~ g - ¥ 2 -r"} I‘:‘\‘,? & CT h, MAST OAK RIDGE NATIONAL LABORATORY operated by UNION CARBIDE CORPORATION W for the : U.5. ATOMIC ENERGY COMMISSION ORNL- TM-326 53// /é INFLUENCE OF VARIOUS GASEOUS ENVIRONMENTS ON THE CREEP-RUPTURE PROPERTIES OF NUCLEAR MATERIALS SELECTED FOR HIGH-TEMPERATURE SERVICE H. E. McCoy NOTICE This document contains information of o preliminery nature and was prepared primarily for internal use ot the Oak Ridge National Leboratery, It is subject to revision or correction and therefore does not represent o final report, The information is not to be cbstracted, reprinted or otherwise given public dis. semination without the approval of the ORNL poatent bronch, Legal and Infor- mation Control Department, LEGAL NOTICE This report was prepored as an account of Goverament sponsored work., Neither the United States, nor the Commission, nor any person acting on behalf of the Commission: A. Mokes any warranty or representation, #xpressed or implied, with respect to the sccuracy, completeness, or usefulness of the information comtained in this report, or that the use of any information, apporatus, method, or process disclosed in this report may not infringe privately owned rights; or B. Assumes any liobilities with respect to the use of, or for damages resvlting from the use of any information, apparatus, method, or process disclesed in this report. As used in the obove, “‘person acting on beholf of the Commission® includes eny employse or contractor of the Commission, or employse of such contractor, to the extant that such employes or contractar of the Commission, or employee of such controctor prepares, disseminates, or prevides access to, gay information pursuant to his employment or contract with the Commissien, or his employment with such contractor. INFLUENCE FROPERTTIES OF ORNL-TM- 326 Copy Contract No. W-7405-eng-26 METALS AND CERAMICS DIVISION OF VARICUS GASEOUS ENVIRONMENTS ON THE CREEP-RUPTURE NUCILEAR MATERTALS SELECTED FOR HIGH-TEMPERATURE SERVICE H. E. McCoy DATE ISSUED SEP 191962 QAK RIDGE NATIONAL LABORATORY Osk Ridge, Tennessee operated by UNION CARBIDE CORPCORATION for the U. 5. ATOMIC ENERGY COMMISSION ABSTRACT In order to increase the operating temperature of gas-cooled reactors one must consider many problems in the selection of structural materials and fuel element claddings. Among these are the interactions of active metals and gases together with the attendant changes in the creep properties, ductility, and rupture l1life. In this paper various types of gas-metal interactions are discussed and their effects upon the service performance of metals are illustrated by experimental data. Pure gases which are potentially suitable for reactor coolants as well as those gases which may be present as impurities were studied. These include Ar, Hj, CO,, CO, Ny, 05, and air. Three basic classes of materlals were evaluated: austenitic stainless steels, nickel-base alloys, and refractory metals. The results of tests of austenitic stainless steels over the temperature range of 700 to 900°C show that their properties sare essentially equivalent in argon and in hydrogen enviromments. Various degrees of strengthening were observed in air, N, 0p, CO, and COp environments. The results of a series of tests at 815°C in enviromments contalning various partial pressures of oxygen in argon are presented which show thaet the properties of stainless steels can be markedly altered by the presence of a few parts per million of oxygen. The respongible mechanisms are postulated in each case, The work on nickel-base alloys was centered around Inconel and "A" nickel. However, supplementary tests were conducted on high-purity nickel and on several laboratory melis to evaluate the influence of composition. In contrast with the sustenitic stalnless steels, these tests show that the creep properties of nlckel-base slloys are inferior in hydrogen as compared with those observed in argon. This effect is manifested through an incresse in the secondary creep rete and & decrease in the time to rupture., No significant changes in ductility were observed. Enviromments which are carburizing, oxidizing, or nitriding are shown to result in &n increase in the creep strength of nilckel-base alloys. Proposed mechanisms for the observed creep behavior are presented, The refractory metals, exemplified by columbium, were found to be strengthened at room temperature by the addition of controlled amounts of oxygen, nitrogen, hydrogen, or carbon. However, serious embrittlement resulted if excessive guantities were added. Allowable concentration limits for each of the above interstitials are presented. The increase in high-temperature creep strength of columbium as a result of the addition of nitrogen and oxygen is demonstrated. INFILUENCE OF VARIOUS GASECUS ENVIRONMENTS ON THE CREEP~RUPTURE PROPERTTES OF NUCLEAR MATERIALS SELECTED FOR HIGH-TEMPERATURE SERVICE H. B. McCoy INTRODUCTTON One currently feasible concepl proposed for gas-cooled reactors involves a nonradicactive coolant, which requires that the nuclear fuel be properly clad to prevent the release of fission gases into the coolant. It is this cladding which impcses a limitation on the maximum operating temperatures and allowable rates of burnup. Because of the present status of knowledge of the fabricaticn and of performance of ceramic-clad fuel elements, the nuclear industry is depending heavily upcn metallic clad- dings. The austenitic stainless steels, the iron-chromium-aluminum alloys, and several nickel-base alloys have all been used as cladding materials. Since all these materials absorb neutrons and hence increase the fuel inventory, the minimum thickness 1s used. The refractory metals, such as niobium, offer some Improvement with respect to higher operating temperatures and better neutron economy. Several gaseous coolants being used in reactors already in operation or under construction include helium, carbon dioxide, air, and Np0.5 vol % 0,. Except for 1ts hazardous nature, the properties of hydrogen [1] establish it as the superior gaseous coolant. Also, impurities in the coolants will be of concern, Air in- leakage results in the introduction of significant quantities of oxygen, nitrogen, and water vapor. The outgassing of components within the system, such as graphite, can also