g ¥ ¥ LY PROGRESS R KL "“-"" . BT - SETLLLT : . Lok f . tor' the Comm-ssmn, nor uny pflsm uchng on l:eholi oi the Commuslon.. . A, Mckas nny wurrcnty or reprqsqmuflom ‘express -or ;mpltad with’ respecf fo !he occurucy, ,completoness, _or usefulness sf the in!ormflhan contamed in |h:s reporf or !hct fl:e use. of' :"_emy mformut:on, upparoms, methcid ofprocess d:scloud ln this. report, - P As -used in the ubove, person acflng on beho" of fhe Commlsslon mcludes nny ernployeg o.-.:‘." g ‘ comrucror of tho Commissnon to the -exte ; S ) ORNL.-2474 - _ UC-81 -~ Reactors—-Power Contract No. W-7405-eng-26 MOLTEN-SALT REACTOR PROGRAM QUARTERLY PROGRESS REPORT For Period Ending January 31, 1958 H. G. MacPherson, Program Director DATE ISSUED lwlfiJ MAY 141958 e omc RIDGE NATIONAL LABORATORY - - Qak Ridge, Tennessee - -operated by UNION CARBIDE CORPORATION for the U.S. ATOMIC ENERGY COMMISSION g 1ok e " .‘"*‘f =) ) L2 o FOREWORD This quarterly progress report of the Molten-Salt Reactor Program records the technical progress of the research at the Laboratory, under its Contract W-7405-eng-26, on power- producing reactors fueled with circulating fused salts. The report is divided into two major parts: 1. Reactor Design Studies and 2. Materials Studies. Until July 1, 1957, the Molten-Salt Reactor Program was largely a design study, with only token expenditures for experimental work, As of July 1, the program was expanded to include experimental work on materials. A further augmentation of the program occurred on October 1, 1957, when personnel and facilities for additional research and experi- mentation became available. As a result of these transitions, this quarterly report has been expanded to include component development and testing, engineering research, metallurgical and chemical investigations, and radiation-damage testing. v " a} N » «) 3Y CONTENTS SUMMARY ettt e st s esa e e et sae sert se bR s abe Sa0m b sste s ot £ b en e s e e e RR e R bt e e s asbeEnR e arA s et e et sasaen s 1 PART 1. REACTOR DESIGN STUDIES Tol. REACTOR DESIGN......ou oot ceererereseeeieretessssesteserenssresasesasessessasasbs sssnsesenssssasassssstas atanssssess sssassenssasanassssaeses 9 Study Layouts of Reactor System........ciriininicncinii it s issess e sassse s ennsnssnns 9 NUCIEAr CalCUIGHIONS ..ottt i s s s rsee et sese s e e esn e s asenases sunsosesbrnnnssnnass mbranaessessens 12 Univac-Ocusol Reactor Calculations ............ eeeeeestteasearaeat et bereaat e et ranses b e b et est s s et s sasae e s 12 Oracle Caloulations. ..t ieerarsreseere s eas s e ssasssteeseemesrse s bessas e sbassssassesarsssnensassees 14 IBMo704 COURS v.veureiireinenreiencreesiereseuree et rerrresssesaesaass seestssasasesss sasasssssassessstonsstaserasstessare ssnenessnrsens 15 Fuel Fill-and-Drain System Design ......cc.vieeeiniiintiiintnisstssssese s ssasss s sesssssasssssnns 15 Design of a Hydrostatic Bearing for a Molten Salt Pump Impeller ..., e 16 1.2, COMPONENT DEVELOPMENT AND TESTING.....cceoriinerrneeneceecrennameeenessssea st scnssss s sasnssnsses 18 Fuel Pump Design and Development .........cccuiiiiiinniiniiiciiinn s svsnsaenssiessssssnsesanseasas 18 Pump Design Studies ... sttt snas st s e e san s e n s e nas 18 Development Tests of Salt-Lubricated Bearings......cccoviviiminiiiiininiiniiessnens 18 Development Tests of Mechanical Shaft Seals ..o 19 Development Tests of Oil-Lubricated Bearings ......c.cccovvrvvnrrercrrnccnnncn. rereerereeteteiaate e sseenans 20 Irradiation and Endurance Tests of Bearings and Seals ....uooeveivcerrniiniiiincinncicencnn, 20 Development of Techniques for Remote Maintenance of the Reactor System ......ccevvnrcenrnencnes 20 Mechanical Joint Development ... ... eriicrinec e e esenae e tisenessssns s s ssessnsnns 20 Remote Manipulation Techniques .....cciericiicncees vt e s b be s s a s aees 24 Heater-Insulation Units for Remote Application ...ccccevirecirceneriiniiirnen e 25 Maintenance Demonstration Facility.....oeeivrrnereinrre ittt st st sre s eessssrsenes 26 Heat Exchanger Development ..........coccmiiiiviniicii st s s st rasss s st s ssesssssenans 27 Design, Construction, and Operation of Materials Testing Loops e reeuseesenasssbesrase s srassnesastsans 27 FOrcedeCirculaHOn LOOPS oeuvrivrrireiteece e srttteesesssiresanearessssesaresasesaes sesssssnmsssessensssssrasassssnsnans 27 in-Pile LLoops .coccvecccirecennens Ceuerusasheneterenaa e e e e R E B SR e sS4 SO A SR R SRR SRR e 08 32 1.3. ENGINEERING RESEARCH -ovevrssecssseserrensern e gen e s 35 Hydrodynamic Studies of MSR Core e rienreneerasase rreseeeserte e irsnerbe e reneher et e b npanesa et se s area sas veeersneas 35 Physical Property Measurements ......ccivoceenivescssencees renenatasainaransnesassansrenin \evesieasasaeiestesarensizasasensns 37 Molten Salt Heut Transfer Studie's torsosentsivatrerebaisRR L ERRs P AL R 0RO 4 RS S8 SR e SRR b 088 ES 37 1.4. ADVANCED REACTOR STUDIES 1vvcerermrserrnes et e et s s eenee 41 A Molten Salt Naturcl—Convechon Reactor ..... eieerieenararessereas eseesesere s rene e et st e st e e ra s be e nanenssaass 41 Salt-Cooled Heat Exchangers .......coviemsrumimuesnssivnnnsnisssonna: eestereancpeses et e rsa ks peR bbb r R et erere st eaas 42 . Helium-Cooled Heat Exchangers for Steam Cycle ..................... 42 Helium-Cooled Heat Exchanger for Gas-Turblne Cycle ................. treresssisvensfosepissnaranssomssasst s 43 Companson of the Various Coohng Sysfems ..... Cisvinevessnsisssasass eterestsessseesseseneesansesrreonrassntoaran, 44 ' Compurlson of Nafura|-Convecflon System with Forced-Convechon SyStem..umvereeriarerrraeens 44 ngh Flux Reacfors .............................................................................................................................. 44 ~ PART 2. MATERIALS STUDIES 2.1 METALLURGY ocerenoerms oottt s sttt s oo 51 B Dynamic Corrosion Studies .......... ................... 51 Thorium-Bearing Salts in Thermal-Convection Loops .....cccccecevireucccrerivenerenens enenerases s annans 51 Uranium-Bearing Salts and Coolant Salts in Thermal-Convection Loops: .....eceeurrvarsrermenenne 52 Results of Examination of Samples Removed from Forced-Clrculaflon Loop CPR ............ 54 General CorroSion STUIes.... ..o vieiiiinnctcse et ce s seeerens setens e st s e asresstsssnsasesnsnsssssnssnasasssasans 54 Effect of Carburization on Reactor Structural Materials .......oooverviiiiiinniniiincccccninee 34 Corrosion of Brazing Alloys by Fuel Salts .ot erere et seenee et aes 59 Corrosion of INOR-8 Welds by NaK and by Fuel Salts....ivriiiisieniriicnnsinnns. 60 Physu:cll Properties of INOR-B ........... 62 Mechanical Properties of INOR=8.......c..coccuiriereierivscsissnssssesmssessessssssssessimssrsssssssessasesssses sesnssssorssn 63 Welding and Brazing Studies.....c..cccviverrnnnes vemesiestsEeneneesnete stssneneiote sermsnneca sbe s bsmnn esisatrest siem e s as st asen 64 Metal Seals for Remote-Disconnect Flanged Jomts ................................................................ 64 Welding of INOR=8 TUDBING ..cccoueveminrcerncieinneestininssccresccsasnest s ssssnsnsessesessssssnsssssssesssssosanssass e 65 Evaluation Tests for Welds ...coeveeverereccnrnnene. ceeaeeeersta et et srennnsenans eereenesaerseesesenransanssasrees 67 Examination of INOR-8 Forced-Clrculoflon Loop That Failed During _ NGl HEGHRG coveveereeriieceecrereneeee e senscesaseresnsnssssness s stsssssssse ssrsasssesssssssnsnsesnsnssessnsans oos SR 71 Fabrication of Test Components ... cienieneenecsenrereseiesieresnsssessesses ssessesassssssons vereseneensneen 73 Development of Nondestructive Testing Techniques ............... eteresener et et s st et sa e e R n s aerasaensanane 75 Evaluation of INOR=8 TubiNgG....ccccceiieerierecrereriimneierssniameesesaessnsssnsssessnsssnsesnessessnsssasssessenssssesses oe 75 Cladding Thickness Measurements and Bond |nspechons ...................................................... 77 INSPECHION RESUIES oottt et e sanens et s e e s assessens oo snasnasennesnssbanesbnsnesesstanses 77 Material INSPECHION ..t re st eesae s e res srasae e basssesaesees e gesmssassssortssresnanss ebesnss 77 Weld Inspection ....cccvvenvvivvnivreesinnenna. fevereestnteterseesaneratetesreeaseeaeanteertsashnetarnaebastea s tesasesnanaernnteanens 78 Failure AnGlYSes ...ccooueiiiieieiiicieciteceescseree st eiae e sebestsnesssesesssesssasses s ssnn st bssassarasmesssnsnensansinssnnnss 78 2.2, RADIATION DAMAGE .....ooceireecreeetetisreres st setestetesss et eassssss et evesbasbeserssssasssssesens evnereree st et rae s 79 In-Pile Thermal-Convection LLoop Tests ... cnseeseerrsnesessssissesssarsesssessesnsresssseseasens 79 INcPile Static Capsule Tests ettt reveessrsessesstsst ssssessesersasnenpessssassrnssessons 80 2.3, CHEMISTRY wooroesoeseevssrssssmssmssesssssmssssmssssssssssnessesssssss s e s s 81 Phase EQuilibrium StUdies ...t s riaee s siasssssstsssssstssenssssnssssssesasssssrnsnens 81 Systems Containing UF, and/or ThF ; oottt 81 Solubility and Stability of PuF, in Molten Flyerides.......... etereearesteneaestanteasereesavares antasesnsassen 88 Fused Chlorides as Secondary Heat Transfer Fluids c.cuviveciniinniiniccsniitcic s 90 Yapor Pressures of LiF-BeF, Mlxtures .......................................................................................... N Fuel Reprocessing....cccomerpeinennininisnennnsicncsianeesssosnnssessesersessssesssass veranresaetaesasreteseresasnssnanes 91 Solubility of Noble Gases in Molten Fluoride MixtUres .......c.veeeemeeemreereesesesssscersessssereensens 91 Solubility of HF in Molten Fluorides ......rencniescncsriecnnenees SO 93 Solubilities of Fission-Product Fluorides ... 94 - Order of Oxide Precipitation in Fluoride Salt Melts......cccocevvverreccriveccrnnne reeveenereeeesnraseres 97 Chemical Reactions of Oxides with Fluorides in LiFeKF .. - 99 Lithium Recovery from NaF-KF-LiF Melts .......ccocovirommreomnerrreessosessmsosessssssenesssssssseees cereeras 101 'Vvi 13 ) 3 ) %) » «) # » *) Chemistry of the CorroSion Process .........iccriieecseeriesessessisssssmessssssisessssssessssssssssssssssassenes 105 Activity Coefficients of CrF, in NoF-ZrF......cooeece... reeierenesrater et enensaereans peesesresenesesanenses 105 Solubility of FeF, in LiFeBeF 5 ceeecs st sttt arnes 107 Use of Cr3! to Study Chromium Migration in Polythermal Inconel— MOoHEn Salt SYSIEMS .o.ceiveirice st s ras s srsasr et e e s easn s s sarens 107 Activity of Nickel in Nickel-Molybdenum Alloys .....ocoirvrerrcineeerecene s m Production of Purified Mixtures ............. reeserenerresteaesrnerasnsenns rereneeenne rerereseessserserrresesiaseenessesnsnnesassane 113 Preparation of Pure Fluorides ..........coiiiicvrivrrinennneirecnemnnsessnsssine s sseesssssssssssssessnens 113 Small-Scale Purification Operations ........ccoeeuverieemssiseeennn s snssesuesnsaresaserenseanssesesensrasennsasrenss 113 Preparation of Material for InaPile Loop ... cseesesesae s ieesasne s 113 Tronsfer and Service Operations ........cucueene. rieaeseeeseiearseseeessestresseestesanasbessseabeshesentennrtstennrennean 113 vii J *) NIRRT RTE S ) " «) N Y N ) 'be possible to exceed 1.0, MOLTEN-SALT REACTOR PROGRAH QUARTERLY PROGRESS REPORT SUMMARY PART 1. REACTOR DESIGN STUDIES 1.1. Reactor Design Preliminary layouts have been prepared of the molten-salt reactor system that are based on the use of five fuel pumps and two blanket-salt pumps, The total thermal power production is assumed to be 638 Mw, with 90% of the power being generated ‘and leakage probabilities. The resulting constants reproduce the initial spectrum ond critical con- = centration exactly, ' S A preliminary fuel flll-cmd-drc:m system desngnr' - 'was completed which satisfies the design criteria - in that (1) it is always in a standby condition in which it is lmmedlately available for drainage of ~ the fuel, (2) it can adequately handle the fuel in the 8-ft-dia core and 10% being generated in the | 2-ft-thick blanket. The fpfopos_ed layouts are being afterheat, and (3) criticality cannot be achieved in - the drain system. The fill-and-drain vessel con- studied in order to achieve simplification and to minimize the fuel inventory, reactor cell are also being studied in order to determine the best possible arrangements of piping and other components that . will provide minimum plant size and yet facilitate remote maintenance.. The parameter study of two-region, homogeneous, molten salt reactors was continued through nuclear calculations on the Univac, the Oracle, and the IBM-704, The Univac calculations indicate that, for the same core diameter and thorium content of the fuel salt, the U233 inventory would be one- half the U?35 inventory., In the U235 case the regeneration ratio is limited to a moximum -of Layouts of the sists of forty-eight 20-ft lengths of 12-in.-dia pipe : arranged in six vertical banks connected on - opposite ends with mitered joints, The system is preheated and maintained at the desired tempera- ture with bayonet-type electrical heaters in tubes located axially in the pipes. Removal of afterheat will be accomplished by radiant heat transfer to ‘banks of wuter-f:lled boiler tubes that will normally be dry. A hydrostahc bearlng was desugned for use in fuel pumps which differs from the conventional bearing in that the pockets rotate on the impeller. - A bearing of this type has the advantage that the about 0.675, and in the U233 case it appears to | It is pointed out, however, that the regeneration ratio for a y23s. fueled reactor will decrease with time as burnup poisons accumulate, while in the U235 fuelec!_,_' reactors with initial regeneration rahos ‘of -about . 0.6 or better the regenerqhon ratio may actially " lmprove with time: and with reasonable chem|c0|-?-‘. . processmg rufes because of the bu:ldup of the_li_f- superror U233 fyel. Comparlsons of - hth:um-bery”lum ond sodlum-;:r_‘i..r'i: E beryllium salts showed that the’ sod:um-berylhumf';'_; o salts’ requtred 1.1 10 2 times the U2 ¥ for computing the constants from the concentra- tions, absorption cross sections, initial spectrum, inventory .~ "ireql.nred with the lithium: berylhum salts, Atso,""';-ffi_;'fi : V,ffhe sod;um-berylllum salts show a dlsadvanfage off '_, S 0.11t0 0.15in regeneruhon rcmo. S T Mod:ficatlons were made in the borghum program-- '3__for the Oracle, which” was: desrgned for computing. - progressive changes in “core . concentration “and. . regeneration ratios. A subroutine was mcorporated“‘ ' ,,?_'.;';-._,::_Pump shah speed pressure of the pumped fluid would maintain the | _centering of the impeller in the pump casing. B 12 Component Developmenf and Tesfmg Prehmmcry pump design studies have mdicated - that five pumps with the followmg characteristics - would be’ suitable. for cnrculchng fuel 130 (BeF - - LiF-UF , 37-62-1 ‘mole %) in the moifen salt reactor bemg consndered Lo - V'Z-tFlow through pump 7 o '_'_":4590 QP"‘ ,;,;Head produ::ecl by pufip ‘ 7'! ft : Ei;ig‘;_Temperature uf pump mlet -':"'1230°F (max) _ | -Sfohc pressure af pump mlet 19. 2 psuu o o o 7..7']160 rpm e f"Pump desngns and arrongements ‘are bemg studled o ‘ ir':“ order to. determme the most favorcble arrange- - - ment ‘of cH elements w:th respect o adapiablllty e “to reactor consfruchon, durability, remote mainte- nance problems, reactor operational probiems, and other considerations, e et e e e mein MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT Facilities are being designed and constructed for development tests of the salt-lubricated bearings being considered for use in fused salt pumps. The oil-lubricated bearings being used at present are unsatisfactory because they must be shielded from both the heat and the radiation from ~the fuel salt in order to avoid damage to the bearings and to the lubricant, Equipment is to be set up for -high-temperature tests of both hydro- dynamic and hydrostatic salt-lubricated bearings. Oil-lubricated mechanical shaoft seals are being _ studied for auxiliary system applications. Modi- fications in purge-gas fiow are being tested as a means of preventing diffusion of process-fluid ~ vapor into the seal region, Also, since oil- lubricated bearings may be required for the upper seals of the fuel pumps, tests of Dowtherm A as a lubricant are under way, This lubricant decom- poses into gaseous and liquid products when irradiated, rather than into carbonaceous deposits of the type that result from the irradiation of hydrocarbon-base lubricants, A fest of a pump rotary element with oil-lubricated bearings is also under way in a radiation field. Further, a bellows- mounted seal is being operated in an endurance test in a sump pump that is circulating NaK at 1250°F, " Techniques for remote maintenance of the com- ponents within the reactor cell are being studied. 1t is considered that the most feasible solution to the maintenance problem is to make all components removable and replaceable by remote manipulation, Therefore means for remotely separating and joining pipes are being investigated. "Tests are under way of three types of flange joints: (1) a freeze-flange joint in which a frozen seal of molten salt is used to prevent leakage of molten salt thraugh the space between the flanges, (2) a joint which has a cast-metal seal between the flanges, and (3) o joint in which a V-shaped tooth on each : flange indents the sealingring. The various types of remote-handling appurafus ‘now ‘available commercially are being studied for applicability to maintenance of this reactor sys- tem, - The remote assembly and disassembly of a pump in a hot cell is being attempted as a means of studying the manipulation problems. Design work has also been initiated on heater-insulation “units that can be remotely removed aond replaced. Plans are being made for a large-scale demon- stration facility in which fo test mechanical joints, the replaceability of components, the adequacy of heater-insulation units, the unitization of wiring harness and service piping, and the application of remote-wewmg ond -hundllng apparatus ond tech- niques, A heat exchanger test facnllty is - bemg operated to obtain heat transfer correlations for predicting the heat transfer performance of the molten salis of interest, Experimental information is required because the heat transfer characteristics of some salts appear to be affected by the type of struc- tural material used and by the wettmg properhes of the salts, " Forced-circulation Joops in which large tem- ~ perature drops can be achieved are being operated to study the corrosion of INOR-8 and of Inconel by the fused salts of interest, The long-time effects are of particular interest, and satisfactory opera- tion of a loop will imply operation for one year or - longer without significant equipment difficulties or changes in operating conditions. Two special loops have been constructed of INOR-8. One of these includes graphite specimens and will be examined after operation to determine the extent to which graphite causes carburization of INOR-8 and the effects of the fused salt mixture on’ graphite, The other loop includes INOR-8 speci- mens for weight-loss studies. A forced-circulation loop is also being assembled that will be operated in a beam hole in the MTR, Operation of the loop will provide information on fuel stability and corrosion of INOR-8 under irradiation at simulated reactor conditions. 1.3, Engineering Research Three entrance-exit systems proposed for the core of the molten salt reactor are being studied in glass models., Preliminary qualitative data on flow patterns and velocity profiles have been ob- tained through visual observation and photographic recording of the motion of phosphorescent particles. The three systems being studied consist of (1) a core with the inlet and outlet diametrically oppo- site each other (straight-through flow), (2) a core with the inlet ond outlet concentric and the fluid entering through the inner pipe and exiting through the outer annulus, and (3) o core with the inlet —and outlet concentric and the fluid entering through the annulus and exiting through the inner pipe. In initial experiments with the concentric system and the fluid entering through the inner pipe, @ ¥ ¥ Hhme (M o) AL e st b v o e s .fli C primary heat exchanger, large toroidal eddy was noted in the region be- tween the high-velocity, central, downward jet and the return stream along the core walis. Such an eddy could engender discrete high-temperature masses within the main fluid flow ond thus give rise to high-frequency thermal cycling of system components. The effect of o vortex generator at the entrance will be studied. ' Experimental determinations are being made of the viscosities and thermal conductivities of several BeF ,-bearing fluoride salt mixtures. Heat transfer studies are being made in order to de- termine the effect of nonwetting and interfacial film formation on heat transfer in Inconel and INOR-8 systems., used for these studies with the salt m|xture LiF-BeF,-UF , (53-46-1 mole %). 