OAK RIDGE NATlONAI. I.ABORATORY - operated by UNION CARBlDE CORPORATION ErfNUCLEAR DIVISION S ST for the s U S ATOMIC ENERGY COMMISSION ORNI. TM—'- - (LTS - 'fi-:':H*E"‘MéC"??f -"'." d IR wear;_ | J L - fmfi flflfiddgffi' HAS Gery Rca 7 . V1 B9 INVENTIONS OF PATSHT 1 EweD. i . -S_HHTIOE Thls document ,--con!ams lnforrnchon of @ prelammory nufure L ond was’ prepared “primarily for’ ‘internal wse at the Ook Ridge Nahonut Rty e " Laboratory. It s ‘subject to revision or correction und therefore &oes T not represent a hnal ‘repor SR P : , e S e b A e by A 2 1 r b o e iy g e e © ot 2 e 5. s T e i ] 1L o A 4 ke e ey af LEGAL ROTICE . This report was prepared as an account of Gé\;cmmont sponsored work, Noiqhor the United Stof;s, _ nor the Commission, nor any person oefing on behalf of the Commission: A. Makes eny warranty or representation, expressed or Implied, with respect to the accufccy, completeness,. or usefulness of the lnformation contained in this report, or that the vse of any information, apparatus, molhod or process disclosed in this _teport may not infringe privct-ly owned rights; or . B, Assumes any liabilities with respect to ihe use of, or for damcgos nsuhlnq from the use of any information, apparatus, methed, or process disclosed in this report. As vused in the cbove, '"person acting on behalf of the Commlulon” includes any employes or - contractor -of the Commission, or employes of such tontractor, to the extent that such employse - or contractor of “the Commission, or empleyee of such contractor prepares, disseminates, or provides access to, any Information pursuant to hu omploymont or contract with the Commission, - Cor hls lmploymem with such contractor. A P s A R e e) q-’ »E *) &Bm* ¥ ‘ ‘ L ORNL~-TM-1854 iy CFSTI PRICES | T rfie. $.3 %% un - és/ Contract No. W-7405-eng-26 METALS AND CERAMICS DIVISION | , LEGAL NOTICE . This report was prepared as an sccount of Guverment sponsored work. Nefther the United i Btates, nor the Commiasion, nor any person acting on behalf of the Commission: ‘ : +"of any information, apparatus, method, or process disclosed in this report may not infringe i, privately owned rights; or ! o . . ‘-« B. Assumes any lHabilities with respect to the use of, or for damages resulting from the ; use of any information, ipparatus, method, or process dizclosed in this report. MATERTALS DEVELOPFIENT FOR MOLTEN-SALT BREEDER REACTORS H. E. McCoy, Jr., and J. R. Weir, Jr. JUNE 1967 " 0AK RIDGE NATIONAL LABORATORY . Oak Ridge, Tennessee - operated by UNION CARBIDE CORPORATION : : .. for the : U.S. ATOMIC ENERGY COMMISSION b AR LSS él‘E%-g . 4)) ‘5? w) fi%fi ¢ 131 - CONTENTS Abstraet e e w Introduction .. | Status of the'Developmentsof'HaStelloy.N General Properties | . .Physical Properties . . . . . C e e Mechanical Properties . . . . . Tensile Properti€s . v v v v v v « o o o o« Creep Properties . . . . . . . | | Fatigue Propertles . . . . . . « .« . . . . . Effects of Irradisation . . « v o v o v v v v o 0 o . Corrosion by Molten Fluorlde Salts : Loop Studies . . o v v v v v u e w0 . MSRE Operatlng'Experience Resistance to Gaseous Contaminants Oxidation Resistance_r.'.b. « . Q e e e Resistance to Nitriding .. .'. e e e e Compatibillty with Superheated ‘Steam f-.-; . Fabrication of Hastelloy N Systems and ComponentSy Raw Material Fabrication Welding and Brazing of Hastelloy N ... c e e Joining for Reactor Component Fabrlcatlon Pressure Vessel and. Piplng S e e e e Heat Exchangers .~.*.*;~5 e e e a e e . Dissimilar Metal Joints . . .. . . . : Remote Joining .. .',_ _ | | _ " Remote - Welding".ffi }-;*. e e e e e e ~ Brazing and Mbchanieel_Joints. C e e e e Remote Inspectlon f}. e e e e e e e e e e Status of Development of Graphite'_,;”._. . | Grade CCB Graphlte _; . e SETUCEUTE o o o o o e e e e . Permeability . « « ¢« « ¢ ¢ ¢ v o ¢ o 4 e . Page VW U W W | w N NN MNNMNNN N MM OWOWOWOW W W WD R P B 5 0 o MK L oW n:_éS 5 0 360 6D RN M = 30, NP iv - - <;;§ Mechanical Properties: . « v v v v v o 4 ¢ o v o s s o . 53 « Isotropic Graphite . « + v v v v o o v o v o . e 54 JOININE v v v e o e o e et e e e e e e e e 56 Graphite-to-Hastelloy'N Joints . . . . ... ... ... 57 | Graphite-to-Graphite Joints . .:.'. .'.,.'.-. e « os . 60 Compatibility of Graphite with Mblten Salts . . ... ... 60 Rediation Effects on Graphite . . . . . .. ........ 6l Nondestructive Testing of Tubing oo e e e e e e .'_64 Materials Development Program for Mblten-Salt | Breeder Reactors . . . « v v v o v o v s s 0 o . Q:, e e e 65 “Hastelloy N PrOGF&Il . o v v v o o o v v ot v v oo n e e 65 Resistance to Irradiation Damagé'-.\. .;.}, . c .. - 65 Corrosion Program . « o & & & o & o & o o o o o o s o o 67 NIEELAINE « « ¢ v v e e e b e e e e e e e, 69 Joining DevelOPmENt . o v o 4 v v v e aie o e o e s .. 69 Inspection Developmeht e e e e e e e e e e e e e e e T4 7 Materials Development for Chemical Processing | -_ | . i; Bquipment . . . . ¢ 00 0 0 s e e e s e e e e e 75 - Graphite Program . . . . . . « ¢« ¢« « « « . e e e e e e 75 Y Graphite Fabrication and Evaluation - . . . « . . « « ¢ 4 76 Irradiation Behavior . . . . . .. e e e e e e e 77 Graphite Joining . . . . . . . o ¢ o o0 e e e e e 79 Permeability Studles . . . . . . .+ o v o v v o o0 . 79 Corrosion and Compatibility of Graphite O (0 Graphite Inspection . . . . . . . . .+ o o o o o o 80 General Development and Project Assistance . . . . .. ... 8L ACKNOVIEdGEENtS . o v o v o v e e e e e e e e e .. Bl Appendix ... . . . ¢ o o & e e e e e e e e e e 83 { .'¢§(»»ug* 4 J’j vl - o ) B MATERIALS DEVEIOPMENT FOR MOLTEN-SALT BREEDER REACTORS H. E. McCoy, Jr., and J. R. Weir, Jr. - ABSTRACT We have described the materials development program that we feel necessary to ensure the successful construction - and operation of a molten-salt breeder reactor, The pro- posed reactor is a two-region system utilizing a uranium- bearing fluoride fuel salt and a thorium-bearing fluoride blanket salt. A third lower melting fluoride salt will be used as a coolant for transferring the heat from the fuel and blanket salts to the supercritical steam. The primary structural materials are graphite and modified Hastelloy N, The individual fuel cells will be constructed of graphite tubes. These tubes must withstand neutron doses of the order of 1072 neutrons/cm and must have very low permeability: to gases (because of 137Xe entrapment) and fused salts. These requirements mean that we need a graphite that is slightly . better than any currently available. We have described in detail what graphites are available and their respective properties. A line of action for obtaining improved grades of graphite is proposed along with a test program for evaluating these new products. Modified Hastelloy N will be used in all parts of the system.except the reactor core. Since standard Hastelloy N is embrittled at elevated temperatures by neutron irradiation, ‘it has been necessary to modify the composition of the alloy with a small addition of titanium., The program necessary for fully developing this modified alloy as an. engineering material is described in detall ; \ , An integral part of the proposed system is a JOlnt between the tubular graphite fuel channels and the modified Hastelloy N. ‘Brazing alloys have been develoPed specifically for this job “and a reasonable design for the joint has been made. The 1ntegr1ty of the goint must be demonstrated by engineering _tests. . Several areas requlre the development of suitable inspec- tion techniques. These techniques are further complicated by the fact that they must be adaptable to remote 1nspection inside the Ieactor cell _ Although numerous problems exist which will require Turther development, none of these appear insolvable. Hence, we feel that the materials development program can proceed at a rate consistent with that proposed for the Molten-Salt Breeder Reactor. 14"[ilv" INTRODUCTION The proposed molten-salt‘breeder'reactorsllwiiljrequire some advances in materiais technology. However, the'construction and operatioh of the _Mbiten-Salt Reactor Experiment have given us iovaluable iasight concerning what advances are necessary. We feel that we can make these advances on a time schedule that is consistent with that pr0posed for the Mblten-Salt . Breeder Reactor A From & materials standpoint, the reactor 1s easily divided into two sections: (l) the core and (2) the reactor veseelrand associated piping. The requirements of the material for use within the core are (1) good moderation, (2) low neutron absorption, (3) compatibility with the molten salt~Hastelloy N system, (4) low permea.bility to both salt and fiseion ga.ses, (5) fabricable into tubular shapes, (6) capable of being jolned to the rest - of the system, and (7) capable of maintaining all the above. properties after A accumulated neutron doses of 1023 neutrons/cm . Graphite 1s the preferred material for the core, and the Grade CGB graphite used in the MSRE satis- fies most of the stated requirements. The main additional requirements. Nh ® are that we develop the technology for producing tubular shapes of a com- parable material with slightly improved gas permeability and then demon— strate that this material will retain its integrity to neutron doses of 1023 neutrons/cm®. Tubes of the desired quality can be produced in the near future. .Available data on the radiation damage of graphite to doses' of 2 to 3 X 1022 neutrons/cm? indicate that the material is capable-of the anticipated doses. , : o | The rest of the system — the core vessel, heataexchahgers, and piping - - will see & considerably lower neutron flux. The requiremeats of.a‘material ~ for this application include (1) resistance to corrosion by fluoride salts, (2) compatibility with'the core material (3) capable of(being fabricated into complicated shapes by conventional processes such as rolling, forging, and welding, (4) good mechanical strength and ductility at temperatures over the expected service range of lOO to 1300°F, and (5) capable of maintaining- ¥ 1p, R. Kasten, E. S. Bettis, and R. C. Robertson,~Design Studies of i~ 1000-Mw(e) Molten-Salt Breeder Reactor, ORNL-3996 (August 1966) A g{ 4 c - oo, ) (n! . “reasonable strength and ductility after eXposure to a neutron environment. Hastelloy N satlsfies most of these requirements This alloy was developed - at the Oak Ridge National Laboratory specifically for. use in molten-salt lsystems.' HOWever, our experlences w1th this alloy in the MSRE indicate ~that 1t has at least two disadvsntages: 8 propensity_forxweld cracking and severe reduction of high-temperature ductility after irradiation. The - first problem can be solved by using vacuum-melted material, but slight changes in the alloy comp051tion will be necessary to minimize the radia- tion damage problem, We have found that a slightly modified alloy con- - taining O 5wt % Ti has good weldability and reasonable resistance to radiation damsge - _ This report presents the present stste of knowledge of these two 'materials, graphite and Hastelloy'N, as they affect an MSBR. The first section discusses Hastelloy N and the second discusses graphlte The final section briefly presents the program requlred to extend the develOp- ment of these materials to_s stage where they may be used in the MSBER Wnilke considerable development and testing must be accomplished to pro- vide the aSSured.performance'necessary for a resctor_system, it appears that adequate materials can'be'obtsined for a molten-salt breeder reactor. STATUS OF THE DEVELOFMENT OF HASTELLOY N JGeneral\PrOperties '3 During the early stages of the Aircraft Nuclear Propulsion Program, '.msny metals and’ alloys were: screened to determine their re81stance to. ';,f'molten fluorides. Primsrlly on' the. basis of these reSults, ‘the nickel—' 'r,molybdenum system was selected as the most: promising for additlonal study. A molybdenum concentration range of 15 to 20% was selected since this Vylelded s1ngle-phase alloys with their inherent metallurgical stability —:varlous other alloying additiOns were studied in order to 1mprove the | ';mechanical prOperties and cxidation resistance of the basic binary alloy. The Hastelloy N alloy reSulted from this program and was used for the MSRE. The chemical comp081t10n is 11sted in Table l Hastelloy N is a nickel-base alloy that is solution strengthened with | molybdenum and has an optimized chromium content to max1m1ze oxidation Table 1. Chemical Composition Requirements for the Nickel-Molybdenum-Chromium Alloy Hastelloy N Element S L o Wfi'%(é) - Nickel T . bal Molybdemum - 15.00-18.00 Chromwm . 6.008.00 ~ Iron [ - 5.00 Carbon o o _tn‘ - 0.04-0.08 Mangenese o SR _-'i;r: ©.L00 Silicon - , - ~ - 1.00 Tungsten o o Lo e 0.50 - Aluminum + titenium o Q.50” ~ Copper B . S 0.35 - Cobalt | | o 0.20 ~ Phosphorus D — - 0.015 Sulfur - - | - 0.020 Boron . | o . 0.010 Others, total . o S 0.50 Single values are maximum percentages unless other- wise specified resistance and to minimize corrosion by fluoride salts. The chemistry | has been controlled to preclude aging embrittlement. The aluminum, - titanium, and carbon contents are limited to minimize severe fabrication and corrosion problems, and the boron content is limited to prevent weld ,cracking. Iron is included to allow more choice of starting materials for melting. The extreme examples of permissible combinations of elements 'allowed by the chemistry specification were studied, and in no case did. any undesirable brittle phases develop. Carbides of the form Mé305 and M¢C exist in the alloy and are stable to at least 1800°F. The MSRE was constructed using conventional practices (comparable to those used for a stainless steel system) and from material obtained from commercial vendors. The major materials problem encountered was one of weld cracking, which was eventually overcome by slight changes in melting - " L)‘ S ) &) o, qt) (“;' '__ '”practice and by strict quality control of the material 'Heats of material subaect to weld cracklng were 1dentif1ed and thereby eliminated by means . of a special weld—cracking test that was included as a part of the | Specifications Physical Properties Several phy51cal properties of Hastelloy N are llsted in Table 2. Specific heat, electrical re51stiv1ty, and thermal conductiv1ty data all »show inflections with respect to temperature at l200°F This is thought to be due to an order-disorder reaction, however, no changes in mechanical v pr0pert1es are detectable as a result of this reaction. - Mechanical Properties The composition and the fabrication procedure for Hastelloy N have been optimized with resPect to. resistance to salt corrosion, ox1dation resistance, and freedom from embrittling aging reactions. The strength of the alloy is greater than the austenltic stainless steels and comparable with the stronger alloys of the "Hastelloy type." | ‘Design data for this alloy were established by performing mechanical prOperty tests on experimental heats of commercial size. Data from this study were reviewed by the. ASME Boller and Pressure Vessel Code Committee and Code approval wa.s obtained under Case 1315 for Unfired Pressure Vessel construction and Case 1345 for Nuclear Vessel constructlon.l The recognized -~ allowable stresses are tabulated 1n Table 3. The" commercial heats used for MSRE construction exhibited strengths equal to or greater than the - experimental heats... "fTEn31le Properties3 Tensile tests were performed on the experimental heats and Spec1fica— tlcns were thereby established for the commercial heats. Flgure l is a 2R 'W. Swindeman, The Mechanical Propertles of INOR-S ORNL-2780 (Jan. 10, 1961). 2. T. Venard, Ten511e and Creep Properties of INOR- 8 for the Molten-_ '_Salt Reactor Experiment ORNL—TM*1017 (February 1965) Table 2. Ifhyaicnl Properties at Va:bims Temperatures | Flectrical | ' Cosfficient of Thermsl Modulus of Temperature Density Resiptivity . Thermal Conductivity Specific Heat Expansion Flasticity (°F) (a/em®) (1b/1in.?) (uohm-cm) (v em™t °C”1) (Btu ™ nr™t °F°Y) (Btu Wt 7Y (°F)-2 . (1b/4n,2) | _ % 10~8 x 108 75-85 g.93 0.320 120.5 ” 1300 : ' 126.0 ‘340 ' 0.098 572 : : : 0.109 . 1000 ‘ _ _ 0 0,115 212-1752 , R , Lo . 7.0 752-1112 ‘ : o , ' C 8.4 1112-1832 ' ‘ Lo ' - 9.9 212-1832 S 8.6 300 0.12 6.94 75 0.14 8.21 825 0.16 9.25 985 0.18 10.40 1165 0.20 - 11,56 1475 - 0.24 13.87 o 55 ‘ 31.5 425 - 29,0 925 . . 27.0 - 1075 -~ 26.3 ns 26.0 1300 ! 2408 - 1475 23.7 1575 2.7 1750 20.7 . 1825 19.1 - 1925 A7 P Cs _ o ¢ Vo O, “C" 13 I a’) "Table 3. Msx1mum.Allowable Stresses for Hastelloy N Reported ~ Dby ASME Boiler and Pressure Vessel Code Maximum Allowable Stress, psi Temperature (°F) | "~ Material Other Cmso . : - than Bolting Boltrng 100 S 25,000 ' 10,000 200 | o 24,000 | o 9,300 300 | ' 23,000 8,600 400 o | - 21,000 ' 8,000 500 A - 20,000 7,700 600 o 20,000 7,500 700 EEE 19,000 | 7,200 800 | 18,000 7,000 900 18,000 - - 6,800 1000 - 17,000 | 6,600 1100 ' 13,000 | 6,000 1200 . - - 6,000 ' | 6,000 1300 o 3,500 3,500 - ) OfiNL-_-Dfi’G 64-4414R2 o ' TEMPERATURE (°C) - ' | O 100 200 300 400 SO0 €600 700 800 900 1000 140 1T TT- T T 1771 1T T1 T 420 SCATTER BAND FOR- EXPERIMENTAL HEATS 100 o o D Q . WEAT 5075 / - ‘ '4PARALLEL TO R. o.‘ : // - ®NORMAL TO R.D. - - _ - % HEAT 5081 - R | 8 e PARALLEL TO R.D. RREEN . e vNORMAL. TO R.O.. H O TENSILE STRENGTH (1000 psi) L o0 b O 200 400 600 8OO 1000 1200 4400 1600 1800 TEMPERATURE. (°F) ' Fig. 1. Tensile Strength at Various Temperatures of Hastelloy N. \ summary of the ultimate strengths at temperatures from ambient to 1800° for both types of material. Similar data on the 0. 