THE UNIVERSITY OF MICHIGAN COLLEGE OF ENGINEERING Department of Chemical and Metallurgical Engineering CONTROL OF CREEP-RUPTUR.E PROPERTRT IES OF TYPE 304 (18 Cr 10 Ni) AUSTENITIC STEELS P^./,Goodeil'" iT. p Cutlen^' - J,. Metallurgy and iping Task Force of the rime overs Committee' " -.:.. o of the Prime Movers Committee of the Edison Electric Institute Project RP-46 administered through:,FFICE OF RESEARCH ADMINISTRATION ANN ARBOR August 1964

.:::...........':ii

ABSTRACT A study has been made of the variables influencing the creep-rupture properties of Type 304 austenitic steel as produced for seamless superheater tube applications, These variables included compositional variations as well as the conditions of heat treatment, This investigation was undertaken to determine the reasons for the higher rupture strength at 12 00'" of present day Type 304 steel as compared with material produced in the early 1950's and to provide information which could be used to insure high levels of strength in Type 304 austenitic steel, The results showed that the creep-rupture strength of Type 304 steel at 1200F is primarily controlled by the levels of carbon and nitrogen in solid solution. Nitrogen was found to be about 25 percent more potent than carbon as a strengthener, The difference in the levels of strength of older material and currently produced material is primarily attributed to an increase in nitrogen content from about 0, 03 percent to about 0. 07 or 0. 08 percent, While the exact cause of this difference in nitrogen contents is uncertain, it is likely that small changes in the levels of manganese and nickel resulted in the increased nitrogen contents in the steel, In low carbon alloys containing up to about 0. 14 percent nitrogen there vas little effect on the creep-rupture properties at l200"F due to varying the temperature of heat treatment between 1750'F to 2050"F. The creep-rupture strength of material containing carbon in excess of about 0, 05 percent was considerably greater when heat treated at 1950' or 2050'F than when heat treated at 1750'F, The influence of small amounts of the various residual elements appeared to depend on their "reactivity" with the carbon and nitrogen in solution in the alloy. Titanium (0. 03 o) reduced the strength of nitrogen bearing heats but increased slightly the strength of low nitrogen, carbon bearing heats Aluminumr (0. 03'/%) slightly reduced the strengtbh of nitrogen bearing heats. XIo ybde -um (0, 20%) and copper (0. 20%) had aImost no effect on the rupture properties at I200"F.O iii

TABLE OF CONTENTS Page LIST OF TABLES o,,. o... vi-ii LIST OF FIGURES......, viii INTRODUCTION I...... I EXPERIMENTAL PR OC EDURE. E..... 2 Materials..... o. 2 Commercial Material..... 3 LJaboratory Heats - Preparation 3 Laboratory Heats - Description 4 Creep-Rupture Tests.. o 5 Structural Examination... 6 Specimen Preparation.. 6 Etch...... 6 Microscopy. 6 Diffraction.7... 7 REULTS.UL....... 7 Creep-Rupture Properties.. 7 Influence of Carbon 7......... 7 Influence of Nitrogen. SI.. 8 Influence of Boron and Titanium ~. ~ 4 e 9 Influence of Manganese. 12 Influence of Aluminum........ 12 Influence of Copper and Molybdenum... 13 Microstructural Examination,.... 14 Grain Structure........ 14 Grain Boundaries 16 Fracture Characteristics... o 20 Precipitate Identification 21 Ferrite... 22 Recrystallization. 23 DIS CUSS I N........ 24 Carbon and Nitrogen..... 24 Titanium and Boron....... 29 Thermodynamic Considerations... 31 Long Time Properties...... 34 "Old' versus "Modernt Material,... 35 v

TABLE OF CONTENTS (concluded) CONCLUSIONS..,......... 39 R ECOMMENDATIONS 42 R EFPERENCES 6 d S O 44 APPENDIX A - Special Creep-Rupture Tests 82 APPENDIX B - Recrystallization.. 87 APPENDIX C - The Influence of Carbon and Nitrogen on the Extrapolated 100, 000hour Rupture Strength of Type 304 Steel - Details of Figure'23........ 97 vi

