i CENTRAL RESEARCH LIBRARY || | |||{"|H || II DOCUMENT COLLECTION 3 4456 0361372 O - ORNL=2723 Reactors=Power MOLTEN-SALT REACTOR PROJECT QUARTERLY PROGRESS REPORT FOR PERIOD ENDING APRIL 30, 1959 CENTRAL RESEARCH LIBRARY DOCUMENT COLLECTION LIBRARY LOAN COPY DO NOT TRANSFER TO ANOTHER PERSON I you wish someone else to see this document, send in name with document and the library will arrange a loan. OAK RIDGE NATIONAL LABORATORY operated by UNION CARBIDE CORPORATION for the U.S. ATOMIC ENERGY COMMISSION /LY ORNL-2723 Reactors — Power TID-4500 (1h4th ed.) Contract No. W-7L05-eng-26 MOLTEN~-SALT REACTOR PROJECT QUARTERLY PROGRESS REPORT For Period Ending April 30, 1959 H. G. MacPherson, Project Coordinator DATE ISSUED JUN 191953 OAK RIDGE NATIONAI LABORATORY Oak Ridge, Tennessee operated by UNION CARBIDE CORPORATION for the U.S. ATOMIC ENERGY COMMISSION GY SYSTEMS LIBRARIES (TR 3 yysk 0363378 0 s CONTENTS SUWARY -o----.o-aooc-o.-o-ooo-.-eo-oo-on-.o-.olc-novnooo-o.oo. Vi l.l. 102. PART 1. REACTOR DESIGN STUDIES NUCLEAR CALCULATIONS AND DESIGN STUDIES .seeesecoscosnses Nuclear CalculationsS eeessseseseersscasssasascsscssacans Effect of lIon-Exchange Processing of Rare-FEarth Fission Products on Performance of Interim Design Reactor Fueled with U233 L..ivevieveennnnns 3 Nuclear Performence of a One-Region, Graphite- Moderated, Unreflected, Thorium-Conversion, Molten-Salt-Fueled Reactor seicecececnscacnsennes 3 Initial Nuclear Performance of One-Region, Graphite-Moderated, Graphite~Reflected, Thorium- Conversion, Molten-Salt-Fueled Reactors ......... 9 Nuclear Performance of Two-Region, Graphite- | Moderated, Molten-Salt~Fueled, Breeder Reactor .. 10 Effects of Fast Neutron Reactions in Bed on Reactivity of Molten-Salt Reactors veeeesececess. 1k Oracle Code MSPR~Cornpone 020 .teseeseccscossacasese 20 Design of 30-Mw Experimental ReacCtor secsieececccssnsccce 21 F.llel Drain System L I SN BN I IR BN B I BN N Y Y IR N S RN R BN B N R BN OEE BN N R B B N N R R Y 22 Gamma Heating in the Core Vessel of the 30-Mw E}(perimental Reactor 4 % ® 58 %9 0 F SR TE SRS S S SR 22 COMPONENT DEVELOPMENT AND TESTING «vecveoveevocscasnenes ol Salt-ILubricated Bearings for Fuel PUMDPS ..coevecrevceeese 2l Hydrodynamic Journal Bearings seeeceecsessasavoness oL Hydrodynamic Thrust Bearings ..eceeesvesscscesnsesss 25 Test of Pump Equipped with One Salt-Iubricated Journal Bearing ® 8 8 8 5 0 PS80 s RS ETEN PSR ea R 25 Bearing Mountings L I B N I BN BN BN DN BN BN BN BN BN OBNE BN IR TN TR BN BN BN NN I N B B B BN BN BN B 25 Mechsnical Seals for Pumps «vvevnees tacestauscacnscserns 27 leEndl}I‘ance Testing LI TN I B N R SR I NN R B I R R N RN I N Y RN R B B N B B RN R N R ) 27 Frozen’-LeadP‘Jmp Seal . 8 B0 B 8 O ¥ & 8N C TSNP R TS S ESeEE ERY e 27 Techniques for Remote Maintenance of the Reactor Sys-tem e & 2 0 0 & X 0 8D 0 O RS I RS SR AN ER RSN s R eSS ST Ne e s 28 Design, Construction, and Operation of Materials Testing I.OOPS LR I N I R N I BN B N B NN NN B N B RN O BN BN R BRI BE B RN R RN I N N BN B 32 Forced-Circulation LOODS tevesesrcosrcscscatanocona 32 In-Pile I-oops & 0 & & 8 O 8 B s 9 B S 3G9 P LB SR RO SRS YRR e 35 iii 1‘30 ENGINEERING RESEARCH * 0 5 & % %5 5 0 e 0 ® © @ © 6 % 6 8 8 3 O PG I OO C OSSO0 s o Physical Property Measurements ...ccevecececccecoscennns Enthalpy and Heat Capacity ececececsssccvconccscsne ViSCOSItY seceocnsecsssesossscscnssassssconcsssoncs Surface Tension +...ee-0a A T Heat-Transfer StudiesS ceececcscsessescscsocesoncssonnee . Hydrodynamic StudieS ceeecsessscsosssssccoccsscsscsscacns 1.k, INSTRUMENTS AND CONTROLS eececenssccasscecsssccannsaceos Molten-Salt-Fuel Level Indicators sceeecccensoscsccscsoas INOR-3 High-Temperature Pressure Transmitters c¢.c.coeo.. PART 2. MATERIALS STUDIES D.1. METATLLURGY «vvevececnnencans ettt et e, Dynamic Corrosion Studies ¢ecceescerscsccrscccscnccccans INOR“'8 Themal—convection LOOPS @ 9 e 35 8 0008 00400 e Inconel Thermal—convectiOn LOOPS ® ® 404000 0e0 0o s General Corrosion StUdiesS ceeeeeccvcrsesencecssacssosnans Penetration of Graphite by Molten Fluoride SALTS covsescsnsorssssassscsssesssosncssaccns sessance Uranium Precipitation from Molten Fluoride Salts in Contact with Graphite ......... cevscssesescana Thermal-Convection-ILoop Tests of Brazing Alloys in Fuel 130 teevecsennenee ceeescescesrecnsenenee Thermal-Convection-Loop Tests of the Compati- bility of INOR-8, Graphite, and Fuel 130 ....... . Mechanical Properties of INOR-8 secveevnronnsocsansannns Creep TesSTS teeescecetecesonsonsoscsscossnsccscsnans Fatigue Studies ceeeeeecnacaeocenne cececanns toesnoas Shrinkage Characteristics of INOR»S ............... Materials Fabrication Studies ....cc.ccc.. cecsscssasnens Effect on INOR-8 of Aging at High Temperatures .... Triplex Heat Exchanger Tubing .cesceces.. ceccacnnna Welding and Brazing StudiesS eceeeeececcsosnsosscscncecrsos Procedures for Welding INOR-8 ...vivescevcccocncas . Mechanical Properties of INOR-8 WeldS .euevevecnesen Fabrication of Apparatus for Testing the Compati- bility of Molten Salts and Graphite e.oeeveeceees 2.2, CHEMISTRY AND RADIATION DAMAGE +'vevenerenennnnsncnensns Phase Equilibrium StudiesS .eeoececsvonsccosoccscsosannsss The System LiF-BeFQ-ThF4 Secs s et usessssnsessnssnnns The System NaF-BeFQ-ThFu Ctessecesasssnessensetsence The SystemNaF-ThFh—UFu ecceccsacsscacace crcenvesaaca iv IIl:h.e System SDFE-NHLLIFQ € 5 8 6 6 08 8 00606008800 0C000CePIROSEIOEILE Solubility of PuF3 in Converter Fuels ......... covas Separation of LifF from LlTF—Bng ............ cosses Fission-Product Behavior ...... cessceen s sesasacaes cesesee PI‘ECipitation Of SmF3 With CEF3 8 @ 9 06 006 00 000 s s 00 ¢ e s 0 Chemical Reactions of Oxides with Fluorides in Molten-Fluoride-Salt Solvents ..cceeceveccecans oo Gas Solubilities in Molten Fluoride Salts ...cvvececccess Solubility of Neon in LiF--BeF2 ....... crsesscssenncs Solubility of COp in NaF-BeZF2 ceciscansecsssssccsse e Chemistry of the Corrosion Process ......ceoee eeasee ceees Samples from Operating Loops ..ceeceeciiereeenaanss .o Radioactive Tracer Analyses for Iron in Molten Fluoride Salts covvveeecceecnnas cesecscensscncsans . Activities in Metal AllOysS .ececcecens cscacsescnass .o - Vapor Pressures of Molten Salts ........ Ceesseassesessens Permeability of Graphite by Molten Fluoride Salts ....... Radiation Damage Studies «.ccievceen crsecsacsnrancens sreane INOR-8 Thermal Convection Loop for Operatlon in the LI'PR- .......................... & 8 9 & & 5O s 00 0t In-Pile Static Corrosion Tests .eeeeeieccens D Preparation of Purified Materials ............ cesceas ceee Purification, Transfer, and Service Operations teces Fuel Replenishment Tests +iceeen... cesccae Pure Compounds Prepared with Molten Ammonlum Blfluorlde ® @ ® 6 8 4 6 % 6 6 S B YV S e e O e s a0 o ¢ & & " & 0 9 6 8 0 Reaction of Chromous Fluoride with Stannous Fluoride .............. ¢ 6 & ® & 2 O & s s 0 e 0 0 & & 8 & % & O & & 0 FUEL PROCESSING ...... ceeieas Ceeeneaas 2.3, | MOLTEN-SALT REACTOR PROGRAM QUARTERLY PROGRESS REPORT SUMMARY Part 1. ZReactor Design Studies 1.1. Nuclear Calculations and Design Studies The effects of fuel processing by the ion-exchange method at a rate of once per year and at a rate of 12 times per year were compared for the Interim Design Reactor fueled with U233. Processing at the higher rate gave insufficient improvement in performence to offer a | substantial economic advantage. The nuclear performasnce of a one-region, graphite-moderated, unreflected, thorium—conversibn, molten-salt-fueled reactor was studied to evaluate the effects of processing the fuel at the rate of once per year and 12 times per year, the effects of decreasing the thickness of the core vessel and adding a blanket, and the effects of using a graphite core vessel and a blanket. The results of the calculations indicated that the nuclear performence of the one-region reactor fueled with Ug35 and processed by either the volatility or ion-exchange methods at the rate of one fuel volume per year is com- petitive with the best performance of the solid fuel systems, and therefore the molten-salt reactor will yield lower over-all fuel costs when the potential economies of continuous, chemical processing of the fluid fuel are realized. The initial nuclear characteristics of a one-region, heterogeneous, - graphite-moderated and -reflected, thorium-conversion, molten-salt- fueled reactor were studied. It was found that with 13 mole % ThF) in the fuel, a maximum regeneration ratio of 0.846 could be obtained with an inventory of approximately 1150 kg of U5°. With 7 mole % ThFu in the fuel, an estimated maximum regeneration ratio of 0.798 at U235 was obtained. an inventory of approximately 900 kg of Calculations were also made of the performance of two-region, graphite-moderated, molten-salt-fueled, breeder reactors. OSpherical vii reactors, 5 ft in diameter, with fuel volume fractions of 0.10 and 0.15% were studied. Results were obtained for reactors with power levels of 125 and 250 Mw. It was found that the only reactor with a doubling time less than 20 years was the 250-Mw reactor having a fuel volume fraction of 0.10. 5 on the reactivity 9 of molten-salt-fueled reactors were studied. It was found that Be The effects of fast-neutron reactions in Be is an appreciable poison in homogeneous, molten-salt-fueled reactors, but it has negligible effect in graphite-moderated, molten-salt- fueled reactors. Work is continuing on the design of a 30-Mw, one-region, experi- mental, molten-salt reactor. 1.2. Component Development and Testing Tests of salt~-lubricated hydrodynamic journal bearings were continued in order to obtain operating data on which to base an optimum design. Assembly of a thrust-bearing tester was nearly com- pleted. An existing centrifugal sump pump is being modified for service tests of a salt-lubricated journal bearing, and means for flexibly mounting the bearing to the pump casing are being investigated. The modified Fulton-Sylphon bellows-mounted seal being subjected to an endurance test in a PK-P type of centrifugal pump has continued to seal satisfactorily forvmore than 12,000 hr of operation. Operation of an MF type of centrifugal pump with fuel 30 as the circulated fluid has continued to be satisfactory through more than 15,500 hr, with the past 13,500 hr of operation being under cavitation damage conditions. A small frozen-lead pump seal on a 3/16-in.-dia shaft has operated since the first 100 hr of 7500 hr of operation with no lead leakage. A similar seal on a 3 1/b-in.-dia shaft has operated 3600 hr, with an average leakage rate of 9O cms/hr. Operational data suggest that the seal should be redesigned to provide better coolant control and packing to decrease the annulus between the seal and the shaft. Construction work continued on the remote maintenance demon- stration facility. The work is on schedule and is to be completed by viii June 30, 1959. The operation of long-term forced-circulation corrosion-testing loops continued in 15 test stands. Tests of two Inconel loops that had operated for one year were terminated. GSalt samples were removed periodically from two loops that contain sampling devices. The in-pile loop operated previously in the MIR was disassembled. It was found that oil which had leaked past the pump shaft seal and filled the oil trap on the purge outlet line was polymerized by radia- tion, and the polymerized oil plugged the outlet line. The second in-pile loop, which was modified to minimize the probability of purged line plugging and activity release, was installed in the MIR on April 27. 1.3. Engineering Research The enthalpy, heat capacity, viscosity, and surface tension have been experimentally obtained for several additional fluoride salt mixtures containing BeF, with varying amounts of UF) and/or ThF), . Altering the salt compoiition toward higher percentages of the high- molecular-weight components resulted in substantial decreases in both the enthalpy and heat capacity. The viscosity showed some increase (of the order of 10%) as the percentage of UF), and ThF) in the mixtures was increased. Initial flow calibrations of a full-scale mockup of the pump system designed for the study of interfacial film formation and heat transfer with BeF,.,~-containing salts have been completed; 2 assembly of the components is proceeding. The flow characteristics of the sintered-metal-filled annulus of a double-walled tube have been obtained for two porosity conditions. 1.k. Instruments and Controls Work has continued on molten-salt-fuel level indicators. Two Inconel "I'"-tube-type elements were prepared for testing, and an INOR-8 element and test vessel are being constructed. Six INOR-8 pressure transmitters and indicating systems were ordered for testing. These units are of the pneumatic~-indicator type. ix Part 2. Materials Studies 2.1. Metallurgy Corrosion studies were completed on four INOR-3 and four Inconel thermal-convection loops. Metallographic examination of the INOR-3 loops, three of which operated for 1000 hr and the other for one year, revealed no observable attack. The Inconel loops, which represented three one-year tests and one 1000-hr test, showed intergranular void attack to depths ranging from L to 15 mils. Twelve thermal-convection- loop tests were initiated during this period. The scheduled tests of two forced-circulation loops were completed, and one new forced- circulation-loop test was started. The effects of penetration of graphite by molten~fluoride-salt fuels are being investigated in order to evaluate the problems associated with the use of unclad graphite as a moderator. In static pressure penetration tests at 150 psia and lSOOOF, partially degassed graphite was penetrated throughout by fuel 30 (NaF-ZrFu—UFh) but was not penetrated by fuel 130 (LiF-Bng-UFu), as indicated by macroscopic examination. Thermal cycling of the graphite that was penetrated by fuel 30 did not result in damage to the graphite. Additional static and dynamic tests are planned. A series of tests was run to further investigate the precipitation of uranium from fuel 130 held in a graphite crucible at lBOOOF. The results of these tests supported the previous conclusion that the uranium precipitation was the result of the fuel reacting with oxygen supplied by degassing of the graphite. Several brazing alloys were tested for compatibility with fuel 130. Coast Metals alloy No. 52 (89% Ni-4% B-5% Si~2% Fe) and pure copper brazed to Inconel and INOR-8 showed good resistance to fuel 130 in 1000-hr tests at 13OOOF in a thermal-convection loop. The following brazing alloys showed some attack or porosity as a result of similar exposure to fuel 130:‘ General Electric alloy No. 81 (70% Ni—19% Cr—11% Si), Coast Metals alloy No. 53 (81% Ni-8% Cri% Si— - L% B-3% Fe), and a gold-nickel alloy (82% Au—18% Ni). A specimen of INOR-8 was examined for evidence of carburization that had been exposed to fuel 130 for 4000 hr in a thermal-convection loop hot leg at a temperature of l3OOOF. The presence of a 10-in. graphite insert in the hot leg of this loop did not cause carburization of the INOR-8. No new creep data for INOR-8 became available during the quarter. Creep tests are presently in progress at 1100 and 1200°F with the fluoride salts of interest as the test environments, and these tests are expected to run in excess of 10,000 hr. The critical results obtalined from rotating-beam fatigue tests at 1500°F indicate that INOR-8 has significantly better fatigue resistance than Inconel. Creep and reiaxation tests indicate that an unstable condition exists in the temperature range from 1100 to 1400°F. There is some indication that a second phase appears which causes contraction of the metal even under load. Based on the results of tensile tests conducted on specimens of INOR-8 aged for 10,000 hr in the temperature range of 1000 to 1400°F, it has been concluded that INOR-8 does not exhibit embrittling tendencies that can be attributed to high-temperature instability. No significant differences were found between the tensile properties of the aged specimens and those of specimens in the annealed condition. An effort is now being made to fabricate triplex heat exchanger tubing containing a prefabricated porous nickel core. Porous nickel tubes have been ordered from Micro Metallic Corp. to determine the feasi- bility of cladding the material with Inconel and INOR-8. A sample piece of porous nickel has been incorporated into the annular space formed between two Inconel tubes for conducting a preliminary cladding experiment. Procedures are being developed for fabricating INOR-8 material ranging in size from thin-walled tubing to heavy plate. The procedures thus developed are being qualified in accordance with methods prescribed by the ASME Boiler Code. The effects of various deoxidation and purification processes on the mechanical properties of INOR-8 weld metal are being studied xi in an effort to improve the high-temperature ductility of INOR-8 weld metal. Several heats of weld metal containing various additives were cast and fabricated into weld wire for mechanical property evaluation. A method was developed for brazing graphite to Inconel. A commercially available brazing alloy composed of silver, titanium, and copper was found that wet vacuum-degassed graphite. Such graphite-to- Inconel joints were used in the fabrication of equipment for studying the penetration of graphite by molten salts in a dynamic, high-pressure system. 