C-46 MIXED~- MEAN FUEL TEMPERATURE, % FUEL ~METAL INTERFACE o~ o Q SN e Lo w 2 g = i - AIR INTERFACE 680 Lt 2077 2 - 40 .. 80. . _ 80 100 120 t40 . . 180 180 200 e R T T L . DISTANCE FROM FUEL ENTRANCE (cm) - Fig42.l9 ACclcfi'luie.d Tempéraiure Pr-ofilersu for LITR Vertical In-Pile Loop (Mark Vill). 249 =% ) et 3 b h":fi ANP PROJECT PROGRESS REPORT ORNL-LR-DWG 15245 73 FUEL COMPQSITION: NoF-ZrF, -UF, (63-25-12 mote %) FUEL REYNOLDS NUMBER: 4000 AIR REYNOLDS NUMBER: 31,500 FLUX: 50 % OF LITR FLUX LITR POSITION: C-46 500 400 300 TEMPERATURE (°C) t00 0 20 40 60 80 100 69 HEAT-TRANSFER FUNCTION, y 1 s n - o o~ RATE OF HEAT TRANSFER (w/cm) 53 L @ COOLING-AIR TEMPERATURE, 7, 45 4 37 120 140 160 180 200 . DISTANCE FROM FUEL ENTRANCE (cm) Fig. 4.2.20. Calculated Air-Temperature and Hect-Transfer-Function Profiles for LITR Verticel In-Pile Loop (Mark VIII), small difference between T? and To, and the large difference between T® and T W especmlly at lower flow rates. The first point suggests that experi- mental values of AT, may be used without correc- tion to characterize filel “AT". The second point results from the fairly small heat extraction rate, around 30 w/cm? or less. The third point is most important in assessing corrosion results. ART RE‘ACTIVITY TRANSIENTS ASSOCIATED - WITH FLUCTUATIONS IN XENON-REMOVAL EFFICIENCY M. T. Robinson J. F. Krause The possibility has been investigated that fluc- tuations in efficiency of the ART xenon-removal equipment might lead to troublesome transients in reactivity. The calculations previously reported 12 250 3} ) & A . -; have been extended for this purpose. Since only short times are of interest here, the equations may be simplified to (1) P = May -, (2) y = fi)tl(x —azy)—Agyt—Bf -Agy ' where _ x = Xe'35 poisoning in the fuel, y = *‘‘equivalent poisoning’’ in the gas phase, = RTS, the product of the universal gas con- stant, R, times the absolute temperatute, T, times the solubility coefficient of xenon in the fuel, S, ‘ 2y T Rob:nson,' A Tf.veorettcal Study of Xel33 Poisoning Kinetics in Fluid-fueled, Gas 500 ev) bish ok - effects with- energy. fabricated to measure the fast flux-by the method of Hurst et al.1® The shield will be used-in con- " junction - ‘with Pu23%, Np237 “and U"'38 fonis In ‘order to compensate for the htgh f:ssion cross sece -at low energles & 1.5cm path of B0 1o lhe foil was required. To check for leckage, o PERIOD ENDING SEPTEMBER 10, 1956 TABLE 4.2.3. INITIAL INCREASES IN ART REACTIVITY FOLLOWING FLUCTUATIONS IN XENON-REMOVAL EFFICIENCY %) Al(sec“) )‘g(sec-l)* Initial Slope Initial Period o7 t<0 >0 <0 t>0 (%/sec) ({sec) 3.98 1 1 0 0.025 1.44 3 0.5 0.5 0 0.025 0.54 3 0.1 S0 0 0.025 0.35 n 0.05 0.05 0 0.025 019 21 0.01 0.01 0 0.025 0.037 106 0.005 0.005 0 0.025 0.019 206 0.74 0,001 0.01 0.025 0.025 0.0069 106 0.001 0.1 0.025 0.025 0.063 12 0.001 R 0.025 0.025 0.27 3 0.75 0.001 0.01 0.015 0.015 ~0,0072 105 | ~0.001 0.1 ~0.015 0.015 0.067 1 0001 1 0.015 0,015 0.27 3 0.82 0.001 0.01 0.005 0.005 0.0075 109 10,001 0.1 0.005 0.005 0.071 12 0.001 B 0.005 0.005 0.27 3 0,97 0,001 0.01 0.002 0.002 0.0095 102 S 0001 01 0.002 0.002 0.080 12 0,001 1 - 0.002 0.002 0.26 4 3.98 0.001 0.01 0 0.038 104 0.001 0.1 0 | 0.35 | n o001 1 | "und for determmmg 1he varaohon of Jrradwhon‘f A boron shield has™ ‘been - ~ of fl'se boron-covered to cudmlum-covered achvmes tion of Pu239 various resonance detectors were irradiated inside 136, s. Hurst et al., Rev. Sci. Instr, 27, 153 (1956). | - *i..The' ih;r_élhg.jas‘:f[o.fi rqte isu g(STP f_iieg\i-rs'/da)-f') !f-,_2427_ X r'_IO5 Ag(sec"‘); - the Bm shield and- msude cadmlum. The ratios for cobalt and manganese detectors were 60 and 20, respechvely. For no shielding at the resonance fenergy, o ratio of cbout 2 would have been ob- erved. - Therefore, 135~ and 340-ev neutrons . are sirongly absorbed by the shield. ¥4p, J. Hu es, Pile Neutron Research, p 138, Addi- son-Wesley, Cambridge, Mass., 1953, 253 018 ANP PROJECT PROGRESS REPORT A photoneutron source was used as part of the study of the variation of displacements with neutron energy before the exact effectiveness of gamma rays in displacing atoms was learned. Depending on the energies, a neutron is 100 to 1000 times as effective as @ gamma ray. This is a sufficient ratio for reactor or accelerator irradiations but not for photoneutron sources. The antimony- beryllium source which was used produced de- tectable effects in n-type germanium with initial concentrations of 10'3 donors/em®, but the ratio of gamma rays fo neutrons was on the order of 105, which made the gamma rays about 100 times as effective as the neutrons. EFFECTS OF RADIATION ON ELECTRONIC COMPONENTS J. C. Pigg C. C. Robinson!’ In the previous report!® the behavior of the characteristic curves of transistor and semicon- ductor diodes was compared. The emitter charac- teristic of the tronsistor was compared with the forward characteristic of the diode., The collector characteristic of the transistor was compared with the reverse characteristic of the diode. Correla- tions were drawn between Co%? gamma irradiation and reactor neutron effects on the characteristics of these devices. It was noted, however, that the barrier change resulting from Co%% gamma irradic- tion showed properties not noted previously with reactor irradiation, Electrical annealing, that is, removal of charged interstitial atoms by the strong electric field present in the barrier, was noted. It was pointed out that because of this effect, the bias applied to a semiconductor component might affect the change observed when the component was operating in @ radiation field. Exposures of Philco surface-barrier transistors have been made in the ORNL Graphite Reactor, but, thus far, attempts to measure the amplification characteristics of these surface-barrier fransistors under irradiation have been unsuccessful. How- ever, the photo-emfs developed across the collector and emitter barriers during reactor irradiation have been measured, and ¢ series of before and after measurements . of collector characteristics have 150, assignment from Wright Air Development Center. 6Jn c. Pl dnd C. c. Roblnsofl. ANP Quafo Prog. Rep. June 10. 1956. ORNL-2104, p 243. 254 -3 G been made. Since a sharp barrier is desirable for an emitter and a diffuse barrier is desirable for a collector, it would be expected that there might be some difference in the behavior of these barriers in ¢ radiation field. The photo-emfs developed across such barriers during irradiation in hole 5IN are shown as a function of integrated dose in Fig. 4.2.23. There is a considerable difference in the photo-emfs observed until after the integrated dose has reached about 1014 nvt, at whlch time ‘the barriers become comparable. Wil The change in collector characteristic with the emitter circuit open, as observed after successive irradiations, is shown in Fig. 4.2.24. As reported previously, this behavior compares with the char- acteristic of a diode biased in the reverse direction. After the third exposure, the sample was allowed to anneal for 18 hr at room temperature. The charac- teristic taken after the 18-hr annea! showed that the saturation current dropped back toward its original value and that the leakage current, as indicated by the slope of the curve, increased. The data agree with previous observations of annealing due to the electrostatic field of the barrier in that the saturation current returns toward its initial value, The change in slope or leakage current can be considered to be due to the same mechanism. Since the base material in this tran- sistor is n-type and the surface of all germanium crystals is p-type, there is a p-n junction between the bulk material and its surface. In the case of a p-n=p junction device, a p-type channe! occurs along the surface through which current may flow without crossing a barrier, as illustrated in Fig. 4.2.25. It would be expected that, as a pn-p device was irradiated and the n-type material changed toward p-type, the p-type surfoce layer would become thicker and more p-type. Both results would lower the resistance of this channel and lower the leakage resistance. It would be expected that an #n-p-n device, however, would require the current to cross a barrier in going from one n-type region to another, and thus no channel would be present. Consequently a p-n-p unit should be more sensitive to surface conditions than would an n-p-n unit, ) The change in the grounded-base characterlshc family of curves after an integrated dose of 9.96 x 1012 nvt, is shown in Fig. 4.2.26. Instecd of an increase in the collector characteristic, there is a decrease, and there is no detectable change in 019 A - gt PERIOD ENDING SEPTEMBER 10, 1956 UNCLASSIFIED ORNL-LR-DWG 45064 8 s ¢ \ P . N, - 7 : 2 2 2 2 e A — =—"-—=—.=-.—-_______ N o 6 5 - ® EMITTER - BASE - A COLLECTOR - BASE &4 | o - o a 3 2 1 0 0 0.5 1.0 1.5 2.0 2.5 30 3.5 4.0 a5 5.0 Fig. 4.2.23., Photo-emfs Integrated Neutron Flux. INTEGRATED NEUTRON FLUX (10" nyt) UNCLASSIFIED NUMBERS INDICATE EXPOSURE R \RRADATIONS 3 = AFTE COLLECTOR CURR T ma) ORNL-LR-DWG 152914 HOURS OF ANNEALING TER THIRD IRRADIATION -2 _ AETER 1 IRRADATION IRR o L . 0 02 04 06 08 . {0 ~ COLLECTOR BIAS (voits) - 1.2 14 Fig. 4.2.24. C-hungeé‘ in Collector Current Char- acteristic of @ Philco Type-L5106 Surface-Barrier Transistor as a Result of a Series of 2-min Ex- posures at a Flux of 8.3 x 10! nv . 0 & & e of a Philco Type-L.5104 Surface-Barrier Transistor as a Function of UNCLASSIFIED ORNL ~LR~DWG 15290 QL Lttt L ORI RN NN Y] Fig. 4.2.25. [llustration of Surface Leakage Mechanism, ‘ ' slope. The separation of the curves for different emitter currents is about the same as it was before bombardment, A circuit designed with normal engineering tolerances would still be operating providing there ‘had been no violent fluctuations of behavior during the bombardment, The smoothness of the transition is illustrated in Fig. 4.2.27, which shows the collector characteristic at 1-v bias at 255 ANP PROJECT PROGRESS REPORT UNCLASSIFIED ORNL-LR-DWG 15292 /g™ 5.85 ma w H L] BEFORE IRRADIATION <=—====AFTER IRRADIATION . COLLECTOR CURRENT {(ma) ™ 0 0.2 04 0.6 0.8 10 1.2 1.4 ‘ COLLECTOR BIAS (volts) Fig. 4.2.26. Change in Collector Characteristic of a Philco Type-L5106 Surface-Barrier Transistor as a Result of Irradiation to an Integrated Dose of 9.96 x 1012 nvt , different emitter currents as a function of the inte- grated dose. These curves are the results of measurements before and after irradiation, however, and may not be indicative of behavior in the presence of the radiation field. IRRADIATION EFFECTS ON THERMAL.