LIST OF TABLES Table Page I Chemical Comeposition of Four Type 304 Austenitic Steely Seamless Tubes... 49.II Chemical Composition of Laboratory Heats 50 III Summary of Creep-Rupture Tests.. 52 IV Properties of the Laboratory Heats and of the Commercial Tube, PT-99. 56 V The Influence of Trace Amounts of Titanium and Boron in Type 304 Laboratory Heats of Varying Carbon and Nitrogen Content... 57 VI Ferrite Content of Type 304 Steels Before and After Cold Deformation..... 58 vii

LIST OF FIGURES FFi gure -.g, I.,'a:"'The influence of carbon content and temperature of heat treatment on the 1000-hour rupture strength at 1200~F of laboratory heats of Type 304 steel with very low nitrogen contents 59 2. The influence of nitrogen on the 1000-hour rupture strength at 1200F of laboratory heats of Type 304 steel with varying carbon contents of 0. 01 to 0(..4 pe rcent.,....,,...... 6 3 Stress-rupture-time curves at 12000 and 13501..' of Laboratory Heats 1341A, 1343A, and 1310A, 4. Stress versus vers inimurn creep rate curves ifor. several laboratory heats containing carbon a.nd/or nitrogen, with and without boron and titanium additions................ 2 5, Rupture time versus minimum creep rate for several laboratory heats containig carbon and/or nitrogen, with and without boron and titanium. additions......,.... 3 6, Stress rupture -time curves at 12000 and 1350~F of Laboratory Heats 1338B and 1339A and the commercial tube, PTM9....... 964 7, The influence of the temperature of heat treatnent on the 1000-hour rupture strength at I.200 1 of several laboratory heats and a conmmercially produced tube of Type 304 steel... 6 8, Stress-rupture-time curves at 1200OF for heats with aluminum or titanium additions..... 66 9. Photomicrographs of Heats 1310A and 1338A after rupture testing at 1200F a.Q 67 10, Photomicrographs of Heat 1.309 B, heat treated at 1950t F, a. nd test:ed at 1t200 F0... 68 i1. Photomicrographs of specimeens from heats of varying nitrogen content with 0, 02 percent or less ca )............. < 1 P:-':lhoi.t:, o.-icr.ographs oi specinlens from Heat 1338)L, tested at 1200^-F0, show1'-ing thle t:tahickeninqa ofi th). bounOldari1e.Zs wi'th in re a sinc time to rupture..l:.13.0, tio:li Phog ra'ph s otf s c: pe n. 38Bi- 1 fI rom nr. e a.t 3...,3; sh.tz.owing two facets of the gfrain't,.bou. nd;.-ar'viil

LIST OF FIGURES (concluded) Figure Page 14, Photomicrographs specimen 38B-1 from Heat 1338B, shoing the occurrence of sigma phase at grain boundaries. 72 15. Photomicrographs of specimens from Heat 1341A, heat treated at 2050~F and tested at 12000F and 1350F..,...,, 73 16, Photomicrographs of specimens from Heats 1357A, 1344B, 1337A, and 1339A, heat treated at 2050.F and tested at 12000F,...., 74 17, Electron micrographs of several specimens showing various aspects of the microstructure,. 75 18, Photomnicrographs of specimens from commercial tube PT-9, tested at 1200 F, showing influence of heat treatment at 1800,I 17000, and 1600OF.. 76 19. Photomicrographs of specimens from Heats 1357A, 1343A, 1341A, and 1339A, heat treated at 17500F and tested at 12000F,..... 77 20, Electron micrographs of particles extracted with HCI-picric, mixed acids, -from various specimens 78 21, Carbon.solubility curve for Type 304,, 18%Cr - 10%Ni, austenitic steel..... 79 22. Miller-Larson parameter representation of the stressrupture time curves of several laboratory heats and of the commercial tube, PT-9. 80 23. The influence of carbon and nitrogen on the 100, 000 hour rupture strength of Type 304 austenitic steel at 1200 F, including data from laboratory and comnmercially produced material... 81 ix