2% offset yield strength are shown in Fig. 2. The values for fracture ductility are presented in Fig. 3. In all cases, the values for the commercial heats were well within the band obtained from the experimental heats. The values from both the .longitudinal and transverse specimens are comparable, show1ng no anisotropy effects. Metallographic data indicate that the heats with low carbon, and | consequently large grain size, tend to exhibit the lower strengths, Tensile tests of notched sPecimenS‘were performed using a'notch o radius of 0.005 in. The notched-to-unnotched strength ratios varied from 1.07 to 1.38 at test temperatures from ambient to 1500°F o | , ORNL-DWG 64-4415R2 ' , TEMPERAT_URE {°C) _ , ' -0 100 200 300 400 S00 €00 700 800 900 {000 0 7 | | 11 T | I 60 S0 /SCATTER BAND FOR / EXPER!MENTAL HEATS . é%fi/% Z 20 ' e — 1 4. HEAT 5075 | ' 4 PARALLEL/TO R.D. ¢ NORMAL TO R.D. 10 I HEAT S081 s PARALLEL TO R,D. vNORMAL TO R.D. 0 1 I - : . ' ‘0 200 400 600 8OO 1000 {200 4400 1600 4800 : TEMPERATURE °F) e 0.2% YIELD STRENGTH (1000 psi) Fig. 2. Yield Strength Values for Hastelloy N. » “C” " “» ‘-J) . " ORNL-DWG €4-4416R2 o : : TEMPERATURE ey o O 100 200 - 300 400 500 600 700 800 900 1000 T 1 I I T 1 ol : : __SCATTER BAND FOR | EXPERIMENTAL‘ HEATS - 70 60 \\\\\\\‘_‘ N & . i i/%//// / /; 8. \\\\\\ i X N N -~ \\ &\ \ o PARALLEL TO R. D v NORMAL -TO- R.D. " 0 1 ‘[, — 1 ' O 200 400 600 - BOO. {4000 1200 400 600 1800 - TEMPERATURE (°F) - Z 30 |- - 7 ol e e % \%z‘ | e e . //// s o Fig. 3. Totai-Elongatidn-Values for HaStelloy'N. Creep. Properties3 Creep*rupture tests were performed on sheet and rod sPec1mens in both air and molten salts Mbst of the testlng was. conflned to the 1100 to 1300°F temperatfire range however, a few teste were_eonducted at ftemperatures up to 1700°F Sfifimary curree'representing etrees vs minimum . creep rate for the MSRE heats of Hastelloy'N are shown in Flg 4. These - data for heat 5055 are plotted to show the tlme to varlous strains at ~ 1300°F in Fig. 5. The rupture life of Hestelloy N plotted as a function of stress is shown in Flg 6. The creep propertles in molten- salt environments were not signlflcantly different from those obtained in air. 10 {x10°) 80 {4100°F) 40 o ORNL-DWG 64-4434R :‘_'_; - v " E (1300°F) w 10 - . {NOR-B s ' AS RECEIVED-TESTED IN AIR | L | MATERIAL FROM MSRE PLATE - 6 ' 1 o HEAT 5055 - A 8 HEAT 5075 (1500°F) - G HEAT s081 . « _ OPEN SYMBOLS ~PARALLEL TO R.D. - CLOSED SYMBOLS ~TRANSVERSE TO R.D. . ! \ : 'o-' K w0 s ot ‘0-3 WMINIMUM CREEP RATE (i '- > —} 1o - ottt Fig. 4. Effect of Stress on Minimum Creep Rate of Ha.stelloy N & 88 H O STRESS {1000 psi) N- o 02% 05 1.0 20 5.0 10 oo .0 0 100 TIME (hr) - Flg. 5. Time to Reach Various Strains for a Giiren' Stress. Numbers in parentheses indicate fracture ductilities. ORNL-OWG 64-4438R 704°C {(1300°F) HEAT 5055 - AIR RUPTURE {000 10,000 (Open symbols — specimens parallel to rolling direction, closed symbols — 8pecimens perpendicular to rolling direction.) ¢ t‘} b1 &) ad (.fl STRESS (1000 psi) 100 50 20 10 ORNL-DWG 64-4418R 593°C (1100°F) ~704°C (1300°F) 816°C_(1500°,F) o'HEAT 5055 & HEAT 5075 0 HEAT 5081 OPEN SYMBOLS - PARALLEL TO R.D. CLOSED SYMBOLS-TRANSVERSE TO R.D. o 0 100 1000 10,000 o 'TIME TO RUPTURE (hr) Fig. 6. Rupture Life of Hastelloy N at Given Stresses and Temperatures. o Fatigue Prop_rties Rotating-beam fatigue tests were conducted on Hastelloy N at 1100 and 1300°F by Battelle MEmorlal Instltute 4 Grain size and frequency were the major variables studied. - The results are shown in Flgs 7 and 8. ) The thermal-fatigue hehav1or of the alloy was investigated at the Univer51ty of Alabama Flgure 9 presents a graph of the plastic straln range Vs cycles to fa:.lure » a.nd the results are seen to obey 8 Coff:m-—type relation The tests 1nvolveé several me.ximum temperatures, as noted, as well as rapid cycling and hold time cycling Furthermore, the data were obtalned from two sPecimen geometries These data show good agreement with isothermal straln-fatigue data on. this alloy. ' An analys1s of these same tests, based on a plastlc strain energy - rcrlterlon, 1nd1cates that the total plastlc work for fallure of Hastelloy N by fatlgue is constant in this temperature range The data have been l_plotted‘in Fig. 10 as plast;c~strain energy Vs cycles.to failure. It is ‘seen that the data for 130Q'afia’1500°r fit’curveS'having the same slope “R. G. Carlsen, Fatigue Studies of INOR-&, BMI-1354 (January 1959), 00 — ' o . ‘ ORNL-LR-DWG 54252 | o' _ : 100 cpm 3000c¢cpm 80 1115.p. 49 COARSE GRAINED o o S.P 19 FINE GRAINED . ~ T0 w ‘ ~ g €0 al . Jouuy _ - ._- - h.~ ~~~~4 s o gl il o | T , < 50 o-HbNoo— " ‘ @ . f %\\ \D#‘--—.._.__ . fl' - B B""'-n-.,, h_# i 5 40 V 0 T . 30 20 1 ! - | 10% ~10* 10° | 108 107 10° . CYCLES TO FAILURE - Fig. 7. Fatigue Properties of Hastelloy N at 1300°F s Rotatirig 2 ORNL LR-DWG 354253 " g0 I H ” TTTTIT 1T T 1711 HHIL 1 1 1 100cem ‘3000cpm S.P.19 COARSE GRAINED o o g0 , S.P. 19 FINE GRAINED . . . . ' 70 o o g @ ol = ?\:\ d ~ | | E 60 ’ Bhiq — e e 3 o g - T~ < 50 m |_ I‘g ) {4 ) Dot b Wt = ‘w40 - 30 ) 20 2 | 103 10* 10° 108 107 108 | ' CYCLES TO FAILURE 'Fig. 8. TFatigue Properties of Hastelloy N at 1100°F; Rotating Beé.m. . S y ' : ' L : - ORNL-DWG 64-4006 10~ @ g £ g 2 [ & 2 10~ g g . 2 5 o . - K z 4 . -2 - E, 0 1600°F E _ bl - -t . & 5 - 9 2 1074 : e — 409 2 -8 0t 2 B -'102 2 -85 - 10 2 s - 0% 2 5 10° N,. CYCLES TO FAILURE Fig. 9. Effect of Stra.in Range on the Thermal Fa‘tigue Life of Hastelloy N. s 3 0% P ‘ . ORNL-Owa €4-4007 - 10° i S - 1300°F W), PLASTIC. STRAIN ENERGY PER CYCLE lin.~ib/in,) o , . 10%. o _ : ‘0% 2 5 1w 2 5 102 2 5 10 2 5. 10* 2 5. 108 ' - N, , CYCLES TO FAILURE . - Fig. lO Relation o:f‘ Plastic Stra,in Energy Absorbed per Cycle to the Fatigue ‘Life of Hastelloy N v ?_ 14 (approximately'—i) but different 1ntercepts. The'intercept'values (at N, = 1/2) are in fair agreement with plastic strain energy values derived from tensfl.e tests at the approPriate temperatures. . | o Effects of Irradiation. Although it has been shown thattHastelloy N'has suitebie:properties for long-time use at high temperatures, a deterioration of high*temperature properties occurs in a hlgh neutron-radiation field 5 we have found that Hastelloy'N is susceptible to a type of high—temperature dirradiation damage that reduces the creep—rupture life and the rupture ductillty Data from 1n-reactor creep-rupture tests run on a. heat of MSRE material are_presented in Fig. 1l. The rupture life has been reduced by a factor | of 10 and the rupture ductility to strains of 1 to 3%. There is,.however, an inuicatiOn that the radistion effect becomes less as the'strese'levels are reduced. W. R. Martin and J. R. Weir, Effect of Elevated Temperature Irradia- tion on the Strength and Ductility of the Nickel-Base Alloy, Hastelloy N, 03NL—TM~1005 (1965) onm.-nws 65-5779R _ 70 o 29 N N\~\ ) €0 . éia , 2199 . 11 | S Negr2 i) | 1 50 N T a—— . 9 . 3 25 UNIRRADIATED o . Y L Il g 40 _ P66 £ 158 fi [ | \\\ . & 30 \33‘.5__,2_5..12.5 _ . LT N L ' ¢ . ’ g, =6x10%m122) N 125 20 Ty <8 '.5‘ "6.... \\i 244>~ 0 30 10° 2 5 o 2 5 0 2 s 100 2 s o ' ' "RUPTURE LIFE (hr) Fig. 11. Creep-Rupture Properties of Hastelloy N at 650°C, Heat 5065. Numbers indicate rupture strains. T ifl ( ) ¥ -x) e 15 . The amount of damage present is a function of the thermal flux and is | essentially independent of fast flux. This supports the hypothesis that the damage is due to the thermal neutrons with the 108 transmutation to lithium and helium thought to be the most probable reaction. The helium | __collects in,the grain boundaries and promotes the formation of intergranu- lar cracks ThlS type of irradiation damage has been found to be quite "'general for iron- and nickel—base structural alloys. Helium.may be formed in Hastelloy'N from two sources. - The normal alloys contain 30 to 100 ppm. B, and, when exposed to thermal neutrons, the transmutation of the 10B Just discussed is by far the predominant source of helium A second source is the (n,q) reactions of fast neutrons with nickel molybdenum, and iron. Thise source of helium predominates after all the 10B has been transmuted (requiring a thermal dose of approximately 1021 neutrons/cmz) or in materials Where the 108 content 1s - extremely low. In the reference MSBR there is no- stressed metal in the core and the 3Hastelloy'N is shielded from the core by the breeder blanket Most of the neutrons reaching the metal will originate as delayed neutrons that are born with t00 low an energy level to produce the fast neutron reactions.\- Therefore, the best way to reduce the amount of helium in the metal is to reduce the boron content . The main source of boron in commercial alloys is the melting crucible, and we have found that with reasonable care ‘commercial heats can be prepared' »,;with boron in the range L—5 ppm. However, the proPerties of irradiated . metal of this low boron content are no better than ‘those of metal contalning fi750 ppm B 1ndicating that the threshold helium levels necessary for damage can ‘be. produced in metal containing below 1 ppm- B. Hence, it appears that ffreducing the boron level alone is not an adequate solution to the prdblem. We have found that the resistance of Hastelloy N to irradiation o damage can be improved by slight changes in chemical composition. A r*reduction in the molybdenum content seems desirable to prevent the forma- tion of mass1ve MSC in the alloy. Flgure 12 shows the 31mpler micro- ldd;structure of the a110y containing 12% Mb Further increases in the 6. R. Martin and J. R Weir, Jr., Solutions to. the.Problem of ’ High-'.I‘emperature Irradiation Embrittlement ORNL-E[M-1544 (June 1966) r. L 3 ' L 14% Mo . - | _ w:“ 'Wfs% M;' Lol Fig. 12. Effect of Molybdenum Content on the Microstructure of Ni=7% Cr—0.2% Mn—0.05% C Alloys After Annealing 1 hr at 2150°F. ” ' ‘ . | S ) ( : ¥ - Sy @y 17 mblybdenum content result'ip_'the' 'formei_tion of the -molybdenumfijich MeC and - have little effect on st’rength’ ' The addition of small amounts (approx 0.5 wt %) of Ti, Zr and HE reduces the irradiation damage problem signif- | 1cant1y F:Lgure 13 illustrates the fact that several alloys have been ‘developed with propertles after irradlatlon that are superior to those of unirradiated starid_ard Hastej_.ley N. V.We are beginning work to optimize the 'c‘:ompos.it.ions and heat treatments forvth_ese alloys. Ve are also initiating the procurement of 1500 1b commercial melts of some of‘the'm‘ore attractive compositions. ' | | The ex-reactor proPerties of the modified Hastelloy N seem very attractive.. Strengths are slightly better than standard Ha.stelloy N and fracture ductilities are about double. The weldability of the titanium- and hafnium-bearing ’a'lloys appears excellent;-'however, ‘we realize that welds need to be made under',_higher reetraint and in 1afg'er sections than ORNL-DWG 67-3523 Nl(22) . ‘\\ J €0 T 1= UNIRRADIATED AIR MELTS ™ -_\\, 50 | e 1IN L Hue0s L) | - . o8 - 408 Hf'“‘. ‘ M (o570 ' §40 : ""'"~\‘ - ”“ N2/ .,” 'i 1 2 1 R _.._\\figs_fl.c)_ \\\ (15 osm L)I l | o IRRADIATED AIR MELTS{ ~ | 8 . | a0 F | \\ ™ ,49051: U | 52 1 TN {3, m-—L) 20 -+ IRRADIATION couomous ——t ey NS 1 | r=esoc T . | K- ¥ -a-suo”m o - L ‘ ‘. - .7._. . : . - : - C e 0 1000 - ',-_';——r—-TlO.OQO : RUPTURE LIFE (br) Fig. 13, Comparison of the Postirradlation Creep PrOperties of Several Hastelloy N Alloys at 1202°F. The first number indicates the fracture strain, the second indicates the alloy addition, and the thlrd indicates the source, C = commercial, L = laboratory have been studied thus far. The zirconium'eddition appears'very detri- ¥ mental to the weldability with extensive weld metal cracking resulting o Thus we feel that the further development of the 21rcon1um-bearing alloy ' must await the development of a suitable filler metal | ' Corrosion by Molten Fluoride Salts A unique feature of the molten-salt\type reactors_ie'the_circuiation of molten salts both as the fuel and the coolant media. The Salts_being " used are mixtures of metal fluorides. Essentially no experience is | available from‘indnstry on handling such salt mixtures at the proposed temperatures,‘tut much information has been produced at ORNL. B o Studies on molten fluoride mixtures for reactor applieatione began . . in(the early 1950's and gave-prinafy consideration to the compatibility of these salt mixtures with stiucturel materials. In the intervening 15 years, an extensive corrosion program has been conducted at,ORNLrwith" several families of fluoride mixtures using‘both commercial and develop- mental high-temperature alloys.”™ 15 As a consequence, the corrosion - : 7L. S. Richardson et al., Corrosion by Molten Fluorides, ORNL-1491 (Mar. 17, 1953). | 8G. M. Adamson et al., Interlm.Report on Corrosion by'Alkali—Metal Fluorides: Work to May 1, 1953, ORNL-2337. °G. M. Adamson et al., Interim Report on Corrosion by Zirconium— Base Fluorides, ORNL-2338 (Jan, 31, 1961). 10y, B. Cottrell et al., Disassembly and Postoperative Examlnation of the Aircraft Reactor Experiment, ORNL-1868 (Apr. 2, 1958). ' o 11y, D. Manly et al., Aircraft Reactor. Experiment —-Metallurglcal Aspects, ORNL-2349 (Dec. 20, 1957) pp. 2-24. 12y, D. Manly et al., Prog. in Nucl. Energy, Ser. IV 2, 1964 (1960). 137, A. Lene et al., pp. 595-604 in Fluid Fuel Reactors, Addison- ~ Wesley, Reading, Pa,, 1958, ~ 14Molten-Salt Reactor Program Status Report ORNL-CF-58-5-3, pp. 112-13 (May 1, 1958). | 153, &, Devan and R. B. Evans III, "Corrosion Behavior of Reactor Materials in Fluoride Salt Mixtures," pp. 557—79 in Conference on Corro- e gion of Reactor Materials, June 4-8, 1962, Vol, II, International Atomic "\Ej Energy Agency, Vienna, 1962. ‘ . 1Y C «} N &) 19 . “technology for the molten-salt'reactor'concept’is’now in an advanced stage of development Furthermore, container materials are. available that have ghown extremely low corrosion rates in fluoride mixtures at temperatures _ considerably above the 1000 to 1300 F range proposed for the MSER. - Unlike more conventional oxidi21ng media, ‘the . products of oxidation of metals by molten;fluorides tend to be completely,soluble in the cor- roding media; hence, passivation is preciuded, and corrosion depends ~ directly on the thermodynamic driving force of the corrosion reactions. Design of a chemically stable system utilizing fluoride salts, therefore, requires the selectiOnVOf:salt'constituentsrthat7are not'appreciably reduced by available StruCtural metals and the-development of containers whose components are in near thermodynamic equllibrlum with the salt ‘medium. -TIn practical applications, the maJor source .of corrosion by fluoride mixtures has derived from trace,impurities in the melt rather than the major Salt'constituents TheSe impurities may originate within »the melt or from oxide films or other contaminants on the metal surface, | Further details concerning corrosion mechanisms in fluoride melts have been presented previously, 2 Corrosion data on Hastelloy N in LiF -BeF; - ‘based salts have been generated in out-of-reactor thermal- and forced convection loops, in- reactor capsules,,and more_recently_the MSRE, Loop Studies The origlnal studies that serve as a base for the MSBR corrOS1on studies were conducted at ORNL a8 part of the Aircraft Nuclear PrOpulsion Project 713 Work on this project wag aimed at a 1500°F maximum tempera- v'ture. Nickel alloys were shcwn to be the most promising for containing the fluorides however, strength considerations restricted the candidate -materials to the commercial nickelvchromium group,_with Inconel 600 ‘receiving the most attention. These‘alloys’incurred corroSion by selective oxidation of chromium through trace impurities and by mass transfer of chromium through a redox reactionrinvolv1ng UF4.' Corrosion‘was in the form of subsurface vbids'withfithe-depth'prop0rtionalrto the test time and temperature. Using information gained in corrosion testing of the commercial | alloys and from_fundamental,interpretations ofthe.corrosionprocess, an . alloy was developed at ORNL to provide improved coerSion resistance as - well as-acceptable mechanical:prOperties; The alloy system used as the,- basis for this'deveIOpment wasrcomposed_ofrnickel with a_prinary_strengthe ening addition of 15 to 20%-Mb.. Evaluations of other'strengthening addi- - tions culminated in the selection of an alloy containing 16% Mo, 7% Cr, and 4% Fe (INOR-8, now Hastelloy N). | . S 'The corrosion testing program for Hastelloy N involved three sequential phases., In the first phase, the corrosion properties of 13 fluoride salt mixtures were compared in Hastelloy N thermal-convection loops operated for 1000 hr. Specific salt mixtures, whose compositions are_listedhin Table 4, were selected_to_provide an evaluation of (1) the corrosion properties of beryllium-bearing fuels, (2) the corrosion properties of beryllium-fluoride mixtures containing large quantities of thorium, and (3) the corrosion properties of nonfuel-bearing fluoride cooclant salts. The second phase of testing, which again used thermal-oonvection 1o0ps, involved more extensive investigations for longer time periods and at two teaperature levels. The third phase of the'testing was conducted'in forced-circulation loops at flow rates and temperature conditions simu- ' lating those of an operating reactor system. These 100ps Operated with maximum fuel-salt temperatures of 1250 to 1500°F, ma.ximum coolant salt temperatures of 1050 to 1200°F, and all with approximately 200°F tempera- - ture drOps. A total of 49 thermal loops and 15 forced-circulation loops - were operated during these phases of the. program | o | Essentially no attack or deposition was found with any of the - fluorides in Hastelloy'N lOOpS in the phase-I studies. The max1mum‘ | attack found after 8760 hr in the phase-II tests was a 1imited surface 7 ‘roughening and pitting to a depth of 1/2 mil Attack in most cases was accompanied by a thin surface 1ayer. ‘The typical appearance of a hot- leg surface is shown in Fig. l4 Nb deposit or other evidence of mass transfer wvag found in any of the cold legs | A - In the third phase of the _progranm, employing forced—convection lOOpS, tubular inserts in the heated sections of some of the 100ps provided infor- mation about the weight losses occurring during the tests. 8Salt samples .’) 4 C 4 [ Xy 8 _ Salt Mixture . - e e s 21 Table 4 Compositions of Mblten-Salt Mixtures Tested for Corrosiveness in Thermal—Convection Loops e - fgcomposition,-moles% - . NeF L KF %P, BeF, R, THR ‘*g_Fuelfsnd,Blanket Salts w2 os7 e 2o . 553 . 4007 4 D am T s s o5 o5 126 s e 1 127 s s 7 128 St 29 e oma e m e 13m0 16 13 134 e . 35 05 1 ‘s s 455051 136:tm m _ t; ti '7Q:i:fl‘:' H. | 107 - . 20 7 Bl +osU e 185 o5 1 _ Coolant Salts | "iftaken from the pump bowls provided a semicontinuous indication of ;;'f "ff;;corrosion-product concentration in‘the circulating systems.11 A summary ':_[_fof the oPeratlng conditions and results of the metallographic examination _ijfof the forced- 1rculation loops are presented in Table 5 the that o .'