2H > ORNL-4829 INTERGRANULAR CRACKING OF INOR-8 IN THE MSRE H. E. McCoy B. McNabb THIS DOCUMENT CONFIRMED UNCLASSIFIED AS DIVISION OF CLASSIFICATION Printed in the United States of America. Available from National Technical Information Service U.S. Department of Commerce - 5285 Port Royal Road, Springfield, Virginia 22151 Price: Printed Copy $3.00; Microfiche $0.95 This report was prepared as an account of work sponsored by the United States Government, Neither the United States nor the United States Atomic Energy Commission, nor any of their employees, nor any of their contractors, subcontractors, or their employees, makes any warranty, express or implied, or assumes any legal liability or responsibility for the accuracy, completeness or usefulness of any information, apparatus, product or process disclosed, or represents that its use would not infringe privately owned rights. oY «) ORNL-4829 . UC-25 — Metals, Ceramics, and Materials | Contract No. W-7405-eng-26 METALS AND CERAMICS DIVISION INTERGRANULAR CRACKING OF INOR-8 IN THE MSRE H. E. McCoy and B. McNabb NOTICE i This report contains information of a preliminary | nature and was prepared primarily for internal use at the originating installation, It is subject to re- vislon or correction and therefore does not repre- sent a final report. It is passed to the recipient in confidence and should not be abstracted or further ! disclosed without the approval of the originating installation or USAEC Technical Information Center, Oak Ridge, T'N 37830 NOVEMBER 1972 OAK RIDGE NATIONAL LABORATORY - 0Oak Ridge, Tennessee 37830 : operated by | UNION CARBIDE_CORPORATION : for the U.S. ATOMIC ENERGY CQMMISSION o\ o) -LJ]. - CONTENTS ABSTRACT « & & v v o v v o v o v v o v o o v et o a a s INTRODUCTION « « « o v 4 o v v e e e e e e e e e e n e THE MSRE AND ITS OPERATION . . . . « &« v & ¢ o & o o | Description . . . « & ¢ ¢« v ¢ ¢ vt 4 h bt e e e e History . . e 4. s e o s e s e s e e & 4 s = e EXAMINATION OF MSRE SURVEILLANCE SPECIMENS . . . . . . . Specimens Exposed Before Power Operation . C e e First Group of Surveillance Specimens . . . . . . . Second Group of Surveillance Specimens . . . . . . Third Group of Surveillance Specimens . . . . . . . Fourth Group of Surveillance Specimens . . . . . . Specimens Exposed to Cell Atmosphere . . . « o Studies Related to Modified Surface Mlcrostructure Summary of Observations on Surveillance Specimens . EXAMINATION OF MSRE COMPONENTS . . .+ + & &« ¢ ¢ o + &« o & Control Rod Thimble . . . & & &« &+ ¢ & &« ¢ o s o o« & Primary Heat Exchanger . . . . . ¢« ¢« ¢« ¢« ¢« &+ « + & Pump Bowl Parts . . « o ¢« v v o o o i o o o o o » Freeze Valve 105 . . &. & ¢ ¢ ¢ ¢ ¢ ¢ ¢ o o o o o EXAMINATION OF INOR-8 FROM IN-REACTOR LOOPS . . . . . . Pum ’ Loops * . . L 2 . . a . » . L] - . * » * . . . - Thermal Convection Loop . + « « ¢« ¢ ¢ + o o « o & Summary of Observations on In-Reactor Loops‘; .« o e CHEMICAL ANALYSES OF METAL REMDVED FROM THE MSRE . . . . DISCUSSION OF OBSERVATIONS ON INORrB FROM THE MSRE AND IN-REACTOR LOOPS . . « & & ¢ v & ¢ o o & o & ~ Summary of Observations . . « « ¢ « « ¢ o 4 o o o @ Possible Mechanisms . . + & ¢« v & ¢ ¢« o « o o o « & POST-MSRE STUDIES . o - . . . - . | . e e o o . @ .o . s s s Corrosion Experiments . . v v v« o « o o ¢ o o & & Difquion Of Te e s s e e s e . u V s & s s e s e Experiments with an Applied Stress . . . . . . . . SUMRY . - e & . ® . - - - . . ‘... - * . . - * » * * . V. Acm OWLEDGMENT .' l ‘ . - . . » . . * . - . . . ) . - - - - * REFERENCES * . » - . * L) - * . . - - * - - . - ® . . iii e M~ N = 14 17 19 - 27 34 50 79 79 83 87 88 106 118 137 138 143 145 149 149 157 157 160 165 165 169 169 170 173 174 o) u} -3’ INTERGRANULAR CRACKING OF INOR-8 IN THE MSRE H. E. McCoy and B. McNabb ABSTRACT The INOR-8 surveillance specimens and components from the MSRE that had been exposed to fuel salt formed shallow intergranular cracks (2 to 10 mils deep in exposures to greater than 20,000 hr). Some of these cracks were visible in fioliéhed-séctions of as-removed materials, but many others were visible after the samples had been deformed. Consider- able evidence indicates that the cracks were due to6 the inward diffusion - of fission products. ‘The fission product cracking mechanism was further substantiated by laboratory tests which.clearly demonstrated that teliurium causes intergranular cracking in INOR-8. These tests have included other méterials, and important variations exist in their respective suscepti- bilities to cracking by tellurium. Several materials, including types 300 and 400 stainless steels,.nickel-_and cobalfi-base alloys containing greater than 15% Cr, copper, mohel; and INOR-8 containing 2% Nb, com- pletely resisted cracking in the tests run thus far. - ] BTy = - INTRODUCTION The MbltenFSalt Réactor Experiment was a unique fluid-fuel reactor.! It operated at temperatures around 650°C for more than 20,000 hr between 1965 and-Décember, 1969. The fuel was a mixture of fluoride salts, cir- culated through a core cf-graphite bars and an external heat exchanger. Except for the graphite, all parts contacting the salt were of a nickel- base alloy known as?INOR—B-and now available commercially under the trade names of Hastelloy N and Allvac N. This alloy, developed at 0Oak Ridge NationalwLaboratory?Spécifically for use in fluoride salts at high temperature,?.has the nominal composition of Ni-16% Mo-7% Cr-5% Fe-0.05% C. . | | . - ' ' INOR-8 in the MSRE behavéd as expected with regard to corrosion by the fluoride salts and the containmentfatmosphere‘(very?little.byfeither), Two prdblems with INOR-8 did appear, however. The first was a drastic reduction in high-temperature creep-rupture life and fracture éttain under creep conditions. The second was the appearance of grain-boundary cracks at iNOR-B,surfaces;eprsed to the fuel salt. | o Embrittlement phenomena have been studied extensively for the past several years in connection with various iron- and nickel-basé alloyss‘ and specifically with regard to INOR-8 by.ORNL. The-embrittlement of . ‘INOR-8 in the MSRE has been attributed mainly to. the helium that is genefated by thermal neutron interaction with 198 present in the alloy as an impurity. We have found that small changes in chemical composition are quite effective in reducing the effects of helium production, and our fiork is well along toward developing a modified INOR-8 with improved resistance to embrittlement by neutron irradiation. This work has been reportéd extens:lvely'*'11 and will be discussed in this document only insofar as it relates to the finding and interpretation of evidence on . the surface cracking problem. - The cause of the surface cracking has not yet been précisely defined, nor can its very long-term behavior be predicted with confidence. The ~ cause must be associafed with fuel system'conditions after the beginning of power operation: numerous cracks or incipient cracks were found on 0 <) ») every INOR-8 surface that was examined after prolonged contact with the -radioactive fuel salt; few or none could be found on INOR-8 surféces (sometimes on the same piece) that had beeh—exposed to the fluoride salt in the coolant system or. to the containment cell atmosphere. The cracks were observed to open up when affected surfaces were strained in tension, but some grain boundary cracks were detectable in polished éections of unstrained specimens. The depth of cracking was 2 to 10 mils, and some sectioned specimens showed as many as 300 cracks per inch of edge. A general trend to more and deeper cracks with increasing exposufe time was evident, but the statistical significénce of the changesAafter the first several thousand hours was poor. Obviously the determination of cause and long-term progression is essential in the development of molten- salt reactors that must operate reliably for many years. | ~ - An intensive effort has been.mofinted within the MoltenQSalt Reactor- Program to investigate the INOR-8 cracking phenomenon and to develop remedies or ways to circumvent the problem. More or less similar effects can be produced in out-of-reactor experiments. Fluoride salts fiithvadded FeF, oxidant cause inter-granular corrosion. Tellurium deposited on INOR-8 and allowed to diffuse at high temperature produces brittle grain " boundaries. Tellurium also affects nickel and stainless steel, but to leséer degrees. However,‘some of the other agents that were suspected of causing the cracking:in the MSRE fuel system have given negative results. This work is currently in progress, and few firm conclusions can yet be.drawn. _ ' - The present document has_%een_prepared to-reportfiand_Summarize all the currently known informatidn obtaified}from.the MSRE, which is the starting point of the investigation.: It also includes pertinent observa- 'tions‘from-other molten-salt systems, brief accounts of current experi- ments, and some tentative conclusions. THE MSRE AND ITS OPERATION | A good, general description of the;MSRE and an account of most of’ its history appear in reference 1. . Reference 12 is a detailed description -of all components and systems. Because these references are widely avail- able, the description ‘here is confined to those portions that are germane to the discussion of the INOR-8 cracking.: 'DescriEtion 'The parts of the MSRE with which we will be concérned are included in the simplified'fiowsheet in Fig. 1. The cracking phenomenon was ob- served on pieces of INOR-8 from the reactor vessel, the heat exchanger, the fuel'pump and a freeze valve near fuel drain tank No. 2. INOR-8 specimens exposed to the containment cell atmosphere outsidé the reactor ‘vessel and surfaces exposed to the coolant salt were examined but did not'show‘the—cracking. The fuel salt composition was LiF-BeF,-ZrF,-UF, (65-30-5-<1 mole %); the -coolant, LiF-BeF, (66-34 mole 7). At full power the 1200-gpm fuel stream normally entered the reactor vessel at 632°C and left at 654°C the maximum outlet temperature at whicthhe reactor operated for any substantial period of ‘time was 663°C (1225°F). When the reactor was at low oower, the salt systems were usually nearly isothermal at about 650°C During extended shutdowns the salt was drained into tanks, where it was kept molten while the circulating loops were allowed to cool. Plugs of salt frozen in flattened sections of pipe ("freeze valves") were used to isolate the drain tanks from the loop. TThe'liquidusftemperature of the fuel salt was about 440°C and that of the coolant salt was 459°C, 'so the | loops were heated to. 600-650°C with external electric heaters before the salt was transferred from the storage tanks. Helium (sometimes argon) was the cover gas over the fuel and coolant salts. During operation, samples of fuel salt were obtained by lowering small copper buckets (capsules) into the pool of salt in the pump bowl. *) ) ' ' K ) . - ' ORNL-DWG 63-15440R —| o 5 r.:siq : 5 psig FUEL - COOLANT ‘ e PUMP $ SAMPLER- . PP V. NsAuPLER LEGEND i )JENRICHER . : sm—— FUEL SALT . ! - : ST T ’ ' ‘ S COOLANT SALT - 1 ! ¢ TO ABSOLUTE FILTERS:ec:eert sveesisessses HELIM COVER GAS i 3 ' 1015 °F i ' ‘ wewemm = RADIDACTIVE OFF -GAS 1 S Lo : 1 850 GPM. ' ' OH':‘-_';‘:E : ’ HEAT EXCHANGER - {210 °F ‘ N OVERFLOW TANK ABSOLUTE 7O *F o AR FLOW: 200,000 cim FILTERS : 1200 G.PM. v ey . - BLDG. REACTOR : : : . . 1075 °F . . B ENTILATION . VESSEL | powER FREEZE FLANGE (TYP) —— — . . 8 Mw y STACK FAN ) . L III - 1 . N A FRoM 1 FREEZE VALVE (TYR) ' : : =i COOLANT } : ’ RADIATOR ‘ ;‘ SYSTEM *_l / _ $ --——-i-r.—--l - - -l - : { oo - 0 S i 1 H, M I WATER STEAM 4 _ . FILTERS . WATER STEAM i ; . . i % Xk 1 MAIN i I CHARCOAL i H K - BED OOOLANT ] | DRAMN TANK \/ e e e e e e b Fig. 1. Design Flow Sheet of the MSRE. The pump bowl served as the surge space.for'the loop and also for sepa- ration'Of‘gaseous fission products from a 50—gpm'3tfeam of salt sprayed out intoithe.gés sbace above the salt pbol. To protect the sample Bucket from the salt spray in the pump bowl, a spiral baffle of'INORrS extended from the top of the bowl down into the salt podl. A cage of INOR-8 rods inside ‘the spiral bafflé guided the sample‘capsule in the pump bowl. The reactor éore was composed of vertical graphite bars with flow paésages bétween them.r In the lattice near the center of the core three thimbles of INOR-8 housed control rods. Nearby,‘and‘accéssible during shutdowns through a flanged nozzle, was an array of graphite and metal specimens, which was exposéd fo the fuel flowing up through fhe core. Throughout most of the MSRE operation the core array‘was as shown in Fig. 2. This array was designed to expose surveillance specimens of grafihite and INOR-8 identical to thé material used in the MSRE core and reactor vessel. 'Later, specimens of modified INOR-8 were included. The assembly was composed of three separable stringers designated RL, RR, and RS. Each stringer included a column of graphite'speciméns and two rods of INOR-8. (Strifiger RS also included a flux monitor tube.) ~ Not shown in Fig. 1, bu; located in the reactof building, was a vessel in which specimen stringers. identical to those in the core could be exposed to fluoride salt having the same nominal composition as the fuel salt. (The stringers in the control facility wvere designated CL, CR, and CS.) Electric heaters on the vessél were controlled to produce a temperature profile along the stringers like the profile in the core. The salt in this control facility did not circulate. - | The fuel system was contained in a cell'in‘which an atmosphere of nitrogen COntéining from 2 to 52 O was maintained. This contéinment atmosphere was recirculated through a system that provided cooling for the control rods and the freeze valves. Arrays of INOR-8 specimens were exposed to the cell atmosphere as shown in Fig. 3. Suspended just ofitside the reactor vessel but inside the vessel furnace,'the specimens were exposed to practically the same neutron flux and temperature as the vessél walls. L1} Fig. 2. ) . i; 7 e N 7 MSRE Surveillance Facility Inside Reactor Vessel. N ") ORNL-DWG 68-8298 THERMAL SHIELD REACTOR VESSEL SURVEILLANCE STRINGER " FLOW DISTRIBUTOR 81in TOP OF LATTICE 1t3in. ELEVATION 828ft | 2% in 1%5-in. LONG NOSE PIECE o 8 16 Lesgdoa] INCHES Fig. 3. MSRE Surveillance Facility Outside the Reactor Vessel. & ¥) s) ol Histogx The history of the MSRE during the four years in which it operated at significant power'isroutlified in Fig. 4. Construction had been fin- ished and salt charged into the tanks late in 1964.£ Prenuciear testing, including 1100 hr of salt circulation, occupied January-May, 1965. During.nuclear startup_experiments'in May—July,.l965, fuel salt was cir- culated for 800 hr. The salt was drainéd and final-preparafions for power operations were made in the fall of 1965. Low-power experiments in December led into the history covered in Fig. 4. (See ref. 1 and MSRP semiannual progress reports for more detéilh) About a year aftér the conclusion of operation, a limited program of examination was carried out. This included INOR-8 pieces from a contfol'fo& thimble, the heat exchanger shell and tubeé, the pump bowl cage and baffle, and a freeze valve. o | | The nuclear fuél was 33%-enriched 235U,'and the UFy concentration in the fuel salt was 0.8 mole % until 1968. Then the uranium was removed by fluorination and 233UFu-waé substituted. The UF, concentration re- quired with 233U was only‘0.13'mole %. The composition of the fuel salt was observed by frequent sampling from the pump bowl.l3 Aside from the 2.33U loading and periodic additions of small increments of uranium or plutonium to sustain the nuclear reactivity, the only other additions to the fuel salt were more or less routine small (v10 g) quantities of beryllium, and, in two of three experiments, a few grams of zirconium and FeF,. The purpose of these additions was to adjuét the U(III)/U(IV) ratio, which affects thé_éorfésion potential and the oxidation state of corrosion~product irqn'and'nickel and fission—prbduct niobium. 7 The primary corrésion mechanism in the fuel salt syétem_fias selective removal of chromium by 2UF, + Cr(in aliby) = 2UF3 + CrF,(in salt) , and the concentration of chromium in salt samples was the primary indi- cator of corrosion. Figure 5 shows chromium concentrations observed in } DYNAMICS TESTS INVESTIGATE . OFFGAS PLUGGING REPLACE VALVES AND FILTERS RAISE POWER ‘REPAR SAMPLER _ ATTAN FULL POWER CHECK CONTAINMENT FULL - POWER RUN ~— MAIN BLOWER FALURE REPLACE MAIN BLOWER MELT SALT FROM GAS LINES REPLACE CORE SAMPLES TEST CONTAINMENT - ~ RUN WITH ONE BLOWER > WNSTALL SECOND BLOWER ROD OUT OFFGAS LINE CHECK CONTAINMENT 30-day RUN AT FULL POWER REPLACE AR LINE DISCONNECTS SUSTAINED OPERATION AT HIGH POWER REPLACE CORE SAMPLES TEST CONTAINMENT } REPAIR SAMPLER 02 46 8% FoeL SSES POWER (Mw) FLusH T} 'Fig. 4. Outline of the Four Years of MSRE Powver 10 SALT N FUEL LOOP POWER 0 2 4 6 8 © el S 0 POWER (Mw) Fuusd ] v ORNL-DWG 69— T293R2 XENON STRIPPING EXPERIMENTS MAINTENANCE 1 } INSPECTION AND REPLACE CORE SAMPLES TEST AND MODIFY FLUORINE DISPOSAL SYSTEM } PROCESS FLUSH SALT - PROCESS FUEL SALT f} LOAD URANIM-233 - T REMOVE LOADING DEVKE . 23y 7tRO-POWER PHYSICS EXPERIMENTS } INVESTIGATE FUEL SALT BEHAVIOR } cLear oFFeas LmEes’ CONTROL ROD DRIVE 253, DYNAMICS TESTS INVESTIGATE GAS N FUEL LOOP ] REPAR SAMPLER AND HIGH-POWER OPERATION TO MEASURE 23y o /o, REPLACE OORE SAMPLES INVESTIGATE COVER GAS, XENON, AND FISSION PRODUCT BEHAVIOR ADD PLUTONIUM IRRADIATE ENCAPSULATED U MAP F.P. DEPOSITION WITH GAMMA SPECTROMETER MEASURE TRITHM, SAMPLE FUEL REMOVE CORE ARRAY PUT REACTOR IN STANDBY Operation. 4 ) u 11 the MSRE fuel over the years of power operation. The step down in chro- mium concentration in the salt in 1968 was effected by pfocessing the salt in 1968 was effected by processing the salt after the 235U fluori- nation. The total increase in chromium in the 4700-kg charge of fuel salt is equivalent to leaching all of the chromium from the 852 ft? of INOR-8 exposed to fuel salt'to|a depth of about 0.4 mil. Throughout the operation of the MSRE a sample array of one kind or another was present in the core. The arrays that were exposed between September 1965 and June 1969 were of the design shown in“Fig.IZ. From the time of construction until August 1965 the‘specimen array in the core contained similar amounts of graphite and INOR—S (to have the same nuclear reactivity effect) but differed in internal configuratlon. During “the last five months of operatlon, an array designed to study the effects of salt veloc1ty on f1531on‘product depositionl" was exposed in the core. Whenever a core specimen assembly of the type shown in Fig. 2 was removed from the core, it was taken to a hot cell, the stringers were taken out of the basket, and a new assembly was prepared, usually in- cluding one or two of the previously exposed stringers. Sometimes the old basket was reused,,sometimes not. The history of exposure of INORrB specimens in this core facility is outlined in Fig. 6. The numbers in- dicate the heats of INOR-8 from which the rods in each stringer were made. Heats 5065, 5085, and 5081 were heats of standard INOR-8 used in fabri- cation of the MSRE.* The other heats were of modified composition designed to improve the resistance to neutron embrittlement. Specimens exposed outs1de the reactor vessel were made of three of the heats that were also exposed in saltt Specimens of heats 5065 and 5085 were exposed from August 1965 to June 1967 and from August 1965 to May 1968; specimens ‘of modified heat 67—504 were exposed from June 1967 to June 1969. | ' | Table 1 lists the chemical compositions of the heats of standard INOR—B that were used in the surveillance specimens and in varlous items ‘that were examined after exposure to the fuel salt. The compositions of *The reactor vessel sides were of 5085; the heads, of 5065. Heat was used for some of the vessel internals. 150 140 130 120 1o 100 I 90 - CHROMIUM (ppm) - o O 40 30 |- RUN FLUSH & D|J FMAMUJJASOND|JFMAM ’ 1966 ’ : 1967 12 0.204 0.05 0.154 Q.10+ L0.35 -0.30 L0.25 -0.20 ol 4-14 15-20 z JJASOND([JFMAM 1968 ORNL-DWG 70-2164 JJASOND|JFMAMUJY JASOND . 1969 Fig. 5. Corrosion of the MSRE Fuel Circuit in 235y and 233y Power Operations. ORNL-DWG T2-7495 - ' 'Fig. 6. Outline of MSRE Core Surveillance Program. STRINGER L- ggf :gg: i | | | 5081 5065 7320 | | 1 ] 5085 21545 67-502 67-554 TRI . STRINGER S 5081 21554 67-504 7320 - ) ] CORE TEMPERATURE r—l‘ - ‘ SALT IN CORE 1. L m000on f:fl | I ] I 11 10 RUN NO. { 2 3 456 7 B9 12 43 15 16 1718 19 20 WA W WA Ylidd 00 W A N R A R YA A 1965 1969 i wh " Table 1. Heats of Standard INOR-8 Examined After Exposure in MSRE Specimens . Content (wt %) ) Heat Exposed Mo Cr Fe Mn c . si .S P - Cu Co - Al v Ti W B 50852 Rods on L1, RL,' 16.7 7.3 3.5 0.67. 0.052 ~ 0.58 - 0,004 0.0043 0.0l 0.15 0.02 0.20 <0.01 ' 0.07 0.0038 L1, L2, R2, ‘ - 2 E - ‘ - o - e - ia, X2 : : o : S - - 5081® Rods on L1, R1, 16.0 7.1 . 3.5 0.65 0.059 0.52 0.002 0.012 - 0.0l 0.10 0.05 0.20 <0.01 0.07 0.0040 st - - ; o ‘ - | ' : : 5065 Rods on L2, R2. 16.5 7.3 3.9 0.55 0.065 0.60 0.007 0.004 ©0.01 . 0,08 - 0.01 0.22 0,01 = 0.04 0.0024 X1, X2 S _ el | _ AR 5055 Straps on R2, 16.1 7.5 3.8 0.43 0.07 0.64 . 0.008 0.003 0.03 ~ 0.11 0.08 0.24 0.02 0.26 0.001 ~ L2,.R3, S4 - o . . ) 5075 - Foil on R3, S4 16.4 6.64 4.0 0.46 0.07 - 0.58 0.006 0.003 0.01 0.06 0,02 0.26 .0.02 0.09 . 0.001 and: Sampler . o ‘ L . . - mist shield | _ 5059 Sampler cage . 16.9 6.62. 3.9 0.35 0.07 0.59 0.003 0.001- 0.07 0.01 0.21 0.01 - 0.04 -Y-8487 Rod thimble 6.8 7.3 4.1 0.3 0.05 0.17 - 0.0075 0.004 0.03 0.1 0.16 - 0.25 ©0.007 5060 Rod thimble 16.4 7.05 3.9 0.45 0.06 0.52 0.006 0.001 0.01 0.07 0.01 0.28 0.01 0.005 sleeve ‘ : - . - ‘ ~ . : : ) 5068 Heat exchanger 16.5 6.45 4.0 0.45 0.05 - 0.58 0.008 0.03 0.02 0.1 .0.01 - 0.27. 0.01 . shell . | _ . o : : N2-5105 Heat exchanger 16.4 6.9 3.9 0.45. 0.06 0.60 0.009 0.001 0.01 0.1 0.01 0.33 0.01 0.06 . 