Boshoku Gijutsu, 28, 82-95 (1979) Report Inconel Alloy 690-A New Corrosion Resistant Material* and A. Cordovi***M. Schultz** J. W. J. A. Sedriks**, Inco Research & Development Center ** *** Inco Limited INCONEL Alloy 690 (Ni-30Cr-9Fe-0.02C, wt. %), is a new high-chromium nickel-base alloy, designed to provide a stable austenitic structure combined with resistance to corrosion and oxida- tion in aggressive industrial environments at elevated temperatures. Extensive laboratory evalua- tion of Alloy 690 during the past 12 years has demonstrated remarkable resistance to sensitization, and to stress corrosive cracking in chloride environments. In high temperature water, this material resists stress corrosion attack over a wide range of tepmeratures, oxygen contents in the presence of crevices, and lead contamination. Alloy 690 has been found to release only a negligible amount of material when exposed to high velocity water at elevated temperatures-a fact that may be of major importance in light water reactor technology. 1. Introduction INCONEL® alloy 690 is a new austenitic nickel-base alloy that was developed to provide excellent corrosion resistance in many demanding high temperature environments. In particular, it shows promise for applications involving high temperature water g., (e. nuclear steam generators) and nitric acid environments g., (e. pickling and nuclear fuel reprocessing). Preliminary results were presented in a paper by Copson et al, at The Fifth International Congress on Metallic Corrosion in Tokyo, May, 1972 (See Fig. 1. The Ni-Cr-Fe phase diagram showing page 376 of proceedings). the location of the r/r+a' solvus from Alloy 690 now is in full-scale production and has 816 1260C. N been fabricated into many forms, including sheet, plate, bar, pipe, tube-shells and thin-wall tubing. Table 1. Effect of high temperature exposure on This paper brings together some of the available the impact resistance of mill annealed data on the corrosion characteristics of alloy 690. alloy 690 It also provides a brief description of its physical metallurgy and mechanical properties. 2. Structure and Properties Alloy 690 is a single phase austenitic alloy with a nominal composition of 60% Ni-30 Cr-9 % Fe-0.02% C. This composition places it well within the austentte (r) field at 816C****, as shown in Figure 1. The a' phase identified in Figure 1 is a chromium-rich phase, very similar to the iron-chromium o phase both in morphology and hardness. It can cause embrittlement when * Presented at Meeting of Stress Corrosion Com- precipitated during prolonged high temperature mittee, Japan Society of Corrosion Engineers, exposure in alloys whose compositions lie in the April 28, 1978 ** Suffern, Y. A N. S. U. r+a' field. Since the composition of alloy 690 *** New York, Y. A. N. U. S. lies well away from the r+a' field, precipitation of **** 816C is the lowest temperature at which the the embrittling a' would not be expected. This is r/r+a' solvus has been determined (1). confirmed by the impact properties listed in Table Vol. 28, No. 2 83 1. These data show that the toughness of alloy 690 virtually is unaffected by 12,000 hour exposures at temperature in the range 760C. N 566 It should be noted that impact resistance is particularly sensitive to precipitation of hard and brittle phases such as a' and c. For example, room temperature impact resistance for Type 310 stainless steel can drop from 270J to less N than 14J due to the forma- tion of a phase at 700C. N Carbide precipitation can take place in alloy 690 provided the carbon content is high enough. The carbide in alloy 690 is Cr23CB. Although carbon solubility has yet to be totally defined, experience indicates that heating to 1150C will dissolve all the carbide in a 0.04%C material. As a result, Fig. 3. Effect of cold work on Tensile properties sensitization due to chromium depletion possible is of 12.7mm hot rolled plate, annealed at 1205C/i h/WQ. on slow cooling. However, if the carbon content is kept at 0.02% or below, alloy 690 cannot be sensitized to the Huey test. Also, levels of 0.02 C can be readily achieved with the argon-oxygen decarburization process used in the refining of alloy 690. The fully austenitic structure of alloy 690 means that it can be fabricated by the same mill practices used to hot and cold work other commercial high- nickel alloys such as INCONEL alloy 600. Be- cause of its higher chromium content, alloy 690 work hardens slightly more than alloy 600, as shown in Figure 2. However, this presents no problems in producing heavily worked products such as sheet and thin-wall tubing. The effect of Fig. 4. Short-time elevated temperature Tensile cold work on the room temperature mechanical