~;;the operating times of these systems range from a minlmnm of 6500 hr to a rs;i’tmaximum of 20 OOO hr., Nine of the loops were - operated for over 14, OOO hr. The corrosion rates of Hastelloy N in ‘the pumped 1oops operating with -'maximum temperatures between 1300 and 1500°F (ref.15) indicate that 22 Fig. 14 Appearance of Metallographic’Specimen from Hot leg (1250°F) of Hastelloy N Thermal-Convection Loop. Operating time: 8760 hr. Salt mixture: LiF-BeFp-UF,-ThF, (62—36.5-0.51 mole %). The small "voids" in the microstructure below the surface layer are microconstituents in the Hastelloy N which have been darkened and partially removed by metallographic etching. corrosion reactions effectively go to completion in the first few thousand hours of loop Operation. As shown in Table 6, weight losses in 10, 000- and lS,OOO—hr testsrin pumped loops containing LiF-Bng-UF4 salts showed no measurable increase'after,the-first_50007hr of 0peration;furthermore, changes in concentrations of corrosion products'in the circulating salt leveled off in the same time period. The temperature dependence of the maximum corrosion rate, as judged from total weight losses of" sPecimens,‘ was less than a factor of 5 over the temperature range investigated | (Table 5). 30wever, the metallographic appearance of the loop surfaces was noticeably affected by the test temperature. At the highest test i temperature (1500°F), Hastelloy-N surfaces were depleted of chromium, as indicated by the appearance of shallow subsurface voids to a maximum depth of 4 mils. At 1300 and 1400°F the surfaces exhibited no evidence of attack or other metallographic changes during the first 5000 hr of 0pera-t tion, at still longer test times a thin continuous 1ntermetallic layer was X) (-" . &X) W stle-5 Hastelloy N Qperating Conditions of Fbrced Convection Loqps and Results of 85ce Table 4. Mstallographic Examinations of Loop Materials P S R Maxinmum | | ' . ‘ - | ~ Duration | ' Flow » L, Loop of Test Salt Mixturea - Fluid-Metal em Reynolds Rate Results of.Mstsllographlc Nunber (hr) - | ~ Interface (°F) Number (gal/min) Examination - Temperature ' & ' o o (- 9354-1 14,563 126 . 1300 200 2000 2.5 Heavy surface roughening and U e B - | plttlng to 1 1/2 mils 9354-3 19,942 "‘ff847f‘ sfl*~ 1200 1100 3000 2.0 No attack, slight trace of _ ' o “'Tfi; j . _;J L | - metallic deposit in cooler coil 9354-4 15,140 . 130 1300 200 3000 2.5 No attack 9354-5 14,503 130 1300 200 /3000 2.5 No attack N MSRP-6 1 20,000 134 1300 200 . 2300 1.5 --Pitted surface layer to 2 mils ' MSRP-7 20,000 133 1300 200 3100 1.8 Pitted surface layer to 1 mil MSRP-8 9,633 124 1300 200 4000 2.0 No attack 'MSRP-9 9,687 134 1300 200 2300 1.8 No attack MSRP<10 20,000 135 1300 = 200 3400 2.0 Pitted surface layer to 1/2 mil MSRP-11 20,000 123 1300 200 3200 2.0 ' Pitted surface layer to 1 mil MSRP-12 14,498 134 1300 200 2300 1.8 No attack MSRP-13 8,085 - 136 1300 200 3900 2.0 Heavy surface roughening and T el = ' pittlng MSRP-14 9,800 - Bu-14 + 0.5 U 1300 200 Pitted surface layer to 1/2 mil MSRP-15 10,200 Bu-14 + 0.5U - 1400 200 Pitted surface layer to 2/3 mil MSRP-16. 6,500 Bu-14 + 0.5 U 1500 200 Moderate subsurface void forma- " tion to 4 mils 24 Table 6. Corrosion Rates of Inserts Located in the Hot Legs of Hastelloy N Forced—Convection Ioops as & Function of Qperating Temperature ' Loop temperature gradient: (360°F) Flow rate: = approximately 2.0 gal/min Reynolds number: approximately 3000 Insert b Time Welght Loss Equivalent Ioss Loop Salt . Temperature per Unit Area in Wall Thickness Numb M & o er Maxture” "m0 Tg/en?) T () 9354-4 130 1300 5,000 1.8 2.0 - ~ | - 10,000 2.1 2.3 o . \ 15,140 1.8 2.0 - 'MSRP-14 Bu-14 1300 2,200 0.7 0.8 - ‘ , 8,460 3.8 4.3 ; 10,570 - 5.1 5,8_' MSRP-15 . Bu-1l4 1400 - 8,770 .11.2c 12.7 T -~ - 10,880 10.0° 11.2° MSRP-16 Bu-14 - 1500 5,250 9.6c 10.9 . 7,240 9.0 9.1 83a1t Compositions: 130 LiF-BeF,-UF; (62-37-1 mole %) Bu-14 LiF-BeF,-ThF,;-UF,; (67-18.5-14—0.5 mole %). Same as maximum wall temperature, Average of two inserts. faintly discernible. PFigure 15 illustrates the metallographic-appearance ~ of hot-leg surfaces from Hastelloy N pumped loops after various operating ‘times at 1300°F. Chemical analyses of exposed surfaces suggested that the layer was an intermetallic transformation product of the nickel-molybdenum system. - Although most of our work was with Hastelloy N, we looked at the behavior of several nickel-base alloys with variations of the Hastelloy N - composition.1® Several alloys were screened by_thermal-convection_loop, tests at 1500°F in salt 107 (composition given in Table 7). 'The composie -tion ef these alloys and the depth of corrosion are shown in Fig.-lé. 165, H. DeVen, Effect of Alloying Additions on Corrosion Behavior of Nickel-Molybdenum Alloys in Fused Fluoride Mlxtures, M.S. Thesis, the Uhlversity of Tennessee, 1960. - ' ‘w 7 " . . ¢ 4 i & e . | s = [ - 0070 - 19459]. 9,700 hr 4,500 hr 20,000 hr Fig. 15. Effect of Operating Time on the Corrosion of Hastelloy N ’ Forced-Convection Loops Operated with Mixtures of LiF-BeF,-UF;-ThF, (62—-36.5-0.5-1 mole. %g Maxlmum_salt-mgtal‘interface temperature: 1300°F. = Loop AT: 200°F. The. differencee in microstructure smong the three specimens below the surface layer are attributable to differences in metallographic etching techniques rather than the test conditions. Table 7. Composition of Fluoride Mixture Used to Evaluate Experimental Nickel—Mblybdenum Alloys Component . Mole% - . Welght® F . 453 24 KF ;;'a;;'41 o 494 'asalt-107jgliQfiiGUS_témperéfure;490°Cr DEPTH OF CORROSION {in. x 103) " ALLOY ADDITION {at%) 26 . UNCLASSIFIED ORNL—LR—DWG 46944 [ @ N 3 N woor- S 7 — />\ \‘ % | S NN Nem ADAD D DN ANN Y vo | 1036|1268 | 1135|1078 | 1400 | 102 | 10.2 [48:77 |14.20| 1042 | 966 | 1005 ol T [es 4.-30 479 sa7| |ss0 654 | Fe - .56 | 0.3 al | 140 |61 [208 [ n30| |24 Ler | ~ |306 |55t ._1.55_ 264 n 202| 271 10| fres| |is| Nb [3.23|087 | 148 | ‘ 291 0.9‘2 1'.57. ) __ 6;39_ Fig. 16. Depths of Corrosion Observed for Nickel-Molybdenum K Alloys with Multiple Alloy Additions Following Exposure to Salt 107. Bars desig- nating corrosion depths appear directly above the alloy compositions which they represent. = (Where bars have both positively sloped and negatively. 'sloped cross-hatching, the height of positively sloped cross-hatching indicates depth of corrosion after 500 hr and combined height of both types indicates depth after 1000 hr. ) Chromium increases the corrosion rate significantly. With chromium pfesént, the addition of about 2 at. % Ti does not appear to increase the corrosion rate. Hence,'we feel that the small titanlum addltlon that we are making “will not be harmful at 1300°F. #) 9 27 ;;~MSRE 0perating ExperiEHCE ;3{:4ff;v The MSRE initially underwent a series of zero- and low-power tests. , sThe chemical analyses of the MSRE fuel and coolant salts showed very little ”-change, indicating a 1ow corrosion rate ‘of the Hastelloy N-and negligible _Ttransfer of impurities from the graphite to the salt, This was substanti- ated by the examination of graphite and metal sPecimens after the initial tests, ' N . » ' The full-power operation of the reactor began in March 1965 The first group of metal and graphite surveillance Specimens was removed in July 1966 after eXposure to the reactor core environment for 7612 Mwhr _ which included a - thermal exposure of 4800 hr at 1200°F The peak thermal dose was 1.3 X 1020 neutrons/cm and the peak fast dose >l 2 Mev was 3 x 10%° neutrons/cm The concentration of chromiwn in the fuel salt increased from an initial value of 37 to 48 ppm where 1t remained fairly steady during the 7612 MWhr run The iron and nickel concentrations remained constant at values of lO and l ppm, respectively Visual exam- ination indicated the metal and graphite specimens to be in ‘excellent condition. Machining marks were quite visible on the graphite gpecimens, Several of the graphite and metal specimens were given to the Reactor Chemistry Division for analysis with respect to fission product content 17 The graphite specimens have heen stored and mechanical property tests will mhe run in the near future.i The Hastelloy N specimens were examined metal- - ;lographically and subjected to tensile and creep—rupture tests.; Control specimens for ccmparison purposes were exposed to MSRE—type fuel salt and - -duplicated the thermal history of the in-reactor specimens Metallographic examination of the Hastelloy'N showed the presence of ‘1fa surface film at’ points where “the graphite and Hastelloy N vere in f'"intimate contact This product is shown in Fig. 17 and was present on _'both the surveillance and the control specimens.r Electron micrOprObe -p;:fexamination showed that the carbon content of the edge of the speCimen '7i{ranged from 0. 3 to 1 2 wt % compared with a value of 0 05% for the inte- 'f;rior of the sample Nb evidence ef attack or carburization was found on 17y, R. Grimes, Chemical Research and Development for Molten-Salt Breeder Reactors, ORNL- TV-1853 (June 1967). Y-78266 = tra 3 T 0.007 INCHE =" 500X T b Fig. 17. Edge of Hastelloy N Specimen Exposed ‘to Molten Salt for 4800 hr at 1200°F. Surface in intimate contact with graphite. Etchant: glyceria regia. metal surfaces that were separated from the graphite 1/16 in. This car- burization is not a concern for the MSRE since we were aware of the problem "before the reactor was built and placed sacrificial shims ‘between the graphite and metal where contact was necessary. Several of the surveillance and controi SPeeimens were'eraluated by tensile tests. The total elongation at fracture when deformed at a strain rate of O. 05 min~?t ig shown as a function of temperature in Fig. 18. Both heats show some refiuction in ductility at lowrtemperatures in the irradiated condition with a greater reduction being observed'for_heat 5085. At temperatures above 932°F, the ductility of the irradiated and the control material decreased with 1ncrea31ng temperature, w1th the 1rrad1ated materlal showing a greater loss in ductility. At temperatures above 1202 to 1292°F, the control material exhibited improved ductility, whereas the ductility of the irrediate&'materialvconfinued to deérease. 29 " ORNL-DWG 67-2452 - CONTROL = IRRADIATED 5 | 1 a © 508 o . A e s063 €0 q ] €=.05 MIN"! A R 1 l | - V. T ' : ‘ ' 4 o , s0 =14 &% 7 - ; . LY e |4 A \ A | - T : ¢ gqo / ' . : \ . : l/, b 9’ 9 —ne S —— 3 : " . % _..*_.‘—. — e s & S g 0 \. \ : ) ,, . ‘ -. - : ) . .)4 a -x_gy L _ 430 1 — < k3 - N B} N 20 * 10 = _ ~ o e o [ | 1 a4 L . I ) 1 2 o 100 200 300 400 800 600 7T00 ~ 800. 900 1000 ‘ TEST TEM P, %c ~ Fig. 18. Comparative Tensile Ductilities ofIMSRE Surveillance - Specimens and Their Controls et a Strain Rate of 0.05 min-1. We compared the ductilities of the surveillance sPecimens with those for specimens irradiated in. other experiments without salt present Heat 5081 had ‘been irradiated previously in the ORR.18 The ‘ORR experiment was run at 1292 F to- a thermal dose of 9 X 1020 neutrons/cm and,the material was in the as-received condition The MSRE surveillance speci- mens were run at 1202° F to a thermal dose of 1.3 X 1020 neutrons/cm and ‘wlthe preirradiation anneal was different - However, none of these differ— '_ences are thought to. be particularly significant and - the results can be ’ compered Figure 19 shows that the postirradiation ductilities of o heat,5081 after both experiments are very. similar. | ' . Several of the surveillance specimens were evaluated.hy‘creep—rupture_ ‘tests. . The results of these tests are shown in Fig.VZO _The specimens 18y, R. Martin and J. R. Weir,'"Effect of Elevated-Temperature Irradiation on Hastelloy'N," Nucl Appl. 1(2), 160-67 (1965). 30 - ~ ORNL-DWG 67-3531 = 10 § T s T z ¢=0002 min™! e ¢ | & THIS STUDY, ¢, =13 510", %' 08 T=650°C g 47 & ORNL-TM-100S, ¢, -9-|oz°m Y r=700°C 2 : "c‘_, 06 + o m - 4 & 5 o4 - v x o A 5 \ | 2 3 o2 7 < & . 400 SO0 600 700 800 800 1000 . TEST TEMPERATURE (*C) Fig. 19. Comparative Effects of Irradiation in MSRE and OBR on the Ductility cf Hastelloy N, Heat 5081 onm.-pwe-cufls - o : \ifl , 1 uoes.@...-zxn’ ORR N : A;oasp..-um'.mnz : \ 1 loss 3_@m-.3no'?nsnt ' €0 N N 1] 80 N UNIRRADIATED ALLOYE %1, tzg} N 3 {9 | | 'H j L § | TROS lus n N 40 L &-H "y O= " 2 AVERAGE CURVE FOR SEVERAL' || ____.3-...__ N : RADIATED ALLOYS (ORR) Ayl R ¢ 4 a8 ' S NN "\ {125] M 20 NS ‘2 N / 10 'Y 0 ) " 00 1000 10,000 RUPTURE TIME (KR) Fig. 20, Comparison of MSRE Surveillance Specimens with Specimens Irradiated in the ORR. Numbers in parentheses indicate ductilities. » 1) ¥ product absorption.-- from the MSRE had propertieS"rery'similar'tO“those observed for specimens - irradiated to comparable doses in the ORR. The most 1mportant question to be answered concerning these data is how - they apply to the operation of the MSRE The surveillance specimens were exposed to a thermal dose of 1.3 x 1020 neutrons/cm ~ (The MSRE vessel will reach this dose after about 150,000 MWhr of operation.) ‘This burned out about 30% of the 10B ~and produced a helium content of about 107 5 atom fractlon 1n both heats. The high- temperature tensile and creep-rupture properties are exactly what - we would expect for this dose Our work indleates a saturation in the | degree of radiation damage at a helium atom fraction of about 10'5 and we feel that the properties of the material will not deteriorate further | The.lowetemperature_ductility reduotionmwas not expected. It is thought tO'be_a‘result'of graineboundary‘precipitates forming“due to the long thermal exposure. Irradlation playa some role in this process that is yet undefined The low-temperature properties are not "brittle“ by any stan- dards, but will be monitored closely when future sets of surveillance speC1mens are removed. L -' ' o . | The MSRE has since operated for a total of approximately 32,500 Mwhr. - The chromium content of the salt has been somewhat higher during this run at approximately 60 ppm we removed a group of surveillanoe speoimens | ,(graphite and Hastelloy N) for study about May 15, 1967. The Hastelloy N _removed was. of the modified type, With small additions of titanlum ‘and- izirconium The graphite and metal specimens will be evaluated with respect ;:to corr031on, metallographic changes, mechanical properties, and fission .Reeistancefto-GaseOustontaminants o VOxidation Resistance'h . The ox1dation resistance of nickel-molybdenum alloys depends on the .l_rservice temperature, the temperature cycle, the molybdenum content - and “'dthe chromium content, . The ox1dat10n rate of the binary nickel-molybdenum --alloy passes through a max1mum for the alloy containing 15% Mo, and the '-scale formed by the ox1dat10n is- N1M004 and NiO. Upon thermal cycling 32 from above 1400°F to below 660°F;_the'N1M004'undergoes a phase transfor- mation that causes the protective scale on the oxidized metal to spall. Subsequent temperature cycles then result in an accelerated oxidation rate. Similarly, the oxidation rete of nickel-molybdemum alloys contein- ing chromium passes through a maximum as the chromium content of the alloys is increased from 2 to 6% Cr. ‘- Alloys containlng more ‘than 6% Cr are insensitive to thermal cycling and to the molybdenum content because B the oxide scale is predomlnantly stable Cr203 An abrupt ‘decrease (by a factor of about 40) in the oxldation rate at 1800°F is dbserved when the chromium content is increased from 5.9 to 6.2%. | - | The oxidation resistance of Hastelloy N is excellént#‘and'centinuous" ~ operation at temperatures up to 1800°F is feasible. Intermittent use at - temperatures as high as 1900°F could be tblerefed.'-For'temperetures;up'__ to 1200°F, theVOxidatiOn'rete is not measurable; it'is*essentielly zero after 1000 br of exposure in static air; as well as in nitrogen containing small quantities of air (thefMSRE'cell environment). It is estimated that oxidation of 0.001 to 0.002 in. would occur in 100,000 hr ef opera- tion at 1200°F. The effect of temperature on the oxidation rate of ‘the alloy is shown in.Fig 21. ' Resistance to Nitriding Hastelloy N does not-react-noticeably with nitrogen at temperatures up to 2200°F in the absence of irradiation. 'waerer, in a neutron environment, the nitrogen is dissociated'and same shalloW‘surface-reac-“ tion (a few mils) is encountered. Although only limited data are avail- able, there appears to be a considerable variation in the depth of such | layers with various lots of material. While such layers are not thick - enough in themselves to affect the strength of the proposed cemponents, - they are brittle and crack eas1ly, 50 that they could possibly act as stress risers. Compatibility'with Superheated Steam In parts of the system Hastelloy N will be exposed to'snpercritical"' steam having & maximum temperature of 1050°F and.a pressure of 3800 psi. Some heat exchange equipment will have coolant salt on one side and steam 1) | (WElG;HT‘. GAI‘~)2 ' (mg/cmé)z‘:_ 33 , ' T S ~ORNL-DWG 64-2855R 4.5 ~ INOR-8 4.0 35 v o N K » o. 1.5 1.0 1800°F o —T"* 0.5 g L ol et VT Y %0 s 10 10 200 250 300 30 400 " Fig. 21. Effect of Temperature on Oxidation of Hastelloy N. 34 on the other side. Because.of the higher pressure of the'steam, a failure in the Hastelloy'N7wou1d resultrin:steam.being forced into the coolant- salt circuit. Thus, we cannot adopt the "cheap material and. frequent maintenance" policy of the power industry. The stressacorrosion’type of failure in boiling water and in superheated steam systems'has_been cbserved for the austenitic stainlese steels in several nuclear systems;l?’zo_ The resistance of high-nickel alloys'to this type of attack has also been established and it seems reasonable that our system should utilize such an alloy. 