1.4. Advunced Reactor Studies A preliminary design study was made of o natural-convection molten salt reactor which could be used in a system in which there was a premium placed on reliability and ease of maintenance. The advantages of eliminating the fuel-circulating pumps and the attendant problems of maintenance are obtained at the cost of the increased fuel volume required for a system in which the pressure losses must ke very low. A gas-turbine cycle or one of several steam cycles that operate efficiently under high-temperature conditions would be used with this system. In this study both helivm and a molten salt were considered as coolants for the A comparison of the natural-convection system with the forced-circula- tion system indicates that, for o 60-Mw . (fhermal)3"'j_" reactor,: the naturcl-conVecflon sysfem requires o =~ fuel inventory about 42% grea’rer than fhat of the*___ forced-cuculchon system. An idealized reactor ‘model -is bemg used in ak,i_»fi‘; study of the influence of vcnous_' - sys tematic - factors on' the power required to obtain a given - Data have been ob- - _ tained -for the ideuhzed model, and - further studlesr"j' - “are under way for'a model modlfled to reflect more - 'frealisnc condmons. - - e flux in a resecrch reactor, PART 2,, MATERIALS STUDIES s - 2 ‘I. Meiu”urgy ' . Cofrd'sion thermal-convection and forced-circulation loops fabricated of INOR-8 and Inconel. of a three-phase test program has been almost - reactor structural materials. A single circular tube will be - _ embrmle the strength of INOR-8.- ':Vperlods are requnred to obtmn dofa on plashc _:‘_,propertles, prellmmary data on fensule properties - are being obtained for use in desugn studies. 'Duta are presenfed for sheet spec:mens that must-. _ 'be considered as _approximate. - expenments are under way w:fh" The first phase PERIOD ENDING JANUARY 31, 1958 completed and part of the second phase of testing is under way, Only Inconel thermal-convection loops have been examined thus far, but the low corrosion rates expected in these 1000-hr tests were found, , Studies are under way for determining the effect of carburization on the mechanical properties of The sodium-graphite system was found to be a rapid and effective medium for carburizing stainless steels, Hastelloy B, and Inconel. On the other hand, Inconel exposed to the graphite—fuel 30 (NaF-ZrF -UF,, 50-46-4 “mole %) system for 100 hr at 1500 and at 1250°F in seesaw-furnace apparatus did not become carburized. Static capsule tests revealed that INOR-8 carburized more readily in sedium than Inconel, The 'INOR-8 tensile specimens to be used for strength and elongation determinations are being prepared by using the sodium-graphite .system, - The precious-metal brazing alloys, 82% Au-18% Ni and 80% Au-20% Cu, being considered for use in the fabrication of fuel-salt—to—coolant-salt heat exchangers were corrosion tested in fuel mixtures. Neither alloy showed attack after 2000 hr at 1200°F in stotic .tests and 500 hr in seesaw- furnace apparatus, Similarly, no attack was found - “on any of the welded INOR-8 plates tested in NaK and in fuel 130 in seesaw-furnace apparatus for 500 he at 1200°F. Various mckel-molybdenum- base welding rods were used., Measurements were made of the modulus™ of elasticity, the thermal conductivity, and the tensile properties of several commercml air-melted “heats - determining whefher INOR-8 has « tendency fto . in the temperature range of 1000 to -~ - 1400°F. Specimens are being aged for penods of 500, 1000 12,000, 5,000, and 10,000 hr, Pre- ||mmury results show that the specumens aged ‘of INOR-8, ~Studies were initiated for 500 hr did not become brittle.- o Tests are under way for obtaining basic data on . ~'Since relatively fong Relaxaflon fests bemg ‘made to determine whether INOR-8 wili deform plashcally under reactor operating con- ditions have indicated that the creep strength will have to be investigated. - different heats of INOR-8 from three different MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT In support of the component development program, presence of the Li® isotope in the fuel mixture metal seals for remote-disconnect flanged joints. (fuel 130) for the in-pile loop. Since the possible were investigated. A series of tests showed that effects of traces of tritium in the loop could not be silver or a_ silver-copper eutectic alloy would evaluated with confidence, it was decided to use make effective cast metal seals on flange joints the best available Li. made of stainless steel, INOR-8, and Inconel. Preparations are being made for the opercmon of Investigations of the welding characteristics of a similar loop in the ORR and for the irradiation of INOR-8 tubing were continued, and evaluation fuel 130 in INOR 8 capsules in the MTR tests were made of various weld metals. Seven sources have been studied, and only one has 2.3. ChemIStry shown cracking tendencies when welded. In these investigations, weld test plates are prepared that provide specimens for mechanical property studies of welded joints, for radiographic, metallographic, and hardness studies, and for obtaining general information on the welding characteristics of the materials under conditions of high restraint, Phase equilibrium studies are being made for determining whether an'l'.iF'-BeFi, mixture will dissolve sufficient ThF, and UF, to provide a fuel for a fused salt breeder reactor, The studies have indicated that the quaternary system LiF- BeF,-ThF,-UF, can be tredted as a temary sysfem and that some interpolations can ke made An examingtion was made of the INOR-8 forced- between the systems LiF-BeF,-UF, and LiF- circulation loop that failed during initial heating Ber'ThF.q with regard to liquidus temperatures in a test stand, The failure occurred near the and phase relationships, Breeder reactor blanket fusion line of the weld joining o Hastelloy B o breeder reactor fuel solvent compositions, nipple and an INOR-8 (Haynes heat SP-16) adapter. whose maximum ThF, concentration is limited to The lack of ductility of the Hastelloy B nipple, that available in salts having less than a 550°C which was part of a finish-machined Hastelloy B liguidus, may be chosen from an area of the phase pump barrel, was the cause of the failure, INOR-8 diagram in which the upper limits of ThF, con- pumps are now being fabricated for use in the centration are obtained in the following compo- forced-circulation loops. sitions: 75 mole % LiF~16 mole % ThF ,~9 mole % BeF,, 69.5 mole % LiF-21 mole % ThF ,-9.5 2.2. Radiation Damage mole % BeF,, 68 mole % LiF-22 mole % ThF ,~10 Preparations are being made for the operation of mole % BeF an INOR-8 thermal-convection loop in the LITR. The 501U5||'*Y and stability of PUF in beryllium- An electrically heated full-scale mockup has been containing fluoride saits are being mveshgafed assembled and filled with fuel to test components The solubility has been shown to increase with and procedures, A new type of thermocouple increasing BeF, concentration in LiF-BeF, mix- assembly has been designed for use with this tures. In the NaF-BeF, system, the solubility of loop. In mockup tests the thermocouple has PuF; is higher in the mixtures with 50 mole % survived dozens of thermal cycles in which the BeF, and 36 mole % BeF, than in the mixture thermol expansion was many times greater than with 43 mole % BeF,. Thus there is an indication that expected in the loop. The new assembly that the solubility of PuF, in this solvent goes allows the thermocouple jacket to move as the through a minimum in the vicinity of mixtures with fuel tube moves during expansion. 43 mole % BeF,. The solubility .of PuF, ot Charcoal for use in a trap for adsorbing xenon 565°C in the binary mixtures studied varied from during operation of the loop was baked in vacuum about 0.2 mole % for the 57 mole % NaF—43 mole % for 48 hr at 500°C in order to decompose organic BeF, mixture to 0.45 mole % for the 51.6 mole % impurities and was then fested with radiokrypton L|F-48 4 mole % BeF. mixture, This concentra- to determine the effect of the heat treatment on tion range would probably be adequate to fuel a its adsorptive qualities, It was found that the molten salt plutonium-burner reactor, in the one _ charcoal was 10% more effective than it was ternary mixture studied, NaF-LiF-BeF, (56-12-28 before the heat treatment, mole %), a value of 1.5 mole % PuF, was obtained Calculations were made of the magmfude of the at 565°C. Thus there is an indication that the undesirable effects that would result from the solubility of PuF, continues to increase with e %) " AN M an sy decreasing BeF, concentration. No evidence of disproportionation of PuF, has been found in these experiments. survey was made of the physu:al chemical, and nuclear properties of fused chlorides of possible interest as secondary heat transfer fluids, The survey showed the eutectic composi- tion 41.7 mole % RbC|-58.3 mole % LiCl to be the most attractive from the standpoints of vapor pressure and corrosion. Thermal-convection loop tests would be required to determine the rate of mass transfer. -Apparatus is being constructed for treating RbCI-LiCl mixtures to remove the water that is always present in the salts, The vapor pressures of LiF-BeF, mixtures are expected to be low at MSR temperatures, and, to determine the magnitude, measurements were made of the vapor pressure of the solvent mixture 64,9 mole % LiF-35,1 mole % BeF,. Since be- havior similar to that of the NaF-BeF sysfem can be expected in the LiF-BeF, system, it is antici- pated that the vapor pressure of a 70 mole % LiF-30 mole % BeF. solution will be about one- half the vapor pressure of the 64.9 mole % LiF- 35.1 mole % BeF, solution at the same tempera- ture, Studies of the solubilities of the noble gases in NaF-KF-LiF mixtures were continved. The studies of the NaF-KF-LiF mixtures were under- taken pending completion of facilities for studying solvents containing BeF,. It is expected that the numerical values of the noble gas solubilities in solvents containing BeF2 will be less than the corresponding values in NaF ZrF, ond “more. fhan “ those in NaF- KF-LlF Studies of the solublhty of HF in NoF-ZrF and ~ NaF- KF-LiF mixtures were continued, and studles‘_f - of L:F-BeF2 mixtures -were - mmated " Both the":-ii HF solubility and the heat: of soluhon vulues for - . the LiF-BeF, (51-49 mole %) mixture are - of the - . “same order of magmtude as the values obtamed . 7'__Wlfl'l correspondmg NaF- Z.rF4 mixtures, _ The solubilities. of - flssmn-producf fluorldes ino- - BeF -containing solvents ‘are being studied by % ;)UF ‘systems is descrlbed The results of actual ‘ ’iloop expenments in which’ Cr3! was used to trace © “the “chromium " migration were “used to check the f}‘valldlty of the calaulations. "usmg radioactive fracer. techmques. “Values™ are - 7"presented for the so?ubn!mes as.a funchon of'-‘.’;_f.‘i- - '_"temperature and’ composmon of CeF3 in NQ::F-BeF2 ~ YF; in NaF- BeF Cer in LiF-BeF; and CeF, - ‘ and LaF in 1he presence of each other in LiF-"“' = of nickel in nickel-molybdenum alloys by using an BeF The precipitation of oxides of flsswn products from fluoride melts is being studied as one of the PERIOD ENDING JANUARY 31, 1958 possible methods for the purification of fuel mixtures. Simple thermodynamic considerations are being applied in order to calculate the relative order of precipitation of oxides. Calculations are descrubed for the oxide precipitation of U**, Ce***, -~ and Be*t from an LiF-BeF., mixture containing UF, and CeF,. As part of this study the solu- bilities and precipitation reactions of a variety of solutes are being studied in the simple binary solvent LiF-KF. The following usefu! generali- zations are based on the information obtained thus far regarding the solubilities of oxides of uranium, zirconium, hafnium, rare earths, alkaline earths, and alkali metals in this solvent, 1. Urenium, zirconium, and hafnium precipitate as the dioxides and are the {east soluble. 2. Rare earths precipitate as R’2O3 and have - very low solubilities. - 3. Beryllium and magnesium oxides are slightly “more soluble than the rare earth oxides but are still quite insoluble, ‘ 4, Barium, strontium, calcium, potassium, sodium, and lithium oxides are more soluble than the oxides specified in items 1, 2, and 3. An experimental study of the activity coeffi- cients of CrF, in NaF-ZrF , (353-47 mole %) was continued, and equilibrium quohents obtained at ~ 850°C for the reaction CrF,(d) + H,(g) == Cr(s) + 2HF(g) are presented. Experlments are under way at ‘other temperatures. ~In.. preparation- for a determination of activity coetf:cuents ‘of Fer in LlF-BeF {63-37 mole %), it was established that ethbrlum measurements »'cun be made with solutions of FeF2 in this solvent “af concentrahons ‘well in excess of 5000 ppm “Fe**, even at 500°C which is the iowesi tempera- ture of interest. - A proposed graphlcal method evolved for calcu- Iatmg ‘the rate (_:md amount of chromium migration (corrosnon) to be expecfed in Inconel ~NoF-ZrF - Afiempts are -being made to measure the actw:ty electromotive-force method, ~ Experimental diffi- culties are being studied. ' e T 1 - 3 b " B "} Part 1 REACTOR DESIGN STUDIES A o €y ¥y +) REACTOR DESIGN H G MacPherson STUDY LAYOUTS OF REACTOR SYSTEM E. J. Breeding | J. Y. Estabrook -~ W. S. Hcms R. E. Helms Prellmmary layouts of the molten-salt reactor (MSR) system are being prepared as a means of - studying the mechanicel design. These layouts are based on the use of five fuel pumps and two blanket-salt pumps. The total thermal power. production is assumed to be 638 Mw, with 90% of the power being genergted in the 8-ft-dia core and 10% being generated in the 2-ft-thick blanket, Each of the layouts, presented here as Figs. 1.1.1 through 1.1.5, has both satisfactory and unsatis- factory features. - It is hoped that through examining many possible arrangements the most simple de- sign can be determined. Layouts of the reactor cell are also being made in order to determine the best possible arrangements of piping and other components that will provide minimum plant size and yet facilitate remote maintenance, | The basic layout shown in Fig. 1.1.1 of a two- pass reactor with a blanket was conceived with the support and fabrication problems as the major influence. The main support ring is, in effect, a channel rolled into @ ring, Both the blanket and core shells are supported by this ring. The fuel pumps are supported by their barrels, which pene- trate the top flange. Both the fuel and the blanket- salt expansion tanks are formed by conventional shell-head shapes. Individual pump suct:on hnes direct to the neck of the reacfor core vessel are prowded to avoid stresses ‘induced by penetrchon. of the biankei shell The ‘method of support of - the pumps - permits . the pump discharges to be - rotated without altering the details of. fabrlcuhon, o of the vessel, Except for the outrigged. blanket-_ e salt pumps,_the unit is symmetrical in order to - - -_\_s|mpltfy fabrication. All parts ‘of the fuel systemi U " with blanket salt, and yet simple shapes have been retained for ease of fabrication. The fuel pumps are equally spaced around ‘the periphery of the reactor, -and the two blonket-salt pumps are located in a central annulus ‘of blanket over the core. This arrangement permits the support of the pumps gnd the blanket pressure shell from a grid type of UNCLASSIFIED ORNL—LR—DWG 27883 FUEL BLANKET-SALT PUMP EXPANSION TANK ~BLANKET-SALT ~ OUTLET are in packages of five in order to lower the cost = -of fabrication. The fuel lnventory of .a reactor of LY ~this ‘type would _be low, that is, ‘approximately = .- 330 ft3 One of lhe main’ ob|echons to this iayoutf S “is that there isno blanket sclt around the neck of .~ l the core vessel, The layout presented in Flg. 1.1.2 has the . - advantage that the core is completely surrounded - ~-BLANKET-SALT INLET Fige" _'l.i.-l. MSR Layout with Outrigged Blanket-Salt Pumps ond No_'Bl_t:fikei Salt Around Neck of Core. MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT UNCLASSIFIED ORNL-LR-DWG 27884 FUEL — QUTLET FUEL INLET —-B —-w= BLANKET-SALY QUTLET BLANKET-SALT INLET —p-=— Fig. l.'l.-2. MSR Layout with Fuel Pumps Equally Spaced Around the Periphery aond the Blanket=Salt Pumps in a Central Annulus of Blanket over the Core. frame above the reactor. The fuel pumps are connected by a gas annulus or torus with space for expansion of the fuel under conditions other than normal. The pumps ore completely covered with fuel at all times, The fuel inventory for this layout is 395 ft3 and could be reduced, since the connecting pipes out of the core and the pump tanks appear to be lorger than necessary. The fuel inlet and outlet pipes pass through the blanket pressure vessel wall and may create undesirable stresses. The temperature difference between the inlet and outlet pipes is expected to be about 135°F. Further study will be required to determine whether the resulting stresses can be tolerated. The layout shown in Fig. 1.1.3 has the five fuel pumps ‘equally spaced around the top and the two 10 UNCLASSIFIED ORNL-LR—DWG 27885 BLANKET-SALT EXPANSION TANK — T FUEL INLET FUEL OUTLET =S / // CORE \ [l o% | | 2% \_j b BLANKE ~-_ | " - L \\ ~~—3- BLANKET-SALT INLET Fig. 1,1.3. MSR Layout with Five Fuel Pumps That Have a Common Header and Offset BlankesSalt Pumps That Draw from a Blonket Chamber Above the Fuel Expansion Tank. blanket-salt pumps offset in a blanket chamber above the fuel expansion tank. The five fuel pumps have a common header, The suction, volute, and discharge of each pump are located in a trough to minimize inventory and to direct the flow. The fuel -returns through the fuel expansion tank, gathers in a plenum, and is then directed to the core. The fuel pumps are supported by their barrels from the fuel expansion tank top. The blanket salt flows upword around the core and passes between the fuel pumps in rectangular ¥} i m e >y " W UNCLASSIFIED ORNL-LR-DWG 27886 L ANKET-SALT PUMP COOLANT OUTLET SALT OUTLET Fig. l.1.4. Single-Pass MSR L.ayout. ducts to a plenum at the center of the blanket deck from which the offset blanket-salt pumps draw.. The blanket-salt: pumps .are suppor'red by;_-__ their barrels from #he blanket expansion tank top.'}_ '_fprobably be required. - fuel. pumps. The fuei mventory for thls Iayout lsf : ’approxsmotely 440 f2, Sl e . One of several prel:mmary onouts of smgie~pass | 'r'fsystems is shown in Fig. 1.1.4. The single-pass - - system” offers the posmblhty of closely ‘coupling * the primary- heat exchangers to the reqctor.‘ Such - ~close couplmg would be of advantage in mmlmlzmg " the size of the reactor ceil. ~Thermal sleeves are _mcorporated to relieve - fhe sfresses | mvolved in - connecting the outer blanket shell to the inlet und__'_"_i, outlet piping. “The top of the reactor is somewhut;- The reactor. supporfs ¢could be located ‘under the simpler than fora two-poss system,- uithough some of the complication was merely transferred to the. bottom of the reactor. This layout required a fuel \BLANKET- PERIOD ENDING JANUARY 31, 1958 UNCLASSIFIED ORNL-LR-DWG 27887 BLANKET-SALT PUMP . BLANKET~SALT . OUTLET FUEL QUTLET FUEL INLET BLANKET- SALT INLET Fige. L1,5 MSR Loyout with Fuel Return Ducts at the Same Place as the Fuel Discharge Ducts, mvenfory of 420 ft3, The reacfor support problem was not consldered in these studies; however, a toad ring ‘and pump bcrrel type of support would "Another two-pass system is shown in Fag. 1 l 5 ' The return fuel ducts are. brought back -to the top _of the reactor at_the same place 'as ‘the pump dis- charge ducts, (Results of flow studies of this ,;}’arrangement are _presented in. Chap 13) ~This ~feature tends to. Iower the over-all helght and ’,hence the pressure - dlfference across the core _shell, The plan view shows much fess blanket '}_materlal in the - reglon of the dlscharge and return’ ducts. than there is in desngns having the return - ducts - above ot below the discharge ducts. As in all the ‘two- -pass” layouts with multiple pumps, the top of the reactor is complex wn‘h regard to fabri- cab:llty._ 11 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT 'NUCLEAR CALCULATIONS L. G. Alexander J. T. Roberts Nuclear calculations relative to parameter studies of molten salt reactors for the production of power were continued, The Univac calculations, for which the Ocusol program is used, have provided information about U235. and U?33-fueled reactors of the reference design type, primarily, but studies have been initiated of plutonium-fueled molten ‘salt reactors and of graphite-moderated molten salt y233 breeder reactors. 