intreduce significant quantities of impurities. In choosing a metallic cladding one must remember that the service environment is an important variable, for no material has a unique set of strength values at a given temperature. The gas-metal interactions which can occur between coclant and cladding could result in significant changes in the creep propertlies, ductility, and rupture life of the cladding material. Consequently, the chemical compatibility of the cladding materisl with the coolant and impurities present in the coolant must be considered. Although numerous other factors are of importance, this paper is concerned primarily with the problem of gas-metal compatibility and the resultant mechanical property changes. The classes of materials studied were austenitic stainless steels, nickel-base alloys, and refractory metals, and the data deal primarily with a few representative materials from each class: type 304 stainless steel, TInconel, high-purity nickel, and niobium. The environments studied were Ar, Hp, C0O, CO,, Nz, 0z, and air. The gases used were of the highest available purity, although the gas composition would not necessarily be analogous to that expected in a reactor, This approach enables a clearer interpretation of the test results and yet reveals potential gas-metal compatiblility problems which may arise in various mixtures of these gases. EXPERTMENTAL DETATLS Type 304 stainless steel and Inconel in the form of 0.060- and 0.020-in, sheet were used in these studlies. Materials from several heats were used. However, the compositions of all materials fell within the commercial specifications given in Table T, Exact compositions are Table I. Nominal Compositions of Test Materials Nominal Composition (wt %) Type 304 Constituent Inconel Stainless Steel Fe 9.0 max Balance Ni 75.0 min 8.0 to 11.0 Cr 12,0 to 15,0 18.0 to 20.0 Mn 1.0 max 2.0 max Si 0.5 max 1.0 max Cu 0.5 max 0.15 max 0.08 max P 0.04 max 0.02 max 0.03 max - 4 - given when they are of importance. The specimens prepared from the 0.060-in., sheet had a gage width of 0.500 in. and a gage length of 4,00 in,; those prepared from the 0,020-in. sheet had gage dimensions of 0.250 x 2.50 in, Sheet specimens were chosen because of their large ratic of surface area to volume., The equipment and experimental procedure for running stress~rupture and tensile tests in controlled environments have been described previously [2]. Small niobium tabs 1.25 x 1.00 x 0.040 in. were contaminated with controlled amounts of carbon and were subjected to bend tests at room temperature. The details of the test procedure are reported elsewhere [3]. RESULTS Creep-Rupture Tests — Type 304 Stainless Steel The results of creep-rupture tests in Ar, CO, CO,, Hjy, Nz, 0y, and alr are given in Figs, 1 and 2 for conditions where sufficient tests were run at equivalent temperatures and stresses for the environmental effects to be compared. In most cases each curve plotted represents the results of at least two tests. The creep behavior in argon is used as the reference with which the properties in other environments are compared. For the test conditions examined, the various test environments influenced the creep properties in the following ways: 1. At 1500 and 1700°F (816 and 927°C) air or nitrogen decreased the creep rate, and CO and COp had a strengthening effect. 2. At 1500°F {816°C) the creep strength was essentially the same in argon and hydrogen but was less in oxygen than in argon. UNCLASSIFIED ORNL—LR-DWG 44447R3 RUPTURED 290 hr, 1013% ¢ DRY AIR DISC. 2061 hr, 12.39% ¢ DISC.3025 hr, T%e¢ STRAIN (%) DISC.1506 hr, 3.74 % ¢ DISC.A73hr, 2.7% ¢ DISC. 0 100 200 300 400 500 600 700 800 300 1000 TIME {hr) Fig., 1, Effect of Environment on the Creep Properties of Type 304 Stainless Steel at 1500°F (816°C) and 3400 psi, UNCLASSIFIED 8 ORNL-LR-DWG 44446R4 * DISC. 1771 br, /7/ Hy 1,37 % ¢ 7 - / * / 6 i& / 5 y Z 4 / = / / DISC. 1005 hr, e 3.39% ¢ / Ne | — " DISC.1554hz/ /// A7 % e > f /f/’/ AIR 1 //'// o 0O 100 200 300 400 500 600 700 800 900 TIME {hr) Fig. 2. Effect of Environment on the Creep Properties of Type 304 Stainless Steel at 1700°F (927°C) and 1200 psi. 1000 -7 - 3. At 1700°F (927°C) the creep life was less in hydrogen than in argon but was strengthened in an oxygen environment., The reduction of strength observed at 1500°F (816°C) in oxygen was somewhat unexpected. In an effort to more fully evaluate the influence of oxygen on the creep properties of type 304 stainless steel, a series of creep tests were run In various argon-oxygen mixtures, The results of these tests are summarized in Figs, 3, 4, and 5. The oxXygen-argon mixtures were prepared in standard gas cylinders, and a flow rate through the test chanmber of approximately 0.1 ftB/hr was used, The relatively large effects of small partial pressures of oxygen on the creep strength probably account for the difficulties encountered in reproducing test results in an argen environment, The crack densitiy plot shown in Fig. 5 wag determined metallographically. Several specimens were examined metallographically. Figure 6 illustrates a typilcal microstructure, except for some variation in grain size, of specimens tested in argon and hydrogen at temperatures between 1500 and 1700°F (816 and 927°C). The steps on the surface resulted from grain-boundary sliding and rotation of the grains and were not observed unless the surface of the specimen was completely free of oxide, Figure 7 is a photomicrograph of the gage section of a specimen tested at 3400 psi in air at 1500°F (816°C) for 3025 hr. Features of importance are the surface oxide, the surface and internal cracks which follow grain boundaries, and the precipitate present. Relevant test data from several specimens tested at various stresses in air at 1700°F (927°C) are given in Table IT. The formation of a nitride precipitate was obvicusly not a unique function of the total test time but was related to UNCLASSIFIED ORNL-LR-DWG 51625 (x102) | 0’ \{ o I ANNEALED fhr AT 1900°F IN H, ! \ 0.060-in. SHEET < l \ ARGON CARRIER GAS £ 14 \ = \ = \ \ r 12 o : / a Ex:) \ [ —— L 10e L LJ o O 8 6 2 3 4q 5 6 10° 10 10 10 10 10 10 OXYGEN CONCENTRATION (ppm) Fig. 3. IEffect of Oxygen Concentration on the Rupture Life of Type 304 Stainless Steel at 1500°F (816°C) and 3400 psi. CREEP RATE (% /hr)x1073 UNCLASSIFIED ORNL-LR-DWG 51633 / - //q S~ ANNEALED 1hr AT 1900°F IN Hp 0.060-in. SHEET ARGON CARRIER GAS Fig. 4. 