19, 20 , Hastelloy N has been included in steam-metal compatibility programs at BNWL, 21522 gn3 some of the results of these studies are sumarized in Table 8. Hastelloy N exhibited much better resistance to corrosion by steam than type 304L stainless steel. It also compared very favorably with the other nickel-base alloys that were included in these evaluation programs (e.g., Hastelloy X, Inconel 600, and Incoloy 800). Although the depth of attack varied slightly with steam pressure and oxygen content, a .rather pessimistic value for the corr081on rate is about 0. l mil attack in 2400 hr. Assuming a linear rate of attack, which again 1s on the pessi- mistic slde, the depth of attack in 20 years would be about 7 mils. Assuming the more realistic parabolic behavior, the expected depth of attack would be only 0.8 mil. Either of these values is quite acceptable | and we feel that Hastelloy N is a very attractive material for use in the steam circuit %W, 1. Pearl, G. -G. Gaul, and G. P. Wozadlo, "Fuel Cladding Corrosion Testing in Simulated Superheater Reactor Environments: - Phase II. Iocal- ized Corrosion of Stainless Steels and High Nickel Alloys," Nucl Sci. Eng. 19 274 (1964) ' 200, N. Spalarls et al Materials for NUclear Superheater Applications, GEAP-3875 (1962). 21p, Claudson and R. E. Westerman, An Evaluation of the Corrosion Resistance of Several High Temperature Alloys for Nuclear Applications, BRWL-155 (November 1965). S 22p, T, Claudson and H. J. Pessl Evaluation of Iron- and Nickel-Base ‘Alloys for Medium and High Temperature Reactor Applications, Part II, BNWL-154 (Nbvember 1965). Table 8. Summery of Data on the Corrosion of Hastelloy N in Steam Test Results . Welight Gain (mg/cm?) Oxide Penetration (mil) Efivironment?éhdfTesfilCon&ifiions '.?-'Héstelldy_NI . Type 304L Stainlegs_Steel_ ] H&Stélldy'N - Type 304L Superheated steam, cos e 1022°F, 3000 psi, 1062 hr : Deoxygenated.steam,‘;‘-. L 932°F, BOOO'psi, 2400 hr 932°F 5000 psi, 2400 hr =~ = (< 50 ‘ppb), 1022°F, 3000 p51,-' 2400 hr - o . Oxygenated steam uto4mm,maw mmpfl 2400 hr Helium plus 15 torrs water vapor,f - 1500°F 300 hr, ~ 1 atm pressure 0.57 0.109 2.20 0.37 . Stainless Steel{f 36 Fabrication of Hastelloy N Systems and Components The fact that Hastelloy N is a'material_from which a complicated reactor system may be constructed has been demonstrated by the'successfnl construction and operation of the MSRE. Evidence of its general recog- nition by industry is demonstrated by its acceptance and.approval by the ASME Boiler and Pressure Vessel Code, | " Raw Material Fabrication Hastelloy N has proven to be readily fabricable into the many raw material forms (pipe, tube, plate,'bar, castings,etc.)‘required in the fabrication of a compdex engineering system such as a nuclearreactor. Fabrication of the various rew material items required for the MSRE was by normal commercial sonrces Forming techniques'deveIOPed'for austen- . itic stainless steels can usually be used for Hastelloy N with a small increase in temperature. Melting and casting is carried out by using conventional practices for nickel and its alloys. Ingots for the MSRE were prepared by both air- and vacuun induction or consumable arc melting Individual.melts up to 10,000 1b and a total quantity of almost 200,000 1b were produced | Castings have been made in molds of water-cooled copper, graphite, rammed , magnesia, cast iron, and sand. While castings are considerably weaker than wrought metal at room temperature, they are slightly stronger at MSBR.operating temperatures. It is easier to control the chemical anal- ysis by vacuum melting and the metal has better mechanical properties and fabricability. Vacuum melting has become & common commercial'practice during the last few years, and this melting practice will probably be used in.preference to air melting in the future for nuclear grade materials In making the many different forms and sizes of material required during the developnent and construction of the MSRE, Hastelloy N was fab- ricated using the normal technigues for nickel-base alloys. The initial fabrication or'breaking down of the ingots was accomplished by forging. or extrusion. Temperatures varied from 1825 to 2250°F. - Secondary fabri- _cation was accomplished hot and—coid; Acceptable techniques included rolling, swaging, tube reducing, and drawing. The alloy work_hardensr w 37 | when‘fabricated cold, but reductions in area of up to'50% are possible 7between anneals.. Temperatures of 2100 to 2200°F are used for hot rolling “with reductions of about -10% per pass. L ~_ The standard finished annealing temperature for this alloy is 2150°F ~All material for the MSRE was annealed after forming or working. After lfinal fabrication, all material was. stress relieved at 1600°F. For any ‘heat treatment the'cooling;rate to 600°F was limited to 400°F per hour per inch of thickness. The slow cooling'helped to impart dimensional stability and improved the ductility in the 1400 to 1800° F range. A necessary adgunct of fabrication is an inspection to demonstrate the acceptance of the product - Products fabricated from Hastelloy N may be inspected With the same nondestructive testing procedures used for other high-nickel alloys or austenitic stainless steels Tubing and pipe received for the MSRE'were inspected using 1iquid penetrants, eddy 'currents, and ultrasonic techniques. Weldingland Brazing of Hastelloy N Extensive experienceuith-flastelloy N has shown-that it exhibits relatively good weldability. mhe MSREtcontains literaily'hundreds of satisfactory tungsten-arc welds in varied section sizes. The welds were - made using procedures developed at ORNL for nuclear quality applications The subsequent inSpection of these welds showed that they met the very tight 'ORNL requirements ' ' | Hastelloy N has also been shown to be readily brazeable. Good :_wetting and flow have been demonstrated using commerCially available brazing alloys and conventional brazing techniques.; o During the course of the investigation of the weldability of : Hastelloy N} welding prdblems were encountered _ Hot cracking of the type. ’occasionally encountered in welding high nickel alloys was sometimes iobserved in heavily restrained JOints both in experimental heats and in ';'later commercial heats. In some of these unacceptable heats, the cause - 7of cracking was established as high boron content, however,'certain lower boron heats were stil1l crack-sensitive for unknown reasons Although the exact cause of the trouble;was never,established,.indications‘were that 38 it was associated with the proprietary melting practice used by the manu- factnrer;. The impurity specifications were relaxed slightly, the melting practice was revised, and weldable heats were produced. ' : Although these changes upgraded the quality of the commercial pro- ducts they did not assure the complete elimination of the propensity toward cracking. A crack test was conceived and used &s the basis for segregating unsatisfactory heats.”” Since the effects'of processing vari- ables and impurity levels on welding were not determined' the‘weldabilr' ity of a particular heat had to be determined by a test weld. Conse-— quently, metallurgical studies aimed at understanding the basic reasons for cracking and-eventually reaching'a more permanent solution were ini- tiated. Cracking was shown to be essociated-with'the segregetionof allofing elements in the heat-affected zone of Hastelloy N weldments. 23 Figure 22 shows the microstructural changes that can occur'ée a result of welding; microprobe analyses of these samples showed that the brittle eutectic-type structure had a markedly different composition from that of the matrix. ' - - _ To date, the gas tungsten-arc welding process is the only technique that haS'been used for the construction of reactors'of this naterial | The MSRE experience has proven the applicability of this process " The weld filler metal used for Joining Hsstelloy N has been of the same basic composition as the wrought product. The room-temperature mechanical properties of Hestelloy N'weldments. have been extensively investigated. In the course of qualifying welding procedures for this alloy in accordance with the ASME Boiler and Pressure Vessel Code, numerous bend and tenS1le tests were conducted The results readily satisfied the Code requirements However, as was expected the ductilities of the welds were not as good as those obtained for the base : metal. Elevated-temperature studies of welds24,25 showed that_theychadr 23R. G. Gilliland, “Microstructures of INOR-8 Weldments, Metalsk | and Ceramics Div. Ann. Progr. Rept. June 30, 1965, ORNL-3870, pp . 24648, 2R, @ Gilliland end J. T. Venard, Welding J. (N Y.) 45(3), 1035-105 (March 1966). - 2%sR Program.Semiann. Progr Rept Aug 31, 1965, ORNL-3872, PP- 94—101 [ 4l 39 - 1N00X - V00D INCHEY iro Tew ] i 000X QUUSD INLHED 1 [ PP g . Fig. 22. Mlcrostructural Changes that Occurred as a’ Result of ~ Welding. (a) Structure of unaffected base metal showing stringer-type - phase; (v) eutectic-ty'pe structure in heat-affected zone. = Electrolytically etched in HaPO,. ' ' 40 lower duotiiity and creep-rupture_strength than the base metal. However, stress relieving at 1600°F improved the properties to where they were comparable with those of the base metal. | ~ The room- and elevated-temperature shear strengths of Hastelloy N joints brazed with Au-18% Ni filler metal nave also been determined.26 Joints tested at room temperature hed-antaverage shear strength of 73, 000 psi, while those tested at 1300°F possessed an average strength of 18,000 psi. Diffusion and mlcrohardness studies of aged brazed joints 'showed that they possessed excellent microstructural stablllty. Joining for Reector Component Fabrication Pressure Vessel and Plping As mentioned prev1ously, only‘weldable heats of Hsstelloy N were used for the MSRE application. The tungsten-arc welding,process, with | filler metal additions, was used throughbut,randltheifabrication was carried out without‘undoeodifficultj A weidifig was ‘done in accor- dance with ORNL-approved ‘specifications by qualifled operators The flnished product readlly met the strict ORNL requirements for nuclear systems. Heat Exchangers: | | 7 The production of reliable‘tsbe-to-header joints for.heat exchange applications_in molten-selt systems-is msndatory.' One means of assuring this reli&biiity, welding and. bhck-brazing; was successfully developed and tested exten51ve1y under severe operating eondltlons for the Aircraft Nuclear Propulsion Project. 27 The technique was further improved and adapted for use in the fabrication of the primary heat 26E, C. Hise, F. W. Cooke, and R. G. Donnelly, "Remote Fabrication “of Brazed Structural Joints in Radioactive Piping," Paper 63-WA-53 of ‘the Winter Annual Meeting, Philadelphia, Pa., November 17—22 1963 of the American Society of Mechanical Englneers 27G. M. Slaughter and P. Patriarca, Welding and Brazing of High- Temperature Radiators and Heat Exchangers, ORNL-TM-147 (February 1962). ¥ X 41 exchanger for the. MSRE 28 F:Lgure 23 shows the comple'bed tube bundle in its handling flxtu_re. Figure 24 is a sketch of the omnt 'I‘o'date, several thousand Jo:p.nts have been fabricated by the welded and back-brazed technlque without a 51ngle reported failure. 28R. @. Donnelly and G. M‘ Slaughter, We.ldi.ng J. (N'.'Y..' ) 43(2), 118-24 (Pebruary 1964). = . T = Fig. 23. Completed Tube Bundle in Fixture. 42 ORNL-LR-DWG 65682R3 —TUBE TREPAN GROOVE WITH / BRAZING ALLOY RING BRAZE SIDE _THREE ‘ /FEEDER HO 7 WELD SIDE ' P S \TREPAN | (a) BEFORE WELD!NG AND BRAZING i N Y N ]4%&%%&2%%%&%4»%%&&%2»%& R\ 27 NS : B 7SS S A ST TSI ST LS LTS TIS SLS S LSS LIS, N [ e | ,.:Z::_'A.\. SR - ‘ L L AT PTII RIS SIS TS S IS S S S S LS LSS SIS, ‘ g@mnmm&&%%%fi%%m&hfi%n%«- (&) AFTER WELDING AND BRAZING Fig. -24 Sketch Showing Joint Design Used for Welding and Back- Brazing MSRE Heat Exchanger ‘ 'Dissimiiar Metal Joints The incorporation of a variety of camponents in a reactor system freqfiently requires the use of dissimilar metal weidS"'Wélding proce- dures have been developed for joining Hastelloy N to several different materlals, 1nclud1ng Inconel 600, austenitic stainless steels, ‘and nickel. r § 43 Remote Joining ‘Molten- salt reactors have: the characteristic that the fuel circulates through much of the external piping system \ During operatlon, residual activity is 1mparted to the walls, and during drainlng some fuel always remains either in the crevices or by sticking to the wall. The resulting radioact1v1ty requires that remote maintenance procedures be used, and they include remote Joining Remote Welding While'many'remote'joining methods'have been-developed for the nuclear: industry,they'were_alwayS”for'special-applicatiOns such as replacing the seal weld on a vessel, —or'making-a closure weld for'a'capsule or a fuel element, %° Commerc1al equlpment is not generally available for such aIgflications.. The nearest approaches to developed remote maintenance systems for a large reactor were one designed and partially developed by Westinghouse Electric Corporation, Atomic Power Departmentjrfor use with the proposed Pennsylvania Advanced Reactor,30 and one at Atomics International for use with'aflarge sodium graphite reactor.3! However, these projects were terminated_before'the developed techniques could be demonstrated. ':Includediwere-teChniqnes for plugging heat exchanger tubes and for butt welding of pressure Piping. | The practicality of remote operations has been- improved by the ' recent rapid advances in the automatic welding field aimed at the nuclear and aerospace 1ndustr1es, an excellent example of an automated machlne is the one developed for welding the HFIR fuel elements. The only function “of the operator is to pos1tion the torch and o press the start button; 'both of these operations could also be done automatically 1f de51rable. 29G M. Slaughter, Pp. 1?3—86 4in Welding and Brazing Techniques for "Nuclear Reactor Components, Rowman and Littlefield, New York, 1964. 30g, H. Siedler et al., Pennsylvania Advanced Reactor —-Reference feDe31gn Two, Layout and Maintenance, Part I, WCAP-llO4 Vol. 4, (March 1959). .31, Newcomb, Calandria Remote Maintenance Tool- Development NAAPSR-11202 (Apr 15, 1966}. b Brazing and Mechanical Joints Studies on remote Joining,hsve also included brszing and'meohsnical joining techniques. Standard ring joint flanges were used at high pres- sures but lower temperatures in the Homogeneous Reactor Expériment.; | Freeze flanges are used to join the major pieces of equipment:in the' MSRE. Similar developments have been made for other reactors and for fuel reprocessing plants. in all cases the jolnts were developed for. use in smaller pipes than are being proposed for the MSBR. The principal deterrent to the use of mechanical joints is the dlfficulty in repeatedly | assuring a seal, particularly in high-temperature cyclic service. | - Equipment was developed and tested for remotely cutting and bra21ng the drain line and the salt piping in the drain tank cell of the.MSRE. The piping is standard 1 1/2-in. sched-40 pipe. These joints were . cafefuiiy inspected _snd fiere determined to be of high qus.lity. ' Remote Inspection A.necessary accompaniment of remote welding must be a remote inspec-__‘ tion. Less development has been done on inspection than on weldlng Sufficient work has been done at ORNL to show that such operations are | feasible and the required techniques should be possible with a_limited B development program. | | Ultrasonic techniques along with remote positloning equipment were developed and used for the remote measurement of the wall thickness of the Homogeneous Reactor Test core vessel. 3? The precision'of messurement vas approximately 1%. ~ Ultrasonic and radiographlc technlques were developed but not trled | for the remote inspection of the EGCR through-tube weldments.Bsflir 32R, W. McClung and K. V. Cook, Develomment of Ultrasonic Techniques for the Remote Measurement of the HRT Core Vessel Wall Thickness, ORNL-TM-lOB (Mar. 15, 1962). 33R. W. McClung and K. V. Cook, Feasibillty Studles for the Non- destructlve Testlng of the EECR Through—Thbe Weldment, ORNL TM,46 (NOV. 14, 1961). ] | S | § oy [ 1] 45 Another class of remote inspection devices has been developed for use in hot cells. Radiographic techniques have been shown to be appli- | ~cable to the evaluation of radioactive materials,in high radiation ) backgrounds. This includes both television®4 and film techniques.’’ STATUS OF. Dm_aomm oF VGRAPHITE In the proposed MSBR, graphite is used in the core as & pipe mate- ‘rial for separating the fuel and blanket salts and also for moderator blocks around the tubes The former is the more critical use. The graphite in the tubes must have low permeability to salt and must be free from flaws that would permit appreC1able seepage of- salt between the fuel and blanket streams. A,value for the permeability to gases has not been specified yet but a very low one is de51rable to minimize the rate of diffusion of xenon- into thefgraphite and thereby minimize the loss of neutrons to +>%Xe. - Although high strength is desirable, 1t need not be ‘higher than has already been obtained readily Techniques sre required - for joining graphite to graphite and graphite to metals. Finally the graphite in the MSEBR will be irradiated to a dose as high as 2 % 1032 neutrons/cm (E >'0 18 Mev) per year and the tubes must perform - satisfactorily at this irradiation level for at least three and poss1b1y: five years to have g practical power reactor Graphites have been used for many years as. the moderator in a ,.variety of power and pdutonium production reactors where the require- ‘ments are less. severe and the radiation levels are much lower - The type ?'CGB graphite used - in the. MSRE approaches the requirements of the MSER - - but the MSRE 1s not. a two—fluid $ystem and is less. demanding Loosely | :assembled bars were used small cracks were not serious faults, and the "._radiation levels were almost two orders of magnitude lower ‘than those 1n‘ "the MSER. Some of the newer isotropic graphites promise greater stabilityr _1-under 1rradiation than does the anisotropicntype CGB. Improvement of o 34J Wallen and R .W.vMeClung, An Ultrasonic, High Resolution, X—Ray Imaging System, 0RNL-2671 (May 1959) 35R. W. McClung, Mater Evaluation 23 (1), 4145 (January 1965). thdse’materials to satisfy other requirements imposed by the usecof'molten- o salt fuels could result in a better graphite for use.in the breeder reac- tors. Extrapolat1on of existing data indicates that the life will probably be adequate Grade CGB Craphite _ Grade CGB graphite was designed for use in molten-salt reactors and was first made in comnercial quantities for the MSRE. dIfi is'basicelly x5 a petroleum needle coke that is bonded withdceei—tar pitch, extruded to rough shape, and heated to 5072°F. Improved dimehsional”stabilifyfunder = irradistion is ensured by not using amorphous carbon ma%erialsrsuch-ase carbon black in the base stock.