0.006 tubes ‘ . : . : 5094 - Freeze valve 105 16.3 7.1 3.8 0.52 0.07 0.76 0.007 0.001 0.01 0.08 0.02 0.39 0.05 0.004 8Less than 0.002% Zt b €Al plus Ti “Less than 0.1% Zr or Hf €T 14 the modified heats that were exposed and then examined are listed in Table 2. These alloYs, besides including-additions of Ti, Hf, W' or Zr, had substantially less Mo, only a fraction as much Fe, and 1ess Mn, V, and Si than did the standard alloys. _ In the correlation of the effects of exposure for different perlods of operation, some common index of exposure would be useful Possible indices are (1) the time that the specimens were exposed to salt, (2)‘the total generation of nuclear heat (and fission products) during the time the specimens were exposed, (3) the total tifie_that.the specimens were at high temperature, and‘(4) the time at high temperature after exposure to salt containing fresn fission products. As will be discussedrlater, some rationale exists for using each of these, so all are included in - the summary of exposures given in Table 3. ) EXAMINATION OF MSRE SURVEILLANCE SPECIMENS INOR-8 specimens removed from the core and from the control facility were subjected to a variety of examinations and tests. First was a visual inspection for evidence of any deposition or corrosion,'particularly any nonuniform or localized corrosion; Metallograpnic examination of selected specimens was used to give more information ‘on the compatibility: of INOR-8 with the salt environment. The major part of this effort was mechanical property testing in connection with the studies on neutron embrittlement. Creep tests to deternine creep rates, rupture life, and rupture strain were conducted at 650°C and stress levels from 17,000 to 55,000 psi. Tensile tests giving yield stress, ultimate stress, and fracture strain were made,at temperatures from 25 to 850°C. Selected samples that had been tensile tested at 25 and at 65050 vere examined metallographicaily.' Similar inspections and tests were given the speci- mens exposed to the cell atmosphere. A tensile spec1men of INORrS being strained at an elevated temper- ature normally develops some fissures before the specxmen fractures The fissures are 1ntergranu1ar and the fracture is intergranular - this is normal at high temperatures. At room temperature, on the other hand, the normal kind of fracture is mostly transgranular and fissures away v “j » wi » _ . " » Table 2. Heats of Modified INOR-8 Examined After Exposure to MSRE Fuel Salt ~ Specimens A . Content (wt %) . Heat Exposed - - Mo Cr Fe ‘ Mn c Si S P Cu Co Al v Ti HE W Zr B 21545 Rods on S2 . 12.0 7.18 0.034 0.29 0.05 0.015 <0.002 0.001 0.04 0.02 0.02 0.06 0.49 <0.01 0.10 0.01 0.0002 21554 Rods on S2 12,4 7.4 0.097 0.16 0.065 0.01 <0.002 ‘0.004 ) 0.03 0.003 ‘ 0.35 0.0002 67-502 Rods on S3 12.7 7.24 0.08 0.14 0.04 '<0.01 0.004 0.003 0.04 0.02 0.12 0.06 0.53 <0.01 2.15 <0.01 0.0001 67-504 Rods on S3 12.4 6.94 0.05 0.12 0.07 0.010 0.003 0.002 0.03 0.02 0.03 0.22 <0.02 0.50 .0.03 0.0 0.0003 67-551 Rods on R3, S3 -12.2 7.0 0.02 0.12 0.028 0.02 <0.002 0.0006 0.01 0.03 <0.05 <0.001 1.1 0.001 <0.01 .0.0002 7320 Rods -on R3,53 12,0 7.2 <0.05 0.17 0.059 0.03 0.003 '0.002 0.02 . 0.01 0.15 <0.02 0.65 <0.05 <0.05 0.00002 ST 16 Table 3. Extent of Exposure of INOR-8 Specimens in MSRE Time at high , temperature? " Time . Reactor Date _ (hr) in fuel heat * ° Thermal Specimen In Out Total With FPb (hx) neuégz:s/cmz) §i32:22° 7 x 1020 Pre-power core array 1964 8/65 3,36 0 1,090 0.0 0.0 Stringers RLl, RRl, RS1 9/8/65 7/28/66 5,550 2,550 2,813 7,980 1.3 Stringer RS2 f 9/16/66 5/15/67" 5,554 5,220 4,112 25,120 41 Stflnger RS3 5/31/67 4/2/68 6,379 6,300 5,877 32,990 5.3 Stringer RR2 9/16/66 4/2/68 11,933 11,600 9,989 58,110 9.4 Stringer RR3, RS4 4/18/68 6/6/69 7,203 3,310 4,868 18,720 5.1 Stringer RL2 9/16/66 6/6/69 19,136 18,800 14,857 76,830 14.6 Final core arfay 7/31/69 12/18/69 2,815 2,360 2,280 11,870 3.2 Components of fuel loop 1964 1970 30,807 24,500 21,040 96,680 0-19 Stringer X1 8/24/65 6/5/67 11,104 0 0 33,100 0.13 Stringer X2 8/24/65 5/7/68 17,483 0 0 66,100 0.26 Stringer X4 6/7/67 6/ /69 13,582 0 0 51,700 0.25 %rime logged above 500°C. All was at 650 + 10°C with the éxception of 100 hr at 760°C in October, 1965, about 750 hr at 500-600°C in February-March, 1966, and 500 hr at 630°C in December, 1967-February, 1968. bTime above 500°C after exposure of specimen to fuel salt containing fresh fission products. ®Fluence of neutrons with E<0.876 eV, based on flux monitor measurements made during 235y operation, with calculated correction for higher flux during 233U operation. : *®] - “‘ 17 from the fracture are abnormal. Thus'a'high incidence of intergranular' cracking in samples deformed at. 25°C is a definite indication of grain boundary embrittlement or weakness. For this reason, in the descrlptions that follow, attention will be focussed primarily on the metallography of _unstrained specimens or those strained at 25°C. The results of the me- chanical property testing have been fully reported*~7, and only those results that are relevant to the surface cracking phenomenon will be ‘mentioned here he;Specimens‘E2posed Before Power Operation The specimen array that was:in the core during the prenuclear testing and nuclear startup experiments was there primarily as a neutronic stand- in for the later surveillance assemblies. Although it contained similar amounts of graphite‘andpmetal; its design was different from that of later assemhlies;'and'theJINOR-S used was from unidentified heats of standard alloy. The post-eXposure examinations were also limited. How- ever, the results are of some interest because the array had been at high temperature and exposed to salt for reasonably long times. at high temperature for 3306 hr, exposed to flush salt for 990 hr, and exposed to uranium-bearing salt for 1090 hr (see Table 3). The only risible'change'in the specimens as a result of their ex- posure was‘that the originally shiny specimens of INOR-8 had developed a bright Vg-ray—white matte surface. 15 The microstructure of .one of Ithe' specimens after it was‘fractored at 25°C is shown in Fig. 7. .No grain; boundary craCRs'are visible. The 1—millsdrfaceflayer that etched more darkly caused some concern at first. However, as will be-brought out later, the modified surface layer was present on control specimens as well as on those exposed in the core and, on a series of specimens, showed no systematic variation with exposure time. Further work showed that it was likelychld—workrfrom'machining causing carbides to precipi- tate'nore"readily near‘the_surface. (See later discussion of.possible- connection to surface cracking.) wn 1w I O Z w0 " o o Y-109870 ':.fi- o gy ity = ; w w X v 2 ) ) o o i i AR i R .:!&:ii : l: ,A.m_ng:nfi,_,Af’: ‘L ! e *.;'J%; g Y-109869 = I~ 100X jw - N 100X Tot . ~ Fig. 7. Microstructure of INOR-8 Sample Held Above 500°C for 3306 { - hr During the "Zero-power" Run of the MSRE. (Fuel was in system 1090 f o hr.) (a) As polished. (b) Etched with glyceria regia. The material is L cold worked, and the grains are not delineated by etching. - ()] 19 The important point to be noted here is that INOR-8 specimens exposed to fuel salt for a considerable period of time before the generation of substantial nuclear power (and fissionwproduct inventories) showed no in- intergranular surface cracks upon testing at 25°C. First Group of Surveillance‘Specimens (Stringers RL1, RR1l, and RS1l) When the first standard specimen array was taken out in July, 1966, portions were found to have been damaged.¥* Because of the damage,inone‘ of the threetstringers could be put back into the core, but.less that one third of the INOR-8 specimens were affected, so there were plenty of each heat (5085 and 5081) that could be tested. Corresponding speci- mens from the control facility were also tested. The extensive results are reported in detail.t | | Corrosion of the core specimens appeared to be-Very minor. By visual inspection the metal surfaces were a uniform dull gray, with no sign of localized corrosion. Metallography also revealed little or no perceptible corrosion. - o " Tensile tests showed a small decrease in the ‘yield strength of the control specimens\compared with the unexposed specimens and those exposed in the core. Ultimate tensile stresses decreased significantly from the unexposed specimens to the control'Specinens to the core specimens, as shown in Table 4. The variatiofl in the ultimate tensile stress is associated\withdecreases'in fracture strain. . (INOR-8 tensile specimens tested at 25°C continue to strain-harden to-nearlfracture, so any reduc- tion in fracture strain would cause failure at a lower ultimate stress.) Examination of:theee and later specimens indicated that the reduction in fracture strain at 25°C was not connected with the surface—cracking ?henomenon, but ‘was due to carbide precipitatlon that occurred upon aging ‘and was enhanced by irradiation. | '*Thefdamage occurred because, when the core was drained; some salt was trapped between the specimens, where it froze and interfered with the differential contraction of the graphite-metal assembly during cooldown.1® This problem was avoided in subsequent assemblies by a slight design change. Table 4. Tensile Properties of Surveillance Samples From First Group at 25°C and a Strain Rate of 0.05/min Heat Cofiditiona Yield Ultimate;. - Uniforfi . - Total Reduction - Stress Tensile Stress Elongation Elongation @ In Area (psi) | (psi) % % - % ©.5081 Annealed 52,600 125,300 56.7 59.5 50.5 5081 Control = 47,700 118,700 55,9 57.6 48.8 5081 Irradiated 51,100 105,500 38.5 38.7 - 31.3 5081 Irradiated 54,100° . 109,000 k2.6 . 4206 25.9 5085 Annealed . 51,500 120, 800 52.3 53.1 42.2 5085 Control = - 46,200 109, 200 | 40.0 400 - 28.6 5085. Control 45,500 111,200 . 46.8 46.8 3L.5 5085 Irradiated 48,100 100,300 - 36.3 34.5 26.0 qAnnealed — Annealed 2 hr at 900°C. Control — Annealed 2 hr at.900°C annealed 4800 hr at 650°C in static salt. Irradiated — Annealed 2 hr at 900°C. irradiated to a thermal fluence of 1.3 x 1020 neutrons/cm? over 5550 hr at 650° C in the MSRE. | 0z ") 21 Effects of Carblde Precipltation An observatlon that connected the decrease in 25°C fracture strain with - carbide prec1p1tat10n was the follow1ng The fracture strain at room temperature of irradiated specimens could be improved by an anneal’ - of 8 hr at 870 C (p. 17, ref. 5) ~ This is a carbide agglomeration anneal and the recovery of ductillty by such.an anneal suggested that the em- brlttiementrwas due to‘the,prec1p1tat10n of copious amounts of carbide. ThepreCipitateluas_observedand identified as MgC, which is brittle at’ room*temperatureak Extraction replicas showed more precipitate in core specimens than in control-specimens. Thus it appeared that, at least in theEStandard alldy, irradiation'enhanced-the nucleation and growth of the prec1pitate that occurs to some extent at hlgh temperature without irra- diation. ) | | _ MEtallography of specimens broken in tension at 25°C revealed differ- ences .in the nature of the fractures in core. and control specimens . These differences are believed to be ‘another manifestation of carbide precipita- tion._ Flgures 8 and 9 show specimens of heat 5081 from the control facility and from the core. The fracture in the control specimen (Fig. 8) is typical of transgranular shear-type failure (cup—cone appearance). In contrast, the fracture of the core specimen (Fig 9) is largely inter- granular. Heat 5085 spec1mens are shown in Figs. 10 and 11. The fracture in the 5085 control specimen is mixed transgranular and intergranular, with the elongated grains attesting to the large amount of strain. The frac- ture in the 5085 core specimen is 1argely intergranular with numerous 1ntergranular cracks in the microstructure Surface -Cracking Spec1mens of both heats exposed in the control facility and tested in'ten51on showed no cracks except very near the fractures. The core specimens of both heats, on the other hand, showed several intergranular cracks along the gage lengths. The difference is clearly shown in Fig. 12, which is a composite of photomicrographs of longitudinal sections of strained\controi and core specimens of heat 5085. Pictures of .- the sections along the entire gage lengths were examined to determine 22 Y-78234 10.040 in. I 0.035 INCHES 100X 10.030 in, A -~ 0.023 INCHES P 100X Fig. 8. Photomicrographs of Control Specimen AC-8 from Heat 5081 Tested at 25°C and at a Strain Rate of 0.05/min., (2) Fracture. (b) Edge of specimen about 1/4 in. from fracture. Etchant: glyceria regia. ) ) 23 Fig. 9. Photomiqrogréphs of Surveillan¢e Specimen D-16 from Heat 5081 Tested at 25°C and at a Strain Rate of 0.05/min. (a) Fracture: (b) Edge of specimen about 1/4 in. from fracture. Etchant! ' glyceria regia. 0.035 INCHES N 100X = e 0.035 INCHES M 100X B v Y-78258 'r_ I 100% 0.025 INCHES e f= U.U2 ) INLHED i 100X Ji Fig. 10. Photomicrographs of Control Specimen DC-24 from Heat 5085 Tested at 25°C and at a Strain Rate of 0.05/min. (a) Fracture. (b) Edge of specimen about 1/4 in. from fracture. Etchant: glyceria regia. 25 ") [~ 0.035 INCHES - N 100X i " 0.035 INCHES N 100X o 'Fig. 11. -Photbmicrographs bf.Sufveillance'Speéimefi'Aflé from Heat 5085 Tested at 25°C and at a Strain Rate of 0.05/min. (a) Fracture. (b) Edge of Specimen about 1/4 in. from fracture. Etchant: glyceria regia. »d i R - 55180 P ATED _.m & e ! e s in c men wa C After 5550 hr Above 500 at 5085 strained at 25 Photomicrographs of He 12 top specimen was in static fuel salt containing depleted uranium, 18- Fi the lower specim and ctively. 070 and 0.090 in., respe The true specimen diameters are 0 \ the MSRE core. o) wi 27 frequency and depth of cracks. In the sectien.Of the control specimen, only one surface crack was found (a frequency of about 1 per inch). Its depth was 5.7 mils. The section of the core specimen showed 19 cracks per inch with an average depth of 2.5 mils and a maximum depth of 8.8 mils. The core specimen has more surface cracks than the control speci- men (and the pre-power core specimen). Second Group of Surveillance Specimens (Stringer RS2) In May 1967 stringer RS2, with specifiens of two modified heats, was removed, and a stringer containing specimens of two other modified heats was installed. At the time it was removed, stringer RS2 had been at high temperature fbr_practically the same length of time as the first group, but had seen 3 times'the neutron dose and fission product concentration. The core specimens from RS2 and corresponding control specimens (stringer CS2) were 1ntensively examined and tested.® Visual inspection showed the core spe01mens to be very slightly dis- colored but otherwise apparently unaffected. Photomicrographs of speci- mens (unstrained) after exposure are shown in Figs. 13'and 14. (The unusually fine grain size is due to the fabrication history of the specimens, which included 100-hr anneals at 870°C.) The as-~polished | views of specimens of both heats show some evidence of grain boundary modifications to a depth of 1 to 2 mils, and-both show eome tendency to etch more readily neaf;the suffaee.' | N Core speeimens.of both heats tested to failure in tension_developed numerous intergranuler cracks along the surfaces, while control specimens tested similarly showed few or no. surface cracks The difference is illustrated in Flgs. 15-18. Figure 15 and . 16 are control ‘and core speci- .mens respectively of heat 21554.— In the core specimen surfaces nearly every grain bbundary_is cracked to a eonsistent depth'of about 2 mils. In the control specimen there is no evidence'of-intergranular edge cracking. Figures 17 and 18 show the same thing for heat 21545. 28 0w w I L9 2 0 ™ © O 0.007 INCHES o Fig. 13. Photomicrographs of Zirconium-Modified INOR-8 (Heat 21554) Removed from the MSRE after 5554 hr above 500°C. Edge exposed to flowing salt. (a) As polished. 500x. aqua regia. 500x. Reduced 30%. (b) Etchant: aqua regia. 100x. (c) Etchant: - 29 . 7 INCHES 00X - Fig. 14. Photomicrographs of Titanium-Modified INOR-8 (Heat 21545) Removed from the MSRE After 5554 hr above 500°C. Edge exposed to flowing salt. (a) As polished. 500x. (b) Etchant: aqua regia. 100x. (c) Etchant: aqua regia. 500x. Reduced 33%. 30 {0.010 in. 0.035 INCHES 100X 10.030 in. Fig. 15. Photomicrograph of the Fracture of a Zirconium-Modified INOR-8 Sample -(Heat 21554) Tested at 25°C at a Strain Rate of 0.05/min. Exposed to a static fluoride salt for 5554 hr above 500°C before testing. Note the shear fracture and the absence of edge cracking. 100x. Etchant. glyceria regia. | 31 u AR : ! R-41482 100X 100X T [re—— e (0.0 35 [NCHE S srrertosrinimos s st sy - (0,035 INCHE § —— 1 o 2] 0,007 INCHES =ttt et ~TR = 1 T o ‘ Fig. 16. Photomicrographs of a Zirconium-Modified INOR-8 Surveillance - Sample (Heat 21554) Tested at 25°C at a Strain Rate of 0.05/min. Exposed | \55/ in the MSRE core for 5554 hr above 500°C to a thermal fluence of 4.1 x 1020 | neutrons/cm?. Etchant: aqua regia. (a) Fracture. 100x. (b) Edge of sample ‘ about 1/2 in. from fracture. 100x. {c) Edge of sample showing edge cracking. - 500x. Reduced 31.5%. ‘ -) 32 100X 0.035 INCHES - i~ o ~ Fig. 17.. Photomicrograph of the Fracture of a Titanium-Modified INOR-8 Surveillance Sample (Heat 21545) Tested at 25°C at a Strain Rate of 0.05/min. Exposed to a static fluoride salt for 5554 hr above 500°C before testing. Note the shear fracture and the absence of edge cracking. 100x. Etchant: glyceria regia. ‘ | | | -) 33 R-41296 0 100X ~— 0.035 INCHES —————————————asr————a = 1 T 1= 0.035 INCHES TR 100% T T 0.007 INCHES B ke S00% o 15 = Fig. 18. Photomicrographs of a Titanium-Modified INOR-8 Surveillance Sample (Heat 21545) Tested at 25°C at a Strain Rate of 0.05/min. Exposed in the MSRE core for 5554 hr above 500°C to a thermal fluence of 4.1 X 1020 neutrons/cm?. Etchant: aqua regia.. (a) Fracture. 100x. (b) Edge of sample about 1/2 in. from fracture. 100x. (c) Edge of sample showing edge cracking. 500x. Reduced 32%. 34 Yield and ultimate strengths at 25°C were not appfééiably different for the control and core speéimens.5 Furthermore, although the edges of core specimens cracked intergranularly; the fractures were predominantly. or entirely transgranular, as seen in Figs. 16 and 18. The fact that at 25°C the ultimate stresses were not diminished by exposure and that the fractures were not intergrafiular in these modified alloys (in contrast to the observations on the standard heats) is attributed to the grain- boundary carbide precipitafion Beifig less embrittling in the modified ~alloy. ' - - | The most important observation relativelto the cracking phenomenon is that the two modified alloys.(heats 21554 and 21545) with much smaller grain sizes exhibited intergranglar:cracking-when deformed after exposure to the fuel salt, Although'the depths of the cracks are less in the two ‘modified alloys the same types of cracks are formed in both the standard and modified samp1es. The alloys involved had significant variations in Fe (4% to <0.1%), Mo (16.7% to 12.0%), Si (0.6% to 0.1%), Zr (<0.1% to 0.35%), and Ti (<0.01% to 0.49%), and the fact that all formed intergran- -ular cracks indicates that these compbsitional variations are not impor- tant in the cracking process. Third Group of Surveillance Specimens (Stringers RR2 and RS3) | At the conclusion of operation with 235U fuel, the core array was removed and specimens from two of the three stfingers were tested. The remaining stringer, containing specimens of vessel heats, was put back into the core for more exposure along with two new stringers containing specimens of modified heats. The stringers from the core and the cor- responding stringers from the control facility included two heats (5065 ~ and 5085) of standard INOR-8 and two different,modifiéd alloys. Complete results of testing of these specimens are reported.® - | . Visual examination showed the INOR-8 to be slightly discolored. but otherwise in very good condition.!? 35 Standard Alloys | Stringers RR2, with,fihe standard heats, had been exposed 2 to 4 timeévés lbng as the stahdard-alloylspecimens in the first group. (The factor depends upon which expésure:index is used; see Table 3.) Effects of the longer exposure were evident in metallographic examinations of unstrained specimens. Figures 19 and 20 ére of specimens of heats 5065 and 5085 exposed on stringer RR2 in the core. Some of the grain bound- . aries fiear the surface are visible in the a34polished condition, and a few appear to be opened as small cracks with a_maximum visible depth of about 1 mil. The etched views show the large amounts of carbide pfe- cipitate that formed along the grain boundaries and the modified struc- ture near the-surface, which is thought to be due to working from machining. In samples of these heats exposed in the control facility, grai%’boundaries Were.not visible in as~polished samples. This is shown for heat 5065 in Fig. 21. As in the core specimens, however, etching brought out the carbide precipitate and the modified stfucture near the surface. The microstructure of ¢ontrdi\spggimens‘of heat 5085 was quite similér; ': _. '_ : ¢ ‘f o . - Cfacking at the surface of unstrainéd material was even more evi- Adent in the examination of the straps that bound stringer RR2. These straps and thoSé around'the contro1 stringer CR2 were of standard INOR-8 heat 5055. As shown in Fig.'22,‘the strap exposed in the core had cracks to a depth of about 1.5 mils. Cracks.were distribfited_uniformly, both on surfaces exposed}to‘flowing sél; and those facing the graphite speci- mens. In contrast, the‘sfiraps exposed in the control facility did not showvany cracks. The straps after being formed, had been annéaled for 1 hr at 1180°C and should not have been stressed thereafter sincé.they expanded more at high temperatfire than did the graphite they enclosed. ‘Samples of heats 5065 and 5085 that had been exposed in the core and in thé control facility were tested_ih tension at 25°C; with the results ~shown in Table 5. Some of these strained specimens were sectioned and examined metallqgraphicé;ly. 36 R-45435 Fig. 19. Photomicrographs of INOR-8 (Heat 5065) Surveillance Specimens Exposed to Fuel Salt for 11,933 hr above 500°C, 500x. Shallow reaction layer is seen near surface. (a) Unetched. (b) Etched (glyceria regia). | 0 o | o o . . T~ U ) g - o g8 2 3% _ &0 ~ ? e . VU~ o W § 5 | 1W Q - = M o - , _ 23 oW N ’ S - Q @ 9 | ~ X : Q0N u o fi o O oo S Do W ~ O [ e T o2 | P~ _ QO O Be . %2 . - 'o . n oM Rov & . 0w e 00 = o~ . O «© oo N O 0 & .mf. , e - . . i QW e *. T e O 2 | g 8 . O M B o Se A ~ o 9 - ) - i n Q SR ‘O W o . & - : ) ] : | & | 38 Y-92950 Fig. 21. Photomicrographs of INOR-8 (Heat 5065) Surveillance Control Specimens Exposed to Static Barren Fuel Salt for 11,933 hr above 900°C. ~ Note the shallow reaction layer near the surface. (a) Etched. 100x. (b) As polished. .500x. (c¢) Etched.. 500x. Etchant: -glyceria regia. 39 o o o o _ A m < mw 500X 9 N ¥ ~ , m L l l 1 i | ! “ | | o) o 109 0 ©l ~ 2 - 160 MICRONS - 1 I : SIHONI L000 - o s e i s ien B S st e i s S S 0T e S e e i D e s e R B e et e S o s ! ¢ - ' 8 (Heat 5055) after Exposure e i : . e 3 Photomicrographs of INOR- the MSRE Core and (b) the MSRE Control Facility for 11 S 22, Fig e P 00°cC. As polished. 933 hr above 5 a2 g 0" 8 U o & — ~ o QO 5 o] = w v = §d - 0 Y 0. & ~ &d v - 0 W e U v o o o = =i o] - - o g 0 £ 500x. Table 5. Tensile Properties of Surveillance Samples From Third Group at 25°C and a Strain Rate of 0.05/min ‘Heat . Condition® Yield , Ultimate Uniform - Total Reduction Stress Tensile Stress Elongation Elongation In Area, (psi) (psi) z : Z % 5085 Annealed 51,500 120,800 52.3 - 53.1 42,2 5085 Outside 46,500 99,100 32.8 . 