10 10 2 10> 10° 10° OXYGEN CONCENTRATION (ppm) Effect of Oxygen Concentration on the Creep Rate of Type 304 Stainless Steel at 3400 psi and 1500°F (816°C). - 10 - - UNCLASSIFIED S ORNL-LR-DWG 51624 S {0 T— o < L = J 8 (08 8 ) X i ) Ef © P‘ | i o / L /I ANNEALED thr AT 1900°F IN H, I 4 ! 0.060 in. SHEET L - 7 L t/ S 2, -~ o 17 L = | D =z O 10° 10! 102 10° 104 109 108 OXYGEN CONCENTRATION (ppm) FPig. 5. Burrace Crack Density of Type 304 Stainless Steel in Argon- Oxygen Environments Stressed in Tension at 3400 psi and 1500°F (816°C), Unclassified Y-41561 INCHES 0.02 0.03 Fig. 6. Structure of Type 304 Stainless Steel Creep Tested for 2869.7 hr at 1500°F (816°C) and 4000 psi Stress in Static Argon. Speci- men ruptured at 15.62% strain. Rtchant: aqua regia. Unclassified Y~37506 LN Tt oy .Y ) ‘ - Y T ; A .018 Foman o B RNIPRE > e S © Unclassified ¥-39586 3 7 w - <) Zz Q02 ] D3 00 a Type 304 Stainless Steel Specimen (a) Gage section, etched with aqua Strain rate was 0.0125 in,/in.-min. annealed for 1 hr at 1900°F {(1038°C) in CO at 1500°F (816°C) for 1731 hr. - 19 - surface., It is belileved that these voids were originally filled by oxides which were lost during polishing. At 1700°F (927°C) the carbides became much coarser and the rate of carburization increased significantly. Specimens tested in CO, over the range 1300 to 1700°F (704 to 927°C) were observed to be carburized, as is illustrated by Fig. 12 of a specimen tested at 3400 psi in CO, at 1500°F (816°C) for 1727 hr, at which time the test was discontinued and the specimen subjected to a standard room- temperature tensile test. In ¥Fig. lZa, showing the gage section of this specimen, there is evidence of considerable carburization although no case 1s visible. Figure 1 Q o w0y . 008 . P \ '\“’ - N :?SQ . c e Loo Fig. 13, Type 304 Stainless Steel Specimen Tested at 7000 psi in CO, at 1500°F (816°C). Specimen was prestrained 10% at room temperature prior to testing. Material failed after 432 hr with 4.66% strain. Etchant: aqua regia. - 22 . e o - Unclassified - - « . 1 » .oy STyt ; ‘ ’ e 4 — s e . w4 3 e, "wi# . i [ :\‘ v s :., . P i - ¥ s ¢ o Do ; ... - o Ay Wog LY b ey T 3] - . - - ~ R Eur e, L N . + “. - 0 fr N T Ay A . g { ~ L5 w e me VO o g R - . S Ju . i . & e r :{ s - ey . Y\ . .*‘m‘ . L N s | . o & PN . ) d . - \ e o T o , > . . o . . . %, - % sl o . - ; 3 . =4 . L . « ' " 3 5 4 M . Y SR v ¥ o . . ot - -, 5 3 wr o ok o * & ¥ Y 4 N 3 e o o, ek ¥ IO !,%("\ %, . . " {:w, - « , oy B w ~ - s & . § ; - X 2, " "t : / . W - . . . - 5 N - .o ~ —_n- b.’.-" i — 5 . w!n . ¥ 4 % LN % 3 y %\ . )‘ £ u“"'?v . Vi e - ‘ -\ . / ‘t‘{’" . v . @ Q07 po— - i . . st = - 2 * { »r = prrr———" Ny ‘e&"i‘-wu [ o Y ok 'X‘ T . ’ ~ 3 W& = re kS b el i ¢ o3 TE M b * i oL s . . < “« o ‘ e, ¥, ‘!\ 1! . 2 B L & s. ¥ i, i L o8 . . u . “ . i . . 1 . 5 i o ok - 1 A 3 a * f'\ } "‘i o A . £ <) e . .t e 003 Nowmm —— - . e * - % w - "\; . By & % ! £ 3oy -t ;T i v o4 & . \ : t WP e . : : o o Sim ‘ A . R ”% 2 ot e L o at (8 ’ w & - . BOaF w0 AT 00 e - @ ¥ ’ - . ¢ & g ,.-4—)'} - j M \ w - . : & s L o U . . - . - . - i ;‘ « L e wow "E\, ! ¥ £ . It ) " * s % 004 . . e gf ., G oG o . d e e -~ Yo . » ¢ {4 L L w . . - T N Y g, N 2 Ty, 008 b 008 Lo Fig. 14, Typical Microstructure in Interior of Specimen Tested for 528 hr at 1700°F (927°C), 3000 psi Stress in Flowing CO,. Carbon content, 0.59 wt %. FEtchant: aqua regia. - 23 - Unclassified Y-45617 INCHES Fig. 15. Fracture of a Specimen Tested in Tension at 1250°F (677°C). Strain rate was 0.025 in./in.-min. Prior to testing, specimen was annealed for 200 hr in flowing COp at 1700°F (927°C). Btchant: aqua regia. Reduced £%. - 24 - Influence of CO,; on the Tensile Properties of Type 304 Stainless Steel Because of tne observed increase in carbon concentration as a result of exposure to COp, a series of tensile tests were run in an effort to evaluate the influence of this carbon increase on the rupture ductility. A group of specimens 0,020 1n, thick were exposed to flowing CO, for times up to 1000 hr and at temperatures ranging from 1100 to 1700°F (593 to 927°C). Control specimens were run 1n argon to separate the influences of thermal treatment and exposure to CO,, Pairs of specimens annealed under duplicate conditions were tested in tension at 75 and 1250°F (25 and 677°C); the results are summarized in Table III. The following obser- vations were made: 1. The contrel specimens showed that subsequent heating in argon at less than the 1900°F (1038°C) homogenization temperature caused a cpange in the material properties. As a result of this instabllity, the room-temperature tensile ductility was reduced. The tensile ductility of the contreols at 1250°F (677°C) was increased by the thermal treatment, the magnitude of the improvement increasing with increasing annealing temperature. 2. At both 75 and 1250°F (25 and 677°C) the tensile ductility was decreased by exposure to COp. The decrease in ductlility seemed, for a given annealing time, to increase with the temperature of annealing. Creep~Rupture Tests — Nickel~Base Alloys The effect of air on the creep properties of Inconel at 1500°F (816°C) is illustrated in Fig. 16. The secondary creep rate was not significantly Table Hl. Effects of Exposure to COz on the Tensile Properties of Type 304 Stainless Steel Specimen Size: 0.020-in.-thick sheet Strain Rate: 0.02 in./in.»min Average Tensile Properties at 75°F {25°C) Tensile Properties at 1250°F (677°C) . Treatment Prior to Tensile Test? Carbon Elongation Strength Yield Elongation Strength Yield Environment Temperature Time {hr) Concentration (%) (psi) Strength (%) (psi) Strength (wt %) (psi) (psi) As annealed 0.02 54.5 80,300 22,300 27.5 34,000 10,100 Co, 1100°F (593°C) 100 55.0 86,100 26,300 23.5 34,800 10,000 300 45.3 83,000 27,200 19.5 34,800 12,800 1000 0.14 41.6 82,900 28,500 21.5 36,600 12,200 Ar 1100°F (593°C) 1003 0.02 55.0 84,800 25,800 29.5 37,700 14,200 Co, 1200°F (649°C) 97.6 53.8 85,100 23,500 27.5 37,200 12,700 329 56.3 87,800 26,700 26.0 39,200 13,800 1002 0.15 45.3 88,400 28,000 17.0 42,100 15,700 Co, 1300°F (704°C) 500 0,21 25.7 81,600 25,600 17.0 40,100 19,400 785 25.7 82,700 27,200 17.0 37,900 18,200 1000 0.27 24.5 83,700 28,200 14.5 39,300 19,300 Ar 1300°F (704°C) 1006 0.02 49.0 83,800 24,300 29.0 37,500 12,000 co, 1400°F (760°C) 303 20.9 70,100 26,900 15.0 35,300 18,000 500 14.8 70,500 29,100 14.0 31,500 15,300 1000 0.25 20.9 77,200 29,000 20,300 co, 1500°F (815°C) 25 0.21 38.0 79,400 23,100 18.5 35,800 13,000 115 27.5 77,900 23,100 19.0 36,400 17,200 500 0.25 19.5 80,600 25,900 16.5 37,700 19,100 Ar 1500°F (815°C) 500 0.02 53.0 84,500 25,100 34.0 38,700 11,300 co, 1700°F (927°C) 16.7 0.12 41.6 64,400 21,400 21.0 36,900 15,100 49.2 22.1 60,100 21,800 17.0 33,000 15,100 200 0.22 i1 45,000 18,400 12.0 28,000 9,500 Ar 1700°F (927°C) 200 0.02 54.5 83,700 24,400 36.0 35,800 11,100 “All specimens annealed for 1 hr at 1900°F (1038°C) in hydrogen prior to receiving indicated treatment. mga- ELONGATION (%) UNCLASSIFIED ORNL—LR—-DWG 14164R / RUPTURED: 2500 hr - ~55 % ELONGATION - . i § o e e '/ DISG. 7400 hr, 52% ELONGATION _~ 40 36 - 28 1 i P . 