INTRODUCT (i ON Austenitic steels have developed a questionable reputa.ticon as a result of a few cases of nlon-r.elia.bility in service'fThe publi c utilities, however, use austenitic steels. in their stearn generators for flexibility of design and for economic service at ternperatures and pressures which require materials itore heat resistant than the low alloy ferritic steels, Particular interest in Type 304 steel has increased in the last few years because of the high strength exhibited, by* rmiost: of the cutrrentl.y produced material which makes it econornically attractive for steann power applications, Many mnlaterials, however, produced prior to about 1955 shsow considerably lower creep-rupture strengths than are characteristic o-f -more recently produced tube material. The specification for the tubing has not been changed in a way which would account for the inc'rease in strength, Fturthermore, it has been known that laboratory heats of Type 304 steel generally have the sane low strengths which were characteristic of the older commercially produced material. These factors raise the question of whether or not unrecognized production conditions cotuld occur which would lead to low strength in the alloy. Accordingly, this investigation has been conducted with the intent of obtaining infor mation and developing general principles relating to the high tetmperatxure properties of Type 304 austenitic steel as produced for superheatebr tubing applications. The results should provide the knowledge necessary to insure high levels of strength in the alloy. The ixit.ial direction'of the research was centered on the possibility that unknown residual elements might have been present in Type 304 steel which gave rise to the increased strength by a mechanism similar to that found active in Type 321 steel by the SP-6 prograrn. ( ) Preliminary emrphasis was placed on the effects of smnall amoou:nts of nitroge n,, titanium and boron on the high tenperature properties of the Type 304 all.oy. The influence o other ele ments was to'be evaluated as developing 4 Figures appearing in parentheses pertain to the referen-ces at the end of the text. I

trends dictated their possible bearing on the problem.. The initial results showed nitrogen to have a very marked influence on the creep-rupture properties of this alloy. They showed that nitrogen alone could account for the differences in the level of properties between the alloy produced in the early 1950's and the present day commercially produced material. Consequently, a concentrated effort was initiated on the influence of this element. This study, however, was complicated by the fact that reliable analyses for nitrogen were not available until late in the program. As a result of the magnitude of the effort directed toward nitrogen, fewer elements were studied than were originally anticipated. Some work was done on the effects of small amounts of aluminum, molybdenum and copper. Using the high temperature data developed, along with thermodynamic considerations, it was possible to speculate as to the types of effects that may result from small amounts of other elements. The OO00-hour rupture strength at 1200F has been the primary basis for the evaluation of:the influence of the experimental variables, Where necessary or desirable, these data have been supplemented with rupture data at 1350"F and with other experiments. This investigation was carried out in the Department of Chemical and Metallurgical Engineering of the University of Michigan College of Engineering under the sponsorship and with the financial assistance of the Edison Electric institute, EXPERIMENTAL PROC EDUR ES Materials Vacuum induction-melted laboratory heats were primarily used to attain the objectives of this investigation. In addition, tubing fromn commercial heats was utiized for purposes of comparing the responses of laboratory and conmmercially produced alloys to specific test conditions, Z