2.2. Chemistry and Radiation Damage A revised phase diagram for the system LiF-BeFE-ThFu was prepared that includes new data obtained from thermal-gradient quenching experi- ments. Quenched samples from experiments in which an equilibration period of 3 weeks was used revealed that the area of single-phase ternary solid solutions involving 3LiF~ThFu is greater than previously reported. A phase diagram showing the progress made thus far in the study of the NaF-Bng-ThFu was also prepared, and the identity and approximate locations of primary phases in the system N&F~ThFu-UFA were determined. | The SnFe--NHuHF2 system was investigated because of its potenti- alities as a strongly oxidizing, low-melting solvent for reprocessing fuels. Reliable data were obtained only in the range O to 40% San. A 15 mole % addition of SnF, gives a mixture with a melting point of - about 100°C. Measurements were made of the solubility of PaF? in LiFnBng- 2 UFh (70-10-20 mole %). The solubility values obtained were all higher than those obtained with LiF-BeF, mixtures having about the same LiF concentration but no UF&' ° The possibility of separating LiF from the mixture LiF-BeF,, (63~ 37 mole %) by adding NaF and decreasing the temperature was investigated. In an initial experiment only 23% of the LiF remained in solution when the temperature of the mixture to which NaF was added at TOOOC was lowered to L490°C. xii Tests were initiated for determining the rate of exchange in a proposed method for decreasing the total rare earth content of molten fluoride salts. The exchange reaction SmF3 (a) + CeF3 (s)“‘fiCer (d) + Sn@3 (s) is utilized by passing the salt through an isothermal bed of solid CeF3 to lower the SmF3 content, and then the temperature of the effluent salt is lowered to decrease the total rate earth content. Two methods for separating uranium from fission products are - being studied. In one method the reaction of UFh with BeO to produce UO2 is utilized. In the other method, the reaction of UFA with water vapor produces UO The sharpness of the separation with water vapor may be advantageois in processing schemes. It is thought that a process can be developed that will eliminate the need for fluorine and that will provide for the simultaneous removal of uranium and thorium. A further step would be required to remove the rare earths. Measurements were made of the solubility of neon in LiF-BeF and CO, in NaF-BeF,.. 2 2 In the study of the chemistry of the corrosion process, further 2 samples of melts from operating INOR-8 and Inconel forced-circulation loops were analyzed for chromium. The chromium concentration in the LiF-BeFE-THthUFu mixture in the INOR-3 loop reached a plateau of about 550 ppm after about 1200 hr of operation. The chromium concen- ration of the same salt mixture in the Inconel loop increased more rapidly than in the INOR-8 loop and after 1700 hr was still increasing. These tests are continuing. Most investigations of corrosion behavior depend on accurate analyses for structural metel ions, and therefore anomalies in the present analytical methods are being studied. Tests of the permeability of graphite by molten fluoride salts have continued. Three types of graphite that were specially treated to meke them impervious were obtained from the National Carbon Company. The types designated ATL-82 and ATJ-82 were resistant to forced xiii impregnation with LiF-MgFg salt and were considerably resistant to penetration by a typical reactor fuel. Unexpectedly high concentrations of uranium in the center of the rods are being investigated. The in-pile thermal-convection loop for testing fused-salt fuel in INOR-8 tubing in the LITR was operated in preliminary fests out-of- pile, and satisfactory circulation was obtained. Final assembly of the loop system is under way. Two fuel-filled INOR-8 capsules were installed in the MTR and are being irradiated at lQBOOF. A device for testing a proposed fuel sampling and enriching mechenism was constructed and tested. It was found that the rate of solution of solid UFM in LiF-BeF, would be adequate for convenient enrichment procedures. ° The use of molten ammonium bifluoride as a reactant for preparing both simple and complex fluorides and the preparation of pure chromous fluoride by the reaction of chromium metal with molten stannous fluoride were studied, 2.3. Fuel Processing Studies of the processing of molten fluoride salt fuels by the fluoride~-volatility process were continued. Further measurements of the solubility of neptunium (1IV) in aqueous solutions saturated with LiF-BeF ““HFM'UFM indicate that a solubility of the order of 0.0002 2 to 0.00005 mole % may be expected in actual processing. Xiv PART 1. REACTOR DESIGN STUDIES 1.1. NUCLEAR CALCULATIONS AND DESIGN STUDIES Nuclear Calculations Effect of Ion-Exchange Processing of Rare-Earth Fission Products on 233 Performance of Interim Design Reactor Fueled with U It was reported previouslyl that substantial savings in fuel burnup and inventory couid be achieved in the Interim Design Reactor2 fueled with U235 by passing the fuel salt rapidly through beds of CeF3 to remove the rare earth fission products. The effect of ion exchange processing on the performance of the same reactor system fueled with U233 has now been studied. The results for two different processing rates — once per year and 12 times per year — are compared in Fig. 1.1.1. Processing at the rate of once per year (1.7 ft3/day) could bé performed either by the fluoride-~volatility methodl or by the ion-exchange method. Because of the high cost associated with the discarding of carrier salt in the volatility method, it is not feasible to use the volatility process for the higher rate. It may be seen that processing at the higher rate gives only a small improvement in performance. It is doubtful therefore that the ion-exchange process offers any substantial economic advantage in the 233 Interim Design Reactor system fueled with U Nuclear Performance of a One-Region, Graphite-Moderated, Unreflected, Thorium-Conversion, Molten-Salt-Fueled Reactor The nuclear performance without processing to remove fission products of a reactor having a spherical core 1 ft in diameter and fuel channels 3.6 in. ID arranged in a square lattice on 8 in. centers, was described previously.3 The performance of the same system with various processing rates and with other modifications has now been MSR Quer. Prog. Rep. Jan. 31, 1950, ORNL-268k, p 25. ®Molten Salt Reactor Program Status Report, ORNL-2634% (Nov. 12, 1958). 3 2681 MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL- , p 31. UNCLASSIFIED ORNL-LR-DWG 38445 1.08 | o \ PROCESSING RATE, FUEL VOLUMES/year 5 1.06 a—— — 0’4 \ 12 = o T — 4 5 1.04 e uJ uZ_l ‘ o 1.02 ! o 1.00 2000 INVENTORY A “w 1500 S— w E A 0 o] 5 .§ 1000 CORE DIAMETER: 8 ft s REACTOR POWER: 600 Mw {th), 250 Mw (e) = PLANT FACTOR: 0.8 > VOLUME OF FUEL: 600 ft° O FUEL: U%®® 5 500 UEL bl > = . < &) = % CUMULATIVE NET BURNUP 0 . m— ——— 1 I 12 -500 0 4 8 12 16 20 TIME OF OPERATION (yr) Fig. 1.1.1. Effect of Fuel Processing Rate on Nuclear Performance of Spherical, Homogeneous, Two- Region, Molten-Salt-Fueled, Breeder Reactor. studied, and typical results are presented in Table 1.1.1 and in Fig. 1.1.2, where regeneration ratio; eritical inventory, and net cumulative burnup are plotted as functions of time of operation. Without processing, the regeneration ratio falls from 0.8 to 0.5 in 20 years. The fuel additions required to override fission-product poisons are large; the critical inventory rises from 800 to 1800 kg of fuel, giving an average specific power of only 0.4 Mw/kg. The burnup is also large, averaging 0.45 g/Mwd. Processing the fuel salt to remove fission products at the rate of one fuel volume per year (2.5 ftB/day) stabilizes the regeneration ratio at above 0.8. The inventory averages about 750 kg. The burn- up 1s reduced to 0.21 g/de. Increasing the processing rate to 12 fuel volumes per year (30 ft3/day) improves the performance only slightly. Although it is uncertain that a core vessel and blanket could be incorporated into this reactor, it was of interest to evaluate the nuclear benefits of such an arrangement. Accordingly, the thick- ness of the INOR-8 reactor vessel wall was reduced to 1/2 in. for the calculations, and a 30-in. blanket containing 13 mole % ThF) in a mixture of LiF and BeiF2 was added. The blanket system was processed 12 times a year by the fluoride-volatility method to remove U233, which was added to the fuel salt. The fuel was processed 12 times a year by the ion-exchange method to remove fission products. The results of the calculations are presented in Table 1.1.2, and the performance is indicated in Fig. 1.1.2 by the lines labeled U235 12,B. The regeneration ratio averaged in excess of 0.96 and approached 1.0 in the steady state. The critical inventory was not decreased significantly, but was shifted toward u?33 in composition (90%). The burnup was reduced to 0.056 g/MWd, and the reactor was practically self-sustaining. However, it was estimated that the savings in mills per kilowatt-hour effected by reducing the burnup costs would be more than offset by the additional capital costs of reactor vessel, blanket materials, pumps, processing equipment, and concomitant operating costs. Table 1.1.1. Effect of Fuel Processing Rate on Nuclear Performance of One-Region, Graphite-Moderated, Unreflected, Thorium-Conversion, Molten-Salt-Fueled Reactor Core diameter: 14 ft Reactor power: 600 Mw(th), 250 Mw(e) Plant factor: 0.8 Fuel volume: 900 3 Fuel channels: 3.6 in. ID Lattice: triangular, 8-in. centers Core vessel: l-in.-thick INOR-8 Blanket: none o s After 20 Years with Fuel After 20 Years with Fuel Processed Once per Year Processed 12 Times per Year Inventory Absorption (kq) Ratio® Inventory Absor?f:]on Inventory Absor.ptti]on (kg) Ratio (kg) Ratio Fissionable isotopes y233 563 0.764 559 0.794 4235 829 1.000 187 0.219 167 0.205 pu2¥ 3 0.017 0.3 0.001 Fertile isotopes Th232 38,438 0.783 38,438 0.776 38,438 0.807 u234 218 0.051 222 0.054 4238 64 0.008 142 0.017 136 0.017 Fuel carrier Li’ « 6,328 0.054 6,328 0.052 6,328 0.055 19 37,571 0.025 37,571 0.025 37,571 10.026 Moderator Be’ 1,840 0.001 1,840 0.001 1,840 0.001 cl? 0.051 0.049 0.052 Fission products 181 0.034 15.3 0.003 Parasitic isotopes U3 and others 197 0.037 184 0.036 Miscellaneous po233 17 0.009 17.8 0.010 Core vessel and 0.149 0.151 0.155 leakage Neutron yield, 1 2.071 2.207 2.217 Total fuel inventory, kg 829 754 726 Cumulative net burnup, kg 0 757 686 Net fuel requirement, kg 829 1,51 1,412 Regeneration ratio 0.791 0.836 | 0.8683 “Neutrons absorbed pet neutron absorbed by fissionable isotopes. UNCLASSIFIED ORNL—LR—DWG 38146 (12) \ W / o ® 7 == T~ REGENERATION RATIO \ CORE DIAMETER: 14 ft ~— REACTOR POWER 600 Mw (th), 250 Mw (e) —— PLANT FACTOR: 0.8 0.4 | FUEL VOLUME: 900 ft3 FUEL CHANNELS: 3.6in. ID 2000 LATTICE: TRIANGULAR, 8-in. CENTERS 1 I | / FUEL AND PROCESSING RATE (fuel volumes /year) U235 (O)V 1500 — . 1000 » - /" A \ S ——— e ———— P —— 500 CUMULATIVE NET BURNUP / ,f/ O P ——c—— N \ / Efi CRITICAL INVENTORY (kg of fissionable isotopes ) c\ [\¥} (&} Pt I \‘\[E ' C N (&) w o w —500 1 8 10 12 14 16 18 20 TIME OF OPERATION (yr) O N N o Fig. 1.1.2. Nuclear Performance of Spherical, Heterogeneous, Graphite-Moderated, One-Region, Unre- flected, Thorium-Conversion, Molten-Salt-Fueled Reactors. Table 1.1.2. Effect of Core Vessel Material and Blanket on Nuclear Performance of One-Region, Graphite-Moderated, ThoriumeConversion, MoltenSalt-Fueled Reactor - Core diameter: 14 ft Reactor power: 600 Mw(th), 250 Mw{e) Plant factor: 0.8 - Fuel volume: 900 t3 Fuel channels: 3.6 in. ID Lattice: triangular, 8-in. centers Blanket thickness: 30 in. Fuel processing rate: 12 times per year Reactor with 1/2-in.-Thick INOR-8 Reactor with Graphite Core Vessel Core Vessel Initial State After 20 Years Initial State After 20 Years Inventory Absorption Inventory Absorption Inventory Absorption Inventory Absorption (kg) Ratio® (kg) Ratio® (kg) Ratio® (kg) Ratio® Fissionable isotopes U233 (fuel) 625 0.917 594 1.000 620 0.936 U233 (blanket) 1.7 3 u23s 829 1.000 65 0.082 48 0.064 Py23 0.2 0.001 Fertile isotopes Th232 (fyel) 38,438 0.783 38,438 0.831 38,438 0.933 38,438 0.852 Th232 (blanket) 76,765 0.095 76,765 0.101 76,765 0.156 76,765 0.149 u234 239 0.060 277 0.072 u2s3e 64.3 0.008 104 0.013 1 Fuel carrier Li’ 6,328 0.054 6,328 0.057 6,328 0.068 6,328 0.060 F19 37,571 0.025 37,571 0.026 37,571 0.028 37,571 0.027 Moderator Be’ 1,840 0.001 1,840 0.001 1,840 0.001 1,840 0.001 c'? 