r— e W s 1 —— o O e ) E——————" o e D e L~ e—— & [ e —— o mIT - Ar— T —————— e A————— ) = 3 445k D359L34 & | MZUN WiARL b ifl Y, T AA s Iveiss s ! _-'\s & R e f"fi":" ,.::’ e Lo e o o ;% . S o S e ZRRae i i ezl R = e 2% e . A & ORNL/TM-5920 Distribution Category UC-76 Contract No. W-7405-eng-26 METALS AND CERAMICS DIVISION STATUS OF MATERIALS DEVELOPMENT FOR MOLTEN SALT REACTORS H. £. McCoy, Jr. Date Published: January 1978 NOTICE This document cantains information of a preliminary nature, It is subject to revision ar correction and therefore does nof represent a fina!l report. GAK RIDGE NATIONAL LABORATORY Oak Ridge. Tennessee 37830 ‘operated by UNION CARBIDE CORPORATION for the DEPARTMENT OF ENERGY TEMS LIBRARIES (TR 3 yy5L 0359634 & CONTENTS ABSTRACT ................................................ INTRODUCTION --------------------------------------------- DEVELOPMENT STRATEGY Irradiation Embriitlement ...................................... -------------------------------------- Tellurivm Embrittlement—Alioy Modilication Tellurium Embrittlement—Salt Modification ---------------------------- ............................. STATUS OF DEVELOPMENT . . . . . . . . e e s e e s e e e e Irradiation Embrittlement—2%-Ti-modified Alloy Fabrication Weldability Creep Strength Alloy Stability Salt Corrosion .......................... ........................................... ........................................... ------------------------------------------ .......................................... .......................................... Postirradiation Mechanical Properties .............................. Microstructural Features . . . . . . . o 0 0 0 i e e e e e e e e e e e e e e [rradiation Embrittlement--Nb-modified Alloys Fabrication Weldability Creep Strength Salt Corrosion . v v v v v o e e e e e e e e e e e e e e e e e e e e Microstructural Features . . . . . . . o v v o e e e e e e e e e e e e Trradiation Embrittlement--Ti/Nb-modified Alloys ......................... ---------------- ............................ ........................... ........................................ .......................................... il STATUS OF MATERIALS DEVELOPMENT FOR MOLTEN SALT REACTORS H. E. McCoy, Jr. ABSTRACT Experience to date has shown that in a molten-salt reactor cnvironment the alloy Hastelloy N is embrittled by irradiation and suffers shallow intergranular cracking due to the fission product tellurium. From January 1974 through September 1976 these problems were actively rescarched. Hastelloy N modified with 1 to 2% Nb was found: to have good resistance to irradiation embrittlement and to intergranular cracking by telluriumn, The: severity of cracking by tellurium was noted to be influénced by the oxidation state of the salt so that cracking could be prevented even in standard Hastelloy N. This observation opened up other possibilities for materials selection. INTRODUCTION Molten-salt reactors were initially considered tor nuclear-powered aircraft in 1947, A small test reactor called the Aircraft Reactor Experiment was constructed in 1954 and was operated successfully for 221 hr. The potential of the concept for civilian power applications was realized, and the Molten-Salt Reactor Program was formalized in 1956. A 74-MW test reactor, called the Molten Salt Reactor Experiment (MSRE), was designed, constructed, and became critical in 1965. The reactor operated successfully for several years, and operation was. terminated in 1969 after the reactor successfully completed its mission.! During the years in which the MSRE was being built and brought into operation, most of the development work on molten-salt reactors was in support of the MSRE. As a result of the success of the MSRE, however, the budget was increased to permit work aimed at molten-salt breeder reactors (MSBR) and the shutdown of the MSRE freed additional funds for this purpose. These reactors would use a LiF/BeF, carrier salt with fissile 233U and fertile 232Th. The salt would flow through passages in a graphite-moderated core, where :fissioning of the uranium and capture of the neutrons by thorium to form uranium would occur. However, numerous fission products would be produced, and their partial removal would be necessary to have an efficient breeder. The latest developments in chemical processing show that the fission products can be removed to acceptable levels by sequentially removing the uranivm by fluoridation and by contacting the salt with Bi/Li solutions. Although a number of areas require further development, this report will deal specifically with the metallic structural material for the primary system. Operation of the MSRE revealed two deficiencies in the Hastelloy N alloy (Ni; 16% Mo; 7% Cr; 5% Fe; 0.5% Si: and 0.05% C) developed specifically for use in molten-salt systems. First, the alloy was embrittled at elevated temperatures by exposure to thermal neutrons. The creep strength of the alloy was not affected, but the strain that could occur betore fracture was reduced. The second problem was that the grain boundaries were embrittled to depths of 5 to 10 mils in all Hastelloy N exposed to the fuel salt and to a lesser extent in material exposed to the vapor above the salt. The embrittled boundaries were opened to form visible cracks only in the heat exchanger. In other M w. Rosenthal, P. N. Haubenreich, and R. B. Briggs, The Development Status of Molten-salt Breeder Reactors, ORNL-4812 (August 1972). b2 parts of the systeimn it was necessary to strain the material to form visible cracks. This intergranular cracking was clearly associated with fission products, and strong circumstantial evidence suggested that tellurium was the cause of the embrittlement. The materials program during the past several years emphasized these two problem areas. The irradiation-embrittlemnent problem was noted several years before the fission-product-related problem, and work has proceeded further in developing an alloy that resists embritilement by neutrons than in developing an alloy that resists embrittlement by