- NEUTRON SHIELD MATERIALS J. G, Mergan R. M. Carroll M. T. Morgan P. E. Reagan The thermal-neutron shield for the ART should have a high absorption cross section, it should be compatible with adjacent reactor materials, and it should retain its bulk structure at elevated tem- peratures under irradiation. Gas release from the material should be at @ minimum. For temperatures up to 1600°F, two cermet shield materials, COB‘- Fe and BN-Ni, clad with stainless steel are being tested. These were fabricated!? by mixing the powdered components for 2 hr in an offset rotory blender, cold pressing the mixed powders at 33 tsi 175, H. Coobs, private communication to J. G. Morgan. 256 ok &> UNCLASSIFIED -3 ()é 107°) ORNL-LR-DWG 15063 TRANSISTOR: PHILCO TYPE L-510C 3 NO. A-1944 BIAS: { volt ROOM TEMPERATURE /p, COLLECTOR CURRENT (ma) 0 14 0 1 2 3 4 5 6 (x10') nvt, INTEGRATED NEUTRON FLUX _(neufrons/cmZ) 'Fig. 4.2.27. Changes in the Collector Current of a Philco Type-L5106 Surface-Barrier Transistor as a Function of Irradiation. ina2x 2'/4-in. steel die, sintering the compacts for 1 hr at 2000°F in o hydrogen atmosphere, coining the compacts at 33 tsi to 0,250 in. thick- ness in a 2 x 2'/4-in. steel die, encapsulating the coined compacts in type 304 stainless steel picture frames, hot rolling the composites ot 2000°F to a total reduction in thickness of 7:1, and machining 0.1875 x 0.500 x 0.050-in.-thick specimens from the rolled composite sheet. The powder particle sizes and the compositions of the cermet core materials are given in Toble 4.2.4, A CaB,-Fe sample was irradiated in helium in the LITR to a thermal-neutron dosage of 5.2 x 1017 nvt, The sample, sealed in quartz, was cocled by the LITR water. Metallographic examination!® of the specimens after irradiation (Fig. 4.2.28) showed that the clad-core interface had retained its con- tinvity, The core material was clad to ensure compatibility with the Inconel to be used as the structural material of the reactor, Many CaB crystals may be seen in the core matrix, ulihougtn some were removed in sample preparation. The samples shown in Fig. 4.2.28 are shown again in- Fig. 4.2.29 at a lower magnification so that the cladding on both faces of the sample may be seen. IaA. E. Richt, private communication to J. G. Morgan. it o o b n 0 °" PERIOD ENDING SEPTEMBER 10, 1956 TABLE 4.2.4. PARTICLE SIZES AND COMPOSITIONS OF CERMET SHIELD MATERIALS L Particle Size Cermet Weight of Component Volume of Component Component P ®) (%) (%) CaBj-Fe CaB, | 44 to 100 7.6 21.0 Fe 147 92.4 79.0 BN-Ni- BN 10 10.3 30 o Ni T 89.7 70 Fig. 4.2.28. Type 304 Stainfess Sfeej Clad -CcB -Fe Cermet Before and After liradiation in the There is no evidence of grb_ss cracking in the core or of deformation of the specimen, Because of the inhomogeneity of the core, hardness tests were’ inconclusive; however, a slight increase in hard- ness was observed, . Lo ~ Samples of BN-Ni are being irradicted in the “LITR for six weeks and will be examined. The specimens are being irradiated in the same facility “as that used for the CaBé-Fe specimens, _In addi- ae,I G o N LiITR toa Thermal-Neutron Dose of 5.2 x 109 nvt. (“), finirradiated. (b) Irradiated. 250X. Reduced 27%. tion, a hid"l-temperu_t'ure irradiation of the clad BN-Ni material is under way in C-48 of the LITR, The sample was sealed under vacuum in an Inconel capsule. A plctinum resistance heater is being used to control the temperature during the irradia- - tion at 1600°F. Gas formed as a result of the & irradiation will be removed and analyzed. Six samples of copper—boron carbide were pre- pared for MTR irradiation. The capsules will be 257 T & ANP PROJECT PROGRESS REPORT Fig. 4.2.29. Type 304 Stainless Steel Clad CaB,-Fe Cermet Spec.imens" Sfiown in Fig. 4.2.28 at @ Lower Magnification. Both cladding-cermet interfaces may be seen. () Unirradiated. (b) lrradiated. 75X. Reduced 27%. ' inserted in an *’A’’ piece and irradiated to a thermal-neutron dosage of 3 x 1020 nvt while the samples are held at a temperature of 1600°F. The instrumentation for the experiment has been con- structed and shipped to the site. The test will be operated to attain a maximum boron burnup equiva- lent to that expected in the ART.!? The samples are 0.1875 x 0.500 x 0.102 in. in thickness and clad on two faces.2? The composition is 6.6 wt % B ,C, =325 mesh, clad with a 3-mil copper diffusion barrier and 7 mils of type 430 stainless steel. The samples are mounted in two capsules, with three samples per capsule, as shown in Fig. 42.30. After irradiation, the samples will be returned to ORNL for postirradiation examination. Another promising neutron shield material is hot- pressed boron nitride. Boron contents of up to 194, M, Perry, private communication to J. G. Morgan, 20y, R, D*Amore, private communication to J. G. Morgan. - 258 -3 O I 44 wt % can be obtained with a density of 2,1 g/cm3 in a molded and self-bonded structure. Boron nitride has a crystalline structure similar to that of graphite and directionalism is present in all physical properties,?! especially thermal expan- sion. Pure (98% and above) boron nitride powder was obtained from two sources, and complete analyses were made. Bodies from these powders will be fabricated for irradiation studies. IRRADIATIONS OF STRESSED SHIELD ING MATERIALS Jo Co Wilson ' o W. E. Brundage W. W. Davis Testing equipment is nearly completed and cali- brated for the irradiation of a 1 wt % boron (B'%)— stainless steel alloy under stress in the LITR, 21k, M. Taylor, Ind. Eng. Chem. 47, 206 (1955). o ) W - PERIOD ENDING SEPTEMBER 10, 1956 UNCLASSIFIED PHQOTO 18123 UNCLASSIFIED PHOTO 18124 {c) () Fig. 4.2.30. Equipment for the lrradiation of Clad Cermet Shielding Materials in the LITR. (a) Spec- imens - in support-grid assembly with thermocouples attached. (b} Inner capsule in which the support- grid assembly will be weld-sealed in @ helium atmosphere. The ceramic hermetic seal through which the thermocouple leads are brought out may be seen at the right. (c) Electrical heater made of platinum coils in an .alumina form and stainless steel can in which mner capsule will be inserted. (d) Aluminum container for assembly.. 259 (o) ) ANP PROJECT PROGRESS REPORT Creep in compression of the specimens will be measured during the tests by a Bourdons-tube de- flectometer, Short-time compression fests on copper—boron carbide cylinders were run at elevated temperc- tures. A special setup in an Instron testing machine permitted load-deformation curves to be recorded directly, Yield stresses were determined 260 for copper and copper plus 20 and 40 wt % B,C at 1300°F. Creep tests were run at the same tem- perature at a stress of 500 psi. Unloading and reloading of short-time creep specimens improved resistance to deformation., This suggests that hot pressing or hot working improves the sirength of the materials. Creep tests were also run at 1450 and 1500°F at a stress of 50 psi, /r’{ %‘} Ack PERIOD ENDING SEPTEMBER 10, 1956 4,3, FUEL RECOVERY AND REPROCESSING R. B. Lindauer D. E. Ferguson W, K. Eister H. E. Goeller VOLATILITY PILOT PLANT DESIGN AND CONSTRUCTION F. N. Browder W. H. Carr W. H. Lewis R, P. Milford Construction of the fused salt—fluoride volatility pilot plant is essentially complete and shakedown tests have been started. Studies have indicated that cooling the UF, to 0°C while the product cylinder is still connected to the recovery cold traps will significantly reduce the UF, holdup in the cold traps after each run. Equipment is there- fore being designed and procured for circulating a blast of cold dry air or nitrogen through the annulus between the product cylinder and its heater after most of the UF, has drained as a liquid into the cylinder. Freon 12 will be used in a direct-expansion extended-surface coil to chill the coolant gas. TABLE 4.3.1. DIAMETER MEASUREMENTS " A project engin€ering report covering the design and construction of the pilot plant is being pre- pared. ENGINEERING DEVELOPMENT J. T. Leng S. H. Stainker Engineering development work has been chiefly concerned with demonstrations of the reliability of freeze valves to be used in the molten salt lines. A freeze valve of the type to be used in the pifot plant showed no leakage in 120 pressure tests with 20-psi N, for 20 min. This valve also showed no leakage that could be detected with a halide leak detector in a test with 100-psi N, and in a test with Freon. Diometer measurements showed slight distortion of the pipe after 99 pres- sure tests (Table 4.3.1). A stub for the junction of electric cable to self- resistance-heated salt transfer lines was designed which allows the temperature of the piping to be OF PIPE IN A MOLTEN SALT FREEZE VALVE BEFORE AND AFTER PRESSURE TESTS Tests carried out in elliptical loop, 9 in. on long diameter, with a 3«in., radius of curvature at the ends, made of %-in. sched-40 Inconel pipe; 2-in. layer of insulation around valve Heating cycle: 200 to 650°C Heating current through pipe: 250 amp Molten salt: NoF-ZrF4-UF4 (56-40-4 mole %) Orientation of - Outside Diameter of Pipe {in.) Point on Loo Measurement with. - — - - P - Respect to Plane Be&_'"f o After 70 ~ After 99 ‘of Loop _ Teshng : . Prgs;ure Tesisr_ - Pressure Tests “9o'clock” Parallel . 0.840 0842 0.845 v .~ Transverse 0.840 0.840 : 0.835 “go’clock® Parallel 0,780 . 0.780 0.780 Transverse 0.880 0.884 0.888 “3e0’clock"” " Parallel 1 0.850 ~ 0.850 0.850 Transverse 0.840 » 0.840 0.840 261 ANP PROJECT PROGRESS REPORT above 525°C while the temperature of the connector to the copper cable is below 85°C, The stub con- sists of 6 in. of sched-40 Inconel pipe welded to the salt transfer line, with 6 in. of 1-in. nickel bar butt-welded to the Inconel pipe. The nickel bar is covered with a 2-in. layer of Superex insula- tion to within 2 in. of the Inconel-nickel junction, The temperature of electrical insulating gaskets in the freeze valve vent lines must be maintained at or below 140°C, It was found experimentally that the temperature of the piping at the location of the gaskets would be 300°C but that two 6-in.