32.8 24.5 5085 Control 53,900 115,900 38.4 38.6 o 29.7 5085 Core 52,300 95,000 28.7 128.9 20.0 5065 Annealed 56,700 126,400 52.9 55.3 50.0 ' 5065 Outside 49,000 | 118,800 57.8 59.7 38.4 5065 - Control 60,900 126,700 46.5 47.4 139.3 5065 Core 51,700 109,300 4l.4 41.5 34.1 ®pnnealed — Annealed 2 hr at 900°C. Outside — Annealed, irradiated to a thermal fluence of 2.6 x 1019 neutrons/cm? over a period of 17,483 hr at 650°C. Control — Annealed, ex- posed to depleted fuel salt for 11,933 hr at 650°C. Core — Annealed, irradiated to a thermal fluence of 9.4 x 1020 neutrons/cm® over a period of 11,933 hr at 650°C. 0% (8] 41 Microstructures of stressed samples of heat 5065 ‘from the control facility and from the core are shown in Figs. 23 and 24, respectively. Numerous intergranular edge cracks are found in the sample from the core (Fig. 24) and very few in the sample from the control facility (Fig 23). The fractures of both of these samples are largely intergranular at all locations and not just near the edge, which is further evidence that the reduction in fracture strain (Table 5)_is due to carbide precipitation along the grain boundaries_and*not related to the intergranular cracking near the surface. Tested samples of heat 5085 from the control facility and_the core are shown in Figs. 25 and 26, respectively. The as-polished views in the latter_figure show some of the large carbide particles that fractured during testing. | S : The samples that were fractured at 25°C‘nefé.repolished at a later date, one-half of each_fractured.sample was photographed. The composite microstructures for heat 5065 are shown in Fig. 27. There are obviously more intergranular cracks in the-sample exposed to the core than inrthe sample exposed.in the control facility. 'There‘were 3 cracks per inch with an average depth of 1.0 mil in the control sample and 230 cracks per inch with an average depth of 1.8 mils in the sample from the core. A composite .photograph of the tested portions of the heat 5085 sample is shown in Fig. 28. There were 134 cracks per inch along the edge of the sample from the core with an average depth of 1.9 mils. None were v1s1ble along the_edge of the control specimen. Modified Alloys Stringer RS3 with the specimens of modified alloy had been exposed about half -as long as stringer RR2 with heats 5065 and 5085. The modified alloy specimens deformed‘more before fracturing than did the standard alloy specimens, hut'they”also developed numerous intergranular edge cracks As with the standard alloys, the samples from the control fa- cility did not show edge cracks Figures 29 and 30 are photomicrographs of strained specimens of heat 67-502 (modified with 0 49% Ti and 2.15% W and containing 0.047 Fe). The specimen exposed,in the control facility (Fig. 29) has its edges | R 42 Y-92917 Y-94391 SR o i | | | | | Fig. 23. Photomicrographs of a INOR-8 (Heat 5065) Sample Exposed to : Statlc Barren Fuel Salt for 11,933 hr above 500°C and Then Tested at 25°C | _ and a Strain Rate of 0.05/min. 100x. (a) Fracture, as polished. (b) - ‘Fracture, etched. (c) Edge of stressed portion, etched. Etchant: glyceria - regia. ' - ' Fig. 24. Photomicrographs of INOR-8 (Heat 5065) Sample Exposed to. Fluoride. Salt in the MSRE for 11,933 hr above 500°C and Then Tested at 25°C and a Strain Rate of 0.05/min. Thermal fluence was 9.4 x 1020 neu- trons/cm?. 100x. (a) Fracture, as polished. (b) Fracture, etched. (c) Edge of stressed portion. Etchant: aqua regia. | . B ‘ : i ‘ v—103181 100X [ttt errnnn. () 038 INCHE S | - 0.030 in, 10.610 in. 1 10.030 in. 10.001 in. n 0.003 0,007 INCHES 500% 10.005 in. 0007 n - Fig. 25. . Photomicrographs of INOR-8 (Heat 5085) Specimen Exposed to | Depleted Static Fuel Salt for 11,933 hr at 650° and Strained at 25°C. (a) - Fracture, as polished, (b) typical edge of gage section, as polished, (c) ' typical unstressed edge, etched with glyceria regia. ‘R-47921 R-47923 R-47922 Fig. 26. Photomicrographs of INOR-8 (Heat 5085) Sample Exposed to Fluoride Salt in the MSRE for 11,933 hr above 500°C and then Tested at 25°C. Thermal fluence was 9.4 x 1020 neutrons/cm?, - (a) Fracture, as polished. 100x. (b) Fracture, as polished. 500x. (c) Fracture, etched. 100x. (d) Edge of stressed portion, etched. 100x. Etchant: aqua regia. -~ for 11,933 hr The fractures are on the left end C After Exposure to Fuel Salt 1 in. o t 25 Diameters of specimens are about 0 it ‘a i . ; s i Sections of Heat 5065 Tested i T 3 P e | Fig. 27. ~above 500°C. D Fig. 28. Photomicrographs of INOR-8 Specimens Strained to Fracture After 11,933 hr above 500°C. The upper sample was in the control facility and the lower sample was in the MSRE core. The sample diameter is about 0.1 in, i : i g [=] O 0.038 INCHES Fig. 29. Photomicrographs of Alloy 67-502 (see Table 1) Exposed to Depleted Fuel Salt for 6379 hr Above 500°C and Tested at 25°C. (a) Edge of unstressed portion, (b) Fracture, (c) Edge of stressed portion. Figs (a) and (c) etched lightly. Fig. (b) etched more heavily with glyceria ; regia. R-476875 8 £ 2 2 "Fig. 30. Photomicrographs of Alloy 67-502 Sample Exposed to Fluoride Salt in the MSRE for 6379 hr Above 500°C and then Tested at 25°C. 100x. (a) Fracture, as polished. (b) Fracture, etched. (c) Edge of stressed portion, etched. Etchant: aqua regia. Reduced 33%. 50 coated with small crystals of almosfi'pfire iron. (The control facility was constructed of material'containing 4lto'5Z Fe, so the'ttansfer of Fe to an alloy that contained only 0.04% Fe is quite reasonable. ) The sample stralned 55/ before fracture, and the fracture was ‘mixed transgranular and intergranular. The edge was uneven from the large defprmation, but there were very few intergranular'sefiarations. The sample from the core (Fig. 30) deformed almost as much (52%) before fracturing; In contrast to the control specimen, edge cracking occurred at almost every grain‘ boundary. The cracks generally extended to a depth of about 5 mils, with the maximum depth being about 7 mils. However, as shown in the view of the fracture, this material tended to crack.intergranfilarly and these internal cracks may have in some cases 1inked together w1th the surface cracks to make them extend deeper. | T Photomicrographs of samples of heat 67-504 tested at -25°C after exposure in the control facility and in the core are shown in Figs. 31 and 32. This heat modified with 0.50% Hf, contained 0.07% Fe. As shown in Fig. 31, iron deposited on the surface of this heat as it did on 67—5Q2. This sample, from the control facility, deformed 55% before failing with a mixed inter- and transgranular fracture. There were no intergranular edge cracks. The sample exposed in the core deformed - 52% before fracture. The fracture was primarily transgranular but there were frequent edge cracks (Fig. 32). Most boundaries were cracked, but the cracks extended ‘only to a‘depth‘of abddt,szils*in the fields that were photographedt= ' Fourth Group of Surveillance Specimens (Stringers RL2, RR3, and RS4) The last regular core surveillance assembly was removed.in June, 1969 to make way for a special experimental array.lqr Stringer RL2, with specimens of standard INOR-8 (heats 5065 and 5085), had been in the core at high temperature almost 12,000 hr during'the 235Uioperation 51 Y -92839 ettt = 1 1000X ™ T 100X 6.038 INCHES mrrerrrrer ety p—— e oo 0.0035 INCHES 0.0 in. 1 . 0,030 in 100X 0.035 INCHES meam s emeeeeeeer—immcmtny 5o m. 1 oo™ — __ Fig. 31. Photomicrographs of Alloy 67-504 (see Table 1) Exposed to . ' Depleted Fuel Salt for 6379 hr Above 500°C and Tested at 25°C. (a) Edge of unstressed portion, (b) Fracture, (c) Edge of stressed portion. Figs (a) and (c¢) as-polished and Fig. (b) etched with glyceria regia. Reduced 33.5%. 52 -47933 Fig. 32. Photomicrographs of INOR-8 (Heat 67-504) Sample Exposed to Fluoride Salt in the MSRE for 6379 hr Above 500°C ‘and Tested at 25°C. 100x. (a) Fracture, as polished. (b) Fracture, etched. ' (c) Edge of Stressed por- tion, etched. Etchant:. aqua regia. Reduced 1l4Z.: S ‘ 53 ‘and 7200 hr during'233U operation. The other stringers, with two modi- fied alloys, had been eiposed only during the 233U operation. Results of examinations of these sPecimens and corresponding specimens from the control facility are reported 7 Visual examination showed all of the INOR-8 specimens from the core to be noticeably more discolored than those in previous arrays, with sur- face films thick enough to be seen in cross section by light microscopy. There were also slight changes in the appearance of the graphite.18 Both of these observations indicate some difference in the exposure conditions during the most recent operation. (From other observations it was known that the fuel salt was relatively more oxidizing during at least part of the7233U operation. See pp. 85-98 of ref. 13.) Standard Alloys (RL2) | Photomicrographs of unstrained specimens of heats 5065 and 5085 from RL2 revealed many grain boundaries near the surfaces that were visible in the as-polished condition. The view of a specimen of heat 5065 in the as-polished condition,e(Fig. 33) shows a thin surface layer (believed to be the}Cause of the visible discoloration)_and some grain boundaries visible to a depth of about 2 mils. Etching this particular sample re- vealed the typical carbide structure plus a narrow band near the surface that seems to have a high density of carbide. Photomicrographs of heat »5085 after removal from the core are shown in Fié. 34, Many of the grain boundaries are visible in the as-polished condition to a depth of about 2 mils. Etching reveals the typical microStructure'with_very ex- tensive carbide precipitation. In the as—polished view of the heat 5065 control specimen (Fig. 35), grain boundaries are also visible, but in this case the appearance is uniform across the Sample:and is due to | polishing long enough that the different grains are at different eleva- tions.t Etching reveals the shallow surface modification and the typlcal carbide structure. The heat 5085 specimen from the control facility, ‘shown in Fig. 36, was also slightly overpolished so that the grains are visible in the as-polished condition. Etching revealed the typical microstructure and the shallow surface modification. 54 } OF i Fig. 33. Typical Photomicrographs of INOR-8 (Heat'5065) Exposed to the { - MSRE Core for 19,136 hr Above 500°C. (a) As polished. (b) Etchant: | . aqua regia. 500x. | ‘ - . . 55 10.001 in. } 10.003 in. P i - 0.007 INCHES 500X 10.005 in. 10,007 in. - - Fig. 34. Typical Phdtomicrogtaphs of INOR-8 (Heat 5085) Exposed to the MSRE Core for 19,136 hr Above 500°C. (a) As polished. (b) Etchant: glyceria regia. 500x. 56 | Fig. 35. Typical Phctomicrogréphs, of Heat 5065 After Exposure to Static Unenriched Fuel Salt for 19,136 hr Above 500°C. 500x. (a) As polished. (b) - ' Etchant: glyceria regia. - | 57 Y-98172 Fig. 36. Typical Photomicrographs of Heat 5085 After_ExppSfire tOVStatic Unenriched Fuel Salt for 19,136 hr above 500°C. 500x. (a) As polished. (b) Etchant: glyceria regia. - o | _ | - 58 As in the previous set of core specimens, the straps of heat 5055 - showed more cracks in the nominally unstrained condition than did the tensile specimens on the same stringer._ A section of a strap from stringer RL2 that had been exposed, along‘w1th the specimens shown in | Figures 33 and 34, is shown in Fig. 37. The cracks in this as-polished' view are visible to a depth of about 3 mils, about twice as deep as in ‘the strap exposed 12,000 hr during 235U operation (Fig. 22). Heat 5055 straps were also used on the stringers exposed only during the 233y operation (7200 hr). A specimen of one of these straps is shown in Fig. 38. 1In the as-polished condition, cracks were visible to a depth of about 1 mil; etching‘made them visible to a depth of about 3 mils. The cracking was quite uniform along alllsnrfaces of the straps, indi- cating that the deformation of the 20-mil straps during removal was not a factor in the appearance of the cracks. Examination of unirradiated control straps failed to reveal a similar type of cracking. | One other interesting spec1men that showed extensive cracking was a piece of thin INOR—S sheet (heat 5075) that had been attached to the_ strap shown in Fig. 38. The sheet had been rolled to 4 mils and annealed 0.5 hr at 1180°C. The edges of the foil were wrapped around the strap to hold thefmaterial in place. Thus the-outside surface of the foil ‘would have been exposed to flowing salt and the underside and:the folded ends to salt that flowed less rapidly. The foil was extremely brittle and broke while it was being removed from the strap. The photbmicrcgraphs in Fig. 39 show that the foil had extensive grain boundary‘cracks; with many appearing_to extend throughout the 4mil thickness. In some areas the cracks were more numerous on the outside where the material was curved and was in contact with rapidly flowing salt. Photomicrographs of a strap and foil that were exposed to static salt in the control facility_ are shcwn in Figs. 40 and 41. The samples showed no evidence ofvinter— granular cracking. | A fracture surface of the foil from the MSRE was examined by the scanning electron microscope (SEM) and Auger spectroscopy. The fracture is shown in Fig. 42 and is intergranular. The SEM with a dispersive x-ray analytical system indicated the presence only of the elements normally C. Q | H , : . . . =5 - - 85 | o | 0 Bt . | < i L ) . ' ) s b 9 « ew . fl 0 . . : , < o : . e . _ : . N o X , . : O O ny HO ) , . , o : Q . ) - ) et 2] o s , ~ U W ) ., SM.M , : . i . .- B \ %o _ _ o ,. | m% , I | e fOM : - | - cQ , - .w w...m S : o A 590 ¥ © _ sxg | _ o - . oY Q , = o : e~ = - , o] o ) . , U H . , i 1SR | . m . Me . « O & ' , n_.l... H , , & M. | | ‘ o0 B W ol i , : 0w - : : ’ , L , &M , , i . , 19 , . . 0 W Co . oo , J ‘. » ) - ¢ . " . [ b n e e e b b o e R-56739 3 10.001 in. 0.003 in. 500X } 10005 in T K.007 n, Fig. 38. Photomicrographs of INOR-8 (Heat above 500°C. ‘(a) As-polished view near cut (b) outside edge (d) Etched view of outside edge. + - 5055) Strap Etched,yiew Etchant: Lactic acid, HNO3, HCI. R=56751 Exposed to the MSRE Core 7203 hr near cut (c) As polished view of Reduced 32%. 0.038 INCHES 00X 500X i 07 INCHES ——————————erm———————ereriiy 0,003 i, | Toootm | 0.0 I 09 R-56725 R- 56729 F0.001 m. 10.003 in. 16,005 n. 500X H———— 0,007 INCHE S et erieeeeeeeeeeegment. 10.007 in. S00X 0.007 INCHES ———— ety 10.003m. 1 0o m ) 0.005 i 10,007 i Fig. 39. Photonmicrographs of Two Specimens From a Thin .INOR-8!(Heat 5075) Foil that was Attached to the Strap Shown in the Previous Figure. The curvature was caused by forming before being inserted into the MSRE. (a) As polished (b) Etched (c) As polished (d) Etched. Etchant: Lactic acid, HNO3, HCl. Reduced 32Z. e ) OO INCHES e ox Ik 501 ¥ 0.007 INCHE § e eeemsemeeeeeeeeeeoeeoeeo——lif 0.003n | 10,004 in. 1 10.005 n. 1 10.004 . 10.003 in. 500X 19 62 i i o -~ Fig. 40. Photomicrographs of INOR-8 (Heat 5055) Straps Exposed to . . ! Static Unenriched Fuel Salt 7203 hr Above 500°C. (a) As-Polished 33x, ' (b) As Polished, 500x, cracked regions are carbide that fractured during initial forming, (c) Etched with glyceria regia, 500x. Reduced 33%. 63 10.001 in. 500X '0.003in. 1 U vtk 10,005 in. 10.007 in. I 10.00t in, = '0.003 in. 0.007 INCHES 500X $0.005 in. - Fig. 41. Photomicrographs of a Thin INOR-8 (Heat 5075) Foil that was attached to the strap shown in the Previous Figure. (a) As Polished, | ~ (b) Etched with glyceria regia. ‘500x. : o 10067 in. : i | | { i i 1 3 ! 1 i i | i i i i i 1 64 Fig. 42. Scanning Electron Micrograph of INOR-8 (Heat 5075) Foil Exposed to the MSRE Core for 7203 hr Above 500°C. (a) Topographical view of fracture showing the surfaces of the fractured grain boundaries. _500x.' (b) 1000x. o 65 found in INOR-8. Auger electron spectroscopy indicated that Te, S, and Mo were concentrated in the boundaries. Definite confirmation of the presence of these elements was difficfilt because many of the spectra overlapped. | | | - Samples of heats 5065 and 5085 that were removed from the control and the surveillance facility after 19,000 hr were subjected to numerous mechanical property tests, and some specimens were exafiined metallo- graphically after having been strained to fracture. The tensile properties measured at 25°C are summerized in Table 6. The changes in the yield stress are small and probably within experimental accuracy. The reduc- tions in the ultimate tensile stress are due primarily to the reduced fracture strain, wfiich, as explained before, is believed to Be due to carbide precipitetion that is dependent upon time at temperature, irradiation, and the particular'hest involved. | The control sample of heat 5065, whose microstructure is shown in Fig. 43, deformedVSOZ before fracture. The fracture is mixed intergranu- lar andrtransgranular and no cracks formed along the edge away from the immediate vicinity of the fracture. The microstructure of a heat 5065 sample from the core is shown[iniFig._44.” This sample deformed 43% and had a mixed fracture. Numerous‘intergrenuiar cracks to a depth of about 5 mils formed along the gage iength. Samples of heat 5085 from the con- trol facility and the core are shown in Figufes 45 and 46. Tfie control sample failed after 41% strain with a mixed fracture. No intergranular cracks were visible in the field shown in Fig. 45. The sample of heat - 5085 from the core that was tested at 25 C failed after much less strain (22%). As shown in Fig. 46, the frécture'is;mixed, and numerous inter- granular cracks formed along‘the\surfaces to a depth of 5 mils. | Composite'photographs of sections of the broken control and sur- veillafice samples of heats 5065 and 5085 were made for overall viewing.' These ere Figures 47 and 48, Thesevpidtures'show very clearly that the intergranular cracks are mOrerfreqfient'and deeper in the specimens exposed in the core than in the control samples; The photograph of the heat 5085 core sample also shows that very small strains are sufficient to make the cracks visible. The sample had a 3/16-in. radius at the Table 6. Tensile Properties of Surveillance Samples From_ Fourth Group at 25 C and a Strain Rate of 0.05 min—1 Heat . Condition? ) Yield . Ultimate" : Uniform f: fl, Total Reductiofi; ' | Stress Tensile Stress Elongation - Elongation In Area (psi) - (psi) % z -. % 5085 EAnfiealedf: 51,500 . 120,800 o523 531 42,2 /5085 Control | 48,100 . 108,500 ~40.5 40.6 . 32.8" 5085 _ Irradiated 53,900 . 89,000 - 22.0 221 19.6 5065 Annealed 56,700 126,400 B 52.9 ~ 55.3 . 50.0 /5065 . Control 161,200 126,500 48.5 49.8 36.8 5065 - Irradiated = 56,700 - 109,500 42,5 42.8 33.2 8 Annealed — Annealed 2 hr at 900°C. Control — Annealed, .exposed to depleted fuel salt for | 19,136 hr at 650°C. Irradiated — Annealed, irradiated in core to a. thermal fluence of 1.5 x 1021 neutrons/cm2 over a period of 19 136 hr at 650°C. 99 | ' T T RIS T T e R Y e Y 9821 - S Fig. 43. Typical Photomicrographs of INOR-8 (Heat 5065) Sample Tested at 25°C After Being Exposed to Static Unenriched Fuel Salt for 19,136 hr . above 500°C. Etchant: glyceria regia. (a) Fracture. 100x. (b) Edge ' near fracture. 100x. (c) Representative unstressed structure. 500x. Reduced 22%. ‘ | D " .. Fig. 44, ‘Photomicrographs of INOR-8 (Heat 5065) Sample Tested at 25°C After BeingExgosed ‘to the MSRE Core for 19,136 hr Above 500°C and Irradiated to a Thermal Fluence of 1.5 x 102! neutrons/cm?. (a) Fracture, etched. 100x. (b) Edge, as polished. 100x. (c) Edge, as polished. '500x. (d) Edge, etched. 100x. Etchant: glyceria regia. Reduced 26%. 89 - Fig. 45. Typical Photomicrographs_of INOR 8 (Heat 5085) Sample Tested at 25°C After Being Exposed to Static Unenriched Fuel Salt for 19,136 hr Above 500°C. (a) Fracture, etched. 100x. (b) Edge near fracture, etched. 100x. (c) Representative unstressed structure, etched. glyceria regia. Reduced 24.5%. 500x. Etchant: B -50900 Fig. 46. Photomicrographs of INOR-8 (Heat 5085) Sample Tested at 25°C After Being Exposed to the MSRE Core for 19,136 hr Above 500°C and Irradiated to a Thermal Fluence of 1.5 x 102! neutrons/cm?. (a) Fracture, etched. 100x. (b) Edge, as polished. 100x. (c) Edge, as polished. 500x. (d) Edge, etched. 100x. Etchant: glyceria regia. Reduced 27%. ' : ' 0L 71 i . s 3 2 i T = ’ i & o . _ = 3 , .m.m , o ,‘ , °c T , | 5mr . , ’ ™~ o | . BLO W . o ¥ s . o o . | 0O & R e | & Q9 Jg B o , , , m,wd - o @ , | o u B W e . umMm a g B ” o ol oeH o W 87 e o on M H O 7~ Y 0 - . : 5 | S5 * wny U W m 4 - ’ w— O | N , : N U Q@ . L] . r R | o O.C_WM e o ‘ H e 4 :mmve 3. o P - RA o . S MWW . Moo oy L o i QO w o W= W g O 0 . , .th . | . _ 0T Q. o | %@ a | : o 0d - : CRaE ‘ ‘.:e.m : _ . Lo~ | 8 | g LB Y | i S0 Mo S EETE %_‘eul . . . P_w . . S O S IRRADIATED £ Fig. 48. Photomicrographs of INOR-8 (Heat 5085) Strained to Fracture at 25°C. The top specimen was exposed to static unenriched fuel salt for 19,136 hr above 500°C and the lower specimen was exposed to the MSRE core for the same time. The diameters are about 0.1 in. ' ¢L 73 ~ end of the gage section and the diameter increased rapidly from 1/8 in. to 1/4 in. Thus the stfess, and hence the strain, decreased rapidly along this radius; Note'in'FigJ 48 that the cracks continue a large distance along the radius, into the region of low strain. An additional experiment was made with a core sample of heat 5085 that had been cut too short for mechanical property testing. The- re—:, maining segment was bent about 30° in a vise, then was sectioned and examined metallographically. The resulting photomicrographs are shownu in Fig; 49. The tension side had intergranular cracks to a depth of about 4 mils, but no cracks were evident on the compression side. Both the tension and COmpression sidestetched ahnormally near the surface-' because'of.the'modification thought to be due to cold wofking. Modified Alloys (RR3 and RS4) , The specimens on stringers RR3 and RS4 were of two different heats of titanium-modified alloy:._heat 7320,'containing 0.65%Z Ti, and heat 67-551, containing 1.17% Ti. The 7200-hr exposure of these specimens had been ehtirelf during the 233y operation, in which the initial oxidation state was relatively high. The specimens were discolored and had much the same appearance as theMSEandard heats. The samples were subjected to Various mechanical prOpefty‘tests:and some were examined metallograph- ically. The microstructnre of heat 73é6 after exposure in the control facility and testing at 25°C is shown in Fig. 50. The sample deformed .50/ before fracturing in a mixed mode.fi No edge cracks were visible. A similar sample that was exposed to theufuel salt is shown in Fig. 51. It'failed afterldeforming‘AS%. Edgeheracks nere present:along almost every grain boundary to a depth of 3?to 5 mils. Note that many of these cracks proceed to where the'crackingfgfain boundary intersects another | boundary. Note also that the‘crack:tips are quite blunt indicating that they formed early dufing'deformation and were simply spread by further " deformation and did not'propagate.'fSimilar'metallographic observations were made on heat 67 551 Control'and core specimens deformed almost: equal amounts before fracturing (524 and 51/) No edge cracks were ob- served in the sample from the control facility (Fig. 52), but numerous . intergranular cracks were evident in the sample from the core (Fig. 53). Fig. 49. Typical Photomicrographs of INOR-8 (Heat 5085) Sample Exposed to the MSRE Core for 19,136 hr Above 500°C. The sample was bent in a vise. (a) As polished, tension . side. 100x. (b) As polished, tension side. . 