tellurium. The timing associated with discovering the two problems also led to the problems being pursued almost independently, and, in retrospect, to actions that do not appear to be aimed at the problem. However, this situation is largely attributable to the problem changing as more information became available. The final objective of this study is to develop a material for construction of the primary system of an MSBR. This development will involve the progression of tests on small laboratory-size melts to tests on production-size melts and, similarly, a progression of test times for a few hundred hours to several thousand hours. This progression would lead to the development of an alloy for which production techniques are readily available and whose properties are well known. DEVELOPMENT STRATEGY The problems of irradiation and fission-product embrittlement seemed best pursued through chemical modifications of the alloy. Previous studies at ORNL had shown that the degree of helium embrittlement of austenitic stainless steels could be reduced markedly by slight chemical modification,? and it was anticipated that this approach would be useful for Hastelloy N. Previous evaluation of components from the MSRE had strongly implicated Te as the fission product responsible for the intergranular cracking.3 There are not much data available on tellurium chemistry, but the approach used was to add elements to Hastelloy N that were reactive with Te, in hopes that the reactive element would form a stable telluride compound. As a compound, the tellurium might be innocuous. A completely separate approach to the tellurium problem would be to modify the salt chemistry to place the tellurium in an innocuous form. Again, this approach was hampered by the lack of data on the many tellurides possible in a salt/fission-product mixture, and it was not viewed with much optimism. Each of these problems will be discussed, and the rationale for the experimental approach will be presented. [t was imperative that these problems be corrected without sacrificing the excellent corrosion resistance of the material to fluoride salts and its ability to be manufactured and formed into useful shapes by conventional methods. Irradiation Embrittlement Iron- and nickel-base alloys can be embrittled in a thermal neutron flux by the transmutation of tramp 10B to helium and lithium. This process generally results in the transmutation of most of the 19B by fluxes of thermal neutrons on the order of 1020 ¢m-2, and usually yields from 1 to 10 atomic ppm of He. With Ni there is a further thermal two-step transmutation involving these reactions: S8Nj + 1 — SINJ, SON; + 41— 3He + S6Fe. 2W. R. Martin and J. R. Weir, Solutions to the Problems of High-temperature lrradiation {fmbrittlement, ORNL-TM-1544 (June 1966). 3H. E. McCoy, Jr., and B. McNabb, Mrergranular Cracking of INOR-8 in the MSRE, ORNL-4829 (November 197 2y, This sequence of reactions does not saturate, and although the cross sections are still in question, it would produce a maximum of 40 atomic ppm of He in the vessel over a 30-year MSBR lifetime.# Helium from both sources collects in the grain boundaries and causes degradation of the mechanical properties at elevated temperatures. The effect manifests itself in reduced rupture life and reduced fracture strain. Previous work with Hastelloy N showed that the predominant carbide formed in air-melted material containing about 0.5% Si was M C where M was 27 9% Ni, 3.3% Si, 0.6% Fe, 56.1% Mo, and 4.0% Cr.5 This carbide was very coarse and was not present on a fine enough scale to influence the transport of He through the individual grains to the grain boundaries. In vacuum-melted material the silicon was lower and the carbide was of the M,C type where M was 80 to 90% Mo with the remainder Cr.® This carbide was fine but coarsened readily at about 700°C. By keeping the silicon low, reducing the molybclénum concentration from 16 to 12%, and adding a small amount of one element from the group Nb, Ti, Zr, and Hf, a finely dispersed carbide of the MC type was formed. Figure 1 is an electron micrograph showing the fineness of 4E, I, Allen, H. T. Kerr, and J. R. Engel, QRNL, personal communication, August 1976, SH. F. McCov, Jr., and R. . Gehlbach, Nucl. Technol. 77,45 (1971), Fig. 1. Typical MC carbide distribution in hatfnium=modified Hastelloy N. The alloy contains 0.7% Hf and 0.06% C and was aged 200 hr at 760°C. Original magnitication: 3000x%. this structure and how it could provide sites for trapping lle and preventing its transport to grain boundaries. Other modifications were made in the specifications for Fe and Mn. These latter changes were made primarily to simplify the alloy for our studies and are not considered critical. Postirradiation creep studies showed that Zr and Hf were effective in producing the desired carbide structure, but that they resulted in very poor weldability.® Zirconium concentrations of as much as 0.05% caused severe weld metal cracking. Because Hf contains some Zr, it was not possible to determine whether the Hf or the residual Zr caused the cracking. However, Hf had a further scale-up problem in that HfO, is more stable than many oxides used in refractories. The Hf added to a mielt in a refractory crucible would be oxidized and not available to form a carbide in the alloy. Thus, the use of Zr and Hf additions was not pursued further for practical reasons. The Ti addition appeared very beneficial, and the development of Ti-modified alloys was pursued. Additions of Nb were studied to a lesser extent. It was not until late in the program when it was discovered that Ti was not only ineffective in preventing intergranular embrittlement by Te, but that it also destroyed the beneficial effects of Nb when both elements were present. However, the 2%-Ti-modified alloy was developed through two large commercial production heats before this fact was discovered. The experimental approach used to evaluate the extent of irradiation embrittlement was to irradiate small samples in the Oak Ridge Reactor (ORR) and then to perform postirradiation creep tests. One-hundred two specimens were irradiated at one time with controlled temperatures in the ORR poolside. Twenty-three hot-cell creep machines provided excellent facilities for the postirradiation creep tests. Control creep tests were run to provide a basis for comparison with the data on irradiated specimens. Selected irradiated and unirradiated specimens were chosen for detailed electron microscopy. The alloys evaluated were made by three techniques. Small 2-1b melts were made at ORNL. They were arc melted on a water cooled Cu hearth, drop cast to ingots 1 in. diam by 6 in. long, and swaged to 1/4-in.