C(on:-iercial Material,. The com-mrnercia-l materials were in the form of /g-inch wall sea:mless tubing and were carried over fron the SP-6. investigation. Detailed ch.emical analyses of these m-u;aterials appear in Table 1 The tube coded PT' -9 was chosen as being representative of the cuirrent high strength commrrnercial material and received prim.mary attention.n this study. A typical fabrication procedure for this material would bet as follows: I. The billet was hot pierced at 2 10 0F to a tube blank approxinartely 4 inches outside diameter by. 3/4-inch wall thickness, 2. The tube blank as then annealed for about S5 rninutes at 1900-2000 F, followed by a water quench. 3, The blank was pickled, straightened and "conditioned" 4. The tube blank was finally "tube redu.ced" or "roto- rocked" to a seam-l less tube 3 0 inches in outside diameter by V. 50-inch wall thickness. Laboratory Heats -- Preparation, The laboratory heats were vacuum induction-xneited in fused magnesia crucibles using a charge consisting of 5000 grams (approximately 10 pounds) of virgin melting stocks The initial melting took place under a pressure of less than 10 microns of mercury. The heats were deoxidized with spectrographic grade carbon which was added to the initial charge in the formr of powdered graphite, In order to avoid excessive losses of manganese through vaporization, this element was added to the charge after the initial meltdown had been complet ed For the purpose of m:aking this addition an atmosphere -was introd uced into the vacu um cham-nber to a pressure of 200-500 micronS. In those heats to which nitrogen -was added, the attmosphere used for the m8anganese addition was dried nitrogen gas. After the nmanganese addition was completed the pressure of nitrogen over the melt was raised to 2 atmosphere aind an addition of CrN was mnade,, The presence of the nitrogen atmosphere over the mtelt helped insure a predictable nitrogen recovery. A short time. A- l a a-llowed after the addition-s were Tnade for honogenization of thke molten -metal, Half of the m2olten n-'tal wias then poured into 3

a 2-inch diameter ingot mold. A further alloy addition was then made to the molten nmetal remaining in the crucible and the second ingot poured. The resultant cropped ingots were approximately four inches in length and weighed about three and one-half pounds. These ingots were rolled into ~4-inch square barstock. This processing was designed to roughly simulate the production of seamless tubing by the "tube reducing" process. The following procedure was used: 1I The ingots were heated for I lhours at the rolling tem perature of 21500'F. 2. The ingots were hot reduced approximately 70 percent to l-inch square bars using 10 to 13 passes with 3 or 4 reheats. 3, The hot rolled bars were then cold reduced 50 percent in cross-sectional area to about 0 7-inch square bars. 4. The cold-rolled bars were heat treated for l-/hour and water quenched, 5. The bars were then given a final cold reduction of 50 percent in crosssectional area to'-inch square barstock. Samples for chemical analyses were taken from the center portion of the hot-rolled barstock. Labratoary Heats - Description. Chemical compositions of the laboratory heats are given in Table II. Complete chemical analyses of all the laboratory heats were not made due to the high cost. Certain key heats were analyzed completely and spot checks were made for various elements in other heats, All analyses were made by commercial analytical laboratories. At least one of the two ingots from each heat was analyzed for carbon. The remainder of the composition data as inferred from the reported analyses, the aim composition and knowledge of recovery efficiencies. These data are tabulated in Tabled IL For the most part, this procedure appears to be quite adequate since the losses of elements such as nickel, chromiun, silicon and manganese are very small (usually less than 3 percent of the aim concentration). Nitrogen analyses were accompl shed ry wet chemical techftnitque and measured total nitrogen content. In the presence of strong nitride formers 4

such as titanium, this techniquge apparently requires considerable care and skill, however, consistent a:d reasonable results were obtained. These results, however, indicated that nitrogen had been picked up from the nitrogen atmnosphere in the furnace. Rapid nitrogen pick-up under atmos(3, 4) pheric pressures was later found to have been noted by several. authors( for chromlium bearing alloys in general and for Type 304 steel in particular. Using the data reported for several heats, the nitrogen contents for the remainder of te heats were estimated. This involves greater error than the estimates of chromium, nickel, etc., The error in the estimated nitrogen contents of Table II is believed to be not more than 0,( 02 percent nitro ge n Creep-Rupture Tests The creep-rupture tests were conducted in simple beam or in direct loaded individual creep-rupture units using test specirmens with a 0 250inch diameter and a I-inch long reduced section,. Three chromel-alu:mel thermocouples were wired to the 1-inch reduced section of the specimens with chromel wire,'Type 304 steel foil 0 005-inches thick was placed between the thermocouple and the specimen body to help reduce the harmful effects (5) occasionally associated with the contact of alumel wire with the specimens. The thermocouples were wrapped tightly with asbestos cord. The teimperature variation along the reduced section was held to less than 3 F while the indicated test temperature was controlled to within ~ 3 F, Test specinens were placed in a hot furnace, brought to temperature and loaded within a period of about I hou:r, This procedure requires close attention during the first hour of the test but minimizes any possible effects of microstructural instability Creep strain rneasurements were taken during the tests using a modified Martens-type mirror extensometer having a sensitivity of 0. 00001 inch. The percent elongation of the rupture speimens was based on the change in th ead-to -th read measurements. The reductioni of area m easurements were based on th-e change in cross sectional area of the specimeins. 5,