0.051 0.054 0.063 0.056 . Fission products 15 0.003 15.3 0.003 Parasitic isotopes U236 and others 122 0.026 20 0.003 . Miscellaneous Pa233 (fuel) 18.1 0.010 18.6 0.011 Pa233 (blanket) 2.2 3 Blanket carrier salt, core 0.055 0.058 0.006 0.061 vessel and leakage Neutron yield, 7 2.071 2.240 2.256 2.241 Total fuel inventory, kg 829 691 594 670 Cumulative net burnup, kg 0 185 0 245 Net fuel requirement, kg 829 876 594 425 Regeneration ratio 0.885 0.995 1.090 1.064 °Neutrons absorbed per neutron absorbed by fissionable isotopes. In a further extension of the calculations, the nuclear benefits to be obtained by the use of a graphite core vessel (in place of INOR-8) 233 U and fueling with were studied. These changes are, of course, impractical in the present state of technology. However, the study was performed in order to define the limiting nuclear performance of this particular core and fuel system. The results of the calculations are presented in Table 1.1.2, and the performance is indicated in Fig. 1.1.2 by the lines labeled U233 12,B. The regeneration ratioc is stabilized above 1.06, the average inventory is about 650 kg, and the cumulative net burnup is slightly negative, amounting to -250 kg of U233 in 20 years. However, 76 kg of this is required as increased inventory in the fuel salt to override fission products and nonfissionable isotopes. It is concluded that the nuclear performance of the one-region reactor fueled with U235 and processed by either the volatility or ion~-exchange methods at the rate of one fuel volume per year is competitive with the best performance of the solid fuel systems, and therefore the molten-salt reactor will yield lower over-all fuel costs when the potential economies of continuous, chemical processing of the fluid fuel are realized. Initial Nuclear Performance of One-Region, Graphite-Moderated, Graphite- Reflected, Thorium-Conversion, Molten-Salt-Fueled Reactors The initial nuclear characteristics of a heterogeneous, graphite- moderated and -reflected, one-region, molten-salt-fueled reactor were studied. The reactor considered in the study consists of a cylindrical core, 15 ft in dia and 15 ft high, surrounded by a 2.5-ft-thick graphite reflector contained in an INOR-8 pressure shell 1.5 in. thick. The core is penetrated by cylindrical fuel passages arranged in an 8-in. triangular lattice parallel to the core axis. The resulting unit cells are hexagonal and 15 ft long. This system has been investigated over a range of fuel volumetric fractions in the core and at two concen- trations of thorium in the fuel salt. The initial nuclear character- 3 istics of the system, with an externsal fuel volume of 673 ft-, are given in Tables 1.1.3 and 1.1.4 for the cases having 13 and 7 mole % ThFu in the fuel, respectively; Fig. 1.1.3 shows the regeneration ratiq g as a function of the system inventory. The modified Oracle program Cornpone was used to calculate the multiplication constant and group disadvantage factors for the fuel and graphite. The results were then used for complete reactor calcu- | lations in spherical geometry on a core having the same volume as the cylindrical core. Mean, homogeneized densities (atoms/cm3) were used for each element. It may be seen from Fig. 1.1.3 that, in the case having 13 mole % ThFu in the fuel, a meximum regeneration ratio of 0.846 is obtained at an inventory of approximately 1150 kg of U235. With 7 mole % ThFh in the fuel, an estimated maximum regeneration ratio of 0.798 at an inventory of approximately 900 kg of U235 was obtained. Nuclear Performance of Two-Region, Graphite-Moderated, Molten-Salt- Fueled, Breeder Reactor The nuclear performance of spherical, graphite-moderated reactors, 5 ft in diameter with fuel volume fractions of 0.10 and 0.15, was studied. The fuel channels were 2.6 in. in diameter and were arranged in an 8-in. trianguler lattice. These reactors were surrounded by 2-in.-thick graphite core vessels and 30-in.-thick blankets. The fuel consisted of 71 mole % LiF, 16 mole % BeF,, and 13 mole % ThFu plus UFM‘ The blanket salt had the same composition as the fuel, but no UFM' The performance of these reactors over a period of 20 years was calculated for power levels of 125 and 250 Mw. The processing rate in all cases was 12 fuel volumes per year. The system volumes used for these calculations are shown in Table 1.1.5; only the heat exchanger volume was increased for higher power levels. The results of the calculations are presented in Tables 1.1.6 and 1.1.7 and are plotted in Figs. 1.1.4, 1.1.5, and 1.1.6. It may be seen that the only reactor with a doubling time less than 20 years is the o 250-Mw reactor having a fuel volume fraction of 0.10. For both cores, 10 1T Table 1.1.3. Initial Nuclear Performance of a One-Region, Graphite-Moderated, Thorium-Conversion, Molten-Salt-Fueled Reactors with 13 mole % ThF4 in the Fuel Fuel Volumetric Fraction in the Core 0.12 0.16 0.20 0.22 0.24 0.28 Inventory Absorption |nventory Absorption Inventory Absorption Inventory Absorption Inventory Absorption Inventory Absorption (kg) Ratio? (kg) Ratio (kg) Ratio (kg) Ratio tkg) Ratio (kg) Ratio Fissionable isotope u 233 798.3 1 925.3 1 1,086 ! 1,168 1 1,299 1 1,526 1 Fertile isotopes Th232 42,338 0.804 46,883 0.820 51,408 0.837 53,671 0.836 55,954 0.831 60,480 0.823 u238 59.9 0.005 69.4 0.006 81.4 0.008 87.6 0.009 97.4 0.011 114.5 0.012 Fuel carrier Li’ 6,974 0.064 7,723 0.061 8,469 0.055 - 8,841 0.055 9,217 0.049 9,963 0.046 19 41,350 0.023 45,789 0.026 50,208 0.026 52,418 0.028 54.648 0.027 59,068 0.030 Moderator Be? 2,021 0.001 2,238 - 0.001 2,454 0.001 2,562 0.001 2,674 0.001 2,887 0.001 cl? 334,900 0.079 328,900 0.056 323,000 0.040 320,000 0.036 317,000 0.030 311,000 0.025 Core vessel capture 0.105 0.104 0.089 0.089 0.093 0.091 plus leakage Neutron yield, 7 2.081 2.074 2.056 2.054 ©2.042 2.028 Regeneration ratio 0.809 0.826 0.845 0.845 0.842 0.835 9Neutrons absorbed per neutron absorbed in fissionable isotope. A Table 1.1.4. Initial Nuclear Performance of a One-Region, Graphite-Moderated, Thorium-Conversion, Molten-Salt-Fueled Reactors with 7 mole % ThF4 in th e Fuel Fuel Volumetric Fraction in the Core 0.12 0.16 0.20 0.22 0.24 Inventory Absorption Inventory Absorption Inventory Absorption Inventory Absorption Inventory Absorption (kg) Ratio® (kg) Ratio (kg) Ratio (kg) Ratio (kg) Ratio Fissionable isotope u 235 472 ] 506 592 1 633 1 677 Fertile isotope Th232 22,926 0.710 25,387 27,837 0.769 29,060 0.771 30,300 u238 35.3 0.003 37.3 44.8 0.005 46.9 0.006 50.3 Fuel carrier Li7 6,641 0.105 7,355 8,065 0.097 8,420 0.099 8,777 F19 39,440 0.028 43,670 47,890 0.032 50,000 0.034 52,120 Moderator Be? 3,284 0.001 3,637 3,988 0.001 4,163 0.001 4,349 cl? 334,900 0.135. 328,900 323,000 0.072 320,000 0.066 317,000 Core vessel capture 0.123 0.100 0.101 plus leakage ‘ Neutron yield, 1 2.105 2.076 2.078 Regeneration ratio 0.713 0.760 0.774 0.777 - 0.787 9Neutrons absorbed per neutron absorbed in the fuel. UNCLASSIFIED ORNL—LR—DWG 38147 0.86 0.84 // \\\ 13 mole % Thf, IN FUEL \o\ 0.82 0.80 = /—/’-— — o 2 i o = / s} £ 078 = /o wJ =2 1] e 7 mole 7 ThF IN FUEL o 0.76 9 CYLINDRICAL CORE: 15 ft DIA, 45 ft HIGH REFLECTOR: 2.5 - ft~THICK GRAPHITE PRESSURE SHELL: 1.5~in.~THICK INOR-8 0.74 EXTERNAL FUEL VOLUME: 673 ft° 072 0.70 0 200 400 600 800 1000 1200 1400 1600 U235 INVENTORY (kg) Fig. 1.1.3. Initial Nuclear Performance of One-Region, Graphite-Moderated, Thorium-Conversion, Molten-Salt-Fueled Reactors. : 13 Table 1.1.5. Characteristics of Two-Region, Graphite-Moderated, Molten-Salt-Fueled, Breeder Reactors Fuel volume fraction in core 0.10 0.15 Power, Mw 125 250 125 - 250 Fuel volume, ft3 Core 6.67 6.67 10 10 Heat exchanger 15 30 15 30 Other 15 15 15 15 Total 36.7 51.7 Lo 55 Average inventory, kg of fissionable isotopes 73 110 76 135 Average regenerstion ratio 1.090 1.075 1.073 1.068 Annual production,? kg 2.15 4 b5 | 1.7 3.2 a . ' 233 Annual production = annual average excess of U 33 based on a reactor lifetime of 20 years. the regeneration ratio drops off at the higher power level. This was to be expected because of the buildup of U234 and U236 ; also, since the processing rate was the same for all power levels, the concentration of the fission fragments was increased at the higher power level. Effects of Fast Neutron Reactions in Be9 on Reactivity of Molten-Salt Reactors | | 4 The net effect of the reactions Be9(n,2n)2He , and Beg(n,a)He6(B)Li6 on the reactivity of molten-salt reactors was studied. In a spherical, two-region, homogeneous core 8 ft in diameter having a carrier salt composed of 58 mole % LiF, 35 mole % BeF,., and 7 mole % ThF), , and ye33 2 9 fuel, the contribution of the reactions in Be” to the neutron multi- plication was estimated to be -0.087. In a spherical, two-region, heterogeneous, graphite-moderated, molten-salt reactor having a core 5 £t in diameter, a carrier salt composed of 71 mole % LiF, 16 mole % BeF,, and 13 mole % ThF) , and fueled with U233, the reactivity effect 1k Table 1.1.6, Nuclear.Performance of Two-Region, Graphite-Moderated, Molten-Salt.Fueled, Breeder Reactors with Fuel Yolume Fraction of 0.10 Core diameter: 5 ft Lattice: triangular, 8-in. center Fuel processing rate: 12 times per year For 125-Mw(th) Plant For 250-Mw(th) Plant After 20 Years Initial State Initial State After 20 Years Inventory Absorption Inventory Absorption Inventory Absorption Inventory Absorption {kg) Ratio® (kg) Ratio?® (kg) Ratio® (kq) Ratio® Fissionable isotopes U233 (fuel) 59.1 1.000 67.9 0.910 83 1.000 99 0.898 U232 (blanket) 2.37 5 u23s 7.8 0.088 13 - 0.098 Py23 0.038 0.002 0.10 0.004 Fertile isotopes : ‘ Th?%2 (fuel) 1,567 0.402 1,567 0.335 2,207 0.402 2,207 0.322 Th2%2 (blanker) 21,527 0.720 21,527 0.660 21,520 0720 21,520 0.648 Y234 39.8 0.092 64 0.100 U238 1.34 0.002 4 0.005 Fuel carrier Li’ 258 0.027 258 0.020 363 0.027 363 0.019 F19 . 1,532 0.013 1,532 0.012 2,158 0.013 2,158 0.012 Moderator Be’ 75 0.001 75 0.001 106 0.001 106 0.001 c? 0.042 0.032 0.042 0.031 Fission products 3.18 0.006 6 0.008 Parasitic isotopes U236 and others 6.21 0.014 13 0.021 Miscellaneous Pa233 (fyel) 1.51 0.008 3 0.011 Pa233 (blanket) 3.05 6 Blanket carrier salt, core 0.060 0.055 0.060 0.054 - vessel, and leakage Neutron yield, 7 2.265 2.238 2.265 2.232 Total fuel inventory, kg 59.1 78.1 83.0 117 Cumulative net burnup, kg 0 ~72.2 0 -123 Net fuel requirement, kg 59.1 5.9 83.0 ~6 Regeneration ratio 1.122 1.122 1.064 1.080 “Neutrons absorbed per neutron absorbed by fissionable isotopes. 15 Table 1.1.7. Nuclear Performance of Two-Region, Graphite-Moderated, Molten-SaltFueled, Breeder Reactors with Fuel Yolume Fraction of 0.15 Core diameter: 5 ft Lattice: triangular, 8-in. center Fuel processing rate: 12 times per year For 125-Mw(th) Plant " For 250-Mw(th) Plant Initial State ' After 20 Years Initial State " After 20 Years Inventory Absorption Inventory Absorption Inventory Absorption Inventory Absorption (kg) Ratio® (kg) Ratio® (kg) Ratio® (kg) Ratio® Fissionable isotopes U233 (fuel) 59 1.000 72 0.907 103 1.000 127 0.901 U233 (blanket) 2 4 u23s 9 0.090 17 0.095 Py2% 0.1 0.003 ‘ 0.2 0.004 Fertile isotopes Th232 (fuel) 1,710 0.472 1,710 0.384 2,987 0.472 2,987 0.378 Th232 (blanket) 23,487 0.642 23,487 0.589 23,527 0.642 23,527 0.585 u234 44 0.095 81 0.099 u23e 2 0.003 4 0.004 Fuel carrier Li’ A 282 0.027 282 0.019 492 0.027 492 0.019 F19 1,671 0.019 1,671 0.017 - 2,920 0.019 2,920 0.017 Moderator Be’ 82 0.001 82 0.001 143 0.001 143 0.001 c'? 0.028 0.020 0.028 0.020 Fission products 3 0.006 6 0.007 Parasitic isotopes U236 and others 6 0.017 13 0.020 Miscellaneous Pa?33 (fuel) 2 0.010 3 0.011 Pa?33 (blanket) Blanket carrier salt, core 0.052 0.048 0.052 0.047 vessel, and leakage Neutron yield, 7 2.241 2.211 2.241 2.209 Total fuel inventory, kg 59 82 103 148 Cumulative net burnup, kg 0 ~59 0 -108 Net fuel requirement, kg 59 23 103 40 Regeneration ratio 1.114 1.061 1.114 1.055 9Neutrons absorbed per neutron absorbed by fissionable isotopes. 16 UNCLASSIFIED ORNL-LR-DWG 38448 | b CORE DIAMETER: 5 ft LATTICE: TRIANGULAR, 8—in. CENTERS 142 FUEL: U233 FERTILE MATERIAL: Th232 9 \ [ % FUEL PROCESSING RATE: 12 TIMES PER YEAR S //WORY FOR 250-Mw PLANT /-’—/ I INVENTORY FOR 125-Mw PLANT | 1 | // CUMULATIVE EXCESS U2 PRODUCTION IN 250-Mw PLANT | —"" | / — / - " CUMULATIVE EXCESS U%** PRODUCTION IN _________;_/ 125-Mw PLANT 2 4 6 8 10 12 14 16 18 20 Fig. 1.1.6. Inventory and Excess U233 Production vs Reactor Operating Time for Two-Region, Graphite- Moderated, Molten-Salt-Fueled, Breeder Reactors with a Fuel Volume Fraction of 0.15. 19 was only =0.00077. It was concluded that Be9 is an appreciable poison in homogeneous, molten-salt-fueled reactors, but has negligible effect in graphite-moderated molten-salt-fueled reactors. Oracle Code MSPR=-Cornpone 020 A revision has been maede of the element absorption edit of Oracle code MSPR-Cornpone 020, which now permits an immediate edit on paper tape. Previously it was necessary to wait for curve-plotter pictures. It is now possible to compute the breeding ratio and prepare the input" for Sorghum as soon as the case is run. The cross-section tape has been cheecked and corrected where necessary, and an edit has been obtained for the cross sections. The elements now included in the cross section library are listed below: Code Number 00 0l 02 03 Ol 05 20 Element 4233 4235 U238* Thorium* 239 Fluorine Oxygen Boron Sodium Beryllium Bismuth Cerium Two Fission Products Deuterium Iron Element Code Number i 1 15 16 17 18 19 1A 1B Nickel Chromium Molybdenum Aluminum Lithium (0.01 % Li Carbon Alloy-8 U23h 1l Thorium (2) Thorium (0) Thorium (1) Thorium (4) Thorium (10) Thorium (25) Thorium (0.25) 6), 26 Hydrogen 2B Thorium (26G) 27 Thorium (0.75) 2C Thorium (29G) 28 Thorium (0.50) 30 Thorium (13) 29 Thorium (7) 31 PthO - 2A Thorium (13G) | o38% There are 32 lethargy groups, with the last at 1180°F. Element UC38 has zero absorption cross sections in all groups. It must be combined with one of the other U238 options listed, which have zeros for all cross sections except absorption, to compose a complete set of cross 238. The options permit the selection of a set of sections for U absorption cross sections properly asdjusted with respect to resonance saturation and Doppler broadening. The numbers in parentheses indicate the concenfiration in mole % of UEBSFA in mixtures of LiI" and BeF2 such as those used in the molten-salt reactors. Thorium is treated simi- larly, and sbsorption cross section options for both mixtures with lithium and beryllium fluorides and graphite (denoted by G in parenthe- ses) are available. Design of 30-Mw Experimental Reactor Work is continuing on the 30-Mw, one-region, experimental, molten- salt reactor described in the previous report. The basic reactor concept has remained unchanged. More details of the equipment have been developed, and various auxiliary system layouts are being studied. A reactor site has been selected for design purposes, and perspective drawings of a test reactor building have been prepared. The bayonet-type fuel-to-coolant heat exchanger is being redesigned in an effort to simplify the fabrication problems. Design data for use in layout studies have been prepared for a U-tube, U-shell, coclant- salt-to-steam superheater. MR Quar. Prog. Rep. Jan. 31, 1959, ORNL-2684, p 3. 21 Fuel Drain System The fuel drain system consists of four 2-1/2-ft-dia vessels approximately 20 ft long. These tanks will be suspended in a vertical position and manifolded to the reactor with pressure-siphon drain lines. Two drain vessels will be required to contain the fuel system inventory, and each pair of tanks will have a separate system to isolate it from the reactor. The two extra drain tanks will provide capacity for a spare salt volume, which may be used for fuel system cleaning or decontamination operations. The vessels will be preheated and afterheat will be removed with a recirculating gas system. This gas system, as in the case of the reactor gas system, will include a blower, heater, and cooler packages enclosed in a loop. The drain system criticality problem is being investigated. A multiplication constant of less than 0.5 was obtained for one drain vessel presumed to contain the entire fuel system UFh inventory in an LiF-BeF,. salt mixture. As a further check, the U235 2 settle out in the vessel as an oxide compound, and multiplication was assumed to constants are being obtained for four different geometries of fuel concentrate, with one edge reflected by the salt mixture. A maximum multiplication constant has not yet been obtained, but, with the fuel | | concentrated in a cube, the multiplication constant was 0.83. It is believed that this value would increase slightly with further settling of the fuel. The bottom section of the drain vessel will therefore be made with a smeller diameter to reduce the potential hazard. Gamma Heating in the Core Vessel of the 30-Mw Experimental Reactor The heat generation in the core vessel of the experimental reactor has been calculated using the Oracle program Ghimsr.s’6 The total heat 5L. G. Alexander and J. W. Miller, Heat Generation in the One- Region, Experimental Molten-Salt Reactor, ORNL~2746 (to be published). D. B. Grimes, Ghimsr, Gamma Heating in Molten-Salt Reactors, ORNL CF 59-5-40 (to be published). 