500x. (c) Etched, tension side. 500x. (d) Etched, compression side. 500x. Etchant: aqua regia. Reduced 27%. ' ‘ Wi 75 Y-98205 ] | 7203 hr Above 500°C. (a) Fracture. 100x. (b) Edge. 100x. unstressed microstructure. 500x.. Etchant: glyceria regia. Fig. 50. Photomicrographs of a Modified INOR-8 (Heat 7320) Sample ‘Tested at 25°C After Being Exposed to Static Unenriched Fuel Salt for (c) Typical Reduced 22.5%. 76 e |R-50961 Fig. 51. Photomicrographs of a Modified INOR-8 (Heat:7320) Sample ‘Tested at 25°C After Being Exposed to the MSRE Core for 7203 hr Above . 500°C and Irradiated to a Fluence of 5.1 x 1020 neut—rons/cmz. (a) Fracture, etched, 100x. (b) Edge, as polished. 100x. {c) Edge, Etched. 100x. Etchant: glyceria regia. Reduced 22%. Y-~98199 Y-98201 Ty pe ~ Fig. 52. Photomicrographs of a Modified INOR-8 (Heat 67-551) Sample _ Tested at 25°C After Being Exposed to Static Unenriched Fuel Salt for 7203 . hr Above 500°C. (a) Fracture. 100x. (b) Edge. 100x. (c) Typical un- stressed microstructure. 500x. . Etchant: glyceria regia. Reduced 20.5%. 78 R-50953 Fig. 53..° Photomicrographs of a Mbdified INORr8 (Heat 67-551)'§§fip1e : Tested at 25°C After Being Exposed to the MSRE Core for 7203 hr Above 500°C and Irradiated to a Fluence of 5.1 x 1020 neutrons/cm?. (a) Fracture, etched. 100x. (b) Edge, as polished. 100x. (c) Edge, etched. 100x. Etchant: Glyceria regia. Reduced 17%. 79 Spécimens Exposed to Cell Atmosphere All of the INOR-8 speciméns exposed to the cell atmosphere outside the reactor vessel (N, with 2-5% 0;) developed a dark gray-green, tena- cious surface film. Examination indicated that the'film‘was oxide, and there was indication of nitriding. _ | | . | ‘Figure 54 is a photqmicrograph‘of_one of the cell specimens exposed the longest (17,483 hr at high témpératurg). It shows a very thin, uni- form layer of oxide on the surface, with internal oxidation extending to a dépth of 1 to 2 mils. 'AISQ-evidént is the usual MgC-type carbide formed during the primafy working and the long thermal aging. Results of 25°C tensile tests on the standard INOR-8 specimens (heats 5065 and 5085) showed reductions (relative to unexposed material) in ultimate stress’and fracture strain consistent With'expectations from the results on control and core specimens. (See Table 5.) Figure 55 shows a sample of heat 5065 exposed to the cell environment ‘and fractured at 25°C. The eldngated grains attest to the large fracture strgin‘(SQZ), and the fracture is mixed transgranular and intergranular.'v Aihéét‘SOBS specimen éxposed the safie 1eng£h of time failed withVOnly 33% fracture strain, in a primarily ifitergranular'mode as shown in Fig. 56. Both specimens showed a few shallow surface cracks, but the oxide layer did not appear to affect the deformation of the specimens. Studies Related to Modified Surface Microstructure Reference waé made in severa1 instances to a modified microétructuré. Metals quite often have different_microstructfires near the surface than deeper in the piece. Processing methbdslcan account for these variations on as-fabricated surfaces. Hot wOrking is often acéompanied by oxidation or decarburization that can produce_either larger or smaller grains near . the surface. Tubing is often made by alterfiatély drawing cold and an- nealing to soften the metal. The lubricants used in drawing are difficult to clean from the inside surface, and residues can diffuse into the'tubing - 80 R-45443 Q.007 INCHES 500X - 10,007 in. 10.001 in. I 10.003 in. [ 10,005 in. I 10.001 in. 0.007 INCHES 500X 10.003 in. 10,005 in, Fig. 54. Photomicrographs of Hastelloy N (Heat 5065) Surveillance Specimens Exposed to the Cell Environment of N, + 2 to 5% 0, for 20,789 hr at 650°C. 500x. (a) Unetched showing surface oxidation (b) Etched (glyceria regia) showing shallow modification of microstructure due to reaction with cell environment. ' 10,007 in. - Ty 81 R-47883 Fig. 55. Photomicrographs of INOR-8 (Heat 5065) Sample Tested at . o 25°C at a Strain Rate of 0.05/min. Sample had been irradiated to a ‘thermal fluence of 2.6 x 1012 neutrons/cm? while being exposed to the | ' cell environment above 500°C for 17,483 hr. 100x. (a) Fracture, as . | polished. (b) Fracture, etched. (c) Edge of stressed portion. Etchant: U aqua regila. 82 R-47891 Fig. 56. Photomicrographs of INOR~8 (Heat 5085) Sample Tested at 25°C at Strain Rate of 0.05/min. Sample had been irradiated to a thermal fluence of 2.6 x‘1019 neutrons/cm? while being exposed to the cell environment above 500°C for 17,483 hr. 100x. (a) Fracture, as polished. (b) Fracture, etched. (c) Edge of stressed portion. Etchant: aqua regia. = - - 83 during annealing, The lubricants are usually high in silicon or carbon and these elements can cause a layer of small grains to form near the affected surface, | 7 | Our surveillance samples weré machined from larger pieces of mate- rial, so none of the explanations that we have given can apply. The 1/4-in.~diam surveillance rods were fabricated by several methods, all involving somé,cold working., The rods Were-annealed‘for 1 hr at 1180°C (standard mill anneal) in argon and then centerless ground to the final dimension of about 0,245 in, Thel/8vin.~diam gage section was machined. The photomicrographs in Fig. 57 show typical surface modifications noted on surveillance and control specimens after long annealifig times at 650°C. The modification is-generally less than 1 mil deep and is not detectably influenced by irradiation. Microprobe examination of the’ first control sample indicated that the modified region was slightly enriched in carbon and that none of the other alloying elements varied. Thus the modification appears to be a region of fine carbides that are dispersed alongrlines. We postulated that these lines were slip lines that resulted from the surface working during the final stages of fabri- cation, These slip.lines would provide preferred sites for carbide pre~ cipitation when the alloy was held at 650°C. We were able to dfiplicate'the surface modification by solution annealing a rod, centerless grinding, and anhealing for a long period at 650°C. The resulting microstructure is shown in Fig. 58. If is quite similar to the microstructures in Fig. 57 and confirms that the modification is due to surface finrking. Thus, this modification is unrelated to the intergranular cracking phenomenon. Summary of Observations on Surveillance Specimens ' This section is intended as .a convenient summary of the observations on the surveillance specimens that we believe are pertinent to the inter- granular surface cracking phenomenon. Discussion and interpretations will come later, after the descriptions of the observations on'MSRE components. 84 Fig. 57. Photomicrogfaphs of'Unstressed INOR-8 (Heat 5085) Control and Irradiated Samples. : ‘ : : N Y-99693 Fig. 58. Photomicrograph of INOR-8 (Heat 5085) Rod that was Annealed 1 hr at 1177°C, Centerless Ground 5.2 mils, and Annealed 4370 hr at 650°C in Argon. 500x. Etchant: glyceria regia. 85 INOR~8 specimens exposed to fuel salt for more than 1000 hr before the beginning'of power operation showed no intergranular surface cracking. All specimens exposed thereafter in the core for periods ranging from 2500 to 19,000 hr did show intergranular surface cracks after being strained in tension, No surface cracks were visible in unstrained speci- mens from the first set}of core surveillance specimens KexPosed for 2500 hr after. the Beginning of ‘power operation), but specimens from.all arrays removed later did show some cracks or incipient cracks in. the unstralned as-polished condition. ' ' In contrast to the core sPecimens, specimens exposed to salt in the control fac1lity for equal times at high temperature and then tested in tension showed only a few surface cracks. Specimens exposed'to the cell atmosphere developed an_oxidizedllayer,-which-cracked upon;being strained; but few or no:cracks extended deeper_than the oxide layer; | The;severity'of surface cracking in core and control specimens of MSRE vessel heats exposed during various periods of time was measured as follows. Photomicrographs were made of polished longitudinal sections of specimens of heats 5065 and 5085 that had been fractured in tension Cracks viS1b1e along the edges of the gage portions of the specimens wvere counted, and the average and maximum depths were determined Table 7 gives the observed ‘crack frequencies (number per inch of gage length) and depths. ' We saw very evident differences between different heatshin the de- gree of surface crackingt Strained'specimens.of heat 5065 had more cracks than_did:sPecimens;of heat SOéS ekposed.simultaneously,.but}average depths were not{very*different.f (See Table 7.) Statistics were not gathered on specimens of modified-heats but the photomicrographs of strained specimens show crack frequencies and depths comparable to those in spec1- mens of heat 5065.' Straps and a f011;of different standard heats showed more cracks than were visible .in unstrained specimens of heats:5065 and 5085. ‘ | Pl o | The surfaces of specimens exposed during the first part of the 233y :'0peration were noticeably more. discolored than the surfaces of specimens exposed only during the 235U operation. Presumably this was ev1dence of Table 7. Crack Formation in Hastelloy N Surveillance Samples' Strained to Failure at 25°C . | Time at High Temperaturea (hr) . Crack Count Degth-gmils) Sample Description '~ Total With FPD Counted (in.”D) Av Max Total | ' Control, Heat 5085 5,550 2,550 1 1 5.7 5.7 Heat 5085 - 5,550 2,550 26 19 2.5 8.8 Control, heat 5085 11,933 11,600 0 - | Heat 5085 11,933 11,600 178 - 134 1.9 6.3 Control, Heat 5065 11,933 11,600 3 3 1.0 2.0 Heat 5065 11,933 11,600 277 . 230 1.8 3.8 Control, Heat 5085 19,136 18,800 4 | 3 1.5 2.8 Heat 5085 19,136 18,800 . 213 176 5.0 7.0 Heat 5085 19,136 18,800 o140 146 3.8 8.8 Control, Heat 5065 19,136 18,800 3 . 3 2.5 4.0 Heat 5065 | 19,136 18,800 240 229 5.0 7.5 471 me logged above 500°C. All was at 650 * . 10°C with the exception of 100 hr at 760° c ~.in 10/65, about 750 hr at 500-600°C in 2-3/66 and 500 hr at 630 C in 12/67-2/68 bTime above 500 C after exposure of specimen to fuel salt containing fresh fission prod- ucts. “One crack that may not be surface connected and probably had a different cause was 15 mils deep, next largest was 7 mils. ag 87 the relatively more oxidizing_condition-of the salt following the pro- cessing. It also indicates the importance of recognizing that corrosion (and possibly the surface cracking)'was not proportional to time. The fluctuation in the concentration of chromium in the fuel salt (Fig. 5) is further evidence of this fact. Aging at high temperature caused MgC—type carbide to precipitate throughout Specimens of heats 5065 and 5085. Irradiation enhanced this process. jThe increased amounts of intergranular Mg C (which is brittle at room'temperature) were attended by reductions in the fracture strains and ultimate stresses at 25°C and a shift in the nature of the fractures at 25°C from largely transgranular to more nearly intergranular. These changes were observed in the surveillance speclmens from the core and the cell and in the control Spec1mens, although the changes were greater in the surveillance specimens. Little or no coarse MgC precipitate formed in specimens of modified compositions, since the modifications were tai- lored to produce fine carbide of the MC type. All aged INOR-8 specimens showed near the surface a layer that etched more darkly. The amount of this modified layer did not vary systemati- cally with time or temperature. Experiments produced evidence that the effect was 1ike1y due to cold work from machining, causing carbide to precipitate more read11y near the surface. ' EXAMINATlON'OF MSRE 'COMPONENTS A major Objective of the postoperation examination of the MSREl9 20 was to supplement the findings on the INOR-8 surveillance specimens by examining pieces of INOR-8 from different parts of the fuel system. All parts had been at high temperature for 31 ,000 hr and in contact with fuel salt for 21,000 hr (about 1.5 times as long as any surveillance specimen) " Otherwise the conditions of exposure were quite varied. ‘l. A control rod thimble. (which had seen the highest neutron fluence of any INOR-8 ever examined) offered three kinds of surface exposure on the same piece. the inside exposed to,N2 + 2% 07 at 650 C, outside sur- face exposed to flowing fuel salt, and outside surface that had been under a loose-fitting INOR-8 sleeve. 88 '7_2. ‘Heat exchanger tubes had been exposed to fuel salt on the out- side and coolant salt on the inside. 3. The heat exchanger shell had been exposed to fuel salt on one side, cell atmosphere on the other. | | 4. The sampler cage and mist shield in the pump bowl extended from beneath the surface of the salt pool, through the surface (where certain classes of fission products tended to concentrate) into'a_gas space. The major part of the examination effort was on determinations of the amount and nature of dep051ts and metallography to reveal evidence of corrosion and surface cracking.rlAlthough the pieces‘were generally ~of awkward shapes, some tensile and'bend tests were run, and the examina- tion of the strained pieces proved‘to‘bermost interesting. Control Rod Thinble The MSRE used three controlrrods21 fabricated‘of Gd,03 and Al503 canned in Inconel 600. The control rods operated inside INORES thimbles made of 2-in.-OD X 0.065—in.-wa11 tubing. The assembly before insertion in the MSRE is shown in Fig.‘59. After being in service for several years (above 500°C for 30,807 hr), the lower portion of control rod_ thimble 3 was severed by electric arc cutting and moved to the hot cells for examination. Figure 60 shows the electric arc cut at the left (about at the midpoint of the core), the spacer sleeves, and the end closure (located at the bottom of the core) The thimble was made of heat Y—8487 - and the spacer sleeves were made of heat 5060 (see Table 1 for chemical 'compositions) . '7 The first cut was made through the sleeve and thimble nearest the electric arc cut. The spacer sleeve had been machined with ribs 0.100 in. high and 0.125 in. wide to position it relative to the graphite moderator. The sleeve had four drilled.holes through which weld beads were depsoited on the thimble to hold the sleeve in place. 'According to the shop'drawings,_the minimum and maximum diametral clearances.bef tween the thimble and the sleeve were 0.000 and 0.015 in., respectively. Thus salt would likely enter this annular region'and be in contact with most of the metal surfaces. ! { 89 . PHOTO TH10 Fig. 59. Control Rod Assembly before Insertion in the MSRE. R =-54116 Fig. Operation 60. Portion of Control Rod Thimble that was Examined After of MSRE was Terminated. The thimble is made of 2-in-0D x 0.065~- in.-wall INOR-8 tubing (Heat Y-8487). The electric arc cut on the left was near the center line of the core. 90 Undeformed Samples' ‘Samples of the control rod thimble and the spacer sleeve were cut and examined to determine their condition at the end of service. Typical phdtomicrographs_of the inside of the thimble tube are shown in Fig. 61. 'This surface was oxidized to a depth of about 2 mils by the cell environ- ment of nitrogen containing 2 to 57'02. The oxidation-process modified the microstructure to a depth of 4 mils, likely due to the selective re- moval of chromlum. _ ' Photomicrographs of one of thesweld beads are shown in Fig. 62, All of the surface shown was exposed to flowing fuel salt. Some dis- lodged grains are near the surface, snd'grain.boundaries are visible in the as-polished condition to a depth of 1 to 1.5 mils. Photomlcrographs involv1ng the interface between the thimble and sleeve are shown in Fig. 63. Figure 63(a) shows the annular region with a separation of about 7 mils and some salt present. Few surface irregu- larities are visible at a magnification ef 100x, indicating that they are considerably below 1'mii. Figfire 63(b) is a 500x view of the thimble and shows some surface cracks to a depth of 0.3 mil. Theloutside of the sleeve is shown in Fig. 63(c); a few grain boundaries to a depth of about -1 mil are visible. Additional photomlcrographs of the sleeve are shown in Fig. 64. The sleeve material does not appear to have received much working since the carbide is very inhomogeneously distribfited. The grain size is larger than usual awsy from the stringers. The inner and duter surfaces both hafie modified structures. A higher magnification view of the inner surface [Fig. 64(b)] shows that much of the modification is a high density of primary carbide. The outer surface [Fig. 64(c)] has a shallow layer of small grains, likely ‘due to a working operation. A second cut was made away from the sleeve, and typical photomicro- : graphs are shown in Fig. 65. This part of the thimble was exposed to flowing salt. Some of the grain boundaries are visible in the as-polished condition to a depth of about 4 mils and there is some surface modifica- tion [Fig. 65(a)]. Etching [Fig. 65(b)] delineates more of the grain strficture and shows the shallow surface modification. Fig. 6l. 91 = 0.007 INCHES 500X T R—54244 T i 300X 0.007 INCHES & 16 10.010 . I 0.035 INCHES 00K 10.030 in, - r Photomicrographs of the Inner Surface of the Control Thimble in the As-removed Condition. This surface was exposed to cell environment of N, containing 2 to 5% 0,. (a) As polished. (b) Etched. (c) Etched, typical microstructure. Etchant: Aqua regia. Reduced 30.5%. 92 e R-54238 In 0.007 INCHES 500X I I Fig. 62. Photomicrographs of Outér Surface of Control Rod Thimble - Showing the Weld Deposit Made to Hold the Spacer Sleeve. The weld surface was exposed to flowing salt. (a) As'polished.. (b) Etched: -Aqua regia. 93 R- 54245 St 100X - 0,038 INCHE S e e ooom 1 10,030 in. + Lie] ] I o) = n o o) o R-54256 T = 500x t——_.-w_.._._.__.—._'o_oo7 INCHES " o Fig. 63. Photomicrographs. of: As~-polished Surfaces of' Control Rod 1 Thimble and Sleeve. (a) Annulus between thimble and sleeve. - Thimble sur- face is in upper part of picture. - (b) Outside surface of control rod thimble, showing presence of some salt and shallow surface cracking. (¢) Outside sur- face of sleeve. Surface exposed to flowing fuel salt, 94 1 ‘R-54251" 10.040 in. 100X 136%™ - O.0C7 INCHES 500% 1) v~ = 7 INCHES soox Fig. 64. Photomicrographs of Control Rod Sleeve. (a) Etched view. showing general microstructure. Inner surface is on left and outer sur- face is on right. (b) Inner surface of sleeve. (c) Outer surface of sleeve., Etchant: Aqua regia. Reduced 30.5% i~ 0.007 INCHES 00X fon l— N 0.007 INCHES 00X fon . a Fig. 65. Photomicrographs of Control Rod Thimble where it was Exposed Directly to Flowing Salt. (a) As polished. (b) Etched: Aqua regia. 96 Miéroprdbe_scané wére'runlon the thimble sampiés that wefe taken frofi-gndEf]thé'sléevéhand outside the slegve.7 In the first case the thimble wall_was exPoéédAto almost static fuel salt, and no g:adients in iron and chromium concentration could be detected within ther3-um fegion of uncertainty near the surface. The sample oufside the sleeve that was exposed to flowing fuel salt was depleted in chromium to a depth of almost 20.um and in iron to a depth of 10 ym (Fig. 66). Thus the amount of corrosion that occurred varied considerably.in the two régions. The measurefiénts were not actually made, but a similar result likely occurred for the spacer sleeve. The inside surface was exposed to almost static salt and was likely not corroded detectably. The outside surface was exposed to flowing salt and likely showed iron and chromium depletion. Deformed Samples i The next step was to deform some of the thimble afid the sleeve to determine whether surface cracks were formed similar to those noted in the surveillance samples. Since the product was tubular, we used a ring test that is relatively quick and cheap. The fixture shown in'Fig;'67 was made by (1) cutting 5/8 in;'through a l-in.-thick carbon steelvplate with a 2-in.-diam hole saw, (2) cutting out the partially cut region with ample clearance around the hole, (3) cutting the plate in two along the diameter and removing 1/8 in. of material on each side of the cut, and (4) tapping a 3/4 in. diam thread into the two pieces. Then rings 1/4' in. wide were cut from the thimble and the sleeve. They fit into the groove and were pulled to failure with the resultant geometry Shown in Fig. 67. The initial loading curves include strain'aSsociatéd with the - ring éonforming tojthe geometry of the grip, so we cannot tell preéisely» how much the sample is deforming. Thus the yield stréngth and the elonga- tion are only relative numbers, but the ultimate tensile strength and reduction in area obtained from these tests are true values. Obviously, this type of test is deficient in giving good meChanidal property data - but is sufficient for.deformingrthe material and observing the incidenée of surface cracking. 97 - | ‘ ' ] — . R - 55205 | _ ‘ I . | _ . | o _ ' | ! ’ . 7 1 . | : ' 1 . : 20 L | ' . v -7—:- Edge exposed to flowing salt i o ' | - - . 15 L l/-\/\/\’\/\/____/ WMo 1 , S ' | _ 7 . - Z | o _ £ @m0} ' o E o & v | : £ _ o 31 _ - s | — Cr z 1 o 5 ;..2 ! o I - Fe -l 1 1 1 1 o -k Q 20 30 . 40 80 60 TRAVERSE DISTANCE — MICRONS 7 N Fig. 66. Electron Microprobe Scan of Sample from Control Rod Thimble. The thimble had been exposed to the fuel salt for 21, 040 hr and had been above 500°C for 30,807 hr. 98 Sleeve. A tested sample is shown on the left. ff3Fig. 672 Fixtufe'for’Testing Rings of Control Rod 'R-54166 Thimble_and Spacer 99 The tensile properties of the rings are shown in Table 8. The test results show the following important facts: 1. All of the unannealed specimens of heat Y-8487 tested at 25°C ‘have a ''yield stress' of 52,000 to 61,000 psi, with these values appearing to be random in the variables involved. 2. At 25°C the crosshead displacement was about 1. 2 in. for the unirradiated tubing and 0.4 to 0.5 in. for the irradiated tubing. Again the variations of 0.4 to 0.5 in. appeared random. 