1 s e e e e o e e e ] | ' : 3 ¢ ; ) —-92‘.‘ 16 - 12 = 8 10 12 14 16 18 20 22 24 26 28 30 TIME (hr x 100) Fig. 16. Inconel Sheet — Heat "B". Tested as~received in argon and air at 1500°F (816°C) and 3500 psi, - 27 - different in air than in argon., However, air seemed to inhibit rupture to the extent that the stress-rupture life cf Inconel was greater in air than in argon. Figure 16 also illustrates the dependence of the environ- mental effect on the size of the test section. Much greater strengthening occurred in the 0,020-in.-thick specimen than in the 0.060-in.-thick one, Metallographic examination revealed the presence of surface and grain- boundary oxidation in the gage section but gave no evidence of nitride formation, A limited number of tests were run to evaluate the intrinsic effects of oxygen and nitrogen on the creep behavior of Inconel. Figure 17 illustrates the strengthening due to nitrogen being absorbed and the formation of a nitride precipitate. The creep behavior in oxygen was not significantly different from that in air. Figure 18a illustrates the extensive oxidation that occurred in the gage section of an Inconel specimen stressed at 3500 psi at 1500°F (816°C) in an oxygen environment, The test was discontinued after 1671 hr with 83,.6% strain, Relatively little oxidation occurred on the shoulder of the test specimen (Fig. 18b). Figure 19 is a contact print of a radiograph of the gage section of a specimen tested under similar conditions for 1285 hr. Although the specimen was elongated 73.44% there was no evidence of necking. However, the gage section was heavily cracked, again illustrating the ability of an oxidizing atmosphere to inhibit crack propagation. Inconel is strengthened in carburizing atmospheres in a manner similar to that illustrated in Fig., 17 for a nitrogen environment. The strength is improved as a result of the solution strengthening provided by socluble carbon and as a result of dispersion strengthening due to a dispersed carbide phase, UNCLASSIFIED ORNL-LR~DWG 14224 INCONEL _SHEET HEAT B ANNEALED 2050°F IN HYDROGEN 2 HRS. | TESTED IN NITROGEN AND l ARGON I500°F AND 3500PSI % ELONGATION | 14 , | | l%r jL “z fL 10 é 1”// % ’ ! / | o / | s i ¥ L l g/ s o / et | ga 4 - i [ / / / | | // ! / s I 1000 2000 3000 4000 5000 6000 7000 TIME IN HOURS Fig. 17. 1Inconel Sheet — Heat "B". Aunnealed at 2050°F (1121°C) in hydrogen for 2 hr. Tested in nitrogen and argon at 1500°F (816°C) and 3500 psi. Fig. 18. Environment at 1500°F (8l6°C). (b) shoulder, etched with aqua regia,. with 83.6% strain, Reduced 18%. ~ 29 _ g% Unclassified INCHES Unclassified ¢ Y-él 242 UE g e Inconel Specimen Stressed at 3500 psi in an Oxygen (a) Gage section, as~-polished; Test discontinued after 1671 hr Specimen tested in the as-received condition. Fig, 19. 3500 psi in an Oxygen Envircmment at 1500°F (B16°C). with 73.44% strain., Specimen tested in the as~received condition. Unclassified Photo 57243 Contact Print of a Radiograph of an Inconel Specimen Stressed at Test discontinued after 1285 hr wog«. - 3] - The characteristics of the influence of hydrogen on the mechanical properties of Inconel are illustrated in Fig. 20, which shows the results of several creep tests run at 1500°F (816°C) in hydrogen and argon environments, The tests were run in pairs — one specimen in argon and another in hydrogen at an equivalent stress. The influence of hydrogen was observed to become greater as the stress level (or strain rate) was decreased. For example, the rupture lives of two specimens tested in argon and in hydrogen at 10,000 psi were essentially the same, but differed by a factor of 3 between the palr of specimens tested at 3500 psi. Another important characteristic of this effect is that the environmental influence was manifested througn both an increase in the secondary creep rate and a decrease in the fime required for tertiary creep to occur, as was obvious from the pair of tests at 3500 psi. The third characteristic is that there was egsentially no difference in the ductilities of two specimens tested at equivalent stress levels., Any ductility differences shown in Fig. 20 are small enough to be attributed to experimental scatter, Representative photomicrographs of the gage sections of the pair of specimens tested at 3500 psi are shown in Figs. 21 and 22, in which both the density and the geometry of the cracks in the two specimens can be seen 1o be essentially equivalent. A very close examination of the cracks shows that many are composed of chains of volds rather than being wedge- shaped with smooth edges. Although the experimental observations in argon and hydrogen environ- mente Jjust described refer specifically to Inconel, similar cobservations were made for several other nickel-base alloys, including "A" nickel and STRAIN (%) 48 40 UNCLASSIFIED ORNL-LR-DWG 61502 T | | T T AWITH 532% | | | L I’O'Ooo?fi | - — TESTED NA L L | ~7500psi | INCONEL ] 6000 psi oo TESTED IN AS-RECEIVED | | T | CONDITION AT 815°C oyl i ‘ A,,mhw_wi__bwhw_w mi 5C¥3F>pss B | B | / \ ! _ o | S IR B S N b § ; / | L R | - | RUPTURED 2169hr | o | WITH 15.6% STRAIN | H, J’,f/Q 1 T 3500pSi e | — | F ] | 400 600 800 4000 1200 {400 1600 1800 2000 TIME (hr) Fig. 20. 1500°F (816°C), Creep-Rupture Behavior of Inconel In Argon and Hydrogen Enviromments at - 33 - . - . e . ; : T i1 7 » - LT e o e D Unclassified . . . ) . 