properties of 3.8mm cold rolled strip, properties of alloy 690 shown is in Figure 3. Both annealed at 1038C. the strength and ductility respond as expected to increasing cold work. Increasing the tensile test temperature decreases the tentile and yield strengths (Figure 4); however, temperatures in excess of 600C are needed to cause a significant decrease in strength. At steam generator temperatures of 316C, N alloy 690 ex- hibits high strength properties which are insensitive to changes in temperature in this region. A de- tailed description of creep and stress-rupture prop- erties for alloy 690, is given in Reference 2. Regarding welding, alloy 690 can be joined using alloy 600 welding products; also a filler of match- ing composition has been developed for gas tungsten-arc and gas metal-arc welding, as well as a covered electrode. These can be made com- mercially available when demand warrants it. There is no evidence of any impairment of stress corrosion resistance due to welding with gas Fig. 2. Work-hardening Alloy 690, Alloy 600 and tungsten-arc welds prepared with a matching filler. Nickel 200. 84 Boshoku Gijutsu Table 2. Environments used to evaluate the corrosion behavior of Alloy 690 Table 3. Composition of Alloy 690 heats used in Table 4. Corrosion rates of Alloy 690 in various corrosion studies nitric-hydrofluoric acid mixtures * Average for duplicate specimens tested at 60C. enviroments known to cause general, localized or stress corrosion attack in other alloys. The com- positions of alloy 690 which were investigated are listed in Table 3. 3.1. HNO3-HF Acid Solutions The reason for testing in mixed nitric-hydro- fluoric acids was to evaluate the potential of alloy 3. General Corrosion 690 as a vessel material for processing spent As indicated in Table 2, the corrosion resistance nuclear fuel elements as well as for the tanks of of alloy 690 has been evaluated in a number of pickling baths. Because of its high chromium con- Vol. 28, No. 2 85 the most severe superheating imaginable on the secondary side of pressurized water reactor (PWR) steam generators. When the temperature of a layer of water at a heat transfer surface exceeds the tem- perature of the main body of slightly alkaline water by a temperature difference AT, a high con- centration of NaOH can develop in this layer3. The maximum attainable d T on the secondary side of PWR steam generators is not known. The 50% NaOH level selected for testing is equivalent to a T of d some 27C (80F), which may be higher than that encountered in practice. The 50% NaOH Fig. 5. Descaled weight loss after 2-week ex- test data, both with regard to general corrosion posure to deaerated 50% NaOH at 316C and stress corrosion cracking, must therefore be (duplicate specimens, Alloy 690 from heat regarded as representing a theoretical extreme. NX10C1H). descaled loss and The number weight of a alloy 690 of other alloys after exposure in deaerated 50 NaOH at 316C is shown in Figure 5. The cor- rosion scales were removed using the alkaline permanganate-acid method4. The results, which are plotted on a logarithmic scale, show that alloy 690 exhibits a corrosion rate intermediate between that of alloy 800 and alloy 600. 3.3. Undeaerated 50% NaOH at 300 and 332C Descaled weight loss resulting from exposure of the same alloys to undeaerated 50% NaOH at 300 and 332C is shown in Figure 6. For these speci- Fig. 6. Descaled weight loss after exposure to mens, the weight loss was determined using the undeaerated 50% NaOH (single speci- electrolytic-sulfuric acid descaling method4'. mens, Alloy 690 from heat Y24A7L). Again, the results show that alloy 690 exhibits a tent, the alloy exhibits excellent resistance to vari- corrosion rate between that of alloy 800 and alloy ous nitric-hydrofluoric acid mixtures, as shown in 600. Table 4. 3.4. Deaerated Sodium Phosphate Solutions at 3.2. Deaerated 50% NaOH at 316C 275 and 325C Testing in high temperature 50% NaOH en- Corrosion test results of alloy 690 and other alloys vironments helped to establish the corrosion re- in high temperature sodium phosphate solutions sistance of alloy 690 under conditions simulating have been recently published by Pessal et g al. and Table 5. Corrosion of various alloys in high temperature sodium phosphate solutions 86 Boshoku Gijutsu Table 6. High velocity loop test parameters * Adjusted with ammonia. are reproduced in part in Table 5. The extent of defined in MACE standard TM-02-74, covering attack on the various alloys is dependent on the the following terms: phosphate concentration and the Na/P ratios, with 1. "Descaled metal loss" (metal consumed): the higher nickel alloys generally exhibiting greater The difference between initial