3. The yield stresses at 650°C were about equivalent for irradiated and unirradiated tubing. However, the crosshead displacement before fracture decreased from about 0.4 in. to 0.1 in. after irradiation. This was due to emhrittlement from helium formed from the 10B(n‘,a)7Li transmu- tation. - ) 4. No material is available of heat 5060 (sleeve material) for un- irradiated tests, but the vendoris certification sheet showed a yield stress of 46,300'psi,aan ultimate tensile sttess‘of_ll?,OOO psi, and a fracture strain of 522. The values that we obtained show higher strengths and lower ductility. Several of the specimens that had been strained were examined metal- 1ographically to determine the extent of cracking during straining. Several photomicrographs of the control rod thimble that was exposed to flowing salt are shown in Fig. 68. The inside surface was oxidized, and the oxide~oracked as the specimen was strained. The oxide should be .- brittle, but it is inportant”that'these cracks did not penetrate the ' metal. The side exposed to the salt exhibited profuse intergranular - cracking., 'Almost every grain boundary cracked and the cracks generally propagated to a depth of one grain, although there are several instances where the crackvdepth exceeded_a_depth of one grain. Several photomicro— , graphs were oombined and‘rephotographed‘to obtain Fig. 69. This sample had 192 cracks per inch with average and. maximum depths of 5.0 and 8.0 mils, respectively. Many of the cracks shown in Fig. 68 and 69 have blunt tips, indicating that they did not propagate into the metal as the specimen was strained Table 8. Tensile Data on Rings From Control RodfThimble 3 (Heat'Y-8487) ' Yield Crosshead ~ Reduction Specimen Condition Postirradiation Test Crbsshead Ultimate - o - Anneal Temperature Speed Stress® Stress Travel in Area (°C) (in. /min) (psi) (psi) (in.) (%) 1 Unirradiated 25 10.05 52,000 114,400 1.20 44.5 2 Unirradiated 25 0.05 56,900 117,500 1.17 46.0 3 Unirradiated 25 0.05 58,000 124,300 1.18 43.0 4 Unirradiated 650 0.05 39,100 76,700 0.37 28.7 5 Unirradiated - 650 0.002 40,700 62,300 0.20 20.6 6 Irradiated None 25 0.05 54,400 105,100 0.54 23.2 7 Irradiated. None 25 0.05 53,300 = 102,500 0.51 29.7 8 Irradiated None 25 0.05 60,500 110,800 0.42 28.5 9 Irradiated None 650 0.05 - 38,300 51,200 - 0.099 12.1° 10 Irradiated ~ None 650 0.002 34,200 38,200 -‘0.0611 ‘ 9.7 11 - Irradiated 8 hr at 871°C 25 0.05 49,300 100,500 0.36 34.9 12 Irradiated - 141 hr at 871°C 25 0.05 48,700 104,900 0.39 3.4 13 Irradiated’ None 25 £ 0.05 51,700 98,600 0.45 32.9 14 Irradiated® None 25 1 0.05 42,400 123,000 0.20 18.0 Located under spacer sleeve. Spacer sleeve, Heat 5060. ZBased on. 0.002 in. offset of crosshead travei. 00T 101 ox fo.05 m. 1C.010 in. T C.038 INCHES 1GOX 10.030 in. + 10,040 in. 0.03% INCHES 100X Fig. 68. Photomicrographs of Control Rod Thimble Specimen Exposed to Flowing Salt and Tested at 25°C. (a) Fracture - dark edge is oxide on inmer surface of tube, and cracks formed on the outer surface; etched. (b) Cracks in outer surface of tubing as polished. (c) Cracks in outer surface of tubing etched. Etchant: Aqua regia. Reduced 30.5%. i il 2 T WALFEE R, ) wAE AT g g Ll T R 3 o " st AR Fig. 69. Photomicrographs of a Deformed Section of the Control Rod Thimble. The fracture is on the left. The upper surface was in contact with flowing salt and the lower surface was exposed to the cell environ- ment of N plus 2 to 5% 02. The sample is about 0.06 in. thick | R-s85182] <01 103 A similar sample was cut from the .thimble under the sleeve, where the salt access was restricted. Photomicrographs of a specifien'tested at 25°C are shown in Fig. 70. The cracks are quite sifiilar to those in Fig. 68 for a sample that was exposed to flowing salt. The composite ~ photograph in Fig. 71 is quite similar to Fig. 69 of the sample exposed to flowing salt. The sample in Fig. 71 had 257 cracks per inch, with average and maximum depths of 4.0 and 8.0 mils,.respectively. The accu- racy of our statistics and possible sampling inhomogenities lead us to conclude that the severity of cracking is equivalent in the samples exposed to flowing and almost static salt. Thus flow rate over the range represented by these two sample 1ocations does not eppear to be an important variable. | : ' | | The spacer sleeve gave enother.opportunity to examine whether a relationship existed between salt velocity and cracking. The photomicro- graph in Fig. 72 shows cracks on both sides, with more being present in the field photographed on the side where the salt was rapidly flowing. However, the composite photograph in Fig. 73 shows that the cracking is about equivalent on both sides. Cracking statistics on the side exposed to rapidly flowing salt revealed 178 cracks per inch with a maximum depth of 7.0 mils. The side exposed to restricted salt'flow had 202 cracks per inch, with a maximum depth of 5.0 mils. The.average depth was 3.0 mils on both sides. These observations again shcw that salt flow rate 1s not a significant factor in intergranular cracking., - Important Observations The examination of the control rod thimble and the'thimble.specer revealed two important charaetefist1CS_of the crackifig..(l) Irradiation_ of the metal is not responsible for the cracking. .The thimble was irrad- iated to a fieak thermal fluence'ofl.Q X 1021_neutrons/cm2. The'flux attenuation across the 0.065 in. wall of the thimble would have been very slight, but -cracks only formed:in the metal on the field side (Fig. 68). Thus' irradiation alone'does nct cause the crackiné, but contact with the fuel salt is required. (2) The severity of cracking is not sensitive to salt flow rate over the range experienced in a fuel channel and under a spacer with restricted flow. = 104. i i i i 0.035 INCHES 100X J== 0.039 INCHES X I 1 Fig. 70. Photomicrographs of Control Rod Thimble Exposed to Fuel RanEl L Salt Under a Spacer Sleeve and Tested at 25°C. - (a) As-polished view of - &h-j | : fracture. 40x. (b) As-polished view of cracks. (c) Etched view of - cracks. Etchant: Aqua regia. Reduced 30.5% | , i i 105 Fig. .71. Composite Photograph Showing the Density of Cracking Along the Edge of a Portion of the Deformed Control Rod Thimble. 9.5%. -The sample is 0.58 in. long. The fracture is on the left, the ‘upper surface was exposed to fuel salt, and the lower surface was exposed to the cell environment of N, plus 2 to 5% 0,. RN R — 56040 0.035 INCHES N 100X | ™) 'Fig.‘72. Photomicrog:aph‘§f a Section of Spacer Sleevé Tested at 25°C. As polished. ‘The upper surface was exposed to almost static fuel salt and the lower surface was exposed to rapidly flowing fuel salt. 106 Primary Heat Exchanger The primary heat exchanger was at temperature for 30 807 hr and oper- ated with fuel salt on the shell side and coolant salt in the tubes. The shell was 16-3/4 in. OD X 1/2 in. wall and constructed of heat 5068. ;The tubing was 1/2 in. OD x 0.042 in. wall from heat N2-5101 (see Table 1 for chemical composition). An oval piece of the shell 10 x 13 in. was cut by plasma torch near the outlet end of'the heat exchanger.‘ This same cut severed a piece of one tube, and pieces of five others were cut by an abrasive wheel. Photographs of these parts are shown in Figs. 74 and 75. The outside of the shell had a dark adherent oxide; the inside surfaces were discolored slightly and had a thin bluish-gray coating probably caused by the plasma cutting operation. Some of this powder was brushed off and found by gamma scanning tocontain;losRuand'125$b. Undeformed‘Samples' Photomicrographs of a cross section through the outer surface of the ~shell are shown'in Fig. 76. The oxide on the outside surface is typical of that obserued on INOR-8 at this temperature and extends to a depth of about 5 fiils; 'Etching attacks the metal in the oxidized layer and etches "the grain boundaries near the surface more rapidly than those in the in-- terior. The inside surface of the shell (Fig. 77) shows some structural’ modifications"to a depth of about 1 mil. No samples of the shell were deformed. o o Photomicrographs of a typicai.cross section of the heat exchanger tubing are shown in Fig. 78. The overall view shows that both sides etch more rapidly than the center to a depth of about 7 mils. This is not un- usual for a tubular product and is often attributed to 1mpurities (pri- ’marily lubricants) that are worked into .the metal’ surface during repeated steps of deformation and annealing._ The higher magnification views in the as-polished conditionhshow'that grain boundary-attack (or cracking)_ occurred on the surface exposed to fuel salt (outer) but not on the cool- ant side (inner). The tenperatures under normal operating conditions were such that the fuel salt side (OD) would have been in compression. | 107 Fig. 73. Composite Photograph Comparing the Density of Cracking of Sides of Spacer Sleeve Exposed to Almost Static Salt (bottom) and to flowing salt (top). 1llx, The sample is 0.52 in. long. R-54182 i 1 i i Fig. 74. Section 10 x 13 in. Cut from the Primé:y Heat Exchanger Shell. The piece was cut with a plasma torch, and the center stud was - used for guiding the torch. | I i | ! ! i i 1 i i 108 R- 54185 Fig. 75. The 1/2-in.-0D Hastelloy N Tubes from the Primary Heat Exchanger. The different shades arise from a dark film that is thought to have been deposited when the shell was being cut. The film was deposited on the side of the tubes facing the shell. 109 o/ . E— . —_—— PR ] - - 5475 o 1 l- in 0.007 INCHES = 00X Jen ™ 10.010 in. 0.035 INCHES 100X 10.C30 in. Fig. 76. Photomicrographs Showing the Outside Edge of the Primary Heat ' - Exchanger Shell. This surface was exposed to 2 to 5% 0O, in Np. (a) As v polished, showing the selective oxidation that occurred. (b) Etched view' showing that the metal in the oxidized layer was completely removed. Etched with aqua regia. 110 l-fi I~ Q.007 INCHES 500X o o JR-54762 ]—o o 0.007 INCHES 500 I ~ Fig. 77. Photomicrographs of the Inside Surface of the Primary Heat Exchanger Shell. The surface was exposed to fuel salt, and the modified structure to a depth of about 1 mil is apparent. (a) As polished. - (b) Etched with aqua regia. - ' 10.040 in. 0.035 INCHES 100X 0.030 in. 500% 0.007 (NCHES memnt—————— W | i L= o 0.007 INCHES S00% T o {c) Fig. 78. Photomicrographs of a Cross Section of ‘an’ INOR-8 Heat Exchanger Tube. (a) Etched with glyceria regia. (b) Inside surface in the as-polished condition after exposure to coolant salt. (c) Out- side surface in the as-polished condition after exposure to fuel salt. Reduced 34%. 112 'Similar'photomicrographs of-a iongitudinal section of tubing are shown in Fig. 79. The featuresAare qnite gsimilar to those discussed for Fig. 78. Both figures show some metallic deposits on the outside surface, which was exposed to fuel salt. These deposits are extremelj small and could not be analyzed with much accuracy with the in-cell microprobe. ‘However, they seemed to he predominately iron. Scanning across the tubing detected:no concentration gradients ofvthe major alloying elements in INOR-8. Deformed Samples _ | Tensile tests were run on three of'the tubes, and the results are compared in Table 9 with those for_as—receined tubing. The tubes were pulled in tension, and we assumed-that the entire.section between the grips was deforming. The grips offer seme end restraint, so our assump- tion is obviously in error, and the errors are such that our yield stresses are high and our fracture strains are low. The ultimate stresses and re- ductions in area are unaffected by this_assunption. The as~received and postoPeration tests were run by the same teehnique, so‘the results should rbe comparable. The largest changes during service were reductions in the yield stress,'fracture strain, and reduetion in area. The tubing was bought-in the "cold drawn and annealed" condition, but such tubing nor- mally is cold worked some during a final straightening operation. ‘The reduction in yield strength during service may have beenrdue to annealing‘ out the effects of‘this'working.The reduetions,in the ductility parane— ters were likely due to the precipitation of carbiaes along the grain boundaries. | - -Photomicrographs of one of the tubes deformed at 25°C_are shown in Fig. 80. Profuse intergranular cracks were formed to a depth of about. 5 mils on‘the side exposed to fuel salt, but none were formed on the coolant salt side. Etching again revealed the abnormal_etching charac- teristics near the surface, but whatever caused this etching characteris- tic did not result in grain boundary embrittlement. Unstressed and stressed sections of tubing were photographed over relatively large distances to get more accurate statistics on cracks Table 9. Tensile Data on Heat Exchanger Tubes (Heat N2—5101)A‘ Crosshead Yield - Ultimate Fracture Reduction. . o - - Speed Temperature Stress Tensile Strain in Specimen - Condition - (ip./min) - (°0) | (psi)fifgi Stress (psi) Az Area (%) 7 From heat exchanger 0.05 650 44,300 67,400 21.3 22.4 6 From heat exchanger 1.0 25 66,300 - 118,000 37.0 20.5 £ S From heat exchanger 0.05 25 64,800 122,000 39.0 29.0 b As received 0.05 650 53,900 74,600 40.0 14.5 2 As received 0.05 25 73,000 127,400 51.1 42,6 3 As received 0.05 25 70,900 = - 126,200 50.1 40.0 1 Vendor certification 25 58,900 - 120,700 47.0 - - | 114 ' R-54776 0.010 in. 100X in. 10. — e 0.038 INCHES oty 00T INCHES =ttty TR | - e [ ———— () N T T 5 o Fig. 79. Photomicrographs of a Longitudinal Section of an INOR-8 Heat | Exchanger Tube. (a) Etched with glyceria regia. (b) Inside surface in the ~ as-polished condition after exposure to coolant salt. - (c) Outside surface in the as-polished condition after exposure to fuel salt. Reduced 31Z%. ") - t~ < 0 | @ Ul OO0 X00L | S3HONI €00 vl €00 M 0004 X00F | SIHONI §£0°0 Tme Heat Exchanger Tube Defo (b) Etched with aqua regia. Photo (a) As poli d to Fracture hs of a crograp shed. mi 80 ig. C. F at 25° 116 numbers and depths. A composite photoéraph of the two samples is shown in'Fig, 8l. The cracks are readily visible in the stressed'sample,but_ may not be apparent in the unstressed sample. A higher fiaéaification print used in making the composite is shown in Fig.'82; the cracks can: - be.more easily seen. The stressed sample had 262 cracks per inch with; an average depth of 5. 0 mils. Two counts were made on the unstressed sample. In the first count only those features that unequivocally were cracks were counted. This gave a crack count of 228 per inch with an 'average depth of 2.5 miis. In another count, we. included some features that were possiBly,cracks and this gafie a crack frequency of 308 per inch. Thus the number of cracks present before straining was'about _ equivalent to that noted after straining. Straining did increase the . vigible depth. Some indirect evidence that the cracks were present initially and only spread open further during the tensile test was ob- tained from eXamination_of the sample in Fig.'81. The sampie length shown deformed 397 (Table 9), and the combined widths of the cracks "account for 35% strain. Thus, thercracks‘widths very7close1y account for the total strain and indicate that the cracks formed with 1itt1e or no deformation and spread open as the sample was deformed. Important Observations The examination of the heat exchanger tubes revealed three . very important characteristics of the cracking. (1) The neutron fluence _received by the tubing was extremely low, but intergranular cracks were formed. The control rod thimble showed that irradiation of the metal | alone was not sufficient to cause cracking since only the side of the - thimble exposed to fuel salt cracked. Examination of the heat exchanger | tubing'suggests the further conclusion that irradiation of the metal is not a factor, since the heat exchanger was exposed to a negligible flu- ence but still cracked. (2) Exposure to fcel salt is a necessary condi-g tion for cracking to occur. Only the outside of the tubing, which was | | exposed to ffiel selt, cracked, and the inside, which was exposed to cool- ‘ant salt, did not crack. (3) The cracks were present in the as-polished tubingras itrwas removed from the MSRE. Straining caused the cracks to | open wider without much penetration in length. 117 Fig. 81. Composite Photographs of Heat Exchanger Tubing Before and After Stressing. The width of the unstressed sample is 0.042 in. - [ o o N O o = 0.07 INCHE S & 50x N | lon oy Fig. 82. Photomicrograph of Heat Exchanger Tubing in the Unstressed Condition. The side with the small cracks was exposed to fuel salt and the other was exposed to coolant salt. I~ 118 Pump Bowl Parts A schematic view of ‘the pufip is shown in Fig.'83.':The components examined were the mist shield and the sampler cage; The.safipler was low- " ered by a windlass ’arrarigelfient into the’ sampler cage v_‘a',vnd’ was used primar- ily for taking salt samples for chemical analysis. The mist shield was provided to mimimiée the amount of salt spray that would reach the sampler. The vérticallsampler.cage rods were 1/4 in;‘diém‘and made of heat 5059. The mist shield was made of 1/8-in. sheet of heat 5057 (éee\Tablé 1 for ‘éhemical'analysis). The mist shield was afispiral with an inside diametér of about 2 in. and outside diameter of about 3 in. ‘Ihe_spiral was about -1 1/4 turns, and the outside was exposed to agitated salt and the inside to salt fldwing.much slower. The outside was exposed to salt up to the " normal liquid level and to salt spfay above this level. The inside was exposed to salt up to the normal salt level and primarily to gas above this level. | 7 - The general appearance of the sampler cage is shown in Fig. 84. The salt normally stayéd at a level near the center of the cage, Bfit the level fluctuated slightly. The amount of material deposited on the surface is much higherlbeldw the liquid interface than above. A gross gamma scan h@d a similar profile with a maximum near the liquid interface. A carbonaceous deposit was present at the top of the Sampler._ The mist shield had some deposité, and these were heavier below the liquid level and on the outer surface of the shield. Samplef'Cage | One of the sampler cage rods was examined metallograpically. Photo- - micrographs of samples from the vapor and liquid regions are shown in Fig. 85 and 86, respectively. There is no intergranular penetration visible at.this magnifigation..,A heavy surface deposit is on thg,sample from the liquid region. This deposit was not examined by the electron microprobe analyzer, but a similar deposit on a copper,cépsule_taken from ", SEAL OIL LEA DRAIN LEAK SAMPLER ENR (Out of Section) {See Inget OPERATING - LEVEL e A2 decsasenmn e Fig. 83. Section of $ \/ SHAFT TER N COOLED \i \ MOTOR SHAFT SEAL : , {Ses Inset) LEAK LUBE QIL BALL BEARINGS LUBE OIL OUT BUBBLE TYPE LEVEL INDICATOR To Overfiow Tonk 119 OANL-LR-DWG-56043-8F 2 OIL BREATHER BALL BEARINGS {Face to Face) HOUSING GAS PURGE IN L} (8ack to Bock) - SHAFT SEAL {See Inset) SHIELD COOLANT PASSAGES {in Parailel With Lube 0il) SHIELD PLUG GAS PURGE OUT (See Inset)— KAGE GAS FILLED EXPANSION SPACE STRIPPER (Spray Ring} SPRAY ICHER MSRE Fuel Pump with Details of Several Areas. The mist shield and the sampler cage that were examined are shown in the lower left insert. Fig. 84. M ; Sampler Cage. interface corresponds with the region of highest material deposition. The assembly is 8.5 in. high. The salt-vapor i ! i i i { i i i i Fig. 85. Located Above aqua regia. 120 R-54657 = 0.035 INCHES N 100X fen 10.040 in. 0.035 INCHES 100X 10.030 in. Photomicrographs of én INOR-8 Rod from the Sampler Assembly the Normal Salt Level. (a) As-polished. (b) Etched with ty 121 R-54651 M, IE bR ST 3 1 e SN L 0.035 INCHES N 100X T 10.040 in. 100X 0.03% INCHES 10.030 in, 1 INOR-8 Rod from the Sampler Assembly. Fig. 86. Photomicrographs of an Normally located below the salt level. (a) As polished. (b) Etched with '~ aqua regia. ' ' ' 122 the sampler region was examined The deposit contained some salt, but had regions that were high in the alloylng elements in INORrB.. Ni, _Cr, Fe, .and Mo. : ' i . | ',“ | | One of the sampler cage rods was pulled in tension at 25°C._ The certified properties of this material were a yield stress of 51,200 psi an ultimate tensile stress of 115, 300 and an elongation of 514. The: measured properties of the sampler cage rod were a yield stress of 42,500 psi, an_ultimate tensile stress of 93,400 psi, and an elongation of 35.7%. The property changes are not large and are likely due to some stress re- lief that decreased the yield and ultimate stress values and carbide pre- cipitation, which reduced the ductility. A composite of the tested sample _in Fig. 87 shows that the fracture occurred below the center near the average liquid-gas interface. One metallographic sample was taken on the ;1iquid side of the fracture. A composite showing the badly cracked edges is shown in Fig. 88. Note ‘that the cracks extended to a depth of 12 mils. A higher magnification view of an area near the fracture shows that the cracks extend'in excess of 3 grains deep and that several grains have -fallen,out-(Fig;789) Another metallographic sample was taken 3/4 in. abouefthe fracture. A compOsite of photographs along the edge shows that the depth of cracking decreases with increasing distance into the . vapor region (Fig. 90). The severity of cracking is different on the two edges of the rod;. Unfortunately, we do not know the-orientation of the rod - relative to the sampler and the mist shield. Mist Shield , , - Figure 91 shows the mist shield after removal from the MSRE. The ‘shield has been split to reveal the inside surface. Four specimens approximately 1 in. (vertical) x 1/2 in. (circumferential) were cut from the mist shield. Their locations were (1) outside the spiral and immersed in'salt, (2) outside the spiral and exposed'to salt spray; (3) inside- the spiral and immersed in salt, and (4) inside the spiral and exposed primarily to gas. Bend tests were performed on these specimens at 25°C using a three point bend fixture.22 They were bent about a line parallel | | [ 123 Fig. 87. (a)vTensile—tested Samplef Cage Rod from Pump Bowl. Rupture occurred near the average liquid-vapor interface. (b) Photograph of rupture area showing extensive surface cracking. Rod diameter is 1/4 in, Fig. 88. Section’of Samp1e‘Cage Rod that was Deformed at 25°C. 9.2x. The fracture (left) occurred at the salt-vapor interface. The rod is immersed further in the liquid from left to right. The length of the sample is 0.63 in. 124 [ R -56030 0.035 INCHES N 100X | Fig. 89. Photomicrograph of Deformed Sampler Cage Rod Near the Fracture. As-polished. - Fig. 90.. Composite.of Photomicrographs of :Sampler Cage Rod that was Strained at 25°C. 8.2x. Left end of sample was 3/4 in. from fracture. The rod progressed further into the vapor region from left to right. The sample length is 0.72 in. ' 125 s o ‘ Fig. 91. In'te'._r:;i.or of Mis_t Shieldj,‘ Right part of right segmentr over- - lapped left part of segment on left. ' ' S , . 