3 generation was 2.30 w/cm” at the inner surface of the core vessel. The individual contributions to the total heating are given below: Source | | - Heating (W/cms) Prompt gemma rays - 0.61 Delayed gamma rays 0.52 Inelastic scattering gamme rays 0.34 Beryllium capture gamma rays 0.02 Fluorine capture gamma rays 0.07 Uranium capture gamma rays | | 0.15 Core vessel capture gemma rays 0.59 Total 2.30 1.2. COMPONENT DEVELOPMENT AND TESTING Salt-Iubricated Bearings for Fuel Pumps Hydrodynamic Journal Bearings The sixth test of a journal bearing operating in molten-salt fuel 130 (LiF-BeF, terminated on schedule. During this test the temperature of the salt -UF), , 62-37-1 mole %), as described previously,l was was increased in EOOF increments from 1200 to lBOOOF. Performance was satisfactory throughout the entire test, which was conducted at a constant journal speed of 1200 rpm and a constant radial load of 200 1b. The total operating time for this test was 192 hr. A seventh test was then performed with the bearing and journal used in the fifth and sixth tests. This test also consisted of steady-state operation at a journal speed of lQOO.rpm and a radial load of 200 1b, but start-stop operations were carried out three times at each test temperature. The temperature was varied as in the sixth test from 1200 to 1400°F in 50°F increments. On attempting a second restart at the 1400°F temperature level, an increase in the power required was noted, and the test was halted. .The total operating time for this test was 216 hr. Inspection revealed a galled spot on the loaded side of the bearing. The eighth test was conducted to study the effect of increasing the bearing radial clearance‘(measured at room temperature) to 0.007 in. All bearings previously tested had a radial clearance of 0.005 in. Molten-salt fuel 130 at 1200°F was used for this test, and the bearing radial load was varied from 50 to 500 1b at each of two journal speeds, 600 and 1200 rpm. After 356 hr of operation, including 39 start-stop sequences, a sudden increase in power halted the test. At the time the power increase occurred the journal was operating at 600 rpm, and the load was being increased from 300 to 400 1b; the load had reached 382 1b. From the standpoint of minimum film-thickness, these conditions were the lMSR Quar, Prog. Rep. Jan. 31, 1959, ORNL-2684, p 39. 2L most stringent yet imposed in these molten-salt~lubricated bearing tests. A carburized INOR-8 journsl was used with an uncarburized INOR-8 bearing for the ninth test. The tester was started and stopped 260 times during a period of operation of 272 hr at 1200°F. The jdurnal - speed was 1200 rpm and the bearing was loaded to 200 1b. Both the bearing and the journal were in good condition, with undamaged surfaces, at the end of the test. The bearing 1s shown in Fig. 1.2.1 and the Journal in Fig. 1.2.2. Hydrodynamic Thrust Bearings Assembly of the thrust-bearing tester described previouslyl is nearly completed. The thrust-load actuator is being given preliminary tests prior to being fitted to the rotary portion of the tester. Test of Pump Equipped with One Salt-Lubricated Journal Bearing Detailed design work is nearly complete on the modifications to be made to a PK type of centrifugal sump pump in order to replace the lower oil-lubricated bearing with a molten-salt-lubricated hydrodynamic journal bearing. The external motor of the PK pump is being replaced with an integral, totally enclosed motor. The configuration of the modified pump is intended to simulate, in large measure, that of the pump proposed for use in an experimental molten-salt reactor. Proposals have been received for the fabrication of the motor and the pump shaft, and purchase orders are being placed. An existing facility is being prepared for tests of this pump. Bearing Mountings Means are being investigated for flexibly mounting INOR-8 bearings to iNOR-B pump casings sSo that thermal distortiofi between the bearing and the pump shaft journal can be accommodated. A diaphragm type of mounting and a mounting that is somewhat similar to the Westinghouse "Thermoflex" mount are being studied. 25 UNCL ASSIFIED PHOTO 33924 9c UNCLASSIFIED PHOTO 33926 0] 1 2 3 ! | | J 0 i 2 3 INCHES | 1 | | INCHES Fig. 1.2.2. Carburized INOR-8 Journal After Operation for 272 hr at 1200°F Fig. 1.2.1. INOR-8 Bearing After Operation for 272 hr at 1200°F with Molten-Salt Lubrication. with Molten-Salt Lubrication. Mechanical Seals for Pumps The modified Fulton-Sylphon bellows-mounted sea12 being subjected to an endurance test in a PK-P type of centrifugal pump has accumulated an additional 1848 hr of operation since the previous report period, for a total of 12,478 hr. The pump has been operating at a temperature of 1200°F, a shaft speed of 2500 rpm, and a NeK flow rate of 1200 gpm. The maximum test~seal leakage rate was 3.5 cmS/day; however, on an averaged basis, the leakage is still negligible. Two pump stoppages occurred — one for reworking-the drive motor commutator and replacing the motor brushes, and one as a precautionary measure during a power outage. Pump Endurance Testing 3 An MF type of centrifugal pump has continued in operation,~ and has logged more than 15,500 hr (epproximately 1 3/4 yr) of continuous operation. No maintenance has been performed on the pump during this period. During the last 13,465 hr, a cavitation endurance test has been under way with the pump operating at the steady-state conditions of 2700 rpm, 645 gpm, and 2.5-psig pump tenk cover gas pressure at 1200°F. During the quarter the pump was stopped five times. Two stops were momentary and were caused by power outages. One stop was of 65 min duration for replacing brushes in the motor-generator set. One stop was of 30 min duration for calibrating the pressure-measuring devices. One stop was for 60 min for freeing the system throttle valve that had become stuck. Frozen-Lead Pump Seal The small frozen-lead pump seal being tested on a 3/16-in.-dia shaft, as described previously, has operated continuously since it was 2MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-2684, p k2. 3MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-2684, p L2, b, B. McDoneld, E. Storto, and J. L. Crowley, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p 23. 27 started on June 13, 1958. The accumulated operating time is more than 7500 hr. Excepfi fbr slight leakage of lead during the first 100 hr of operation, there has been no further leakage. | The large lead pump seal with a 3 1/4-in.-dia rotating shaft has been operating for 3600 hr. Leskage of solid lead from the seal is sporadic. The average leakage rate for the time operated is 9 cms/hr. Operational data suggest that the seal should be redesigned to provide better coolant control and to incorporate a short, resilient packing to decrease the annulus between the seal and the shaft. A detailed description of the equipment was presented previously.5 At the present time, the operating'conditions are the following Shaft speed 975-1000 rpm Argon pressure over molten lead in tank 0—-3 psizg Seal cooling water flow rate 0.8 gpm Temperature rise in cooling o water ~25°F Temperature where seal is formed .l8O~19OOF Techniques for Remote Maintenance of the Reactor System Progress has been made in the design and construction, Fig. 1.2.3, of a remote maintenance demonstration facility. The scheduled completion date is June 30, 1959, and all phases of the work are on schedule. The General Mills mechanical arm manipulator, 3 1/2- and 6-in. Inconel piping subassemblies, dummy heat exchangers, salt dump tank, dunmy reactor, two closed-circuit television systems (Fig. 1.2.4), and the spool for Jjoining the pump and motor have been received. One of the 20 sets of freeze-flange pipe joints (Fig. 1.2.5) is on hand, and the modifications required to adapt an existing PK type of pump for use in this facility have been completed. Eleven pipe preheater and insulation units that were designed to MSR Quar. Prog. Rep. Jan. 31, 1950, ORNL-268%, p 43. 28 - ok um:mw%m PHOTG 33887 ‘ Fig. 1.2.3. Remote Maintenance Demonstration Facility Under Construction. ot Fig. 1.2.4. Closed-Circuit Stereo Television Equipment for Remote Maintenance Demonstration Facility. UMCLASSIFIED FHOTO 33885 i R R S UNCLASSIFIED - e i : B PHOTO 33886 . ' i - I i : = ~ Fig. 1.2.5. A Freeze-Flange Joint for a é-in. Pipe in the Remote Maintenance Demonstration Facility. 31 contain electric heaters and insulation in one package are being fabri- cated. These units are to be removeble and replaceable in short sections with the use of the General Mills manipulator. Specially designed spring-type pipe supports are being procured. Three gas and compressed- air control cabinets are being fabricated. Installation of the General Mills manipulator, heat exchangers, dummy reactor, salt-circulating pump, salt piping, service piping, pump- lube-oil piping, and electric heater control cabinets 1s in progress. Design, Construction, and Operation of Materials Testing Loops Forced-Circulation Loops The operation of long-term forced-circulation corrosion-testing loops was continued. Fourteen test loops are presently in operation, two Inconel loops were terminated as scheduled during the quarter, and one new Inconel loop was started. A loop fabricated of INOR-8 is presently being installed. The two Inconel loops, designated 9344-2 and 9377-3 in Table 1.2.1, were terminated after one year of operation. The facilities from which these loops were removed have now been revised according to the latest design6 for long-term operation. Of the 15 available loop facilities, 11 are of the improved design, and the remaining four loops are in various stages of improvements. Several momentary power failures caused interruption of power to the loops during the quarter, but flow resumed in all the loops after power was restored, and there were no freeze-ups. The power failures that have occurred since the beginning of this molten-salt-reactor corrosion-testing program and the results of such failures are described in Table 1.2.2. ' The loops designated MSRP-12 and 9377-5 in Table 1.2.1 contain molten-salt-sampling devices. It was necessary to add 600 g of salt 63. L. Crowley, MSR Quar. Prog. Rep. June 30, 1958, ORNL-2551, p 36. Table 1.2.1. For " Composition | A imat i Loop Loop Material Number of pproximate Approxim Designation and Size Circulated Flow Rate Reynol. Comments 1o (gpm) Numbe Fluid 9354-3 INOR-8 84 2.8 ration Hot leg, % in. sched 40 4500 Cold leg, % in. OD, 5400 0.045 in. wall ‘ 9344-2 Inconel, }'2 in. OD, 0.045 12 2.5 8200 Jan. 28, 1959 after one year of in., wall 9377-3 Inconel, }'2 in. OD, 0.045 131 2 3400 ! March 18, 1959 after one year in. wall on 9354-1 INOR-8, J, in. OD, 0.045 126 2.5 2000 ration in, wall 19354-5 INOR-8, % in. OD, 0.035 130 1 2200 ration in, wall 9354-4 INOR-8 130 2.5 y insert removed Feb, 17, 1959 Hot leg, 98 in. sched 40 3000} hr of operation; loop restarted Cold leg, }'2 in. OD, ‘ 3500, fluid; normal operation other- , 0.045 in. wall MSRP-7 INOR-8, J in. OD, 0.045 133 1.8 3100 ration in. wall 9377-4 Inconel, % in. OD, 0.045 130 1.8 2600 ration in. wall MSRP-6 INOR-8, %, in. OD, 0.045 134 1.8 2300 ration in. wall MSRP-8 INOR-8, J, in. OD, 0.045 124 2 4000 rqtion in. wall MSRP-9 INOR-8, J, in. OD, 0.045 134 1.8 2300 ration in. wall ‘ MSRP-10 INOR-8, , in. OD, 0.045 135 2 3400 ration in. wall o MSRP-11 INOR-8, % in. OD, 0.045 123 2 3200 (ation in. wall MSRP-12 INOR-8, /, in. OD, 0.045 134 1.8 2300 cation in. wall 9377-5 Inconel, J, in. OD, 0.045 134 1.8 2300 /qtion in, wall 9377-6 Inconel, J in. OD, 0.045 133 1.8 3100 ation Feb. 26, 1959 in. wall 9Composition 12: NaF-KF-LiF (11.5-42-46.5 mole %) - Composition 130: Composition 84: NaF-LiF-BeF, (27-35-38 mole %) Composition 131:: Composition 123: NaF-BeF y-UF ; (53-46-1 mole %) Composition 133:, Composition 124: NaF-BeF ,-ThF 4 (58-35-7 mole %) Composition 134: Composition 126: LiF-BeF ,-UF ; (53-46-1 mole %) Composition 135:. bJ, L. Crowley, MSR Quar. Prog. Rep. Jan. 31, 1958, ORNL-2474, p 31. Ibid., p 32. d). L. Crowley, MSR Quar. Prog. Rep. June 30, 1958, ORNL-2551, p 36, Fig. 1 ®J. L. Crowley, A Sampling Device for Molten Salt Systems, ORNL-2688 (to be 33 Table 1.2.2. Summary of Power Faillures That Have Occurred During Operation of Forced-Circulation Corrosion~Testing Loops Number of Number of Number of Number of Loops Date Loops Loops Loops in That Failed as Operating Affected Which Salt Result of Power Froze Failure December 17, 1957 5 5 5 1 April 6, 1958% 9 6 6 3 August 5, 1958 7 7 éb None November T, 1958 1k 14 None None January 15, 1959 15 15 None None March 15, 1959 15 12 None None ®putomatic controls installed on most locops between April and August. bThese loops did not have automatic controls; only part of the salt froze. to replenish the inventory and thus raise the fluid level to the normal height in loop 9377-5 after 27 samples had been taken. The molten salt was introduced into the pump bowl through the pump flange without interruption of the flow or operation of the loop. Analyses of the samples removed from these loops periodically are presented in Section 2.2. In-Pile Loops The in-pile loop which was operated in the MIR during the previous quarter was disassembled, and the cause of the partially plugged purge line, which resulted in the release of activity, was determined. Oil which leaked past the pump shaft seal and filled the oil trap on the purge outlet line was polymerized by radiation, and the polymerized oil plugged the outlet line. Chemical and metallographic analyses of the fuel and container materials have not been completed. The second in-pile loop was completed and inserted during the MTR shutdown of April 27. Modifications were made to the internal 35 purge system to minimize the probability of purge line plugging. In addition, changes were made to the external purge system which should eliminate the possibility of a repetition of the activity release. These changes included the installation of a charcoal trap in the purge inlet line to prevent back-diffusion of fission gas into un- shielded lines. A prototype in-pile pump identical to that installed in the second loop has accumulated over 3600 hr of satisfactory operation. 36 1.3. ENGINEERING RESEARCH Physical Property Measurements Enthalpy and Heat Capacity Determinations were made of the enthélpies, heat capacities, and heats of fusion of three additional beryllium-containing fluoride salt mixtures: salt 133 (LiF-BeF2~ThFh, 71-16-13 mole %), salt 134 (LiF- BeF ,-ThF), -UF) , 62-36.5-1-0.5 mole %), and salt 136 (LiF-BeFe—UFu, 70- 10-20 mole %). The results of these measurements are presented in Table 1.3.1 as the constants a, b, and ¢ appearing in the correlating equations, 2 HT - HBOOC = a + bT + T and c._=Db+ 2T , p where H is the enthalpy (cal/g); Cy the heat capacity (cal/g.