-diam rods. Melts 50 to 100 1b were made by commercial vendors and formed into 1/2-in.-thick plates. These plates were used for weldability tests, and strips were cut for making test specimens. Two large commercial melts of 8000- and 10,000-1b sizes were obtained. They were made by standard melting practice and were formed into a number of products. Tellurium Embrittlement—Alloy Modification Examination of components from the MSRE suggested that the fission-product tellurium was responsible for the intergranular cracking that occurred in Hastelloy N. Some early laboratory experiments demonstrated our ability to produce similar attack by exposing Hastelloy N to small concentrations of Tein the laboratory.3 These experiments revealed several important differences in the degree of embrittlement of various alloys by Te. However, these tests were not a good simulation of reactor conditions in that they exposed the metal to a relatively high flux of Te for a short time at the beginning of the test, and no further Te addition was made during the thermal anneal. In a reactor the metal would be exposed to a very small but almost constant flux of Te throughout its period of operation. Thus, better methods of exposure were needed. The development of Auger spectroscopy methods at ORNL allowed a very definitive experiment to be performed recently on material retained from the MSRE.7 In this experiment a thin Hastelloy N foil from the MSRE was broken in the Auger spectrometer to expose a fresh grain boundary. As shown in Fig. 2, 64. E. McCoy, Jr., Influence of Titanium, Zirconium, and Hafnium Additions on the Resistance of Modified Hastelloy N to Irradiation Damage at High Temperature—Phase I, ORNIL/TM-3064 (January 1971). L. E. McNeese, Molten-Salt Reactor Program Semignnu. Prog. Rep. Feb. 29, 1976, ORNL-5132, pp. 88-95. ORNL-OWG 76-T598 (a2} SPOT {-FRESH FRACTURE SPOT{-SPUTTERED 5 min (b} N E)/oE (orbitrary units) ENERGY —mmectm Fig. 2. Auger spectra of fracture surface of Hastelloy N specimen from MSRE. Auger spectroscopy showed the presence of Te in a brittle grain boundary and did not reveal another fission product in detectable concentrations. This experiment confirmed previous circumstantial evidence that indicated that the embritt].emem was due to the fission product tellurium. The sparsity of data on the chemistry of tellurides makes it possible only to postulate a mechanism for embrittlement. The phase diagram for the Ni/Te system is shown in Fig. 3. The nickel-rich side is characterized by very low solubility of tellurium in nickel and several intermetallic nickel/tellurium compounds. Such compounds are generally quite brittle and there is no reason to suspect that this system will not be the same. However, these compounds occur at reasonably high concentrations of Te, and one must postulate a mechanisin: for the enrichment of Te that would lead to the formation of brittle compounds. A schematic diagram of a grain boundary is shown in Fig. 4. Grain boundaries are regions of poor crystalline perfection between adjacent grains of different orientations. This region is generally assumed to be about 3 atomic diameters wide and can have diffusional and chemical characteristics much different from those of the bulk material. It is quite plausible that tellurium diffuses along the grain boundaries and that the concentrations reach levels as high as:those required to form the brittle intermetallic compounds. In fact, Auger analysis of the specimen in Fig. 2 revealed about 25 at. % Te, a concentration quite adequate to form NiyTe, (Fig. 3). The question of whether grain-boundary compounds were formed was not answered conclusively. There was litile evidence to support the formation of massive precipitates, but the presence of rather large carbides in the grain boundaries made the analysis rather difficult. Generally, the Te was located within a very few atom layers of grain boundary or within a few atom layers of the carbide/bulk interface when a carbide 7500 - i T i T I ) T B 1 i \\ - 7300 - - \ Z 7200 |- \‘ . { N \ , ’fifl - ‘\\ — 70045° \10215° 7008 M*W*Z - ' o Zemperatur (°C/ % S S 700 500 S 400+ &ele ™ Jog 1 L | ] | 1 { 1 g W 20 2 W S0 &0 W i S0 Ar-% e Fig. 3. Phase diagram for Ni/Te system. Source: K. O. Klepp and K. L. Komarek, Monatsh. Chem. 103,941 (1972). ORNL DWG. 77-5132 ) D A OO Iig. 4. Schematic of grainboundary. particle was present in the erain boundary. Sufficient data of the correct type to determine whether very thin grain boundary telluride phases formed or whether the presence of elemental Te (not as a compound) was sufficient to cause embrittlement did not exist. With this picture of Te embrittlement, il became necessary to control the surface Te flux at some level typical of reactor operation. In the MSRE the flux of Te atoms reaching the metal was 10% atoms cm™ 2 sec” !, and this value would be 1010 atoms cm™2 sec™ ! for a high-performance breeder. (These values were calculated based on many assumptions and are given only as order-of-magnitude estimates.) Even the value for a high-performance breeder is very small from the experimental standpoint. For example, this flux would require that a total of 7.6 X 10°® g of Te be transferred to a sample having a surface area of 10 cm? in 1000 hr. Several Te delivery systems based on the vapor pressure of Te metal and the disassociation pressure of several metal tellurides were used, and