126 to the 1/2 in. dimension and so that the outer surface was in tension. The information obtained from such a test is a load-deflection curve. The equations normally used to convert thisjto a stress-strain curve do not take into account the plastic deformation of the part, and the stresses 'dbtalned by this method become progressively in error (too hlgh) as the ~ deformation progresses. ‘ "Table 10 shows the results of bend tests en the specimens from the mist shield. Note that the’yield and ultimate stresses are about double those reported for uniaxial tension. The fracture strain is the parameter ~of primary interest. The test of unexposed INOR-8 (heat N3¥5106) did not fail after 40.5% strain in the outer fibers. (No material was available of heat 5075 the material used in fabricating the mist shield ) Although all of the samples strained more than 10% before failure, the two samples - from the outside of the-spiral'were more brittle than the other samples. ‘The bendrsampies of the mist shield were exemined'metallographically. Photomicrographs of the sample from the inside vapor region are shown in Fig. 92, and a composite of several photbmicrbgraphs is shown in Fig. 93. The edge cracks were intergrsnular and about 1 mil deep. The photomicro- graphs of the sample from the liquid region (Fig. 94) show that.the cracks extended to a depth of about 8 mils in the liquid region. The composite photograph in Fig. 95 also shows the increased frequency and depth of ~cracking in the liquid region. . _ The sample from the outside 1iquid region was quite similar metal- lographically to that from the inside liquid region. Photomicrographs of the fracture and a typical region of the tension near the fracture are shown in Figs. 96 and 97, respectively. A composite of several photomicrographs is shown in Fig. 98. Photomicrcgraphs of the outside vapor (salt spray) ‘bend specimen at the fracture and on the tension side neer the fracture are shown.in Figs. 99 and 100, respectively. The com- posite of several photomicrographs in Fig. 101 shows that-the cracks are not much deeper nor more frequent than‘in the_liquid regions (comparerA Figs.'95,'985'and 101), but the tendency for grains to fall out is much greater._ This is even more significant when one notes that the strain: was the least in the sample from the outside vapor region (Table 10). Table 10. Bend Tests at 25°C on Parts of the MSRE Mist Shield? Maximum® Yield b Tensile - : Stress Stress Strain | | Specimen . - (psi) (psi) - (%) - Environment‘” | | S x 103 x 103 | ,_ Do 5-52 - Mist shield top inside 97 - 269 46.9d | VapQr'region, shielded S-62 Mist shield top outside = 155 224 10.79 Vapor region, salt spray S-60 . Mist shield bottom inside 161 292 31.6d' Liquid region, shielded,' L - o o | salt flow ~ S-68 Mist shield bottom outside 60 187 17.7d Liquid region, rapid 4 | o o o o o salt flow.. N3-5106 Unirradiated control’ 128 238 40.5 test, 1/8 in. thick %The mist shisld was fabrlcafied of heat 5075 with certified room-temperature pfoperties of 53 000 psi yield stress, 116, 000 psi ultimate stress, and 49Z elongatlon. bBased on 0.002 in. offset of crosshead travel. CMaximum tensile stress was controlled by fracture of sample or by strain limitation of test fixture. : : dSpecimen broke. 1T 128 0.035 INCHES X 0.035 INCHES 100X R-55993 ([T 0.035 INCHES X ~ Fig. 92. Photomicrographs of Bend Specimen of Mist Shield from Inside Vapor Region. (a) Fracture and tension side, as polished. (b) Fracture ; and tension side, etched. (c) Tension side, etched. Etchant: Aqua regia. | - Reduced 31%. : | | Side of a Bend Sample he Tension 93. from Inside Vapor Region of Mist S Fig. Composite Photograph of t not the compression Lower edge 1is hield. side of the sample. 130 0.038 INCHE § ettty - 1 ¥ 100K e srsne: 1 e e et N = 1 INCHEE 500X Q007 B T Fig. 94. Photomicrographs of Bend Specimen of Mist Shield from Inside Liquid Region. (a) Fracture and tension side, as polished. (b) Fracture and tension side, etched. (c) Fracture and tension side, etched. Etchant: Aqua Regia. ' ' from side Fng 95. Cofiposite Photograph of the Tension Side of a Bend Sample Inside Liquid Region of Mist Shield. Lower Edge is not the compression of the sample. | TeT 132 R- 56767 R-56769 I= 0.035 INCHES i 100X o e 10.001 in. i 10.003 in. ~=0.007 INCHES 10.005 in, 500X 10.007 in, Fig. 96. Photomicrographs of the Fracture of a Bend Specimen from the Qutside Liquid Region of the Mist Shield. Etchant: Lactic acid, HNO3, HCl. - (a) 100x. (b) 500x. " & L - X0 ed] ") | | w000 | Wwgoo0p X0O0S | ‘W G000 | w000 Lw S3HONI 6€0'0 ! (== - S3IHONI 200°0 - d . 56770 ‘(a) 100x. x "HC1. 133 HN03’ Photomiéfdgraphs of the Tension VS'ide Near the Fracture of a .the Outside Liquid Region of the Mist Shield. o L o 3] o s ES O 3] - 8 w o g Moo - o 4] ont ~ oK o H o [ & ] - o0 Q X - O = wn O : i o a8~ Q .0 m - 7T Fig. 98. Composite of Several Photomicrographs of a Bend Specimen from the Outside Liquid Region of the Mist Shield. Top portion shows the tension side and the lower portion shows the compression side. 24x, o 135 0.035 INCHES N 100X T 0.007 INCHES 10.004 in. 10,003 in, | 500X 10,005 in. Fig. 99. Photomicrographs of the Fracture of a Bend Specimen from the Outside Vapor Portion of the Mist Shield. (a) 100x. (b) 500x. Etchant: Lactic acid, HNO;, HCL. : 10.007 in. 136 R-5e7ss] 0.035 INCHES 10.004 in. b 0.007 INCHES S00X I {0005 =™ T 10.007 in. Fig. 100. Photomicrographs of the‘Edge Near the Fracture of a Bend Specimen from the Outside Vapor Portion of the Mist Shield. (a) 100x.. (b) 500x. Etchant: Lactic acid, HNOj3, HC1. 16.663 m. 137 The reduced ductility of these samples remains uneXplained The maximum crack: depth in the liquid region was only 8 mils, and it is quite unlikely that an 8 mil ‘reduction in a sample thickness of 125 mils would cause much embrittlement. The embrittlement may have been partially due to the extensive carbide precipitation shown in Figs. 93, 95, 96, 98, and 99. However, the outside and inside of the shield received identical thermal treatments, but the samples from the outside had lower ductilities. Thus, attributing the embrittlement solely to carbide formation does not seem appropriate. Important Observations The only free surface in the primary circuit was in the pump bowl, and there appears to have been a collection of material at the surface. Visual observation, gamma scanning, and the electron microprobe show an accumulation of salt, fission products, and corrosion products at this ‘surface. Besides having a free surface, the region around the'sampler was often quite reducing because of the addition of beryllium metal at this location. Fluorides of the structural metals would have been reduced . to the metallic form and accumulated at the free surface or ‘deposited on nearby metal surfaces. Some of ‘the less stable fission product fluorides ‘would have 1ikely reacted in a similar way. Thus, it is not surprising that the intergranular cracks are most severe near the free surface. The most important observation relative_to'the cracking is that its severity - was much less in material-exposed primarily to gas. This allows the con- . clusion that exposure torliquid salt is necessary'for the cracking to occur. Freeze Valve 105 Freeze valve 105 failediduring the final thermal cycle involved with terminating operation of the MSRE. The failure was attributed to fatigue from a modification.23 This valve consisted of a section of 1 l/2-in. " sched. 40 INOR-8 pipe (heat 5094) with a jacket for air cooling. It was used to isolate the drain tanks and was frozen only when salt was in the 138 drain tanks. The line was filled with salt from the drain tanks, so the fission product concentration would have been relatively low. The line was filled with salt and above 500°C for about 21,000 hr. The salt was also statlc except when the system was being filled Thus, th1s partic-‘ ular component was subjected to a unique set of conditions. Rings 3/16 in. wide were cut, away from the flattened portion of the pipe. They were subjected to the ring test described previously. A bend specimen and a.specimen'for chemical analysis were cut from adjacent regions. The results of the mechanical property tests are given in Table 11. As shown in Fig. 102 both the bend specimen and the ring specimens showed some surface cracking when viewediat low magnification. A metallographic section of one of the ring specimens is shown in | Figs.‘103‘and 104. The fracture and both edges are visible in Fig. 103. The oxide is on the outside surface that was exposed to air (top) and shallow cracks formed on the salt side. Typical edges near the fracture are shown in Fig. 104. The cracks on the air side followed the oxide and did not.penetrate'further. The cracks on the salt side were intergranular and penetrated about 1 mil. | h | | The most important observation on this component is that the cracklng was less severe under conditions where the flssion product concentration was less. However, intergranular cracking occurred on the surfaces that were eiposed to fuel salt. The corrosion'(selective removal of chromium) ‘that occurred under these conditions should have been extremely small since the salt was static most of the time. Thus, corrosion does not seem to be a requirement for intergranular cracking although it may ac- celerate the process. EXAMINATION OF INOR-8 FROM IN-REACTOR LOOPS The_obserVation of intergranular cracking in the MSRE caused us(toh reexamine the available results on three in-reactor loops. Two pumped -loOps were run by Trauger and Conlinzu 25 in 1959, but the material from these 100ps had been discarded and only the reports and photomicrographs remained for examination. 1A:more recent thermal convection loop was run Table 11. :Resnits of Mechanical Property Tests on Specimens From Freeze Valve 105 o (Heat 5094 at 25°C and a Deformation Rate of 0.05 in./min Yield | Ultimate . Crbéshead Reduction | . o Stress " Tensile Stress Travel = in Area Type of Test o (psi) _(psi) o - (in.) A7) Vendor's, tensile 45,800 . 106,800 | 52.6 ‘Ring, temsile - - 45,800 | 89,700 0.72 25 Ring, tensile 48,900 90,100 ©0.59 | 29 Ring, tensile 41,900 90,300 0.73 37 Wall_'.segm'ent',be_nd 71,300 o o 0.41 332 'éMaximum strain in outer fibers. 6€T 140 Fig. 101. Composite of Several Photomicrographs of a Bend Specimen From the Qutside Vapor Region of the Mist Shield. Top portion is the tension side and the lower part is the compression side. 15x. Fig. 102, The Surfaces Exposed to Salt in Freeze Valve 105 After Deformation at 25°C. (a) Fracture of ring specimen pulled in temnsion. Note surface cracks near the fracture. 4x (b) Surface of bend specimen. Note some cracks on surface and edge cracks. 7x. Reduced 18.5Z. i [ | | ! | I | I : i ; 141 R-56857 'R—56865 Fig. 103. Photomicrographs of the Fracture of a Ring Specimen from Freeze Valve 105 that was Deformed at 25°C. (a) As polished. (b) Etchant: Lactic acid, HNO3, HCI. . 40x. Reduced 29.5%. ; 142 R-56853 B To.oatn. 10.003 in. b 0.007 INCHES S00X 1 1~.005 . 1C.007 in. R-56856 [ ~10.001 in. 10.003 in. Q.007 INCHES S00X ~ 1F.005 . I I6.067 in. |- 1506 1 i 10,003 . 0.007 INCHES S00X Fig. 104. Photomicrographs Near the Fracture of a Ring Sample from ’ | Freeze Valve 105 Deformed at 25°C. (a) Edge exposed to air, as polished. P (b) Edge exposed to salt, as polished. (c) Edge exposed to salt, etched : : - with lactic acid, HNO3y, HCl. Reduced 307%. ' \ - 143 by'Compere et al.,?® and some of the pieces were still available and their location in the loop identified. Some samples of these pieces were de—-h formed and examined metallographically. PumE LooEs Trauger and Conlin ran two INOR-8 pump loops in the MTR.2%525 These operated at a peak temperature of 704°C at an average power density of 66 W/cm3. The loops both contained salt of composition7 LiF-Ber-UFq (62f37-1 mole %). The first loop ran at a peak Reynolds number of 4100 and operated for 638 hr. The second loop had a peak Reynolds number of 3100 and operated for 766 hr. The operation of both loops was termin- ‘ated by leaks in the heat exchanger. A detailed metallographic examin- ation was made of the first loop. 27 The loop had a nose cone that was closest to the reactor, and it was here that the type of attack shown in Fig. 105 was noted. The attack is actually a surface roughening in which grains were removed‘from the inside of the INOR-8 tubing. The second loop did not show this type of attack.28 The leak in the heat exchanger was not located in either loop The fission density in these loops was 66 W/cm » with salt volumes of 135 cm3 and surface areas of about,650 cm? each.- One of the fission - products that will be discussed further_is tellurium, and a‘comparison of the amount produced in thesezloops‘nith that produced in'the MSRE is useful. About 1l.4 X 1017 atoms of Tellurium were produced per unit of . metal surface area in the MSRE and only 0.2 x 1017 atoms/cm in these two loops. The time at temperature was also much smaller for these 1oops than for even the first group of surveillance specimens from the MSRE, in wh}ch cracking was hardly detectable (Fig. 12). - The uneven inner surface: of the loop tubing may or may not have been ‘due to the same phenomenon that caused the intergranular cracking in the MSRE. Note in Fig. 105 that the material was removed 1n increments of single grains. However, there is no evidence of partially cracked grains that had not been completely removed. ‘The explanation officavitation is also not very acceptable since the peak velocity was only about 1 ft/sec. 144 o ] 0.07 INCHES 50X I~ 0.008 in. R- nges Fig. 105.. Inside Tube Wall of INOR-8 from In-pile Eoop MTR 44-1. J-}The_coating is nickel plating used for edge preservation. A 145 Being able to deform a piece of the tubing and then examine it for cracks should be enlightening, but no material from either loop remains. Thus, the observations on these loops are not very helpful in the present anal- ysis. Thermal Convection Loop A more recent in-reactor thermal convection looptfias run by Compere et al.?6 A schematic of the loop is shown in Fig. 106, and the pertinent operating statistics are given in Table 12. The loop was constructed of INOR-8, contained salt of compostiion 7LiF-BeF2—ZrFq-UF4(65.3-28.2-4.8—1.7 mole %), and had a graphite region where power densities of 150 W/cm3 of salt were attained. Tne loop had maximum and minimum operating‘tempera— tures of 720 and 545°C, respectively. The loop failed at the hottest - portion after 1366thr of nuclear operation. The failure (Fig. 107) is "quite typical of high~tenperature failures noted previously in this ma- terial after irradiation.“~7 Note that the inside surface is free of cracks except very near the point of failure. Two pieces of the loop were retrieved and tested. One sample was from the flat sheet used to fabricate the top of the "core section' where the graphite was located A small sheet was cut, bent so that the sur- face exposed to the fuel salt was. in tension, and examined metallograph- ically. A typical photomlcrograph is shown in Fig. 108. Surface cracks were very infrequent and extended to a maximum depth of 0.5 mil. A piece of tubing fias available from the coldest section, and it was deformed by crushing in a tensile machine. A typical view of the inside surface is shown in Fig. 109. Cracks occurred along almost every grain boundary, but they only extended to a maximum depth of 1 mil. The amount of tellurium produced in this loop per unit area of metal surface-was 4.2 x 1016 atoms/cm,r(reference 29), compared with 1.4 x 1017 atoms/cm2 in the MSRE Thus, the amount of tellurium was one—fourth that in the test loop, but the system was above 400°C only 937 hr when fission products were present. Table 12. Sfimmary of Operating Periods for In-Reactor Molten-Salt Loop 2 Operating Period (hr) Total Irraaiation Full Power Dose Equivalent Out-of-Reactor | Flush - 77.8 Solvent Salt 171.9 . In-Reactor Preirradiation 73.7 Solvent Salt - | 343.8 339.5 - 136.0 Fueled salt 1101.9 937.4 547.0 Retracted-fuel removal® 435.0 428.3 11.2 Total -~ 2204.1 1705.2 694.2 aMaintained at 350 to 400 c (frozen) except during salt-removal operations and fission product leak. investigations 9oHT 147 ORNL-OWG 67-11832 UINE SALT LEVEL OUTLET PIPE S w e seremen aAFFLE GAS SEPARATION TANK : INCONEL COOLING THERMOCOUPLE WELL V/q-in.~diom FUEL CHANNEL (8) 5% in. RETURN LINE {COLD LEG) SALT SAMPLE LINE 5 Fig. 106. Diagrém of Molten-Salt Iq—Reactor_Loop-Z.' R- 36583 F.ig. 107.- Photomicrograph of a Cross Section Showing the Inner Surface and Crack in Core Outlet Pipe of In-Reactor Loop 2. Located on top side of tubing. Etchant: Aqua regia. N CORE BODY & I I~ = fn = = 250% o i i i i : i R-56010 [ | L . . X ; gx a8 : 3 SR - ? | _ Fig. 108. Bend Specimén from Section of In-Reactor Loop Which T ; Operated at 720°C. The tension side was exposed to the fuel salt. . l—.— ! - 1, 2h w|2 alz, . o ; 3 | | ' Fig. 109. Tension Side of Crushed Tubing from Coldest Part of | In-Reactor Loop, Which Operated at 545°C. ' R 149 Summary of Observations on In-Reactor Loops The loops just discussed were exposed to fission product concentra- tions much lower than those observed in the MSRE., The time at tempera- ture after fission products were produced was also much less than for the MSRE. The loop times were all less than 1000 hr, and the first group of MSRE surveillance samples was'femoved'aftEr'ZSSO hr exposure (Table 3) to fission products at elevated temperatures. The number of cracks in this first group of surveillance samples was quite small (Fig. 12, Table 7). Thus, it is questionable whether these loops had adequate exposure to cause detectable cracking. The-loofis run by Trauger and Conlin had some regions from which grains were removed, but no intergranular cracks were visible. Strained samples of Compere's lobp had shallow intergranular cracks similar to those noted in samples from the MSRE, particularly the sémple exposed at 545°C (Fig. 109). | . CHEMICAL ANALYSES OF METAL REMOVED FROM THE MSRE We fdund the‘electrbn microprobe analyzer to be useful for measuring chromium gradiénts in INOR-8, but we'wefe not successful in locating any fission proddcts. We then used the technique of eléctroljtically removing surface layers and analyzing the solutiofis."Specifically, the method involved a methanol-30% HNog,solution at -15 to -20°C, a platinum cathode, 'énd the specimen as the anode. The electric potential was 6 V, a level that had been-shéwn,by,laboratory‘eXperiménts'to polish rather than etch.r The solutions were analyzed by two methods. The'coficentrations of the- stable elements were obtained by evaporating 2 cc of the solution into a mass spectrometer and then analyzing the vapor by mass—numbers: Numerous experimental difficulties were encountered in'taking the samples. Surface deposits (formed in the pump bowl) and thin oxide films . (found on many parts from the core) acted as inhibitors and made the methanol-30% HNOj solution attack nonuniformly or not at all. The voltage usually had to be increased to attack these surface barriers. Thus the 150 removal was not uniform on most samples until a fefi_mils had been removed. A second problem was in sample preparation. The round geometry of the - surveillancersample and the heat exchanger tube was desirable, but strips had to be cut from components suéh és the control rod thimbie and the . mist shield. We were not successful in masking the surfaces not exposed to fuel salt, so the ratios of fission product cbncentration to that . of nickel are lower than w&uld have been obtained had all the surfaces been exboSed to fission products. | Two of the better profiles are shown in Fig. 110 and 111. The sample in Fig. 110 fias_a segfient.of heat‘Exchanger.tubing. Epoxy was cast in- ‘side to mask the surface that was exposed to coolant salt. The sample ifi Fig. 111 was a surveillance sample. We made diametral measurements with a micrometer, but these were not very accurate. We actually obtained the depth'meésurements by calculation from the measured nickel concentrations using the supplemental knowledgé that the alloy was 70% Ni and that the volume of the solution was 140 cc. A | Compere comparéd the amounts of fission firoducts that we actually found with the total inventory in the MSRE. He assumed that the fission products were deposited unifofmly on the total system metal area of 7.9 x 10% cm?. (Inclusion of the graphite sfirface area increases the deposition area to 2.3 x 10% em?.) The results of this type of analysis are summarized in Table 13 for the most highly concentrated fission pro- ducts. The various samples are listed in the order that théy.wquld occur around the primary circuit, beginding at the bottom of the core and pro- gressing up through the core, through the pump, into the heat exchanger, and BaCk to the core. The values for the second sample should be quite | low because of the shallow sampling, and the samples from the sarveillance - specimen and the heat exchanger tube shouid yield the most accurate date. . However, there seem ‘to be many inconsistencies and no sysfiematic variation of the elements around the circuit. | One of the most interesting sets of chemical results was obtained from a surveillance éample of heat 5058 from the fourth group. The -sample was oxidized for a few hours in air at 650°C before beingrstrained. A very thin oxide film was formed and should have‘acted as a barrier to 151 ORNL-DWG 71- 7106 HEAT EXCHANGER TUBE { 2 .3 100 = . ° _ o—t-0 .---—-Ni._ g T T " M(? | 0" bgm—pa——omre—1=Cr = o o—z— Fe 10-2 = . —~ - *—tre——e— Mn .Te O . Y I THITHIT | /,’/ . o - o 6l H T L4t 1 1 LLiid L LElitl L1 LU L iul T TTHTIT b d// he) p = 3 1oL EELL .6' A P T T TTTEN / 4 yd 7 | / of o/o * o B __-.___-a 2 45 Cooonl 07 CONCENTRATION RELATIVE -TO Ni - d o 0° B “1ppb —= ~10 - \o o 1 _ . O TS PONb S o { =2 3 4 | DEPTH (mils) Fig. 110. "Concentr'atibn'_ Profiles from the _' Fuel Side of a MSRE Heat Exchanger Tube. Arrows indicate that analytical values were reported as being less than the indicated point. = | ' | . 152 ~ ORNL-DWG 72-2934 163 CONCENTRATION RELATIVE TO Ni 166 16 ——|——=110ppb 169 0 B 2 3 DEPTH (mils) Fig. 111. Concentration Profilesxfrom a Surveillance Sémple Exposed in the MSRE for 19,136 hr. Arrows indicate that analytical values were reported as being less than the indicated point. _‘:Tdble 13. Concentrations of Several Fission Prodficfs on the Surfaces B - of Hastelloy N, Compared with the Total Inventory Sample Location | 'Pzzgtgagiop ___Concentration of Nuclide Compared with Inventory R (mils) 277, 134gg o 125g, 103y 106 Ry 955, 997 Control rod thimble (bottom) 2.4 0.43 0.84 0.85 0.40 0.13 0.37 0.32 Control rod thimble (middle) 0.1 0.14 0.24 0.35 0.15 0.1 0.26 ~ 0.19 Surveillance specimen 3.5 0.001 0.35 1.04 0.006 0.087 0.006 0.30 Mist shield outside, liquid ~ 6.0 0.23 0.035 ~ 0.74 0.069 0.10 0.067 0.19 Heat exchanger shell 4.2 0.35 . 