5 L > & T i St o Tl T N Y-41461 L - ‘ - 2 b ‘ #_ o L C o u.w—d I Q Z 0.02 0.03 z i ,,}7 . MSM;} ,l = Ia E ,2wj.xféf‘fi{ ...8... e )2 CO A Fig. 21. Inconel Specimen Tesgted at 3500 psi in Argon at 1500°F (816°C). Failed after 2169 hr with 15.6% strain., Photograph taken 0,25 in., from fracture. Etchant: aqua regia. - 34 .y“**»\J . Unclassified T w’“ Y-41458 0 o 1 § s p o [ ] z 0.02 0.03 > e {0 o Fig 22. Inconel Specimen Tested at 3500 psi in Hydrogen at 1500°F (816°C), Failed after 825 hr with 14.06% strain. Photograph taken 0.25 in. from fracture. Etchant: aqua regia. - 35 - laboratory melts of Ni-—5 Fe, Ni-15 Cr, and Ni~1l5 Cr-5 Fe., A limited number of tests were run with nickel refined by an electron-beam-zone process, The results ol one sguch test are given in Table IV. The test environment wag cycled periodically between argon and hydrogen; the creep rate in hydrogen was greater by a factor of 2 than that in an argon envi- ronment., These data also illustrate the reversible nature of the environ- mental effect, Effect of Carbon on the Room-Temperature Bend Properties of Nicbium The results of tests to evaluate the effects of nitrogen, oxygen, and hydrogen on the room-temperature bend properties of niobium have been described previously [3]. The results of additional tests to illustrate the influence of carbon additions are given in Table V., The stress values in Table V are based on an elastic-stress analysis of the test specimen, Therefore the yileld and vitimate stresses were of value only for comparison since plastic deformation occurred. Although carbon improved the strength slightly, it was not as effective as equivalent amounts of oxygen and nitrogen. Carbon concentrations as great as 2100 ppm did not seriously reduce the room-temperature ductility of niobium. Figure 23 illustrates the microstructure of a bend specimen containing 1000 ppn ¢ (No. 106). DISCUSSION OF RESULTS Type 304 Stainless Steel The influences of various environments on the creep propertiesg of type 304 stainless steel stimulated a desire to understand the mechanisms _ 36 - Table IV. Creep Results of Polycrystalline High-~Purity Nickel Test Temperature: 1700°F (927°C) Test Enviromment: Cycled between Ar and H» Stress: 1000 peil Duration of Exposure Creep Rate Environment (hr) (in./in.-hr) x 1077 H, 90.9 7,00 Ar 126.,2 3.48 Ho 119.8 7.40 Ar 97.6 3.10 H, 168.1 7.15 Ar 167.1 3.60 Hy 172.0 6,82 Ar 118.4 3.93 Table V. Effect of Carbon on the Room~-Temperature Bend Properties of Pure INiobium from Heal 238135B Analysis Stress at Specimen? (ppm) Proportional Yield Maximum Max imum Specimen Annealing Limit StrengthP? Deflection Deflection Number Time O Nz g2 C {(psi) (psi) (in.) (psi) Cracked 24 As-annealed 340 48 < 1 140 29,600 41,000 0.25 59,500 O 104 1 hr 360 34 21 310 35,300 43,200 0.25 ©9,100 o 105 2 hr 33,600 43,000 0.25 67,200 o 106 5nr 330 25 10 1000 31,300 42,000 0.25 65,700 No 107 24 hr 310 28 8 2100 38,500 477,400 U.25 66, 800 Mo w-ngm %A1l specimens were first annealed in vacuum at 2372°F (1300°C), then annealed at 2192°F (1200°C) in HpCgHg for the time specified, and then homogenized. b . . , : . . . : . , The stress at the intersection of a line with its origin at zero stress and 0,005-in. deflection and drawn parallel to the elastic portion of the stress~deflection curve is called the "yield strength." - 38 - Unclassified Xw41599 INCHES Fig. 23. Niobium Specimen Bent at Room Temperature. Prior to bending, specimen (No. 106) was annealed for 5 hr at 2192°F (1200°C) in Hy-CgHg followed by a Z2-hr homogenizing treatment, Carbon content 0.10 wt %. Reduced 6%. - 39 . by which the properties are altered. Most of the changes in properties that occur from carburizing or nitriding seemingly can be explained by the theories of dispersion strengthening. The enhancement of creep properties cbtained through carburization and nitriding has been demonstrated by Swindeman and Douglas [5]. The current results show that the magnitude of the enhancement of creep properties in a carburizing atmosphere is a function of the quantity o carbon absorbed by the specimen, At 1500°F (816°C) the creep rate was higher in (0, than in CO {Fig. 1), a behavior consistent with expectations since the carburizatiocon rate was higher in CO than in CC,. Also, the creep rates in both environments decreased with time, again indicating a dependence on the carbon concentration. At 1700°F (927°C) the creep- rupture behavior was similar to that Jjust described. Although these data clearly demonstrate the influence of the quantity of carbon present, there are more subtle factors which may be of importance, for example, the size and distribution of carbides. Carbide distribution of specimens exposed to CO, was affected by stress (cf. Figs. 14 and 15) and by prestraining (cf. Figs. 12 and 13}, It was also found that rather complex arrays of carbides could be formed in as-received material by cold working and subsequent thermal treatments. This is illustrated quite well by the work of Garafalo [6] on the ability of carbides to retard the move- ment of dislocations in type 316 stainless steel during creep at 1100, 1300, and 1500°F (593, 704, and 816°C). Specimens were soluticn~annealed, prestrained, annealed at 900°F (482°C) to precipitate the carbides along dislocations, and creep-tested in the above temperature range, This treat- ment resulted in an enhancement of creep properties., However, the amount - 40 - of prestraining was found to be important since agglomeration occurred at prestrains in excess of 25%, Another factor of possible importance is the change in chemistry of the metal brought about by the formation of carbides, If the carbide formed is Cry;Cg, 1.67 wt % Cr will be taken out of solution for every 0.1 wt % C added. The depletion of the matrix in chromium probably increases the solubility of the interstitial elements oxygen, nitrogen, and carbon, Work is currently in progress toc evaluate these secondary processes, It is felt that environments which cause nitriding enhance the creep properties through the same mechanisms Just discussed Tor carburization. However, the conditions under which nitrogen is absorbed are important. During tests in pure nitrogen nitrides were formed at 1500 and 1700°F (816 and 927°C) but those formed at 1500°F (816°C) were only in areas that had undergone large deformation, At 1700°F (927°C), nitride formation was more general and large quantities