these systems are shown schematically in Fig. 5. Te metal has a vapor pressure of about 104 tog at 300°C, and one method of exposure involves a Te metal temperature of 300°C and a specimen temperature of 700°C in a sealed quartz vacuum capsule. Several metal tellurides were synthesized; CryTey seemed to have the tellurium partial pressure at 700°C that most closely approximated the desited flux. This telluride was used in a vapor capsule, as a packing media, and in a fluoride salt. The last system appeared most desirable, and several hundred specimens were exposed in 1his manner. The exposed specimens were strained to failure in a room-temperature tensile test, and a longitudinal metallographic section was prepared. An optical system using a filar eyepiece with a position transducer was used to count the depth and frequency of cracking from which several related statistical parameters were deduced. The attractiveness of forming a surface telluride reaction product in preference to having the telluriunm diffuse into the metal can be illustrated by a rather simple calculation. With a tellurium flux in an MSBR of 3.8 X 1019 atoms cm™2 sec” !, the amount of tellurium deposited on the metal surfaces over a period of 30 years would be 2.9 X 1012 atoms/cm?2. 1f the tellurium reacted with the nicke! in the ratio of two atoms of tellurium per three atoms nickel (Ni;Te,), the nickel telluride would be contained in a layer of reaction ORNL-DWG 75-15991R TENSLE SPECIMENS : {l [ noeC Too~C Gty ' ELECTROCHEMICAL ‘ PROBES 1 o . STIRREA -~ GUARTZ JuTTe }/ o ot P e SPECIMENS . e I 1 ’ o i b o | T ulr’l"lri’s < lel phcuing vy, ot VACLILM Vrog s 8 - L areoN eyt /lf - Lk r’: HASTELLOY N ) ! SALY et mnotc ad METAL TELLURIDE ~J—‘r-l—-«:scrc) \<:57L Ta(300%) DIFFUSION PRESSURE CONTROLLED CONTROLLED Fig. 5. Laboratory methods tor exposing metal specimens to tetluriiimi. product 4.7 X 10~ 4 cm thick. Experimental evidence shows that tellurium does not react with nickel in this way, but diffuses preferentially along the grain boundaries. However, the addition of alloying elements to nickel that would form such a surface reaction product is a possible solution to the problem. The most realistic method for exposing metal samples to Te was an irradiated fuel capsule made of the material of interest. Four experiments of this type, designated TeGen-1, TeGen-2, TeGen-3, and TeGen-4 were performed in the ORR at 700°C. Sections of the wall were strained after irradiation, and the severity of cracking was determined by metallographic means. Chemical analyses were performed on samples of the salt and the metal. Teluriuim Embrittlement--Salt Modification The behavior of tellurium in salt containing fission products is complex and open to speculation. Experience with the MSRE indicated that Te was deposited on the metal and graphite surfaces, and that only a very small amount resided in the salt at any given time.? Brynestad and others? suggested that it may be possible to make the salt reducing enough to tie the tellurium up as an innocuous telluride (e.g., “CrTe’’) by a reaction of the type: CrF, + Te + 2UF; -+ 2UF, +“CrTe.” Manning and Mamantov!? performed an electrochemical experiment that indicated that the ratio of UF, to UF; would have to be about 150 to favor the existence of telluride species over elemental tellurium in molten LiF/BeF, /ThF, at 650°C. The work of Toth and Gilpatrick!! sets a lower practical limit for how reducing the salt can be during reactor operation. At reasonable reactor inlet temperatures of 500 to 550°C, the ratio of UF, to UF; must be about 10 to form uranium carbide. These observations lend some support for the approach of altering the oxidation state of the salt, but it is questionable whether the amount of UF, required is practical and whether it is so high as to support the formation of uranium carbide. Recent work by Keiser offered more hope that a stable telluride could be formed at reasonable Ut/U3+ ratios.}2 Keiser used CryTe, as a tellurium source in LiF/BeF,/UF, salt and varied the U**/U3* ratio of the salt by adding NiF, (oxidizing} or Be (reducing). Small tensile specimens of standard Hastelloy N were exposed to the melt for about 200 hr at 700°C at each condition. The specimens were strained to failure and the crack severity determined metallographically. STATUS OF DEVELOPMENT Irradiation Embrittlement--2%-Ti-Modified Alloy Fabrication The 2%-Ti-modified Hastelloy N has been developed to rather advanced stages. About one hundred 2-b lab melts, about twenty 50- to 100-lb commercial melts, and two large {10,000- and 8,000-1b) commercial melts have been made and processed to a large number of products.!3:14 Fabrication experience with these Sw. R. Grimes, Chemical Research and Development for Molren-Salt Breeder Reactors, ORNL/TM-1853 (June 1967). 9] Brynestad, ORNL, personal communication, 1975. °L E. McNeese, Molten-Salt Reactor Program Semiannu. Prog. Rep. Feb. 29, 1976, ORNL-5132, pp. 38-39. - M. Toth and L. O. Gilpatrick, The Equilibrium of Dilure UF 3 Solution Contained in Graphite, ORNL/TM-4056 (December 1972). J R. Keiser, Status of Tellurium-Hastelloy N Studies in Molten Fluoride Salts, ORNL/TM-6002 (October 1977). I E. McNeese, Molten-Salt Reactor Program Semiannu. Prog. Rep. Feb. 29, 1976, ORNL-5132, pp. 4245, ‘1K McNeese, Molten-Salt Reactor Program Semiannu. Prog. Rep. Feb. 28, 1975, ORNL-5047, pp. 60-68. 