weight and resistance to attack than Type 304 stainless steel. weight after removal of adherent corrosion 3.5. Deaerated Ammoniated and Borated water film. at 316C 2. "Corrosion film weight" (adherent corro- In pressurized water reactors, a factor of parti- sion film): The difference between the cular interest is the corrosion resistance of steam weight before descaling and the weight generator tubing materials in flowing primary after descaling. water. Material lost to the stream can become The difference between the descaled metal loss radioactive by coming in contact with the reactor and the corrosion film weight represents material core and then redeposit on the primary circuit surfaces. A recent evaluation of the corrosion (a) AMMONIATED WATER, 1000 HOURS behavior of alloy 600 and alloy 800 in a CANDU ○ Descaied Metal Loss reactor outlet autoclave during pressurized lithiated ◎ Corrosion Film Weight ● MGteriol Lost to Sfream operation has shown that these two alloys exhibit comparable corrosion resistances. The corrosion resistance of alloy 690 under these and similar primary coolant conditions is not known. In an attempt to obtain some insight into how alloy 690 is likely to behave, a laboratory test program was carried out employing a high velocity test loop. Alloy 690 was evaluated in two simulated primary water environments, referred to here as "am- moniated" and "borated". The test parameters (b) BORATED WATER, 2250 HOURS in these environments are shown in Table 6. The alloy 690 test specimens and those of con- trol materials were heat treated for 0.5h at 980C/ AC. The surfaces were prepared by grinding on a wet 120 grit silicon carbide belt to a 0.75pm finish. The specimens were weighed and then ex- posed for extended periods of time under the con- ditions shown in Table 6. After exposure, the specimens from the ammoniated water tests were descaled by cathodic charging in 5% HZS04 in- hibited with quinoline ethiodide4. The specimens Fig. 7. Effect of chromium content on material from the borated water tests were descaled by the lost to stream in 316C deaerated water alkaline permanganate acid method4. The method flowing at a velocity of 5.5 m/s. Alloy 690 of reporting results and terminology adopted are from Heat Y24A7L. Vol. 28, No. 2 87 lost to the stream. Alloys showing the least loss 690 determined by the Huey test was found to be of material to the stream would be expected to critically dependent on the carbon content of the produce less activity on the primary water side of alloy. As shown in Figure S, alloy 690 containing steam generators. 0.02% C shows very low corrosion rates in boiling The behavior of alloy 690 in ammoniated and borated water shown is in Figure 7. It can be seen that alloy 690 loses less material to the stream than any of the other materials tested. The films formed on alloy 690 as a result of the ammoniated and borated water exposures were of the thin tarnish type and appeared to be extremely adherent to the metal. The results suggest that the crud release rates of alloy 690 in an operating steam generator could be very low. The use of this alloy in steam generators could therefore reduce the amount of activity built ▲ 3h/595C up on the primary side. This is assuming that an ● 3h/650C ■ 3h/705C important source of the active species is the ma- terial being corroded from the tube wall. A note of caution is warranted here; the data shown in Figure 7 are subject to a standard error of about 25% and should be regarded as a guide rather than Fig. 8. The effect of carbon content on the cor- a precise measurement. rosion rate of Alloy 690 in the huey test. Material annealed 1h/1150C/WQ before 4. Localized Corrosion sensitizing. 4.1. Intergranular Corrosion Tests in Boiling 65% HNO3 Table 7. Corrosion rates of Alloy 690 welded The intergranular corrosion resistance of alloy with matching filler in boiling 65 690 was evaluated in boiling 65% HN03. This nitric acid test, known as the Huey test, is standardized as ASTM A262, Practice C. It is a very severe test which for stainless steels attacks chromium de- pleted areas, carbides, and a, when present. The intergranular corrosion resistance of alloy Fig. 9. Time-temperature-sensitization diagran for Alloy 690 in boiling 65% Nitric acid. Heat T-57177 containing 0.05% c, solution annealed 1 h/1150C/WQ. 88 Boshoku Gijutsu Table 8. Polythionic acid and nitric acid test results a) Test duration=720h; a-bend test specimens b) Average of five 48-h periods C) Cracks within 1h, qualifying test * Heat Y24A7L 65% HN03, irrespective of heat treatment. For 4.3. Pitting Tests in Chloride Solutions the 0.02 and 0.03% C alloys, the corrosion rates Alloy 690 specimens from heats NX10C1H, represent