of nitrogen were introduced, A series of tests at 1700°F (927°C) in air showed that the formation of nitrides was a definite function of the strain rate and probably depended also upon the time of exposure (Table IT). At 1500°F (816°C) no nitride formaticon was observed in specimens tested for as long as 3025 hr in air. However, Dulis and Smith [4] describe a type 304 stainless steel specimen tested in air at 1500°F (816°C) for 10,000 hr in which nitrides were formed, On the basis of the results just described 1t was concluded that the formation of nitrides in type 304 stainless steel tested in air was a function of the strain rate, the temperature, and the time of exposure, The combination of factors necessary to produce nitriding is one which will make a c¢lean surface available to the nitrogen. If the - 4] - strain rate is not sufficlent for the reactive metal surface to be exposed, extremely long times are required for the nitrogen to permeate or diffuse through the surface oxide, The influence of oxygen on the creep properties was one of strength- ening at 1700°F (927°C), the mechanism probably being that of internal oxidation, but was somewhat deleterious at 1500°F (816°C) (cf. Figs. 1 and 2). In an effort to better understand the influence of oxygen at 1500°F (816°C) a series of tests at various oxygen partial pressures were run. A marked strengthening was observed at very low oxygen levels but subsequently became reduced at higher levels (Figs. 3 and 4). This difference was manifested both through an influence on the secondary creep rate and by the rupture life, The crack density increased as the oxygen concentration was increased (Fig. 5). The results of these tests are currently unexplainable since all of the oxygen pressures except the lowest resulted in surface oxidation. However, they cause strengthening by the mechanism of intermal oxidation in this material at 1500°F (816°C) and oxygen pressures near 1 atm to be viewed with skepticism. In light of the observations made regarding the behavior of materials in pure oxygen at 1500°F (816°C), there is some doubt as to the mechanism responsible for the enhancement of creep properties cobserved in air at this temperature. It was polinted out previously that no nitrides were visible in specimens tested in air at 1500°F (816°C). Although vacuum- fusion analyses do not indicate a substantial increase in nitrogen, they do represent the average of a cross section of the specimen; consequently, the nitrogen concentration could be relatively high near the surface with- out ralsing the average. Therefore it is presently postulated that the - 42 - observed strengthening in air at 1500°F (816°C) was due to nitrogen near the surface inhibiting the formation and propagation of cracks. At 1700°F (927°C), however, oxidation did seem to provide some degree of strenthening, and the enhancement of creep properties in air may be a consequence of both nitriding and oxidation. Oxidation also occurred in CO and CO, environments. In light of the tests run in pure oxygen, it is felt that such strengthening was due primarily to carburization. At 1500°F (816°C) the creep properties were essentially the same in argon and hydrogen; at 1700°F (927°C), however, they were inferior in hydrogen, probably because of the removal of carbon by the hydrogen. An improvement in the creep resistance of type 304 stainless steel wvas obtained by the addition of carbon but was accompanied’by a reduction in rupture ductility. The following generalizations concerning the influence of carbon content seem to be warranted: L. The ductility was reduced as the carbon content was increased, <. Burface layers high in carbon served as nucleation sites for cracks which propagated across the test piece. Hence, for a given average carben content, the ductility was higher if the carbon was uniformly distributed. Although the ductility was not reduced to an alarming level by moderate carbon additions, the problem is worthy of serious consideration, Tt is conceivable that carbon concentrations in excess of 1 or 2 wt % may result over long periods of use in a carburizing environment., Another reason for concern is that the role of carbon distributicon has not been - 43 - evaluated., Carbon distributions more embrittling in nature than those studied might easily be obtained as a result of the mechanical and thermal cycling that might be encountered during the operation of a reactor. Nickel-Base Alloys The creep-rupture results for Inconel (Fig. 16) showed that the gtrengthening effect of an oxidizing enviromment was manifested principally through increasing the time of tertiary creep. It was shown further (Fig. 19) that severe cracking occurred in an oxidizing atmosphere but that none of the cracks were able to propagate to failure. Hence it was concluded that the strenthening effect in this material was due to the internal oxidation of impurities and the resulting dispersion of oxide particles, wnich retarded the propagation of nucleated cracks, Similar obgervations were made of "A" nickel and reported previously [7]. A significant observation made during the current work was that the creep benhavior of Inconel is the same in air and in oxygen at 1500°F (816°C). Different rates were observed for type 304 stainless steel under similar conditions. Since Inconel can be carburized and nitrided in & mamner similar to type 304 stainless steel, the introduction of significant quantities of carbon and nitrogen results in strengthening., Pure nickel and nickel-base alloys which do not contain carbide-forming elements cannot be significant- ly strengthened by these processes because of the very low solubility of carbon and nitrogen in these materials. The creep properties of Inconel in hydrogen were inferior to those observed in argon (Fig. 20). It was also noted that the envirommental effect becomes greater as the strain rate decreases and that it does not result in any embrittlement of the metal, Data were obtained (Table IV) which showed that the envirommental influence is reversible, at least for high-purity nickel. Following these observations, proposal of a mechanism to account for the observed behavior is needed. Two obvious possibilities deserve atten= tion: (1) some chemical change occurs; for example, subsurface oxides may be reduced by