- dia copper fins spaced 1 in. apart and silver- soldered to the vent lines just above the vent furnace would reduce the temperature at the gasket location to the required 140°C, A device for sampling the molten salt (Fig. 4.3.1) was tested and found to be suitable for pilot plant use., A minimum sample of 2 g was desired, and it was found to be possible to obtain a 2.9-g sample. The sample cup used for the tests was fabricated of low-density graphite, and some difficulty was experienced in removing the salt from the pores of the graphite. A high-density-graphite sample -cup will therefore be tested in an attempt to improve recovery of the sample. CHEMICAL DEVELOPMENT M. R. Bennett G. l. Cathers R. L, Jolley A study of the decomposition of the UF («3NaF complex at temperatures of 245°C and h|d1er has confirmed the belief that uranium retention ort the NaF bed will be excessive in the NoF desorption step if the temperature and sweep gas flow rate are not properly controlled. The retention results from decomposition of the UF6-3NaF complex to a complex of NaF with UF,, which is not volatile, Maximum decomposition rotes of about 0.01, 0.09 and 0.5% per minute would be incurred at 250, 300 and 350°C, respectively, in the absence of fluorine if all the vranium were in the form of the solid complex UF_,.3NaF. Under optimum conditions, UF, desorption from the NaF bed! competes favor- ably with the decomposition effect, and the final uranium retention is small. The presence of fluorine appears to be essential in inhibiting the decomposition reaction and, 1H, K. Jackson et al., ANP Quar. Prog. Rep. March 262 UNCLASSIFIED ORNL~-LR-DWG 15857 SET SCREWS ROD (¥g-in. DIA, 23/, in. LONG) CONNECTOR TO OUT- SIDE OF SHIELD = PIN VISE WIRE ROD (%—m DiA, 7/gin. LONG) 0.272-in. ID— HIGH-DENSITY GRAPHITE ROD ( %gin. DIA , 2in.LONG} ~er AT T T HIGH-DENSITY GRAPHITE |* Hi€ ROD ( Ya=in. DIA, Y in. LONG) —pmtfi? ROD {¥g-in. DIA, t in. LONG) SAMPLE CUP Fig. 4.3.1. Molten-Salt Sampler for Volatility Pilot Plant, possibly, in promoting refluorination of the non- volatile compound formed by the decomposition. The possibility of uranium retention by the decom- position mechanism in the absorption step at 100°C appears to be insignificant, even over extended periods of time. The dependence of the rate of decomposition of UF ;«3NaF was determined in a series of runs over the temperature range 245 to 355°C (Fig. 4.3. 2) The reaction involved is probably UF ;-3NaF (solid) —> UF ;+xNaF (solid) + + 0.5F, {gas) x'fil .4 - FRACTIONAL DECOMPOSITION RATE, 7 {min™') 3 " UNCLASSIFIED . é; 1077) ORNL-LR-DWG 16495 50 log r = 6.09— (5.22 x (0%T) N Q0 wn S5 e AT 18 1.9 © 20 {(x107%) rir S Fig. 43.2, Dependence of Rate of Decompo- sition of UF ;«3NaF on Tempemiure, S The dependence of the decomposctlon rufe on tem- - percture is gwen by the expressnon SR Iogr = 6.09 - (5.22 X, 103/'1') v _i o ",where r.is fbe fracflonnl decomposmon rote in - reciprocal minutes and T is the cbsolute fempera- - ~.ture. - The rate was’ calculnfed ‘on the bas:s of an- 1 -_absorphon cupocuty of 1.33 g of UF ‘pet gram of NaF, The energy of activation was calculcfed as - 'a . 423.9 keal/mole of UF, o3NaF comp!ex.., Atist C possibly slgmflccnt thaf fl'IlS energy change is "_}- PERIOD ENDING SEPTEMBER 10, 1956 opproximately the same as the enthalpy change of +23.2 kcal/mole involved in the volcatlllzahon of UF from the UF ' 3NaF complex. The decomposmon data were obtained with 4- to 9-g samples of NuF (Harshaw grcde, classified to 12-20 mesh) held in a U-tube (/ in.~dia stainless steel tubing), through which gaseous UF, was passed at atmospheric pressure. An oil bath was used for manually controlling the temperature to 13°C during the course of each experiment, The runs were terminated by removing the oil bath and cooling the sample rapidly, The UF ¢*3NaF was formed at the beginning by saturating the NaF at a high UF ; flow rate, after which it was decreased to a rate of 0,1 to 1 g/min for the remaining time, The lengths of the runs at various temperatures were adjusted to obtain a uranium(V) content in the final product of 1 to 10%. The excess UF, still _absorbed on the NaF at the end of each test was not desorbed because of the difficulty of achieving this without changing the uranium(V) content, The umeunt of UF ¢*3NaF complex affected by the re- action was determined from uranium(V) and uranium(Vl) analyses. The temperature was not increased to above 355°C, since at the higher temperatures it would have been difficult to main- tain saturation of the NaF with the UF . without the use of a pressurized system, In prellmmary work on the decomposition reaction - it was determined that a uranium(V) content of 20 to 26% represented a limit which could not be ~exceeded in one cycle of saturation of the NaF with UF at 100°C followed by heating as a closed - system fo 350 10 400°C. Usuadlly in these runs the NaF: weight increase corresponded closely to the " estimate based on the assumption that the decom- “position product is a complex of UF, with NoF, Xeray - crystallography ‘data mdlcated that the '-'UFsoxNaF complex in the decomposmon product has an orthorhombic structure with ce!l dimensions ag = 4904, by = 547 &, and ¢ = 3.87 A The x-ray ‘patterns of y- and B-UF; were not observed in the material. 263 ANP PROJECT PROGRESS REPORT 4.4, CRITICAL EXPERIMENTS A. D- C(I“ihan REFLECTOR-MODERATED REACTOR EXPERIMENTS D. Scott E. Demski D. E. McCarty W. J. Fader! "W, C. Tunnell D. A, Harvey! E. V. Sandin! Jo Jo Lynn Experiments ot Room Temperatute One of the room-temperature critical assembiies representing the reflector-moderated reactor with circulating fuel has been reconstructed in order to extend the earlier studies in this program.2 The fuel region contains olternate lamince of Teflon and enriched uranium foil. In this particular varia- tion, assembly CA-22.2, one of the fuel end ducts is somewhat thicker than the other to allow fUrfher investigation of end-duct thickness as a variable,> The array was critical with a loading of 23.4 kg of U235 with about 0.5% excess reactivity. The estimated critical mass of 22.5 kg is in good agree- ment with the value® of 23.3 + 1,0 kg, which was based on the excess reactivity observed in the previous loading of 28.35 kg in the same geometry. Reflector Evaluation. = The effect on reactivity of replacing a part of the outermost layer of the beryllium reflector with stainless steel was meas- ured. This peripheral layer, which is approximately cylindrical and is separated by 8/ in. of beryllium from the outer core shell, is 2% m. thick and 19% in. long and is centered about the mid-plane. Re- moval of 27.3% of the beryllium in this layer gave a 45.9¢ loss in reactivity, whereas the addition of stainless steel increased the reactivity 16.7¢. The beryllium is, therefore, 2.75 times more effective than stainless steel in this region. Reactivity Coefficients. — The reactivity co- efficients of a number of materials of engineering interest were evaluated at various positions along 10n assignment from Pratt & Whitney Aircraft. 2. D. Callihan et al., ANP Quar. Prog. Rep. Sept. 10, 1955, ORNL-1947, p 58. , 3p. Scott, J. J. Lynn, E. V. Sandin, S. Snyder, Three Region Reflector Moderated Critical Assembly with End Ducts — Experimental Results with CA-22, Enlarged End Duct Modzfzcanon. ORNL CF-56-1-96; see also A. D. Calllhcn, op. cit,, p 59. 264 the radius at the mid-plape of the reactor, and the results are given in Figs. 4.4.1 and 4.4.2, in each case the value given is the change in reactivity introduced by filling a void with the material. An etror appeared in the results of the beryllium reactivity coefficient measurements in the compact- | core reflector-moderated reactor experiments re- ported previously.4 The error resulted from a mistake in dimensioning a sample, and the cor- rected data are shown in Fig. 4.4.3. Experiments at Elevated Temperature A number of design features and nuclear charac- teristics of the circulating-fuel reflector-moderated reactor being designed by Pratt & Whitey Aircraft - are to be investigated in a critical experiment to be performed at ORNL late in 1956. The equipment and the experimental program are being prepared in close collaboration with Pratt & Whitney Aircraft personnel, who are fabricating most of the com- ponents for assembly at ORNL, The experiment will be performed at about 1300°F, and a molten U235.enriched NaF-ZrF ,-UF , mixture will be used as the fuel. This salt mixture will not be circu- lated, but it will be transferred pneumatically from the storage reservoir to the annulor fuel region. The purposes of the experiment are fo determine the critical uranium concentration, the effectiveness of control materiais, and an over-all temperature coefficient. A limited number of neutron-flux and fission-rate distributions will be measured with appropriate detectors. The design of this experiment follows quite closely that of the ART high-temperature critig:al test performed last year and reported previously,? although the dimensions and shape of the core region will be somewhat different, The central - beryllium island of the critical assembly (Fig. 4,4.4) will be a right circular cylinder (8.0-in.-dia) enclosed in a Hastelloy X shell (8.5-|n.-OD Zmine= thick)y A 2.94-in.-dia axial hole in the lsiand lined with a 0.12-in.-thick Hastelloy X thimble (2.88-in.«0D), will accommodate the control" assembly. The poison section of the control rod, which will be enclosed in Inconel, will be an 4A. D, Callihan et al, ANP Quar. Pfog. Rep. Marcb 10. 19569 0RN|—‘206], P 68' F'gc 3-6. ¥ ) H.oy -, » w [*J - PERIOD ENDING SEPTEMBER 10, 1956 “SeerrT ORNL-LR-DWG 16196 40 1o LiF, 2.689 OF Li, 2% x1%¢ x 0.20 in. « ‘ ® 20% Li—80% Mg ALLOY, 0.328 g OF Li, 2% x 17 x 0.02 in. 8 16 SAMPLE POSITION; RADIAL DISTANCE FROM AXIS {in.) » § A INCONEL, 56.28¢, 2 x1%¢ x»O.lOB in. w & SODIUM, 122 g, 2Va x 2% x1in, g s < o W SO s g FUEL REGION BERYLLIUM REFLECTOR T} : - - @ L a a = < : > - 5 = DIVIDE SCALE BY 100 E‘ : . Z f\ ‘ 2 / 5 / 4 . ,/\\ 10 15 20 25 Fig. 4, 4.1. Reactivity Changes Effected by the Addition of Vanous Materials to the RMR Critical Assembly CA-22-2 Boener 8 ORNL~-LR-DWG 16197 e DT 4 U(9347% UZ3%), 1.31g, 2% x 76 x 0.001 in. 7 \ - o A U(93.17% UZ33),43.81q, 4 x 24 x 0.004 in. ' e & BERYLLIUM, 254g, 273 x 27 x¥in. : % _ o BERYLLIUM, 1259, 2% x1%g x1in. | o L ) . - B i -t Q. = . o “ g — , : S - ' FUEL REGION © " BERYLLIUM REFLECTOR . E - T & .a '>: 7.-. - =3 \ -_ § | owioe scALE A\ W .| BY 100 -\ ® N z 2 \ . PLoga o AN ‘> |, NN | /l a DIVIDE SCALE \ DIVIDE SCALE BY 10 & BY 10 / Zz 10 AN 2 \_J/ \\ < . o 5 : i ) \.., 0 o : 5 10 15 20 25 SAMPLE POSITION; RADIAL DISTANCE FROM AXIS (in.) Fig. 4.4.3. Reactivity Changes Effected by the Addition of Various Materials in the Compact-Core RMR Critical Assembly. annulus (2.46-in.-0D, 1.97-in.