- High density and low permeation are achieved fhroughfmultiple'impregnatiens'and'heat treatments. All compo- nents are fired to 5072°F or higher. The product is essentially a well- graphitized material and.is'highly"anisotropic.w Its properties are tabulated in Table 9. LT e Tl Grade CGB graphite represents significant progress in the develop- “, ment of a low-permeability radiation-resistant graphite. Ekperiments ' have demonstrated that the multiple 1mpregnations required to dbtain the low permesbility and small open-pore size do not appreciably alter the | dimensional stability compared to conventional needle-coke graphites. In manufacturing the material for the MSRE, - experimental equipment and processes were used on a commercial scale for -the first time by the - Carbon Products Division of Union Carbide Corporation to‘pfoduCe'bars 2.50 in. square and 72 in. long which were machined to the shapes - required for the MSRE. The bars were the 1argest that could be produced by the then-current technology to meet the requirements of. low permeabil—- ity to salt and gases end resistance to't&diatibn'effects. Assembly of the MSRE core from bars of the graphite is shown in Fig. 25 - Structure The MSRE graphite has an accessible void sPace of only 4. O% of: 1ts bulk volume. The pore entrance diameters to the accessible voids are L QO b g g~ . — o 47 T&ble 9. PropertiéS'of'MSRECo:e'Graphite;Grade.CGB Phjsiéalrpropértiés | - : : 8 /.m3 : : _ Bulk density, g/em® R el Range - | o L — L 1.82-1.89 - Average : - 1.86 .-Porosity,b'% | | o - c Accessible Inaccessible Total . o | . Thermal conductivity, ‘Btu £EL hr'17(°F)’1.- ‘With grain . : _ - At 9Q°F. 112 At 2200°F - L Normal to grain [ | S At 90°F . SRR 63 At 1200°F L o 34 | Temperature coefficlent of expan31on,c (‘M- | With grain = ; o At 68°F 0.27 x 10~6 b3 sJNJCD Normal to grain o S . | , | At 68°F oL o 1.6 x 10-6 Specific heat;d ‘Btu 11 (oF)-l T D e At O°F R o 0,14 - At 200°F '.,' _171;_ -;‘7 o 0,22 At 600°F L 70,33 CAt 1000°F - o - 0.39 : At 1200°F . e | - - 0.42 Matrix coefficient of permeability to helium 3 x1074 ~ at 70°F, cm /sec :1' S e . "Salt absorption at 150 psig, vol % o 5 ;iff 0,20 Mbchanlcal strength at 68°F ' e T | - Tensile strength, psi With grain o ,;a,,-;;-'; L el e "Range xf{ TS o 15006200 Average Q*irinjgj;f‘ o S 11900 NOrmal togrein © . R Range 43'f¢a”5iTiifff' S '1 :,1 L ;fl;mf 5{{1100fi4500 -~ Average o 1_;:_¥'_‘ afj‘_‘ S 7-14OQ Flexural strength, psi LT Ll With grain ,'__-,=;*;]gr* B S -~ Range . o . -~ 3000-5000 Average o o 0 4300 S48 : o ‘Table 9. (continued) - | ¥ Mechanical strengthb at 68°F Flexural strength, psi. - Normal to grain | | | IR Range - | - .2200-3650 - Average o 7 . 3400 Modulus of elasticity, psi T With grain | . -+ 3,2 X 108 Normal to grain - ' : ' 1.0 x 106 Compressive stren'gth,dpsi' ConT T 8600 Chemical purityf - | | L Ash, wt % - - | | - 0,0041 Boron, wt % , o 0.00009 . Venadium, wt % . - - - 0.0005 - Sulfur, wt % | | | - 0.0013 Oxygen, cm? of CO per 100 cm of graphite 9.0 Irradiation data® e IR [ Exposure: 1.0 x 102° neutrons/cm , (E>2.9 Mev) . Shrinkage, % (350-475°C)] | | - | r Ao . | | 0,10 | aMeasurements made by ORNL. ’ blesed on measurements made by the Carbon Products Division and ORNL.. “Measurements made by the Carbon Products Division. | d'Representsti\are data from the Carbon Products Division. 7 ®Based on measurements made by the Carbon Products Division on pilot-production MSRE graphite. _ - f1'.}zatta. from the Carbon Products Division small, over 95% of the entrances are < 0.4 u in diameter.3_6 'I.’he molten | salts do not wet the graphite, and calculations indicate. that pressures in excess of 300 psig would be required to force salt into these acces- sible voids. The microstructures of a conventional nuclear graphite and the graphite in the MSRE are compared in Fig 26. ' 24, H. Cook, MSRP Semiann. Progr. Rept. July 31, 1964 ORNL-3708 . — 49 Photo 70797 ) " $t : i L - Fig. 25. Anisotropic; I | | -~ for MSRE Core. o ngh-density and small-pore structures are des;rable characteristics ; - - of unclad graphite for molten-salt reactors, HOwever, for the MSRE bars, -this particular structure was perfected to such a degree that in the ’final fabrication operations_gases could not escape rapldly enough through the natural pores. in the’ graphite.' Consequently, the matrix'of thefgrephitéferacked'as*thefgaseS'fOrCed their way out. However, 0.50-1in. -thick sections (heevieet wall for MSBR fueletubeS)~have been made successfully without cracks. 50 “PHOTC 67670 asoT ceB 1.68 g/cm3 * "f 186gmm3 247 % ACCESSIBLE VOID VOL. 4.0 % ACCESSIBLE VOID VOL. 3.7 % INACCESSIBLE VOID VOL. 14.0 % INACCESSIBLE VOID VOL. 25.4% TOTAL VOID VOL. = _"1 e.o' o TO_TAL VOID VOL. Flg 26 The Mlcrostructures of a Conventional Nuclear Graphlte, Grade AGOT, and of the MSRE Graphlte, Grade CGB Mechanical tests indicate that"the-—'prbc'essing-cracks, in the graphite resist propagation.37 Since it is probable that cfacks'ih'gréfihite will £fi11 with salt, tests were made to determine if repeated melting and freezing of the salt would propagate the cracks. _Imfiregfiatédépecimens were cycled from 390 to 1830°F at various rates of temperature rise. Specimens and cracks were not detectably changed by the thermal qycles,38’39 The maximum temperature in the tests considerably exceeded the 1350°F expected in the MSRE or the MSEHR. 37c. R. Kennedy, GCRP Semiann. Progr Rept. Mar. 31, 1963, ORNL-3445, . 221, 38y, H. Cook, MSRP Semiann. Progr. . R@t July 31, 1963, ORNL-3529 p. 76. 39y, H. Cook, MSRP Semiann. Progr. Rept July 31, 1964 ORNL-B'?OS p. 382. L1 » " 51 Permeability 'Some'of the more”stringent specifications on graphite for MSER will _ be on permeability. 4O Values will be specified for perrieability to both: molten salts and to helium The screening test for salt permeability is to expose evacuated specimens for 100 hr to molten salt at 1300°F under 8 pressure of 150 ps1g.4 The MSRE de51gn spec1f1cations reqnired that f-less than O. 5% of the bulk’ volume of the graphite should be filled with salt and that the salt should not penetrate the graphite matrix Speci~ ‘mens of grade CGB had an average of less than 0. 2% of their bulk volume permeated by salt in these screening tests, and . the salt distribution -~ . was limited to shallow penetrations at the surfaces and along cracks " which intersected ‘the exterior surfaces. The salt that entered the cracks was confined to the cracks and a8 shown in Fig 27, did not .. permeate the matrix. The permeation characteristics of the graphite as determined by test should be representative since irradiation does not appear to change the wetting characteristics of the salt and the screening pressures are three times the maximum design preBSure for the MSRE core. The limit on gas permeability ‘has played a major. role in establishing the fabrication sequence for the graphite Low permeability to gases is desirable to prevent gaseous fission products from entering the graphite. We would like a graphite for the 'MSER with a helium permeability of 10"7 2/sec. While it has been possible to obtain graphites meeting the o salt-permeability requirements, 1t has not been possible to reduce the - gas . permeability to the lO 7 level | Crack-free CGB graphite in the MSRE f;pexhibited average helium permeability values of 3 X 10'4 cmz/sec As "jpjpreviously discussed this graphite was fabricated using special proce—_. f;fdures aimed at low porOSity.. The graphite tubes for MSER'will have much _"ihthinner cross sections and should be less permeable. ‘At present, it pf'fiwould appear. that values of O s cmz/sec would ‘be about 2ll that cen be ,fgdproduced by present technology Experimental and other spec1a1 graphite | .'“_.ypipes have been produced w1th permeabilities less than 10'5 cmz/sec 40Hielium is a convenient medium to use for quality control to establish permeabllity 11mits to gaseous fission products. 520 | B L ' Fig. 27,1_Radibgréphs of Thin'Seétiofis Cut from an Unimpregnated Specimen and Three Salt-Impregnated Specimens. (a) Control (no salt = - . present); (b), (c), and (d) salt-impregnated specimens; white phase is . . o salt. _ , 7 - o | ) = o - i} " ” 53 (refs. 41, 42). The 1atter~contained carbon black in the base stock which would significantly decrease itsdimensional stahility-under radiation. The former had amorphous‘carhon impregnantS‘ but'this'appears to cause only slight decrease in its. dimens1onal stability under irradiation. %3 While & homogeneous graphite is desirable, one w1th a surface seal must also be considered Such 8 seal may be & penetrating, integral skin of graphite or a thin coating of 92Mo or Nb The metallic coatings would be applied hyrdeposition from the metal chlorideggas at elevated temperatures. The surface COating_would_be quite thin (approx 0.0001 in.) but would be applied under.conditions where the metal sould be linked into the pore structure so that‘good.adherencefWOuid be. ensured, Mechanical Properties The tensile and flexural strength.measurements on grade CGB graphite for the MSRE averaged71900 and 4300,p31, reSpectively,_despite the pres- ence of cracks in the graphitert There'fiere no values less than the minima of 1500 and 3000 psi spec1fied for ‘the ten511e and flexural strengths, respectively. These strength ‘values were spe01f1ed primarily to assure a good quality of . graphite.‘ The stresses expected in the 'reactor are very low in the absence of irradiation However, under fast- neutron 1rrad1ation, the differences in shrlnkage rates across the 1/2 -in. wall of a pipe caused by flux gradients can produce high stresses. The tensile strength of ‘the MSRE graphite is. shown in Flg 28 The band with the single cross-hatching indicates the stress strain behav1or : of the material However, some of the material contained cracks and | lfallure occurred at lower stresses and strains as indicated by the solid __;points ‘The crack- free material represented by the solid p01nts, failed 'd:at higher stresses Spec1mens that demonstrated definite effects of | cracks (the ‘black- points) had a minimum strength of 1510 pe1 and an average 4*K Worth Techniques and Procedures for Evaluating Low-Permeability _Craphite Properties for Reactor Application, GA~3559 . 7 (March 1, 1963). 421, W. Grahsm and M.S.T. Price, "Special Graphite for the Dragon -Reactor Core," Atompraxis 11 549-54 (September—October 1965). 43G. B. Engle et al 3 Irradlation of Graphite Impregnated with’ Fu.rfuryl Alcohol, GA-6670, p. 2 @ct 5, 1965). _54 ORNL-DWG $3-J354R © NORMAL FRACTURE ® STAR-STEP FRACTURE WITH A RISER 5 ¥4 in. 0 0.0% ) 45 0.20 028 0% ' STRAIN (%) S Fig. 28. Tensile Properties of Grade CGB Graphite Parallel with - Grain (Axis of Extrusion); Bar 45, 2 X 2 in. strength of 2940 psi. Specimens that were not appreciably affected by cracks (the open circles) yielded an average strength of 5440 psi and g modulus of elasticity of 3.2 X 106 psi, corresponding to an unusually | strong graphite. These results indicate the variation of properties.within' a single bar (No 45); however, there was algo some yariation between various bars. The double cross-hatched portion indicates the overall range of prOperties obtained by sampling various bars of the MSRE graphite Isotropio cfapme | There has been recent progress in the development of isotropic ' grades of graphite that offer greater dimensional stability under 1rra~g | diation and potentially can be used as a ‘base for the 1mpregnations to Obtain the necessary low permeability. Development of this material has occurred in the last few years and.much of the data are proprietary and have not been released for publication. However, properties of some available grades are listed in Tatleilo : . Table 10, ‘D1men31onal 1nstabi11ty at 700°C 2 x1072 . 1/2 to 1/3 that of AGOT | Properties of Advanced Reactor‘GradeS'of G-raphitea o v _ - Gradet - Property : — ' — | o - H-207-85 H-315-A ‘H-319 . -Density, g/cm3 o 1.80 - 1.85 1,80 Bend strength, psi j\.761“ - >6000 4500 ~>4000 ;Mbdulus of elastlclty, psi S o 1.351. 65 X 106 R ;Thermal conductiv1ty,.\ | ‘i“fiffi; 22 22 22 Btu ft'l hr~t °pt (1832°F) a o | - | / n }Thermal expansion,(l L - 5.2-5.7 4.8-5.8 3.7%.4 \ 108/°¢ (22-700°C) | o | Electrical resist1v1ty, 8.9 9.4 9.9 ~ ohm-cm X lO E 'Isotropy factor, CTE”/CTEL 1.11 1.20 l;l8 Permesbility to helium, c /sec‘_ g x 1072 Adapted from G. B. Engle, The Effeet of Fast Neutron Irradiatlon from 495°C to 1035°C on Reactor Grephite, GA-6888 (February 11, 1966) . All propertles measured at room temperature unless stated otherW1se. Based on very prelimlnary results on H-207-85 and type B isotropic data to 2 X 1022 neutrons/cm 56 The isotropic graphite is made in various shapes with conventional carbonraad'graphite molding and extruding equipment. Bulk densities can range from < 1,40 to 1.90 g/cm3. The strengths of various grades tend to exceed those of grades of conventional, anisotropic graphite. Past work has not been directed toward producing isotropic graphite with the low permeabilitles required for molten-salt breeder reactors. Pipes of grade H-315-A (Table 10) have been made that are approximately 4.5 in. in diameter with O. 50-in.-wall thickness which would correspond to the large MSBR pipes. These pipes had an average bulk den81ty of 1.83 g/cm3 and radlally they had a permeability to helium of 2 X 10‘3 to 1 X 10”2 ecm?/sec. The lower value refers to the unmachined pipe and the higher value to pieces machined from the pipe. walls. Pore spectrum measurements indicated that approximately 50% of the pore volume had pore entrance diameters betweeh 0.5 and 3.5 p and less than'lfi% of the pore volume had pore entrance dismeters > 3.5 p. This pore spectrum ' suggests that this material, although it was not designed for it, might be suitable base stock to impregnate to preduce 1cw-permeabi1ityrgraphite. of ceurse, in fabricating graphite for MSBR use one woald want to opti- mize the base stock for impregnation which would mean that the base stock would be fabricated with pore entrances less than 1 p in diameter. | In the following program, the isotropic graphite will be considered as the reference material. However, until its development has proceeded to the point that meeting the MSBR requirements is reasonsbly assured, it will be necessary to continue the development of the anisotropic needle-coke graphite as backup material. Most of the data obtained to date have been with needle-coke graphite; however, much of this technology is transferable to the 1sotropic graphite. Joining The MSBR is unique among reactors in that it uses graphite as an engineering material and the present conceptual design calls for graphite- to-graphite and graphite-to-Hastelloy N ,joints.44 These JOints,are oni . 44P, R. Kasten et al., Summary of Molten-Salt Breeder Reactor De51gn' Studies, ORNL-TM-1467 (March 24, 1966). ambesine a1 2o gt o g ey £k . 461-69 (1962). g pp 101-04. 27 ithe core pieces but in a region where the neutron flux dose is reduced \con81derably With all the tons of graphite that have been used in | ',nuclear reactors, most of the work on graphite—to-metal Joining seems to have been done in small programs at ORNL" 46 and Atomics International Graphite-to-Hastelloy N JOints o Only 8 few brazing alloys are suitable for brazing graphite for molten-salt.serv1ce,.since, in addition to,wetting abillty, the braze material mustfbecorrOSion:resistantiandstable under irradiation. To promote.bonding,;somechemicaiflreactiontwith the graphite is desirable; therefore, most proposed:graphite brazes contain some "active" metal. However, titaniumfand'Zirconium ;;the‘tw0'u8ually_pr0p65ed — may not be suitable as far as corrosion resistance to molten fluoride salts is con- . cerned. Similar‘statements can'he made about alloys containing large quantities of chromium Thus,_brazing development in support of molten- salt concepts has ‘been cOncerned'with developing an alloy that not only will braze graphite but also,will ‘be corrosion resistant, melt_high ~ enough to be strong in nse‘butdlow enough to be brazed with conventional vacuum or inert-stmosphere equipment, andjbe stablefunder irradiation. 'The‘low.cOefficientfof;thermal'expansion_and lowrtenSile strength of the graphite may make it difficult to join it directly to-the' ‘Hastelloy N or otherlstructural materials. - The problem is further com- . aplicated by -the- respective shrinkage and expan81on expected in the graphite as 1t is exposed to the fast neutrons ‘ Two of several approaches are receiving the greatest study at this 'timé; One is to braze ‘the Hastelloy N to the graphite in such a way ‘_that the Hastelloy N applies ‘a compressive 1oading to the graphlte as the Joint is cooled from the brazing temperature This is ‘desirable 'dfld_51nce graphite is stronger and more ductile in compression than in _f,_tens1on.; The other approach is to 301n the graphite to materials that "have coefficients of thermal expans1on approaching or equal to that of 43R, G. Donnelly and G M Slaughter, Welding J. (N Y, ) 41(5), 46p. B, Briggs, MSRP Semiann Progr. Rept Feb 28, 1966, ORNL-3936, 58 the graphite. These transition materials_would, in\turn,”be joined to the Hastelloy N by brazing. Therefore, refractory metals, such as molybdefium and tungsten, which have thermal expansion coefficients more closely matched to graphite as well as much higher strengths, have been selected as prcmising transition materials. The joint utilizing the refractory metal transition piece ‘has .received considerable attention. The ductile and corrosion-resistant alloy, 82%;Au—18% Ni, was chosen:aé.a-starting pcint for mclfen-salt 1 applications. To this base meterial has been added the corrosion- ‘c,- _ resistant, strong carbide formers, molybdenum and tantalum. As a result, ‘alloys in both the Au-Ni-Mo and Au-Ni-Ta systems were developed4’ which, fulfilléd;the above requirements wifh the'possibleexcepticn of hucléar stability in a high neutron fluxi(gold_will_transmute to mercury:in high_ , flux fields). The most promising alloy contains 35 Au-35 Ni-30 Mo . (wt %) and has been designated as ANM-16. Graphite-to-graphite and | graphite-to-molybdenum joints have been made with ANM-16 by induction braiing in an inert atmosphére and vacuum brazing in.a muffle furnace. A brazing temperature of 2282 to 2372°F was used. Several componeuts have been brazed with this alloy, the largest of which have been o 1.50-in. OD X 0.31-in.