°C); and T, the temperature (OC). The experimental enthalpy data for these three mixtures, along with earlier_resultsl for related mixtures, are summarized in Fig. 1.3.1. It may be seen by comparing the results for salts 133 and 136 with the curve for mixture 130 (LiF-BeFe-UFu, 62-37-1 mole %) that altering the salt composition toward larger percentages of the high-molecular-weight fluorides (ThFu and UFA) causes a substantial decrease in both the enthalpy and heat capacity (on a unit weight basis). Replacement of the LiF in salt 126 (LiF-BeFQ-UFh, 53=46-1 mole %) with NeF (mixture 123; NaF-BeF,-UF), 53-46-1 mole %) results in similar, though smaller, reductions in the enthalpy and heat capacity. This effect will be studied further in order to establish a general correlation for the prediction of enthalpies and heat capacities in the LiF-BeFe— UFM-ThFZ system. lW. D. Powers and R. H. Nimmo, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p uh. 37 o3 Table 1.3.1. Enthalpy Equation Coefficients and Heats of Fusion of Several Fused Salt Mixtures Enthalpy Equation Heat of Fusion (cal/g) Salt Mixture Phase = Temperature Coefficients® Temperature - HS Range (°C) a b c (oc) x 1077 LiF-BeF ~ThF) -UF), Solid 100-350 -b.9 +0.22k +3k4.65 450 6L (62-36.5-1-0.5 mole %) Liquid 500-800 -2.1 +0.545 -4.03 LiF-BeF,~ThF) | Solid 100-400 -7.7 +0.209 +4.58 500 h (71-16-13 mole %) Liquid 550-800 b7 404730 -11.90 LiF-BeF ~UF), Solid 100-400 -5.6 +0.153 +3.50 450 53 (70-10-20 mole %) Liquid 550-800 -17.8 +0.289 -3.21 SThe enthalpy is given as HT - H3 ¢ =b + 2¢T. P = a + bT + CTQ; the heat capacity is then evaluated as Viscosity The viscosities of salt mixtures 133 and 136 have been obtained over the temperature range from 550 to 800°C with the "skirted" capillary efflux viscometer previously déscribed.2 The viscometer cups used for the measurements were those employed in the earlier studies with salts 130 and 134%. Recalibration of these cups with the NaNOZ-NaNO?)-KNO3 (40-7-53 wt %) mixture showed negligible changes in the kinematic viscosity—efflux time relation. The results, correlated in the form where u is the viscosity (centipoise), T is the temperature (%K), and A and B are experimentally determined constants, are as follows: Salt 133 A = 0.0526; B = L4838 , Salt 136 A = 0.0489; B = L4847 The date are plotted in Fig. 1.3.2 and compared with data from measurements of salts 123, 126, 130, and l3h.3 The "high" results with i i mixtures 123 and 126 will be rechecked with the new type of viscometer. In Fig. 1.3.3, the data for salt 136 are compared with Mound Laboratory results for the same mixture that were obtained with a Margules design viscometer consisting of two concentric Inconel cylinders. The outer of the two cylinders contalned the fused salt and was rotated at a fixed speed with reference to the inner stationary cylinder. The viscosity was determined from the torque exerted on the inner cylinder. The reason for the discrepancy between the two sets of data has not yet been established. Surface Tension The surface tension of molten-salt mixture 134 was experimentally determined using the maximum-bubble-pressure technique.3 The results, °MSR Quer. Prog. Rep. Jan. 31, 1959, ORNL-268k, p 65. 3W. D. Powers, MSR Quar. Prog. Rep. June 30, 1958, ORNL-2551, p 38. l;B. C. Blanke et al., Density and Viscosity of Fused Mixtures of Lithium, Beryllium, and Uranium Fluorides, MIM-1086 (March 23, 1959). 39 1, VISCOSITY (centipoise) 450 500 550 600 650 700 750 800 825 100 s — | e /I//p 90 CR 60 \3 %o 47‘(//.? ~ 65 LQUIDUS >~ ~TEMPERATURE (° \\ PN \\\ \ \\ \ 10— S~ — SALT COMPOSITION (mole %) e ———— N NO. LiF NoF BeFp ThFs UFg e N 8 — = 126 53 46 f — T~ . 123 53 46 1 \\s \\ 133 7 16 13 7 \\\ 136 70 10 20 BSOS \ ~ 4l— 134 82 365 10 05 BN 130 62 37 f TN 2_ | N 00014 00013 0.0012 0.0014 0.0010 00009 40 UNCLASSIFIED ORNL-LR-DWG 38454 [ T L!QUIDUS TEMPERATURE (°C) COMPOSITION {mole o) LiF NaF Befp, Thigy UF4 62 37 i 134 62 365 1 05 /A 126 Pl 53 46 ! / / 53 46 f P } 33 74 s 13 - 70 10 20 T i / — / 13024134 — T —— SALT NO. 130 500 400 _ SALT o < NO. g ° 130 = 126 = 123 & 133 o 136 £ 200 — i2] T i < 100 o} 0 Fig. 1.3.1. 400 500 TEMPERATURE (°C) TEMPERATURE (°C) 600 700 800 Enthalpy-Tempetature Relations for Several Beryllium-Containing Salt Mixtures. UNCLASSIFIED ORNL-LR-DWG 38152 Fig. 1.3.2. Viscosity-Temperature Relations for Several Beryllium-Containing Salt Mixtures. given in Fig. 1.3.4, can be represented by the equation, o (dynes/cm) = 272.2 - 0.143 T (°cC), to within +5% over the temperature range from 500 to 800°C. The salt 134 data are essentially identical with the results obtained for salt 130, except for a slight difference in the temperature dependence. Heat~Transfer Studies Flow calibration of the system designed for determihing surface film formation in molten-salt systems by heat transfer coefficient measurements has been completed. A Tygon-coupled mockup was used, and water was the working fluid. The test unit is shown in Fig. 1.3.5, and the key components are indicated. The over-all system pressure drop was obtained with a 100-in. mercury manometer connected at the pump inlet and outlet lines. The data indicate a pressure difference of about 48 psig for a flow of 2.8 gpm. This corresponds to a test-section Reynolds modulus of 11,000 at 1200°F with a salt such as mixture 130 (LiF-BeF, UF), 5 62-37-1 mole %). Initial attempts to determine the flow rate vs pump speed characteristic by a weight method were unsuccessful, and a rotameter was used in the final calibration. A magnetic-pulse pickup coupled with a Hewlett-Packard counter yielded the pump speed to within +1 rpm over any lO-sec counting period. The characteristic curve obtained was very nearly linear between 2000 and 6000 rpm, ranging from 1 to 3 gpm, respectively. Design work is continuing on the system support structure to ensure against buckling of the test sections due to thermal expansion and to prevent electrical short-circuiting of the test-section heating currents. Welding of system components is in progress. L VISCOSITY (centipoise) UNCLASSIFIED ORNL—LR-DWG 38453 TEMPERATURE (°C) 500 550 600 650 700 750 800 % 7 T [T T T ] AN REANN N s N O MOUND LABORATORY L 1.4 1.3 1.2 1.1 1.0 0.9 x1073 1/T K} Fig. 1.3.3. Comparison of Independent Measurements of the Viscosity of Salt Mixture 136 (LlF-Ber-UF4, 70-10-20 mole %). UNCLASSIFIED ORNL-LR~-DWG 38454 250 } SALT 434\ —. 200 = e | £ \'-7-_=.§ < SALT 130 * —— o 130 ) o~ = 150 ! g 134 .9 = jun] }_ w 100 Q E o 2 1] b 50 0 400 450 500 550 600 650 700 750 800 850 TEMPERATURE (°C) Fig. 1.3.4. Surface Tension of LiF-Ber-ThF4-UF4 (62-36.5-1-0.5 mole %). Lo UNCLASSIFIED PHOTO 464147 HEAT EXCHANGE €4 Fig. 1.3.5. Flow Calibration Mockup of Molten-Salt Heat Transfer System. Hydrodynamic Studies The pressure drop through the sintered-metal-filled annulus of & double-walled tube has been determined.5 A tube of this construction has been proposed for use in steam-generation systems associated with molten-salt power reactors. It is hoped to provide good thermal contact between the low-pressure primary salt and the high-pressure steam and at fihe same time effectively isolate the two fluids in the event of s wall failure. The sintered medium would be filled with static helium, at an intermediate pressure, which would give an indication of a bresk in either wall. | The results of this study of the flow of helium through two speci- mens of double-walled Inconel tubing containing sintered Inconel of different porosities are given in Fig. 1.3.6. The correlating lines can be expressed by the equations: | 0.979 AP = 0.342 ( L) for specimen 1, with a porosity of approximately 59%, and 0.960 G = 0.0046 (FAIM) g L for specimen 2, with a porosity of approximately 25%. In these corre- lating expressions, G is the mass velocity based on the geometric cross-sectional area (1b /ft3), P is the mean helium den81ty (1p /ft3), AP is the pressure drop through the sintered material (1b /ft ), and L is the length of the sintered section. The experimental precision was estimated to be +8%. The two curves in Fig. 1.3.6 are nearly parallel, with slopes only slightly less than unity, and are in accord with the statement of Darcy's law for the flow of gases in porous media, °J. L. Wantland, H. W. Hoffmen, and R. L. Miller, Flow Through Sintered Inconel Annuli, ORNL CF 58-9-36 (Sept. 17, 1953). 6 As opposed to an estimated pore void area at any cross section. Ll UNCLASSIFIED ORNL—LR-DWG 32943 A 100 (] & A 50 & SPECIMEN NO. {; 0® POROSITY, ~ 59 % g8y — o | / & X 20 y 9 o o® /t‘ ’“) —~ © N y. e - / £ ® o2 £ o .DE ,// ® - - P > SPECIMEN NO. 2 4 £ 0 POROSITY,v25%, ~ ¢® o [ .| ,' L) > “. a 7 a yd 2' " S 2 / °/ /| P 1 L / v L 4 0.5 0.2 04 7 to 100 1000 5000 _ _ 6 PAP/L by 1o, [17) Fig. 1.3.6. Flow vs Pressure Drop Characteristics for Two Annuli Filled with Sintered Inconel. hs If it is assumed that the data follow Darcy's law exactly, the specific permeability coefficient, Bo’ can be obtained for the two samples by using the given porosities and mid-range vaelues of G, and PAP/L. In -12 .2 this manner, B_ was calculated to be ~62 x 10 £1° for specimen 1 and o ~164 x 10-14 £t2 for specimen 2. These results can be compared with B_ values of 216 x 10~1° £t° for an unconsolidated send of 37% porosity and 68 x lO"lh porosity. In addition, the slope of approximately unity indicates that £t° for consolidated small alundum particles with a 24% the flow is in the laminar regime. L6 1.4, INSTRUMENTS AND CONTROLS Molten-Salt-Fuel lLevel Indicators Visual and x-ray examination of a level probe used in the tests described previouslyl have been completed. The probe appears to be in excellent condition, both mechanically and electrically. Disassembly of the probe is in progress, and specimens have been sent to the metallurgy department for corrosion examination. 'Two new Inconel "I'"-tube-type level elements have been completed and installed in test vessels. Two identical level-measuring systems are now complete, and prefilling checkouts of the systems are in progress. Fabrication of an INOR-8 level element and test vessel is approxi-' mately 80% complete. INOR-8 High-Temperature Pressure Transmitters An order was placed with the Taylor Instrument Company for six INOR-8 pressure transmitters and indicating systems. The units are scheduled for delivery in early June. The transmitter body will be fabricated from INOR-8 plate. The diaphragm material will be 0.005-in.-thick INOR-8 shim stock. The manufacturer believes on the basis of preliminary tests that satisfactory diaphragms can be made of this material. No unusual difficulties are expected in the fabrication of these units. The six units will be composed of three having a range of O to 50 psig, and three with a range of O to 100 psig. All units are of the pneumatic-indicator type. 1R. F. Hyland, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p L47. L7 PART 2. MATERIALS STUDIES 2.1. METALLURGY Dynamic Corrosion Studies Corrosion studies were completed of four Inconel and four INOR-8 thermal-convection loops, and 12 thermal-convection-loop tests were initiated. The operating conditions for the new loops are given in Table 2.1.1. Two forced-~circulation loops have completed the scheduled Table 2.1.1. Operating Conditions for New Thermal Convection-Loop Tests Composition Number Maximum Fluid-Metal Scheduled Loop of Salt Being Interface Operating No. Material Circulated® Temperature (OF) Period 1236 Inconel 134 1250 1 yr 1237 Inconel 134 1350 1 yr 1238 INOR-3 134 1350 1yr 1239 Inconel 133 1350 1 yr 1240 INOR-8 133 | 1350 1yr 1247 INOR-8 136 1250 1000 hr 12k2 INOR-8 136 1350 1000 hr 1243 Inconel 135 1250 1000 hr 124 INOR-8 135 1250 1 yr 1245 Inconel 135 1250 1yr 1246 INOR-8 135 1350 1lyr 1247 Inconel 135 1350 1l yr fComposition 134: LiF-BeF ~ThF) -UF) (62-36.5-1-0.5 mole %). Composition 133: LiF-BeF,-ThF) (71-16-13 mole %). Composition 136: LiF-BeF -UFM4(70-10-20 mole %). Composition 135: NaF-BeF,-ThF) -UF) (53-45.5-1-0.5 mole %). N oM 51 test program but have not yet been examined, and one new forced- circulation loop test was started. The present status of all forced- circulation-loop tests now in progress is given in Chapter 1.2 of this report. INOR-8 Thermal-Convection Loops Metallographic examination of INOR-8 loop 1185, which circulated salt 126 (LiF-BeFe-UFh, 53-46-1 mole %) for one year at a hot-leg temperature of 12500F, revealed no observable attack in either the hot- or cold-leg sections. A photomicrograph of a typical hot-leg section from this loop is shown in Fig. 2.1.1. Further, no evidence of mass transfer was found in examinations of cold-leg sections or analyses of after-test salt samples. Results of salt analyses (presented in Table 2.1.2) indicate the level of metallic impurities Table 2.1.2. Analysis of Salt Mixture 126 Before and After Circulation in Loop 1185 for One Year Major Constituents Minor Constituents (3t %) | (ppm) Sample Taken U Be Ni Cr Fe Before test 6.53 10.1 30 40 235 After test | 6.4o 10.6 35 140 250 in after-test salt samples to be effectively the same as in before- test samples. Metalloéraphic examination results for the INOR-8 loops 1227, 1228, and 1229 are summarized in Table 2.1.3. These three loops were each operated for 1000 hr, and the maximum fluid-metal interface temperature was 12500F. Metallographic examinations of the hot and cold legs of these loops showed no evidences of surface defects of the types normally associated with fluoride salt attack; however, well-developed intergranular cracks from 0.5 to 2 mils deep appeared 52 €S Table 2.1.3. Results of Metallographic Examinations of Inconel and INOR-8 Thermal Convection Loops Meximum Fluid- Metal Interface Metallographic Results Loop Test Tempgrature Salt Hot-Leg Cold-lLeg No. Material Period (°F) No.? Appearance Appearance 1182 Inconel 1 yr 1350 126 Heavy intergranular voids Surface pitting 0.5 to a depth of 15 mils mil deep 1188 Inconel 1 yr 1250 8k Heavy intergranular voids Grain-boundary to a depth of 9 mils penetrations to a depth of 1 mil 1189 Inconel 1yr 1250 130 Heavy intergranular voids Grain-boundary to a depth of 7 mils penetrations to a , depth of 0.5 mil 1230 Inconel 1000 hr 1350 130 Moderate intergranular Few penetrations voids to a depth of 4 < 1 mil deep mils 1227 INOR-8 1000 hr 1250 134 No attack No attack 1228 INOR-8 1000 hr 1250 133 No attack; fabrication No attack flaws found in both legs 1229 INOR-8 1000 hr 1250 135 Same as loop 1228 Same as loop 1228 qComposition 126: Composition 8k: Composition 130: Composition 13k4: Composition 133: Composition 135: LiF-BeF,-UF), (53-46-1 mole %). LiF-NaF-BeFs (35-27-38 mole %). LiF-BeF,-UF), (62-37-1 mole %). LiF-BeFo-ThF) -UF), (62-36.5-1-0.5 mole %). LiF-BeF,-ThF), (71-16-13 mole %). NeF-BeF 5-ThF), -UF), (53-45.5-1-0.5 mole %). at random intervals along the exposed surfaces of all three loops. An examination of as-received specimens of tubing used for these loops was subsequently made, and similar cracks were detected in these specimens. A typical flaw is shown in Fig. 2.1.2. The flaws were apparently not detected Dby x-ray inspection of the tubing before loop fabrication. Therefore ultrasonic inspection of loop tubing is now being utilized to supplement existing inspection practices. Inconel Thermal-Convection Loops Three of the four Inconel loops for which metallographic exami- nation results have been recently obtained were operated for one year, while the fourth was operated 1000 hr. The examination results are summarized in Table 2.1.3. As indicated, the hot legs of all the Inconel loops operated for one year showed heavy intergranular void formation. Of particular interest is loop 1182, which circulated salt 126 for one year at 13500F. Examination of this loop revealed heavy intergranular void formation to a depth of 15 mils, as may be seen in Fig. 2.1.3. Chemical analyses of the salt showed no significant changes in the impurity content. However, examination of the trap area, which was the coldest part of the loop, revealed a large quantity of metallic particles. These particles are being analyzed. General Corrosion Studies Penetration of Graphite by Molten Fluoride Salts The problems associated with the use of unclad graphite as a moderator in a molten-salt reactor system are being studied. The possibility exists that fuel salts would enter the pore spaces in the graphite and create problems associated with (1) reactor fuel inventory, (2) hot spots, (3) effects from fission-gas release associated with fuel in the pores, and (4) effects of thermasl expansion of the salt as a result of thermal cycling of the reactor. The large volume change which fluoride salts exhibit on melting could cause cracking or spalling of the graphite if sufficient pressure were built up by salt entrapped 54 Fig. 2.1.1. Specimen Taken from Hot Leg of INOR-8 Thermal-Convection Loop 1185 at Point of Maximum | ed ) UNCLASSIFIED | X~ RN B UNCLASSIFIED T-16554 e ' ' : 4 T-16822 007 008 Q08 010 Reik] 012 013 Loop Temperature (1250°F). Loop was operated for 1229. 