7Unrivc‘lc-0cuso| Reactor Calculations J. T. Roberts The parameter study of two-region, homogeneous, molten salt reactors described in the previous report! has been extended and refined. The calcu- lations based on the use of the 69 mole % LiF-=31 mole % BeF, salt plus ThF4 and UF , in the core and of the 75 mole % LiF-25 mole % ThF, salt in the blanket were completed, and a new study based on the use of a 63 mole % LiF~37 mole % BeF, salt plus ThF, and UF, in the core and of a 71 mole % LiF-16 mole % BeF,~13mole % ThF, salt in the blanket was initiated. The effects of changing the salts are expected to be a small reduction in critical mass and a reduction of approximately 12% in the external regeneration ratio. Blanket thickness will also be a variable in the new parameter study. In the calculations just completed a comparison was made of U233- and U?35-fueled reactors. The results of these calculations are presented in Fig. 1.1.6, which shows the fissionable material in- ventory at startup as a function of core diameter. The effect of various amounts of thorium, in the “range 0 to 1 mole % in the fuel salt, is also shown. For the some core diameter and thorium content of the fuel salt, the U233 inventory for the cases shown is less than half the U?3% inventory., This is due to both the higher ¢, and the lower a for y233 compared with those for U233 in intermediate reactors, In Figs. 1.1.7 and 1.1.8, the 5-1, the initial total regenercmon ratio, and the initial blanket regeneration ratio for U235. and UZ33. fueled reactors are plotted as functions of thorium content of the fuel salt for core diameters in the L. G. Alexander and J. T. Roberts, MSR Quar. Prog. Rep. Oct. 31, 1957, ORNL-2431, p 5. 12 - UNCLASSIFIED ORNL—-LR-DWG 27888 5000 —— 1T SALT HOLDUP OUTSIDE CORE ASSUMED TO BE 339 #3 - FUEL SALT: 69mole % LiF -3 mole % BeF, + ThE, + UF, BLANKET SALT: 75mole % LiF~25mole % ThE, oo LN R 8 A 1N I 1 AN : 4 \\ R s U235, {mole % Th |— IN FUEL SALT N N CRITICAL INVENTORY (kg of U3 o y23%) 200 ‘L\\\ \\ \: . U235, No Th "k‘ ' U233, 4mole % Th —]—_AU?35 025 mole % Th N\ N \ - L \ T U235, 0.5mole % Th 100 U233 NO Th 50 3 4 5 6 7 g 9 {0 CORE DIAMETER{ft) Fig. 1.1.& Clean, Critical Inventory of U233 gnd ' U235 {or 600-Mw Reactors Fueled with Lithiume Beryllium-Base Sclts. range 5to 10 ft. In the U235 cases the regenera- tion ratio is limited to a maximum of about 0.675 by the comparatively low value of 5 and a practical minimum loss to capture in the salt and in the core shell of about 0,1+, The U233 cases were not carried for enough to establish a *‘ceiling’’ on the regeneration ratio, but, judging from the n—1values, it would seem to be possible to exceed 1.0 by increasing the thorium content of the fuel salt to, perhaps, 1.5 to 2% and, at the same time, in- creasing the U233 inventory to, perhaps, 300 to 900 kg, depending on the core diameter. It should be remembered, however, that the regeneration ratios shown for the U233-fueled reactors will decrease with time as burnup poisons accumulate in the system, while in the U235.fueled reactors, with initial regeneration ratios of about 0.6 or better, the regeneration ratio may actually improve with time and with reasonable chemical- process'ang rates because of the buildup of the superior U%33 fuel, Calculations were also made in order to compare lithium-beryllium ond sodium-beryllium salts as fuels for reactors of the reference design type.: Figure 1.1.9 presents a comparison of the y23s inventories at startup for reactors fueled with U233, 0.25mole % Th ~—1- L3 0 F. P P ) ‘ UNCLASSIFIED 'ORNL-LR-DWG 27889 T T T T FUEL SALT: 69 mole % LiF-31 mole % BeF,+ T, + UF, 41" BLANKET SALT: 75 mole % LiF 25 mole % ThF, 1.0 . \ CORE 0.9 DIAMETER - \ \ 10 ft1 N & 0.8 ™ : = x \ N (" © r— 5 ft o , G ft L 07 o —1_ ot TOTAL z S| == 81t Y REGENERATION S ' ft 2 os ?__‘e < et 2 1| 1 RATIO & 6 / % N/ A g o0s % |1 04 Ny —{5+# . \ T 6 ft| |p o3 . BLANKET | ' I REGENERATION ~——_ ~——_ RATIO ~—_ 8 ft 04 0 Yq s 3 4 THORIUM IN FUEL SALT (mole %) Fige 1.1.7. _Initial Regeneration Characteristics of Molten Salt Reactors Fueled with thhlum-Beryllium- Base Salts Containing U235, 69 mole %" L|F—3I mole % BeF plus ThF and,' UF in- ‘the core and 75 mole % LiF-25 mole % - ThF, in the blanket and for reactors fueled wnh 53 mole % NaF=47 mole % BeF, plus’ ThF and ,'_-{UF in the core and .58 mole % ‘NaF-35 mole % ' L BeF —7 mole % ThF in the bIanket. For-the same’ - core’ dsameter and thoruum confent of the fuel salt L - the sodlum-beryllwm-base -salts requnred Llto. 2 times the U?3 mventory requured for the lithium- beryllium-base salts. In Fig. 1.1.10 the 7-1,the - - initial - total regenerahon rcmo, and - the mmql' ~ -blanket regenerunon ratio are ‘plotted for: reuctors' ' fueled with Na-Be salts, - In' comparison’ ‘with the ‘Li-Be-base " fuels (Fig. 1. 1.7), “the ‘Na- Be*base'fl_,g. fuels show a disadvantage of 0.1 to 0.15 in regeneration ratio. Since Na-Be salts cost ap- proximately one-half as much as Li 7.Be salts, _ concentmt:on of mute, together with the PERIOD ENDING JANUARY 31, 1958 , UNCLASSIFIED : . ORNL—LR—DWG 27830 1.3 | | _ o 0.25mole % Th w2 b -NOTh - Pl _ ]1;— . [ m—— imole % Th £ 10 _ 7 \ = imole % Th g \ \ TOTAL REGENERATION RATIO £ a h = o8 N _ ! g p \ \ _—025mole % Th = N\ 25mole Y% \ - h 567 \ - NS TOTAL REGENERATION __| § N, | ToTAL T 086 , L REGENERATION RATIO N (ALSO BLANKET : ™. REGENERATION RATIO) 0.5 h —y i \\ [ S /—0 25mole % Th 0.4 L BLANKET ’ <. REGENERATION | - I~ RATIO 03 tmole ToTh—Sw_ | T~y : BLANKET REGENERATION RATIO S Y ] ) 0.2 v 3 4 5 6 7 8 a 10 CORE DIAMETER (ft) Fig. 1.1.8. Initia! Regeneration Charocteristics of Molten Salt Reactors Fueled with Lithium-Beryilivne Base Salts Containing u233, some further studies of them will be made, even though they do not compare very favorably on a nucleor basm. A comparlson of flux-averaged cfoss sections of U235 U236, and U238 in reactors of the reference - deSIgn type . indicated that - at steady state- the Y23’ plus U238 will- be about 60 to 90% of the U235 concentration. This esti- U233 critical concentration “estimates, indicates that the use of 1 mole % UF in corrosion tests is conservuhvely high if data cppllcabie to the reactors of greatest interest are " to_be obtained. The data may not, however, be appllcable to. grophnte-moderated thermal reactors, since the steudy-state ratios of even to odd uranium “isotopes will be greater in thermal thon in mter- medsate reactors. o Cross sections for Ng, Be, Pu239 Pu‘uo C, ond Bi have been added to the Ocusol sigma tape (for 13 //‘ MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT UNCLASSIFIED - ORNL-LR~-DWG 2789 5000 " T T T T 1 T T T Na-Be FUEL SALT: 53 mole % NaF-47 mole 7 BeFy — PLUS ThF4 + UF4 - No-Be BLANKET SALT: 58 mole % NaF-35mole 7 BeF,- | 7mole % ThF, Li~Be SALT: 69mole % LiF-31mole 7 BeFy PLUS ThFg +UFy . Ll-Be BLANKET SALT: 75 mole % LiF-25mole 7o ThF, 2000 |— SALT HOLDUP OUTSIDE CORE ASSUMED TO BE ' 339 113 y o s o ~ * 0\ B : NN~ - y No-Be, {mole % Th £\ N i L 1000 - o IO R -:-‘: \ \ \\ \ \m\ S \ 1\ N g Li~Be, imole % Th — > AV N > \ . . N\ = : o 500 [—e- \ N - \ \ T T~—g¢ Li-Be, 0.75mole % Th VNN ] | \ | ¢ Li-Be, 05 mole % Th ' | | L y No-Be, NO Th RN~ 200 1 | : 9 \“ . I ! ’ \ ~ & Li-Be,0.25 rpole % Th \u.. b Li-Be, NO Th 100 . : 4 5 6 7 8 9 {0 CORE DIAMETER ({f1) Fige. 1.1.9. Clean, Critical Inventory of U235 for 600-Mw Reactors Fueled with Lithlum-BerylHum- or with’ Sodium-Beryllium-Buse Salts. cross sections available previously see ref 2), Also, thorium cross sections with o adjusted for resonance saturation effects have been inserted for 0.25, 0.5, 0.75, 7, and 13 mole % thorium in " lithium-beryllium-base salts. Pseudo cross sec- - tions “for combined- Li-Be-F in various core and blanket salts have been inserted so that these may be used as a single ‘‘element” in Ocusol ' culculatlons, whlch are limited to seven elements in g reglon. ' dmcle.talcul-utions : L. G. Alexander " The Sorghum program was designed for computing progressive ‘chonges in core concentrations and = 2J. T. Roberts and L. G. Alexander, Cross Sections for OCUSOL-A Program, ORNL CF+57-6-5{June 11, 1957). 14 UNCLASSIFIED ORNL~LR-DWG 27892 -2 T T T T 1 | FUEL SALT: 53 mole % NoF -47mole % BeF, + ThF, +UF4 1.| |—— BLANKET SALT: 58 mole % NaF - 35mole%Ber 7moie% ThF, — - 1.0 THORIUM IN FUEL SALT | _A—qnoTh 0.9 — _ u/ : 2= o8 tmole% Th b tmole — : o q REGENERATION RATIO OR 7 -1 o [ d 5 et tmole % Th] 0.4 TOTAL b REGENERAT!ON - , 0.3 V\ RATIO 02 ' ‘\\ S —tro T~ [ fhom BLANKET O.f [~~——g {mole % Th 1| REGENERATION ___| l , RATIO l 0 S 6 7 8 9 10 CORE DIAMETER (ft) Figs, 1.1.10. Initicl Regeneration Characteristics of Molten Salt Reactors Fueled with Sodium-Beryllium- Base Salts Containing u23s, regeneration ratios in molten salt reactors, with Ocusol or Cornpone outputs being used for the initial states, It is based on the assumption that the scattering and leckage probabilities are in- sensitive functions of the varying concentrations. The program wos operated with constants that characterized the spectrum, leckage, and scattering in Ocusol Case 89-2, for which the data were pre- sented previously.® The code operated satis- factorily but disclosed the hitherto unrecognized fact that the spectrum and leakage output from Ocusol are not consistent with the scattering properties. The source of the inconsistency is not known, but may be due to a biased round-off of the leakages at the highest energies, which, being small, are known to only two significant figures. Since the spectrum in Sorghum is computed in a stepwise manner, beginning with the high-energy groups, the round-off bias tends to accumulate and to be amplified in the lower groups, 3L. G. Alexander and J. T. Roberts, MSR Quar. Prog. Rep. Oct. 31, 1957, ORNL-2431, p 6, Table 1.2, O v . 1 14 3 - Y “GM-GNU,” & In order to remedy this defect, a subroutine for computing the constants for Sorghum from the con- centrations, ubsorphon cross sechons, initial spectrum, and jeakage probabilities was written and incorporated in the program. The resulting constants, which differ only slightly from those based on the scattering cross sections computed by Ocusol, reproduce the initial spectrum and critical concenfrahon exacfly. Absorphon cross sections for Pa233, U234, U236 Np237, and Pu?3? were estimated and were inserted into the program, together with fission cross sections for Pu?3?, The modafled program has been operofed successfully on a test case. Tests to determine limitations imposed on the magnitude of the time ancrement by convergence requirements and the drift in group leakage proba- bilities are in progress. Efforts to bring the Oracle program Cornpone into use for molten salt reactors have been con- tinved. Initial tests, for which the Ocusol group structure and cross sections were used, showed disagreement in the critical mass; also, the com- puting time was excessive. The effects of modify- ing the group structure, scaling factors, number of space points in thin-shell regions, and method of treating the principal scatterers are being studied, A decision to normalize the fractional absorption of neutrons in each elemental species with respect to the reactor as a whole, rather than with respect to the various regions, has necessitated revisions of the edit subroutine. Provision has been made to edrt the mtegrated flux -in each group and each L region for use in Sorghum and other per?urbchon o calculafrons. IBM-704 Codes and the *‘Curtiss-Wright 075," o two-dnmensnonui multigroup, multiregion code. - PERIOD ENDING JANUARY 31, 1958 FUEL FILL-AND-DRAIN SYSTEM DESIGN G. D. Whitman A preliminary design study has been made of a fuel fill-and-drain system for the MSR. This system must meet the following three major design criteria; 1. A preheating system must be provided that is capable of maintaining the drain vesse! and its connecting plumbing at 1200°F, 2. A reliable heat-removal system must be pro- vided that has sufficient capacity to handle the fuel afterheat, 3. The drain vessel must be ‘‘ever-safe’’ so that criticality cannot be achieved when the fuel is drained. : The fuel draining operation has not been con- sidered as an emergency procedure which must be accomplished in a relatively short period of time in order to prevent a catastrophe, |t is considered that, if all forced circulation ceased during. high- power operation, the thermal capacity and heat losses of the system would prevent prompt thermal transients that would be capable of causing failure of the primary containment vessels. There are, however, other incentives for rapid removal of the fluid from the fuel circuit. [f, for example, there was a leak in the fuel system it would be _important to drain the fuel in order to minimize the cell contamination and cross contamination of ‘systems. Further, rapid removal of the fuel at the time of a shutdown for maintenance would have an ‘economic advantage in reducing the power outage 'hme. A consnderahon of these factors mdlcated that e the maximum afterheat design load should be 10 Mw " - for ' a 600-Mw reactor that had been operating. 'for one year- und had been shut down' for 10 min e ~ before the fuel was drained. No credit was taken L D’ Baxter ST e g f'lssmn-gus removal during operation, - It was The appllcablhfy of existing: lBM-704 codes 10 il ' molten salt reactor calculahons was - mveshgated ‘and three promising codes were selected for further -~ ~study. These are: “the * CURE,”l a general:zed o o ;,two-dlmensnoncl : mulhgroup program “suitable for - ff_mvesngatmg the effect of ‘the north hecd of tfie-:;:_:" - reactor on’ H1e reachvaty und breedmg ratio; the-;*-;l “multigroup, . muihregton, -one- dimensional program “similar to Eyewash-Ocusel; - determined that 15 min would be reqmred to remove f.'the fuel from the reactor. -~ » _ “For the drain vessel design calculuhons, it was 31-:\,ussumed that af 3200°F the fuel system volume "~ would be 600 #3,~ “The design: cupacnty of the ‘drain vessel was’ therefore set at 750 #3 in order - to ‘G“OW for jemperature excursuons and a residual ‘heel. An array of 12-in.-dia pipes was selected as ‘the primary containment vessel of the drain system in order to obtain a large surface-area-to-volume ratio for heat transfer efficiency and to provide a 15 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT large amount of nuclear poison material, Forty- eight 20-ft lengths of 12-in.-dia pipe would be arronged in six vertical banks connected on opposite ends with mitered joints. The six banks of pipe would be connected at the bottom with a common drain line that would connect with the fuel system, The drain system would be preheated and maintained ot the desired temperature with electric heaters installed in small-diameter pipes located axially inside the 12-in.-dia pipes. These bayonet-type heaters could all be removed or in- stalled from one face of the pipe array to facilitate maintenance. The entire system would be installed in an msuiated room or furnace to minimize heat losses. The removal of the fuel afterheat would be accomplished by filling boiler tubes installed be- tween the 12-in,-dia fuel-contdining pipes with cold water from a header that would normally be full. The boiler tubes would normally be dry and at the ambient temperature of about 1200°F, Cooling would be accomplished by radiant heat transfer from the fuel-containing pipes to the low- pressure, low-temperature water in the boiler tubes. For the peak afterheat load, about 150 gpm of water would be required to supply the boiler tubes, This fill-and-drain system design satisfies the design criteria in that it is always in a standby condition in which it is immediotely available for drainage of the fuel, it can adequately handle the fuel afterheat, and it provides double containment of the fuel. The heat-removal scheme is essen- tially self-regulating in that the amount of heat removal is determined by the rate of heat transfer. Both the water and the fuel systems are at low pressure, and a double failure would be required “for the two flvids to be mixed. The drain system tank could be enclosed and sealed from the atmos- phere because there are no large gas-cooling ducts or other major external systems connected to it, A stainless steel tray would be placed below each bank of pipes to catch the fuel if a leak - developed. These trays would be cooled by water walls to prevent any possibility of meltdown and destruction of the cell, DESIGN OF A HYDROSTATIC BEARING FOR A - MOLTEN SALT PUMP IMPELLER | B. W. Kmyon A hydrostatic bearing was investigated for use as a pilot bearing for a molten salt pump because 16 the pressure of the pumped fluid could be used fo maintain the centering of the impeller in the pump casing. Two designs are possible: one in which the pockets are located around the .inside of the bearing ond one-in which. the pockets are on the impeller. The latter design offers the advantage that the supply passages for fluid flow to the pockets are in the impeller and are thus replace- able. ' ' The chief difference between the conventional _ hydrostatic bearing with stationary pockets and the bearing in which the pockets rotate on the impeller is that in the latter case the fluid flow through a given pocket is variable except when the impeller is centered, which only occurs under the improbable situation that there is no side-load on the shaft. The variation in flow velocity, v, of the fluid leaving the pockets of the two dessgns is shown-in Fig. L.1.11,, The pressure in the pocket achng between the shaft and the journal is proportional to v2 if the pockets are stationary, but the proportionality must be adjusted by an acceleration factor if the pockets are rotating. The curves of Fig. 1.1.11 indicate that the load-supporting ability of the two designs UNCLASSIFIED . ORNL=—LR—DWG 27893 1.2 : o X 10 7N 2 7N © ROTATING J/ v STATIONARY S POCKET—7/ \\ POCKET 4 08 ! N\ ' a IMPELLER 4 \ 3 CLEARANCE {/; \ o CLOSING / ) L A \ o A “ & 06 ff IMPELLER S i/ CLEARANCE o y OPENING > y / E 04 - X 1 . w o \\ 0.2 0 60 120 180 240 - 300 360 ANGULAR POSITION OF POCKET {deg) Fig. 1.L11. Comparison of Velocities of Fluid Leaving Pockets of Stationary and Rotating Hydrostatic Bearing Systems, r) 1) i $ .’\ o is nearly the same, but there is a sllght angular displacement between the curves. - Preliminary calculations for a 16-in. - Z g& x . ,%m 200 — of : / et | . | o Sz @ o W 100 / .0 - o 0005 0.0410 0.015 ' IMPELLER DISPLACEMENT (ln) Flg. LL12, Becring Ccpuchy vs Displucemenf of Impeller Having 12 Packets Around Periphery and Opercflng at 1175 pm. 17 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT - 1.2. COMPONENT DEVELOPMENT AND TESTING . Ho w. chage , FUEL PUMP DESIGN AND DEVELOPMENT W. F. Boudreau A. G, Grindell Pump Design Studies W. G. Cobb ‘M. E. Lackey The optimum design of a fuel pump for a circulating-fuel reactor is, of course, dependent on the general configuration of the reactor and its major components. In turn, the design of the reactor can be substantially influenced by the pump characteristics. This. mutual interdependence dictates that pump design must be continuously coordinated with the general reactor design. One of the major considerations in the design of a centrifugal pump is the problem of cavitation, The onset of cavitotion is primarily governed by the net total pressure available at the impeller inlet to lift the salt into the impeller. Cavitation must be avoided, since it can cause physical damage to the impeller or to other parts of the pump. - ‘Cavitation data obtained for fused salt pumps at ORNL have been studied in relation to pressures, flow rates, and pressure drops in the heat exchangers and in other portions of the primary fuel circuit of the reference-design reactor, These preliminary studies indicate that five pumps with the following design parameters will be suitable for circulating fuel 130 (BeF,-LiF-UF,, 37-62-1 - mole %): Flow through pump 4950 gpm Heqd prcduced_ by pump 71 ft Tempe’rature at pump inlet 1230°F ‘(max) : Static pressure ot pump inlet 19.2 psia Pump shaft speed ' 1160 rpm o W. Bo .MC,DOHOH . A pump design is being prepared that is based on these design values, but it is realized that further studies of the reactor design may require changes. It will not be difficult to translate the resulfs of these studies to pumps of other ccpacmes. The bdsic" design of an impeller and a double volute for the pump wos developed that is serving as the basis for a series of pump layouts. Various types and arrangements of bearings (salt-lubricated, oil-lubricated, and gas-lubricated) and seals are being considered in conjunction with at least two different motor arrangements (totally enclosed and ‘conventional). These layouts will permit a de- termination of the most favorable arrangement of | all elements from the standpoints of adaptability to ‘reactor construction, durability, adaptability fo remote maintenance operations, reactor operutlonal problems, and so on, A number of commercial organizations have built pumps for use with liquid metals under service conditions which approach those of the molten salt reactor. In order to utilize their experience, study contracts are to be negotiated with several - manufacturers. These companies will be asked to study the fuel pump application in the fight of their own experience and to formulate conceptual designs that incorporate their recommendations. Preliminary discussions have been held with representatives of Allis FLOW METER 3 : COLD TRAP | - ( J.._._ i ‘ . ' \/ NaK FLOW SALT FLOW Ifi THERMOCOUPLES | I{I — QPRESSURE—MEASUQING DEVlces'/_) — - ~ NoK - . o SALT - SUMP TANK SUMP TANK Fig. 1.2.11. Schematic Flow Diagram m‘ Small Heat Exchanger Test Stand C. o L¥0dI Y $53490dd WYII0Id ¥OLIVIY LIVS-NILTOW 6¢C 107in. {REF} MATERIAL: INCONEL . : S 26 TUBES (0.1875-in. OD, 0.025~In. WALLS} FUSED SALT OUTLET FUSED SALT INLET SECTION A-A FUSED SALT HEADER HEAT EXCHANGER TUBES | NoK HEADER— 'HEAT EXCHANGER SHELL SEfS COMB SPACERS . SPACER WIRE 0.03% x 0.055 in. UNCLASSIFIED ORNL-LR-DWG 15289R ISOMETRIC OF HEAT EXCHANGER _ NaK INLET | .,‘Fi;q..’ 