-ID} made of a mix- ture of 30% rare earth oxides (samarium and gadolinium) and 70% nickel. The outer core shell, also of Hastelloy X, will have an inside diameter varying from 14.85 to 21.05 in. The wall thickness will be 0,16 in. at the center, increasing to 0.25 in. at the top and bottom. The bottom of the fuel section will be flared to represent the region in the power reactor where the fuel will be directed radially ocutward and upward into the heat ex- changers. The reflector will be built up of 13 annular slabs of beryllium with outside diameters varying from 22,5 in. at the top to 48.0 in. ot the mid-plane and 32.0 in. at the bottom. The central holes will be ‘tapered to conform approximately with the shape of the outer core shell. The beryllium refiector, therefore, will be slightly over 13 in. thick for @ distance of 8 in. on each side of the mid-plane and will decrease to 3.3 in. at the top and 6.5 in. at the bottom. To mock up the coolant passages, 832 vertical holes, 0.25 in. in diameter, will be 266 drilled through the reflector, and 125 holes, also 0.25 in. in diameter, will be drilled through the island. The coolent which would normally occupy these holes was not mocked up because of opera- tional and safety considerations. Fast-neutron leakage from the entrance and exit ducts of the power reactor is to be reduced by suppressing the fission rates in these regions. In current designs, this suppression is to be accom- plished by shielding the fuel from neutrons re- flected by the beryllium. In the critical experiment, shields composed of a mixture of boron. carbide and copper will be placed in the beryllium regions at the bottom of the reactor adjacent to both the core shells. In similar locations at the top of the reactor, cylindrical shelis of B19, 0.1 in. thick, in Hastelloy X annular.containers, can be inserted vertically to a position 6 in. below the top of the _ beryllium. The volume of the reflector, island, and fuel region will be 36.7 3, and the volume of the fuel region alone will be 5.3 3, C 575, || 481n.. 'PERIOD ENDING SEPTEMBER 10, 1956 Eenee ORNL-LR-DWG 16199 ROD THIMBLE 62.07 In. STAINLESS STEEL [ HELIUM | ATMOSPHERE - 3/8 in, FUEL ANNULUS BERYLLIUM REFLECTOR FUEL ANNULUS - . 18.0in. 4. DETECTOR THIMBLES STAINLESS STEEL ‘8-Cu Fig. 4.4.4. Schematic Diagram of High-Temperature RMR Critical Assembly. 267 ‘.) 5.1. SHIELDING THEORY A MONTE CARLO STUDY OF THE GAMMA-RAY ENERGY FLUX, DOSE RATE, AND BUILDUP FACTORS IN'A LEAD-WATER SLAB SHIELD OF FINITE THICKNESS S. Auslender! The gamma-ray energy flux, dose rate, and buildup factors in a lead-water shield of finite thickness have been calculated by a Monte Carlo method.? The calculation included 1-, 3-, ond 6-Mev photons incident on the slab both along a normal and at an angle of 60 deg. {Calculations for an angle of 75Y, deg were also performed, but they are not included here since the attenuation was quite large in spite of the large buildup fczctor. For the same reason the results for 1-Mev photons incident at an angle of 60 deg were also omitted,) The buildup factors for energy and dose obtained in this calculation were compared with those obtained by the use of the moments method® for monoenergetic, plane monodirectional sources normally . incident upon a semi-infinite, homogeneous medium, The thicknesses of the lead-water shield in ‘centimeters and in-mean free paths for the various incident gamma-ray . energies . are given in Table 5.1.1, and the tissue dose equivalents for the incident fluxes (no shield present) are given in Table 5.]-20 IOn asmgnment from Pratt & Whitney Alrcraft. 2g, Auslender, ANP Quar. Prog. Rep, March 10, -1956. ORNL-2061, p 223. 3H, Goldstein -and 'J. E. Wilkins, Jr., C'alculatzons of the Peretration of Gamma Rays. Fmal Report. NYO- .3075 (June 30, ]954) _ . 7 o TABLE 5.1 'I. NORMAL THICKNESSES OF A LEAD-WATER SLAB SHIELD Thn:kness - e Mecm Free Paths (mfp) . . Region™ o o ’ sre For 'I-Mev For 3Mev For 6-Mev- o Phofons - Photons Phctons P, ss" _9.0739 S 5';'5_'54?: 5.878 _ ; Hzol, 3581 2502 1382 . 096 Total 47.40 11.630 7.035 6.874 TABLE 5.1.2. TISSUE DOSE EQUIVALENT FOR INCIDENT FLUX Photon Energy ( mr /hr Do/q5 1 (Mev) . y/cmzo sec 1 _ - 0.001923 3 0.004368 6 0.007206 The results of the calculations for 1-Mev photons are presented in Figs. 5.1.1 through 5.1.4, those for 3-Mev photons in Figs. 5.1.5 through 5,1.10, and those for 6-Mev photons in Figs. 5,1.11 through 5.1.16. The normalized energy fluxes for the 1-, 3-, and 6-Mev energy groups are plotted in Figs. 5.1.1, 5.1.5, and 5.1.11, respectively, as functions of the normal thickness of the shield (in centimeters). The uncollided energy flux, normalized to unity at the initial boundary, is also plotted. The third curve in each figure is the energy buildup factor, Bp, which is the ratio of the two energy flux curves: bg/E oby ¢k BE = e—t/h . qusle—t/h where , _, 5 .qSE(Mev/cm?osec) = energy flux at the point of _ - . interest in the shield, Ed(Mev/y) ‘= initial photon energy, - -'lc;Sl(y'S/CmZ-Sec')' :number flux incident von . _sh:eld o ’dlstance befween the mmul ' - t{em) .~ _boundary and the :point of e ‘--.,m?erest in the shleld Alem) = relaxahon Iength e“t/A= :'uncolllded energy flux at o ‘the pomf of mteresf In a s:mllar manner the normuhzed dose rates and dose buildup factors, B, for the three energies 271 ANP PROJECT PROGRESS REPORT are plotted in Figs. 5.1.3, 5.1.7, 5.1.9, 5.1.13, and 5.1.15. Here D/¢y D B = = ’ Doe =lse¢e G/A/‘#l Doe-t sec 8/A where D(mr/hr) = dose rate (tissue) at the point of interest, Dy(mr/hr). = dose rate (tissue) of the uncollided incident photons at the initial boundary. The various energy buildup factors as functions of oblique thickness, in mean free paths, are pre- sented in Figs. 5.1.2, 51.6, oand 5.1.12. Corre- sponding dose buildup factors are given in Figs. 5. 1.4, 5.1.8, 5.1.10, 5.1.14, and 51.16, Data from the resuvits of the moments method solutiond are also plotted for purposes of comparison, although it must be remembered that those calculations were for infinite homogeneous media and are not directly comparable to the present calculations for a finite two-region slab. The calculations for lead do agree reasonably well for the first few reloxation lengths where the effects stemming from the dissimilarity of the slabs should be least. The dose rate and dose buildup factors resulting from an earlier Monte Carlo calculation for 3-Mev photons normally incident upon a one-region 8-mfp-~ thick lead shield® are also presented (Figs. 5.1.17 and 5.1.18) as an example of a case intermediate between the Monte Corlo calculation for a two- region finite shield oand the moments method solu- tion for the one-region semi-infinite shield. A comparison of the dose rate in the one-region finite shield with the dose rate in the one-region semi-infinite shield (calculated from buildup fac- tors reported in ref 3) shows good agreement. The characteristic dip near the final boundary of the finite lead shield is apporent. The dose buildup factors for the 3-Mev incident photons in the two-region and one-region finite shields (Figs. 5.1.8 end 5.1.18) are identical to within 2 mfp of the lead-water interface. It is apparent that at this energy and angle of incidence the back scattering is important in lead for a distance of about 4 em (Ap, = 2,05 em at 3 Mev). The scattering in the water tends to compensate 4s, Auslender, unpublished work. 272 G644 for the finite thickness of the lead-water slab, It can also be concluded that the back-scattered flux at that interface is about 12% of the total flux and about 20% of the scattered flux. The dif- ferences for the finite and the semi-infinite one- region shields (Fig. 5.1.18) are particlly due to differences in the cross-section data used for the calculations. An extensive discussion of the errors in the calculations from the moments method is included in ref 3, Consistent errors may exist that will bias the answers obtained with that method. The Monte Carlo data is much less sub- ject to bias thon to statistical fluctuation. (The large fluctuation of the data in Fig. 51.9 may indicate ‘a relatively large error, but this is stlll probably less than 10%.) Since the Monte Carlo calculations for the two- region finite shield and the moments method solu- tion for the one-region semi-infinite shield were in general agreement, the striking differences in the buildup factors near the lead-water interface and near the final boundary should be enlightening, In Figs. 5.1.2 and 5.1.4, for example, the energy and dose buildup factors for 1-Mev photons in a finite two-tegion shield increase sharply in water to a valve about half way between those for semi- infinite one-region shields of water and lead. At this energy the finite water region is 2‘/2 mfp thick, and, at first, the buildup increoses almost at the same rate as it does initially in the semi-infinite water shield. This confirms that at this energy, which is below the minimum in the total cross section, the uncollided radiation dominates in the penetration of the lead, as it should. In Figs. 51.6 and 51.8 the increase of the buildup factors in the water is very small, since, for the primary 3-Mev photons for both lead and water, the major portion of the total cross section is attributable to scattering. Hence, the difference in the behavior of the flux in water and in lead is due to that small portion of the flux which is absorbed in the lead but is scattered in the water, For 6-Mev photons (Figs. 5.1.12 and 5.1.14) the absorption cross section of lead is no longer so small that the absorption in it is almost negligible. In this case the buildup factor increases rather abruptly in the water, almost to the buildup factor for 6'/ mfp of water calculated by using the moments method, It is evident from the results of the calculations for radiation incident at 60 deg (Figs. 5.1.9, 5.1.10, 11 Part 5 REACTOR SHIELDING E. P, Blizard 5.1.15, and 5.1,16) that the practice of using only normal incidence data for shield designs can lead to a poor approximation. This problem becomes most acute when the number of meon free paths across the shield is small or when the angular distribution is such that a large portion of the UNCLASSIFIED - 2=-01-059—104 EO = INITIAL PHOTON ENERGY = INCIDENT NUMBER FLUX & = ENERGY FLUX Apy = 4.37cm Ay, 0 = 14.09cm By = ENERGY BUILOUP FACTOR ‘o 10 20 30 a0 50 60 1. NORMAL THICKNESS (cm} Fig. 5.1.1. Gamma-Ray Energy Flux and Energy Buildup Factor as a Function of the Normal Thick- ness (Centimeters) of a Finite Lead-Water Slab’ Shield: 1-Mev Normally Incident Photons. iy s boma PERIOD ENDING SEPTEMBER 10, 1956 radiation is not normal or nearly normal to the slab. Figure 5.1.16, which shows the flux de- pression near the initial boundary and the large buildup factor into the slab, is an excellent ex- ample of streaming (or short-circuiting). UNCLASSIFIED . 201~ 059-102 L | — MONTE CARLO, FINITE SLAB © H,0] MOMENTS METHQD, = SEMI- INFINITE SLAB, NORMAL INCIDENCE 2 8 Pb ) (Ref.: NYO-3075) < [+ 4 e W £ 0 ° o > Pb H,0 = o 2 2 3 / § 5 / Ldpeor = 1.63 mfp o / z ] . K / © § 1 » 2 ° /' ./ 0 2 a4 6 8 10 12 14 6 fof» OBLIQUE THICKNESS (mfp} Fig. 5.1.2. Gamma-Ray Energy Buildup Facter as a Function of the Oblique Thickness (Mzan Free Paths) of a Finite Lead-Water Slab Shield: 1-Mev Normally Incident Photons. 273 ANP PROJECT PROGRESS REPORT UNCLASSIFED 2-01—059—103 O = DOSE RATE O = INITIAL DOSE RATE INCIDENT NUMBER FLUX Do/, = 0001323 (e /he)/y/eme-sec) App ™ £.37cm My,0 = $4.09 cm &, = DOSE BUILDUP FACTOR 0 10 20 30 40 50 €0 7, NORMAL THICKNESS (cm) Figo 50‘03. 