1] materials was generally quite favorable, except for the occurrence under some conditions of cracking during hot working. Surface cracking occurred in the first heat in almost all products, and the problem was pursued by T. K. Roche (ORNL) and R. W. Bonn (Cabot Corporation).13:16 Gleeble evaluation tests were used to show that the large heat had a much narrower working temperature range than small heats that had been produced previously. The problem was partially solved, and some useful products were obtained. A second heat (8000 lb) was made with only slight modification in the melting practice by the vendor. Although the chemical analysis of the second heat differed only slightly from that of the first heat (Table 1), the second heat had a much wider working temperature range than the first heat (Fig. 6). Experiences showed that the second heat fabricated much better than the first, but the cracking reappeared in drawing products having diameters below 1 in. The problem was partially alleviated by flexing the worked product' to relieve residual stresses before annealing or by using 1120°C as an intermediate annealing temperature rather than the usual 1175°C. Products including welding wire down to 3/32 in. diam, and several sizes of plate and bar were produced. However, the hot cracking problem still exists as a partially unsolved black mark against this alloy. This type of cracking is due to the alloy not having sufficient hot ductility to prevent cracks forming in material that contains residual stresses due fo a prior working operation. This behavior is usually associated with some of the residual elements in nickel. Because, within limits, titanium is not reputed to embrittle nickel-base alloys, it is possible that the hot-working problem is associated with some residual impurity rather than the major alloying elements in this alloy. It is unfortunate, but not too surprising, that fabrication problems were experienced with the first two large heats produced. Clearly this problem must be solved before the 2%-titanium-modified alloy could be produced in large quantities. LST K. Roche, ORNL, personal communication. 16R. W. Bonn, Cabot Corp., personal communication. Table 1. Chemical analysis of production ORNL-DWG 75- 7197 heats of 2%-Ti-modified Hastelloy N7 , I [ Composition (wt %) 2% Ti- MODIFIED HASTELLDY N Flemen o S e Heat Heat : 28104-79017 8918-5-7421°¢ ‘L L ’ 7 ><_“-—-"'"""’ \ Al 0.10 0.12 5 % 7% 5 B <0.002 <0.001 < . /f/ S C 0.06 0.07 & —1 1 % Co 0.02 0.02 Z 6o ! - Cr 6.97 7.10 5 { \ Cu 0.02 <0.01 5 / \\ Fe 0.08 0.06 8 40 ! | A Mn 0.02 0.12 & \ Mo 12.97 11.93 e e ™ i N 0.003 0.007 o T Ni Balance Balance p 0.002 0.004 TEST SPEED: 5in. par sec S 0.002 <0.002 | | | Si 0.03 0.04 1800 1900 2000 2100 2200 2300 °F Ti 1.80 1.90 942 1038 1093 150 1204 1260 °C w 0.01 <0.08 TEST TEMPERATURE Vendor analyses. I'ig. 6. High-temperature ductility of two production £10,000 Ib. heats (2810-4-7901 and 8918-5-7421) of 2%-Ti-modified 8000 Ib. Hastelloy N, determined by Gleeble testing. 10 Weldability Weldability tests were performed on 1/2-n.-thick plate of the small and large commercial heats of the 2%-Ti-modified alloy. Two plates of alloy were welded with filler wirc of the same material with the plates under high restraint. Bend straps were cut from the platc and subjected to bend tests according to specifications of the American Society of Mechanical Engineers. All heats, including the two large commercial heats, passed the tests and were noted by the welders to have excellent weldability. One important observation was made concerning postweld heat treatment. A postweld heat treatment of 2 to 8 hr at 870°C was established for standard Hastelloy N, and this treatment was sufficient to recover the weld properties to those of the hase metal. Qur limited studies indicated that welds in the 2%-Ti-modified alloy must be annealed at 1175°C (the normat solution annealing temperature) to bring the weld properties back to those of the base metal. Creep Strength Creep tests were run on most of the heats of the 2%-Ti-modified alloy. The stress-rupture properties are compared in Fig. 7 with those of standard Hastelloy N, and the Ti-modified alloy is superior at all three temperatures. The minimum creep rate of the Ti-modified alloy is compared in Fig. 8 with that of standard Ilastelloy N. The modified alloy is superior at 650°C, and standard Hastelloy N is slightly superior at 704 and 760°C. Thus, the modified alloy has adequate creep strength. ORNL-DWG 77-9222 ORNL-DWG 77-9223 TO o VeI T T T T T T 70 T T T T T T N — STANDARD HASTELLOY N o7 60 \\ — STANDARD HASTELLOY N 60 | ~=2% Ti MODIFIED HASTELLOY N ,’/ B ” "~ 2% Ti MODIFIED HASTELLOY N 50 % 40 — Z - v i @ 0 & a5 . - x 530 o | ol s | \ I | | | | J \ | | ol vvnd s J_J_.L_ilJ_ o b vl vl il 10! 10 103 107 1072 T 10° RUPTURE TIME (hr} MINIMUM CREEP RATE (%/hr) Fig. 7. Comparison of the stress-rupture properties of Fig. 8. Comparison of the creep strengths of standard standard and 2%-Ti-modified Hastclloy N. and 2%-Ti-modificd Hastelloy N. Alloy Stability Samples of Ti-modified alloys were aged to 10,000 hr, but very few were tested. However, the few specimens that were tested had excellent properties and did not show any evidence of deterioration of properties with aging. Roche prepared several experimental alloys to cvaluate the influence of Al and Ti content on the formation of gamma prime in this alloy base. This question came up because of the use of 11 Al for deoxidation of the melt. Rochel7 found at 650°C that this alloy must contain more than 2.7% Ti to form v' and that an alloy containing 2% Ti could contain about 0.6% Al before forming +'. Thus, the composition of the 2%-Ti-modified alloy is such that it is a comfortable distance away from the phase boundaries of embrittling ¥’ Salt Corrosion Because titanium is readily converied to a fluoride by salt, the addition of this element to the alloy raises the question of accelerated corrosion by fluoride salt. Actual measurements in thermal-convection loops have indicated that the corrosion rate of the Ti-modified alloy is lower than that of standard Hastelloy N.I8 Two factors are involved. First, the iron content of the modified alloy is only a few tenths of a percent, whereas the iron content of the standard alloy is 4 to 5%. Because iron is almost as easily oxidized by the salt as chromium, iron contributes significantly to the corrosion rate in salt of reasonable oxidation potential. Thus, the lower iron content of the modified alloy would significantly reduce the corrosion rate of the modified alioy. The second factor is that titanium is not accessible to the salt unless it can diffuse to the alloy surface. Measurements of the diffusion