'l‘.2‘.'l2. Diogffim. of Small Heat Exchanger of the Type Being Used To Obtain Fused Salt Heat Transfer Data. 8541 ‘L€ AYVNANYF 9NIGN3 Q01¥3d i j 1 i i | I I i 1 i | [ i { | i 30 RESISTANCE _ " HEATER Fig. 1.2.14. Piping Arrdnigemerfl. of Sma "Nak-TO-AIR RADIATOR 3 H Hedf'Exchu.nger Tést Stand :C.' PHOTO 24914 UNCLASSIFIED - L] & S UNCLASSIFIED & éfiNL—LR DWG 27904 NoK-TO-AIR | RADIATOR NaK PUMP NaK FLOW, 18,600 Ib/hr . Rttt ——————————————r H52°F 944 °F 36.0 psig FUSED-SALT-TO-NoK 106.0psig 1246°F |- HEAT EXCHANGER 06 °F 94.2 psu; 20.25 psig o FUSED SALT FLOW 15,900 Ib/hr \ FUSED SALT PUMP RESISTANCE HEATER - 280 kw Fig.fl 1. 2.15 Block Diugrutn Showing Typical Oper; ating Conditions for @ Hent Transfer Run in Test Stond SHE-C. without significant equipment difficulties or changes in operdting conditions. The scope of the planned tests requires at least 15 test facilities, nine of which were available at the start of the quarter and two of which became available during the period. The remaining loops are being constructed. The operating experience thus far has shown the need for improvements in test stand rellobthty, : and therefore more precise methods for inspecting, maintaining, and assuring continuity of operation fnconel and three INOR alloy loops are c:rculatmg - salt mlxtures, and one lnconel ‘bifluid- loop ‘s in. - operation - that has separqte sodium “and salt _ circuits connected through a U-tube heat exchanger.' : AH the loops - ‘ceased to operate and. the salts. “froze during an_ unanhctpcted unprotected power - _ plates hold’ the tods in place and distribute the flow. - The - ,5-m. rods are. Separated from each other. by wire ‘spacers wound ‘around them at 3-|n._; o lmtervais. The rods ‘were corefully weighed after - - _the 'spacers were . mstaHed ‘and their- posmons outage, and two Inconel Ioops, mcludmg ‘the = ~ bifluid-loop, failed as a result, - For the INOR-8_ _ Ioops fabrlcated ‘to dute, Husteiloy B pumps -have ~ been used because. INOR-§- “pumps were not - . -,ovallable., “Three of these. iNOR 8- loops - hove:..ij_: | * failed in the sections made of Hastelloy B, which' ~-w does not weld readily and becomes brittle, INOR-8 pumps will be used in future loops. (Discussions - of solutions to fabrication problems are presented ‘in Chap. 2.1 of thls report.) PERIOD ENDING JANUARY 31, 1958 The ovatlablltty of good quality INOR-8 is improving, 'and two special INOR-8 loops have been constructed. One of these includes graphite specnmens and the other includes specml INOR-8 specimens for weight loss studies. In the weight loss studies an effort will be made to determine accurately the corrosion rate of INOR-8 at a wall temperature of 1300°F, when exposed to a fused salt, by carefully measuring and weighing the inserts. Provisions will be - made to assure that the specimens are not exposed to the atmosphere before and after exposure. Three samples will be exposed, each for a different time, in an effort to establish a corrosion rate which may be extrapolated to operating periods of severol years. The loop for the weight loss studies was designed, as shown in Fig. 1.2.16, to incorporate the three samples at the hot end of the heater section with a separate power source to control the wall temperatures. The samples were machined to fit inside a sleeve. The samples.and the sleeve ~were fused together at the ends when welded into the loop piping so that current would pass through the samples. The loop which contains graphite specimens will be examined after operation to determine the _extent to which graphite causes carburization of INOR-8 aond the effects of the fused salt on graphite,. Specimens of impervious graphite ‘especially prepared by the National Carbon - Company are being used.” The graphite-specimen assembly is shown in Fig. 1.2.17. The graphite container was designed to give a graphite-to- of test stand components are being provnded. Four - INOR-8- surface area ratio of 0.67. It has been installed at the ‘outlet of the heater section of a stqndard forced-cnrculatlon “foop - fubr;ccted of /-ln.-dla, _ 0.035-|n.-wa|| INOR-8 tubing. "The ‘ gruphlte rods are 11 in.. long, 32 of them are 4 in. Cin dlameter and 24 are / in, in diamefer. At both “ends ~of the. contamer, retainer and baffle were noted before the box - was sealed. " The bifluid loop that was placed in. operat:on in "_':March 1957 had accumulated a total of 6673 hr of . | operation at the specified temperatures and flow rates when operation of the loop was terminated 31 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT . Salts. by the power failure des'cribed above. Freezing of the salt ‘in the loop caused a section of the first heater leg to rupture, This damaged tubing section and the pump and pump motor have been replaced. The salt circuit of the loop was refilled with a new charge of fuel 122, but the original sodium charge’ is being used. - Operation was resumed on January 13, 1958. (The results of examinations of samples removed from the damaged portion of this loop are reported in Chap. 2.1 of this report.) The loops that have been operated or are now' operating are listed in Table 1.2.2, which also gives operating data. A maximum pump" speed of 3000 rpm has been established on the basis of test - ‘experlence in-order to ensure long Ilfe. At this speed, full turbulence in the fluid is difficult to attain, however. The Reynolds numbers are low 32 UNCLASSIFIED ~ORNL-~LR-DWG 279302 3g-in., SCHED -40 INOR-8 SAMPLE INSERT Fig 1.2. '|6 Dlogrum of Forced- Circuluhon Loop for Weighi Loss Studies of INOR-8 Exposed to Circulcfing Fused and vary widely for different flunds becouse of dlfferences in viscosities. In-Pile Loops Do Bo Tl'cwger : . : J. A. Conlin - P. A. Gnadt - Authorization has been requested for wradlatidn of a forced-circulation loop: m the MTR, The loop . is to consist of a hairpin of é-m., sched-40 pipe with addlhonal coils in the nose end near the reactor face, a miniature pump to circulate the ~ fuel, and ‘a salt-to-air heat exchanger. Shielding, hux:hary cooling, pump drive mechanisms, ‘and " instrumentation will also be ‘provided. Oper- - ation of the loop will provide - information on. fuel sfablllty and on the corrosion of INOR 8 under £e . i r? iy s n‘ o o ‘Table 1.2.2. Forced-Circulation Loop Operations Summary as of January 31, 1958 ; ST Couifio3|t|5n oo Maximum Minimum Maximum Hours of Loop " Loop,. . Number of F'low Rate Approximate Wail Fluid Fluid Operation ot c Designation - Mmflol and Snze Clrculmd - {gpm) - Reynolds Number Temperature Temperature Temperature = Conditions omments aa . - Flmd* o (°F) (°F) (°F) Given 9344-1 lnt:ldi;lelg ‘4 in.m, ‘I 123':}.“_:‘ 2“‘.0" . 3250 - 1300 1100 1210 2400 . Operation ‘c.mtmuing, salt hfis' fiozon twice ‘ 0.045.in.| :wcll_- LT e T e ‘ T ‘ .because of contral d:fficuhy and o power T L . failure 93442 Tngnel, ,5sn.coj Coom a2l 8200 1200 1000 1100 1328 Loop failed on thaw-out attempt after power i 0 045 in. wull R - failure; repairs undor wuy, operation to be : o ‘ resumed . 9377-1 - |ncone| 6 in.d) " 126 f '_ 2.0 1600 1300 1100 1175 1890 Operation continving; salt froze once be- Do . 0.045 in.wall ; S . i PO _ cause of power failure 9733-2 l'w?'jlncoml !5 in. CD. L y"l3cl o 2,0 | 3000 1300 - 1100 1210 1510 Operation cofifinuing; salt has frozen mic; ‘ L 0.045 in. woll L ' because of motor control difficulties and o . e | \ _ power failure 9354-3 “INOR‘B L 84 “- 2.75 o 1200 1070 1150 1160 Operation continuing; salt has frozen twice Hot leg 3, m b 4500 because of motor control difficulties and a o sched'40 L : ) power failure Cold leg, .n.'_ ! 5400 OD 00045 i". . . wuli - S | 9354-1 -f_:INOR-a, 35 0D, 126 25 2000 1300 1100 1210 184 Failed on stortup; repaired and restarted, L 0.045 in. wull ‘ - to ‘ : but flow stoppages because of plugs finally resulted in second failure; repairs . ‘ being made 9354-2 INOR-B, 4 in. OD 12 o ; 2.6 6500 1200 1050 1140 12 - Hastelloy pump bowl failed after 112 hr of L 0.045 ine wall T ' : ‘operation; repairs being made CPR - lnconel s 122 :,':‘ ' .'l o 5000** 1250 1095 1190 | 6673 Operation'confihuing; salt has frozen twice; Sodiom T port of salt tubing domaged second time 97,700 | 1085 n3s and replaced; pump replaced and fuel re- placed after second freeze *Composition 123: NoF-BeF UF (53-46-1 mole %) Composition 12:° NoF-LiF-f(F (11.5-46.5-42 mole %) - " Composition 126: LiF-BeF —UF (53-46-1 mole %) Composition 130: . LiF-BeF UF (62-37-1 mole %) Composition 84z NoF-LanBeF (27-35-38 mote %) Composition 122 NaF-ZrF -UF (57-42-1 molo %) **|n heot exchangar 8S61L ‘L€ AYVYANYI 9NIAN3T QO1d3d MOLTEN.SALT REACTOR PROGRAM PROGRESS REP-O.RT simulated reactor service conditions. The operating conditions specified are listed below: UNCLASSIFIED PHOTO 30631 Average loop power density . - 50 w/cm® | Nose coil power density | 187 w/_c:m3 Maximum wall temperature I. 1300°F | Maximum temperature difference ~ 150°F (obtained by the addition of 4 kw of electric heat) Pump speed | 1700 rpm Flow velocity ' 2.7 fps Propovsed test duration 2000 hr Reynolds number 1600 Fission power 7 _ 7.8 kw A section of the heat exchanger and the heater ~ have been fabricated, and out-of-pile tests of these components will be initiated as soon as they can be incorporated into an existing heat transfer test focp. Fig. j.2.i_7. G.ruphite Specimen Assembly for In.sertion into INOR-S ququ_-Circulction Loop. 34 ¥ ») PERIOD ENDING JANUARY 31, 1958 1.3. ENGINEERING RESEARCH H. W. Hoffman HYDRdDYNAMIC_ STUDIES OF MSR CORE F. E. Lynch The molten salt reactor as presently designed has a spherical core. Three entrance-exit systems have been proposed for this core, namely, inlet and outlet diametrically opposite each other with straight-through flow, inlet and outlet concentric with the fluid entering through the inner pipe and exiting through the outer annulus, and inlet and outlet concentric with the fluid entering through the annulus and exiting through the inner pipe. With each of these entrance-exit systems, stagna- tion and flow separation are possible in the spherical core region and could cause excess heating of the core wals and thermal cycl:ng of reactor components. Glass models of each of the 1hree core configura- - tiens have been fabricated from large (13.7-in.-ID) Pyrex reaction vessels. Two of the models are shown in Figs. 1.3.1 and 1.3.2. The straight- Fig. 1.3.1. MSR Straight-Through Flow Model.’ through core has 6-in.-dia entrance and exit pipes. The concentric system has a 4-in.-dia central pipe within a 6-in.-dia 90-deg elbow. In the model for annular flow studies, the central tube is flared ‘outward at the bottom to direct the flow along the spherical surface. Initial qualitative data have been obtained for the concentric system with fluid entering through the central pipe by using the phosphorescent- particle flow-visualization technique. Both visual observation of particle motion within a continuously illuminated plane and photographic recording of instantaneous velocity profiles were employed obtaining the data. Figure 1.3.3 shows the results obtained visually at an inlet Reynolds modulus of 80,000 (2.58 fps in the 4-in. pipe). The inlet jet was of sufficient strength to retain its shape (a 4-in.-dia cylinder) to the bottom of the vessel. At this point the fluid turned and moved upward in a‘_l/2~ to %-in.-wide annular region immediately UMCLASSIFIED 3 PHOTO 30531 Fig. 1.3.2, MSR Concentric Entrance-Exit Core Model Designed for Fluid to Enter Through Inner Pipe. 35 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT Y 11 I FLUID LI UNCLASSIFIED ORNL-LR-DWG 27903 [ao— 7 FLOW UP OUTER WALL IN LAYER ABOUT {in. TO Yin. WIDE. " LARGE TURBULENT EDDY WHICH BROKE UP INTO TWO EDDIES, WITHIN THE LARGE EDDY, AS THE FLOW PULSATED. . S —_— FLUID FLOW QUT — 6-in.~-1D GLASS PIPE ————— : . — - s Y, “— LOW VELOCITY AREA PARTICLES SETTLE OUT VELOCITY PROFILE OF INCOMING FLUID; AS THE FLOW PULSATED, THE FLUID FLOWING IN FANNED QUT INTQO THE SPHERE, AND THUS DECREASED THE SIZE OF THE LARGE EDDY, OR THE LARGE EDDY EXTENDED INTO THE FLUID FLOWING DOWN THE CENTER OF THE SPHERE. Fig. 1.3.3. Flow Patterns Observed Within an {lluminated Plane of the Concentric Entrance-Exit Core Model with Fluid Entering Through Inner Pipe. adjacent to’ the core wall. The volume between these two flow regions was filled with a large toroidal ‘eddy, with its axis of rotation lying near the main upward flow. Because of unsteadiness in the flow-generating system, the flow patterns fluctuated. The main effect of these pulsations was in the structure and size of the large eddy. During periods of pulsation, the central core of ‘the eddy broke up into two smaller eddies sur- rounded by the larger rotating mass. It can be - 36 “seen that the eddy removes fluid from the exiting stream and recirculates this liquid through the high heat generation regions of the core. If good mixing occurs, there could result a uniform increase of the -exit temperature that corresponded to a longer fluid residence time in the core. However, there exists a strong possibility that the outgoing stream will contain discrete fluid bodies that will be at temperatures considerably higher than the mixed-mean temperature. Under these conditions, i o Hy high-frequency thermal cychng of system com- ~ ponents could occur. - Several typical mstantuneous proflle photogruphs of the flow in the concentric system are shown. in Fig. 1.3.4. Since grid photographs are not avail- able, quantitative velocety analyses of the data cannot be made. In all cases the time lapse be- tween the initial excitation and the recording of the profile was approximately 0.1 sec. In Fig. 1.3.4a the velocity profile at the exit of the inlet pipe is shown. The straight light line above the profile indicates the initial excitation, or zero-time line. The zero-time line is either dlrectly marked or indicated by a small triangle near the outer sphere wall on all the photographs. The profile in the outer annulus is also visible in Fig. 1.3.4a. Figure 1.3.4b shows the profile at a position just cbove the equator of the sphere, and both the - downward central flow and the upward flow along the wall may be seen. The effect of flow pulsation may be seen clearly in Figs. 1.3.4c and 1.3.44 for excitations at the same position but at different _times. Figure 1.3.4¢ shows a profile obtained ot a 45-deg angie near the sphere bottom. = This photograph indicates a very rapid drop- off in velocity in moving rud:olly away from the surface ond defines clearly the extent of the low velocity - region (high llghf-mtenslty region). Studies will be continued with this model, as well as with the straight-through and the annular entrance systems. The effects on the velocity pattern of inducing vortex motion by a generator Iocuted at the entrance will also be defermlned PHYS|CAL PROPERTY MEASUREMENTS ' W, D Powers e 7 Experlmentul determmuhons are. bemg que o!--,- | rthe viscosities -and thermal conductivutles of - several BeF -beurmg fluoride salt’ mixtures.” The o composmons to be studied have been: prepared ] " and preliminary ‘measurements are being made on " _one mixture. The tox;c:ty of the beryllium salts . "‘_‘dactuted that oll meusurements be made -in |nerf—‘_, - cup (capillary) viscometer will be used. " This: :f‘mstrument ‘consists’ of a 1 m.—long ccp:llary tube . . - _through “which the fluid- dralns from a’reservoir ~ with a capacity of about 6 cm®. The time required for the reservoir to empty ,through the tube is directly proportional to the kinematic viscosity. The cups to be used have been calibrated with: "PERIOD ENDING JANUARY 31, 1958 glycerine-water solutions whose viscosities have been accurately obtained wnh Cannon-Fenske- Ostwald viscometers. The thermal conductivities of several Ber- confmmng fluorlde salt mixtures will be measured with the -use of a varicble-gap device. In this apparatus the molten salt is contained in the gop between two parallel plates, the upper plate being movable. A series of measurements are made for various salt thicknesses. This technique permits the separation of interfacial and metal thermal resistances so that the thermal conductivity of the salt alone is obtained. MOLTEN SALT HEAT TRANSFER STUDIES H. W. Hoffman Heat transfer studies of molten scltsl'z have indicated that the general correlations for heat transfer of ordinary fluids (0.5 < NP < 100) also apply to the salts. -However, these investigations _have also shown that for some metal-salt combina- tions marked reductions in system heat transfer occur because of interfacial film formation and nonwetting.. Therefore, data.on these phenomena are needed for the desngn of crmccl heat trunsfer “components.’ ' : The apparent heat trunsfer coeffncnenf experi- 'mentully determined for flow through a single _circular tube has been found to be a sensitive _indication of the presence of nonwetting or inter- ~facial film formation in molten salt systems. The _experimental apparatus developed for such heat - transfer coefficient determinations is shown in. Fig. '1.3.5, and @ schematic diagram of the system " _is shown in Fig. 1.3.6. It is proposed to use this "'-.;’apparufus to study the salt mixture LiF- Ber-UF4 " {53-46-1 _mole %) ‘in- both . Inconel ‘and INOR-8 ~ systems. - The molten salt, contained in the two ‘tanks, is ¢ycled through an electrical-resistance- _heated test section by pressurizing one of the “tanks with the helium ‘blanket gas while venting - -the other tank. ‘When the salt has been transferreds S mto fhe vented tonk 1he pressures are uutomahcu“y " gas dry boxes.. For the vuscosnfy study, the efflux o T - IH. ‘W. Hoffindn, 'Turbflul‘erirt Fd}cea-Cbnvectzbn H’eat 7 ; Transler in Circular Tubes - Contammg Molten Sodmm Hydroxide, ORNL-1370 (Oct. 20 1952), 2H W, Hoffmon and S. I, Cohen, Fused Salt Heat Transfer Part lI; Forced-Convection Heat Transfer in Circular Tubes Containing the Salt Mixture NaNO ,- NaNOB-KNO ORNL.-2433 (to be published). 37 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT @ W c) - )y (e) : Fig. 1.3.4. Insfantuheous-VeIocity-Pffi?ile 'Phofogrdph# in 'Concenfr'i'c.;E_iiirdncé-Exit_ Core _Modd with' Fluid. Entering Through lnné; Ffipe.‘ : o ' S 38 Y O af ¥ » b Fig. -1.3.5. Experimentol sttem for Mole %). Defermifiing Heat T.l"unsfer PERIOD ENDING JANUARY 31, 1958 UNCLASSIFIED PHOTO 28430 Coefficients for LiF-Ber-UF4 (53-46-1 39 MOLTEN-SALT REACTOR PROGRAM PROGRESS.REPORT reversed, and the flu:d is coused to flow in the opposite direction through the test section. The fluid flow rate is determined by observmg the deflection of the steel beam that supports one of the tanks as the fluid flows into or out of the tank. Test section inlet and outlet mixed-mean tempera- ) tures, as well as tube outside surface temperatures, are recorded. - Because of the toxicity of the beryllium salt, traps or absorbers on the system gas vent lines. it will be necessary to provide -Sufficient salt has been prepared to charge this system, and experimental measurements will be _made in the near future. Since film formation, if such should occur, may “be a relatively slow - process, an apparatus is _bemg designed which will allow long-time, continu- ous exposure of test section tubes to flowing salt. _ The tubés will then be removed from this system and welded into the heat m:msf_er coefficient apparatus for study. . " UNCLASSIFIED ORNL—LR—DWG 27904 CYLINDRICAL STEEL BEAM _ VENT 4 * He SUPPLY ¢ :En —~—BEAM DEFLECTION INDICATOR VENT 1 : Lb—SOLENOID- VALVE ” SOLENOID- VALVE—|J|] E LECT RODES — FLEXIBLE HOSE O SWIVEL JOINT . CONCRETE 3 - Po T :'-> — SUMP TANK NS TEST ~ SECTION \Muxms CHAMBERS FLEXIBLE BELLOWS bk . ———— L/ "WEIGH TANK Fig. 1.3.6. BeF,-UF, (53-46-1 Mole %).. 