7 Gamma-Ray Dose Rate and Dose Buildup Factor as a Function of the Normal Thick- ness (Centimeters) of a Finite Lead-Water Slab Shield: 1-Mev Normally Incident Photons. UNCLASSIFIED 2-01-059-104 —— MONTE CARLO, FINITE o H,0] MOMENTS METHOD, SEMI- INFINITE SLAB NORMAL INCIDENCE ® Pb ) (Ref.: NYO-3075) 8,, DOSE BUIL JUP FACTOR 0 2 4 6 8 0 {2 14 46 #o”» OBLIQUE THICKNESS (mfp) Fig. 5.1.4. Gamma-Ray Dose Buildup Factor as a Function of Obligue Thickness (Mean Free Paths) of a Finite Lead-Water Slab Shield: 1-Mev Normally Incident Photons. 274 UNCLASSIFIED 2-01-059-105 = {NITIAL ¢, =INCIDENT NUMBER FLUX ¢ = ENERGY FLUX kpp= 2.05¢cm Aun=25.91cm 8 = ENERGY BUILDUP FACTOR 102 1073 0 10 20 30 40 30 60 f, NORMAL THICKNESS {cm) Fig. 5.1.5. Gamma-Ray Emergy Flux and Energy Buildup Factor as a Function of the Normal Thick- ness (Centimeters) of a Finite Lead-Water Slab Shield: 3-Mev Normally Incident Photons. UNCLASSIFIED 2-01-059-106 — MONTE CARLO, FINITE SLAB © H,0] MOMENTS METHOD, SEMI—INFINITE SLAB, NORMAL INCIDENCE ¢ Pb | ‘(Ref.: NYO—3075) B, ENERGY BUILDUP FACTOR o -2 4 6 8 10 12 14 16 por+ OBLIQUE THICKNESS (mfp) Fig. 5.1.6. Gamma-Ray Energy Buildup Factor as a Function of the Oblique Thickness (Mean Free Paths) of a Finite Lead-Water Slab Shield: 3-Mev Normally Incident Photons. 014 UNCLASSIFIED 2-0¢-059-107 O = INITIAL DOSE RATE ¢; = INCIDENT NUMBER = 0.004368 (mr, Apy = 2.05¢m Ayyo = 25.91cm 8, = DOSE BUILOUP FACTOR 0 10 20 30 40 30 €0 ' #, NORMAL, THICKNESS {cm} Fig. 5.1.7. Gomma-Ray Dose Reate and Dose Bmldup Factor as a Function of the Normal Thick- ness (Centimeters) of a Finite Lead-Water Slab Shield: 3-Mev Normally Incident Photons. - o : UNCLASSIFIED 2-01~059-108 i i = MONTE CARLO, FINITE SLAB - . o ° HZO MOMENTS METHOD, SEMI- INFINITE SLAB NORMAL INCIDENCE 2 ® Pb. ) (Ref.: NYO-3075) S 8;, DOSE BUILDUP FACTOR o _‘-aj_'_;'-4 6 B, 10 12 e L p.,r oauouETmcxNEssfmfp) Lo a Function of Oblique Thickness (Mean Free Paths) of a Finite Lead-Water Slab Shield: 3-Mev Normully incident Photons, 015 - Ba4 PERIOD ENDING SEPTEMBER 10, 1956 UNCLASSIFIED 2-01-059-109 1072 -3 10 10 D = DOSE RATE Dg= INITIAL DOSE RATE 4, = INCIDENT NUMBER FLUX 105 0y /%, = 6.004368 {mr/hr)fly scm?- w0 App= 2.05¢cm ngO' 25.9t cm 8,= DOSE BUILDUP FACTOR 10" o5 10° w08 o/%, 10° w07 —24/) / /4,[ 0w '0-8 . w0 ® 0 {0 20 30 40 50 60 £, NORMAL THICKNESS (cm) - Fig. 5.1.9. Gamma-Ray Dose Rate and Dose Buildup Factor as ¢ Function of the Normal Thick- ness (Centimeters) of a Finite Lead-Water Slab Shield: 3-Mev Photons Incident at a 6§0-deg Angle. UNCLASSIFIED 2-01-059—110 102 © H,0 | MOMENTS METHOD, - SEMI-INFINITE SLAB, 5 NORMAL INCIDENCE ® Pb ] (Ref.: NYO-3075) S N 8, DOSE BUILDUP FACTOR o 1 - o 2 4 € . 8 o 2 14 16 #ot, OBLIQUE THICKNESS (mfp) Fig. 5. 1.8., Gnmmu-Ruy Dose Buildup Fuctor us‘ -Fig. 5.1.10. Gamma-Ray Dose Buildup Factor as ¢ Function of Obliqgue Thickness (Mean Free Paths) of a Finite Lead-Water Slab Shield: 3-Mev Photons Incident at ¢ 60-deg Angle. 275 ANP PROJECT PROGRESS REPORT I.IICLAgSl:lED 107 gl 10 5 5 2 2 - £4 = INITIAL PHOTON ENERGY ' ¢, = INCIDENT NUMBER FLUX 5 & = ENERGY FLUX 2 Ay, = 1.97em 2 108 =L Mo = 3596 cm - 1ot F= 5, = ENERGY BUILDUP FACTOR g 5 0 10 20 30 40 50 60 7, NORMAL THICKNESS (cm) Fig. 5.1.11. Gamma-Rey Energy Flux and Energy Buildup Factor as a Function of the Normal Thick- ness (Centimeters) of ¢ Finite Lead-Water Slob Shield: &-Mev Normally Incident Photons. UNCLASSIFIED 2-04-059—-142 —= MONTE CARLO, FINITE SLAB © H,0] MOMENTS METHOD, SEMi— INFINITE SLAB), NORMAL INCIDENCE * Pb (Ref.: NYO—-3075) 8., ENERGY BUILDUP FACTOR 0 2 4 6 8 10 12 14 16 RBo’ s OBLIQUE THICKNESS (mfp) Fig. 5.1.12. Gamma-Ray Energy Buildup Factor as a Function of the Oblique Thickness (Mean Free Paths) of a Finite Lead-Water Slab Shield: 6-Mev Normally Incident Photons. 276 UNGLASSIFIED 2-01-059-43 _f/y¢! &= DOSE RATE Op =INFIAL DOSE RATE _ #, =INCIDENT NUMSER FLUX 1078 4, /$, =0.007206 (me/hr)/(y 7em®. sec) Apy=1.97cm Hgo = 3596 cm 8,= DOSE BUILDUP FACTOR A 1077 0 10 20 30 40 50 60 #, NORMAL THICKNESS (cm} Fig. 5.1.13. Gomma-Ray Dose Rate and Dose Buildup Factor s a Function of the Normal Thick- ness (Centimeters) of a Finite Lead-Water Slab Shield: 6-Mev Normally Incident Photons. UNCLASSIFIED 2-01—-059~1{14 == MONTE CARLC, FINITE o H,0 MOMENTS METHOD, SEMi-INFINITE SLAB, NORMAL INCIDENCE ® Pb J (Ref.. NYO-3075) 8,, DOSE BUILDUP FACTOR o 2 4 € 8 10 12 14 16 #of, OBLIQUE THICKNESS (mfp) Fig. 5.1.14. Gamma-Ray Dose Buildup Factor as a Function of Oblique Thickness (Mean Free Paths) of a Finite Lead-Water Slab Sl-ueld 6-Nbv Normully Incident Photons. N Y oy e "\ Fig. 5.1.16. Gamma-Ray UNCLASSIFIED 2-01<059-115 1072 D = DOSE RATE Do = INITIAL DOSE RATE 7 = INCIDENT NUMBER FLUX 1073 By /¢y = 0.007208 tme/hri/iy femP.sec) 5 Apy = 1.97cm )LHao = 35.96 ¢m & = DOSE BUILDUP FACTOR o 10 20 30 40 % NORMAL THICKNESS {cm) 50 60 Fig. 5.1.15. Gamma-Ray Dose Rate and Dose Buildup Facfor as a Function of the Normal Thick- ness (Centimeters) of a. Finite Lead-Water Slab Shield: 6-Mev Photons Incident at a §0-deg Angle. UNCLASSIFIED 2-01~-059-116 == MONTE CARLQ, FINITE SLAB 9 Hy0f MOMENTS METHOD, SEMI-INFINITE SLAB, NORMAL INCIDENCE ® Pb | (Ref.: NYO-3075) 8-, DOSE BUILDUP FACTOR 0o 2 4 6 8 0. 12 14 6. " pory OBLIQUE THICKNESS (mfp} " =~ = as a Function of Oblique Thickness (Mean Free Paths) of a Finite Lead-Water Slab Shield: 6-Mev Photons Incident at a 60-deg Angle. PERIOD ENDING SEPTEMBER 10, 1956 UNCL ASSIFIED 2-01-058-#17 O = DOSE RATE fp = INITIAL DOSE RATE $, = INCIDENT NUMBER FLUX D°/¢f =0.004368 (mr/hr)/(y!cmz- sec) Apy=2.05¢cm ® MONTE CARLO, FINITE SLAB & MOMENTS METHOD, SEMI-INFiNfTE SLAB (REF: NYQ-3075) 2 3 ) ] 5 7 8 Po7 OBLIQUE THICKNESS (mfp) Fig. 5.1.17. Gamma-Ray Dose Rate as a Func- tion of Normal Thickness (Mecan Free Paths) of a Finite Lead Slab Shield: 3-Mev Normally Incident Photons. UNCLASSIFIED 2 . 2-04-059-118 ® MONTE FINITE SLAB © MOMENTS METHOD, SEMI- SLAB (Ref.: NYO-3075) 8-, DOSE BUILDUP FACTOR 2 6 8 10 © 19 6 Hols THICKNESS {mfp) DoseBuildupFuctor Flg. 5.1.18. Gummu-'Rcly'Dose- Buildup Factor as a Function of Oblique Thickness (Mean Free Paths) of a Finite Lead Slab Shield: 3-Mev Normally Incident Photons. ' 277 ffiffifigfi’ ANP PROJECT PROGRESS REPORT 5.2. LID TANK SHIELDING FACILITY R. W. Peelle STUDY OF ADVANCED SHIELDING MATERIALS V. R. Burrus! J. M. Miller J. R, Smolen? D. R. Otis? P. B. Hemmig! The studies of advanced shielding materials45 have now included tests on mockups consisting of 10n assignment from U, S. Air Force. On assignment from Pratt & Whitney Aircraft, On assignment from Convair, San Diego. “These experiments have been designed to be of particular interest to GE-ANPD, 5R. W, Peblle et al, ANP Quar. Prog. Rep. June 10, 1956, ORNL.-210¢4, p 269. a beryllium moderator region, a lead or depleted uranium gamma-ray shield, and a lithium hydride and oil neutron shield. In some tests a boral sheet was inserted outside the beryllium layer to prevent thermal neutrons from entering the gamma-ray shield. The specific configurations tested are summarized in Table 5,2.1. The compositions and dimensions of all the materials except the beryl- lium aond boral were previously reported. The metallic beryllium slab is 48 x 49 x 4 in, and the boral sheet is 48 x 48 x |n. The latter has an average density of 2.6 g/?: ‘and contains about 20 wt % boron, or 0.66 g of boron per square centi- meter. TABLE 5.2.1. SUMMARY OF THE CONFIGURATIONS USED FOR L TSF MOCKUP TESTS OF ADVANCED SHIELDING MATERIALS Configuration No. Composition 6%-3 4 in. of Be in oil 4 in. of Be + 12 in. of LiH in oil 4 in. of Be + 24 in. of LiH in oil 69f'| 2 4 in. of Be + 3 in, of Pb in oil 4 in. of Be + 3 in. of Pb + 1 ft of LiH in oil 4 in. of Be + 3 in. of Pb + 2 ft of LiH in oil 4 in. of Be +3 in. of Pb + 3 ft of LiH in oil 69-13 4 in. of Be + % in. of boral +3 in. of Pb in oil 4 in. of Be + % in. of boral +3 in. of Pb+ 1 f1 of LiH in oil 4 in. of Be + }’2 in. of boral +3 in. of Pb + 2 ft of LiH in oil 4 in. of Be + % in. of boral +3 in. of Pb+ 3 fi'of LiH in _oi_l 69-14 4 in. of Be + 3 in, of U”s* in oil 4 in. of Be +3 in. of U238 4+ 1 f1 of LiH in oil 4 in. of Be +3 in. of U238 + 2 ft of LiH in oil 4 in. of Be +3 in. of U238 4+ 3 ft of LiH in oil 69-15 4 in. of Be + z In. of boral + 3 in. of U238 in oil 4 in. of Be + g in. of boral +3 in. of Uz’38 4+ 1 f1 of LiH in oil 4 in. of Be + z in. of boral + 3 in. of U238 + 2 £t of LiH in oil 4 in. of Be + ,'2 in. of boral + 3 in. of U238 -|-3, #t of LiH in oil *Uranium depleted to 0,24 wt % in U235. 278 The materials were placed within the oil medium as close as possible to the Lid Tank converter plate, The thermal-neutron fluxes quoted from the data are equal to the neutron density times 2200 m/sec. The gamma-ray tissue dose rate measurements were obtained with an anthracene scintillation detector calibrated against a stand- ardized radium source. The gamma-ray dose rates beyond the various configurations “are shown in Figs. 5.2.1 through 5.25. It is not yet clear to what extent these results were influenced by the thin loyers of oil between the slabs in the configuraflons. Neutrons slowed down in these oil layers may give rise to secondary production- in the adjoining slabs of materials, as well as in the oil itself. The average - total thickness of these layers was 3 ¢m for each conflguratlon, and tests to study thelr effects are being planned. ‘ : : Since there was' a noticeable decrease in the gamma-ray dose rate whenever lithium hydride was added to the configuration immediately behind the gamma-ray shield material, an investigation was performed to determine how much of this reduction could be attributed to the elimination of oil capture - gamma rays. Measurements were taken beyond a configuration in which there was a I-ft layer of oil between a lead gamma-ray shield (followed by boral) and the lithium hydride. The gamma-ray dose rate was considerably increased (Fig. 5.2.6), PERIOD ENDING SEPTEMBER 10, 1956 ond a curve of the difference between the two curves should largely represent the attenuation curve of cil capture gamma rays. There was a strong reduction in the gamma-ray dose rate when the depleted uranium slab was moved to a region of relatively low neutron flux (Fig. 5.2.5). The preliminary interpretation of this reduction is that fissions caused by intermediate and fast neutrons were largely eliminated by the additional neutron shielding. The thermal-neutron trcverses for the various configurations are shown in Figs. 5.2.7 through - 5.2.11, Only the traverse in oil was checked against gold-foil measurements, Figure 5.2.9 shows that the thermal-neutron flux is independent of the order of the lithium hydride and lead, except for the difference at the beginning of each traverse caused by variations in the amount of oil trapped between the various slabs. In Fig, 5.2.11, how- ever, for a similar arrangement containing depleted uranium, moving the uranium out of the intense neutron field reduced the thermal-neutron flux. This seems to be the result of the intermediate or " fast fissions discussed above in connection with the gamma-ray measurements. Studies are being continued on configurations similor to those reported here. In addition, fast- neutron dose rate traverses will be made available in a future report. 