-wall graphite pipes with mclybdenum.end caps and sleeves. dJoints of this type as well as T-joints have -been examined metallographically and found to be sound in all cases. If brazing conditions are controlled closely, the joints appear sound in all respects. However, if the assembly is maintained at brazing: temperature longer than necessary or if too high a brazingftemperature is used, the alloy in the fillet spreads onto the surface of ‘the graphite in a very thin layer. Upon cooling from the brazing temperature, this .thifi leyer will crack. This behavior is aggravated when the surface of .. the graphlte is smooth and free of porosity, such as with the desired _ | MSER graphite. Joints made with high-density CEY graphite pipe exhibited this behavior, whereas it had not beeh noticed when brazing the.vefy | pofous grades sfich as AGOT. 'MetallOgraphic examination,'hcqcver,wchqqs_”__ that the cracks do not extend into the body of the £illet, e et s ikt i [ chromium will not cause a corr051on problem 59 A short section-of'graphite pipe with a brazedggraphite-to-metal ‘end closure successfully"contained molten saltsfiforVSOO hr at 1292°F and a pressure of 150 psig. The graphite pipe was machined to 1, 25~ 1n OD X 0. 75-1n. ID frOm a bar of MSRE graphite ~ Molybdenum end ~ caps were brazed to the ends of the pipe using braze metal ANM-16. No" leak occurred during the 500 hr test period and posttest radiographlc rexamlnation 1nd1cated that the salt was contained as planned Metallo- graphic examination revealed no penetratlon of salt into the graphite and no detectable corrosion of the braze material Recently, an alloy w1thout gold has been developed to overcome the poss1ble problems of transmutation This alloy has & composition of ’35% Ni—60% Efirfi% Cr and has exhibited brazing characteristics very much ~ like the 35% Au-35% Ni—BO% Mo alloy The chromium content was kept low in the hope that acceptable corros1on resistance would be obtalned. The brazing alloy wet both materials and appeared to form a joint as good as that obtained with ANM-16. Joints of graphite to molybdenum have been brazed w1th this alloy and survived a 1300°F corrosion test (LiF-Bng-ZrF4-TthfUF4) of 10,000 hr with no attack.,.There wae_a'thin palladium-rich layer_on the surface of'the_brazing'alloy;IEA.2Q,QOO-hr corrosion test has accumulated 10,700 hr to date. Chromium, the cerbide former, may be tied up in the form of a corrosion-resistant carblde, Cr3C,, to such an extent that the - 47 To date, no mechanical property tests have been run on brazed ) graphlte 301nts at ORNL but graphite—to—molybdenum specimens ‘brazed W1th the 35% Ni-60% Pd—fi% Cr alloy ‘have successfully withstood ten thermal _;'cycles fram 1300 F to ambient., Metallographic examination revealed no _cracks after this test o 47W, H. Cook, MEtallurgy de Semlann Progr. Rept April 10 1956, | _-1‘GRNL-2080 pp. 44 and 46. 60 | - R Graphite-to-Graphite'Joints | Techniques are commercially available for Joining graphite fs'ff' itself.48 Most techniques are based on the use of furfural alcohol and graphite powder The mechanism of sealing is by polymerization of the - aleohol at temperatures to 1600°F The Joint ‘must be held under pressure during the curing to prevent gross porosity The por051ty arises frOm 5 the gas that must escape during curing. It may be minimized by close Joint fitups and slow heating rates The polymerization would be followed by a graphitizing treatment at about 4892°F waever, it may be more desirable to use & metallic brazing alloy to obtain a Joint with lower permeability and higher strength Small graphite-to-graphite Joints have been made at ORNL using the Au-Ni-Mb and Au-Ni-Ta brazing alloys, however, | such joints were most effective when graphite that was much more porous than the CGB grade was used Several other alloys are being investigated that wet the graphite and are resistant to corrosion by fused salts. Compatibility of Graphite with Molten Salts Since the MSBR uses grephite in direct contact with the molten-salt - fuel and the breeder blanket, good compatibility between graphite and Hastelloy N in these molten fluorides is needed. Tests performed have indicated that excellent compatibility exists. The tendency for Hastelloy N to be carburizediwas investigated'in static pots containing LiF-BeF,-UF; (67-32-1 mole %) and a graphite at 1300°F for times ranging from 2000 to 12 000 hr. No carburization was metallographically detected on specimens from.any of the above tests.49 Tensile properties were' " determined on specimens included in each test. A comparison of the tensile strengths and elongation values of the tested specimens with those of the control specimens indicated that no slgnificant change | occurred in the test specimens. _ 48R, E. nghtingale, Nuclear Graphite, Pp. 62-65, Academic Press, New York, 1962. “%H. G. MacPherson, MSRP Quar. Progr. Rept. July 31, 1960, ORNL-3014 o/ p. 70. b ™ '5gthe graphite under such conditions.g,_nf" el 'For the'purpOSeioftevaluating thelconpatibility of graphite and'- convection loop was operated for 8850 hr The 1oop operated at a max1- ‘ir;mum temperature of" 1300 F and circulated a fluoride mixture of , ;-:LiF-Bng-UF4. On completion of the test, components from the loop were 7 nchecked metallographically and chemically, and specimens were checked : for dimensional changes and'weight changes. The tests indicated that (1) there was no corrosion or erosion of the graphite by the flowing salt; (2) there.was.veryfilittle permeation of the graphite hy the salt and the permeation'thatfloccurred]was uniform,throughout'the‘graphite ,rodspc(BJ the'various Hastellov’flrloop componentsfexposedlto"the salt were not carburized; (4) the Hastelloy‘N components:expOSed‘to'the'salt and graphite were negligibly attacked, and (5) with ‘the possible excep- tion of oxygen contamination, the salt appeared to have undergone no chemical changes as a result of exposure to the graphite test ‘specimens. In-reactor tests 50 confirmed this good compatibility - Radiation;Effects'on Graphite - B Although graphite has heen used successfully in reactors for over 20 years, it is only within the past 10 years that the effects of irra- idiation have been studied inten51vely These studles have been ‘responsi- : Vble for 51gn1ficant improvements in- both the quality of the graphite and {the development of reactor concepts that fully utilize the unique prop- -erties of graphite ' The molten-salt reactor systems are very good : *; exampdes of reactors that fully utillze the advantageous properties of graphite. These systems will however, subject. the graphite to much "Cff'higher neutron doses [for 8’ lO-year life, 2 % 1023 neutrons/cm ,_j'(>-o 18 Mev)] than have been or are being obtained in any other reactor ___l;fljsystem. These Large doses give rise to questions of growth strength | ”nzifliand dimen51ona1 stability that will determine the serv1ceable life of 5°w R. Grimes," Chemical Research and Development for MOlten-Salt Breeder Reactors, ORNL-TM-1853 (June 1967) ' - 62 The thin-wall tubular désign used in the MSER is one of the better configurations in reducing the differential growth problem to within the capabilities of the'graphité. The tubular shape also has the advantages ' of good fabricability and‘reliable nondestructuve testing'techniquéé to ensure maximum integrity initially. Extrapolation of irradiation damage data for doses to 2 X 1022 neutrons/cm?® leads to the conclusion that graphite should have an adequate life in the MSER. waever, we cannot conclude this with certainty until'we have exposed graphlte to the doses" anticipated for the MSBR. Although theére is no experimental evidence beyond _ 2 x 1022 neutrons/cm s, several factors lead to optimism. about ‘the abillty of graphite to .sustain much greater‘damage. ‘The first factor is the_ recognition that in the 1292°F temperature range_the:cqntracturalvdimen_ sional changes for the first 10?2 neutrons/cm? are due to the repair of the damage caused by cooling the graphite from its graphitization temper- ature. Doses above lOzz_fieu.trons/cni2 produce an expansion in the “c" direction and a contraction in the "a" direction. 1In effect, the Stresses produced-normal to the basal planes will be compressivé rather than tensile. This, of course, is very much preferred in that the crystal can sustain this type of loading more readily without cracking. Even so, measurements obtained from-irradiated pyrolytic graphites indicate that a deformation rate of about 1% shear strain per 1021 neutrons/cem? must be absorbed. Thus, for a 5-year life, the graph- ite tubes operating in a flux region of 6 X 1014 neutrons cm™? sec™ absorb an internal deformation of about 100% shear strain. There is evi- must dence, howeVer, that graphite can sustain this type of internal déformation without creating micrbcracks.‘ Very fine crystallite pyrocarbons have been irradiated under conditions prbdubing crystallite shearing'rates of 60% per 10?2 neutrons/cm® (ref. 51) and have demonstrated the ability to absorb 160% shear strain without observable internal cracking or loss of integrity. Thus, 1t is difficult to predict whether the anisotropic 51The dimensional changes in graphite are conventionally expressed as strain per dose. This is referred to as a strain rate although the variation is with respect to dose rather than time. H » a3 63 erystal growth Will-producemicrocracksin-the,structure; andlif.the | | microcracks are formed, whetherrthey_Willlbe more;detrimental to the structure than the initial microcraoks produced by cooling from'the graphltlzation temperature | | o As prev1ously mentloned, the relatively thln-wall tube des1gn minimizes stresses due to the dlfferential growth, ‘The magnltude of the maxlmum stress produced can be fairly well calculated with the main _uncertalnty being the flux difference across the ‘tube wall, ~ Using a conservative estimate of 2.4.% 10724 cmz/neutron as the growth rate and a 10% flux drop across the wall a differentlal growth rate of 2.4 X 10725 cma/neutron is Obtalned The restralnt is internal; there- | fore, about half of the differentlal growth is restrained to produce a strain rate. Thus, & strain rate of 1.2 x 10723 cmz/neutron is obtained for comparison to an estimated creep rate coefficient of 4 x 10727 (psi)™t The stress is s1mply and dlrectly calculated to be: - Thus, the equllibrlum stress level is very low, probably'much below that required for feilure even at hlgh radistion dose levels. ' | -Failure,rhowever, could result from inability of the graphite to -ebsorb the creep deformation even though the stress level is much less ~than the fracture stress. For 1ifetimes of 1 X 1023, 2 X 10%%, and 4 x 1023 neu.trons/cm2 (correspondlng to 5-, 10-, and 20-year llfetlmes),'” _the strain to ‘be absorbed would be 1.2, 2.4, ‘and 4. 8%, resPectively ,The cons1derat10n of a straln 11mit for failure is. realistlc, ‘however, “the strain limit for fracture has not been establlshed It has'been | "__demonstrated that graphlte can absorb strains in excess of 2% in = . 1022 neutrons/cm? under stresses in excess of 3000 psi. W1thout fracture. rfThere is also some ev1dence that -growth rates of isotroplc graphltes Cwill diminish after a dose of 1022 neutron/cm - Thus, the- graphlte o "'will not be forced to absorb the quantltles of straln calculated | 64 Although there is no direct evidence that"- graphite can, sustain 1.2% ‘strain in 5 years of very high dose and’ low stress, the- indirectfev1-' dence does indicate that a failure is improbable Nondestructive Testingfof Tubing Thererhas'not been extensive development of nondestructive tests for graphite at ORNL and other AEC installations52 53 or by industry However, past efforts at ORNL aimed at specific tests have produced confidence that several NDT methods are applicable. These programs have produced background knowledge that will aid in subsequent development programs. ' . Several. low-voltage radiographic techniques have,been developed for use on 1ow-density materials.f?4 . These innovations. have produced superior- " achievements in sensitivity and have been applied to numerous. graphite shapes. Sufficient sensitivity has been obtained to make it possible to permit examination of details as small as l u in carbon-coated fuel particles. 55,56 The eddy-current method has been shown to be applicable to graphite | inSpection. One of the earlier applications of_this‘technique_was for inspecting the graphite support sleeves for the EGCR fuel assembliesv57__ Because of the local porosity in.the,graphite, the sensitivityiinithis 52R. W. Wallouch, "Adaptation of Radiographic Principles to the - Quality Control of Graphite," Research and Development on Advanced Graphite Materials, Vol. IV, WADD Tech. Rept. 61-72 (October 1961). 53G, R. Tulley, Jr., and B. F. Disselhorst, The Pore Structure of Graphite as It Affects and Is Affected by Impregnation Processes, GA-3194, pp. 40, 47-48 (April 30, 1963). 54R. W. MbClung, "Techniques for Low-Vbltage Radiography," Nondestructive Testing 20(4), 248-53 (1962). " 3°R. W. McClung, "Studies in Contact Microradiography," Mater. Res. Std. 4(2), 66-69 (1964). DR | 56R. W. McClung, E. S. Bomar, and R.- 3. Gray, "Evaluating Coated Particles of Nuclear Fuel," Metal.frogr 86(1), 90-93 (1964). S7R. W. MCClung,'"DeveIOpment of Nondestructive Tests for the EGCR Fuel Assembly,” Nondestructive Testing 19(5), 352-58 (1961). " 65 -test_waS-limited to detection of discontinuities of approximately 10% ~of the l-in. thickness. | More recent ORNL eddy-current work has included develOpment of the phase-sensitive instrument which overcomes some of the disadvantages of , conventional equipment. This‘device:has been used to measure thicknesseS- ofigraphite58-and to detect.flaws in the unfueled graphite shells of fuel spheres58'Such as are proposed for advanced gas-cooled reactors. Other testing with graphite has included studies of infrared o methods59-to detect laminations or unbonded areas in the fueled spheres - and ultrasonic methods to detect flaws and to measure elastic properties. - _MATERIALS_DEVELOPMENT PROGRAM FOR MOLTEN-SALT BREEDER REACTORS In the prev1ous sections of this report, we have shown that we have & strong technical background in working with the: two prlmary structural terials in the MSER, graphite and Hastelloy N However, we feel that certain advancements in technology must be made with reSpect to both materials to ensure the successful Operation of the MSBR We will briefly outline the areas of concern and indicate the work necessary for | develOping suitable technology in these areas. The cost estimate for the materlals development for the MBBR is appended - | - 7 Hastelloy N Program Resistance to Irradiation Damage | f The major problem area with Hastelloy N requirlng additional develOp— ' 7 :ment before it may be: used in the MSBR 18 that of 1mproving its resistance fito neutron irradiation The present alloy is susceptible to a type of ._high—temperature radiation damage that reduces the creep-rupture life and zthe rupture ductility Solving this problem w1ll be a maJOr consideratiOn e 580, V. Dodd, "Applications of o Phase-Sensitive Eddy Current ,Instrument " Mater. Evalustion 22(6), 260-62 and 272 (1964). 59a, V. Dodd, GCR Semiann. Progr Rept ‘Sept. 30 1963 ORNL 3523 PP. 318—24 and GCR Semiann. Progr. Bept Mar. 31, 1964 ORNL-36l9 pp. 75-77 L R . { 66 .in establishing the schedule for the MSBR. The problem is complicated by the long lead time required to obtain in-reactor mechanical property data and by the fact that the solution appears to lie in a change in . eomposition This'means'that'once a modified radiation—resistant'alloy is developed, it will be necessary to determine if the changes in composi- tion have affected any other properties. Another complication is that it has been shown that the radiation damage is sensitive to fabrication practice, so to be truly representative, material used will have to be _gtaken‘from large commercial heatsr' Obviously, a compromise*is required in which small laboratory heats will be used initially for screening with the results being confirmed with material from the large heats. The fabrication practices for the large and. small heats will be kept as nearly alike 8s possible Our work has shown that titanium, zirconium, and hafnium are effec- tive additions that will reduce the radiation damage of Hastelloy N. We have not established the exact mechanism re5ponsible forrthe ifiprofiement, but feel that it is associated with the reactivity of these elements with boron and other impurities in Hastelloy N. Con During the first year, screening-type tests are being run. | The major goal is to determine which of the additives appears ‘to be- the most effective and to begin obtaining an indication of the optimum‘composition Machined specimens are being irradiated in capsules at elevated tempera- -tures; on removal they are used to determine the tensile and creep-rupture properties of the material. During this period, most testing will be on laboratory-size vacuum-melts and on small (100 lb) commercial melts. These irradiations are being conducted in the ORNL Research Reactor and ,the Engineering Test Reactor. _ . , A limited number of ccmpositions will be evaluated by being inserted into the core of the MSRE. These samples will be added as the surveil- lance specimens are removed for testing. These samples Wlll be tested to obtain postirradiation'mechanical properties and studied metallograph- ically to determine whether corrosion has occurred P - The ‘second year will be spent in determining the optimum comp031tion, heat treatment, and fabrication practice for the most promising additives. | Since these composition changes may adversely affect other prcperties, T a [ & o 67 alloys based on &t least two different alloy additions will be carried through this step. The majority of the testing Wlll still be postirradia— tion tensile and creep-rupture. The results will, however, be confirmed. by running a few'in reactorrstress-rupture tests. Both laboratory and -'“small melts commercialLy fabricatedfiwill be tested with a shift gradually being made to the latter type | _ : & - , , Iarge commercial heats (1500 lb) of a few compositions will be tested during the second and ‘third years.' This material will also be available “ for -other testing programs., These ingots will show if. scaling-up in size has any adverse effects and also if materials from different vendors are _Vcomparable When the. composition is firmly established we shall procure & few full-scale commercial heats (10,000 lb) , Specifications must be issued during the first year of testing; they will be continually upgraded by incorporating new data as they become available. The speCifications for materials for & full—51ze mockup and for the MSER will be issued during the third year. As both the mockup and reactor'material'areireceived they will be irradiation tested. Corrosion Program Molten Salt. — Since the reference design of the MSBR primary coolant circult incdrporates,the same,basic‘fuel salts and construction materials as the MSRE, an extensive-corrosion program will not be required. The large volume of data'generated during the development and operation of .rthe MSRE is directly applicable to the MSBR and has demonstrated that an acceptable system has been developed | _ Corrosion testing for the primary system will be mainly in the area 'of proof testing As compos1tional adjustments are made to the Hastelloy N or to the graphite, their effects on oorrosion will have to. be determined | :The active metals being proposed as a solution to the radiation- 'e,embrittlement problem may inorease the corr031on rate, although our pre-‘ . vious experience indicates that the effect will be small.: Alloys of _modified comp081tion will be evaluated initially in thermal-oonvection | loops. As The final composition 1s established more firmly, the corrosion ~ behavior Will be better establlshed by pump loops. 