0.017 0.68 0.027 0.05 0.085 0.27 Heat exchanger tube 4.3 0.67 . 0.006 1.13 0.028 _ 0.14 0.070 0.3l €ST 154 chemical dissolution. The sample was deformed about halfway to failure, and two electrolytic dissolutions were made. The material should have been selectively removed from the_regioné that were freshly cracked since these surfaces were,not‘oxidiied. The solutions were analyzed, and the results divided By the nickel content."Significant enrichments were found in Te, Ce, Sb, Sr, and Cs. These solutions were afialyzed for sev- eral other elements that had not been defietmined previousiy. Uranium-235 was present at a concentration of 37 x 107% in the second layer; These concentratibns cofrespond to 26 and 7 ppm,_respectiVE1y. Determinations vere made of Al (0.23), B (0.01), Co (0.23), K (0.05), Na (1.2), P (0.23), V (0.23), and S (1.2)-w1thithe amounts shown in pafenthesis being the concentrations in percent found in the outer layer. The inner layer showed very little change. Samplés of distilled water and unused meth- anol-30% HNOj had eXtremély,low 1eveis of the élemen;s, and we know of no source of contamination in the safipling'operation. The high concentrations of sulfur and phosphorus are indeed reasons for concern, since these elements are known to embrittle nickel-base | alloys.30 Haubenreich calculated that the total sulfur introducedvby- pump oil inleakage was 27 g and fihat the initial fuel Charge contained <5 ppm S or a2 maximum of 24 g. Thus, sulfur was in the éystem and it may héve concefitrated along the grain boundaries. The sulfur (and the other elements) also could have been impurities,in the INOR-8 and segre- gated in the grain boundéries during the long exposure at 650°C. Another group of specimens waé collected, and a different approach was taken to the analysis of the concentration of fission pfoducts.‘ These samfiles were first exposed to a solution of Versene, boric acid, and citric acid. As shown in Table 14 this solution is not very aggressive toward the metal and very small weight changes were noted. Thus, this ' solution should be enriched in the material that was on or very near the surface of the metal. The sample was then completély dissolved in a separate solution. Both solutions were counted for various fission products, and the results are given in Table 14. Several observations can be made from thesé data. w Table 14. Concentrations of Fission Products'round on Several Samples From the MBREa-b . Surface Weight Weight Surface Concentration (atoms/cm?) Sample Description Area Before - After K - - - P : ’ ‘ - (em?) Leaching Leaching 127mpg 125gp,. _ ?OSr 137¢5 - 134¢cg . i 144 e ‘ 1°§Ru N 7e - (g) (g ) ‘ ' D o : : Spacer sleeve, smooth 12.2. 2.45489 2.44200 3.5 x 1011 1.5 x 1013 2.0 x 10!* 5.5 x 1013 6.6 x 101! 8.9 x 101! ¢ 1.6 x 1016 - o - 9.6 x 1012 4.6 x 101* 2.1 x 10} <4.8 x 10}* <5 x 1013 3.1 x 10'* 9,4 x 1016 Space sleeve, with rib 8.4 5.36317 ° 5.33873 2.2 x 10!} 4.0 x 10'? -~ 3.3 x 101" 5.0 x 10! 1.8 x 101! 1,8 x 1012 ¢ 2.3 x 1018 ' T 2 x 1083 <15 x 10150 5.8 x 101 <3.5 x 10l% <1 x 1013 <5 x 1013 6.7 x 101" 2,4 x 107 Thimble under spacer 6.6 3.80443 3.78641 3.1 x 101} 1.8x 1013 3.6 x 101" 7.5 x.10!3 2.1 x 10 1.7 x 1012 8.6 x 1012 -1.7 x 106 o : S 5.9 x 10} <3 x10* 1.8 x 10!* <2 x 1015 <2 x 1013 <1 x 101% <2 x 10'* 2.2 x 10!7 Thimble under spacer 6.3 3.65780 3.64322 3.2 x 10!! 1.4 x 10!* 3.6 x 10!1* 9.0 x 10!3 1.5 x 1012 1.3 x 1012 c 1.0 x 1018 ' - ‘ | R 4.2 x 1012 <9 x 101% 1.7 x 101 <3 x 1015 <1 x 10l% <1 x 10M* <3 x 10!* 1.9 x 1017 Thimble, bare 4,5 2.59565 2.59272 - 3.6 x 101} 4 x 1012 - 7.5 x 1013 3.9 x 10! 7.8 x 10}! <5,9 x 101! 6.9 x 102 4.2 x 10!5 - : : o 1.8 x 1013 <1.5.x 1015 1.4 x 101 <1 x 1015 < x 1013 <3 x 10} 1.1 x 1017 Thimble, bare 6.1 3.55622 3.55265° 1.5 x 10} 3.7.x 1013 5.9 x 1013 4.0 x 1013 4.8 x 10! 2.7 x 101! 6.4 x 1012 2 x 10! ) ’ 1.6 x 1013 <2 x 10!5 1.1 x 10M* <8 x 1015 <1 x 1013 ... <5 x 10 9.4 x 1018 Foil on fourth group- 1.2 '0.02047 0.02012 9.3 x 1011 5,6.x 1013 1.1 x10!3 2.6 x10}3 <5 x 10!l 4.9 x 1013 1 x 1018 surveillance \ L6 x 1013 <7 x10! 2.8 x1013 2,6 x 1015 <1 x 1013 <8 x 101% 2.7 x 101* - 1 x 10!8 Strap on fourth group 1.3 0.11191 0.11178 1.6 x 1011 1.6 x 1013 6.4 x 10}2 4.0 x.1012 <2 x 101! 1.1 x 1013 4 x 1013 surveillance . : 4.9.x 1011 2.2 x 101* 2.6 x 10'3 3 x 101 <2 x 1013 <1 x10!3 9.4 x 10'F 2.4 x 1016 Freeze valve 105 6.2 10.2255 ° 6.0 x 10° 7.5 x 10} 2.6 x 10}3 4.1 x 1013 <9 x 1010 6 x 1010 8.6 x 1012 6 x 101* ; 2.0 x 10} 9,1 x10}2 ° 1.0 x1012. 5.9 x10'% <5 x 10!0 6 x 1010 7,9 x 1012 5 x 1ol Concentration at end 8.9 x 101 1.4 x 10'* 2,5 x 10}7 1.8 x 107 6.6 x 1016 4,7 x 1015 2,7 x 1017 of operation ’aSamples counted about 2 years after end of MSRE operation. The foil and strap samples were removed 6 months before operation was terminated. The half-lives were 109 days for 127mpe . 2.7 years for 125gph, 28 years for 29sr, 30 years‘for'137Cs, 2.3 years for 139Cs; 284 days for 11+"Ce, 1 year for 108Ry, 5 x 105 years for 29Te. ' S bThe sémples were first leached in a solution of Versene, Boric acid, and citric acid. This solution was analyzed, and the first number under each isotope is the result. The remainder of the sample was dissolved and the second number under each isotope is the analytical result on the dissolved gsample, o . ) ‘ ®Present in, particulate form, but not in:splution. GST 156 1. Generally the concentration of fission products per unit surface ‘area exposed to fuel salt is greater in the metal than near the surface. However, the difference varies considerably for different isotopes.. For example, the 127Te concentration is abput,two orders of magnitude higher in the metal rhan_on-the eurface,'whereas gy is onlyrslightly concen- . ‘trated beneath the surface. | - ' ' 2, - The thimble samples that were exposed to flowing salt (desig- nated "bare" in Table 13) and those that were covered by a spacer allow some comparison of the effects of flow rate ofi the depesition_of fission Aproducte; The sample exposed to restricted salt flow consistently has a lower concentration of fission produets,’but only by a factor of 4 or ‘less. | a | ‘ ' -3, Freeze velve.105;(FV-105) had a lower concentration of fission products. This was expected because of its operating conditions (p. 30b) and is consistent with the observation rhat intergranular cracking'wae less severe in this component (Table 7). ' 4. With due consideration of the half lives, appreciable fractions of the total inventory of 127Te_, 1'ZSSb, 106Ru,_and 997¢ are present on the metal surfaces. This generally agrees with the indications of the data in Teble 13 except for the behavior of 134cs, Some of the 1ncrementa1 dissolution data in Table 13 indicate that large amounts - of 13%Cs were present, whereas the data in Table 14 do not support this observation. These results are interesting but must be viewed with reservations. The technique of electrolytically removing sections in the hot cells has not been used previously (to our knowledge) at ORNL, and numerous experimental difficulties arose. Uneven removal of material and deteri- oration of contacts, leads, etc. were some of the main problems. The results for radioactive species were obtained by proven methods. -However, the level of gamma radiation from ®9Co often masked the activity from elements of 1fitereSt. The technique of.evaporating 2 ml of the solution into the mass spectrograph is not new, but the complexity of the spectra - presented caused many problems in interpretatiofi. The different species are identified only by mass number by this method, and the preeence of L U AL S At a7 T A 157 the alloying‘elements in INOR-B, numerous fission products, constituents of the salt, and possible compoundsrbetween these elements and the elec- trolytic'solufiion made it difficult to interpret the patterns. The method used to obtain the data in Table 14 involved only radio- chemistry and dissolution techniques that were better established. How- -ever, the chemical reactivity'of the first solufiion (leach) with the . surface material leaves an uncertainity whether the materials removed were simply salt residues or small amounts of the metal. The surfsce areas were uncertain in most cases because of the complex geometry. With the qualifications that have been made the chemical analyses indicate several important'points. : 1. Several of the fission products penetrated the metal to depths of a few mils (Flgs. 110 and 111). 2. The fission products Te, Ce, Sb, Sr, and Cs were concentrated in the cracked regions of a ‘strained survelllance specimen Sulfur and phosphorous were also concentrated in these same regions, It is possible that the segregation of sulfur and phosphorus to the grain boundaries in this alloy is a normal phenomenon. | 3. Slgniflcant fractions of the total amounts of Te, Sb, Ru, and Tc were deposited on the metal surfaces. These experiments were good introductions to the types of studies that could be performed. More work was needed to develop confidence in them, but termination of operation of the MSRE stopped our'best’souree of,ekperimental material.‘ The full significance of the cracking problem ‘was only fully realized in the_postoperation examination of the MSRE. DISCUSSION OF OBSERVATIONS ON INOR-8 FROM ) - THE MSRE AND IN-REACTOR LOOPS '_Summary ofVObservhtions Observations have been pfesented on three basic types of samples.' 'The'firSt samples,wete the surveillance samples that were'present in the MSRE primarily for the purpose of following the corrosion and radiation 158 damage to the metal. The second set of samples consisted of several components from the MSRE that were removed for examination after termi- ‘nation of its operation. The third set of samples came from in-reactor ~pump and thermal convection loops. Several importent'observations were made and these will be diécussed,f These observations will then be used to propose some mechanisms that may be responsible for the cracking. The surVeillence samples included two heats of material that were carried throughout the program‘end'removed‘periodically for examination and testing. The samples were usually slightly discolored, but there was no evidence of corrosion beyond the slow rate of chromium removal, indicated by the analysis of the salt to be equivalent to removing all the chromium to a depth of only 0.4 mil or by the microprobe to be a concentration gradient in the thimble extending to a depth of 20 um (0.8 mil). When the samples were deformed at 25°C intergranular cracks formed to depths of a few mils. The frequency of cracking increased w1th time, but the maximnm depth did not increase detectably (Table 7). Examination of several components showed that shallow intergranular cracks were present in all materials that had been exposed to fuel salt. The statistics on the number and depth of cracks are given in Table 15 aleng with the tellumium‘concentrations based on.the chemical analyses in Table 14. Very thin surface eracks were often noted when the various components were removed from the MSRE. This was particularly true of the heat'exehanger tubing. Samples from the pump bowl where there was a liquid interface gave an opportunity to observe the decrease in crack severity in traversing from the liquid to the gas region. Allrof’the surfaces that were exposed to fuel salt were coated with fission.products. ‘Electrolytic sectioning showed that the fission products penetrated a few mils into the metal, but it is not definite whether they had diffused into the metal or whether they had-pleted on.thifi intergranular cracks. Chemica1 ana1yses on a sample that was oxidized to pessivate the surface and strained to expose the reactive cracked regions showed that the cfacked regions were enriched in Te, Ce,_SB, Sr,Cs, S, ahd P. ‘Ihe only sample that showed a definite dependence of crack severity on fission Table 15. Crack Formation in Various Samples from the MSRE Strained at 25°C ~ (>500°C for 30,807 hr and Exposed to Fission Products 24,500 hr) Freeze valve 105" aMeasuféd. bCalculated“from measured 127m7e, relative isotopic ratios, and half 1ife of 127Mpe, ‘cdne crack was 12 mils deep, next largest was 5 mils. - Cracks Depth (mils) 127mpea Total TeP Sample Description Counted Per Inch Av Max Atpms[cm? Atoms /cm? | | % 1015 x 1017 Exposed thimble 9 192 5.0 8.0 1.8,1.8 2.9,2.9 Thimble under spacer sleeve 148 257 4.0 8.0 0.59,0.46 0.95,0.74 lThimbié spacér, Qfiter Surfacé 88 178 3.0 7.0 D - Thihblé_spacer,'Inner Surfacg 106 202 3.0 S.Oj} 1'9’2'8 1.6.4.5 Mist shield, inside vépprfi_ 47 192 1.0 - 2.0 Mist shield, inside liquid ~ 33 150 4.0 6.5 . Mist'sfiiéld; outside vapor 80 363 4.0 5.0 Mist shield, outside liquid 54 300 3.0 . 5.0° 0.55 0.89 Safipler éaée rod, vapor - 100 143 2.5 5.0 Safipiei cage tod;-vaPOr | 170 237 . 3.2 10.0 Sampler cage rod, liquid . 102 165 3.7 10.0 . Sampiér cage:rod,.liquid 131 238 7.5 12.5 | - 131 260 0.75 1.5 0.04 0.06 6ST 160 product concentration was the one from FV 105. The fission product con- centrations were generally less than on the other samples by an order of magnitude (Table 14) and the cracks were very shallow (Fig. 104). Possible Mechanisms These observations raise several important questions including the cause ofgthefcracking‘and how rapidly these cracks propagate. Unfortu~ nately, the information obtained from the MSRE and the in-reactor 1ooosi is only sufficient to allow speculation on these questions, and further experiments such as those summarized in the next section will be required for complete evaluation. , g The first mechanism that‘must_be considered is that the cracking is due to some formlof_corrosion. The most likely form‘ofdcorrOSion‘under‘ the MSRE operating conditions would be the selective removal of chromiun by the reaction. ' 2UF,, (salt) + Cr(metel)'$=CrF2(salt) + 2UF3 (salt) .* Some impurities in the salt, particularly at the beginning of operation with 233U, wonld-have removed chromium from the metal by reactions such . - as FeF,(salt) + Cr(metal) = CrF,(salt) + Fe(deposited)' NiF,(salt) + Cr(metal) = CrF,(salt) + Ni(deposited) The‘net result of all of the corrosion reactions is that chromium ~ is selectively removed from the metal. This process is'controlled by the diffusion of chromium to the surface of the metal, where it is avail- able for reaction. DeVan3! measured the rate of diffusion of-chrominm in INOR-8, and his measurements can be used to calculate the depth to which *It is unlikely that the salt was ever oxidizing enough to form fluorides of Ni and Mo. Even if this had occurred, metal would be uniformly removed. As the salt became less oxidizing, the less stable . fluorides of Ni, Mo, and Fe would react with Cr in the alloy. It is likely that the Cr would reside in the salt -as CrF, and that the other metals would be deposited in the metallic form.) : i 161 chromium could have been removed. The most extreme case would occur when the chromium concentration at the surface was maintained at zero so that ‘the driving force for chromium diffusion toward the surface would be a maximum. The calculated‘chromium concentration profiles based on the assumption of zero surface concentration and the measured diffusion co- efficient of 1.5 x 10'1'+ cm2/sec are shown in Fig. 112 for various times. - The measured chromium profile for the thimble (Fig. 66) showed depletion to.acdepth of only 0.8 mil (0.0008 in.) which is somewhat less than the calculated depletion depth shown in Fig. 112. However, the assumption of zero concentration of chromium used in calculating the curves in Fig. 112 is too severe, and it is reasonable that the measured chromium deple- tion profiles would be less than those calculated. The rate of diffusion along grain boundaries, as we.wili discuss_more in detail iater, is more rapid than through the grains. Thus, the depth of chromium depletion along the grainrboundaries could be much greater than the d,B mil measured within the grains. | | - | | The evidence that has just been examined shows that the chromium could be depleted along the grain boundaries to depths approaching those of theobserved.cracks. However, two significant pieces of evidence sug- gest that chromium depletion alone does not cause the crscking. First, thousands of hours of loop corrosion tests were run involving several fluoride salts and INOR-8, with no 1ntergranu1ar cracks being ob- served,32>33,3% The second and most convincing evidence is that chromium . dell:»let.ion could not be detected in 'samples'- from theMSRE heat exchanger and in the section of the control rod thimble under a spacer sleeve. .h-However, these samples were cracked as severely as those (e ges the bare 'control rod thimble) in which chromium depletion was detectable (Table 15). Thus3.it seems'unlikely that chromium depletion alone can account.for'the observed cracking. | | | | | . Another mechanism to be considered is the diffusion of.some element(s) into the material, preferentially along the”grain boundaries. A reaction ofvthis type could cause (1) the fOrmation of a compound that- is very brittle,-or.(Z) a change in composition_along the grain boundaries so that they are liquid, or solid but very weak. Some deformation would 1likely be needed to form the cracks in all cases. - 162 " | ORNL-DWG 72-8704 l I I T . T | o 2500 hr . 25,000 hr . I 250,000 hr 10 Sy ~ | : . - e "/." a | /" MY T 5 0 - 0.8 , ¢ o \ "/ | | ° /./ c/c, o o \ .‘ .\ ] D e v oa -4 / Wy o 02 Gany) _6_ " 0002 0004 0006 0.008 0.010 0.012 0014 'o.o1_75 | - DISTANCE FROM SURFACE ( in.) - fig 112. Calculated Chromium Profiles at 650 c in INORrB Assuming Zero Surface Concentration of Chromium. 163 It is extremely important that the element(s) responsible for the cracking be identified and the mechanism determined. Examination of the analytical data in Tables 13 and 14 shows that all of the fission products with sufficient half-lives to be detectable afterVZ years were present, The data on the sample_that'was preoxidized show enrichments in several fission products as well as sulfer and phOSphorus. These data offer some vindication of the elements that are present but the detection limits on the nonradioactive elements were not sufficiently low to ensure that some of the stable fission products were not present in even larger con- centrations than the radioactive elements. Thus, it seemed profitable to look at all of. the elements in the fission spectrum with sufficient half-life to diffuse into the metal. Possible effects of these elements on INOR-8 are listed in Table 16. Many factors may be important, but only the information that we felt to be most relevant has been included Bieber and Decker3® have summa- rized the observations on pure Ni and Wood. and Cook36 have examined the effects of several elements on relatively complex niobium-base elloys. The thermodynamic properties and the behavior in the salt have been accumulated by W. R. Grimesret al37 from the literature, research, and studies of the MSRE. Several pieces of information are available from our current research and will be summarized in the next section. ' The’elements sulfur and selenium have detrimental effects under some tests conditions, but Te has had a more pronounced effect in all ~ types of tests run to date, These three elements form relatively un- stable fluorides and would likely be deposited on the metal and graphite - surfaces. Also As,'Sb, and Sr would,be deposited, but no deleterious - effects of these elements on the mechanical properties of nickel alloys have been noted. Zinc and cadmium may be deposited or present in the salt, depending on the oxidation state of the salt. Both of these ele- . ments are reported to be insoluble in nickel and we have not observed any deleterious effects in our test. Although Ru, Tc, Mb and Rh should be deposited we have seen no deleterious effects in our tests. Since -Zr, Sr, Cs, and Ce form very stable fluorides, they should remain in the salt. We have no evidenice, positive or negative, on the effects of stron- tium and cesium on the mechani cal behavior, but we presently feel that Table 16. Possible Effects of Several Elements on the Cracking of Hastelloy N2 Free Energy -Melting Concentrated Vapor and Effect on Effect on " Effect on . Effect on - of Formation Expected Element Point Near Cracks Electroplated Tensile Properties Creep Properties Tensile Properties Creep Properties of Fluoride Location Overall (°C) Specimens of NickelC on Nickel Alloyd " of Hastelloy Nb of Hastelloy Nb at 1000°K of Rating : | ' - (kcal/mole F)€ Element .~ s 119 - - + -- + - -34 , Depbsitgdf - Se 217 + - - + + - =27 Deposited - Te 450 - - -- - - - - - -39 ‘Deposited ~ - As 817 + - + + + -62 Deposited + sb 630 - + - - + + 4+ -55. Deposited + Sn 232 - - + + + =60 Deposited + Zn 420 , Insoluble + -68 g - cd 321 + Insoluble + 64 g 4 ‘Ru . 