rate of titanium in Hastelloy N showed that it moved about a factor of 10 more slowly than chromium.!? Thus, it is quite reasonable that the modified alloy with its lower iron ¢ontent and its 2% titanium corrodes at a slower rate in salt than does standard Hastelloy N. Postirradiation Mechanical Properties Many specimens of the 2%-Ti-moditied alloy base have been irradiated in the ORR to fluxes of thermal neutrons of 2 to 3 X 1029 ¢m™2 and have been subjected to postirradiation creep testing. These tests have shown that the 2%-Ti-modified alloys are very resistant to irradiation embrittlement at 700°C and lower, but that most heats of the modified alloy are embrittled at 760°C. The postirradiation creep properties at 650°C of several heats of the 2% modified alloy are compared in Figs. 9 and 10 with those of standard Hastelloy N. The modified alloys are far superior to standard Hastelloy N in all regards. The data for three heats are shown in Fig. 11, where irradiation teruperature is evaluated as the test variable. These data show that the properties deteriorated as the irradiation temperature was increased. The 2%-Ti-modified alloy has good properties after irradiation at 650°C, acceptable properties after irradiation at 704°C, and unaccepiable properties after irradiation at 760°C. Although there is a plimmer of hope that the alloy can be made stable at 760°C, it does not appear very promising. Reactor systems that operate at 650°C have good thermal efficiency, so there seems little incentive to operate at 700 to 760°C. However, an alloy capable of operation at these temperatures would give more margin {for temperature excursions. Microstructural Features Microstructural studies showed that the carhide formed in the 2%-Ti-modified alloy was of the MC type, but the tine points of the behavior of these carbides were not studied sufficiently. Braski only began to scratch the surface in his efforts Lo produce a more homogeneous structure.?? Bands of carbides form in this alloy (as they did in standard Hastelloy N} during early fabrication, and the resuit is that there are alternative regions of high and low densities of carbides. 171, E. McNeese, Molten-Salt Reactor Program Semianru. Prog. Rep. Feb. 29, 1974, ORNL-5132, p. 51. V8], Koger, Alloy Compatibility With LiF-Bel 5 Salts Contalning ThE yand UF , ORNL/TM-4286 (December 1972). 19 ¢ E. Sessions and T. S. Lundy, J. Nucl. Mater. 3/(3), 316 (1969). 20p N, Braski and J. M. Leitnaker, Production of Homogeneous Titanium-flastelloy N Alloys, ORNL{TM-5697 {February 1977). 12 ORNL-DWG 75-7204 70 T T T T TTTTT T 1177 60 A7{—583 =, ~ 9.4 10.8 "‘“\\ A o 19.88 19.8 50 T ,8-4 12.5 T ase Wi — . 71 1.8 — S~ o Y g.g 8.1 = 40 \\\ 6.7 fi%f‘: 5.3 5580203 —— 472 - A \""-.., 129088, A11.9 :.‘_:J 30 STANDARD HASTELLOY, IRRADIATED ~~_ %m\ » | [ ST HEAT NUMBER 20 o 471114 | 470 727 A 471-583 O 474 -533 O 472-503 & 474 -534 ® 473008 O 474 -539 10 | & 474-301 $474~-535 T T e — 0 | IIIIlJ_l‘ L L Lot Lot Nl 107! 100 10! 102 103 10t RUPTURE LIFE (hr) Fig. 9. Postirradiation stress-rupture properties at 650°C of several heats of Ti-modified Ttastelloy N. Samples were irradiated at 650°C to a flux of thermal neutrons of 3 X 1029 ¢m™ 2. The numbers by the individual data points are the fracture strains in percent. The solid lines are for the heats in the unirradiated condition. ORML-DWG 75— 7203 70 . T T 7T T T T 1T TR el 11T HEAT NUMBER 471-583 6o l— 0a7i-114 pd & 471 - 583 o 472~ 503 oA ® 473~ 50 |— ?:5dooa . / B 470727 / /{ a o 2 a0 4 .- 0 @ /A & 3 30 $ P ~ STANDARD HASTELLOY N 20 — IRRADIATED AND UNIRRADIATED 10 0 L LA Lo i Ll L iy Lo L 107* 1073 1072 107! 1P 10! MINIMUM CREEP RATE (% /hr) Fig. 10. Postirradiation creep properties at 650°C of several heats of Ti-modified Hastelloy N. Samples were irradiated at 650°C to a flux of thermal neutrons of 3 X 102% ¢cm™2. The solid lines are for heats in the unirradiated condition unless noted otherwise. 13 ORNL—DWE TS5~ 153R - ']' YT r | i : STANDARD HASTELLOY N | T gy L IRRADIATED AT B30°C. -1 4o 1 , b N ! ! «‘fi.”'_‘:\._. AAAAA . ) . -~ Tt - cdeelemis P | | = a L ; . S5 a0 — 2 0 T1-M4 (196 % Ti) LA TI-583 (479 % Ti) . _Q f ‘ 1 2 - el St O 8 72503194 % Ti) ] © 450 (2% T i } ‘ | 10 |- OFPEN POINTS IRRADIATED AT 650°C _ . EILLED POINTS IRRADIATED AT 760°C %mmmao HASTELLOY N CROSSED OFEN POINTS IRRADIATED AT 704°c IRRADIATED AT 760 °C e e L 10t 2 5 409 2 5 100 2 5 10 2 5 107 2 5 10° RUPTURE TIME (hr) Fig. 11. Postirradiation creep propertes of Ti-modified alloys at 650°C after irradiation at indicated temperature to a flux of thermal neutrons of 3 X 1029 ¢ 2. Numbers by points indicate fracture strains. Arrows indicace that test was still in progress when figure was made. The practical aspects of the influence of carbide structure can be appreciated by examining samples of two heats of material that responded quite differently to irradiation. Comparative transmission electron micrographs are given for heats 71-114 and 72-502 in Figs. 12 to 14. After irradiation of 650°C there is not an appreciable difference in microstructure, but as the irradiation temperature was increased the carbides remained much finer in heat 72-303 than they did in heat 71-1 4. Similayly, the properties of heat YE-11087 (a} ) Fig. 12. Flection micrographs of two alloys after irradiation at 650°C for 1200 hr. (2) Alloy 563. (b) Alloy 114. - ;,fflwmwbx% Lo e Iig. 14. Electron micrographs of two alloys alter irradiation at 760°C tor 1200 hr. () Alloy 503.(h) Alloy 114, 72-503 were far supetior to those of heat 71-114. Preirradiation annealing had a marked influence on the properties of this alloy. As shown in Table 2, it was possible to cause dramatic changes in the postirradiation creep properties by very small changes in the preirradiation annealing tenmperature. Further study of the precipitation morphology and kinetics is needed for this alloy. These studies are not likely to result in any surprises, but simply will reveal methods for controlling properties by closer control of heat treating during processing. 