Structural Examination The techniques used in this investigation for structural examination included optical and electron microscopy as well as X-ray and electron diffraction analyses of extracted residues. Specimen Preparation. Specimens for microscopic examination were sectioned longitudinally along the center line of the barstock or the creeprupture specimens using a water-cooled cut-off wheel. The specimen was then mounted in Bakelite and mechanically polished on wet silicon carbide paper wheels. Final polishing was accomplished using a 6 and 0. 5 micron diamond pastes, in turn, on cloth-covered rotating wheels, followed by Linde ina pde der in aqueous suspension on a vibratory polisher. Etchants. The following etchants were used for various purposes throughout the investigation: 1. 60-percent HN 3, electrolytic. This etch satisfactorily revealed carbides, sigma phase, grain and twin boundaries and concentration variations in 18Cr-IONi alloys, 2, 10-percent chromic acid, electrolytic. This etch was useful for revealing carbide phases. 3. 60-percent phosphoric acid, electrolytic, This etch was particularly suited for high resolution of fine carbide precipitates. 4. ION-KOH, electrolytic. This etch was primarily used to identify sigma phase, ( 5. Vilella's reagent (HC1 and picric acid in alcohol), immersion etch. This etch was used to reveal sigma phase and carbides, Electrolytic etching was accomplished using a platinum electrode with an applied potential I to 3 volts and a current density of about 0l 5 amperes per square inch for between 5 and 30 seconds. Microscopy. Conventional methods were employed for optical examination. Electron microscopy was carried out using collodion replicas 6

of the etched. rnetal surface.. The replicas, supported by fine -nickel screens, were shadow cast with paladiu:m to increase contrast. A JE.M -6A electron imicroscope was used in this investigation for exax ina..ti.on of. the rep.icas. Diffraction, Extractions of:ninor phases from. th.e specimens were made using a mixture of HCI and picric acid as the extracting m.edium Portions of residues fr. the xtra tione se f r a teeed for selected area electron diffraction studies. So03.e of the residues were "rolled" into filanments using a Duco cemenit binder and used to make X-ray power patterns for the identification of the extracted phases. R ES UL TS Creep-: Rupture Properties The creep and rupture data accnumulated during this program are tabulated in Table IIl. Fronr these data the 1000-hour rupture strength, the slope of the stress-rupture time cu.rve and the estim.ated elongation and reduction in area at rupture in 1000 hours have been detertriined for a.ll the nmaterials and are presented in T able IV. The results, as a function of comnpositional. variables, can be sumn.:mar ized as follows: Influence of Carbon,. Laboratory heats available fromn other sourc es' ) supplemented by heats made in the present investigation provided a series of heats with variations in carbon contefnt fro m. 009 to 0. 14 percent in the base alloy (18. 5%Cr - 10. 5%oNi - 1 5%Mn - 0 5L/oSi). No addition of nitrogen was made to any of these heats, The 1000-hour rupture strengths and the estimated ductilities (elongation will be used as the mneasu re of ductility unless otherwise stated) at rupture in 1000-hours of these heats are shown in Figure 1. The following points concerning th.e role of carbon are considered to be significant:'7