the average values of consecutive five 48- NX4460H, and NX4458H (see Table 3) were ex- hour exposures, whereas for the 0.05 and 0.07%C posed for 96 weeks at 316C in water containing alloys, the weight loss is an average of two 48-hour 500 ppm chloride, and daerated with N2H4, to exposure periods. The extremely low corrosion determine whether pitting would occur. No evi- rate of the 0.02% C alloy, together with the fact dence of pitting or any other type of attack was that no detectable intergranular attack could be found using optical and scanning electron micro- detected by conventional metallographic examina- scopes at magnifications up to 3000X. tion of sections after corrosion testing, suggest that the 0.02%C alloy resistant is to sensitization. 5. Stress Corrosion Cracking Alloy 690 containing 0.02 and 0.03% C was also 5.1. Chloride Environments examined in the welded condition. Test coupons, Alloy 690 has been tested for resistance to chlo- 0.5cm thick, were butt welded using the matching ride stress corrosion cracking in a wide variety of filler and exposed for five 48-hour test periods to high temperature chloride environments. The con- boiling 65% HN03. Again, as shown in Table 7, trols used in these tests were alloy 800 and Type 304 very low rates of corrosion were found, with the stainless steel. Both of these alloys are known to only attack observed being light interdendritic crack in elevated temperature chloride environ- etching of the matching filler. ments, with alloy 800 being more resistant than, Higher carbon versions (e.g., 0.05% C) of alloy Type 304 stainless steel. In all cases, alloy 690 690 can be furnace sensitized to give significant was tested for periods significantly in excess of intergranular attack in boiling nitric acid, as shown those required to crack alloy 800. The following in the time-temperature-sensitization diagram in tests were employed: Figure 9. Effects of welding on the sensitization 1. Boiling MgC12, 154C, 9 weeks. 13. behavior of such higher carbon materials still need 2. Undeaerated water +500ppm chloride to be evaluated. (NaCI), 260C, 18 weeks. 4.2. Polythionic Acid Tests 3. Vapor phase above environment No. 2, 8 In conjunction with the boiling nitric acid tests, weeks. alloy 690 was also evaluated for resistance to Poly- 4. Undeaerated water + ppm 875 chloride thionic acid cracking. The standard polythionic (NaCI), 260C, 8 weeks. acid test was employed, with the potency of the 5. Vapor phase above environment No. 4, 8 acid solution being demonstrated by cracking a weeks. sensitized type 304 stainless steel within 1 hour. 6. Undeaerated water + 660 ppm chloride Alloy 690 exhibited complete resistance to cracking (NaCI) +150ppm Na2HP04, 260C, 8 in this solution, both in the annealed condition and weeks. after prolonged aging at 316C (Table 8). 7. Deaerated N2H4 water +500ppm chlo- Vol. 28, No. 2 89 Table 9. Stress corrosion tests at 260C in undeaerated water containing 500 ppm chloride (NaCI) to evaluate the effects of heat treatment, cold work, welding and crevices Double U-bend (crevice) test specimens for alloy 690 Single U-bend test specimens for alloy 800 and 304 SS Code: MA= Mill Annealed CR=Cold Rolled 40 Al=1h @ 1093C/WQ A2=1h @ 1150C/WQ A3=1h @ 1205C/WQ A4=1h @ 1065C/WQ La=5h @ 704C/AC L=2h @ 650C/AC W (thickness, mm)=manual gas tungsten-arc welded at indicated thickness with matching filler * Heat Y24A7L ride (NaCI), 316C, 96 weeks. and the presence of crevices did not reveal any In tests using undeaerated water plus chloride or detrimental variable. Since no evidence of any chloride plus phosphate, the solution was prepared stress corrosion susceptibility was found, alloy 690 by dissolving NaCI in distilled-deionized water. must be regarded as highly resistant to chloride The pH was not adjusted and the autoclave was cracking. sealed with 1 atmosphere of air in the headspace. 5.2. Caustic Environments The oxygen content of the solution at temperature Alloy 690 has been tested for resistance to could not be measured, but was estimated to be at caustic stress corrosion cracking in both un- least ppm 6 at the start of each test period. Speci- deaerated and deaerated NaOH solutions. The mens were either immersed in the solution or sus- controls used in the caustic tests were Type 304 pended in the vapor phase above the solution. stainless steel, alloy 800, alloy 600 and com- In the test using deaerated water plus chloride, mercially pure nickel, Nickel 201. The latter was the solution was prepared by dissolving NaCI in included when it was realized that all commercially distilled-deionized water to which 65% N2H4 was available Fe-Cr-Ni alloys exhibit some degree of added (0.2ml of N2H4 per 