2500 + + + =51 Deposited 4+ B fé' 213b + + F + -46 Deposited -\¥+ Nb 2468 + + + + -70 g ++ Zr 1852 - - + 4 -99. Salt + - Mo 2610 + + + + -57 Deposited ++ sr 768 - - -125 ° sale L " Cs 29 - Insoluble ' -106 Salt Ce 804 - + -120 salt + Rh 1966 + ~42 vbeposi;ed + aThg symbols used in this table should be interpreted in the following way: A Mp" refers to good behavior and a "_" jndicates detfimental-effegts. bResults of current research: ‘ - ‘ 7 o c - d 7, 109, C. G. Bieber and R. F. Decker, "The Melting of Malleable Nickel and Nickel Alloys," Trans. AIME 221, 629 (1961). D. R. wood and R. M. Cook, (1963). ®private communication, W. R. Grimes, ORNL. fMay appear as HpS 1if HF concentration of melt is appreciable. fngMay appear in salt if salt mixture is sufficiently oxidizing. "Effects of Trace Contents of Impurity Elements on the Creep-Rupture Properties of Nickel-Base Elements," Mbtallurgiq %91 165 these elements will stay in the salt and not enter the metal. Zircofiitm and cerium do not have adverse effects when added to INOR-8. Niobium can be in the salt or depoéited, but it has very favorable effects on the mechanical properties. Thus, although some exploratdfy work remains, it appears that the cracking could be caused'by'the inward diffusion of ele- ments of the S, Se,lTe'family, with tellurium hafiingrthe most.adverse effects in the experimefits run to date. Our studiés have'conéentrated on Te on the basis that Te has had the most adverse effects. Because these elements all behave similarly, an understandihg of-how tellurium causes-cracking‘Should lead to an understanding of the role of the other elements in the cracking process as well. POST-MSRE STUDIES Since the surveillance SampleS'and parts of the MSRE were examined, numerous laboratory experiments have been conducted in an effort to better understand the cause of the cracking and its effects on the operation of a reactor. - Most of our work has concentrated on the fission product tellurium, and the rationale for this choice was discussed in the previous section. The experiments fall into the general categories of (1) corrosion in salt, (2) exposuré to several fission products to.compare the tendencies to produce cracks, (3) diffusion of Te, and (4) éxperiments with an applied ~stress. These experiments have involved materials other than INOR-8 in an effort to better understand the cracking phenomenon and to find a mate— rial that is more resistant to cracking. These experiments are continuing and only the most important findings will be summarized in this report. . Corrosion Experiments Arguments have already been presented that indicate our belief that the intergranular cracking in the MSRE could not have been due solely to chromium depletion. Wé'did-run'one further experiment. A thermal convec- tion loop containing a fuel salt had operated for about 30,000 hr with a very low corrosion rate. The bxidation potential of the salt was in- creased by two additions of 500 ppm FeF;, and the specimens were examined 166 periodically. The corrosion rate increased and some selective grain - boundary attack occurred but the grain boundaries d1d not crack when the _corrosion samples from the 1oop were deformed. Thus, we conclude that corrosion alone cannot account for the ob~ served cracking. However, chromium depletion by corrosion could make the material more susceptible to cracking by the inward diffusion of fission products, and this will be discussed later in more detail. Exposure to Fission Products Three types of experiments have been run in which samples are exposed “to fission products: (1) exposure of mechanical property samples to vapors of fission products, (2)’electroplating'tellurium on mechanical property safiples, and (3) studying-the mechanical properties of alloYS that contain small alloying additions of the fission products. Sulfur has been included in this work even though it is not a fission product, since it is intro— duced in a reactor by lubricant inleakage and is reputed to be detrimental to nickel-base alloys. The experiments with fission product'vapors have included S, Se, Te, I, As, Sb, and Cd.; These materials have sufficient vapor pressure at 650°C to transport to mechanical property samples. The samples were annealed in a quartz capsule with each fission product for various times | at 650°C, deformed at 25°C, and sectioned for metallographic examination.‘ Examination of the samples exposed in this way has revealed several important facts 1. Of the samples of INOR-8, type 304L stainless steel, and nickel— 200 exposed to all seven elements for 2000 hr at 650°C, only INOR-8 and nickel 200 exposed to tellurium had intergranular cracks after deformation ‘at 25°C, | | 2, The cracks produced in INOR-8 by exposure to tellurium look quite similar to those formed in the material from the MSRE (Fig. 113). 3. Exposure to tellurium over the temperature range of 550 to 700° and concentration range of two orders of magnitude showed that the severity of cracking in INOR-8 and nickel 200 increased‘with increasing temperature and tellurium concentration. Type'304L stainless steel did not crack under any of the conditions investigated. 167 Y~ 114523 | e }--——-o 0314 MSRE Thlmble - 3] 000 hr in Fuel Salt at 650°C - 1.4x10I atoms/cm? of Te l-——o 03"—’—1 Vapor Plated with 5.4 x 1018 afoms/cm of Te Annealed 1000 hr at 650°C Fig. 113. Samples of INOR-8 Strained to Failure at 25°C. The upper photograph was made of a piece of the MSRE thimble. The upper side was exposed to fuel salt and cracked when strained. The lower side was exposed to the cell environment and the oxide that formed cracked during straining. The lower specimen was exposed to tellurium vapor on both sides and deformed. The cracks are similar to those formed on the surface of the upper sample that had been exposed to fuel salt. 168 4. Inconel 600 and a heat of INOR-8 modified with 2% Ti showed less severe cracking than INOR~8 under similar exposure conditions. About 60 commercial alloys have been electroplated with_tellfirium'to compare the'relative‘tendencies-to form intergranular cracks. One set of samples was annealed for 1000 hr at’650°¢ and the other set was annealed - 200 hr at 7006C.‘_The materials included consisted of (1) iron and sever- al stainless steels, (2) nickel and several hickel-base alloys, (3) copper and monel, (4) two cobalt-base alloys, and (5) severai heats of INOR-8 with fiodified chemical compositions. The‘samples were eiectroplated, welded in a metal capsule, evacuated and backfilled fiith argon, and - welded closed. The ehvirbnmefit was impure enough that some oXides'were | detected by x rays, and the results may be influenced by the preeence of the oxide. .The samples were deformed at 25°C and examined metal- lographically'to—determine whether cracks were present. The,fesults of the two experiments were quite consistent and several important observations were made. 1. Cracks did not form in any of the iron-base alloys. 2. Nickel, Hastelloy B (1% Maximum Cr), Hastelloy W (5% Cr), and _INOR-8 (7% Cr) formed eracks, but several alloys containing 15% or more Cr did not crack (e.g., Inconel 600, Hastelloy C, and Hastelloy X). 3. Copper and Monel did not crack. B | 4. The two cobalt-base alloys contained about 20% Cr and did not crack. . _ . _ | . | | 5. Several of the modified heats of INOR-8 cracked less severly ‘than the standard. alloy. The better alloys seemed to contain more nidbium, | although there were usuallyrother additions also. Two_alloys containing 2% Nb did not have any cracks. | i " The resfilts of the.two types of eiperiments; one ekposing the mater- ial to fission product vapors and the other to electroplated tellurium, generally agree. Not all of the alloys have been tesfed by both methods, -Small melts have been made containing nominal additions of 0.01% ~each of S, Se, Te, Sr, Tc, Ru, Sn, Sb, and As. Tests on these alloys show no measurable effects on their mechenicel properties at 25°C., In 169 creep tests at 650°C only S, Se, and Te have deleterious effects. These elements reduce the rUpturerlife‘and the fracture strain. Alloys con- taining sulfur and fellurium have been fiade with and without chromium. The effects of sulfur and tellurium are more.deleterious when chromium is not present. This suggests a possible tie between C6rrosion-by se- lective chromium removal and interg:anular cracking due to tellurium. The regions depleted in chromium should be much less tolerant of tel- lurium than those containing,chromium; Diffusion of Te The rate of diffusion of tellurium in.INOR~8, nickel 200, anfl type 304L stainless steel has been méasured. The penetration depths were quite small in some samples, but the data generally give the bulk ahd grain boundary diffusion coefficients at 650°C and 760°C accurately enough to make predictions of theidepth-of penetration in service. The depth was very shaliow in type 304L stainless steel, at least twice as déep in INOR-8, and many times greater in nickel 200. - ' ' The Fisher model for grain boundary diffusiofi relates thg depth of penetration to the diffusion time to the one-fourth power.38 Thus, the inérease in the depth of penetration for a System at temperature'for 30 years (an MSBR) is only 1.8 times as great as that for a system at tem- peratures for three years (the MSRE). However, these calculated values are lower limits and make no allowance'for-the diffusion front advancing as intergranuiar cracks form and propagate. Experiments with an Applied Stress Several types'of expefimentS'with'ah applied stress have beén run. ‘One is a standard creep test with the sample in an envirpnment-of’argon ~and tellurium. The tellurium is'preplaced'in a quartz vial at a location where it will have a vapor pressure of aobut 0.5 torr. INOR-8 cracks very badly at 650°C at relativelyfihigh stress levels. The cratks extend 35 , mils, compared with about 5 mils in the tests that were stressed after the 170 exposure to tellurium. Tests near reasonable design stresses have been in progress several thousand hours. Type . 304L stainless steel, nickel-200, and Inconel 600 have been in test-several thousand hours, but have not been examined metallegraphically; The-reaction products of tellurium with nickel-200 and INOR-8 are basically nickel tellurides, but the only detect- able ptoduct on stainless steel is an oxide of the Fe304 type, | Tube burst tests of INOR-8 were run with the outside of the tubes in helium or salt enviromments. Some of the tubes were electroplated with tellurium before testing. The test environment had no detectable effect, but the tubes plated with tellurium failed in ghorter times than those not plated. Mbtallographic examination revealed intergranular ctacks'about 10 mils deep along almost every grain boundary of the plated - specimens. Some of.the stainless steel tubes are still in test but the f rupture lives of those that have failed are equlvalent for plated and unplated specimens. ' The other type of stressed test that has been run is one with con- trolled strain 1imits. Thermal stresses commonly occur in‘components such as heat exchangers where high heat fluxes develop and transients are frequent. Thermal strains of +0.3% are anticipated for MSBR heat exchangers.' Our first test has utilized a Te-plated Hastelloy N speci- men strained between limits of +0.16%. The rupture 1ife-wae'shorter than anticipated on the basis of tests with tellurium, and numerour inter- granular cracks were present on the outside where the tellurium was plated. SUMMARY These tests show the tellurium causes the formation of intergranular cracks in INOR-8 and that these cracks are deeper if the material is stressed and exposed to tellurium simultaneously than they would be if the material-wete exposed to tellurium.and then stressed. Although the diffusion rate of.tellurium in INOR-8 has been measured, the role of - .stress in aceelerating crack propagation makes the diffusion measurements of questionable value in estimating the extent of cracking in service. Only elements of the S, Se, and Te family‘were found to cause inter- granular cracking in INOR-8. | | o \f 171 Exploratory experiments indicate that several materials are more resistant to intergranular'cracking than INOR-8. Iron-base alloys, copper, and Monel seem completely resistant., Nickel- or cobalt-base alloys with about 207 Cr seem résistant, but the results for alloys such as Inconel 600 with 15% Cr are inconclusive. Some modified'cbmpositions of INOR-8 have exhibited improved resistance to crécking. 'The INOR-8 from the MSRE, which had been exposed to fuel salt, was noted to contain'intergranfilar cracks to depths of -a few mils. These .cracks were visible in some materials in as-polished metallographic sec- tions as they were removed from the MSRE. When the samples were deformed at room temperature, the cracks‘became more visible but still reached a limiting depth of about 10 mils. The severity of cracking could not be related with the amount of chromium removal but was most severe at the liquid interface in the pump bowl and least severe in regions of the pump bowl exposed to gas. Samples were sectioned electrochemicélly, and some fission produéts were found td depths of several mils, although it was not clear whether the fission producté.had diffused through the metal to these depths or simply piated on the surfaces of cracks that already existed. | Not being-able to relate the intergranular cracking to corrosion (chromium leaching) in either the.MSRE or in laboratory experiments, we investigated the possible effects of fission products on the material. | The MSRE had been shut down and the work was continued in laboratory ex- 'periments. This work included the exposure of INOR-8 to small amounts of vapor or electroplated fission products, the study of several alloys with fission products added, exPefiments with tellurium and an applied stress present simultaneously, and the measurement of the diff#sion of tellurium in iNOR—B. These experiments have shown clearly that of the many_eléments tested,‘only tellurium caused intergranular cracking sim- ilar to that noted in samples from the MSRE. Several other materials were also included in these experiements to determine whether they might be more resistant to embrittlement by tellu- rium. Several alloys, ifiClfiding 300 and 400 series stainless steéls, cobalt- and nickel-base alloys containing more than 15% Ce, copper, Monel, 172 and some modified compositions of INOR-8 are resistant to cracking in the ‘tests run to date. Further work will be necessary to show unequivocally that these materials resist cracking in nuclear environments, including in-reactor capsule tests. % 3 173 ACKNOWLEDGMENT The observations in this repqrt.cover several years and include contributions from many individuals. W.’H, Cook and A. Taboda designed the surveillance fixture, and W. H. Cook was responsible for its assembly and disaséembly.,_The MSRE opeiations staff, héaded by P. N. Haubenreich, exercised extreme'care_in handling the surveillance fixture and in re- moving the wvarious cdmponents for examination. The Hot Cell Operation Stéff, headed by E. M. King, developed several special-toolé and techniques for various examinations and tests of materials from the MSRE. The metal-~ lography was pérformed by H. R;-Tinch, E. Lee, N. M. Atchley, and E. R. Boyd., The microprobe scans were made by T. J. Henson and R. S. Crouse. The technique for electrolytically dissolvifig samples in the hot cells was developed by R. E. Gehlbach and S. W. Cook. The meghaniéal property tests were performed by B. C. William, H. W..Kline, J. W. Chumley, L. G. - Rardon, and J. C. Fettner. The chemical analyses were performed under the supervision:of W. R. Laing, E. I. Wyatt, and J. Carter. Assistance was received from several members of the Reactor Chemistry Divisiqn'in several phases of this work.-,J. V. Cathcart, J. R. DiStefano, P. N. Haubenreich, and J. R. Weir fevieWed_the manuscript of this report and made many helpful suggéstions,' Kathy Gardner made the original drafts of this report, and the drawings were prepared by the_Graphic Arts De- partment. 7. 10. 11. 12. 14, 15. 16. 174 - REFERENCES P. N. Haubenreich and J. R. Engel, "Experience with the MSRE," Nucl. H. E. McCoy, "The INOR-8 Story," ORNL Review 3(2), 35 (1969). S. H. Bush Irradiation Effects in C'Zadd'mg and Stmctural Matemals s ~ Rowman and Littlefield, Inc., New York (1965) . H. E, McCoy, An Evaluation of the Molten-Salt Reactor Expemment Hastelloy N Surveillance Specmens — First Group, ORNL-TM-1997 (November 1967). ' ~ H. E. McCoy, 4n E’valuatwn of the Molten-Salt Reactor Expemment Hastelloy N Surveillance Specimens — Th'z,rd Group, RNL-TM—2359 (February 1969) _H. E. McCoy, An E’valuatwn of the MoZten-SaZt Reactor Experiment Hastelloy N Surveillance Specimens — Third Group, ORNL-TM-2647 . —(January 1970). ' . H. E. McCoy, 4n Evaluation of the Molten-Salt Reactor Experiment Hastelloy N Surveillance Specimens — Fourth Group, ORNL-TM~-3063 (March 1971). : W. R..Martin and J. R. Weir, "Effect of Elevated Temperature Irradi- ation on the Strength and Dectility of the Nickel-Base Alloy Hastelloy N," Nuel. Appl. 1, 160 (1965). W. R, Martin and J. R. Weir, "Postirradiation Creep and Stress Rup- . ture of Hastelloy N," Nucl Appl. 3, 167 (1967). H. E. McCoy and J. R. Weir, "Stress~Rupture Properties of Irradiated and Unirradiated Hastelloy N Tubes," Nucl. Appl. 4, 96 (1968). H. E., McCoy, "Variation of the Mechanical Properties of Irradiated Hastelloy N with Strain Rate," J. Nucl. Mater. 31, 67 (1969). T. C. Robertson, MSRE Design and Operafions Report, Part I — | Descmlption of Reactor Design, ORNL-TM-728 (January 1965). R. E. Thoma, Chemical Aspects of MSRE Operatwn, ORNL-4658 (December 1971). C. H. Gabbard, Design and Cons truction of Core Irradtatwn-Speczmen Array for MS'RE' Runs 19 and 20 ORNL-TM-2743 (December 1969) . MSR Program Semiannu. Progr. Rep. Feb. 28, 1966, ORNL-3936, pp. 96-99. MSR Program Semiannu. Progr. Rep. Aug. 31, 1966, ORNL-4037, pp. 97-102. O S - 17. 18. 19. 20. 21. 22, 23. 24, 25. 26. 27, - 28. 29. 30' 31. 32. 33. 175 MSR Program Semiannu. Progr. Rep Aug 31, 1968, ORNL-4344, PP 211-14. o JMSR Program Sbmzannu Progr. R@p. Aug. 31 1969, ORNL-4449, pp. 165-68. ‘ , P. N. Haubenreich and M. Richardson, Plans fbr Pbst—Operatzon Ezxam- ination of the MSRE, ORNL-TM-2974 (April 1970). MSR Program Semiannu. Progr. Rep. Feb. 28,‘1971, ORNL-4676, pp. 1-16. G. M. Tolson and A. Taiboé.da, MSRE Control Elements: Manufacture, Inspection, Drawings, and Specifications, ORNL-4123, (July 1967). “H. E. McCoy and J. R. Weir, The Effect of Irradiation on the Bend Transition Temperatures of Molybdenum- and Niobium-Base Alloys, . ORNL-TM-880, pp. 7-10 (July 1964). B. McNabb and H. E. McCoy, MSR Program Semiannu. Progr. Rep. Feb. 28 1971, 0RNL-4676, pp. 160-166. D, B, Trauger and J. A. Conlin, Jr., "Circnlating Fused-Salt Fuel Irradiation Test Loop," Nucl. Sei. Eng. 9, 346 (1961). J. A. Conlin and ‘B. H. Montgomery, MSRP In-Pile Loop ORNL-MTR-44 Design and Operating Conditions, unpublished. E. L. Compere H. C. Savage, and J. M. Baker, MSR Program Semiannu. Progr. Rep. Aug. 31, 1967, ORNL-4191, p. 176. E. J. Manthos, "Disassembly and Examination at In-Pile Loop, ORNL- MIR-44-1," unpublished. E. J. Manthos, "Disassembly and Examination at In—Pile Loop, 0RNL—MTR—44 -2, " unpublished. , E. L. Compere, private communication; C. G. Bieber and R. F. Decker, "The Melting of Malleable Nickel and Nickel Alloys, " Trans. AIME, 221, 629 (1961). J. H. DeVan, Efféct of Alloying on Corroston Behavior of'NzckeZ— Molybdenun Alloys in Fused Fluoride Mixtures, ORNL—TM-ZOZI (May 1969). H., E. McCoy and J. R. Weilr, Materials Development for MbZten-SaZt Breeder Reactors, ORNL-TM-1854 (1967) G. M. Adamson, T. S. Crouse, and W. D, Manly, Interim Report on - Corrosion by Alkali-Metal FZuomdes. Work to May 1953, ORNL-2337 (1959). 34, 35, 36. 37. 38." 176 G. M. Adamson R. S. Crouse, and W. D, Manly, Interim R@port on Corrosion by Zirconium-Base Fluorzdés, ORNL-2338 (1961). C. G. Biéber and R. F. Decker "The Melting of Malleable Nickel and Nickel Alloys," Trans. AIME, 221, 629 (1961) ' D. R. Wood and R. M. Cook, "Effects of Trace Contents ofAImpurlty Elements on the Creep-Rupture Properties of Nickel-Base Elements, Metallurgia 72, 109 (1963). W. R. Grimes et al., private communication._ . J. C. Fisher, J. Appl. Phys. 22, 74-77 (1951). 9 4 ) . 4. - 18, “ 1-3. 4. 5-39. 40, 41. 42, 43. 44, 45. 46. 47. 48. 49. 50. 51. 52. 53. 55. 56. - 57. ' 58. 59. ~ 60. '61. 62, 63. 64. 65. 66. 67. 68. 69. 70, 71. 72. 73. 74, 75. 76. 77. 79. 80. 81. Central Research Library | 83. R. 177 ORNL-4829. UC-25 — Metals, Ceramics, and Materials ~ INTERNAL DISTRIBUTION E. Gehlbach ORNL — Y-12 Technical Library \ 84, L. 0. Gilpatrick. Document Reference Section 85. W. R. Grimes Laboratory Records Department 86. A. G. Grindell Laboratory Records, ORNL R.C. - 87. R. H. Guymon ORNL Patent Office , ' - 88. W. 0. Harms G. M. Adamson, Jr. 89. P. N. Haubenreich J. L. Anderson L ' 90. R. E. Helms W. E. Atkinson | | | 91. J. R. Hightower C. F. Baes | ' 92. R. F. Hibbs C. E. Bamberger 93-95. M. R. Hill C. J. Barton : 96. E. C. Hise H. F. Bauman- 97. W. R. Huntley S. E. Beall . 98. H. Inouye C. E. Bettis ' 99. W. H. Jordan D. S. Billington 100. P. R. Kasten F. F. Blankenship \ ) - 101. H. G. MacPherson. E. E. Bloom | | | ) 102, R. J. Kedl R. Blumberg ' - 103. C. R. Kennedy E. G. Bohlmann- . ~ 104. R. T. King J. Braunstein - i ' 105. S. S. Kirslis M. A. Bredig - 106. J. W. Koger R. B. Briggs 107. A. 1. Krakoviak H. R. Bronstein - o 108. T. S. Kress G. D. Brunton _ - 109, J. A. Lane S. Cantor - 110. R. B. Lindauer D. W. Cardwell - o _ 111. E. L. Long, Jr. J. V. Cathcart - 112. A. L. Lotts 0. B. Cavin , | 113. M. I. Lundin Nancy Cole o , 114, R. N. Lyon C. W. Collins - . -115. R. E. MacPherson E. L. Compere 7 S 116. D. L. Manning W. H. Cook o | "117. W. R. Martin J. L. Crowley o ~118. R. W. McClung F. L. Culler - . 119-124, H. E. McCoy D. R. Cuneo - - . 125. D. L. McElroy J. E. Cunningham | S 126. C. J. McHargue J. M. Dale o 127. H. A. McLain. J. H. DeVan - 128, B. McNabb J. R. DiStefano 129, L. E. McNeese S. J. Ditto | 130. J. R. McWherter W. P. Eatherly - o - 131. A. S. Meyer J. R. Engel | " S 132. R. L. Moore D. E. Ferguson - ) "133. D. M. Moulton J. H Frye, Jr. | o ~ 134, J. P. Nichols W. K. Furlong _ 135. E. L. Nicholson C. H. Gabbard o ' 136. R. B. Parker 137, - 138. 139. 140. 141. 142, 143, 144, 145. 146, 147. '148. 149. 150. 151. 152. 153. 154. 155. 156. 176. 177. 178. 179. 180. 181. 182. 183. 184, 185, 186. 187. 188. 189. 190-191. 192. 193. 194. 195. 196. 197. 198. P. A, C. B. G: Do' R. K. M. H. W.. J. J. G. G. I. R. R. J. J. 178 Patriarca M B. E. L. c. A. W. c. F. L‘ . M. P. A. W. R. A' . Perry Pollock Prince Ragan Richardson Robertson Bomberger Rosenthal Savage Schaffer Dunlap Scott Scott Shaffer Slaughter Smith ‘Spiewak Strehlow Swindeman Tallackson 157. 158. 160. 161. 162. 163. 164. 166. 167 L] 168. 169, 170. 171. 172. 173. 174, 175, EXTERNAL DISTRIBUTION R. E. Thoma D. B. Trauger W. E. Unger G. M. Watson J. S. Watson -H, L. Watts C. F. Weaver B. H. Webster "A. M. Weinberg J. R. Weir J. C. White R. P. Wichner L. V. Wilson E. L. Youngblood F. C. Zapp ‘Leo Brewer (consultant) Walter Kohn (consultant) G. V. Smith (consultant) W. S. Williams (consultant) Acciarri, Continental 0il Co., Ponca City, Oklahoma 74601 R. M. Bushong, UCC Carbon Products Division, 12900 Snow Rd., Parma, Ohio 44130 G. C. Clasby, Byron Jackson, P.0. Box 2017, Terminal Annex, Los Angeles, California 90054 D. F. Cope, RDT, SSR, AEC, ORNL D. R. deBoisblanc, Ebasco Services, Inc., 2 Rector St., New York, N.Y. C. B. Deering, Black and Veatch, Kansas City, Missouri David Elias, AEC, Washington, D.C. A. Giambusso, AEC, Washington, D.C. T. A. Flynn, Jr., Ebasco Services, Inc., 2 Rector St., New York N.Y. : J. E. Fox, AEC, Washington, D.C. 10006 10006 C. E. Johnson, AEC, Washington, D.C. Kermit Laughon, AEC, OSR, ORNL C. L. Matthews, AEC, OSR, ORNL J. M. Martin, The International Nickel Company, Bloomington, West - Virginia N. Haberman, AEC Washington, D.C. T. C. Reuther, AEC, Washington, D.C. B. Mong, Babcock & Wilcox Co., P.0. Box 1260, 1201 Kemper St. ‘Lynchburg, Virginia 24505 T. K. Roche, Cabot Corporation, Kokomo, Indiana 46901 S. Rosen, AEC, Washington, D.C. . . M. -Shaw, AEC, Washington, D.C. J. M. Simmons, AEC, Washington, D.C. ‘E. 0. Smith, Black and Veatch,. P.0. Box 8+05 Kansas City, Missouri 64114 \,‘ -9 s bt o b el b e " A L A e v e+ ok ek e A T 199. 200. 201. 202. 203. 204-396., 179 E. E. Stansbury; The University of Tennessee, Knoxville, Tennessee D. K. Stevens, AEC, Washington, D.C. - J. A. Swartout, Union Carbide Corporation, New York, N.Y. 10017 Research and Technical Support Division, AEC, ORO Patent Office, AEC, ORO - \ Given distribution as shown in TID-4500 under Metals, Ceramics, and Materials category (25 copies — NTIS) |