279 ANP PROJECT PROGRESS REPORT SR 404 OIL ONLY 10* 5 4 in. OF Be + OIL 4in. OF Be + 4 ft OF LiH + OIL 102 4 in, OF'_Be + 2 ft OF LiH + OIL GAMMA-RAY TISSUE DOSE RATE (ergs/g-hrew)' 10 5 2 BASED ON SOURCE PLATE POWER OF 5.24 w 4 5 2 107! 20 40 60 80 100 120 140 Zg, DISTANCE FROM SOURCE PLATE (cm) Fig. 5.2.1. Gamma-Ray Tissue Dose Rate Traverses Beyond Configuration 69-3. 280 2-04-057-69-275 iy 460 GAMMA-RAY TISSUE DOSE RATE (ergs/g-hr-w) PERIOD ENDING SEPTEMBER 10, 1956 “ in.OF BORAL + 3in.OF U + 1t OF LiH ¢+ OIL Cro - NEUTRON FLUX ( neutrons/u:,m2 -gsec-w) w» o o > THERMAL o - 4in.OF Be + )2 in.OF BORAL + 3in OF U + 2 ft OF LiH +OIL 1 4in.OF Be +1 ft OF Lit + Y5 in.OF L+ '3in, OF U + 11 OF LiH + OIL 5 4in.OF Be + Y5 in.OF BORAL + 3in. OF U + 31t QF LiH + OIL 250 DISTANCE FROM SOURCE PLATE {cm) Fig. 5.2.11, Thermal-Reutron Flux Traverses Beyond Configuration 69-15. 290 o 10 20 . 30 40 50 60 70 80 90 100 1o 120 130 140 150 160 1707 4 PERIOD ENDING SEPTEMBER 10, 1956 fiéé’ 5.3. BULK SHIELDING FACILITY F. C. Maienschein THE FISSION-PRODUCT GAMMA-RAY ENERGY SPECTRUM W. Zobel T. A. Love The importance of obtaining information about the gamma-ray energy spectrum and time decay characteristics of the fission products of U235 as related to circulating-fuel reactors has been re- peatedly emphasized.'=3 The preliminary results on the time decay characteristics (Phase | of the experiment) were presented in a previous report,3 and equally preliminary results on the energy spectrum (Phase |l) of these fission-product gamma rays at different lengths of time after flssuon are presented here. The equipment used in the expernment has been described elsewhere.4 Samples of enriched vranium weighing ebout 2, 7, 15, and 32 mg, re- spectively, were irradiated in the ORNL Graphite Reactor for periods varying between 1 and 64 sec, depending on the run. The weight cnd bombarding time were chosen so that the counting rate during a run did not exceed a manageable value. The energy calibration of the spectrometer was carried out in the usual manner, with Na22, Y88 qand ThC” sources. Sources of Hg203, Cs137, Na22,. Zn65, Cob9, and Na24, whose absolute source strengths were determined with the aid of the high-pressure ion chamber of the ORNL Radio- isotopes Control Laboratory, were used for the calibration of the efficiency of the spectrometer as a function of photon energy. The efficiency thus obtained “was the total effncuency “of the spectrometer, ~rather _than the peak effu.nency,_ which is sometimes used - To. correct the data for counting losses, the average count rate over - The countlng" ' 532 between 0.28 and 5.0 Mev and between 1.25 -~ and 1600 sec gave a total of 2,81 photons emitted * per fission with a. total energy of 3.22 Mev per fission. These values carry. an estimated error of about 125%, -They should be compared with the values obtained in Phase | of the experi- ments,3 which, when corrected for differences 3w, Zobel T. A. Love, and R.- W. Peelle, ANP Quar. the counting period was. used,: _ ‘R. W. Peelle, T. A Love. ond F. C M:cil.enScHelln, : "ANP - Quar. Prog Rep ]une 10 (1955, - 0RNL-1896,_ : p 203. 2R W, Peelie, W. Zobel und T A. Love, ANP Quar.' Prog. Rep. Dec. 10, 1955, 0RNL-20|2 p 223, Prog. Rep Marcb 10, 1956, ORNL-2061. P 250. . 41. A Love, R W. Peelle, and F. C. Molenschem, Electronic Instrumentation for a Multzple-Crystal Gam- ma-Ray Spectrometer, ORNL.-1929 (Oct. 3, 1955). losses for a given count rate were determined experimentally, The determination of the number of fissions taking place in a sample depends on the length of time the sample is exposed to the thermal neutrons and the intensity of the flux. In an experiment of this type, where the sample is moved into and out of the flux, the bombarding times and fluxes used in the calculations must be the ‘‘effective’ bombarding times and neutron flux. The effective thermal-neutron flux was obtained by exposing bare and cadmium-covered gold foils for a nominal 32 sec at the time of each run and computing the flux from the cadmium difference. The effective bombarding time was computed from the activities of gold foils, which were also exposed for the nominal bombarding times used in the experiment. It was found that the flux differed from the previously used value by about 27%, Also, the bombarding times differed some- what from the earlier ones. The energy spectra obtained at 1.7, 6.2, ]07 40, 70, 100, 250, 700, 1000, and 1550 sec after fission are shown in Figs. 5.3.1 and 5.3.2. They have been separated to avoid any confusion which ‘might arise owing to the scatter of the points ‘belonging to adjacent curves. It should be pointed out that the peaks merely represent an attempt to ‘'show some of the structure in the curves. The error on the points has not yet been computed, so that these peaks are not necessarily final. They do appear on successives curves, however, which lends some credence to their existence. No attempt has yet been made to assign these - peaks to specific |sotopes An integration of the spectra in Figs. 5.3.1 and discovered in the evaluation, were 2,92 (1£25%) “photons per fission and 3,23 (125%) Mev per - fission. The agreement between the two phases is good; this is also shown in Figs. 5.3.3 and 291 ----- ANP PROJECT PROGRESS REPORT Fig. 5.3.1. Fission-Product Photon Energy Spectrum at 1.7, 10.7, 70, 250, and 1000 sec After Fission. 292 PHOTON INTENSITY {photons/Mev-fission -sec) 5, N o N 5, o 16° 167 * COMPTON SPECTROMETER ** PAIR SPECTROMETER 10.7 sec 250 sec 1.0 : 20 30 PHOTON ENERGY (Mev} o2 kv | S g ek 40 UNCLASSIFIED 2-0{-058-0-57 50 -4) W ' 100 sec* PHOTON INTENSITY ( photons /Mev - fission -sec) - * COMPTON SPECTROMETER 2o PAIR o 100 200 . 30 - ' . " PHOTON ENERGY (Mev) PERIOD ENDING SEPTEMBER 10, 1956 40 UNCLASSIFIED 2—-01-058—0~-58 5.0 Fig.5.3.2. Fission-Product Photon Energy Spectrum ot 6.2, 40, 100, 700, and 1550 sec After Fission. 293 ANP PROJECT PROGRESS REPORT UNGL ASSIFIED - 2-01—058—0—59 10.7 sec PHOTON INTENSITY ( photons /Mev-fission:sec) 0 1 2 3 4 5 6 PHOTON ENERGY (Mev) Fig. 5.3.3. Fission-Product Photon Energy Spectrum from Phase !l of Experiment with Superimposed Spectrum from Phase |. ' | 294 6i4 (21 ) PERIOD ENDING SEPTEMBER 10, 1956 5.3.4, where cross plots of the data from one here. The values obtained in the LTSF experiment phase of this experiment are shown with the were 4.2 (£20%) photons per fission and 4.8 appropriate points from the other phase. Another (£20%) Mev per fission. comparison may be made with the spectra obtained in the experiment carried out at the LTSF.2 It should be kept in mind, however, that there are Further refinement of the evaluation of the data systematic differences in the evaluation of the is contemplated. It is expected that such re- data in that experiment and the one described evaluation will show better agreement between UNCLASSIFIED 2—01—058—0-60 1072 DECAY RATE { photons/fission-sec) o PHASE | OF E! ° PHASE Il OF 10—3 1074 . 1 2 5 10 2 5 102 2 5 10 2 5 10 r, TIME AFTER FISSION (sec) Fig.‘ 5.3.4. Decay Rote of Fission-Product Gamma Rays from Phase | of Experiment for a 0.28- to 5.0-Mev Energy Range with Superimposed Points from Phase 1. 295 ANP PROJECT PROGRESS REPORT the different experiments and lead to a more defi- nite answer. GAMMA-RAY STREAMING THROUGH THE NaK PIPES THAT PENETRATE THE ART SHIELD T. V. Blosser D. K. Trubey A mockup experiment®+é to investigate the in- crease in the dose rate outside the ART lead shield caused by gamma-ray streaming through the NaK-filled pipes that penetrate the shield has now included measurements beyond a duct placed through a mockup of the ART shield at an angle of sT. V. Blasser, ANP Quar, Prog. Rep. March 10, 1956, ORNL-2061, p 249. , : b1, Blosser and D. K. Trubey, ANP Quar. Prog. Rep. June 10, 1956, ORNL-2106, p 274. 4.25-in, THICK LEAD ALUMINUM-FILLED DucCT TWO INCONEL PIPES ' AIR 51 deg 30 min (Fig. 5.3.5). The duct consists of two concentric Inconel pipes (separated by a thin air ‘layer) filled with aluminum turnings and fitted with a tungsten collar at the top. The axis of the duct in the horizontal plane was at a 45-deg angle from the x axis. For this particular arrange- ment a specific coordinate system using x’ and y” axes was adopted to designate the points of measurement. The x’, y’ axes were at an angle of 45 deg to the x, y axes as shown in Fig. 5.3.6. . Measurements were made beyond the ducts along several x’ traverses (labeled A through E in Fig. 5.3.6) for two z distances (10.5 and 20,5 cm) from the lead shield. The ratio of the dose rate beyond the water-filled hole (that is, with the duct not in position) to the UNCLASSIFIED ORNL-LR-DWG 16200 M LINER TUNGSTEN COLLAR 51deg 30min Fig. 5.3.5. Mockup of an ART Duct Penetrating the Lead Shield ot an Angle of 51 deg 30 min. 296 GLs #} ‘dose rate beyond the solid lead shield is plotied in Fig. 5.3.7, as well as the actual dose rate beyond the solid lead. (For convenience -in plotting, the curve actually represents twice the measured dose rate.) The ratios of the dose rafes beyond the aluminum-filled duct to the dose rate .40 20 PERIOD ENDING SEPTEMBER 10, 1956 beyond the solid lead are shawn in Figs. 5.3.8 and 5.3.9. Again the actual dose rates beyond the solid lead are plotted on the lower portions of the figures. The reduction in dése due to tilting the duct is about a factor of 4. This indicates that the UNCLASSIFIED ORNL-LR—-DWG 16868 DISTANCE ALONG y AXIS (cm) - 10 -20. -0 L— =40 © ~10 O POINTS OF MEASUREMENTS 40 10 0 20 30 " 'DISTANCE ALONG ¥ AXIS (cm) = Fig. 536Coordmcrte Sysiemfor ART 51-deg 30-min _Dfic_f Mockup Tests. ; 297 Fa 4 g .ANP PROJECT PROGRESS REPORT - ENET . ORNL—LR~DWG 16204 S oC,y'=0 oD, y/=Tcm S E, y'= 4icm AA y= -141em AB, y'==Tcm TO DOSE RATE BEYOND SOLID LEAD SHIELD W . RATIO OF DOSE RATE BEYOND WATER-FILLED HOLE . N -/hr) o - DOSE RATE (x2) BEYOND SOLID LEAD SHIELD {r, X’ {cm) Fig. 5.3.7. | :qumu;Ruy Dose Rates Beyénd a Water-Filled Hole Penetrating ¢ Mockup of the ART Shield at an Angle of 51 deg 30 min. HSSRTT ORNL—LR—-DWG 16202 10 P o C,y/=0 e ® D, y’~Tcm Fws 5 8 E,y/=14.4¢m n:§t_-'1 A A y'=—{41cm gee ABy=-T7 B 30 3g 2 o £2% cPm w 1 7x10': DOSE RATE (x2) BEYOND SOLID LEAD SHIELD (r/hr) x'(em) Fig, 5.3.8. Gamma-Ray Dose Rates Beyond Aluminum-Filled Ducts Penetrating a Mockup of the ART Lead Shield ot an Angle of 51 deg 30 min (z = 10,5 cm). 208 644 -20 -0 0 10 20 30 40 SO mockup gamma-ray source has an approximate distribution proportional to cos?