1S Table 2. Influence of preirradiation annealing treatment on the postirradiation creep properties of modified Hastelloy N (heat 471-114) Irradiated at 760°C and tested at 650°C. . Rupture Minirhum Total Test Preirradiation Stress . creep . no.? anncal? (107 psi) i}n? rate elunia tion (hr) (@) (%0) 1691 A 33 4.7 (0.014 0.33 1712 A 30 6.0 0.010 048 1736 A 25 96.2 0.0022 0.78 1692 B 35 0.7 0.20 0.22 1746 B 30 34 0.029 0.36 1719 B 23 171.5 (.0009 1.8 1683 C 35 1.6 0.17 2.3 1706 C 30 23.6 0.029 3.6 1786 C 3¢ - - 1715 C 25 1463.6¢ 0.0G42 8.6 1792 C 20 2850.0 0.0010 8.1 1777 D 47 27.3 0.090 7.1 1737 D 40 223.0 0.011 36 1693 i 35 663.3 06.0044 4.0 1694 E 35 0.3 0.3 2.7 1720 E 25 60.0 0.0028 0.35 17594 E 15 2782.t 0.0006 2.1 IR series. PA = annealed 1 hr at 1038°C in argon; B = anncaled 1 hr at 1093°C in argon; C = annealed 1 hr at 1177°C in argon; D = annealed | hrat 1204°C in argon; £ = annealed 1 hy at 1260°C in argon. “Priscontinued before failure. Irradiation Embrittlement—Nb-modified Alloys Fabrication Because work began later on the Nb-modified alloys than on the titanium-modified alloys, the fabrication experience was not as extensive. About fifty 2-Ib laboratory melts containing up to 4.4% Nb were melted and fabricated into 1/4-in.«diam rod, and five commercial melts about 50 lb each were melted and fabricated into 1/2-in.-thick plates. These alloys fabricated well by using the same annealing and working temperatures used for standard Hastelloy N. Weldability Test welds (gas tungsten arc) were made in all five of the commercial Nb-containing heats. They were prepared by using the same welding parameters used for standard Hastelloy N. Bend specimens from all heats bent without evidence of cracking. Thus the weldability of these materials appears excellent. Creep Strength The creep testing program on the Nb-modified alloys was still in the beginning stages, but sufficient data were obtained to show that Nb had very beneficial effects. The influence of Nb content on the 100-hr rupture stress is illustrated in Fig. 15. The properties ot a typical heat of standard Hastelloy N are shown by the horizontal lines on the right-hand side of the figure. At 650°C the effects of Nb are linear up to about 1.5% Nb, and further additions have a lesser effect. Irradiation at 650 and 704°C with subsequent testing at 16 ORNL-0OWG 77-5144 60 50 100 hr RUPTURE 40 e STRESS (ksiy =0 20 — Fig. 15. Intluence of Nb content on 10 | }\] the stress-rupture properties of modified 0 1 2 3 4 5 STD Hastelloy N at 650°C. Nb CONTENT (%) 650°C resulted in progressively lower strengths, but the strength variation with niobium content paralleled that of the unirradiated material. Irradiation at 760°C resulted in drastic degradation of properties for all alloys except those containing 3.5% or more Nb. The influence of Nb on creep strain is shown in IFig. 16. Niobium additions up to approximately 1.5% have a large strengthening effect with a much lesser effect at higher niobium concentrations. Alloys containing 0.5% or more Nb are stronger than standard Hastelloy N. The fracture strain is the parameter of most concern. The influence of Nb on this parameter is shown in Fig. 17. In the unirradiated condition Nb has a beneficial effect up to about 3.3%, and the ductility diminishes at higher Nb concentrations. The ductility of specimens irradiated and tested at 650°C was the same as that of unirradiated specimens. Irradiation at 704 and 760°C resulted in much lower fracture strains, but niobium had a beneficial effect. Thus, these data show that niobium improves the resistance of Hastelloy N to irradiation embrittlement, but that the alloys are likely not useful at operating temperatuyes much above 650°C. ORNL-DWG 77-3145 55 — 56 UNIRRADIATED, UNAGED STRESS IRRADIATE D TO PRODUCE a5 i : 1% STRAIN F IN 100 hr 40 — {(ksi) 35 m— Fig. 16. Influence of Nb content on 30 I 1 I I \/\—m the creep properties of modified Has- o 1 2 3 4 STD telloy N at 650°C. Nb CONTENT (%) 17 ORNL-DWG T7-54486 CREEP 20 STRAIN v 15 | UNIRRAD, foohr 650°C ot CREEP TEST 704°C (%) 760°C 0 I . 1 l \J\ Fig. 17. Influence of Nb content on 0 1 2 3 4 S STD the fracture strain of modified Hastelloy Nb CONTENT (%) N at 650°C. Salt Corrosion Nicbium forms a very stable fluoride, so Nb would be removed from the alloy by the satt. However, the process would be limited by the diffusion of Nb to the specimen surface. It is very unlikely that Nb diffuses more rapidly than Cr; so, at worst, the presence of 1 to 2% Nb can be likened to increasing the chromium content from 7% to 8 to 9%. Limited corrosion measurements in thermal convection and forced circufation loops indicated very good corrosion resistance of Nb-modified alloys.2l Although further corrosion measurements need to be made, it is very unlikely that corrosion in salt will pose a problem with the Nb-modified alloys. Microstructural Features Very limited microstructual studies showed that the carbides in the Nb-rnodified alloys bebave qualitatively the same as those in the Ti-modified alloys. Solution annealing at 11757C seemed to be nearer a true solution anneal for the Nb-modified alloys, with most of the carbon being in solid solution at this temperature. During subsequent exposure at typical service temperatures {e.g., 5350°C) fine carbides of the MC type precipitated. Fig. |8 shows the precipitates that formed at 650°C, and Fig. 19 shows how 21y, R. Keiser, Compatibility Studies of Potential Molten-Salt Breeder Reactor Materials in Molten Salts, ORNL/IM-5783 (May 1977). (1179 {(171) (M1 [022] [615] Fig. 18. Unstressed portion of 1%-Nb-modified Has- Fig. 19. Specimen of 1%-Nb-modified Hastelloy N telloy N solution annealed at 650°C for 1175 hr. Original solution annealed and siressed at 40,000 psi and 650°C for magnification: 25,000X. 11735 hr. Original magnification: 25,000X. 18 dislocations interacted with these precipitates in a stressed sample. These precipitates would also act as very effective pinning sites for He. The precipitates were extracted from some of these specimens and were found to consist of two face-centered