hydrogen to form water such as has been observed in copper [8], although this results in a loss of ductility, which is contrary to experimental observations in the present case; (2) carbon is removed by surface decarburization. Chemical analyses indicated that this second process does not occur at a detectable rate, The fact that an environmen- tal effect was observed with high-purity nickel dces not substantiate an impurity effect, nor does the reversible nature of the effect., However, sufficient work has not been done to completely eliminate the possibility of an impurity effect. Another possibility, resulting from the mechanism proposed for low- temperature embrittlement [9], is that segregation of hydrogen arcund a crack nuclei is regquired for a crack to form. For example, Livshitz et al. [10] observed that welds in steels were embrittled by prolonged heating in hydrogen in the range 500 to 600°C. They proposed that this loss in toughness must be associated not only with the possibil- ity of iron-carbide decomposition and the formation of methane but also with the hydrogen atoms diffused in the metal combining into molecules, causing stresges in microregions and embrittlement of the material. The quite high mobility of hydrogen [11] in nickel at elevated temperatures, - 45 - however, makes any mechanism which depends upon the segregation of hydrogen seem unrealistic. A further discussion of how hydrogen influences the creep behavior of nickel-base alloys would require a clear picture of the creep process in the absence of hydrogen. Any review of the huge volume of work carried out in an effort to define and understand high-~temperature creep processes is beyond the scope of this paper, and the reader is referred to the excel- lent review by Davies and Dennison [12]. However, the concept of high- temperature stress rupture, as presented by Cottrell [13], is worthy of brief mention because of its relevance., Cottrell envisions two processes which ultimately lead to intergranular failure: one occurring at high strain rates and involving the formation of cracks at triple points or at other points of high stress concentrations, and the other becoming opera- tive when small stresses are applied for long periods of time. In the latter process small holes nucleate at grain boundaries and grow as a result of vacancy condensation., It 1s currently proposed that the first deformation preocess 1s not significantly influenced by the presence of hydrogen but that the second process is affected, possibly through a reduction of the surface energy. This would in turn increase the driving force for the growth of voilds by vacancy condensation. There are numerous details of this hypothesis which must be filled in and work toward this end is currently in progress. Niobium Work reported previously [3] illustrating the influence of oxygen, nitrogen, and hydrogen on the room-temperature bend strength of pure niobium 1s reviewed briefly here, Although oxygen is an effective - 46 - strengthener, concentrations in excess of 3000 ppm result 1ln serious embrittlement, Nitrogen is a very potent strengthener. For example, increasing the nitrogen content from 48 ppm to 460 ppm changes the yield strength from 41,000 psi to 84,200 psi. However, a concentration as high as 1000 ppm drastically reduces the rupture ductility. This may be due in part to the presence of a very thin nitride film on the surface. The homogenization treatment was not sufficient to effect complete homogenization, Hydrogen concentrations in excess of 500 ppm seriously embrittle niobium at room temperature. A slight amount of strengthening is observed with the addition of small concentrations of hydrogen., The absorption rate of hydrogen in niobium seems to be greatest in the range 500 to 600°C, where g stable hydride is formed. The kinetics of formation of this hydride were observed to be quite rapid and the hydride was formed in all specimens cooled through the above temperature range in the presence of hydrogen or water vapor. A very limited number of creep and tensile tests at approximately 1800°F (982°C) showed that moderate oxygen and nitrogen additions improved the high=-temperature strength of nicbium. The strength was decreased by the presence of hydrcgen or water vapor. The present studies showed that carbon additions did not result in gross improvements in the room-temperature strength of niobium, For example, increasing the carbon concentration from 140 to 2100 ppm increased the yield strength only from 41,000 to 47,400 psi. The material was not embrittled by this large quantity of carbon. - 47 . SUMMARY AND CONCLUSTONS The experimental observations which have been presented illustrate the effects of several gaseous enviromments on the mechanical properties of type 304 stainless steel, Inconel, high-purity nickel, and niobium, Type 304 stainless steel was studied over the range 1500 to 1700°F (8lé to 927°C) in environments of Ar, H,, €O, COp, Ns, air, and O,. To evaluate the influence of CO, on rupture ductility, a series of tensile tests were run on samples which had been exposed to CO, for various times over the range 1100 to 1700°F (593 to 927°C). Inconel was studied in Ar, Hs, Np, Op, and air environments at 1500°F (8lé°C). High-purity nickel was studied in argon and hydrogen enviropments at 1500 and 1700°F (816 and 927°C). A review is presented on the influence of additions of oxygen, nitrogen, and hydrogen on the room-temperature properties of niobium as evaluated by bend tests on samples containing controlled amounts of each impurity. Also reviewed are the high-temperature creep properties of niobium as evaluated in environments of dry hydrogen, wet hydrogen, wet argon, argon plus alr, nitrogen, and pure argon. New data are presented on the effect of carbon additions to niobium on the room~temperature bend properties. Type 304 Stainless Steel., — l. Creep data determined in pure argon are counservative values Tor deslgn purposes compared with data obtained in all enviromments except hydrogen and possibly pure oxygen. 