® 6. This is in agreement (within' the uncertainty) wnh the esti- mate of the distribution reported previously.b It must be remembered, however, that the source distribution of the ART will probably not corre- spond to that of the mockup source, even though the angle of penetration is the same as that of the mockup reported here. That is, the fuel in the heat exchanger is directly in line with the " axis of the duct in the ART, but the fBurce of “the mockup is believed to be a cos2. @ distribution about ‘the normal to-the lead. For this reason it- “would probably be more realisfic to apply the previously reporfed2 results - for the -straight- through penetrations, which mocked . up ‘portions of the north-head ducts. Some measurements have been made beyond - . straight-through penetrations mocking up portions of the south-head ducts, and the results will be published later. The effects of various com- ponients in the mockups and the effects of patches are now being investigated, b ' ORNL—LR—DWG 16203 Wwg 10 0 ] oo ! =3 & @3 oY x5 b &Y w o g3 —20 -$Q ] 10 20 30 40 50 x* {cm) - Fige 5.3.9. Gamma-Ray Dose Rates Beyond Aluminum-Filled Ducts Penetrating a Mockup of the ART Lead Shield at an Angle of 51 deg 30 min (z = 20.5 cm), ' -} -] e o e, . PERIOD ENDING SEPTEMBER 10, 1956 . - 5.4, SHIELD MOCKUP CORE ‘ C. E. Clifford The design of the Shield Mockup Core (SMC),’ a 5-Mw fixed-fuel reactor proposed for shielding studies of the circulating-fuel reflector-moderated reactor (CFRMR), has been gltered somewhat as dictated by nuclear and shielding calculations per- formed by the Engineering Physics Group at Pratt & Whitney Aircraft.2: Other changes also became necessary for engineering or ecbnomicul reasons. The SMC has l:een des;gned so that the leakage fluxes, with the exception of the fission-product radiation from the heat exchanger, will be the ..same as those from s CFRMR (PWAR-1) now contemplated for aircraft propulsion.. The core will consist of 204 modifie_d'M_TR-type fuel plates lc. E. Clllford and L. B. Holland ANP Quar. Prog. Rep. June 10, 1956, 0RNL-2ID6, P 279 Calculohons performed by F. C. Mornman and J. B, Dee, Pratt & Whitney Alrcraft. L.. B. Holland placed radially around the island reflector. The shape of the plates will be such that when they are all in position the geometry of the CFRMR will be simulated. Numerous calculations have been performed to compare the SMC with the CFRMR, and each deviation from the CFRMR design was considered. A summary of the calcu- lations which [ed to the current SMC des;gn (Flg. 5.4.1) is given below. SE LECTION ‘OF CALCULATIONAL MODELS The nonregular geometry of the CFRMR core, which is somewhere between a sphere and o cylinder, necessitated the use of several cy- lindrical and - spherical calculational models to determine the upper and lower limits of the critical mass. The validity of the nuclear models was checked by using them in calculations of the critical mass of the ART high-temperature critical experiment. As shown in Table 5.4.1, the results '» - TABLE 5.4.1. COMPARISON OF CALCULATED AND EXPERIMENTAL CRITICAL MASS AND VOLUME OF THE ART HIGH-TEMPERATURE CRITICAL EXPERIMENT * Critical Mass Core Voluhe ond acwal outer radivs of reflector, core rodlus de- ‘ ' e iermmecl by welghflng ‘that volume of the core which :l;’ffidev:otes frorn ‘a cyllncler** by a cosune funcflon and oddmg to the cylmdrlcol core. volume (cnhcol mass - _ 'defermmed with uctual core volume) - {kg) (liters) E_xperinlentcl 18.2 _Calculated Spherical models Equatorial crosscut’ , S : 176 64 ‘Equal core volume, ocfuol micl-plone tl'uckrless of lslcmd, . 18.6* _ 95.2 and actual mld-plane outer dmmeter of feflecfor - Equal volumos for all reglons - o 21.8 o 96.3 Cylindncal models ) : Equutonal crosscuf S . LT 29.6 Equal core volume, acfuul msd-plane tlnckness of island 22.8 . and actual mld-plane outer diameter of reflector - Equal importance:’ octual mtd-plone tl'm:kness ol |slond - . 18.7* - *Models used in SMC calculutions. **That is, region 5 in Fig. 5.4.7. 299 ANP PROJECT PROGRESS REPORT SUPPCRT ASSEMBLY REMOVABLE IONIZATION CHAMBER REMOVABLE PLUG — . SECTION B-B 101234567 el v SCALE IN INCHES LEAD LIFTING HOOK REMOVABLE PLUG LEAD CONTROL ROD DRIVE - 2-01-059-79 — FUEL PLATE SPACERS GRAPHITE ISLAND S ::\‘\ “ N :\ N R COOLING NN NN LINE NN ' NN SRR -.\\\‘\\ SN CONTROL, RN e L ‘\\‘\\\\\‘_ S CHAMBER AR FyEL PLATES . N \\\\\\.\\\\ 0 \\‘ A P2 SECTION A-A Y IR N 101234567 g COOLING LINE I § SCALE N INCHES “fi‘ § GRAPHITE ISLAND N FUEL BERYLLIUM BORON LAYER INCONEL SALT REGION INCONEL BORON LAYER : INCONEL (CFR PRESSURE SHELL) INCONEL NEUTRON SHIELD CONTAINER (ALUMINUM) ' 2 3 4 SCALE IN FEET Fig. 5.4.1. The Shield Mockup Core (SMC). 300 ) of both a spherical and a eylindrical model were in agreement with the experimental critical mass. Both a continuous slowing-down model and a Goertzel-Selengut slowing-down model were used to account for the energy degradation of neutrons in the light water used as the coolant in the core. The agreement was good enough to permit the use of the continuous slowing-down model for para- metric studies to determine a design point. CALCULATIONAL PROCEDURE The calculational procedure first inciuded a determination of the fuel loading for a critical condition of a given core configuration by a multigroup machine calculation, Parametric studies were then carried out to match the core power distribution and neutron-absorption distribution in the beryllium of the SMC with those in the CFRMR. Once these distributions were in agreement the contributions of the various radiation sources in both the SMC and CFRMR to the total dose in the crew compartment were calculated,” The SMC configuration which gave the proper dose contri- butions was then selected. In the method used to determine whether the doses from the gamma-ray sources in the SMC and the CFRMR were matched, both reactors were divided into 35 spherical shells, and the dose rate from each material of each shell at a point 50 ft from the reactor was calculated. The total contri- bution from each material in a particular region was then obtained by adding the dose rate from . that material in each shell in that region, The calculations were based on the neutron absorptions from the critical calculations and published data® on capture gamma rays. .In the first calculations an average capture gamma-ray energy was assumed, but: in later colculatlons a . capture gamma-ray spectrum was approxlmated by using the ap- propriate number of 2-, ;6= 7. and 9-Mev photons indicated for a parhcu!or material. Since the operating temperature of the CFRMR will be ~ approximately 1300° F higher than that of the SMC, . and ' therefore the thermal-neutron "cross section - will be much smaller, it was necessary fo match’ ‘the neutron absorptions, rather than the thermal- neutron flux, in the beryllium and other materials 3P. S. Mittleman and R. A. Liedtke, Nucleénics 13(5), 50 (1955). tolerated. PERIOD ENDING SEPTEMBER 10, 1956 of the two reactors to obtain the proper capture gamma-ray distribution, The fission neutrons which contribute to the neutron dose in the crew compartment and thus are of concern to the shield designer are those born at high energies. (Fast neutrons thermalized in the core or reflector will not pass through the boral ¢urtain outside the reflector. They con- tribute to the dose in the crew compartment through the capture gamma rays which they produce in the core-reflector region.) It is assumed that if the removal cross sections of the materials in the SMC core are very nearly the same for fast neutrons as the removal cross section of the fused salt fuel in the CFRMR, that if the normalized core power distributions are matched, and that if the regions outside the core are correctly mocked up in the SMC, the number and energy spectrum of neutrons leaving the SMC shield should be the same as those leaving the CFRMR shield. CALCULATIONS FOR SMC CONFIGURATIONS USING UO,-STAINLESS STEEL FUEL ELEMENTS It was assumed originally that the CFRMR leakage fluxes could best be mocked up in the SMC with UO_—_stainless steel fuel elements cooled with D20. The results of early calcu- tations confirmed that a better match of the core power distribution and the neutron absorptions in the beryllium could be obtained if D20 rather than H O were used (Figs. 5.4.2 and 5.4.3).4 This resulf was to be expected, since the moderation of neutrons by D20 is closer to that of the fused ~ salts than is the moderation by H20. However, . the use of a closed D_O system in an H,O storage pool at the TSF or the BSF posed many handling problems and a safety problem that could not be It became apparent from a plot of the critical mass of the SMC for all mixtures of D,0 ~and H 0 (Flg 5.4.4) that a supercrmcal condition --could arise almost instantaneously if H20 were inadvertently pumped into the D,O-cooled core ‘as a result of component failure or mishandling. It was therefore decided that H20 should be used as the coolant in the SMC. 4All curves are normalized to one fission per cubic centimeter in the core region. 301 ANP PROJECT PROGRESS REFORT orcren 7 2-0t-059-98 - 3.0x407 FUEL ELEMENTS: U0,—STAINLESS STEEL WITH UNIFORM FUEL DISTRIBUTION. N ’0-7 /’\ _ / N © O k SMC, Be + Al, 100% D,0 IN VOIDS 3 . W & 20x107 _ & /\ \ PBW CFRMR, Be + No | % = o : 5 ' N s SMC, Be ONLY, {00 % D,0 IN VOIDS & 1.5%407 & / i S ‘ £ | / PBW CFRMR, Be ONLY o e ] \ QO . L72] 3 #/ >< 2 1.0x410-7 3\ o / - K\ m . - . = W z / 0.5x10-7 \\ SMC, Be + Al, 100 % H,0 IN VOIDS \ ! k ~ SMC, Be ONLY, 100 % H,0 IN VOIDS _ ' x10-7 - 35 40 45 50 55 60 65 70 SPACE POINTS IN REFLECTOR (SPHERICAL MODEL) Fig. 5.4.2. Neutron Absorptions in Reflector Region of an SMC Using UO,.Stainless Steel Fuel Elements and Hzo or DZO Coolant. Because of the excessive moderation of neutrons in the core by H20 and the consequent reduction of neutron absorptions in the reflector of the SMC, it was necessary to devise some means of reducing the amount of H,O in the core. A parametric study ~ was therefore carried out to determine the effect of substituting aluminum (by means of wedges between the fuel plates) for some of the H,LO. As shown in Fig. 5.4.5, the combination which re- sulted in a neutron absorption distribution in the SMC reflector that most nearly matched that in the CFRMR reflector was 25% H,O and 75% 302 aluminum in the spaces between the fuel plates. [The calculations were for absorptions in the beryllium, aluminum, and H20 (ref 5) in the SMC reflector and for beryllium and sodium in the CFRMR reflector.] The agreement for the region near the core-reflector interface is poor, but this difference can be minimized by adjusting the amount of aluminum ‘in the SMC reflector. 5At the time these calculations were performed it was felt that it would be necessary to cool the beryllium with H,0. ' : A} v} PERIOD ENDING SEPTEMBER 10, 1956 ST 2-0t-059-91 2.4 P&W GFRMR 2.0 / / / o /[ ) / / /7 / H,0, CONTINUOUS SLOWING DOWN / 1 | / \\/ N POWER DENSITY (NORMALIZED TO 1 fission/cm>) . Ho,0, G-5 SLOWING DOWN o » 15 20 25 30 35 40 45 50 55 SPACE POINTS IN CORE (SPHERICAL MODEL) Fig. 5. 