'Séfiemati;: Diagram of E‘xperimehtu.l;System for Detrerminingyl'leui Transfer Coefficients for LiF- PERIOD ENDING JANUARY 31, 1958 1.4. - ADVANCED REACTOR STUDIES A MOLTEN SALT NATURAL-CONVECTION | B UNCLASSIFIED _ REACTOR = ORNL-LR-DWG 27943 F. E. Romie B. W. Kinyon EXPANSION DOME One of the problems of a circulating-liquid-fuel reactor is the provision of reliable iong-lived fuel- circulation pumps. This problem would be elimi- nated if the liquid fuel circulated by natural convection through the reactor core, through vertical convection risers, and through the primary heat exchangers. A molten salt fuel is well suited to such a reactor because high temperatures could be attained without pressurization, and therefore a gas-turbine cycle or one of several steam cycles that operate efficiently under high-temperature conditions could be utilized. The temperatures would not be so high as to be inconsistent with the desired long corrosion life. The advantages of eliminating the fuel-circulation pump and the attendant problems of maintenance and replacement are obtained at the cost of the. increased fuel volume required for a system in which the pressure losses must be very low. There are, however, applications for a reactor system in which the premium placed on reliability and ease of mainte- nance could make the natural-convection. system attractive. The results of a preliminary design study of such a system are described bruefly here, and a detailed report is being published.” A schematic diagram . of the 60-Mw (thermal) ‘molten salt natural-convection reactor system that _ . was unalyzed is shown in Fig. 1.4.1. The. temper-. L . ~ ature of ‘the fuel suh enterlng the: exchanger was_f*-' R ~specified to. ‘be - T225°F “a temperature that s .. . ‘consistent w:th long ‘corrosion life, ‘and the exlf'f;;-:;}.'g Ny o fuel-salt- 1emperature was varied from 975 t0'1025°F.. Ly "' The fuel salt used in'this study was LiF-BeF, UF Y A e - (62-37-1 mole %, mixture 130). . Two med:ums a‘ R - molten - salt (NcF-LlF BeF .27-35- 38 mole %o, - mixture 84) and helium, were cons:dered as cooldnts S _for the. prlmary exchanger. - The exchanger entrance . ~and ‘exit temperatures: for the ‘molten salt coolant - - were fixed at 875 and. 1025°F, respectively. These'j-’: T ‘ jtemperutures are consistent. with the generation of ~850-psia, 900°F steam in a sysi‘em in which heat- . . -0 HEAT EXCHANGER COOLANT 1 DOWNCOMER " - s tronsferred from the fuel salt to a- coolunt sald - F. E. Romie and B. W. Kinyon, A Molten-Salt Natural~ : : u Convection Reactor System, ORNL CF 58-2-46 (t;‘ ‘ie Fig. 1.4.1. Schematic Diagram of o Molten Salt published). " Natural-Convection Reactor. 41 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT to sodium to steam. A similar system has been descnbed elsewhere.? For a 60-Mw thermal output, " the 850-psia, 900°F steam would give a generator output of about 22 Mw, with a thermal effncnency of 37%. Two sets of terminal temperatures were specified for the helium coolant. The first set, 850 to . 1025°F, was selected for generation of 850-psia, 900°F steam in a system in which helium would be the only intermediate medium between the fuel salt and steam. The second set of helium terminal temperatures, 676 to 1100°F, was selected for a helium gas-turbine cycle with an estimated output of roughly 18 Mw. ' For a specified flow rate of the fuel salt, temper- ature change of the fuel salt, and pressure drop across the exchanger, there is one combination of riser diameter and height of exchanger above the core for which the salt contained in the risers is a minimum. - These optimum combinations were used in heat exchanger calculations carried out for exchanger heights above the reactor of 5, 10, and 20 ft. Salt-Cooled Heat Exchongers , Design values for the salt-cooled heat exchanger are summarized in Table 1.4.1. The designs are based on 0.634-in.-ID, 0.059-in.-wall tubes. Smaller- diameter tubes would give a lower total fuel-salt volume, but the number of tubes required would be 2B. W. Kinyon and F, E. Romie, Two Power Genera- tion Systems for a Molten Fluoride Reactor, paper to be presented at the 4th Nuclear Engineering and Science Conference of the 1958 Nuclear Congress, Chicago, lll., March 17-21. larger. For example, 0.42-in.-ID tubes in an ex- ~‘changer 10 ft above the reactor would involve a total external fuel-salt volume of 123 #3, but it would require 7200 tubes . The design valves presented in Table 1 4.1 are based on a fuel-salt temperature change of 225°F. Varying the temperature change from 200 to 250°F does not change the total fuel-salt volume appreci- ably. ' Helium-Cooled Heat Exéhufigers for Steam Cycle Reference design values for a helium-cooled heat exchanger in which the helium terminal temperature would be suitable for the generation of 900°F steam are presented in Table 1.4.2. It was assumed Table 1.4.2. Design Values for a Hélium-Coolé_d Heat Exchanger for a Steam Cycle Helium temperature range: 850 to 1025°F Helium pressure level: 100 psia Height of heat exchanger above 10 reactor, ft Number of tubes 3240 Length of each tube, ft 4 1.5 Distance from front to rear of tube 0.77 array, ft Transverse dimension of heat 78 exchanger, ft Generator output used to pump 2 helium through exchanger, % Total fuel-salt volume outside core, ft3 172 Table 1.4.1. Design Values for a Salt-Cooled Heat Exchanger for a Natural-Convection Molten Salt Reactor Coolant salt temperature range: 875 to 1025°F Height of heat exchanger above reactor, ft Number of tubes Diameter of tube bundle, ft Length of each tube, ft Fuel-salt volume in risers, ft3 Fuel-salt volume in headers and tubes, ft° Total fuel-salt volume outside reactor core, ft3 5 w20 4570 3150 2200 5.1 | 4.2 3.5 7.8 ns - 16 . 55 68 89 97 91 | 86 152 159 s 42 o i 4@ that copper fins would be used on the helium-cooled heat exchanger tubes. The use of lower-conduc- tivity fins would lead to a considerably increased fuel inventory. The dimensions given for the heat exchanger are those for a three-helium-pass cross- flow exchanger. - Doubling the helium pressure level to 200 psia while maintaining a pumping expenditure of 2% of generator output would not change the total salt volume appreciably but would give a more compact heat exchanger. A similar result would be obtained if at 100 psia the pumping expenditure were doubled. , - A cross section of a possible helium-cooled heat exchanger configuration is shown in Fig. 1.4.2. The configuration shown permits attachment of the riser pipes to the salt headers, with sufficient flexibility to minimize thermal stresses, and pro- vides a cylindrical container for the pressurized helium. . Use of two such cylinders to contain the reference-design heat exchanger would give two PERIOD ENDING JANUARY 31, 1958 heat exchdnflgers,'eoch 19.5 # lon.g.r If the helium pressure were increased to 200 psia, the length ‘of each H_ed_t exchanger would be reduced to 12 ft. Helium-Cooled Heat Exchanger fbr Gas-Turbine Cycle The temperature of the helium returned to the heat exchanger from the gas-turbine-system regen- erator is estimated to be 676°F, which is 174°F below the fusion temperature of the fuel salt. Therefore a counterflow heat exchanger with fongitudinally finned tubes was selected for the gas-turbine cycle. The temperature of the interface ‘between the fuel salt and the tube wall can be maintained at, or above, any desired value by adjustment of the gas-side thermal resistance per unit tube length. The design values given ‘in ~ Table 1.4.3 are based on a minimum interface temperature of 900°F. For a longitudinally _finned_ UNCLASSIFIED ORNL-LR- DWG 27948 _-PRESSURE SHELL Fig. 1.4.2, Schematic D.iugrum of a Possible -FueI-Suit—to-Heliun'.l Heat Exchanger with Three Gas Passes. 43 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT Table 1.4.3. Design Values for a Helium-Cooled | Heat Exchanger for a Gas-Turbine Cycle Helium temperature range: 676 to 1100°F Height of heat exchanger above 10 reactor, ft Number of tubes 3300 - Length per tube, fi ' 1 Total fuel-salt volume outside | 160 core, ft3- : tube the required gas-side thermal resistance variation along the tube length can be obtained by an axial variation of the fin height. No data were available on the heat transfer and flow friction characteristics of longitudinally finned tubes in a ‘tube bundle, and the exchanger designs are thus’ not complete. ‘However, the number of tubes, length per tube, and fuel-salt holdup volume can be estimated with good accuracy if it is assumed, as seems likely, that the same gas-side thermal resistance is realizable with uncut longitudinally fined tubes as with circumferentially finned tubes. Comparison of the Yarious Cooling Systems The fuel-salt holdup volume differs less than 10% for the three cooling systems considered and is therefore not a determining factor in the selection of the coolant medium for the reactor. In general, the helium-cooled heat exchangers have much larger over-all dimensions than the salt-cooled heat exchangers. The increased bulk is caused by the larger spacing required by the finned tubes and also by the large volume required by the helium headers. An increase in helium pressure would decrease the helium header volume and also should decrease the fuel-salt header volume. An upper limit on the helium pressure would probably be set by consideration of the effects of a tube rupture in the fuel-salt system. It is of interest to note that, in other studies of gas cooling in which such considerations were apparently not limiting, gas pressures as high as 1000 psia have been recommended. Comj:nrison of Natural-Convection System with Forced-Convection System The design of a forced-convection reactor sys- tem? led to an estimate of 0.56 #3 of fuel salt 44 external to a molten salt reactor per thermal “megawatt. For the free-convection system using a 160-ft3 external salt volume in the generation of 60 Mw, the corresponding specific volume is 2.67 3 /Mw. Calculations based on these numbers and initial, clean, critical-mass data for an un- ‘reflected molten salt reactor indicate that the fuel inventory for both the free- and forced-convection systems is ot a minimum with a core diameter of about 8 ft for a thermal output of 60 Mw. For an 8-ft-dia core the specific powers ar'e'895rand 1275 kw of heat per kilogram of U235 for the free- and forced-convection systems, respectively. The - free-convection system is thus estimated to require ‘a fuel inventory about 42% greater than that of the forced-convection system. With higher thermal outputs the specific powers would be larger, but ~ the specific power for the forced-convection reactor increases more rapidly with increasing power than does that of the free-convection reactor. HIGH-FLUX REACTORS W. K. Ergen A systematic study® was started to determine the influence of various factors on the power ‘required to obtain a given flux. At the beginning of the study, a reactor idealized in ‘the manner described below was considered.? The fuel is concentrated in a spherical shell embedded in an infinite moderator; that is, the moderator occupies the space inside as well as outside the shell. The shell is “infinitely thin'’ but ‘““black’’ for thermal neutrons; that is, the thickness of the shell is small compared with its radius but large compared with the diffusion length in the fuel. The shell emits nf fission neutrons for each thermal neutron absorbed. Absorptions at epithermal energies, both in the fuel and the moderator, are neglected. In this model, the fission neutrons emitted from the shell slow down according to the age kernel® and then diffuse ~ Chain Reactions: 3w, K. Ergen, Preliminary Design Data for a Circulating Fluoride~Fuel High«Flux Reactor, ORNL CF-56-6-9, 7 Revision No. 2 (Jan. 28, 1958). 4W. K. Ergen, Flux Distribution in a Reactor Con- sisting of a Spherical Shell of Fuel in a Infinite Moderator, ORNL. CF-57-12-100 (Dec. 24, 1957). sA. M. Weinberg and L. C. Noderer, Theory of Neutron Volume 1, Digus:'on and Slowing ?ggm” of Neutrons, ORNL CF-51-5-98, p 111-38 (qu 15, ny n [} v according to diffusion theory, with the boundary condition created by the ‘‘black’’ shell. For this model, it is possible to compute the number, J, of neutrons which return, at thermal energy, to the shell per fission neutron emitted. The neutrons returning to the shell give Jnf fission neutrons, and if the reactor is to be critical Inf = 1. From this equation, the required 75f has been computed‘ for D20, Be, BeO, and C moderators and for various values of shell radius. The results are plotted in Figs. 1.4.3 through 1.4.6. The abscissa of each plot gives the shell radius, r, in cm, aond also in dimensionless units p = r/\NT, where /T is the slowing-down length. In the sphere enclosed by the shell (the island, or internal thermal column, or flux trap) the neutrons 5w, K. Ergen, Fluxes Obtainable in a Flux-Trap Reactor, ORNL CF-58-1-4 (to be published). PERIOD ENDING JANUARY 31, 1958 reach thermal energies at some distance from the shell, and, in the process of diffusing toward the shell, these neutrons set up a flux gradient, which results in high fluxes at the center. These center fluxes are also plotted in Figs. 1.4.3 through 1.4.6. It may be seen that for Be and BeO a flux of 10-3 neutrons/cm? per fission neutron emitted is ob- tained, which corresponds to about 140 Mw for 10'¢ neutrons/cm?.sec. For D,0 and C the flux per fission neutron is smaller, and hence the power required for 1016 neutrons/cm?.sec is higher. This is primarily due to the large diffusion constants of D,O and C, which make it possible for the neutrons to reach the shell without setting up a very large flux gradient. The curves are, of course, applicable only to the idealized model. Present investigations cover the modifications caused by more realistic assumptions, such as holes in the shell, finite thickness of the shell, moderation in the shell, and flux depression in the flux trap as a result of insertion of absorbing samples. UNELASSIFIED ORNL—LR—DWG 27505 ricm) 12 19 16 i8 20 22 24 26 28 30 32 34 36 38 40 42 44 (X 1074 | I | B T ] | 1 11 I ] T T3 g, 2-8 o 2 2 8 2.6 o § 37 2.4 c = s - T8 T T —— 22 & g 1 1o L & €0 N0 | T LT~ : g s - ~ : o e - . & @ _ ~ B : - . \ . - = - E4 - T : - T _"‘\ — 18 5 < < b Tao - S g =5 - W . . = h S R — ¢ o " Ssel g Ee : 2 -.:.‘.',-,__:7__,;. 1 ’ ‘ = ‘ Ee = i : S ) 4 wo ) . o e . s o o " - - —— 1 i Y : 1.2 o 1.0 D90 42 0 44 8 18 20 2224 26, 28 030 32 .34 36 38 40 ;- pIDIMENSIONLESS) L ' ' Fig. 1.4.3. Central Flux and Multiplication Factor as Functions of Shell Radius of 1dealized Flux Trap Reactor witha D 207 Moderator. 45 B MOLTEN.SALT REACTOR PROGRAM PROGRESS REPORT UNCLASSIFIED ORNL~LR —DWG 27906 r{cm) . 2 {0 2 4 16 i8 . 20 22 24 26 28 30 32 34 36 38 30" X107 ] l I I T I I I [ [ I | I ! I - 9 7 ~~. ' 28 : \ " N -8 - \ / \@c 2 N ' _‘_-.E: 8 \\ / ™ - 26 @ . \ ' ) g N\ \ £ N\ a7 - 2.4 s c g “\ - & £ e > AN 22 F g N \ g o \\ \ = E b, " g { 5 ‘\\ . \ 2-0 - g “wb ) \ . 8 .E.‘ S : o E =2 "-..- £ 4 = P~ 18 F OQ- ) - "‘-l-.--, .,___.----- i g § 3 = 16 i . | a . Eo2 1.4 E wl o 1 12 0 i0 1.0 2 14 16 18 2.0 2.2 24 26 28 3.0 32 3.4 36 38 4.0 p (DIMENSIONLESS) Fig. 1.4.4. Central Flux and Multiplication Factor as Functions of Shell Radius of Ideuiized.Fqu. Trap Reactor ‘with a Be Moderator, UNCLASSIFIED ORNL—LR-DWG 27907 ricm) 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 : —4 (x107%) r I T | | | T 1 | | [ | | I ] 3° 9 s - 28 / \ ¢c 3 pd , £8 26 § \ 4 5 \/ .; 7 \\ - 2.4 “ g S ’ = 5 € " 26 S oS 2.2 - 2 Y = : N S o~ \\ = E ~ .o {5 < 20 E E ‘\~\,qf g - \\ - 3 ~. _ g 2a S ~ 48 3 efli ..-"-.,_-‘- ) i = § 3 -—--- = m e _.___.__----“--- 1_6 It " s epam -l & A Ee 14 =z . Lt Q 4 12 .0 1.0 10 2 14 16 1.8 20 22 24 26 2.8 30 32 34 36 38 4.0 ' p{DIMENSIONLESS) - o ' ' o Fig. 1.4,5. Central Flux and Multiplica - with a BeO Moderator. 46 tion .Factor' as Functions of Shell Radius of |dealized Flux Trap Rec.cfor PERIOD ENDING JANUARY 31, 1958 UNCLASSIFIED ORNL—LR—DWG 27908 ':: rlem) 16 20 24 28 . 32 36 40 44 48 52 56 60 —g . - 30 (xt0 7 F] l J ] ] | | I | 1 i - 9 2.8 3 € B 26 L4 s \ 2.7 LS 2.4 < \\ - o B &= £ \, < 5 8 2.2 © . AN 5 NE . & O LY F =5 N7 205 g \*s '<-[ =2 ~\~ 9 < e & :&, 4 L= “%\_ 1.8 5 . / . ) "-...,__..‘~ \& g >3< . . —~ ~~a.l_ P — 6 [ '--..---..--..___J}‘ ;I‘ T .--Q'-—--- E 2 P — 1.4 =z T iy i O 4 1.2 o . 1.0 i 12 1.4 .6 1.8 20 22 2.4 26 28 30 32 3.4 36 38 40 p (DIMENSIONLESS} Fig. 1.4.6. Central Fiux and Multiplication Factor as Functions of Shell Radius of {dealized Fiux Trap Reactor with a C Moderator. » i ! ! . 11 . | - | S 47 j. . .- ‘ . . " Part 2 MATERIALS STUDIES . . L] W. D. Manly D YNAMIC CORROSION STUDIES Jl Ho Devan J. R. DiStefano. R. S. Crouse Corrosion experiments are under way for which thermal-convection loops and forced-circulation loops were fabricated of Inconel (nominal compo- sition: 15 wt % Cr, 7 wt % Fe, bal Ni) and of INOR-8 (nominal composition: 17 wt % Mo, 6 wt % Fe, 6 wt % Cr, bal Ni). As discussed previously, the test program is being conducted in three distinct phases,! In the first phase, relative corrosion properties of 12 fluoride sait mixtures are being determined in thermal-convection loops operated for 1000 hr. During the quarter, tests of Inconel loops were completed for all the salts except salt 131 (LiF-BeF,-UF ,, 60-36-4 mole %); four salts were tested in INOR-8 loops, which are now being examined. As part of phase 2 of the program, three salts are being circulated in Inconel loops and five in INOR-8 loops. The tests will be run for extended periods at two temperature levels, 1250 and 1350°F. In the third and final phase of the test program, salts are being tested in forced- circulation loops under simulated reactor operating conditions. Four salts are presently being tested in Inconel loops and three in INOR-8 loops, as described in Chap. 1.2 of this report. The results of postoperative examinations of these loops will be described in this chapter as they are completed. Ten of i'he Inconel loops operated in phase 1 of the progrom have been examined metallographically, ;'After successful .operation for ‘1000 ‘hr, the salt mixture - in each’ loop was allowed ‘to freeze in = ~ place. -Samples were then cut from each loop at - - the locations shown in Fig. 211, and the salt. _ . 'contamed in"each sample was ‘melted ouf under a = helivm utmosphere. The results of metc:llographlc,j o ;excmsnatlori of the samples are presented in Table_';'rij - 21 land are dlscussed below. ' ' ' Thonum-Bearmg Sulis in 'I'hermul-Convechon : : Loops R L The loops whlch cnrculated thorsum-bearmg saltsv'_rr' comprlse three pairs on- the ‘basis -of slmllarmes 1J.' H. DeVan, J. R. DiStefano, and R. S. Crduse, MSR- Quar. Prog. Rep. Oct. 31, 1957, ORNL-2431, p 23. A. Taboada in the fuels they circulated, Loops 1169 and 1177, the first pair listed in Table 2.1.1, circu- lated salt 128 (LiF-ThF,, 71-29 mole %). Loop 1169 showed less than 1 mil of attack throughout (Fig. 2.1.2), while loop 1177 was, in general, attacked to a depth of less than 1 mil (Fig. 2.1.3) but had scattered pits to a depth of 1.5 mils in some areas. UNCLASSIFIED ORNL-LR-DWG 23451 ~ | METALLOGRAPHIC SAMPLES HOT—LEG SIX 6-in. CLAM- CHEMISTRY SAMPLE SHELL HEATERS COLD—LEG CHEMISTRY SAMPLE ~ | METALLOGRAPHIC SAMPLES Fig. 2.1.1. Diagram of a Standard Inconel Thermal- Convection Loop Showing Location of Metallographic - ‘Samples. The sécond pair of loops, 1173 and 1176, was .. - operated to evaluate salt mixtures 124 (NaF-BeF ,- - ThF,, 58-35-7 mole %) and 127 (LlF-BeF -ThF4 © 58-35-7 mole %). . As may be seen in. Table 2.1.1, loop 1173, in which sait 124 was circulated, was - - attacked substanhally more than foop 1176, which . circulated” salt 127, * This result is in conflict, - -however, with the - corrosuon properties normally exhibited by NaF-BeF ‘and LlF-BeF2 mixtures; ~in previous tests the LlF-BeF mixtures. tended :to produce more initial corroslon in fnconel sys- tems than NaF-BeF2 mixtures produced, Thus the test results for loop 1173 are questionable, and o repeat test under similar conditions has been 51 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT Tuble 2.'l 1. Results of Memllogrnphlc Examinations of Inconel Thermcll-COrwecflon Loops ' Operuted in Phuse 1of Ccrrosion Test Progrom o Sqlt - Mefo_!logrdphig: Results H p_f;:L"e g Loop NS; Designation | Salt-:vC“o_m-posifion PR , ‘ _ _ o - Attack - - Cold-Leg Appearance (mils) - _ o | > The'rium-Bea;'ing So't_s . nee 128 LiF-ThF,, 71-29 ;5;'1; % <1 Light surface roughness 177 128 LiF-ThF, 7129 mole % s Attack 1o o depth of | mil 1’1_73_' - 124 NaF-BeF,-ThF 4,'53-"35-7 mole % 4 - Attack to a depth of 1 mil | 1176 ,lé7 | LiF-Ber-ThF4,V"758-35-,7 n;lole % <1 | Surfuce roughness and surfoce pits 1174 . 125 NaF- BeF'2~U F4-th4; 53-46-0.5-0.5 md_le'%; 2 . Surface roughness and surface pits 1163 .' i23 . Na F-Be‘Fé-UF4,53f4e;'|mole % : - the 80% Au=20% Cu alloy was exposed showed © 123 ppm “gold. - No attack was observed on either : D H Jonsen i The precnous-metcl base bruzmg al!oys, 82% Au—- “alley in’ any . of the fests. The 82% Au-18% Ni g +y 18% Ni and 80% Au-—20% Cu, which are being con - sidered for use in the fabrication of fuel-sait—to-fr_ o coolant—suh heat exchongers, have been corrosion tested in NaF-KF LiF-UF, (11. 2-41-45.3-2,5 mole - - observed on the same alloy when tested in fuel %, fuel 107) and in LIF-BeF ,-UF, (62-37-1 mole %, fuel 130). - Static tests were run on both the alloys - alloy - specimen " that was tested ‘in fuel 130 for - 500 hr in the seesow-furnace apparctus is shown in- - “Fig.- 21,15, The thin layer on the surface was 107. X-ray analysis on this layer indicates that it is composed of approximately 70% Ni—30% Au. 59 MOLTE'N-SALT REACTOR PROGRAM PROGRESS REPORT Fig. 2.1.14, 100 hr ot 1500°F. Etchant: copper regia. 