1 year with salt mixture LiF-BeF,-UF, (53-46-1 mole %). UNCLASSIFIED fl T-16577 T Q L 206 F 0 Y FE _ n (& = E ,il|||1|l||t|i| N—_———"_=————————__— N RN/ /7 : 3 V% ,, \\\ 3 //fl////// ////\\\\ /7 = SN S5 7 e e ————————\ /7 Nt — N\ \ Nyl — — Q) ,/6, / / W ///N/W/, /% = /,, N S = b e e S Fig. 2.1.4. Test Setups Used to Study the Precipitation of uo, from Fuel 130 in Graphite Crucibles. 59 in contact with a small area of a small quantity of graphite to determine, in conjunction with the results of test D, whether the quantity of graphite affected the quantity of uranium precipitation from the fuel. Radiographic examinations of each of the test setups at total accumulated test times of 5, 10, and 100 hr were made. No uranium precipitation was observed for the control, test A (Fig. 2.1.k). A small amount of uranium precipitation occurred in test E. A moderate and approximately equal amount of uranium precipitation was found in each of the other three tests. These results support the conclusion arrived at previously2 that the uranium precipitation observed in fuel 130 was the result of the fuel reacting with oxygen supplied by degassing of the graphite. Tests are in progress to determine whether a suitable flush can degas the graphite to such a degree that it will contain fuel 130 without causing uranium to precipitate from the fuel. Thermal-Convection-Loop Tests of Brazing Alloys in Fuel 130 A scheduled 1000-hr test of Inconel and INOR-8 lap joints brazed with various alloys and exposed to fuel 130 in the hot leg of a thermal- convection loop has been completed. Similar tests scheduled for 5,000- and 10,000-hr periods are continuing. The configuration of these thermal-convection loops and the manner in which the brazing alloys are incorporated in them was described previously.3 The brazing alloys tested were: Alloy Composition Coast Metals No. 52 89% Ni—5% Si-4% B—2% Fe Coast Metals No. 53 81% Ni-8% Cr—4% Si—4% B-3% Fe General Electric No. 81 70% Ni—20% Cr—10% Si Gold—Nickel Alloy 82% Au—-18% Ni Copper 100% Cu W. H. Cook, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p 62. : 3E. E. Hoffman and D. H., Jansen, MSR Quar. Prog. Rep. June 30, 1958 ORNL-2551, p 62. . 60 The temperature of the circulating salt in the region of the test s?ecimens was l3OOOF. Metallographic examination showed that all the brazing alloys had good flowability on both Inconel and INOR-8 base metals. There was a tendency for the formation of diffusion voids in the fillets of the joints brazed with the gold—nickel alloy. The General Electric No. 81 alloy was heavily attacked on Inconel base material. Coast Metals Nos. 52 and 53 were depleted at the fillet surface to a depth of 1 to 2 mils, Some slight cracking of the alloys occurred at the brazing alloy-base metal interface. Pure copper showed good corrosion resistance and no cracking. Thermal-Convection-Loop Tests of the Compatibility of INOR-3, Graphite, and Fuel 130 : Fuel mixture 130 (LiF-BéF2-UFu, 62-37-1 mole %) was circulated in an INOR-8 thermsl-convection loop containing a 10-in. tube of TSF graphite in the hot leg in order to ascertain whether the graphite would cause carburization of the INOR-8. The loop operated for 4000 hr with a hot-leg temperature of 1300°F. | Metallographic examination of a loop section adjacent to the graphite insert revealed that the carbide precipitates in the grain boundaries were slightly larger near the salt-metal surface than throughout the rest of the specimen (Fig. 2.1.5); however, similar precipitates have been observed in INOR-8 tubing from corrosion loop tests with no graphite present. Microhardness measurements of the area 2 mils from the salt-metal surface were of the same magnitude as those observed in the rest of the specimen. These values ranged randomly from 160 to 180 DPH. A carbon content of 0.037% was found by chemical analyses of two successive 5-mil layers taken at the inner surface of the INOR-8 tubing. An additional series of layers to a depth of 35 mils was analyzed and found to contain a range of carbon values from 0.01% to 0.026%. Consequently, these different values are not considered to be significant, especially in view of the large quantities of inclusions 61 003 04 UNMCLASSIFIED Y -29049 0 UNCLASSIFIED 62 'Ber' te cap- iF ion Loop that Circulated L Convect 8 Specimen Taken from a Thermal fuel 130) for 4000 hr at a Hot-Zone Temperature of 1300°F The (a) Outer and (&) Inner Surfaces of an INOR Fig. 2.1.5 UF, (62-37 jacent to a graphi ion ad is from a sect imen The spec -1 mole %, in the hot leg. aqua regid. Etchant sule incorporated in this heat of material, as shown in Fig. 2.1.6. Mechanical Properties of INOR-8 Creep Tests Most of the creep tests presently being conducted in fused salts are low-stress, long-time tests. The results of these tests will pro- vide accurate data in terms of stress, total strain, and creep rates for INOR-8 at times in excess of 10,000 hr. None of the test specimens failed during the past quarter and therefore no data are available to report. These experiments are being run at 1100 and 1200°F, Fatigue Studies The fatigue properties of INOR-8 at 1100 and 1500°F are being studied at Battelle Memorial Institute under a subcontract from the Metallurgy Division. A rotating-beam-type of test is used at frequencies of 100 and 3000 rpm. The initial results reported were for the 100 rpm and lBOOOF conditions. The data are summarized below: Stress (psi) Cycles to Failure 45 x 107 8 x 103 40 | 34 x 10 35 160 x 103 30 730 x 10° 30 930 x 10° 27 72 x 10° (discontinued) Thus, the stress to produce fatigue failure in 1 x 106 cycles is slightly over 29,500 psi. Under similar test conditions, the stress to produce fatigue failure of Inconel in 1 x 106 cycles is about 18,000 psi. Shrinkage Characteristics of INOR-8 A certain peculiarity in the behavior of INOR-8 has beeh noted almost since the start of the current testing program. In creep tests, 63 6L B e & aiaey P INCHES CLASSIFIED “¥-29052 e Fig. 2.1.6. Cross Section of INOR-8 Tube that was Adjacent to Graphite Insert in 4000-hr Thermal-Convection Loop Test. there is an initial period during which virtually no creep occurs. This is followed by normal creep. In relaxation tests, the load must be increased during the early stages in order to maintain a constant strain. These observations are important, since at low stresses the effects are manifest over a thousand or more hours and the results in terms of creep rate and total strain vs time measurements are attenuated. The behavior pattern suggested that the metal might be contracting as a result of metallurgical instabilities. Isothermal dilatometry measurements indicate that there is contraction in certain heats of INOR-8 but not in others, notably SP-16 and SP-19. Metallographic examinations showed that those specimens which contracted contained many large areas of second phase material. The stable heats had equi- axed grain structures with fine precipitates concentrated mainly in the grain boundaries. There appear to be no significant differences in either the major or minor alloying constituents in the heats examined, except for the carbon content. However, stable heats have been found that range from a special low-carbon melt up to heats containing carbon in excess of 0.1%. The specimens which contracted had carbon contents which fell within this range. Currently the contraction phenomenon is being studied by means of resistivity measurements to determine temper- asture-dependence and reaction rates. This study will be continued, since all heats have shown a plateau in the creep and relaxation values. Materials Fabrication Studies Effect on INOR-8 of Aging at High Temperatures The experiments which were being run to determine whether INOR-8 exhibits a tendency to embrittle in the temperature range between 1000 and 1400°F have been completed. ©Specimens which had been aged for 10,000 hr have been tensile tested, and no significant differences from annealed specimens have been found. The data from these tests are presented in Table 2.1.k. 65 99 Table 2.1.Lk, from Heat SP-19 (0.06% C) Effect of Aging Upon the Tensile Properties of INOR-8 Specimens Specimen Annealed 1 hr at 2100°F Specimen Annealed and Aged 10,000 hr at Test Temperature Test - - - - - Tbmfg;§ture gi?:;;ih éii:igth El°?§§ti°n Temfiiifiiure gifi:;éih si;:igth El°?§§ti°n (psi) (psi) (°F) (psi) (psi) ’ Room 114,400 Lk, 700 50 1000 117,700 45,100 50 1100 120,000 47,500 49 1200 116,000 46,800 46 1300 115,500 45,800 43 1400 115,000 43,600 40 1000 93,000 28,300 L6 1000 97,700 L6 1100 93,000 28,900 50 1100 90, 500 Lo 1200 82,400 27,500 37 1200 81,900 32 1300 69,900 28,000 2l 1300 76,100 2l 1400 61,800 26, 200 21 1400 65,100 20 It is concluded from this group'of tests and from the previous tests'?D of specimens aged for 500, 1000, 2000, and 5000 hr that INOR-8 does not exhibit any embrittling tendencies that can be attributed to high-temperature instability. Triplex Heat Exchanger Tubing Work is continuing on the development of techniques for fabricating a heat exchanger tube made up of two concentric tubes with a porous annulus between to permit the passage of a gas for leak detection. The materials of construction presently being investigated are a porous nickel core clad on the outer and inner surfaces with Inconel or INOR-8. Two methods of incorporating the porous core into the annular space of the concentric tubes are.being considered: first, tamping loose nickel powder into the annulus and developing a porous core bonded to the - outer and inner tubes by suitable drawing and sintering operations; and, second, bonding a prefabricated porous core to the outer and inner tubes by suitable drawing and sintering operations. Results obtained by the first method were described previously. Because of difficulties encountered with an oxide layer being formed at the core-to-cladding interface and causing poor bonding, this method of fabrication has been discontinued in favor of the second method. Porous nickel sheet is available from Micro Metallic Corp. in four common grades, which differ in mean pore opening, as follows: Grade Mean Pore Opening (in.) E 0.0015 F | 0.0008 G ' 0.000k4 H 0.0002 ILH. Inouye, MSR Quar. Prog. Rep. June 30, 1958, ORNL-2551, p 67. °H. Inouye, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p 67. 6MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-2684, p 85. 67 The sheet material has a density of approximately 50 to 60% of theoretical, and 1/16-in.-thick stock can be rolled into a 3/Wk-in.-OD tube. Twenty-five feet of both grades E and G rolled into tubular shape have been ordered for cladding studies. An attempt is presently being made to clad with Inconel a small, sample, porous nickel strip received from Micro Metallic. The as- received strip was annealed for 1 hr at 1800°F and then bent around the outside of a 0.625-in.-0D Inconel tube. The porous nature of the resulting ring can be seen in Fig. 2.1.7, which shows an enlarged view of the surface. Channeled spacers were tack-welded to the Inconel tube adjacent to the porous ring to prevent slippage of the ring during drawing and to provide a means of filling the annular space. The assembly is shown in Fig. 2.1.8. An Inconel tube was placed over the assembly and the resulting triplex assembly was cold drawn from 1.000 in. OD x 0.500 in. ID to 0.855 in. OD x 0.468 in. ID. The tri- plex aséembly is now being preparéd for sintering to promote bonding of the porous ring to the Inconel cladding. A vacuum will be maintained in the annulus of the triplex assembly during sintering. An evaluation of the porosity and bond between ring and cladding will subsequently be made. Similar cladding experiments are to be performed by Superior Tube Co. on Subcontract No. 1112. It is anticipated that these results will help determine the feasibility of fabrication of a triplex tube by a commercial vendor. Welding and Brazing Studies Procedures for Welding INOR-8 The fabrication of INOR-8 components for the wide variety of MSR applications requires that suitable welding procedures be developed for material ranging in size from thin-walled tubing to heavy plate. An inert-arc welding specification, RMWS-12 (ref 7) is available for "R. M. Evens (comp. and ed.), Reactor Material Specifications, TID-7017, p 195 (Oct. 29, 1958). . 63 Fig. 2.1.7. Surface of Prefabricated Porous Nickel Strip Received from Bending Around a 0.625-in.-OD Inconel Tube. 7]/2X. INCHES 1 2 I I I | { Micro Metallic Corp. After UNCLASSIFIED Y-29097 Fig. 2.1.8. Assembly for Cladding Porous Nickel Ring with Inconel. Channeled spacers provided for restraining ring, facilitating evacuation of assembly, and filling annular space. 69 INOR-8 material up to 0.100 in. thick, and additional procedures are being qualified for thicker sections in accordance with the methods prescribed by the ASME Boiler Cod.e.8 Several lengths of 5-in. sched-4O pipe have been fabricated from l/h-in.-thick plate, and a welding procedure will be developed which will be applicable over the thickness range 1/8 to 1/2 in. A welding procedure for l/2~in. plate will also be developed, which will be applicable for thicknesses up to 1 in. The tests to be conducted include reduced-section tension, side-bend, face-bend, and root-bend tests, as well as radiographic and metallographic examinations. It is expected that these welding procedures will be extremely useful in fabricating containment vessels and test components from INOR-8. These procedures will serve also as an essential supplement to the mechanical property data which are being accumulated for the presentation of INOR-3 as a case before the ASME Boiler and Pressure Vessel Code Committee. - Mechanical Properties of INOR-8 Welds Studies are continuing in an effort to improve the high-temperature ductility of INOR-8 weld metal. These include the investigation of techniques to deoxidize and purify weld filler metal during the casting of original ingots. One promising method involves the use of a vacuum- | melted basic charge with additions of small quantities of aluminum, titanium, manganese, silicon, boron, and magnesium. Seversl different heats of weld metal éontaining these and other additions have been cast and fabricated into weld wire. Analyses of these heats are presented in Table: 2.1.5, and analyses of commercial materials are included for comparison. The mechanical property values obtained to date for these materials are summarized in Table 2.1.6, and properties of other materials ére included for comparison. The results of the mechanical property tests on these filler metals will permit an evaluation of the relative merits of the various deoxidation and purification practices. BASME Boiler and Pressure Vessel Code, Section IX (1956). 