4.3. Power Distribution in an SMC Usnng UO,-Stainless Steel Fuel Elements and H 0.D0,0 Coolant. In order to further investigate the effect of substituting aluminum for H,O in" the core, the power distribution in the core was determined for - each H o O-aluminum_ combination (see Fig. 5.4.6). Once ugam ‘the power distribution for the 25% H,0-75% aluminum’ combination” most nearly ap- be attributed to-the fcct that the rodml arrangement of the SMC “fuel plates will cause areduction -in the uranium density with increasing radlus, where- as the homogeneous fused salts in the CFRMR ‘will give a uniform uranium density. |f the uranium loading in the SMC fuel plates is varied so that only the outer regions of some of the plates _contain fuel, as in Fig. 5.4,7, for example, the fuel density in the core will become more nearly - constant. For calculational purposes the core was - proached that of the. CFRMR, . but the agreement = was net good ‘enough. It was felt that this could - divided into five .concentric regions and the fuel ':"‘Ioadmg in_ the outer three .regions was varied. - As shown in Fig. 5.4.8, the power density which - most nearly approached that of the CFRMR was for @ 1:1:1:2:3 fuel distribution. A comparison of neutron absorptions in the reflector of the 303 LANP PROJECT PROGRESS REPORT SECRET 2-01~059-92 Ha0 (%) {00 90 80 70 60 50 40 30 20 10 o 9 g 8 = S > o < E7 o / o g T 4 ' / = 6 2 / T O . 5 .4 0O 10 .20. 30 40 50 60 70 80 90 400 0,0 (%) Fig. 5.4.4. Critical Mass of an SMC Using uo,- Stainless Steel Fuel Elemenfs uqd D20-H 20 Coolant. SEoREP _7 2-01-059-TTA "1 SMC, 12.5% H0, B7.5% Al -P & W CFRMR - SMC, 25% H,0, 75% Al 2x10°7 TN N TV 7 VI i ~ N —_— 07 SMC, 50% H,0, 50% Al | SMC, 100% H,0 o L 3 40 45. 50 85 60 65 70 SPACE POINTS IN REFLECTOR [SPHERICAL MODEL) NEUTRON ABSORPTION (FRACTION/cmOF REFLECTOR) ‘ Fig. 5.4.5. Fraction of Neutron Absorption in . Entire Reflector Regien (Be + Al + H,0) for Various Core Configurations of en SMC Using UO,-Stainless Steel Fuel Elements and H,0 Coolant. ' ' orCRTT 2-04-059-76A i @© n > SMC, H,0 SMC, 50% H,0, 50% Al SMC, 25% Ho0, 75% Al ro o SMC, 12.5% H,0, 87.5% Al ‘ POWER DENSITY (NORMALIZED TO { FISSION/¢m’) N -~ n o @ 8 W CFRMR o > o 12 16 20 24 28 32 SPACE POINTS IN CORE (SPHERICAL MODEL) Fig. 5.4.6. Power Distribution in Various Core Configurations of an SMC Using UOz-»Stuinless Steel Fuel Elements and H,0 Coolant. SMC using this configuration with those in the reflector of the CFRMR also showed gocd agree- ment. On the basis of the above results, the first shielding calculations, that is, the first calcu- lations of contributions to the gamma-ray dose by the various regions of the reactor, were per- formed for an SMC using UQ_~stainless steel fuel elements with a 1:1:1:2:3 fuel distribution and H,0 as the coolant filling 25% of the voids between the fuel plates. It immediately became apparent that for the lower operating temperature of the SMC the capture gamma-ray production in the Inconel core shells was too high. Since the shell at the core-reflector interface was the im- portant contributor, succeeding calculations were made for an SMC with a half-thickness Inconel shell at that interface. The results of these calculations are shown in Table 5.4.2 for two different shield configurations, The effect that the thinner shell had on the power distribution was then calculated. As can be seen in Fig. 5.4.8, the results were in closer agreement with the power distribution for the CFRMR than any of the other SMC configurations. ol “w )] - r NOTE: REGIONS 4 THROUGH 5 INDICATE THEORETICAL DIVISIONS OF THE CORE VOLUME FOR NUCLEAR CALCULATIONS S PERIOD ENDING SEPTEMBER 10, 1956 S EGRE 2-04-059-100 II [~ L/ u-a1 NOT TO SCALE Fig. 5.4.7. Proéosed Uranium Loading of SMC Fuel Plates. When ' the - neutron absorptions in the reflector regions also corresponded (Fig. 5.4.9), it was felt that the best configuration for an SMC using UO - stainless steel fuel elements had been attained. At the same time, it was realized (Table 5.4.2) that the capture gamma-ray dose rate from the stainless - steel in the fuel elements was ex- cessive, even though the first calculations were made for a single capture- gamma-ray energy for each element rather than for an energy spectrum. It was then decided to convert to uranium-aluminum fuel elements and to repeat all the calculations ) usihg the spectrum of capture gammaeray energies previously indicated for the dose rate calculations. 'CALCULATIONS FOR SMC CONF IGURATIONS USING URANIUM-ALUMINUM FUEL ELEMENTS The use of uranium and aluminum in the core will necessitate a thicker fuel plate design; however, the thickness of the aluminum wedges between the plates will be adjusted so that the 305 ANP PROJECT PROGRESS REPORT amount of H,O in the core region will remain the same and it will continue to be referred to as a 25% H,0-75% aluminum case. The power distribution in an SMC configuration using vranium-aluminum fuel elements is shown in Fig. 5.4.10 for three different fuel distributions. The 1:1:1:2:3 distribution again gave the best agreement with the CFRMR. This distribution was then used in the calculations of the neutron ab- sorptions in the reflector region (Fig. 5.4.11). A comparison of Fig. 5.4.11 with Fig. 5.4.9 shows that the substitution of aluminum for stainless steel in the core will not appreciably change the number of neutron absorptions in the reflector. In view of the agreement in the nuclear calcu- lations, shielding calculations were then performed 'TABLE 5.4,2. COMPARISON OF CALCULATED GAMMA-RAY DOSE RATES? FROM CORE-REFLECTOR - REGIONS OF THE CFRMR AND THE smc? usING UO,—STAINLESS STEEL FUEL ELEMENTS | Dose Rate at 50 ft (r/hr) 32-cm Alkylbenzene 5.in,-Pb=43.cm-Alkylbenzene ! * ' - Regwn Material Reactor Shield (No Pb) Reactor Shield . SMC CFRMR SMC ‘ CFRMR Isla n_d-co;'e shell Inconel® 332 96.17 0.2330 0.0544 H,0¢ 11.37 0.00079 | Sodium 0.1377 0.0002 ‘ Subtotal 343.0 96.0 0,2338 0.0546 Core Stainless steel 1.785 1.8969 Hzo 11.53 0.0169 Aluminum 417.9 0.442 Fuel 3,385 3979 6.456 7.588 Subtotal 5,599 3979 8.8118 7.588 Corereflector shell Inconel 1,570 1454 0.8100 0.7714 H20 0.54 0.0166 . Sodium 2.017 0.0029 Subtotal 1,571 1456 0.8266 0.7743 Reflector Berylliuvm 4,238 3868 4,272 3.883 Aluminum 1,708 1.735 Sodium 262.8 0.382 Subtotal 5,946 4131 6.007 4.265 Total 13,459 9662 15.879 12.682 'aAn average capture Qummo-roy energy was assumed. bSMC configuration includes UO —stainless steel fuel plates with 1:1:1:2:3 fuel distribution, 25% H,0 and 75% alu- minum in core voids, a half-thlckness Inconel shell at the core-reflector interface, and a ful|-fh|cl\. N - .‘f//}. / q - ., “ > [ - s "N \. N/ . § b, .'.‘v\ :.’ / /..-. 2 * 2 \\s...\ T . AT -'.: 1 6 ( '---4—..4" ” = . . " e, e — e lt o ' 2 A S - . ....d a ‘- . -.:%" e o e V ..“"\. ":::.. ) . l \ . v e, "*0d b, - ~ = P& W CFRMR o V . poss ..."'--‘. 0.8 - N s S SPACE BETWEEN ELEMENTS | - ' |- " CONTAINS 25 %5 H,0, 75% Al ' o4 L — L || | A . - 17 19 21 23 25 27 29 39 33 35 37 ° T o o . SPACE POINTS IN CORE (CYLINDRICAL MODEL) ' F|g. 5.4. 8 Power Distribution for Vurious Fuel Distributions in an SMC Using UO -Stmnless Steel Fuel Elements and H O Coolant. THReRT. 2-01-059—99 ’ e e ':é'Mc:B + Al | = R - | ——SMC, Be - o7 g 2_5 e —— < L .7 7 . - w / R P & W CFRMR, Be +Na | SMC: STAINLESS STEEL-~UQz FUEL PLATES g N . - WITH 14:1:2:3 FUEL DISTRIBUTION, = 2.0 - - — -~ 25% HpO AND 75 % Al IN CORE vOIDS, — w o f o \ N HALF ~THICKNESS INCONEL SHELL AT m SMC, Be ONLY~, | . : CORE-REFLECTOR INTERFACE .- Z 15— . _— - 5 | \ \\ 7 S 0 E 0 e —— —— - \ L 2. : - \

Ef:wugERT M T.S. SHEVLIN, QHIO STATE EXPERIMENT L ENKES by : G. L. CATHERS oT T. V. BLOSSER AP M, L, RUEFF*, SEC, AP F. KERTESZ c 1R SITES 51 R. G. SHOOSTER M STATION M T, ROB!NFON 55 M. R. BENNETT 1 G. T. CHAPMAN AP E. DEMSKI PWA H. J. BUTTRAM c L. C WILLIAMS " ! R. L. JOLLEY cT- I G. DESAUSSURE AP J. FLELLIS* AP F. A. KNOX ¢ T FLUX MEASUREMENTS €. J. SHIPMAN or G. ESTABROOK AP W. J, FADER PWA G F. SCHENCK PWA . L K. 4. HENRY AP V. G, HARNESS* AP O PUCKER a MECHANICAL PROPERTIES . BINDER | 55 ; E. B. JOHNSON AP D. A, HARVEY PWA 1. M. DIDLAKE ¢ D. A. DOUGLAS M J. E. KRAUSE PWA - NIT OPERATIONS 1. D. KINGTON AP J. 3. LYNN AP € . DOLLING " 1:: . NI [ J. LEWIN GLM R. E, MALENFANT AP PRODUCTION OF PURIFIED FUELS C. R, KENNEDY PWA ELECTRI OMPONENTS | iW. K. EISTER* I T. A. LOVE* AP E. R. ROHRER* AP J. R WEIR, JR. M ME TALLURGY CONSULTANTS 1 G PIGE | w 1T LONG 4 E. 6. $ILVER AP E. V. SANDIN PWA 6. J. NESSLE C W, WOODS M N. CABRERA, UNIVERSITY OF VIRGINIA Sy s S M. STAINKER k- W. ZOBEL™ AP D. SCOTT, JR. ARE M. R. SWAN, SEC. < K. W. BOLING M N. J. GRANT, MASSACRUSETTS INSTITUTE OF e ROBISON 5o C. L. WHITMARSH i 0. W. CHRISTIAN AP W. C. TUNNELL* AP C.R. CROFT C E. BOLLING M TECHNOLOGY 0. E‘ SCHOW & G. JONES, JR. . o g. 'é filRNBEYw fitcp: g. 'é aLg:LGTHyT' AAE i é\- SS&ZAN g J.T.EAST M J. L. GREGG, CORNELL UMIVERSITY Ml DESICN - L . E. Mg - E. MC - E. J. D. HUDSON M W. 0. JORDAN, UNIVERSITY OF ALABAMA . : F. E. RICHARDSON AP 1. TRUITT c V. G. LANE M : ENGINEERING PROPERTIES . J. SELLERS Ic C. N, ADKINS C | o E. F. NIPPES, RENSSELAER POLYTECHNIC H. E. GOELLER* R M, SIMKONS AP R. T. ATKINS c i E. b. PATTON, IR M INSTITUTE O s R. P. MILFORD D. SMIDDIE 1 CONSULTANTS J. P. EUBANKS c ! B STOWERS JR M W, F. SAVAGE, RENSSELAER POLYTECHNIC J “;- nORc(;E;OLL g - ‘ G. G. STOUT Ic H. A. BETHE, CORNELL UNIVERSITY W. K. R. FINNELL c ‘ C. K. THOMAS ~ M INSTITUTE P. E. REAGAN 55 : ‘ PILOT PLANT : H. WEAVER AP J. A. NOHEL, GEORGIA INSTITUTE OF B. F. HITCH ~ *PART TIME C. W, WALKER M | | R.B.LINDAUER* +. = . cT TECHNOLOGY E.HOLT < T : W.H. LEWIS* . cT . JENNINGS c MASS TRANSFER STUDY ! : WM. CARR JeT W, M, JOHNSON C N BROWDER. . o L. MASSENGILL. C J. L. SCOTT M B Beion p CONTRACTORS R. G. WILEY C H. W, LEAVENWORTH PWA 2 & KESLY ot i METALLURGY CONTRACTORS - B METAL HYDRIDES, INC, F. W. MILES cT NUCLEAR DEVELOPMENT ASSOCIATES, INC, MOLTEN SALT THERMODYNAMICS . NONDESTRUCTIVE TESTING BATTELLE MEMORIAL INSTITUTE TECHNICAL RESEARCH GROUP F, F, BLANKENSHIP C : R. B. OLIVER M BRUSH BERYLLIUM COMPANY M. BLANDER c ; J. W, ALLEN M GLENN L. MARTIN COMPANY 5 S, CANTOR c R. W. MCCLUNG M METAL HYDRIDES, INC. i S, LANGER ¢ ! 0. E. CONNER M NEW ENGLAND MATERIALS TESTING | , M, B, PANISH c . R. A, CUNNINGHAM M LABORATORY | . L. E. TOPOL C ; W. J. MASON M RENSSELAER POLYTECHNIC INSTITUTE i ; 1 SUPERIOR TUBE COMPANY i 4 SUPPORTING STUDIES | INSPECTION UNIVERSITY OF ALABAMA UNIVERSITY OF FLORIDA P. A. AGRON c L A. TABOADA W UNIVERSITY OF TENNESSEE J. E. SUTHERLAND C A. E. GOLDMAN M M. D, DANFORD C ; R. L. HEESTAND PWA 3 G. M. TOLSON M i R. M, EVANS M CONSULTANTS % H. INSLEY 1 T. N. MCYAY CONTRACTORS 1 BATTELLE MEMORIAL INSTITUTE CARTER LABORATORIES T T T— .33 - -,