2. Air and nitrogen increase the creep resistance at 1500 and 1700°F (816 and 927°C). It is Telt that nitrogen absorption is primarily responsible for the observed strengthening in air. - 48 - 3. Carburigzation occurs in environments of CO and CO, and both enviromments increase the creep resistance. Various types of carbide dispersions can be obtained by different mechanical and thermal treatments, 4, The ductility is reduced as a result of carburization, with the extent dependent upon the quantity of carbon absorbed by the material and the distribution of the carbon. Inconel, — 1. The creep-rupture life is improved by oxidizing conditions, manifested through an increase in the duration of tertiary creep. Hguivalent creep rates are observed in air and in oxygen. 2. The creep strength is improved in a nitrogen environment as =z result of nitrogen absorption. 3. The creep properties of several nickel-base alloys in a hydrogen environment were observed to be inferior to those in argon. This effect wvas manifested through both an increase in the secondary creep rate and a decrease in the time required Tor the initiation of tertiary creep. The rupture ductility is not significantly affected. Tests on high-purity nickel have shown the effect to be reversible, Niobium. — 1. The room~temperature strength of pure niobium is im- proved by additions of oxygen, nitrogen, carbon, and hydrogen. Nitrogen seems to be the most potent strengthener. Embrittlement results from additions of nitrogen greater than 1000 ppm, of oxygen in excess of 3000 ppm, and of hydrogen greater than 500 ppm. Carbon additions as high as 2100 ppm do not seriously reduce the ductility. 2. The creep properties of nicbium at elevated temperatures are im- roved by enviromments which result in the absorption of oxygen or nitrogen. P v rp yEg £ - 49 - The ductility reductions are not so severe as those observed at room temperature. 3. The high-temperature creep strengtn in a hydrogen environment is inferior to that observed 1n argon. These results indicate that no material nas a unique set of strength values, but rather that various properties are exnibited as tne service environment is changed. They als¢ show that all environmental influences are not undesirable. In fact, most materials exnibit their minimum strength in neutral or reducing conditions., Since environmental influences cannot be eliminated, it 1is important that the reactor designer understand these influences and use them to nis advantage. It is hoped that the findings reported herein will contribute fo a general understanding of gas-metal interactions and their influence on the mechanical behavior of metals, ACKNOWLEDGMENT The auvthor would like to acknowledge the able assistance of several service groups at the Osk Ridge National Laboratory in carrying out this researcn and in preparing this paper: the Analytical Chemistry Division, the Graphic Arts Department, and the Metallography and the Reports Office groups of the Metals and Ceramics Division. The mechanical property tests were conducted by E. Bolling, J. EFast, V. (. Lane, and C. Dollins. An- nealing and metal compatibility tests were conducted by B. McNabb and K. W. Boling. The assistance through consultation of co-workers W, R. Martin and J. R. Weir is gratefully acknowledged. - 50 - REFERENCES P. N. Garay, "Hydrogen as a Reactor Coolant," Nuclear Power 5(51), 96 (July 1960). H. E. McCoy, W. R. Martin, and J. R. Weir, "Effect of ®nviromment on the Mechanical Properties of Metals,” p 163 in 1961 Proceedings of the Institute of Environmental Sciences Hational Meeting April 5, 6, 7, 1961, Washington, D. C. H. E. McCoy and D. A. Douglas, "Effect of Various Gaseous Contaminants on the Strength and Formability of Columbium,” p 85 in Columbium Metallurgy, D. L. Douglass and ¥F. W. Kunz, editors, Interscience Publishers, Inc., New York, 196l1. BE. J. Dulis and G. V. Smith, "Formation of Nitrides from Atmospheric Exposure During Creep Rupture of 18% Cr, &% Ni Steel,” Transactions of the American Institute of Mining and Metallurgical Engineers 194, —_— 1083 (1952). R. W. Swindeman and D, A. Douglas, "Improvement of High-Temperature Properties of Reactor Materials After Fabrication," Journal of luclear Materials 1, 49-57 (1959). F. Garafalo et al., "The Creep Behavior of an Austenitic Stainless Steel as Effected by Carbides Precipitated on Dislocations,” Transactions of the American Society for Metals 54, 430 (1961). D. A. Douglas, "The Effect of Environment on High-Temperature Creep Properties of Metals and Alloys," pp 429447 in High-Temperature Materlials, R. R. Hehemann and G. M. Ault, editors, John Wiley and Sons, New York, 19359, 10, 1l. 12, 13. - 51 - F. N. Rhines, "Hydrogen Embrittlement of Pure Copper and of Dilute Copper Alloys by Alternate Oxidation and Reduction," Transactions of the American Institute of Mining and Metallurgical Engineers 143, 312 (1941). R. Troiano, "The Role of Hydrogen and Other Interstitials in the Mechanical Behavior of Metals," Transactions of the American Society for Metals 52, 54 (1960). L. S. Livshitz et al., "Effect of Heating in Hydrogen at High Temperature on the Embrittlement of Welds," Metallovedenie i Obrabotka Metallev 1959, No. 9, pp 52-55. M. L. Hill and E. W. Johnson, "The Diffusivity of Hydrogen in Nickel," Acta Metallurgica 3, 566 (1955). P. W. Davies and J. P. Dennison, "A Review of Intergranular Fracture Processes in Creep," Journal of the Institute of Metals 87, 119 (1958-59). A, H. Cottrell, "Theoretical Aspects of Fracture,”" p 38 in Fracture, ed. by B. L. Averback et al., Technology Press of Massachusetts Institute of Technology and John Wiley and Sons, New York, 196l. 6~-15. 16. 17. 18. 19. 20. 21, 22. 23. 24 . 25. 26. 27. - 53 - DISTRIBUTICN Biology Library 28. Central Research Library 29. Reactor Division Library 30. ORNL — Y-12 Technical Iibrary 31, Document Reference Section 35 Laboratory Records 3335, Iaboratory Records, ORNL RC 36. ORNL Patent Office 37. G. M. Adamson 18, R. J. Beaver 39. M. Bender 40. B. Borie 41, J. V. Cathcart oy J. H. Coobs 4. J. L. Cook L8 J. E. Cunningham 49 J. H. DeVan 50, D. A. Douglas 51 52. R. L. Stephenson 53. W. C, Thurber 54, G. M. Tolson 55. J. Venard 56. J. R. Weir 57. W. J. Werner 58. J. W. Woods 59-73. Division of Technical Information Extension {DITIE) PR oE A EEF@Ea2R 2B D0 74. Research and Development Division AEC-ORO 75-76. D. F. Cope, AEC-ORO H E J L. 0 R B & B = 33U Frye, Jr. Goldman Gray Greenstreet . Harms . Hill . Hinkle Inouye G. MacPherson Manly Martin McClung MeCoy McHargue Patriarca L. Sisman G, M. Scott Slaughter