70 T. Table 2.1.5. Containing Alloying Additions 72% Ni—16% Mo—-4% Fe-8% Cr Basic melt charge: Analyses of INOR-8 Weld Metal Heats Carbon . Heat Content Analysis Additive (wt %) Number (wt %) Source Mn Si Al i R Mg Hastelloy W 0.03 ORNL 0.58 0.07 0.15 0.08 0.02 (25% Mo—T7% Fe—5% Cr— 2.5% Co—bal Ni) Westinghouse M-5 0.08 Vendor 0.79 0.19 0.02 Haynes SP-19 0.06 Vendor 0.48 0.0k ORNIL, MP-3 0.06 Intended 0.50 0.10 0.20 0.20 0.005 0.03 ORNL 0.36 0.25 0.25 0.20 0.00k 0.02 ORNL MP-k 0.06 Intended 0.50 0.10 0.20 0.20 0.03 ORNL 0.3+ 0.05 0.20 0.24 0.002 cl Table 2.1.6. Mechanical Property Studies on INOR-8 Weld Metal Tensile Strength (psi) Flongation (% in 1 in.) Heat At Room At Room Number Temperature At 1200°F At 1500°F Temperature At 1200°F At 1500°F Hastelloy W 127,000 94,800 70,400 37 33 25 Westinghouse M-5 118,900 4, 000 55,800 36 18 5 Haynes SP-19 115,900 70,400 53,000 38 18 10 ORNL MP-3 108,000 67,800 49, 200 46 o5 13 ORNL MP-4 107,600 68,000 50,100 45 22 9 Fabrication of Apparatus for Testing the Compatibility of Molten Salts and Graphite Equipment was fabricated which will be used by the General Corrosion Group to study molten salt penetration of graphite in a dynamic, high- pressure system. Before this apparatus could be fabricated, it was necessary to develop a‘method of attaching an Inconel tube to a hollow cylindrical graphite specimen and thereby forming a leaktight connection. This was accomplished by brazing with a commercially available braze alloy composed of silver, titanium, and copper that was found to wet vacuum-degassed graphite. The large difference in thermal expansion of the graphite and the ‘brazing alloy caused some shear cracking in the graphite. However, | metallographic sectioning of sample joints indicated that the cracks were rather limited in length, and pressure testing of.the completed joint with air indicated no leakage other than that associated with the inherent porosity of the base graphite. | The graphite-to-Inconel assembly prior to brazing is showvn in Fig. 2.1.9, and the completed specimen is shown in Fig. 2.1.10. The completed Inconel test rig with associated entry, drain, and purge lines is showvn in Fig. 2.1.11. 3 UNCLASS‘ FIED Y.28696 Fig. 2.1.9. Grcp‘nite-to-\ncone\ Assemb\y Befote Brazing UNCLASS\F!ED Y.28695 Fig. 2.1.11% Comp\e'ted inconel Test Rig: Fig. 2.1.10. Graphife-io-lncone\ Assembly Afrer Brozing. 2.2. CHEMISTRY AND RADIATION DAMAGE Phase Equilibrium Studies The System LiF-BeF _-ThF, A revised phase diagram for the system LiF-BeFE-ThFu is presented in Fig. 2.2.1 that includes data obtained by increasing the equilibration period in thermal-gradient quenching experiments to three weeks. Quenched samples from such experiments revealed that the area of single-phase ternary solid solutions involVing 3LiF~ThFh is greater than previously reported;l’2 it occupies approximately the triangle indicated by cross hatching in Fig. 2.2.1. Invariant equilibria in the system LiF»BeFE-ThFh are listed in Table 2.2.1. The limits of the area of the single-phase Teble 2.2.1. Invariant Equilibria in the System LiF«BeFQ-IhFu Composition (mole %) Invariant Type of Solids Present at LiF BeF2 ThFh Temperature Equilibrium Invariant Point (ec) 17 81 2 Lot Peritectic ~ ThFj, LiF-4ThF), » and BeFp 33.5 6b 2.5 451 Peritectic LiF.A4ThF), LiF-2ThF), \ and BeF2 L7 51.5 1.5 355 Eutectic 2LiF.BeF,, LiF.2ThF),, and BeF, 60.5 36.5 3 431 Peritectic LiF-2ThF), 3LiF-ThF) (ss), and 2LiF-BeF, 66 2.5 4.5 448 Peritectic LiF, 2LiF-BeF,, and 3LiF.ThF) (ss) 63 29 8 452 Peritectic 3LiF.ThF), (ss), TLiF-6ThF),, and LiF. 2ThF), 1R. E. Thoms et al., MSR Quar. Prog. Rep. June 30, 1958, ORNL-2551, 2R. E. Thoma et al., MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, P 9. 75 UNCLASSIFIED ORNL-LR-DWG 37420 Thi, 11H4 TEMPERATURE IN °C COMPOSITION IN mole %o LiF-4ThE, 7LiF-6ThE T A LiF-2The, 3LiF~ThF4 sS 448 LiF-2ThE, ‘ P 897 A 2LiF-BeF, [N 7TLIF-6ThE N \ P 597 P 565 \ 3LiF-ThE, FAN \‘ SORS S, e LiF AN 845 2LiF-BeF2/5 ¥ Ce*F3 (a) + LaF3 (s)= LaF3 (a) + Ce F3 (s) (2) _ is an analogue of exchange (1). Thus Ce*F3 (1abeled), which has desirable tracer characteristics, can be used in exchange (2) to approximate the behavior of SmF3 in exchange (1). A proposed method of salt purification utilizing exchange (1) would be to pass the salt through an isothermal bed of solid Ce}?3 to lower the S.m’F3 content and then to lower the temperature of the effluent salt to decrease the total rare earth content. To determine the feasibility of the method, information is required on the rate of exchange. The first exploratory rate test was conducted at 5OOOC with an agitated CeF3-LiF-BeF2 melt (1500 ppm CeF3, 62 mole % LiF, 38 mole % BeFE) to which an excess of LaF3 was added. Within 1 min, the CeF content was decreased to 400 ppm; within 5 min the system was at 3 equilibrium (300 ppm CeFB) with respect to exchange (2). The second test employed a 1lk-in. horizontal column of 3/k-in. tubing packed with 100 g of +18 mesh LaF.. The packed colum was charged at 300°C with LiF-BeF,, (62-38 mole %) containing 1000 ppm CeFB, and 12 g of melt, not counting the holdup volume, was transferred through the LaF. at 3 18 gpm with a pressure drop of 3 psi. The CeF. content of the effluent ranged from 80 to 30 ppm. Attempts to continuz the experiment at a ) later time with the same packing gave evidence that the liquid was bypassing the LalF3 (s) and that future tests with vertical columns should give better performance. Chemical Reactions of Oxides with Fluorides in Molten-Fluoride-Salt i Solvents The chemical reactions of oxides with UFM in molten fluoride solvents have indicated that uranium can be separated from fission products by fractional oxide precipitation. As mentioned previously, 8J. H. Shaffer, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p 92. 82 BeO and water vapor appear to be desirable precipitating agents because they do not introduce extraneous constituents into BeF2 mixtures. Investigations have continued on the removal of uranium from solution by reaction of UFH with BeO in a column-extraction process and on the fractional precipitation of uranium with water vapor. Extraction of Uranium with BeO. The reaction of UF), with BeO to produce UO2 has been studied in several experiments.8’9 Observations of the reaction in a packed column show several interesting characteris- tiecs. A eutectic mixture of LiF-KF (50-50 mole %) containing 1 mole % UFh was twice passed unldirectionally through a small column packed with extruded BeO (1/16 in. in diameter and 1/8 in. long) at 600°C. The column contained approximately three times the amount of Be( o+ During the first pass, 34 .6% of the uranium was removed from 2 kg of liquid mixture at an average required for converting the UFA to U0 flow rate of 700 gpm under a pressure drop of 5 psi. A second pass removed an additional 30% of the initial uranium from solution. How- ever, the average flow rate during the second extraction was consider- ably slower (30 gpm), since a pressure drop of approximately 15 psi developed. A third extraction was not attempted. Calculations based on the experimental data show that 16% of the beryllium oxide in the column reacted with UF&‘ It is interesting to note, however, that only 18% of the reacted beryllium was present in the column effluent at the completion of the experiment. A similar retention of Bng,_probably on the surface of the BeO pellets, was previocusly noted in the conversion of Zth to ZrOe. In future experiments, the effectiveness of BeO beds for removing lower concentrations of uranium from molten fluoride mixtures will be investigated. Precipitation with Water Vapor. The effective removal of uranium as UO, from an LiF-BeF, (63-37 mole %) solvent by the reaction at 600°C 2 95. H. Shaffer, MSR Quar. Prog. Rep. June 30, 1958, ORNL-2551, 83 of water vapor with UFA has been demonstrated.B In reprocessing molten fluoride reactor fuels, however, the separation of uranium from the fission-product rare earths would be a primary requisite. An experiment has shown that the reaction of water vapor with CeF3 in LiF-BeF, (63-37 mole %) is too slow for detection by a fairly sensitive radiochemical tracer technique (precision, <:ip.2%). In two subsequent experiments, water vapor was allowed to react at 600°C with UF) (1 mole %) in LiF-BeF, (63-37 mole %) containing 1.25 wt % CeF3. In each case, UO2 was formed without any detectable precipitation of Ce203 or BeO. The urenium concentration in scolution at the stoichi- ometric point of the reaction was on the order of 100 ppm in each experiment. The sharpness of this separation is very favorsble from the standpoint of reprocessing schemes, and it points to a method for removing uranium without using fluorine. It also leads to a method for removing uranium and thorium simultaneously that would permit the rare earths to be selectively precipitated in another step that could be followed by redissolution of the uranium and the thorium in reclaimed barren solvent. An experiment for studying the reaction of water wvapor with'UFu dissolved in LiF-KF (50-50 mole %) indicated a much lower rate of reaction as a result of strong complexing of both 0 and Uh+ ions in the melt. A negligible quantity of HF was evolvéd, and x-ray | diffraction examinations of filtrate samples taken from the reaction mixture showed the presence of KF-2H20. Treatment with 8500 meq of water resulted in the precipitation of 64 meq of uranium from solution and the evolution of 30 meq of HF. A comparison with the reaction of water vapor with UF, dissolved in LiF-NaF (60-40 mole %) is planned as a further study. 8l Gas Solubilities in Molten Fluoride Salts Solubility of Neon in LiF-—BeF2 The investigation of the solubility of neon in a mixture of LiF-BeF2 (64-36 mole %) has been completed at 500, 600, 700, and 800°C covering the pressure range of O to 2 atm. As in &ll previous studies of the solubility of noble gases in molten fluoride salts, Henry's law is Obeyed, and the solubility increases with temperature. The solubility decreases with increasing atomic weight of gas, and the enthalpy of solution increases. Henry's law constants for this system are (3.09 + 0.09) x 10-8, (k.63 + 0.01) x 10‘8, (6.80 + 0.09) x 10*8, and (9.01 + 0.15) x 107 moles of He per e of melt per atm at 500, 600, 700, and SOOOC, respec- tively. A heat of solution of 5880 cal/mole was calculated from the temperature dependence of these constants. Solubility of CO, in Na'F-BeF2 The solubility of CO, in molten NaF-BeF, (57-43 mole %) has been examined at 500, 600, 700, and 800°C at pressures from O to 2 atm. The experimental data illustrate the linear dependence of solubility with saturating pressures. Henry's law constants, K, at these temperatures are (7.93 + 0.36) x 10'8, (7.05 + 0.1k4) x 10'8, (8.32 + 0.08) x 10"8 moles of CO2 per.cms of melt per atm. Of all the gas solubility data for molten salt systems obtained to date, these eare the first to show a nonlinear log X versus 1/T plot. A minimum is observed on this plot in the vieinity of 600°C. -Preliminéry measurements of 002 solubility in the eutectic mixture LiF-NaF-KF (46.5-11.5-42 mole %) were questionable. It appears that the presence of dissolved oxygen-containing species increases the solubility of 002; lously purified solvent before and after adding known quantities of a This will be checked by measurements in a meticu- soluble oxide. 85 Chemistry of the Corrosion Process Samples from Operating Loops Periodic sampling of the melts in forced-circulation corrosion- test loops, as described in the previous report,lo has continued. One loop is fabricated from INOR-8 and the other from Inconel. Both loops were charged with the selt composition LiF-BeF-ThF)-UF), (62-36.5-1;0.5 mole %) containing about 400 ppm of chromium, and the maximum wall temperature is 1300°F. | The chromium concentration in the INOR-8 loop (MSRP No. 12) reached a plateau at about 550 ppm after about 1200 hr of opération and has maintained this level for the last 1000 hr. In the Inconel loop (9377-5) the chromium éoncentration in the salt rose gradusally during the first 400 hr of operation to a value of 550 ppm. The increase was more rapid than in the INOR-8 loop but was uniform. However, after the first 400 hr, the chromium concentration began increasing very rapidly, as indicated below: Chromium Time Concentration (hr) (ppm) 500 800 650 | 950 ’ 800 1100 1700 2000 ) Since sample removals decreased the quantity of salt in the loop, more salt was added after 2250 hr to permit further monitoring. Samples taken before and after the addition of 0.6 kg of fresh salt showed 2100 and 1900 ppm of chromium, respectively. The loop was allowed to circu- late for three days (72 hr) and again sampled, and 2050 ppm of chromium was found; after 2600 hr, 2350 ppm of chromium was present. The latter value may represent an approach to a steady-state level, but the chromium 19MsR Quar. Prog. Rep. Jan. 31, 1959, ORNL-268L4, p 106. ) 86 concentration is consideréblyvhigher than was expected in the absence of oxidizing impurities. Radioactive Tracer Analyses for Iron in Molten Fluoride Salts Most investigations of corrosion behavior depend on accurate analyses at low dilution of Fe't and other structural metal ions. For example, the validity of chromium diffusion experiments in systems consisting of a molten salt and a chromium alloy depends on the purity of the salt with respect to NiF, and FeF,.,. Both compounds form CrF 2 2 2 from chromium; for example, FeF,, + CrCe— CrF, + e . Experimental equilibrium constants show that the equilibrium concen- tration of FeF2 in the presence of chromium should be too low to detect. However, analyses obtained in the diffusion experiments indicated that FeF2 was present at concentrations ranging from 100 to 200 ppm in NeF-ZrF), (53-47 mole %) which had been in contact with Inconel ([Cro]z 0.16 wt fraction). This anomaly suggested that finely divided iron might be passing through the sample stick filters under certain con- ditions. To investigate this possibility, 800 ppm of labeled FeF,. was dissolved in a nickel container filled with NaF-ZrF) (53-47 mole ;) solvent and then completely removed from solution by reduction with zireconium. Standard counting procedures verified that no labeled Fe++ or iron was present in filtered portions of the melt after the zirconium addition, although an average wet analysis of 205 ppm of iron was reported for the same material. It seems definite that currently used procedures result in misleading results from wet chemical analysis and that the presence of iron in the melt as sampled is not the explanation. The discrepancy is being investigated. Activities in Metal Alloys The measurements of thermodynamic activities for nickel in the Ni~NiO system, as obtained from an electrode concentration cell with a molten electrolyte, have been delayed by a series of experiments 87 designed to test the approach to equilibrium obtained by various annealing procedures for the electrodes. These experiments have not given definite results. Another series of experiments has established that the results are not being invalidated by thermoelectric potentials. Vapor Pressures of Molten Salts The lowering of the vapor pressure of CsF (vapor préssure, 83 mm Hg) at 1000°C by 20 mole % additions of alkaline earth fluorides has been measured as part of a study of the effect of cation size and charge on the thermodynamic properties of fluoride melts. The smallest cation Mg++ (radius, 0.783) gives a lowering of 30 mm Hg; ca’ (radius, 1.063), 26 mm Hg; and Ba (radius, 1.433), 18 mm Hg, following the expected order. Freezing-point depressions that reflect the same effects are also being measured. Permeability of Graphite by Molten Fluoride Salts Permeability tests on various types of reactor-grade graphite with molten fluoride salts have continued. In the routine procedure for impregnating graphite with a molten salt, the graphite samples were degassed under vacuum at 900 to 950°C and then treated with a molten salt, such as 1iF~MgF2‘(67.5—32.5 mole %), while still held under vacuum. After the graphite samples were completely covered with salt the vacuum was relieved and a pressure of 15 psig of helium was applied to the system for 48 hr. At the end of 48 hr, the pressure was relieved, and the molten salt was transferred away from the graphite. In previous tests, graphite rods 3-in. long and 1/4, 1/2, and 1 in. in diameter had been used. For routine testing, l-in.