68 . ' = L | 7' ' Sufficient information is ‘available to indicate that corrosion prob-,- : .. lems are not to be expected in the blanket salt system; however, because | ‘of fewer tests, such a conclusion.is not as well substantiated as that for the primary system. It will be necessary to operate thermal-convection and pump loops with the actual pr0posed salt composition~as'proof'tests , As with the primary system, any'major changes in Hastelloy N or graphite will be checked with these salts. ' o Another area of corrosion testing is anticipated in association w1th the development of a low-melting coolant salt for the MSBR. Studies of . both fluorcborate and stannous fluoride Systemsrare currently under way;_ for this application. Neither salt'system has been subjected to;evaluaef tions in prior corrosion studies, and corrosion data will cbviously be required in the overall assessment of their properties relative to the Additional corrosion studies are being planned in support of goals of longer range than the MSER. In particular, the improvements in purity ‘ and chemical stablility of flnoride systems have brought us to a stage where it appears reasonable to consider the use of austenitic stainless 4 steels as salt-containment materials. The transition fromta nickel- to - iron-base alloy offers important economic advantages in larger sized molten-salt reactor plants, and there is a strong incentive for eXaminingi the utility of iron-base systems in the coolant-salt region alone, L Accordingly, we plan to investigate the behavior of stalnless steels in the presence of the LiF-BeF, salt system both as a function of salt- processing techniques and exposure temperatures, Steam. Although Hastelloy N looks very attractive for use in the steam circult we shall run some proof tests to demonstrate the compat- ibility of Hastelloy N and supercritical steam Engineering experiments | - are planned that involve steam and coolant salt separated by Hastelloy N Gb These tests should yvield useful heat transfer data as well as prov1de metal corrosion data 60Dunlap Scott Components and Systems Development for Molten-Salt | P Breeder Reactors, ORNL-TM-1855 @une 1967). - o/ » 69 . Witriding To mlnimize any exp1081ve hazard a nitrogen blanket is prOposed for 'use around the reactor vessel Preliminary studles have shown that nitrogen - is dissociated by the nuclear env1ronment and msy react w1th the Hastelloy N. The rate of reaction varies from ‘heat to heat of the alloy.-i ‘Specimens of Hastelloy N will be exposed to NH3, in the absence of irradistion, and the effects on mechanical properties determined. :Samples from various heats of msterial will be exposed to reveal the rate- controlling element in the alloy. If any deleterious effects are found, _additional samples will Dbe: exposed in in-reactor experiments - If nitriding is indeed a problem, the cell atmosphere w1ll ‘be changed to another gas, such as srgon. Joining,@evelopment | While a detalled examination of the joining problems for theiMSBR ‘cannot be msde until designs have progressed beyond the conceptual phase, a tentative evaluation has been- conducted to reveal the ma jor problem areas. The fabrication problems: fall into two categories: (1) the original conStruction'snd»(2)fmsintenence of‘the system. The'letter- will be by far the more: dlfficult of the two since, in most cases, it must be maintained remotely. ' o R Welding will be encountered in the reactor care, reactor vessel fab- ,-rication, heat exchangers for: the fuel and blenket systems, and in the assembly of the piping system.s Fortunstely, the large nuMber of Joining ;ftasks may be combined 1nto s few general types, at- least for the initial development Thls combining of pmoblems w1ll greatly reduce the effort : required for the first 2 years The general prdblems are-?rf" ,Ll} Joining graphite to—itself ' ' 2. joining graphite to Hsstelloy N, e 3;:fgeneral welding develoPment of Hastelloy N, | 4. 'Joining Hastelloy N tubes to Hsstelloy N headers, r,ifr L o 5. }development of remote welding techniques for a. variety of Hastelloy N JOlnts from 4. 1n. to 6 Tt in diameter,' 6, remote capplng or plugging of tubes. The proposed programs for. solving the graphite Joining prdblems are’ o included in the graphite section of this report The_programs for the ‘other problems will be presented in the following section, . General Welding Development of Hastelloy N. — The welding of Hastelloy N has been developed to the point that it was possible to con- ~ struct the MSRE using a wide variety of welded joints. . However, before a MSER is eonstructed -additional welding development is desirable, This general welding development will be aimed primarily at a better under— \ _standing of the weld-cracklng problem which -has been encountered with Hastelloy N but is also common to all high-nickel -alloys. Weld-cracking problems have: sporadically occurred, in commercial _ heats of Hastelloy N manufactured to the same specification and bygthe samé procedure. While we were able to work around this problem by using a weld-cracking test to select the material, the cause was not established. To avoid this problem, welding and metallurgical studies aimed at under- standing the baeic reasons -for cracking will be conducted. . Previeus-, indications that cracking'was-associated with segregation-of~alloying elements need to be expanded to . a widersvariety ofecompoaitions. These studies will rely heavily on the use of the electron microprobe to = delineate the segregation. Welds will be made in material prepared-fbrj‘ the radiation damage program and also in 5pecial‘small'heats specifically for welding studies. The various compositions will be welded at several heat inpdts with restraint applied to the pieces. Heats showiag tenden- cles to crack will be studied on the "Gleeble"61’which will be programmed to magnify any effects resulting from heating. As work progresses .on .the development of,a=radiationrresistant alloy,- it will be necessary to conduct weldability studies on the likely candi- dates.r Zirconium, one of the attractive alloying additions, is knowfi-to be detrimental to welding. Attempts will therefore be made to;finda set of welding parameters that will be adaptable to the zirdoniumecontaihing ‘ 617rade name for a machine designed to simulate the ty?e of heating obtained in the heat-affected zone of a weld. The specimen can be . fractured after heating to determine the resultant ductility » n alloys. It will also be necessary to make welds that can be fabricated into 1rradiation samples to demonstrate adequate radiation res1stance of the weldments - Currently, the filler metals used for welding Hastelloy N are of the ' same ‘composition as the base material. Results in the two previous | programs will be examined forlindications-of ways to improve the welds ,hy changing the composition-pf'the filler'metals. Possible improvements will be sought by.eliminating impurities'known to be harmful or by the addition of other elements'to alter the form of the harmful impurity. Preliminary tests have indicated that the addition of either niobium or tungsten to the filler metal.will 1mprove weldability ThlS lead will be pursued and the composition optimized - Joining Development for Components.'— As the engineering components' get larger and the reactor power increases, it w1ll be necessary to use heav1er sections An’ construction. With such sections, there w1ll be an econonic incentlve to use welding processes capable of high depos1tion rates. All previous welding of Hastelloy N has been with the tungsten- arc process which is reliable but slow. Processes of potential 1nterest for the higher depositiOn rates are gas metal-arc, submerged-arc and plasma welding. These processes by their very natures utilize high heat inputs and, therefore, studies of their heat-affected zones will be ‘required Welding parameters will be established for any process which in the prelimunary testing, looks promising These processes will also be studied as to their suitability for use in remote welding - Procedures for making tube to-header Joints for small tubes (O 50 in.) m;were developed and used successfully for the MSRE heat exchanger These 1Joints were welded and then baek-brazed For the MSBR, a- large number | '.of such Joints, many with much larger tubes, will be required The ma jor welding development effort in- this area therefore will be to adapt the \'present techniques to. the new geometries and to automate them. While 'f manual welding is. possible, the large number of sueh welds make automation, appear imperative | Because of the larger size components required and o limitetionS'in sizes of trazing or ennealing’furnaces, it msy be necessary _tothendle the bundles in subsections whichiare‘subsequently welded together. | . To reduce the costs and to simplify the designs, the'elimination of back-brazing will be explored. Conventional mechanical-expanding tech- niques; such as rolling or plug drawing, will be investigated and tested ‘for reiiability. Use of high energy forming for expanding and bonding _ would appeer to be & promising technique. We have the equipcnent at ORNL for these processes and will investigate them further. -Remote Joining. -A.meaor progrem will be required to develop remote welding and brazing techniques. It will be necessary to first develop techniques for the specific geometry end then to gdapt them to remote operation. The development of the positioning end guiding fixtures will be a major‘part of the program. These phases of the program will be part 'of the maintenance equipment development 62 Several different types of remote welds will be required and they will be sufflciently different to require separate development o . Although the welds differ in 'size and geometry, the programs for their development will attempt to answer 51milar questions. The general questions ares 1. Which welding technique will be most reliable? 2. What are the necessary welding parameters, such as voltage, current, wire feed rate, and torch speed, and over vhat limits may they vary? This will include sequences for start, Operation, and stop. 3. What is the recommended Joint design, and how much misalignment of the pieces can be permitted? . o o 4, How pure an inert atmDSphere will be required both inside and outside the pipe? 5. What are the effects of small amounts of fluoride contsminetion?,‘ 62p, Blumberg, Maintenance Development for Molten-Salt Breeder Reactors, ORNL-TM-1859 (June 1967) . _ Bhr o3 6. Wnat is the permissible misalignment of the welding torch? '7? How will the pipes and the torch be positioned? The major problem with the Hastelloy N- to-Hastelloy‘N butt JOints in tubes will be the limited space around each pipe; this is-approx1mately 1 in. Present plans are to develop a s1ngle procedure for the initial assembly and for remote repairs Several_weld joint designs appear possible and will be Investigated. .A-concurrent'effort_will be conducted on remote brazing for joining these tubes. This techniqne, which was developed for the MSRE, makes use of a conical- -type joint in which one member must either twist or slide as the braze metal melts .8 ‘A single development effort should suffice for all -the large (approx 12-in.-diam) remote pipe joints. Many of'these Joints are containment members, so a high degree of reliability will be reqnired This makes remote welding much more attractive than mechanical-goining= . techniques,- While these pipes are of large diameter,-they_dolnot have heavy walls, 'However,:the'additiOn\of fillerumetalrwillibe necessary. After completion of exploratory studies, a selection of the applicable welding process will be made. . Welds similar to these were developed by Westinghouse for the Pennsylvania Advanced Reactor,64 and their procedures will serve as a starting point A complication with these welds will be the tight 1imit;onrpipe,lengths required to minimize fuel inventory. The positioning and aligning-present difficulties.since‘short;;large- - diameter pipes are very rigid,— An'insert may be.requiredato gnide,the g _pipes together. - If one tube of a heat exchanger develOps & 1eak, it w1ll have to . be plugged or the whole unit will have to be replaced.; It Wlll be neces- lsary to plug the ends of the failed tube through a small access port in . each.end of.the exchanger Such plugging procedures are standard on - 63, C. Hise, F. W. caské; and R G. Donnelly, "Remote Fabrication of Brazed Structural Joints in Radioactive Piping," Paper 63-WA-53 of the Winter Annual Meeting, Philadelphia, Pa., November 17—22 1963 ~ of. the American Society of ‘Mechanical Engineers SR 64F. Y, Seidler, Pennsylvania Advanced Reactor —-Reference Design Two, Layout and Maintenance, Part I, WCAP-1104, Vol. 4 {March 1959]), % commercial heat exchangers, but are not for remote applicationsrl'The positioning equipment for such & job would be similar to that used to cut small pieces from the HRT core tank. ¢ | : L | For such an application, a plug must be developed which can be inserted‘into the tube and then fused to the header. A major problem will be-that'snch a weld is by nature highly restrained and is, therefdre, subject to cracking. This may require trepanning the head before the plug is inserted and welded. | | - - Inspection Development While a complete evaluation of testing and inSpection problems cannot be made until a firm design is- available, some are apparent from the con-'r_ ceptual design. The ones requiring most attention are those that will have to be made remotely. Inspection development for the Hastelloy N tubing. and pipe will consist of adapting available techniques to the necessary configurations and to the required sensitivity levels. The techniqpesu developed for inspecting the MSRE heat exchanger should also be adaptable " to the MSER heat exchangers. ‘Techniques to be used would include penetrants, radiography, and ultrasonics. A Remote inspection will be required for (1) Hastelloy N tube Joints at the fuel header, (2) large butt-weld pipe joints to disconnect “the main piping from both the reactor and the heat exchangers, and (3) plugs in the heat exchanger tubes. ‘Until develo@ment of the joining techniques has progressed further, it is impractical to speculate on inspection | . procedures. Several testing techniques will be evaluated In developing the nondestructive testing techniques, three phases ‘must be completed (not necessarily in sequence). They are (1) demonstra- tion of feasibility, (2) determination of test sensitivity and establish- ment of reference standards, and (3) development of detailed techniques and eqpipment 65p, P. Holz, Description of Manipulator System,'Heliarc Underwater Cutting Torch, and Procedure for Cutting the HRE-2 Core, 0RNL-TM—175 -~ {(Nov. 5, 1962). R _ _,"Zfl, ks ' ' L PN O 75. For the remotely inspected joints, each of the steps must first be _ performed in'a'coldrlaborétory (but with cognizance of the need to progress to remote operation) and then the necessary mechanlcal devices must be,developed for the remote,performance. It may be possible to devise accessories which can~beiattached»to the manlpulators used to fabricate the joints. Materials Development for Chemical Processing Equipment The fuel salt will have to be continually reprocessed to remove fission products. The present concept of this processing calls for dis-- tilling the salt at a temperature of about 1800°F. The strength require- ments are quite modeet,;but-the'material will be exposed to salt on one side and to some gaseous atmosphere on the other side. The easiest gas- eous atmosphere to obtain is the cell environment which is N, + ~2% Oy . This would require that the material be fairly resistant to oxidation and nitriding.