150X, Cortosnon of INOR-B Welds by NaK and by ~ Fuel Salts D. H. Jansen INOR-8 plates from heat SP-16 that were welded with various nickel-molybdenum-base welding rods have been corrosion tested in seesaw-furnace apparatus at a hot-zone temperature of 1200°F in 60 UNCLASSIFIED Y¥.24915 Lol po 0 . Y, LhE - e "1, A e -3"‘_‘ . INOR-8 from Heat SP-16 (a) As Received and (b) Aher Exposure to o Sodwm-Graphfle Sysfem for NaK (56-44 wt %) and in fuel 130 for 500 hr, These tests were conducted- because the weld metals contained some minor constituents that have shown limited .corrosion resistance to these mixtures; however, no attack was found on any of the welds. The compositions of the welding rods are given in Table 2.1.4, along with the observed weight changes and the results of metallographic examinations of the tested welds, : v h L3 ay PERIOD ENDING JANUARY 31, 1958 UNCLASSIFIED Y-25064 ' . - LY Y ,- S0 » . fi'* ‘ \ a ‘. ‘ -. xR " - v " 4 b"»g‘ ‘.; *‘. 1 ' "‘ L% ,'- o e ""f‘*“‘“" R, sy el Sl oo r':fi?azfi’u"'{f}'_'z"" ¥ u 0¥ 73w’ Se ST re 3 . .' oot & . TN IO oo AR AR E Lo Beb i SN A g S e 1 etk o gpot Ay L Sy .-‘fiff,v'. BN s T L O - "Ql‘\"' ” '20.."‘ s ’. AT R TE IR T L "§'~‘:0f$¥,’3 J ¥ ,‘l_ + B "‘;"" .t, o g ." 2 ; .:..‘f - [ SRR QO R, (0 ot vyt St L BT ] "‘.’:.‘ j ‘. "."")-;., .,‘-'g‘- a%." & !3 ‘a ,..'.': - ... O‘. v “‘ OIZ e d Ay £ o g . Byt ¥, e e b a8 -"&"?4‘- S8 T .*.;,-‘f’f-:el‘ o0 6o & o . ! . - -, - gt & . o 5 g At AVPR Y ..,“-‘3-,'3,.».;.-3-(4}'\ e Ny Tjd\ s’ ’ ‘mg.g'fi o."'\ ',"2' ‘:::-95“ u .’ "':é Ci4 . L . - - y r - & Al :..:. <%0 f;"..' o~ o~ o i a3 :,fl >:'s!\~: Y Nt et i % AN e g T s TR NN LG, T R T s e Yoo \ ¢ ¥yt -1 Coiaatis ey y- (il e Y P e Ry L ARY - Paig ot I 0o s o0& denr w"!{ 3 & ML P SRR Ly L e 8 G i X .f',»'l- A% L 1, S A RS ke doil o "?';."'..;{"4 Jw;f dp: ’, m:: ot ‘..»‘ %{,‘\;___‘;...: v 3/. . :“:..3.. T e A T AR RN T AR 2 Ty Lxts n... A e a st N fy. AWM Y - Fig. 2.1.15. Speclfien of en 82% Au-18% Ni Brazing Alloy After Exposure to Fuel 130 for 500 hr in Seescw-. Furnace Apparatus with a Hot-Zone Temperature of 1200°F. Etchant: KCN. 250X. Table 2.1.4. Results of Corrosion Tests of INOR-8 Welds in NaK and in Fuel 130 Time period: 500 hr -Hot-zone temperature: "1200°F g Specimen 'T-estéd"-ifi Nol(.-_- RO Specilfiéh Tesfed in Fuel 130 : Nommul Composi!uon of : e e L - - e . A F:Her Rod ~ Weight Change = - Metallographic _, Weight Change = Metallographic TR - (®) - _oNotes " (W . Notes 74% Ni=15% Mo—6% Cr-S% Fe = +0.015 Noattack; cracks Iin . -0.035 No attack (heaf 30-38 INOR 8) . - = . heat-affected zone S 70% Ni-16% Mo—?% Cr—S% Fe 0,035 Noarockicracks'in —0.037 N attack; cracks ~“(heat SP- 16 lNDR-B) 0. < heat-affected zone - .. - - in welds . ”“60%N..25% Mo—7% Fe-—S% G- 40016 No attack on weld 0047 No attack on weld 2.5% Co (Ha sfellpy W) 61 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT PHYSICAL PROPERTIES OF INOR- 8- T. K. Roche - ' Measurements were made of the modulus of H. Inouye elasticity, the thermal conductivity, and the tensile properties of several commercial air-melted heats of INOR-8 fabricated at Haynes Stellite Company or at the Westinghouse Electric Corp. The compo- sitions of the heats studied are given in Table . 2.1,5. The SP designations indicate material from- Haynes Stellite Company and 8M designates material from Weshnghouse. Preliminary | thermal conductivity data for heat SP-16 in the annealed condition are presented in Table 2.1.6. The data wére obtained in vacuum, ~and Armco iron was used as the standard, Values are glven for Inconel for comparison. 'The values of Young's modulus glven in Table 2,1.7- were obtained by senic methods.® An annealed bar of heat SP-19 was used. Studies were initiated for determining whether INOR-8 has a tendency to embrittle in the tempera- 3Datu obtained by S. Fulkerson of ORNL wnth sonic equipment ot the Bureau of Standards. Table 2.1.5. Chemical Analyses of INOR-8 Heats ture range of 1000 to 1400°F. Specnmens are being ' 'uged for perlods of 0, 500, 1,000, 2,000, -5,000, “and 10,000 hr. - properties obtained. after 500-hr aging heat treat- ‘The " room-temperature " tensile ments at several temperatures are presented in " Table 2.1.8. These data show that the alloy does not become embrittled during a 500-hr aging period, - The higher strength and lower ductility of heat 8M-1 in comparison with heat SP-19 are ascribed to a difference of 0.08% in the carbon content. " Table 2,1.6, Thermal Conductivities of INOR-8 - "{Heat SP-16) and Incqnel at Various Temperctures Thermal Conductivity Tem;;:(r:fl;fl-!re ,[cal/Cfn2°sec-(°C/cm)] o INOR-8 Inconel 100 0,023 - 0.039 200 0028 0.041 300 0.042 o 0.043 400 0.050 | 0.045 500 0.059 - 0.048 600 0.067 0.050 (extrapolated) 700 0.075 0,052 - {extrapolated) ~ Amount Found (wt %) Table 2,1.7. Modqlus of Elasticity for INOR-8 at Temperatuies Up to 1050°C Element - _ Heat SP-19 Heat 8M-1 "_Heaf SP-16 Mo . 16.65 16.20 15.82 Cr 7.43 7.47 6.99 Fe 4.83 6.1 4.85 c 0,06 0.14 0.02 Si 0.04 0.21 0.32 W Trace * 0.35 Mn - 0.48 0.69 0.34 P 0010 0.009 0.009 ‘s 0,015 0.006 0.014 Cu o 0.02 * . 0.03 vV .10 . * * B S * 0.04 Co - 0,51 o+ . 0.51 Ti . o o Ni 7000 - 70.6 . 70.50 \ *No'_f dna[yzed f_'or. Temperature Young's Modulus (°C) (psi) Cx 108 14 L 3.7 223 o 29.3 - 412 o 27.8 501 27.1 576 26,3 636 262 701 . 24.8 800 = - 23.7 857 ' 22.7 902 219 953 : 20.7 1000 , 194 1050 | 7.7 AL ] i ap L1 PERIOD ENDING JANUARY 31, 1958 Table 2,1.8. Room-=Temperature Tensile Properties of INOR-8 Yield Strength at Tensi:e S.;rength 0.2% Offset EIor(n;c;tion si % Heat Treatment P (psi) Heat Heat Heat Heat Heat Heat SP-19 8M-1 SP.19 8M-1 SP-19 8M-1 Annealed 114,400 117,100 44,700 51,900 50 39 Annealed ond aged 500 hr at 1000°F 112,000 115,700 | 42,500 47,200 53 43 Annealed and aged 500 hr at 1100°F 112,600 114,500 44,000 48,000 51 43 Annealed and aged 500 hr at 1200°F 112,300 114,600 44,700 48,600 51 43 ; Ar'meo’ed and uged 500 hr at 1300°F 112,000 113,400 44,500 47,600 49 4 Annealed and aged 500 hr at 1400°F 112,400 116,000 43,900 47,000 50 40 MECHANICAL PROPERTIES OF INOR-8 - D. A, Douglas Tests are under way for obtaining the basic data on the strength of INOR-8 required for design calculations. The confidence level which can be applied to the data and the varicbles which affect ~ the reproducibility of the data will be defined, Studies will also be made of the behavior of the metal under both static and dynamic loadings in order to more accurafely predict fhe service life of various component parts, Since relatively long pericds are required to obtain. data on ploshc properties, preliminary data on fensile properties are being. obtained for use - ~in design studies. The yield srrength at 0.2% - 45 be mVeshgated " Creep tests are therefore “offset and . the rupture strengths -of _INOR-8 were - under way at stresses. ‘of 12,000 to 30,000 psi with measured - in the temperature range of 1000 to ~ 1300°F ond at room temperature, The results of '-_the meosurements are summorlzed in Table 2. 1 9 L o The dora presented in Table 2.1.9 were obtamed- - ~-on sheeir specimens .and must be consudered as - - '_opprox:rnote because of expenmentai errors in the ~ elastic-portion of the stress-strain data, Errors’ ‘occur _partly ‘because it -is difficult to. ochreve--i accirate alignment with a sheet. specimen and 'pcrfly becuuse subsize’ $pecimens are sensitive to . _ expemr-eniel vonohons._ Conventional specimens 0.505 in. in dlumerer are being machined from a wrought bar so that more accurate values can be obtained. The study of the plastic properties of INOR-8 is being made in order to determine whether INOR-8 will deform plastically under reactor operating conditions. In the relaxation tests used for this study, a specimen is loaded to a fixed amount of strain and the resulting elongation is maintained either by adding to the load or by subtracting from it. The need to remove the load to maintain the fixed strain would indicate that the material de- formed plastically. The results of a series of tests at 1200 and 1300°F are summarized in Tabie 2.1.10. The large decrease in stress with time indicated that the plastic properties of the INOR-8 were important and that rhe creep strength would have the specimens exposed to fuet 107 -at 1100, 1200, ~and 1300°F, The status of these tests is pre- sented in Toble 2.1.11, Tests were also performed at femperotures con- ‘gislderob!y above the anticipated operating tempera- ture of the reactor as a means of gaging the damage - to the metal which might occur through occrdentol rremperoture ‘excursions. - These - resulfs, are .pre- -sented in Fig. 2.1.16. ' “Ancther series of resrs will be run in order to ;obrom a siorrshcol estimate of the reproducibility of‘ creep results from one test to another and from one heat of material to another. These tests will be conducted in air at 1250°F, 63 MOLTEN-SALT REAC_TOR PROGRAM PROGRESS REPORT Table 2.1.9. Elastic Pr_operties of INOR-8 Yield Strength at Ultimate Strength ' ‘ Temperature Ductilit INOR-8 Heaf | :’:F) 0.2?p:)i|;fs_et o) - e Y Haynes SP-16 Room 45,000 106,000 58 Haynes SP-19 Room 45,000 ) 114,000 39 Westinghouse " Room 52,000 117,000 50 ‘Haynes SP-19 1000 27,000 90,000 e Westinghouse 1000 36,000 100,000 o Haynes SP-19 1100 29,000 93,000 50 Westinghouse 1100 38,000 103,000 ) 37 " "Haynes SP-16 1200 25,000 67,000 44 " Haynes SP-19 1200 27,000 82,000 36 Westinghouse 1200 38,000 83,000 16 Haynes SP-16 1300 24,000 58,000 v - Haynes SP-19 1300 28,000 70,000 24 Westinghouse 1300 | 38,000 70,000 ST Table 2.1.10. Relaxation Data for INOR-8 Tempérotufe Strain Initial Stress Stress (psi) for Constant Elongation (°F) . (psi) (psi) At 1 hr " At10hr At 100 hr 1300 © 0.05 11,000 11,500 10,000 6,000 0.1 21,500 21,500 16,000 5,500 0.2 . .29,750 . 20,500 - 10,500 , 4,500 1200 0.05 12,000 12,500 12,000 10,000 0.1 22,500 23,000 22,000 - 17,000 WELDING AND BRAZING STUDIES ' - G, M. Slaughter Metal Seals for Remote-Disconnect Flanged Joints In the development of mechanical joints that can be disconnected by remotely controlled mecha- nisms, it has been necessary to investigate various ‘types of seals. Tests have been made of the feasibility of heating a sealing material in an ‘annulus to make or break the seal or of pouring molten metal into a preheated and cldmped flange assembly, The results of mockup tests of such a 64 joint, designated “‘cast-metal-sealed flange joint,"” I gn ge | _ are presented in Chap, 1.2, 7 A survey of phase diagrams of possible seal and flange materials which should be relatively immiscible in each other indicated that the silver- nickel system might be useful. Silver and iron ~were found to be virtually insoluble in each other, and copper and iron possess only limited solu- bility in each other. It also seems probable that the silver-copper eutectic alloy, Handy & Harman alloy BT, which melts at 1435°F, could be used with iron-base flange materials, v AL ] wh o Toble 2.1,11. Creep Data for INOR-8 Temperature Stress Strain Time Creep Rate (°F) {psi) (%) (hr). {%/hr) 1300 30,000 15.33 110 Ruptured 25,000 6.53 187 * 20,000 9.56 882 Ruptured 15,000 10.99 2894 Ruptured 1200 30,000 4.7 9** Ruptured 25,000 2.84 1195 2x10-3 20,000 2.70 1894 2 x 10~4 15000 0.81 1863 2% 104 12,000 0.97 1172 6x 103 1100 25,000 0.70 164 * 15,000 - 0.58 68 * 12,000 0.3% 43 - *Test under way. **Tegt to be repeated; low elongation at failure is not considered to be typical. . UNCLASSIFIED ORNL—LR—=DWG 27909 [ ' A "-1650‘F / ) —{500°F 2 S STRAIN (%) o > o Q L \%,\ 606095\ / P N, N R 0 50 100 150 200 250 300 . 3/0 400 450 - 500 L ,TIME(hr) e Flg | 2.1, 16.‘ Creep Curves for Soluiion-Annealed. ) ‘ INOR-B Tested in’ Fuel 107 ut Varlous Temperutures' R : ;rond Sfl-esses. R o : , PR ‘ The speczmens ‘are shown in- Flg. 2.1.17. Meta“ographlc ‘examinations were mode of dupll- cate samples after solidification and after holdmg - at elevated temperatures for extended periods of time in order to determine the extent and type of - PERIOD ENDING JANUARY 31, 1958 diffusion. The excellent wetting of nickel by silver in dry helium is indicated by the small contact angle shown in Fig. 2.1.18, An even smaller contact angle was found on the nickel|-BT alloy sample. Metallographic examination of the silver-nickel interface revealed no penetration of silver into the nickel after brazing and after sub- sequent aging for 500 hr at 1200°F. Only slight penetration of nickel by the silver-copper alloy was observed ofter aging. The marginal wetting of INOR-8 and Inconel by silver in dry helium is illustrated by the contact angle of approximately 90 deg shown in Fig. 2.1.19. The BT alloy ex- hibited very poor wetting on both INOR-8 and Inconel. The wetting of iron by silver and the BT alloy in dry helium was poor and intermittent, The deposition of an electrolytic nickel plate on Inconel and INOR-.8 samples was found, however, . to promote good wetting by both silver and the BT alloy. The nickel plate was not penetrated by silver, but there was some solution of the nickel plate by the BT alloy. The deposition of an electrolytic silver plate on the electrolytic nickel plate did not improve the wetting by either alloy, Aging experiments at 1200°F are now under way on electroplated materials. A preliminary sealing test specimen was de- veloped that consisted of a nickel tube and a nickel container in which the seal could be made with silver under a dry helium atmosphere. The joint was made and broken four times, and the joint was leaktight after each sealing. Thus there appears to be excellent wetting of nickel by silver under repeated heating ond cooling cycles. As ~ stated above, mockups of cast-metal-sealed flange ~_joints ‘are bemg tested fhat uhhze the results of these studles. - Weld ing of INOR-8 Tubing G M. Slaughter Before the fabncahon of INOR-8 tublng from heat 7 A series of tests was fhen conducted in d,-yi - SP-16 into the various corrosion loops of interest - hehum in order to study the wetting character;st:cs‘, of silver and the silver-copper eutectic ailoy on “nickel, iron, and other possible flange materia)s — - - Vr'type 316 stamless steel, INOR-8, -and Inconel ‘ was mmated, it was decided that a metallurglcal iinvestigation of typical welds should be conducted A preliminary mvesngahon, described prevnously, _ ?_‘_mdtcated that fhe use of SP-]6 filler wire for the _welding of /2 in. plate from heat SP-16 was, in "general, | ‘unsatisfactory because of weld-metal cracking. Also, welds made on similar specimens 4P; Patriarca and G. M. Slaughter, MSR Quar. Prcg. Rep. Oct. 31, 1957, ORNL-2431, p 18. 65 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT SILVER ' BRAZING TEMP. 1800°F INOR 8 HELIUM ATMOSPHERE BT SILVER SOLDER i - BRAZING TEMP. 1480°F HELIUM ATMOSPHERE 70N SP16 ~ .UNCLASSIFIED T Ya24465 stess - INCONEL . NICKEL I T 36 SS - INCONEL -~ NICKEL Fig. 2.1.18. Specimen Showing Excélle:rlrWeflihg of a Nickel Tube by Silver. Unetched. 50X. Reduced 30%. under . condmons of high restrumt resulfed in severe bcse-metal cracking. Test welds have since been made on fhe / in,- OD, 0.035-in.-wall tubing from heat .SP-16 under low-restraint conditions. Since the previous * experiments had indicated that SP-16 filler wire was not satisfactory for this application, material from an ORNL heat bf INOR-8 (heat 30-38) which had shown promise in weld-cracking tests was fabrscated mio wire and utilized as flller metal in 66 . :Fig.' 2.1.19. Specimén S}‘:c.'w'ing.'-flarginq-l Wefling of INOR-8 by Silver. Unetched. 50X. Reduced 30%. ~ the brodbc_tion of the test welds. Heat 30-38 has the nominal composition 15% Mo—6% Cr—5% Fe-— 0.5% Mn=0.5% Al-0,06% C—bal Ni. - A photomicro- graph of a typ:cal welded |omt is shown in Flg. 2.1.20. “The results of visual, rudlographlc, and metallo-r graphic examinations .and mechanical .tests at room temperature and at 1300°F on as-welded- .specimens indicate -that sound, low-restraint butt welds of SP-16 tubing con be made with heat 30-38 o e e R ALl b PERIOD ENDING JANUARY 31, 1958 UNGLASSIFIED «24329 Fig. 2.1.20. Typical Joint in SP-16 Tubing Welded with Heat 30-38 Filler Wire. Etchant: chromic and hydro- chloric acids. 25X, filler metal. No base-metal or weld-metal cracks were found, and the room- and elevated-temperature mechanical tests of the as-welded joints indicated satisfactory characteristics. The properties of the joints after aging ot elevated temperatures are being investigated as a means of determining their over-all suitability for high- temperafure applica- tions, Since no significant qucmhty of heat 30-38 flller , Numerous saddle’ welds were exommed durmg,p the development of the weldmg procedures, and " infergranular - base-metal ‘cracks fo the extent - of 20% of the tube wall thickness were oceasionally i observed in the microsections, The higher re- straint conditions involved in the fabrication of meicl ~was -ovailable in the optimum. wire size, ' maierlol from ‘another ORNL heat (30-72) of the - some “nominal - composutlon was - processed into- ~wire, Sample butt welds of the /B-m.-OD 0.035-- “in.awall $P-16 tubing were ‘made as “before, and a - - . ‘were described prevnously.‘l_ preliminary evaluation made to ‘obtain data for the ‘ - Ctest, whlch -utilizes ‘an inert-arc fusion. pass on a ~“development of a procedure specification indicated” " -3 ' ~ -that the welds were sut:sfactory. _ “mental work has been. completed on which to base " a procedure specification ‘and an opemror s quoll-j" L flcahon test spec:hcuhon. . T ‘The develop-;f_-‘ these welds undoubtedly explain the presence of base-metal cracks in the saddle welds aond the freedom from cracks in the lower restraint butt welds. Since the detection of these defects cannot be ‘ensured by radiography or dye-penetrant in- spection, it is recommended that the SP-16 tubing '}nof be used for critical applications, such as in- pile loops, Evnluutlon Tests for Welds - Prehminary results in the development of screen- ,-'_._mg tests for determmmg the relative susceptl- ' ‘-l)llll’les of various alloys to weld-metal cracking ‘The circu lor-groove ’,,szqn.-wude, 2-|n.-dta, }g-ln.-deep clrcular ‘groove ~machined 'in-a "4 % 4-in, specimen of /2 in, plate, “has now been used fo test the weld-metal cracking "tendencues ‘of the - INOR-8 alloys Westmghouse‘ 8M-1-and Haynes SP-l9 {see Table 2.1.5, above, for- H1e composmon of these alloys). "No weld- __i'metol “cracks -were. found in either: alloy, the "_'-Weshnghouse 8M-'l spec:men is shown in- Fig. 2121, | Metollographic 'secfi_oning_ of this type of speci- men can also be used to determine the susceptibility 67 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT of the material to base-metal cracking during weld- ing. Base-metal cracks |dent|ca| to those found in htghly restrained test welds? ‘were found in a Haynes SP-16 specimen,-as shown in Fig. 2,1.22, Fig. 2.1.21, Weshnghouse 8M-1 Circulur—Groove Weld- Crockmg-Test Specimen. ‘No base-metal defects - were observed in the Westtnghouse 8M-1 matenal “a typical area’ of which is shown in Fig. 2.1 23. ‘No strmgers or “inclusions are evident. Ahhough no ‘base-metal ~.cracks .were found in the Hoynes ‘SP-I9_specnmen, the presence ‘of slight, occasional, fusion-line porosnty was noted Thls condmon |s shown in ‘Fig. 2.1.24, A crack test for materials avullable onIy in rod form was also deveioped “A longitudinal slot, 3/16 in. deep and / in. wide, is machined in ¢ z-ln.--dla bar and a fusion weld is made along the slot. Weld-metal cracks were found in a Haynes SP-16 bar, but there were none in c Westmghouse 8M-9 bar. The two specnmens ‘are. shown in ‘Fig. 2.1.25. The weldmg charocienshcs of seven’ dlfferent - heats of INOR-8 from three different sources have been studied, and ‘only the Huynes SP-16 heat “exhibited crccklng tendencies. This crockmg has ' been ‘attributed to the melting practice and should not be considered as mdlcahve of fhe properhes of the alloy., Weld test plates are bemg prepared as a means . of evaluating the weldability of heavy sections of nickel-molybdenum alloys. Three stages in the W.mfim;rnwm UNCLASSIFIED, - Y«24908 0, p- b J x o - z 0.02 0,03 x O Fug. 2.1.22. Buse-Meta! Cracks Found by Metuilogrophic Examinoflon of a Huynes SP 16 Curculcr-Groove Weld- Crackmg-Test Specimen. Etchant: HCl + CuCl + alcohol. 100X. 68 PERIOD ENDING JANUARY 31, 1958 UNCLASSIFIED| Y24910 S INCHE 0.02 Fig. 2.1.23. Typical Area of Westinghouse 8M-1 Circular-Groove Weld-Cr‘acking-Tes.t‘Specimen. Etchant: HCI' + CuCl + alcohol. 100X, - . A UNGLASSIFIED | INCHES iy 6.02 e ng. 2.1.24. Haynes SP-19 Circular-Groove Weld-Crccldng-Tes! Specimen Showing Slight, Occasional, Fusion- Line Porosity. Etchant: HC! + CuCl + aleohol. 