-dia rods have been adopted. By using welded container vessels, shortening vacuum lines, and using extreme care in assembly, the efficiency of impregnation has been increased, as indicated by an average weight gain of 12.5% compared with 8.5% in earlier runs on reactor-grade TSF and AGOT graphite samples. 88 The objectives of the current exploratory tests are to determine (1) how effective salt impregnation of graphite is in preventing or decreasing penetration of possible molten-salt reactor fuels at the maximum reactor temperature, (2) the degree of penetration into untreated graphite of typical reactor fuels under forced impregnation (vacuunm and pressure) conditions at the maximum reactor temperature (1250°F), (3) the degree of penetration of typical reactor fuels into untreated graphite under normal reactor operating conditions (in a 1000-hr test in a circulating-salt system), and (4) the distribution of the main fuel components (uranium, thorium, and beryllium) in the graphite when penetration occurs. Samples of three special types of graphite were obtained from the National Carbon Company for testing. These were identified as follows: ATT-82, ATL-82, and AGOT-82. The number 82 after each type refers to a special process used by National Carbon to make impervious graphite. Because of the geometry of the available samples, the rods were machined to 3/h in. OD and 3 in. long. Attempts to impreghate these rods of special grephite with LiF-MgF2 in the usual manner gave the following results. The ATJ-82 graphite showed a loss in weight of 0.9%. The ATL-82 graphite showed a loss in weight of about 0.1%. The AGOT-82 graphite showed a gain of 2.0%, and samples of TSF graphite, included for comparison, gained 5.5%. The rods were then subjected to a 1000-hr soaking test at 12500F under a static helium pressure of 1 psig in stirred LiF-BeF,-UF) (62- 37-1 mole %). The ATJ-82 gained 0.2% in weight, the ATL-82 gained 0.4%, the AGOT-82 gained 2.1%, and the TSF gained 2.8%. These gains are based on the weights of the rods before and after the 1000-hr test and do not include the forced-impregnation gains or losses. To determine the penetration of uranium and beryllium into these rods, successive 1/32-in. layers were machined from the rods and sub- mitted for chemical analysis. When the rods became too small to machine without breaking, the ends were cut off and the center section was submitted for analysis. Typical results are listed in Table 2.2.3. 89 Table 2.2.3. Analyses of Cuts Taken from Graphite Rods Impregnated with LiF—MgF2 and Then Soaked in LiF-Ber-UFu Type of Graphite TSF AGOT ATL-32 ' ATT-82 e i Be T Be U Be U B ) (ppm) (ppm) (ppm) (ppm) (ppm) (ppm) (ppm) (ppm) 1 3500 6400 2200 4300 1000 1800 600 1000 2 3300 6400 1500 3800 700 1200 150 270 L 3100 6300 1300 3300 550 900 25 85 8 3500 7000 940 2900 Ls0 600 10 100 Center 3700 5200 1200 1400 1900 450 9200 350 The cut numbers were assigned so that cut No. 1 was the first cut taken on the rod, cut No. 10 was the last cut taken, and "center' indicates the remaining center section, which was also submitted for analysis. It appears that the ATL-82 and ATJ-82 grapites are resistant to forced impregnation with LiF-MgF2 salt and are considerably resistant to penetration by a typical reactor fuel. The most startling observa- tion, however, is the unexpectedly high concentration of uranium in the center section of the ATL-82 and ATJ-82 rods. Subsequent long-term soaking tests of reactor-grade graphites TSF and AGOT in a fuel mixture of the same IiF-BngeUFu composition also showed disproportionately high urenium concentrations in the center in every case. Ifforts are being made to determine the mechanism of this phenomenon and to further verify it. An experiment was carried out to determine whether untreated TSF graphite could be forcibly impregnated with the LiF-BeFE-UFh (63-37-1 mole %) fuel mixture of current interest by using vacuum and pressure techniques at the maximum reactor temperature of 125OOF. Six TSF graphite rods were used in this experiment, and weight gains that ranged from 0.1% to 1.85% were obtained. To determine depth of penetration, these rods were machined, as previously described, and 90 the machine cuttings plus the center section were submitted for analysis. In all cases, the bulk of the salt penetration was in the first and second machine cuttings. HoweVer, it was noted again.that, in general, the center sections contaified several times (occasionally 200 times) more uranium than the last cutting taken at about 1/4 in. from the center. Radiastion Damage Studies INOR-8 Thermal Convection Loop for Operation in the LITR The in-pile thermal-convection loop for testing fused-salt fuel in INOR-8 tubing in the LITR was operated in preliminary tests outside the reactor, and satisfactory circulation of the salt (IiF-BeFE-UFu, 62-37-1 mole %) was obtained. Radiography of the fuel tube after these tests demonstrated that no opaque material had deposited at the bottom, as in a prev1ous loop test. Ll Thermocouples have been installed on the fuel tube of this loop, and the air annulus tube has been assembled around the loop. Detailed examinations of the thermocouples were made, and complete records of their condition, inecluding individual phbtomicrographs of each thermo- couple, are available. This examination record will be used to correlate thermocouple performance during operation in the reactor with the initial conditions of the thermocouples. The information obtained will be useful in the construction of Ffuture in-pile loops. Modifications were made in the cooling-air control system at the reactor to prbvide for indiVidual temperature control of various portions of the loop. After resistance-heating elements have been installed and the loop inserted in the outer can, the loop will be operafied in the LITR. In-Pile Static Corrosion.Tests Two fuel-filled INOR-8.capsules, which were described previously,ll were installed in the MIR and are being irradiated at a temperature of MMSR quar. Prog. Rep. Jan. 31, 1959, ORNL-268k, p 112. 91 1250°F. The fuel is LiF-BeF ,-UF), (62-37-1 mole %), and the power density in the fuel is 1200 w/cm3. Two additional capsules of the same type are being prepared for irradiation in the ORR. Preparation of Purified Materials Purification, Transfer, and Service Operations The processing of small batches of various molten-fluoride-salt compositions for use in corrosion tests, physical property studies, and small-scale component testing decreased considerably during the first half of this quarter. However, demands have gradually increased during the past month, and a normal rate of operation is now being maintained. A total of 35 kg of mixtures not containing beryllium and 85 kg of beryllium-containing mixtures was processed during the quarter. Transfer and service operations to assist engineering groups in handling high-temperature fluids were continued at a slightly increased rate. Fuel Replenishment Tests A simple device for testing the proposed fuel sampling and enriching mechanism;g was constructed and tested. The device employs two salt storage pots interconnected by a transfer line, with the enriching device on one pot and the sampling device on the other pot. The molten salt is moved alternately from one pot to the other with - semiautomatic controls by using differential gas pressures. - Solid, cast, UFM fuel slugs weighing 40 g each were introduced to the dissolution pot in a copper basket and the salt, LiF-BeF, (50-50 mole %), was cycled between the two pots. The rate of dissolution was measured by withdrawing and weighing the amount of UFh remaining; samples of the solution were also taken for chemical analysis. A 4O-g slug of UFh dissolved in 20 cycles (1 1/2 hr), but it was not completely dissolved in 15 cycles (1 hr). This rate of solution 12MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-268L, p Q. 92 appears to be adequate for convenient enrichment procedures. Pure Compounds Prepared with Molten Ammonium Bifluoride Molten emmonium bifluoride (mp, 125°C) has served as a reactant for preparing both 51mple and complex fluorides. 13,1k Recent trials show that commerc1al chromlc oxide is not noticeably attacked by molten ammonium bifluoride. Magne51um.ox1de is converted into NHhMgF3, which can be decomposed 1nto pure magne51um fluoride. Beryllium oxide (calc1ned at 900 C) is also converted to an ammonium complex. ~ When commercial "Co 03” powder is treated for 1 hr with molten ammonium bifluoride, a small amount of oxide is left unchanged. The reactibn product has an x-ray diffraction pattern Somewhat shifted from that of KCOF3, so it is presumed to be NHhCoF3 Nickel oxide was found to be more resistant to attack than ”00203 A l-hr treatment left almost all the oxide unreacted. The oxides Ge02, UO,, ThO,, and CeO, seem to react completely with molten ammonium bifluoride, but the products have not yet been identified. Vanadium pentoxide reacts with molten ammonium bifluoride to evolve yellow fumes. Presumably, Vanadium>pentafluoride is formed; it is reported to react with moist air to form;yellow oxyfluorides.l5 | For all the reactions with oxides, at least a 100% excess of ammonium bifluoride was used in order to form a mixture which was sufficiently fluid to stir. | | Electrolytic iron powder reacted with molten ammonium bifluoride to form ammonium hexafluoferrste (III). Chromium metal reacted to forfi ammonium hexafluochromate (III), but the reaction was slow. Granulated metal, finer than 100 mesh, was treated with the melt for 70 min, and the unreacted ammonium bifluoride was volatilized. Extraction of this 138, J. Stuwrm end C. W. Sheridan, Preparation of Vanadium Tri- fluoride by the Thermsl Decomposition of Ammonium Hexafluovanadate (III), ORNL CF 56-5-95 (May 28, 1953). My . Sturm, MSR Quar. Prog. Rep. Oct. 31, 1958, ORNL-2626, p 107. 0. Ruff and H. Lickfett, Ber. deut. chem. Ges. ki, 2539 (1911). 93 product with aqueous nitric acid to remove the ammonium hexafluochromate (III) showed that 60% of the chromium metal had not reacted with the ammonium bifluoride. Reaction of Chromous Fluoride with Stannous Fluoride The reaction of chromium metal with molten stannous fluoride pro- vides a useful method of preparing pure chromous Jf‘luozc'n‘.d.e;llF however, an excess of stannous fluoride can further oxidize the chromium. The predominant phase resulting from the reaction of equal molar portions of stannous fluoride and chromous fluoride is a compound thought to be Cr3(CrF6)2.l6 The product obtained by fusing 1 mole of chromous fluoride with 2 moles of stannous fluoride appears to be stable at the boiling point of stannous fluoride (~700°C). At 1000°C, all the stannous fluoride volatilizes and leaves a residue of chromic fluoride. The probable intermediate, Sn3(CrF6)2, has not been definitely identified, but the reaction is presumably: 00°C 3 uSnF2 + 20rF2 ~—-———>Sn3(CrF6)2 + Sn 1000°C SnS(CrF6)2 — 2CrF, + 35nF, 1 1°MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-268k, p 113. ok 2.3. TFUEL PROCESSING Processing of molten-fluoride-salt reactor fuel by volatilization of the uranium as UF6 appears to be feasible. The barren LiF-BeF2 can be recovered for reuse by treatment with nearly anhydrous HF, in which it is appreciably soluble, and in which rare earth neutron poisons and most other polyvalent element fluorides are insoluble. Development work on the process has continued with a study of the behavior of neptunium, vhich is possibly the most important neutron poison other than the rare earths.l Further measurements of the solubility of Np(IV) have indicated that its solubility (Table 2.3.1) in aqueous HF solutions saturated with LiF-BeF-ThF) -UF) (61.5-37-1-0.5 mole %) will be considerably less than reported previously. Apparently the Np(IV) solubility is decreased by the presence of excess ThFu and UFM' The addition of nickel and iron metals to solutions saturated with LiF, Ber, and Nthralso caused a significant reduction in the neptunium solubility, except in anhydrous HF. This is probably the result of reduction of Np(IV) to Np(III). In a reactor processing system the solution will contain much larger amounts of rare earths, plutonium, uranium, and thorium than of nep- tunium, and the container will be a metal, probably a nickel alloy. Therefore & neptunium solubility of the order of 0.0002 to 0.00005 mole % is expected in actual processing. 1MSR Quar. Prog. Rep. Jan. 31, 1959, ORNL-268k, p 115. 95 96 Table 2.3.1 Solubility of Np(IV) in Aqueous HF Np(IV) Solubility™ Np(IV) Solubility in HF Saturated HE in HF Saturated . . Concentration with LiF-BeFo with LiF-BeF,-ThF)-UF) (%) mg/g of mole % relative Fe and Ni Added solution to salt mg/g of mole % relative mg/g of mole % relative solution to salt solution to salt 80 0.026 0.0031 0.0013 0.00012 0.0043 0.00052 90 0.011 0.0012 0.00046 0.00005 0.001k 0.00016 95 0.0086 0.00072 0.00054 0.00005 0.0029 0.00025 100 0.0029 0.00024 0.0047 0.00039 0.0029 0.00025 ®Values previously reported in ref. 1 for LiF-BeF2 salt in absence of UFM and ThFu. » * L L] * L b E=EQsQrE-HdugEBHRD gt -y - . * > L] . * * ] L ] * GEPpUONGEIHOSTE S0 00 INTERNAL DISTRIBUTION G. Affel G. Alexander S. Bettis Billington Blankenship Blizard Boch Borkowski Boudreau Boyd Bredig Breeding Briggs Browning Campbell Carr Cathers Center (K-25) Charpie Coobs Culler . DeVan . Douglas Emlet (K-25) Ergen Estabrook Ferguson Fraas A. Franco-Ferreira H. Frye, Jr. R. Gall T. Gresky L L * fi?fi?dbmg"fflbbjfi.momsflQ?m‘#C{t‘*Uuflm o . L. Gregg R. Grimes Guth S. Harrill R. Hill W. Hoffman Hollaender S. Householder H. Jordan W. Keilholtz P. Keim L6, 47, L8, 4o, 50. 51. 52. 53. 54, 55. 56. L] . il 3 el = o JRSS =] . c—({?jf;dtrfc—{ =00 G. * L] - . L} - - L 4 . * L - '#U':UM‘PC-IOMP:E.’ME{QPMZQ*U ORNL=-2723 Reactors — Power TID-4500 (14th ed.) : Livingston MacPherson Manly Mann Mann McDonald McDuffie McNally Metz Milford Miller Miller Morgan Murray (Y-12) Nelson Nessle . Osborn Patriarca M. Perry Phillips M. Reyling T. Roberts T. Robinson W. Savage W. Savolainen L. Scott E. Seagren D J * e e ] L] » * * L 2 munNzowa?wwpwumm>m . Shipley . Skinner H. Snell Storto D. Susano A, Swartout Teboada H. Taylor E. Thomsa B. Trauger C. VonderLage o7 89. G. M. Watson 90, A. M. Weinberg 91. M. E. Whatley 92. J. C. White 93. G. D. Whitman ok, G. C. Williams : g5. C. E. Winters 96. J. Zasler 97-100. ORNL - Y-12 Technical Library, Document Reference Section 101-140. Laboratory Records Department 141. Laboratory Records, ORNL R.C. 1so-1kly, Central Research Library EXTERNAL DISTRIBUTION 145. D. H. Groelsema, AEC, Washington 146. Division of Research and Development, AEC, ORO 147-734. Given distribution as shown in TID-4500 (1kth ed.) under Reactors-Power category (75 copies - OTS) | 98 Reports previously issued in this series are as follows: ORNL-2373 ORNL~2431 ORNL~ 2L T4 ORNL-2551 ORNL-2626 ORNI-2684 Period Ending September 1, 1957 Period Ending October 31, 1957 Period Ending January 31, 1958 Period Ending June 30, 1958 Period Ending October 31, 1958 Period Ending January 31, 1959 99