1.6l of solution). The cracking in high temperature caustic environments. solution was then deaerated within the closed The following test environments were employed: autoclave by three alternate pressurizing (3.5 MPa- 1. Undeaerated 50% NaOH, 300C. argon)/aspirating (5.7 KPa) cycles, and the auto- 2. Deaerated 50% NaOH, 284, 300, 316 and clave sealed with 1 atmosphere of argon in the 332C. headspace. The oxygen content after deaeration 3. Deaerated 10% NaOH , 316C. was <20 ppb. For these caustic tests, the autoclave solution was Alloy 690 did not exhibit stress corrosion crack- prepared by dissolving ACS certified reagent grade ing in any of the environments used. Further- electrolytic NaOH pellets in distilled water. For more in Test No. 2 (Table 9), extensive evaluation tests in undeaerated NaOH, the autoclave was of the effects of heat treatment, cold work, welding, sealed with an 0.35 MPa over-pressure of air in 90 Boshoku Gijutsu Table 10. Stress corrosion tests in undeaerated 50% NaOH at 300C to evaluate the effects of heat treatment, cold work, and welding U-bend test specimens Code: MA=mill annealed CR = cold rolled 40 Al=1h @ 1093C/WQ A2=1h @ 1150C/WQ A3=1h @ 1205 C/WQ L=2h @ 650C/AC W (thickness, mm)=manual gas tungsten-arc welded at indicated thickness with matching filler * Heat Y24A7L the headspace. The specimens were examined Figure 10). However, the differences are small after each 3-day exposure period, with a fresh solu- and dependent on test temperatureT. tion being used for each subsequent period. For In deaerated 50% NaOH, alloy 600 appears to tests in deaerated NaOH, deaeration was accom- be somewhat more resistant to stress corrosion plished by three alternate pressurizing (3.5 MPa- cracking than alloy 690. The reverse is true in argon)/aspirating (5.7KPa) cycles, followed by a deaerated 10% NaOH, as shown in Figure 10. 2-hour argon bubbling treatment. The autoclave Here, with the exception of Nickel 201, alloy 690 was sealed with 1 atmosphere of argon in the was the most resistant alloy tested, although again headspace. the difference between alloy 600 and alloy 690 is The results from the undeaerated 50% NaOH slight. tests are shown in Table 10. With the exception The caustic stress corrosion test data can be of one of two cold rolled specimens, alloy 690 did summarized as indicating that alloy 690 is the most not exhibit cracking in the 27-day test period, ir- resistant of the alloys tested in undeaerated 50 respective of heat treatment, cold work, or presence NaOH and deaerated 10% NaOH, while its re- of welds. sistance is slightly inferior to that of alloy 600 in The results from the deaerated 50% NaOH tests deaerated 50% NaOH. have been previously reported 8. At 316C, 5.3. High Temperature Water Containing Oxy- alloy 690 appears to have a slightly lower re- gen sistance to stress corrosion cracking than alloy 600 It has been known for a long time9 that most in this environment (see scatterband shown in commercially available Fe-Cr-Ni alloys exhibit Vol. 28, No. 2 91 stress corrosion cracking in oxygen-containing (undeaerated) water at elevated temperatures in the presence of a crevice. In fact, stress corrosion resistance under these conditions was a major pa- rameter in the development of the composition of ● 0.5h/1066-1121C/WQ alloy 690. The "oxygen-plus-crevice" test used ○ 0.5h/1010C/AC □ 4.5h/1066-1121℃/WQ here employs double U-bend specimens, with the +2h/649-677C/AC crevice being formed between the inner and outer members of the specimen. Cracking, when it occurs, is confined to the outer surface of the inner U-bend member, having been initiated within the crevice. In these tests, the environment was dis- tilled and deionized water adjusted to pH 10 with NH4OH. The autoclave was sealed with 1 atmos- phere of air in the headspace, which provides a minimum dissolved oxygen content of ppm. 6 As shown in Table 11, cracking was readily ob- tained in the control alloys, whereas alloy 690 survived the 48-week test period without cracking. Fig. 10. Stress corrosion cracking of U-bend The cracking resistance of alloy 690 in this environ- specimens in deaerated 10% NaOH at ment was not impaired by various heat treatments, 316C. Exposure time=6 weeks. cold work, or the presence of welds. Table 11. Stress corrosion tests in undeaerated water at 316C to evaluate the role of heat treat- ment, cold work, and welding on the "oxygen-plus-crevice" effect Double U-Bend (crevice) test specimens Code: MA=Mill Annealed CR = Cold Rolled 40% Al=1h @ 1093C/WQ A2=1h @ 1150C/WQ A3=1h @ 1205C/WQ L=2h @ 650C/AC L1=4h @ 593C/AC L2=5h @ 704C/AC W (thickness, mm)=manual gas tungsten-arc welded at indicated thickness with matching filler Heat Y24A7L