100X. 69 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT ~ UNCLASSIFIED O Ye24776 Fig. 2.1.25. Rod Weld-Cracking-Test Specimens Showing a Crack in a Haynes SP-16 Bar ond No Cracks in o Vestinghouse 8M-9 Bar. preparation of a typical test plate are illustrated in Figs. 2.1.26, 2.1,27, and 2.1.28. These plates provide specimens for mechanical property studies of welded joints ond are useful for radiographic, metallographic, and hardness studies, as well as for obtaining general information pertaining to the welding characteristics of the materials under conditions of high restraint, A summary of the status of the evaluations of the INOR-8 weld test plates made to date is presented in Table 2,1.12 " For hardness studies, samples of the welded test plates are removed and measurements are made on the weld metal ond base metal in the gs-_&vélded “condition and after aging. ' Preliminary work has consisted of determining the Rockwell B “hardniess of the two zones, but complete hardness traverses across the weld heat-affected zones are "being made to obtain a more complete understanding 70 of the influence of welding. The results of the preliminary hardness measurements on test plates 38 and 40 are presented in Table 2.1.13. in com- parison, the hardness of Hastelloy B and W weld metals in the as-welded condition is opproxlmately ' 20 on the Rockwell C scale, which is approxi- mately equal to a value of 98 on the Rockwell B scale.. After aging at 1300°F for 200 hr, the hard- ness of Hastelloy W weld metal rises to approxi- mately 32 on the Rockwell C st’:ale,lrwh_ileithat of Hastelloy B rises to approximately 41, It may be seen that heat 30-38 weld ‘metal is slightly- harder in both the as-welded and aged conditions ‘than Haynes SP-16 weld metal. How- ever, no significant hardening that can be attributed to aging is evident for either alloy, Both these alloys are- consuiercbly softer after aging than either Hastelloy BorW, 'y PERIOD ENDING JANUARY 31, 1958 BT = T - \ : B UNCLASSIFIED)| PHOTO 42156 ] £ OA K HIDGE HATIONAL L ABORATOR lg:chTAgsgggy - Approxlmateiy 40 bend tests have been made on ' Hasfelloy W weld metal and heat 30-38 weld metal at room ond elevated temperatures. These tests ‘were ‘made on.specimens - in the as-welded condi- “tion ” qnd on : aged specmens._ The testing and '-agmg temperatures used were room temperature and 1100, 1_200': ,1300 1500, and 1650°F, The data ; o '-shown in Fig. 2.1.29, The Hastelloy B mpple was n - . R I ~ part of the finish-machined Hastelloy B pump b 8 152 L~ = /'/‘I 150 e 1.48 - - ThE, 20 40 60 80 UR : UR {mole %) Fig. . 23.5 Indices of Refraction in the System The 7LiF-6ThF -7LiF6UF, solid solution is green and uniaxial. lts birefringence is low-and _varies from 0 to 0.006. The indices of refraction ~ are ~shown .in Fig. 2.3.9 as functions of UF, PERIOD ENDING JANUARY 31, 1958 : : _ ~ UNCLASSIFIED ' _ UNCLASSIFIED : . - ORNL-LR-DWG 27915 . ORNL-LR-DWG 27916 1050 1.62 ' . 1000 \'\\ LIQUID 1.60 \ 950 e — \ _ 1.58 LIQUID + Thfy~ UF SOLID SOLUTION | \.\ EQ'N‘ .'\ LIQUID + ThFy—UF, SOLID" 7k" . y 800 I"SOLUTION + LiF - 4ThF, = T LiF- 4UF, SOLID SOLUTION—] ‘ =~ 750 . - 50 LiF-4ThF4—LiF-4JUF4 SOLIID SOLUTION - ./ / 9200 1.56 | "] ) /l o INDEX OF REFRACTION \\ 1.54 TEMPERATURE (°C) 1.52 o - 7 0 20 30 40 50 60 70 80 UF, (mole %) 700 0 10 20 30 40 50 60 70 80 UF, (mole %) Fig. 23.7. Indices of Refraction Along the Join Fig. 23.6. The Join LIF-4ThF -LiF-4UF . LiF-4ThF -LiF-4UF . UNCLASSIFIED ORNL-LR-DWG 27917 850 800 - LIQUID ~. | | T 750 - *— ' NN ) | NN - LIQUID + LiF: 4ThF,~ LiF - 4UF, q TO00 p————S—PN———— SOLID'SOLUTION = —p— ] LIQUID+LIF-2ThE, N o L , - SOLID SOLUTION * Lk | S - LiF :2ThF, SOLID SOLUTION 1 - TEMPERATURE (°C) 650 —L— LIQUID + LiF - 4Th, ~LiF - 4UF, SOLID SOLUTION +—— = -———— —————————f_—_. ‘—-—-7' 600 fu | LIQUID + LiF -2Thf, SOLID SOLUTION + Y- | = 1{Lic - -2 Thi N T '_-"-Luo’u:_o +LiF- 4ThF, ~ LiF - 4UF, SOLID SOLUTION | - 7LIF+6ThF, — 7LiF - 6UF, SOLID SOLUTION/- *{ = |~ £ 7LiF - 6ThF, - 7LiF - 6UF, SOLID SOLUTION | - . TLIF-6THF,“7LIF 6UF, SOLID SOLUTION 500 0 5 .- A0 15 ZQ : 25 30 35 40 45 UF, (mole %) Fig. 2.3.8, The Join 7LiF-6ThF,-7LiF-6UF,. . L 85 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT UNCLASSIFIED ORNL-LR-DWG 27948 1.62 1.60 1.58 1.56 \ 1.54 ' o / 4/ INDEX OF REFRACTION 152 2 1.50 0 5 410 5 20 25 30 35 40 45 UF, (mole %) Fige 23.9. Indices of Refraction Along the Join 7LiF'6ThF4-7LiF-6UF4‘. ThF, content. As the UF, content increases, the solid solution changes from uniaxial positive to uniaxial negative. Further, for mixtures containing about 31 mole % UF4, the solid solution should be isotropic. Quenches of such mixtures have produced 7LiF.6ThF 7LiF.6UF, solid solutions that have birefringences which are too low to measure microscopically. The curvatures of the ThF -UF , solid-solution primary-phase field ond the L‘iF-41'hF4-LiF-4UF4 solid-solution primary-phase field, which were mentioned above, indicated that their boundary curve should bend away from the LiF vertex of the LiF-ThF‘-UF4 diagram, Experimehtal values confirm this indication, as shown in Fig. 2.3.10. UNCLASSIFIED . ORNL~LR—DWG 27949 a0 80 70 /\/ ThF,- UF, A SOLID SOLUTION 6\; 60 3 0\0 - . £ Qb & 80 40 LiF - 2ThE, SOLID SOLUTION 3Lif - ThF, . SOLID SOLUTION LiF~4ThF4—LiF-4UF‘w/\/\ SOLID SOLUTION \V /N LiF & ' - {0 20 \ 20 40 50 4LIF- UF, & 60 70 UF, (mole %) Fig. 2.3.10. The Primary-Phase Fields in the System LiF-ThF4-UF4. 86 \J In the system diagrom presented in Fig. 2.3.11, the temperatures of points a and ¢ are 500 and 488°C, respectively. Consequently the LiF- 4LiF.UF, boundory curve temperatures drop from point a to point c. The temperatures of points ¢ and 4 are 488 and :500°C, respectively. Conse- quently the boundary curve between the LiF primary-phase field and the 7Li F-6ThF ,-7LiF.6UF solid-solution primary-phase field drops in temper-- .ature from point d to point ¢. In addition, evidence exists (Table 2.3.2) that 4LiF.UF , appears as a third solid phase in compositions along this PERIOD ENDING JANUARY 31, 1958 boundary curve. There is also evidence (Table 2.,3.2) that LiF appears as a third solid phase in compositions along the boundary curve between the 4LiF-UF , primary-phase field and the 7LiF 6 ThF,- 7LiF-6UF ; solid-solution primary-phase field. Thus the boundary curve temperatures must decrease from point & to point c. Point c is, then, a eutfectic. The solid phases in equilibrium in compositions near point ¢ at the invariant temperature are LiF, 4LiF-UF,, and 7LiF-6ThF -7LiF.6UF ; solid solution containing about 41.5 mole % UF ,. From UNCLASSIFIED ORN{—-LR—-DWG 27920 LiF-2ThF, SOLID SOLUTION LiF-4ThF, - LiF-4UF, SOLID SOLUTION \ ' 7LiF-6ThE,— N 7LIF-6UFR, 'UI-;," (mole %) Y SOLID'SOLUTION -7 \/ L _ . | « 30 . 35 40 45 50 T 4LIF - UF, ' o R Fig. 2.3.11. Primary-Phase Fields in the .System LiF-ThF4-UF4 for the Region Representing Mixtures Con- taining 50 to 100 Mole % LiF. 87 MOLTEN.SALT REACTOR PROGRAM PROGRESS REPORT - Table 2.3.2. Results of Thermal-Gradient Quenching Experiments in the System LiFf_ThF“-UF‘ Composition Temperature of (mole %) : Phase Change Phusés Above Phase Change Phases Below Phase Change ‘ 7LiF-6ThF4 sqiid solution 1 e . . : o LiF | ThF4 l.lF4 ("C) 40 10 50 808 Liquid Liquid + LiF+4UF ,«LiF-4ThF , solid - o - solution 33.3 33;3 33;3 862 Liquid + UF4-ThF4 solid solution 'Liquid + LiF'4UF4-LiF°4ThF4 solid ' . o sofution - 40 50 10 880 Liquid Liquid + LiF°4UF4-Li F+4ThF, solid : solution 725 1 26.5 485 Liquid + 4LIF-UF, + 7LiF+6UF ;- LiF +4LiF-UF , + 7LiF-6UF -7LiF+6ThF , C 7LiF°6ThF4 solid solution solid solution 71 2 7 27 491 Liquid + LiF + 7LiF-6UF4- LiF ~l-~4Li-F-UF4 + 7LiF'6UF4-7LiF-6ThF4 solid solution this information the compatibility triangle at the invariant temperafure can be drawn as shown in Fig.2.3.11. The invariant falls within this triangle, and consequently, as stated above, point ¢ must be a eutectic. Solubility and Stability of PuF, ' Molten Fluorides C. J. Barton R. A. Strehlow Values for the solubility of PuF, in NaF-BeF, (57-43 mole %) at three temperatures were given in the previous report.d During the past quarter, values were obtained for the solubility of PuF two additional NaF-BeF mixtures, three LiF- BeF2 composmons, and one NaF- LlF -BeF, mixture ds a function of temperature. The range of BeF, concentrations covered in each of the binary solvent systems, approximately 36 to 50 mole %, includes _the concentration range likely to be of interest in the MSR program. Binary mixtures with lower BeF, concentrations have too high liquidus - B temperatures to be useful at power reactor temper- atures, and the viscosity of mixtures containing more than 50 mole % BeF, is too high to make them attractive. The ternary solvent mixture was included in the ‘study to permit solubility de- terminations with a lower BeF, concentration than would have been possible in the binary systems at the lower limit of the temperature range studied. 3¢, J. Barton et al,, MSR Quar. Prog. Rep. Oct. 31, 1957, ORNL-2431, p 39. 88 The results obtained to date in these studies are presented in Table 2.3.3. Some of the results are almost certainly incorrect and will be checked in new filtration experiments ‘as soon as possible. The possibility of analytical error was minimized in most of the doubtful results by repeating the analysis, either with a new portion of the same sample or a separate portion of the origineal sample., Except for the first few samples sub- mitted for analysis, which were ground but not sieved, all filtrates were ground either to ~65 or ~200 mesh and mixed thoroughly to avoid the possibility of inhomogeneity. A possible cause of the observed discrepancies in the data would be an error in the measurement of the temperature of -the liquid salt mixture. A fairly large vertical temperature gradient is known to exist in the heated zone occupied by the fused salt container, and it is possible that the thermocouple was accidentally shifted in the course of the experi- ‘ments, An additional uncertainty in the determi- nation of the melt temperatures resulted from the thermocouple wells in most of the filter bottles used in these experiments being shorter than the deéigh"'cu“ed for; the ends of the wells were not in contact with the melts. Improved filter bottles that have thinner walled thermocouple wells of the proper length to permit the " thermocouple junction to be surrounded by the salt mixture became available near the end of the quarter and were used for ‘the last two measurements made, . IJ' PERIOD ENDING JANUARY 31, 1958 Table 2.3.3. Solubility of PuFy in Alkali Fluoride—Beryllium Fluoride Mixtures Solvent Composition - Filtrate Analysis | Filfrdfien (mole %) Temperature: Pu l:'uF3 NaF LiF BeF, (°c) (wt %) (mole %) 64 36 550 .54 0.29 598 2.43 0.46 650 4.40 0.85 57 43 538 1.17 022 ' 600 1.36 0,26 652 2.16 0.41 50 _ | 50 552 1.77 0.34 ' 600 1.97 0.37 651 272 0.52 64 36 532 1.15 0.16 ' 600 1,82 0:27 643 4.30 0.63 56 44 550 | 1.98 0.30 ' ' 649 6424 | 0,98 5146 48.4 463 1.02 0.16 - | 549 2.44 0.38 599 2.88 0.45 654 5.76 0,93 56 16 28 554 7468 1.4 ~ mole % LiF-28 mole % Ber " nmproved boflles wlll be used to check some’ of thef:_f:'fi The solvent composmons given in The -actual compo-- . - “sifions’ have ‘not yet been estabhshed by chemical - ~analysis, except for ‘the 51.6 mole- % LiF-48.4 composmon,whlch is_the some mixture as that used for the determmuhon of the _which _is. dlscussed m a..ieii' -~ mole % BeF a solubnln‘y of CeF “subsequent secflon of this chapter. S - The data _given in ‘Table 2,3.3 show thut for" LiF- BeF2 ' that is, the measurements on the 51.6 mole % LLiF-48.4 mole % BeF and 56 mole % NaF 17 earlier data. Table 2.3.3 are theoretlccll B mixtures solubility of PuF is - hlgher in the mixtures with 50 mole % BeF qnd 36 mole % BeF than in the mixture with 43 mole % BeF Thus there is. an mdlccmon that the solubility of F’uF . ‘solvent goes through a minimum in the wcm:ty of “mixtures ‘with 43 mole % BeF.,. ,_‘vafue obtamed with the ternary mixture, which is -~ probably low because of the lack of a sufficient ~amount of PuF, to saturate the solutlon, mdlcates mlxtures. _ The - '.‘wnh decreasmg BeF plutonium in thls mlxture wasfound fo be combmed . -f‘os NaPuF 2 in the composmon ranger'-» "studled ‘the.. so]ublllty of l'-’uF3 increases with - _increasing - BeF2 concentration, “at . least at ‘the lower temperatures. - In. the NaF- Ber system, “the in this ‘The solubility that the’ solubfifly of PuF continves to. increase concentrcmon. All the “The - solubtllty of . PuF “at’ 565°C_ in_the blnary mixtures studied varied’ from about - 0.2 mole% for the 57 mole % NaF ~43 mole % BeF, mixture ~to 0,45 mole % for the -51.6 mole %_ -_!_fLiF--48.4 ‘mole % Ber mixture. tration. range is believed, at the present time, to - “be: more .than - adequate to fuei a molten: salt -fplutomum-burner reactors 2 Th:s concen- Another purpose of this investigation was the observation of the stability of PuF, in fused salt 89 MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT mixtures. This has been accomplished by de- termining the plutonium species present in cooled samples of filtrates and unfiltered residues produced by filtering mixtures that had been heated at temperatures of 550 to 650°C for periods that were usually in excess of 2 hr. The plutonium compounds present in the mixtures were identified by petrographic examination, and valence de- terminations were made by spectrophotometric examination of dissolved portions of the samples. Only trivalent plutonium has been found in the mixtures examined to date. In addition, the bottom portions of two nickel filter bottles used in ~ filtration experiments were submitted for analysis after removal of all except a trace of fused salt. One sample had been in contact with the NaF- BeF ,-PuF, mixture ot 600°C for about 2 hr and the other had contained the LiF-BeF ,-PuF “mixture at temperatures varying from 500 to 650°8 for approximately 8 hr. The plutonium found in the two samples amounted to 0.05 and 0.17 mg, respectively, and could be accounted for by the trace of fused salt remaining on the nickel. No evidence of disproportionation of PuF, in fused alkali fluoride-beryllium fluoride melts under the conditions maintained in these experiments has been observed to date. FUSED CHLORIDES AS SECONDARY HEAT TRANSFER FLUIDS R. E. Moore The physical, chemical, and nuclear properties of a secondary coolant for a molten.salt power reactor must satisfy a number of requirements. The following characteristics are desired: 1. low melting point, 2. low viscosity, 3. high heat capacity, 4. high thermal conductivity, 5. low vapor pressure, 6. stability toward structural metals, 7. -reasonably low thermal-neutron cross section, 8. freedom from activation in a radiation field, _Fluorldes and chlorides seem to offer the best ~choices among possible salt systems. Melting points of 300°C and lower can be obtained with fluoride mixtures containing ‘BeF,, but the viscosities of these mixtures are hngh. If a low melting point accompanied by a low viscosity is to be obtained it is necessary to turn to chloride systems, ) 90 - A large number of chloride systems have been described in the literature, but, after an elimination process - based on gross deviations from the properties listed above, there remained only a few systems for further consideration. The eutectic compositions (mole %) and melting temperatures of these systems are listed below: Melt.ing VComposiiion_ Temperature (°c) 41.7 mole % RbC|-58.3 mole % LiCl 318 © 64 mole % ZnC|2-36 mole % SnCI2 : 171 - 38 mole % KCl—62 mole % SnCl, 180 29 mole % KCI-71 mole % ZflCl2 262 23 mole % LiCl~77 mole % ZnCl, 294 47.5 mole % RbCl-52.5 mole % ZnCl, 249 The very low melting points of the last five mixtures listed would seem to make these compo- sitions especially attractive. All these mixtures contain, however, either .‘:'mCl2 or ZnCl,, which have high vapor pressures, although the pressures are possibly not high enough to interfere with pump operation in the coolant circuit. A calcu- lation of the ideal vapor pressure of the 29 mole % KCI-71 mole % ZnCl, mixture, based on literature values for ZnCl,, gave values of 140 mm Hg at 650°C and 1.4 mm Hg at 450°C. The actual pressures may be somewhat lower than the ideal. Yapor pressures of the mixtures containing SnCl, would be expected to be much higher than those' for mixtures containing ZnCI2 because the boiling point of SnCl (623°C) is considerably lower than the boiling pomf of ZnCl (732°C) - Another 1mportant questlon concerning compo- sitions containing SnCl, or ZnCl, is the possibility of reaction with structural metals. Calculations based on vqlues of free energies of formation? indicate that considerable reaction may occur, especially in the case o{:'SnCI2 mixfures. There is enough uncertainty in the calculdtions, however, that static corrosion tests should be made. Plans are being made for experiments in_which Inconel will be exposed to molten mixtures containing ZnCl, and SnCl, in seoled containers of fused silica. 4A. Glassner, A Survey of the Free Energies of Formation of the Fluondes, Chlorides, and Oxides of the Elements to 2500°K, ANL.-5107 (Aug. 1953) 1) ,!f\ N The eutectic composition41.7 mole % RbCI-58.3 mole % LiCl appears to be the most attractive from the standpoints of vapor pressure and corrosion. The boiling points of RbCl and LiCl (1390 and 1353°C) are sufficiently high that the vapor pressure of the mixture should be negligible at reactor temperatures. Calculations based on values of the free energies of formation? indicate that the mixture should be extremely stable in contact with Inconel. Thermal-convection loop tests should be made to determine the rate of mass transfer. The new low prices of rubidium salts (quoted by American Potash & Chemical Corp. at $13.00 to $27.50 per pound) may remove the cost objection to the use of rubidium chloride. The treatment of RbCI-LiCl mixtures to remove water, which is always present in the salts, requires vacuum drying, grinding, and contacting with HCl during slow heating. Once the hydrolysis reaction Cl= + H,O==0H" + HCI has proceeded to the right, the reaction is not readily shifted to the left.®> An apparatus for carrying out such treatments is to be constructed. VAPOR PRESSURES OF LiF-BeF, MIXTURES F. F. Blankenship It was anticipated that at MSR operating temper- S. Cantor atures the vapor pressure of the fuel mixture would - be low, but, since it is important to know the magnitude of the vapor pressure, the total vapor -pressure of a fuel solvent composed of 64.9 mole % LiF and 35.1 mole % BeF, was measured by using the Rodebush Dixon® method. The equcmon thot Hits the dato, whlch are gwen in Toble 2.3.4, is. 1 0,050 logp(mmHg)---rsm o .‘.there T is temperotu:e in’ °K, A Ilneor extropo-‘ ~ lation of the data to temperatures of reoctor mferest i is presented in Tobfe 23.5. Ik , From’ the - vapor. pressures of the NoF BeF ; Lo system,? it was found that a change of 5 mole % . . m the h|gher bodmg component lowered the 1otal," 5H A, Lumnen, W S. Ferguson, ond R. A, Osteryoung, ' J. Electrochem. Soc. 104, 516~20 (1957). SW. H. Rodebush and A. L. Dixon, Phys. Rev. 26, 851 (1925), . PERIOD ENDING JANUARY 31, 1958 Table 2.3.4. Vapor Pressures of the 64.9 Mole % LiF.35.1 Mole % BeF, Mixture ~ Temperature Pressure (°C) 4 (mm Hg) 96846 4.9 1017 ‘ 9.4 1039 : 12.9 1066 18.1 1095 25.8 - 1128 40.0 1146 47.3 1160 55.4 1182 71.8 1194 83.0 1203 88.8 Table 2.3.5. Extropo_loted VYapor Pressures of the 64.9 Mole % LiF~35.1 Mole % Ber Mixture Temperature Pressure (°C) (mm Hg) 500 0.000058 550 0.00036 600 0.0018 650 0.0076 700 0.027 pressure by approximately one-half. Similar behavior can be expected in the LiF-BeF, system. For instance, the vapor pressure of a 70 mole % LiF-30 mole % BeF, soclution would be about ~one-half the vapor pressure’ of -the 64.9 mole % LiF=35.1 mole % Ber soluhon ot fhe same. temperature. ' S ' S FUEL REPROCESSING _ G.M Wofson - F.F. Blankenshlp Solubilnty of Noble Goses in Molten Fluorlde ' M:xtures P : "NV, szth Numerlcol values of the solublhhes of hellum, “neon, argon, and xenon in NaF-ZrF (53-47 mole %), - expressed as Henry s Iow constonts, were presented 7K A, Sense, R. W, Sfone, and R. B. Fiibert, Jr., Vapor Pressure and Equilibrium Studies of the Sodium f‘éu;;zde-—Beryllzum Fluoride System, BMI-1186 (May 27, 5 AN - previously.8 Measurements on the solubility of - helium and of neon in NaF-KF-LiF (11.5-42-46.5 - mole %) have now been concluded, and the experi- - mental results are summarized in Tables 2.3.6 and 2.3.7 and are presented graphically in Figs. 2.3.12, 2.3.13, and 2.3.14. The solubility study was undertaken with the NoF-KF-LiF mixture as the solvent pending the completion of facilities for studying solvents containing BeF . The solvent NaF-KF-LiF has a liquid Structufé which is quite different from that of the’ NaF-ZrF mixture, and the structures of mixtures contcm:ng BeF , probably range in between. Accordingly, it may be safe to expect that the numencal values 8J. H. Shaffer et al., MS'R Quar, Prog. Rep Oct. 31, 1957, ORNL 2431, p 47, MOLTEN-SALT REACTOR PROGRAM PROGRESS REPO_RT of the noble gas solubilities in solvents con- taining BeF will ‘be less than the corresponding values . f\lcF ZrF ‘but more than those in NaF-KF- L|F. , ‘ The results presented in Tubles 2.3.6 and 2.3.7 show the same trends as those prev:ously observed for solubilities in mixtures containing ZrF The solubilities follow Henry's law, mcrease with - increasing - temperature, and decrease with in- . . creasing molecular weight of the gos. The heats “of solution for the helium and neon gases in this solvent were calculated to be 8000 and 8900 _calories per gram-mole of gas. The numerical _magnitudes ‘of the solubilities in NaF-KF-LiF ‘are -roughly” 50% of the corresponding values in ~ NaF-ZrF . ~ A more detailed analysis will be made when the measurements ‘of the solublhty of argon in thls solvent are completed Taoble 2.3.6: Solubility of Helium in NaF-KF-LiF ('lslo5442-46-'5 Moie'%) . Saturating S Temperature Pressure Solubility K* (°C) (atm) -+ (moles of helium/cm® of melt) ‘ x 10~8 x 10-8 600 2.08 21 106 1,77 190 - 10.7 1.51 16,9 - ' 1.2 1.00 1.0 | 11,0 1.00 12.8 - 128 AV 13t 07 650 2,08 285 o 13g 2.04 34.8 - A 1.75 30.8 178 0.98 176 ' 1749 | | | AV 17,5 1 0.2 L S 2.06 483 235 | 2.04 48.2 ) 23.6 1.77 a8 - 2346 1.51 336 o 22.3 0.99 S 217 21 Av 230 £ 0.7 *K = c/p in moles of gas per cubic centimeter of melt per atmospheres wi 2} " PERIOD ENDING JANUARY 31, 1958 Table 2.3.7. S&I_dbflity of Neon in NaF-KFaLiF (11.5-42.46.5 Mole %) " Saturating s'| . olubility Temperature Pressure K* _ (°C) (atm) (moles of neon/em’ of melt) x 10~8 x 10~8 600 207 9.57 4.61 1.49 6.42 4,30 1.01 4.20 4.16 ' Av 436 t 0.20 700 2.05 15.06 7:36 ' 1451 11.04 7.33 1.02 7.99 7.84 Av 7.51 t 0.22 800 : 2.07 2252 10.89 1.50 16+53 11.00 1.03 12.05 11.66 ——————n. Av 11.18 * 0.26 - SOLUBILITY (moles of He/em of meit) *K = ¢/p in moles of gas per cubic centimeter of melt per atmosphere, _ UNCLASSIFIED 16 UNCLASSIFIED (x10~7 ORNL—LR—DWG 27922 0 ORNL—LR--DWG 27921 4.0 6.0 S$ y 3.0 S ’ oG 5.0 «° / ° G 109 4.0 SOLUBILITY (moles of Ne/cm3 of melt}) e N o o / .0 .