--Theioxidation resistance is improved by increasing the chro- mium content of the alloy, bfit:this.in turn inereases the corrosion rate on the salt-side; We feel that several possibilities exist for materials for constructing this vessel. - - | 1. Hastelloy N with an inert cover gas, 2. a molybdenum-base alloyflwith an inert cover gas, | 3. a duplex system with arnickel-molybdenum alloy, exposed to the salt and an oxldatlon-resistant materlal such as Haynes alloy No. 25 _exposed to the: cell env1ronment -'4.! a vessel made of an oxidatlon-resistant alloy with a graphlte liner. - These possibllities and. others will have to be explored by screening | ftests to determlne re51stance to. ox1dat10n, nitriding, and salt corrosion. ”_The ease of fabricability and cost of the candidate materials will also - ',be considered in making a flnal choice. ‘5,1;Gféfifiite_Program;'«f The graphlte that we need for this reactor requires that some advances be made in present technology Therefore, our first concern is to procure material that meets our75pe01f1catlons for study. The ma jor 1tems to be 76 evaluated for this material are (1) determining its behavior at radiation dose levels as high as possible, (2) joining of graphite-to-itself_and'to structural materials, (3) fabrication development to yield a low gas permeability, (4) evaluation and charaCterization of the~m0dified'or' improved graphite, (5) determining its compatlbllity when used with Hastelloy'N in a system circulating fused salts at elevated temperatures, and (6) fabrication of the graphite into engineering systems and evaluating - its performance. Because of its greater potential for reducing radiation damage effects, the major effort will be on the development of isotropic graphite, However, this material is so new and so little is known about it, development of anisotropic graphite such as the needle-coke'types will also be continued. The proposed development program is- discussed in the f01lowing sections. - ' Graphite Fabrication and Evaluation - The grade CGB graphite used in the MSRE was produced in experimental i equipment and its properties still leave much to be desired for a MSBR. - No graphite from any source is now available that w1ll meet all of the e MSBR requirements, Since there is no present industriel need for low= . permeability graphite, we are not likely to obtain much allied develop- ment help from industry. An adequate supply of the highest'quality. graphite will be purchased to supply material for use in test rigs, to establish reasonable specifications, and for evaluation. Continuing orders will be placed as improvements, or potential improvements, are made in the quality. The intent is to progress in order size from the laboratory to the pilot plant and then to a.production;size order. Until msteriel is actually produced in production equipment and is teSted' - dodbt will exist as to how representative the smaller. batches really are, It is expected that grade CGB or a similar grsphite will be used in the first tests. The various problems to be faced will include: 1. As soon as p0381ble, consult with msnufacturers end.prepare a specification for small lots (1aboratory quantltles) of graphite of as - e near MSBR quelity as it is pOSS1ble to obtaln immedietely | ' ' kaj LG il 7 R. Place small orders for both 1sotropic and anisotrqpic grades of graphite in the form of block and pipe.-f o - 3. Measure pertinent physical and.mechanlcal properties of representative graphites. - - - _ 4, In about 1 year, prepare a Specification for a pllOt-Slze batch - of the most prom1S1ng type, or types, of graphite | It 1s not expected _ that any significant new radiation-damage 1nformation w1ll be available at this time. e _ | L 5. If suitable graphite cannot be purchased from out31de sources, an in-house pilot-production facility will have to be established . Irradiation Behavior In laylng out any graphite development program, an enigma rapidly ‘becomes apparent One of the major unknowns is the effect of massive | neutron doses. To Obtain the desired doses requires irradiation periods of two to three years, and we,would like to firmly establish in three years that graphite will have an acceptable 1ife in the MSBR._ The problem therefore 1is that, if during the development period wmajor changes are made in the graphite, the graphite being irradiated will no longer be repre- sentative and the results will therefore be suspect. The most satisfactory. solution appears to be starting irradiation tests as soon as possible with samples fabricated from the highest quality material availahle and to fol- low these with other tests of improved material The purpose of the work would be to determine first the dimenszonal stability of the graphite for __.the temperatures and high-radiation doses that would have to be sustained ~in the MSBR plus the effects of irradiation on the. graphite properties, isuch as creep ductility, mechanical.pmoperties, accessible voids, and "'-permeability. - The. inclusion of ‘both anisotropic and isotropic graphite is warranted since the anisotropic has more technological development ~behind it, but the isotrOpic graphite which has been under development for -approximately three years shows more potential These experiments should ,-Ebe performed on graphlte, graphite~to-graph1te Joints, and graphite-to- ' Hastelloy N joints. | o 78 To demonstrate the ability of graphite to sucCessfully ret&in_its | integrity after exposure to dose levels of 1 >< 1023 neutrons/cm? (5-year life), we plan'fo irradiate gfephite to as near this dose as possible. This will require the irradiations to be performed in reactors that have fast fluxes of 1015 neutrons cm=? sec™l or greater to obtain the data in a reasonable time. At present,-there are only tworfast reactors, EBR-1I and Dbunfeay, and one thermal reactor, HFIR, that approach the'fast-flux requirements.' Approximate values of the time required to reach a dose ‘of 1 x 1023 neutrons/cm , cost per year, and facility size are listed in, Table 11. Table 11. Comparison of Reactors EBR-TT Dounreay I-IFIR - Iflme to 1l X 1023 neutrons/cm year 6 4B 'i 2 IR Cost per year, $ . ~ 20,000 300, 000 15,000 Facility size, in. 0.75-diem 0,75-diam 0.50-diem TEWOEEER S s X 20 a2’70-operating power days per year It is, therefore, proposed to use HFIR to demonstrate the ebility 7 of the graphite to retain its integrity after an exposure of - 1 x 102> neutrons/cm®. Due to the size limitation of HFIR, the irradi- ation will be restricted to simply prepost type of-testing. To | demonstrate the ability of the MSBR graphite to absorb creep"Strein, restrained growth type experiments will be performed in EBR-II. By enforcing & creep rate equal to the growth rate, the creep strain aceumulation.will be forced into the material ten times faster than. under MSBR conditions. Although en equivalent neutron exposure will not be’achieved,'the enforced strain will be-greater than expected in the MSER, - I . O 79 Graphite Joining Since,the graphite-to-Hastelloy N joint is a_very important part of the system, we will carry}parallel efforts on at least two different types of joint;-fBraze'joints.ofnseveral designs will be evaluated using the 35%‘N1—60%'Pd%5% Cr allcy\and'pure copper as brazing,materials.; These. - joints will be designed so that graphite is initially in compression where it is strongest and has the ability to undergo small smounts of plastic strain. This will.minimiZe the'tensile stress that will develop ‘in the graphite as it shrinks due to irradiation.. A mechanical type of joint will also be evaluated as a second preference. “These joints will be made u31ng the 5-1n.i0D'Xel/2-in,-wall.graphite pipes that are,presently in the MSBR design. These joints will be subjected to thermal cycling, - evaluated'for'corrosion'resistafice'to the'fuel-and'blanket,salts, and irradiated to determine integrity under service conditions, - o 'The’graphite;to-graphite joint'will be studied in'detail | Based on this study the choice will be made between a graphitized Jjoint and a | brazed joint. : | o Permeability Studies | As'"imprcred“'grades”dfigraphite are received from producers, it will be necessary to determine their permeability for molten fluorides, helium,:and'fissianegases.z'AS'was pointed out in the‘previcUS'diScussion, | obtaining graphite with the very 1OW'permeabillty of 1 x 10~7 '2/sec “will be very difficult ‘When the data on xenon stripping and’ the designs " of the MSBR become firmer, realietic Values for gas permeability will have to be established This program will include: s 1. »studies of permeability to gases and.mclten salt of"- the variousfl o ie-grades of graphite | "I,"” o ' EREE R '2a-iinvest1gation of the prcperties of graphite which affect the perme- - rability and how it may be minimized "_' o 33.; measurement of the effects of various coatings on. the permeahility _ _ _,of gra,phlte, I e o .fi R S 4., determination of the permeability of graphite joxnts 5. determinationrof the effects of fluorides and solid flssicn_products on the graphite. 80 Corrosion and Compatibility of Graphite Some additional data will be obtalned on the compatibility of graphite with Hastelloy N in systems c1rculat1ng high-temperature fused salts. The testing must also include the brazed 301nts. Initial tests will be simple capsule tests containing both materials. Then small—scale grephlte and graphite-Hastelloy N thermal-convection loops will be fabrlcated and operated with fused salts to determlne the compatibllity and corrosion ' resistance of these materials. The final stage will,be_testing in—the . engineering loops to evaluate-full-scale.conponents.'.Theseflarge'loops : are e part of the component developnent progranm, 'The functionrof'this . task will be to thoroughly examine the components after exposure in the loops. | | o S Date on effects of 1rradiation on corrosion and compatibllity Will be obtained from in-reactor tests in the Chemical Research and Development Program.®® Various types of graphite are being ineluded in the Surveil- lance Program with subsequent evaluation of fission—product retention and changes in mechanical and physical properties. Graphite InsPection The necessary nondestructive testing program for the graphite tubes proposed to be used for MSBR would cover severel test methods It is 7anticipated.that,eddy-current techniqnes would be used for dimensional kgaging.and both lowsvoltage.rsdiography.and eddy-current“techniques would be used to detect discontinuities, If laminations are present and unde- sirable, it may be necessary to use nltresonics or infrared techniqnes; to detect them, For geometries more complex than concentrlc tubes, the seme methods would likely be pursued. However, there would be more requirement for dimensional gaging and the flafiafinding techniques WOuld | be more difficult to develop. The development would proceediin_the_;; following steps (not necessarily in sequence): é6W. R. Grimes, Chemical Research and Development for Molten-Salt Breeder Reactors, ORNL-TM-1853 (June 1967). L ad 8 1, Vdetermination of feasibility-of application of the test method to ~ the. conflguratlon and grade of graphite, 2. _determlnatlon of characterlstlc flaws and relative test sen31t1v1ty - attalnable - 3, establishment of reference,standards,_ b, development of detailed'techniques_(perhaps including equipment). General Development and Proaect A851stance - Many details of the f1nal design will depend upon the properties of the two primary structural materials -~ graphlte and modified Hastelloy N. . As we progress with the development of these materials, we will keep the designers informed, so that this information can be 1ncorporated in the _de51gn. Several of the planned engineering tests‘w1ll,also require | assistanCe. We shall be called upon for design assistance and fabrica- ftion technology. The evaluation of many of the test results will be made by'materials"personnel. ‘After working with the vendors to obtain the desired products and With.thepdesigners to finalize the MSBR design, we will take an active part in writing thedSpecifications, procuring the materials, and assisting;in}the'actual oonstruction. ‘Several parts of the proposed molten-salt systems could probably be made'of'cheaper materials;~such as the-austenitic stainless steels. We do not have adequate corrosion data on these materials, but tests will be initiated to obtaln this informatlon on a second prlorlty basis e'We w1ll crltlcally assess the properties of these materials for thiS' o program and 1ntroduce them where they appear advantageous. - f hnmomsmrmNTs | The authors are indebted to G M .Adamson, Jr,, and A Taboada for_ 'r*assembllng the flrst draft of thls report, 0RNL-CF-66-7-42 Several ' persons contributed informatlon for the various sections-' J H. DeVan, G. M. Slaughter, R. G Donnelly, D A Canonico, E A. Franco~Ferre1ra 'R, W. McClung; W. H. COOk andC R Kennedy - APPENDIX 2 ' ' ' ¢ ~ Cost Estimate for Materials Development for the MSBR" Tasks and Type Costs FY-6¢ FY-69 FY-70 FY-71L FY-72 FY-73 FY-74 | Manpowver Costf' | Hastelloy N i | o L | B ‘Irradlatlon testlng " j ' . 250 250 200 200 100 100 <8 Joining . e 120 120 1200 120 90 60 Corrosion and compatibllity o 120 120 . 120 90 60 30 Nitridlng o | - 20 L RN o Inspectlon develqpment - . 60 60 60 - - 30 30 , subtotal _‘f“j) fi3‘¢‘"f" .;' 570 550 600 440 280 190 Graphite _ _ "A:f* "_;;j f - | L S Irradiation testing R 100 100 100 1100 . 35 35 Joining S :,; o 75 75 60 45 30 30 Permeability oo 60 60 - 30 15 15 Procurement and characterization 120 120 90 60 30 System development ‘ - 60 a0 60 60 30 . Corrosion and compatibility 30 30 30 30 S ‘Inspection develqpment | - ‘ 45 60 60 30 30 Subtotal 490 535 430 340 170 65 General o 'f' R o ) | | | | | MEtallurglcal serv1ce e 60 90 90 90 30 Outside service . = 30 60 - 60 30 | | Supervision, secretary, ete. ~60. 60 60 60 30 30 30 Project assistance =~ 60 90 9% 90 - 60 60 60 Subtotal 210 300 300 270 120 90 - 90 Total 1270 1385 1330 1050 570 345 90 Cost Estimate (continued) | Tasks and Type Costs - FY-68 FY-69 FY-70 FY-71 FY-72 - FY-73 FY-74 Unusual Cost 'Hastelloy N Reactor . | 50 50 50 40 40 | Hot cells 60 60 60 60 60 - 30 Experiment and specimen f&brication 110 110 140 100 s - o Materials 85 100 100 50 30 - | ‘ Analytical chemistry _' | 40 30 30 30 10 5 : Subtotal 345 350 380 280 190 35 Graphite o o B S B - Reactor o 100 100 50 100 100 100 & Hot cells - ‘ ' 10 20 20 20 10 10 | Experiment and specimen fabrication 60 90 100 100 40 : Materials 80 120 100 10 10 10 _ Analytical chemistry 20 20 20 10 5 Subcontracts 15 35 25 o - | Subtotal . 285 405 315 240 165 120 Equipment | . L 100 50 50 - 50 10 . Totel 730 805 45 570 365 155 | Surmary of Costs ) | | Manpower total 1270 1385 = 1330 . 1050 570 345 90 | | Uhusua1 £ota1 o o | | ,"730_ - 805 ‘745 i'f" 57Q . 365 155 :" ‘ | Grend total 2000 2190 2075 . 1620 935 500 90 gAll‘costs in thousands of dollars. e g oy S ot 4 k. 1-2. 413, 14, 15. 16. 17. 18. 19. 20, 21. 22. 23, 24 . 25, 26. 27. 28. 29, 30, 31. 32. 33. 34, 35. 36. 37. 38. 39, 40, 41. 42. 43, - 44, 46, 47 .. - 48, 49, 50, 51. - 52, 53. 54. 55. >mpmuwgwaMMmzmupmwggpomwgwwwgmpzmaowflwnw - INTERNAL DISTRIBUTION Central Research Library - ORNL — Y-12 Technical Library - Document Reference Section Laboratory Records: Department Ieboratory Records, ORNL R C ORNL Patent Office Adams Adamson , Jr. .—Affel Alexander Apple ‘ Baes. Baker Ball = Barthold Bauman ' Beall ‘Bender - 'S, Bettis - F. Blankenship -E. Blanco 0. Blomeke Blumberg G. Bohlmann J. Borkowski E. Boyd . Braunstein A. Bredig ' B, Briggs - oo R. Bronstein = -l D. Brunton A. Canonico - » - - - - * . - - - mmpqzfiwpmgfi - T . . Cantor . L. Carter . I. Cathers . M, Chandler . L. Compere . H. Cock - . T. Corbin . L. Crovley . L. Culler . E. Cunningham -~ . M Dale , G. Davis . J. Ditto « S. Dworkin 69. 70. 56. 57. 58. 59. 60. 61. 62. 63. 64. - 65. 66. 67‘. 68. 71, 72. 73, Yler 75. 76. 7719, 80, - 81. 82. 83. 84, 85, 86, - 87. 89, .90, 91. 92, © 93, 94 . 95, 96. 97. 98. 99. - 100. 101. 102. 103, spawmznwwmnuwwwb;mmoym?pdhuf Q”“”F?”P““?mfififigs?fi ORNIL~TM-1854 .- R. Engel . P. Epler .. E. Ferguson M. Ferris P. Fraas . A, Friedman H Frye, Jr. . H. Gabbard . B, Gallsher . BE. Goeller R. Grimes . G. Grindell . H. Guymon . A. Hannaford . H, Harley .. G. Harman . S. Harrill .- N. Haubenreich . A, Heddleson . G. Herndon . R, Hightower . R, Hi1l1 . ‘W. Hoffman . W. Horton . L. Hudson - Inouye . H., Jordan "R. Kasten J. Kedl T. Kelley cJ. Kelly . R. Kennedy . W. Kerlin . T. Kerr . S. Kirslis ;- T. Krakoviak . W. Krewson . BE. Lamb -A. Iane. . B. Lindauer P. Litman - -I. Lundin . N. Iyon . G. MacPherson . E. MacPherson . D. Martin &8 - | 104. C. E. Mathews . 184. - C. E. Sessions - - 105. €. L. Matthews 185. J. H. Shaffer | ~ 106. R. W. McClung o | -186. G. M. Slaughter 107-111. H. E. MeCoy, Jr. o - 187. -A. N, Smith 112. H. F. McDuffie o ' '188. F. J. Smith 113, C. K. McGlothlan - 189. G. P. Smith 114. C. J. McHargue =~ = - 190, O. L. Smith 115. L. E, McNeese - . 191. P. G. Smith - 116. A. S. Meyer ~ 192. W. F. Spencer 117. R. L. Moore . .. 193. I. Spiewek - 118, J. P, Nichols . . - 194, R. C. Steffy - 119. E. L. Nicholson IR '195. H. H. Stone ~ 120, L. C. Oskes o o 196. J. R. Tallackson - 121. P. Patrierca , © 197, ‘E. H. Taylor 122. A. M. Perry o - 198. R, E. Thoma - 123. H. B. Piper I 199. J. S. Watson 124. B. E. Prince - ... . 200. €. F. Weaver 125. J. L. Redford . .- | 201. B. H. Webster - 126, M. Richardson o o 202. A. M. VWeinberg 127. R. C. Robinson ~ o 203—207 J. R, Weir, Jr. 128. H. C. Roller ' | 208. W. J. Werner '129-178. ‘M. W. Rosenthal 209, K. W. West 179. H. C. Savage ‘ ' | - 210, M. E. Whatley : 180, W. F. Schaffer 211, J. C. White | T 181, C. E. Schilling - 212. L. V, Wilson ' : - 182. Dunlap Scott. - ) 213, G. Young - . ‘H. C. Young ' _183 H. E. Seagren B 214, EXTERNAL DISTRIBUTION 215-216. D. F. Cope, RDT, SSR, AEC, Osk Ridge National Leboratory 217. A. Giambusso, AEC, Washington 218. J. L. Gregg, Bard Hall, Cornell University 219. W. J. larkin, AEC, Oak Ridge Operations 220-234. T. W. McIntosh, AEC, Washington - 235. R. M. Roth, AEC Oak Ridge Operations 236—237. M. Shaw, AEC Washington - 238. J. M Simmons, AEC, . Washington 239. W . L. Smalley, AEC, Oak Ridge Operations - 240. D. K. Stevens, AEC, Washington | - 241, R. F. Sweek, AEC, Washingbon ' 242-256. Division of Technical Information Extension . O