cubic carbides having lattice parameters of 4.280 and 4.175 A. X-ray fluorescence analyses of the extracted precipitates showed the presence of Mo, Nb, Cr, Ni, and Zr in the order of decreasing concentration. By transmission electron microscopy it was found that both carbides were epitaxially oriented with the matrix the [1 1 1] planes of the carbide and of the matrix being parallel to each other.??2 Irradiation Embrittlement—Ti/Nb-imodified Alloys Alloys containing Nb and Ti were not pursued adequately to obtain a very complete understanding, but they had a number of interesting features. Alloys with small concentrations of Nb and Ti (e.g., 1% each) had very attractive pre- and postirradiation properties. They were strong and had fracture strains from 10 to 25% in the irradiated condition. Alloys that contained higher amounts of Nb and Ti (e.g., 2% each} were very strong. They formed a stress-induced precipitate, probably ', that made them strong with moderate ductility in the unirradiated condition. Some of the features of these alloys were presented previously.23 In the irradiated conditions these alloys were strong but brittle. Alloys with medium concentrations of nicbium and titanium (e.g., 1% each) have very attractive properties in the irradiated and unirradiated conditions. Alloys with higher amounts of Nb aunid Ti are very strong and may be useful for nonnuclear applications. In the medium-concentration alloys no ¥’ was found, and the strengthening was attributed to combined solid solution and carbide strengthening mechanisms. In the higher concentration alloys, ¥' was formed and its formation was accelerated by stress. These alloys are very complex and were not studied sufficiently to fully understand the phase relationships. Tellurium Embrittlement--Screening Tests on Ti/Nb-medified Alloys Approximately 50 experiments involving sonie 2000 specimens were performed in which several alloys were exposed to tellurium. Several of these experiments were performed in learning how to expose specimens under reasonable conditions, and they did not provide useful materials information. The most realistic and reproducible method involved exposing metal specimens to salt containing CryTey and CrgTeg at 700°C. This method is shown schematically in the right-hand side of Fig. 5. The last specimens examined from this experiment were exposed for 2500 hours. Selected specimens from this experiment illustrate our general findings in this program. Photomicrographs of these specimens are shown in Fig. 20, and the chemical analyses are given in Table 3. Standard Hastelloy N [Fig. 20(¢}] was embritiled under these conditions. Type 304 stainless steel [Fig. 20(b)] was roughened because of corcosion by the salt, but there is no evidence of intergranular embrittlement. Hastelloy S [Fig. 20(c)} and Inconel 600 [Fig. 20(d)] both contain nominally 15% of Cr, but they were embrittled by the tellurium. Hastelloy N modified with 1.75% Ti was embrittled [Fig. 20(¢)] . Alloy 506 contained 14.5% Cr [Fig. 20(f)] and was embrittled as severely as Hastelloy N containing 7% Cr. Alloys 516 [Fig. 20(g)], 421543 [Fig. 20(h)], 517 {Fig. 20())], 525 [Fig. 20}, and S28 [Fig. 20(k)} contained <0.10, 0.70, 1.10, 1.5, and 4.4% Nb, respectively. The severity of 22D N. Braski, ORNL, personal communication, 1976, 234, F. McCoy, Ir., R. E. Gehlbach, and B. McNabb, “Development of New Nickel-Base Alloys for High-Temperature Service,” pp. 245-58 in Space Shuitle Materials, National SAMPE Techaical Confercnce (proceedings), Society of Acrospace Matcerial and Process Engineers, Huntsville, Alabama, Oct. 5-7, 1971. ¥.140918 » o {a) ¥-140917 bi | Y-140916 j ic) % ¥-340923 SR {d) {a! 13 ‘!(?Ci E?O | MICRONS 6(?0 ?O;O O.OTO:S O.\’}HO INCHES O.dZO 0.0i25 Fig. 20. Photomicrographs of several alloys exposed to salt containing CrgTeg and CryTe, for 2500 hr at 700°C and strained to failure at room temperature. (z) Standard Hastelloy N. (b) 304 stainless sieel. (¢) Hastelloy S. (@) inconel 600. (e) Alloy 474533, (f) Alloy 506. ! N v-140919 {h) Y-140926 ¥-140929 Y-140927 {1 fiCl)O 2C1JO | MICRONS 6Ci)0 ?(l)O f ¥ - I I 0.C05 0.010 INCRES 0020 0.025 Fig. 20 (concluded). Photomicrographs of several alloys exposed to salt containing CrgTe, and Cr3Te, for 2560 hir at 700°C and strained to failure at room temperature. (g} Alloy 516. (h) Alloy 421543, (/) Alloy 517. (/) Alloy 525. (k) Alloy 528. (/) Alloy 518. : Alloys werc exposed to sait containing Cryle, and CryTe; for 2500 hr at 700°C. Table 3. Chemical anatyses of alloys Concentration of element? (wt % Alloy Ailloy no. Ni Mo Cr Fe Mn C Si Ti Nb Al Co Cu La Ni W Type 304 stainless steel 139969 9.3 b 18.8 Bal 1.51 0.027 0.45 b b b fnconel 600 NX6372G Bul - 15.13¢ 9.17¢ 0.27¢ 0.10° 0.09¢ 0.31¢ Hastelloy S 476316 Bal 14.16° 15.05€ 0.47¢ 0.52¢ 0.003¢ 0.38¢ 0.27¢ 0.18¢ Hastelloy N 405065 Bal 16.0 7.1 4.0 0.55 0.06 0.57 <0.01 b 0.03 Modified Hastelloy N 474533 Bal 11.37 7.3 0.03¢€ 0.04 0.091 0.15 1.75 b 0.54 0.03¢ 0.14¢ 421543 Bal 124 7.31 0.038 0.08 0.05 0.014 0.003 070 0.02 411 Bal 11.71 6.78 b 0.1 0.043 b b 1.15 b 413 Bal 11.82 6.75 b 0.1 0.045 b 0.9 1.13 b 506 Bal 11.97 4.5 b 0.22 0.037 b <0.01 <0.02 0.10 516 Bal 10.58 7.3 <0.05 0.20 0.049 comwhoonZfl:wrfiacowmwwom 100 101 . E. McCoy, Jr. . J. McHargue . McNabb . E. McNeese . . MacPherson . E. MacPherson Mamaniov L Manning . R. Martin . L. Matthews . Maya . Patriarca . Postma . K. Roche . W. Rosenthal . C. Savage . J. Skinner . M. Slaughter . P. Smith Spiewak O. Stiegler E. Thoma B. Trauger Y. Valentine M. Weinberg R. Weir C. White . L. Youngblood 102. Research and Technical Support Division, ERDA, Oak Ridge Operations Office, Post Office Box E, Oak Ridge, Tenn. 37830 103. Director, Reactor Division, ERDA, Oak Ridge Operations Office, Post Office Box E, Qak Ridge, Tenn. 37830 104--105. Director, Division of Nuclear Research and Applications, ERDA, Washington, D.C. 20545 106--216. For distribution as shown in TID-4500 under UC-76, Molten-Salt Reactor Technology category (25 copies—NTIS) -+ US.GOVERNMENT PRINTING OFFICE : 1977-748-189/356