- 080 100 150 200 - 250 3.00 SATURATING NEON PRESSURE (atm) o 3.0 20 Fig. ?.3.13. Solul:llify of Neon in Molten NoF-KF- "-"_LIF (11.5-42-45.5 Mole % o Solublllty of HF in Molten Fluorldes o | : s _ , " J.H. Shaffer | O - 1050 CThoo o 200 260 Meosurements are being made of the solubility o .SATURATING HELIUM PRESSURE (atm) . o of HF “in various. molten fluoride solvents as a o s TR ald funchon of temperature, pressure, and solvenf . Fig. 23.12, Solubilny of Hellum in Molten NuF-KF- composition. The effects of composition of mixtures LiF (11,5-42-46.5 Mole %). in the NaF-ZrF4 system are shown graphically in 93 -~ MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT UNCLASSIFIED ORNL—LR—DWG 27923 TEMPERATURE (°C) _; 800 700 600 3X10 — 2 ey ny = 8000 cal/mole .N€ON 8900 cal/mole K [moles of gas/(cm® of melt) (atm)] L] o 9.00 40.00 11.00 10Y/7 (°k) 12.00 Fig. .2.3. 14, Temperature Dependence of Sclubilities in NaF6.5 R fi . Radiochemical *Compositions of saturating phases calculated from indicated end points of titration curves obtained with CaO as the titrating agents 99 ' MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT 2,3.23 show that uranium is precipitated essentially quantitatively as UO, before any of the Ce203 is formed. experiments wete determined radiochemically, while the uranium concentrations ‘were determined by chenical analysis. In the titration of the mixture ZrF CeF, with CaO, as shown in UNCLASSIFIED 5 ' ORNL-LR -DWG 27931 e 4 ._———.-"— R _Ce (DETERMINED RADIO- _| ©7 CHEMICALLY) N\ \ 0 0.1 0.2 0.3 04 0.5 0.6 ' EQUIVALENTS OF CoO ADDED ) \ ' METAL FOUND (wt %) UF, END POINT——- — .| U (DETERMINED "CHEMICALLY) \ / Fig. 2.3.22. Titration of CeF3 and UF4 with Ca0 in LiF-KF (50-50 Mole %). UNCLASSIFIED ORNL-LR-DWG 27932 5 | | l o —U(DETERMINED CHEMICALLY) 4 \\ UF, END POINT ———= @ z3 ‘ N o ) ® = =2 O L. 32 '.— b N I 1 |————— Ce {DETERMINED . / RADIOCHEMICALLY) ¢ _ T Q T '<\ : . @ . i 0 , . .| 0 1005 O0t0 - 0Of5 0.20 0.25 " EQUIVALENTS OF CaQ ADDED. Fig. 2.3.23. Titration of CeF, and UF, with CaO in LiF-KF (50-50 Mole %). 100 The concenfratmns of cerium in these - | Flg. 2.3.24, the seporatlon does not appeor to be quite as sharp as in the case of the UF ,-CeF mixture, as shown in Figs. 2.3.22° and 2.3.23. The htrahon with CaO of ZrF in the presence of nonradsouchve CeF3 is shown in Fig. .3.25 UNCLASSIFIED ORNL-LR-DWG 27933 2.0 — 9 Zr (DETERMINED CHEMICALLY) . 1.5 : , 3 \ = L E‘ I er4 END POINT —= 2 10 _ 4 = W E. Ce{DETERMINED RADIOCHEMICALLY) 0 - 0.05 - 040 0.45 0.20 EQUIVALENTS OF CaO ADDED - Fig. 2.3.24. Titration of Cer in LiF-KF (50-50 Mole %). and ZrF4 with Ca0 UNCLASSIFIED ORNL-LR—DWG 27934 2'06 I : aa—2Zr(DETERMINED RADIOCHEMICALLY) 5 , ZrF4 END POINT —= _ \ - B2 | ‘\ : 2 > e S 10 ‘ = 2 = Q £ xz . N 05 =\ o 005 ‘040 015 0.20 ‘EQUIVALENTS OF CaO ADDED - Fig. 2.3.25, Titration of ZrF in the Presence of CeF with Ca0 in LiF-KF (50 50 Mole %) 7 In order to determine the concentration of zir- conium radiochemically, the zirconium was labeled with Hf181 and the hofnium was Gssumed to behave like zirconium. The titration curve shows the assumption to be sound. The titration of the mixture Cer-BeF with CaO is shown in Fig. 2.3.26. From the curve it may be seen that cerium and beryllium are precipitated simultaneously but not in definite proportions; they probably precipitate as o solid solution. The results are summarized in Table 2.3.13. [t has been suggested!? that the results of this experiment indicate that an attempt should be made to precipitate cerium from the solution by adsorption on an excess of BeO. Accordingly, attempts are being made to precipitate not only cerium but also uranium. Lithium Recovery from NaF‘Cr(5) + QHF(g) -—80x10' . Comparison of these constants with the average value that was obtained for the equilibrium ratio, (1.45 £ 0.17) x 10-3, yielded values of activity coefficients of CrF of 0.284 and of 0.181 at 16, Brewer et al,, NatlL Nuclear Erzergy Ser. Div. IV 198, 65, 109, 201 (1950). +% . iyt PERIOD ENDING JANUARY 31, 1958 -3 UNCLASSIFIED (x 1077) . ORNL— LR—DWG 27937 4.0 l ‘ | 3.5 - Ky =(1.452047)x10 ® DETERMINED UNDER REDUCING CONDITIONS —~3 & DETERMINED UNDER OXIDIZING CONDITIONS 30 Ky(s)=51 x 10 Ko (/1 =18.0 x 1073 y{s}= 0.284 2.5 y(/}=0.181 % o 2.0 " i £.5 e e 1.0 A 0.5 0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 (x1072 X Cer Fig. 2.3.28. Equilibrium Quotients at 850°C of the Reaction Cer(a') + Hz(g) = Cr(s) + 2HF(g) in NuF-Z‘rF‘. (53-47 Mole %). 850°C with respect to the solid and the super- cooled-liquid standard states, respectively, Ex- perimental work is under way at other temperatures. Solubility of F"eF.2 in LiF-BeF, R. J. Sheil Before attempting to determine activity coef- ficients of FeF, ~tration range below which the FeF2 does not precipitate as @ ‘complex compound or in the pure - For this mvestlgahon, 'known amounts offifi ~ Sstate. _iron were added, as FeF,, to the solvent, ond filtered samples of the solz “added. separafe oddrtlons of FeF, Table 2.3.17. The tobulotecf results indicate that -equ:llbrwm measurements can be made w:th -$0- lutions of FeF, well in excess of 5000 ppm Fe'* which is the lowest temperature contemplated for the experiment. To_ble 2.3.17. Concentration of Iron Dissolved in LiF-BeF2 {63-37 Mole %) ot 500, 600, and 700°C Fe++ Found in Filtrate (ppm) in molten LiF-BeF, (63-37 " mole %), it was necessary to estabhsh a concen- - Fe'' Added (ppm) At700°C At 600°C At 500°C 5,000 4,925 4,835 4,845 10,000 9,386 9,125 9,390 141,700 33,100 * 50,000 49,100* ution were obtained at = _ different temperatures. The concentrations of iron - - in the filtrates; as determined by chemical analysis, =~~~ ‘were then compored with the ‘amounts of Fer":"" The results of the compartsons for these - in this solvent ot concentranons' . even at 500°C, - *Pefi-ographie examination of this filtrate revealed that ._the sufurnhng phase was pure FeF2 Use of Crs ' To Study Cbromaum Mlgrofion in Polythermul Inconel-Molten Sulf Systems Ro Bt Evans - . A proposed grophlcal method for colculatmg the - ‘are summarized in. rate and amount “of : migration: (corrosion) to be " expected ’ becouse of the reochon C " Inconel4 ‘ = o (-d’ 7+« dat Cr0 < It Jeox o 967 | olo | 303 [/0.2527¢ 14,{{6”3': t+6.9xi0TH I aM : ’ ! (d}') Q o \er : - ! . 77777 STTTIT V ‘952 952 | 252 /025 A 1x10-87 . ) i _ ////// ; am - l (df) >1 l T —> . o ¥ = X (.2m & - , | (—;,—) = NET DIFFUSION RATE OF ALL B E r CHROMIUM | - - |/ , . % MELT IN CONTACT WITH WALL. (—d—,—) = NET DIFFUSION RATE QF UNLABELED S , ce® CHROMIUM SURFACE LAYER ASSUMED TO BE - \q? .= NET DIFFUSION RATE OF LABELED ’ Cr°* . CHROMIUM ~Fig. 2.3.30. Loop Wall Segments, . curves as those obtained experimentally (Flg. 2.3.29) because (1) Cr®* is always diffusing into the walls along the entire loop, regardless of the sign or value of the other flow rates, and (2) the Cr® diffusion rate is highest in the high-temper- ature region,. since (dM/dt)Cro, is proportional to (AC o,oD‘/z) and D172 jncreases with temper- ature. It may be concluded that experiments similar to the first three listed in Table 2,3.18 are of little value for corrosion studies because = ~ the Cr% distribution pattern is controlled by the wall exchange reaction and wall temperature - rather than by the corrosion reaction of interest. In the fourth experiment, Cr5! was mmally present both in the salt, as Cr*F o and in.the walls, as Cr°, The labeled chromium migration “pattern in this case tended to follow the over-all " migration poflern of the . corrosion, as shown grcphlculiy on Fig. 2 3. 3] Comparisons of ~at the inner surface of the tubing is very close _to the salt temperoture and that this temperature controls * the value of the ethbnum constant IN EQUILIBRIUM WITH MELT i1 INFINITE INCONEL TUBE Theoretical Concentration Relationships for Chrcmlum in Vurlous Inconel Thermul-Convechon '.Figfi.: 2.3.31 and 2.3.32 - demonstrate that the - corrosion pattern calculated by assuming that T, att)x=g = Tyqy is in fair agreement with the pattern -obtained experimentally. - mental curve of Fig. 2.3.31 is not symmetrlcal - and the point of maximum deposition appears to be . displaced downstream - from the maximum wall ‘_femperature at ‘section A.- features characterize the curve of Fig. 2.3.32. This strongly suggests that the wall temperature ' for the corrosion reuchon. The ability toreproduce the shape of the activity vs wall-temperature plots for the Battelle thermal loop experiments — particularly for the fourth ‘experiment - by means of calculations based on - the predlcted behavior suggests that the proposed The experi- - In general, the same O - of !nconel. i iy PERIOD ENDING JANUARY 31, 1958 UNCLASSIFIED ORNL-LR—-DWG 27940 f'mlgratlon mechamsms gwe a reasonable expla-'_‘-fil.?-' ' nation of the - steady-state chromwm mlgratlon "'wnhtn thermai I°°Ps",;-,r.~' . s T Actuvrty of Nrckel m NlekeI-Molybdenum Alloys ‘_>;_ : 2. mechamcal ag|tat|on has no . effect on- the 3. Longer Measurements are bemg made of the actwnty ef nrckel in mckel-moiybcfenum u”oys by using an -The: rehab:llty af T -',47, 138 (]951) electromotive - force method. easurements made wnth emf cells of the type NlCl'z(] wt ’%) in ~ Ni(metal) AR NaCl-KCl eutectic NI-MO alloy Ly . B’ A c! 2 160 — Z T 1 ~ : | o l L |8 & 150 l ~ Z 140 5 1, - a 4100 |- WALL ™=, - 3 g 'r. - - SALT 2 % 50 / el S 120 2 \/ / . \ O ) o . o O 25 50 715 100 125 § % OF LOOP LENGTH % 100 F——x% — = — N - FLOW /,/ —— 3 / \\ < L - / O \ 2 ao 1 A A s - /7 5 - / o : : ' 7 N\ \ / q [ curee - THERMAL ASSUMPTION —|" / O - \ / g Twarr] = TeaLt \ . - g 0 mee——= Toarr] = TwaLL] -~ ¥ 7 _ / . = ‘X =0 DENOTES INTERNAL TUBING SURFACE ] LJ | -X—=o DENOTES EXTERNAL TUBING SURFACE \\l / . g' 20 — ; - 7 8 \ / L) . = . .2 . : ] \ 5 1 - & 0O . - 5 0 20 - 40 60 80 100 BO 60 40 20 0 _ PERCENTAGE OF LOOP TEMPERATURE DIFFERENCE T(mm) S ‘ T(max) : ‘ T(min) Flg. _23 31 Calculated Loop Wali Dlstribut:on af Chromium After Diffusion-Controlled UF -UF Corrosion _f;'.-ls dependent upon the reversrblhty of the ce!l . | "The “usual criteria18-20_ accepted as evidence '~ that the cell is’ ‘behaving reversibly are that Yithe: potentcal remains, constant w:th time at - ;-.flconstant temperature, ' G '5_' j;k—i’potentral 7 3. the . potential - of the cell returns to its “equi- ""”f:ilubrlum value rapldly after palarlzatlon, 184 F, Elllot nnd J Chipman. TranS- Fafaday Soc. : 9K W, Wagner, Tbermodynamzcs of Alloys, p 921 —97 Addison-Wesley, Caombridge, 1952. 2G5, F. A. Kortum and J. O'M Bockris, Textbook of Electrecbem:stry. p 581, Elsewer, New York, 1951, m MOLTEN-SALT REACTOR PROGRAM PROGRESS REPORT UNCLASSIFIED ‘ B A C ORNL~LR-DWG 27941 120 : ] | 3 100 _ e w o< ® e gg | p . FLOW & . 80 ) \\ . ZI._L [ O > \ @ EE | | . aZ S ® ¢ ok 60 o - < : ® tn = | 33 = = 40 : / zs 1 ¢ o2 EXPERIMENT 4: Cr INITIALLY _-n‘_)?z PRESENT IN WALL AND IN SALT _ g E \ o Z 20 .‘ /‘ . 3 0 | &L 0. 20 40 60 80 {00 80 60 40 20 0 : PERCENTAGE OF LOOP TEMPERATURE DIFFERENCE , T(mvin) 7 (max) 7 (min) Fig. 2.3.32. Effect of UF4-UI"'3 Corrosion on Cr° | Distribution in Inconel Loop. . 4, the observed potentials are reversible with _ . respect to temperature changes, 5. there is no difference of potential between two apparently identical electrodes. The first condition has been met by all cells tested. Also, mechanical agitation of the elec- trodes has had no effect on the observed emf. The cells have also returned to their equilibrium potentials within 30 min ofter all four electrodes have been shorted together. In most cases, the electrodes are within 2 mv of their equilibrium potential within 5 min after shorting for as long as 15 min. It was found that concentration cells in the nickel-chromium system were not readily ~ reversible with respect to temperature changes.?! " However, the first attempt to run a cell at two different temperatures in this study was suc- cessfuls A cell with alloy electrodes containing 2Vyork done by M: 8. Panish, ORNL. 112 10 oand 20 ot. % molybdenum gave constant po- tentials at 800°C for 15 days and for 6 more days at 700°C. It is planned, in the future, to ascend and descend the temperature scale between 750 and 1000°C with all cells. . The criterion of reversibility, however, has been the cause of the early termination of several cell experiments. Each cell as presently run contains two alloy electrodes (usually of different compo- sition) and two pure nickel electrodes that were annealed at 2200°F for 16 hr after fabrication. in general, one nickel electrode becomes erratic either shortly after the start of a run or after behaving properly for as long as 12 days. Potentials as high as 70 mv have been observed between two nickel electrodes which two days: previously 'had been behaving properly ond exhibited no potential difference. = Cells con- taining four nickel electrodes from different sources are now being run in an effort to discover the cause -of this anomalous behavier, No explanation is evident to date. S i contained BeF PRODUCTION OF PURIFIED MIXTURES Jq Po Blakely Gl J- Ness‘e Preparation of Pure Fluorides Some transition-metal fluorides of high purity were prepared for use in studies of corrosion, chemical equilibria, and phase relationships, ond for use as reactants in the preparation of other flucrides, including complexes. Nearly a kilogram each of ferric, ferrous, nickelous, and cobaltous fluorides were prepared. Ferric fluoride was prepared by passing anhydrous hydrogen fluoride gas over anhydrous ferric chloride at about 300°C. For preparing the ferrous, nickelous, and cobaltous fluorides, the commercially available hydrated chlorides were partially dehydrated at about 100°C and treated with hydrogen fluoride at about 400°C, Small+Scale Purification Operations C. R. Croft The experimental facilities were used for the processing of 530 kg of salt mixtures in 55 batches that consisted of 5-, 10-, and 50-lb quantities.. Twenty-eight of these batches were materials that g+ Included in the total were 11 batches of the NoF-KF-LiF-UF, mixture for use by the Metallurgy Division. Requests for molten ‘salts and especrally for BeF ,-bearing mixtures have reached a level which is beyond the production capabilities of the experimental facilities. It is anticipated that the 250-1b-batch production equipment will be operated at intervals during the ‘next severul months to e 'prowde the material reqmred ; “Four. batches totaling 270 kg of LIF-Ber-UF‘ - mixture were prepared during a trial run of thef' ~.250-1b- batch production equipment for two weeks : durmg December. This operation, on a three- , o shift, flve-day-week b05|s,' was - conducted - o .~ cooperation ~ with the ~Y-12 industrial Hyg:enel_ ':"'Deportmenr, and face. masks and complete pro- " tective - clothlng were - used by the operating " personnel, Careful monitoring by the ‘industrial - hygienists-: reveuled no,_significant exposure of the - - personnel to air-borne berylilum compounds.r The .- occurrence . of . severe “dermatitis among the operators, however, indicated that routine operation of the equipment under the conditions of this test PERIOD ENDING JANUARY 31, 1958 was not feasible. It will, apparently, be necessary to enclose the equipment and to provide an improved ventilation system before its continued use can be odjudged saofe in all procticable circumstances. Anclyses indicated only 3 to 4 wt % uranium in the final product, rather than the 6.4 wt % charged to the reactor. ' Examination of the residue in the reactor showed high concentrations of uranium as UO,. It is apparent that contact between HF and the liquid in the large reactor was insufficient to remove H,O and BeO from the mixture and that precipitation of UO, by the reaction UF, + 2BeO —>UO, + 2BeF, resulted. Experiments are under way to develop a modified technique for preparing satisfactory material with this equipment. Most of the available material con be used in experiments for which the precise compositions requested are not essential. -Preparation of Material for InsFile Loop - F. A. Doss A single batch was prepared that totaled about 300 g of LiF-BeF ,-UF , mixture in which Li7 and U235 were used. The solvent mixture of Li’F and BeF, was purified in Building 9928 at the Y-12 Plant. When analyses of specimens from “this preparation had indicated its purity to be “satisfactory, the material, in its original con- tainer, was taken to Building 9212 at the Y-12 Plant where the U235F was added and the final _purification was occompllshed Chemical analyses . .made by Y-12 and ORNL laborutorles were in good agreement as ‘fo. composmon .of the melt for '_7accountab||sty purposes. . The material has, - ~accordingly, been trunsferred to the requestor (see Lo Chap. 2.2 of thls report) Trunsfer and Servrce Gperatlons F. A Doss - Flfty-flve flllmg and . dramlng ‘operations were ':'.',performed during the quarter. Requests for such “services declined to a low of 12 during December ‘and -then increased to 25 during January. About 640 kg of salt and about 300 kg of alkali metals were transferred in these operations. 113 10 1. 12. 13. 14, 135. - 16. 17. 18. 19. 20. 21. 22. 23. 24, 25. 26. 27. 28. 29. 30. - 31-33. Y 35, e o 39 S A 42, S s, \\ VOENO LA LN a0 . Alexander . Bettis . Billington . Blakely . Biankenship lizard . B. Briggs . 0. Campbell . W. Cardwell . H. Carr . |l. Cathers E. Center (K-25) A. Charpie J. H. Coobs J F E. A C ., G M. E. J. Breeding R D D W G C. R. 'F. L. Culler J- Ho DeVan L. B. Emlet (K-25) V. K. Ergen J. Y. Estabrook 'D. E. Ferguson A. P. Fraas E. A. Franco-Ferreira J. H. Frye, Jr. A. T. Gresky J. L. Gregg W. R. Grimes E. Guth - - "C. S. Harrill 36,0 | .A. Hollaender -~ - = -A.S. Householder s < W.H.Jordan . = G. W, Keilholtz H. W. :Hoffmuh C.P.Keim M. T, Kel!ey 7 F. Kerfesz o | B. W, Kmyon *_ e M E. Lackey INTERNAL DISTRIBUTION 46. 47. 48. 49. 50. 51. 52. 33. 4. 55. 56. 57. 58. 59. 60. 61. 62. é3. 64. 6é5. 66. 7. é8. &9. 70, 71. 72. 73. 74. 78. 76. 77. 78, 79, 80. 8l 82 83, . .100.- 101 ]03._ EXTERNAL DISTRIBUTION o 104. Division of Research and Development, AEC, ORO v 105-681. Given distribution as shown in T1D-4500 (13th ed. Rev.) under Reactors-Power category ’ . (75 copies = OTS) ' : ORNL-2474 UC-81 = Reactors=Power TID-4500 (13th ed., Rev.) February 15, 1958 vingston ac Pherson . Morgan . Murray (Y-12) L Nelson . Patriarca A. M. Perry D. Phillips. P. M. Reyling J. T. Roberts M. T. Robinsen H. W. Savage A. W. Savolainen J. L. Scott E. D. Shipley M. J. Skinner A. H. Snell J. A. Swartout A. Taboada E. H. Tayler R. E. Thoma F. C. VonderLage G. M. Watson A. M. Weinberg :" ' M. E. Whatley _G. D. Whitman - G. C. Williams .C. E. VWinters ~Jo Zasler ORNL Y-12 Technical Librury Document Reference Section Laboratory Records Department - Laboratory Records, ORNL-RC Cenfrd | Research Library 115 r S ) ORNL.-2474 - _ UC-81 -~ Reactors—-Power Contract No. W-7405-eng-26 MOLTEN-SALT REACTOR PROGRAM QUARTERLY PROGRESS REPORT For Period Ending January 31, 1958 H. G. MacPherson, Program Director DATE ISSUED lwlfiJ MAY 141958 e omc RIDGE NATIONAL LABORATORY - - Qak Ridge, Tennessee - -operated by UNION CARBIDE CORPORATION for the U.S. ATOMIC ENERGY COMMISSION g 1ok e "