(a) There was a general increase in the level of strength with increasing carbon content up to about 0. 10 percent carbon for heat treatments at 1950" and 2050\'F. The data indicated no significant increase in strength of the alloy when the carbon connent exceeded this level, (b) There was no appreciable difference in rupture strength or ductility of the alloys with variable carbon contents when heat treated at 1950" or g050 OF. (c) At carbon contents less than about 0. 05 percent, the 1000-hour rupture strengths were slightly lower when heat treated at 1750'F than when heat treated at 1950 F. Material heat treated at 1750~F tended to increase in strength with increasing carbon content, but only up to about 0, 05 percent carbon. (d) The rupture strength of the alloys heat treated at 1750^F tended to decrease slightly with increasing carbon content above 0. 05 percent. (e) The estimated rupture ductility at 1000 hours decreased with increasing carbon content for heat treatments of 1950 F and 2050 F, The decrease was most rapid for the first few hundredths percent carbon, At higher carbon levels there was little change in ductility. (f) The ductility of the materials heat treated at 1750FT was greater than that of the materials treated at 19500 or 2050 "F The rupture ductility, however, followed the same decreasing trend for the first few hundredths percent carbon, as was noted for the solution treatments at 1950 and 2050~" F. Above 0. 05 percent carbon the rupture ductility increased with increasing carbon content, Influence of Nitrogen. The 1000-hour rupture strengths of laboratory heats with varying nitrogen contents heat treated at 2050'F are shown in (7) Figure 2. The data of Nakagawa and Otoguro ) are also shown for comparison, Figure 3 shows the influence of various heat treatments on the stress-rupture time curves of heats containing high nitrogen with low and normal carbon (Heats 1343A with 0 145%N, 0 00ZC and 1341A with 00. 8%N and 0,. 06oC). Data from Eeat 1310A(with 0 01%OoN and 0. 009%C:) have a.i.so 8

been included and serve as a base for comparison with data from other materials' The minimum creep rates of several of these alloys are plotted versus stress in Figure 4 and versus rupture tire in Figure 5. These data indicate the following: (a) Increasing the nitrogen content of the laboratory heats greatly increased the 1000-hour rupture strength, This was true for all carbon levels studied, ie e between 0, 01 and 0, 14 percent carbon. (b) The increase in strength arising from a given nitrogen addition was greater than that due to the same amount of carbon. (c) Nitrogen, up to at least 0, 14 percent, appears to be beneficial as a strengthene r No well-defined limit to the increase in rupture strength due to increasing nitrogen content nas been established, however, (d) The ductility of alloys decreased with increasing nitrogen content, The magnitude of this decrease appeared to be only slightly greater than that resulting from similar increases in carbon contents The combined effects of nitrogen and carbon in decreasing the ductility of the laboratory heats were somewhat greater than that which would be accounted for by the independent effects of these elements' (e) Heat treatment at 1750^F was less detrimental to the creep-rupture strength of an alloy containing nitrogen and low carbon (Heat 1343A) than to heats with higher levels of carbon, This is reflected both in the stress-rupture time curves (Figure 3) and in the stress-creep rate curves (Figures 4 and 5) (f) The slopes of the log stress-log rupture time curves of heats containing nitrogen and carbon did not vary significantly among different heat treatments and testing temperaturesu Influence of Boron and Titanium. The 1000-hour rupture strengths and the estinated rupture ductilities at 1000 hours are tabulated in Table V for the laboratory heats containing small amounts of titanium (0. 02 - $0 05%) and/or boron (15 ppm) in addition to carbon and/or nitrogen_ The properties of these heats are compared in this table with the properties of similar heats 9

with no boron or titanium additions. The data shown in Table V are for materials tested at 12000 F after heat treatrment for -'hour at 20500F', W. Q, The influence of varying heat treatment;s on the stress-rupture tin.me curves of alloys with boron and titanium additions is shown in Figure 6., The response to heat treatment of the alloys containing nitrogen plus low carbon, nitrogen plus normal cabon nitrogen plus boron plus titanium and nitrogen plus normnal carbon plus boron plus titanium are comupared in Figure 7. The creep rate data for several alloys containing additions of only carbon and nitrogen, heat treated at Z050" and 1750"F, are presented as a function of stress in Figure 4; these data can be compared with data obtained from other alloys containing carbon, nitrogen, boron and titaniun which are also shown in Figure 4. The following observations can be nmade regarding the influence of boron and titanium: (a) The alloys with boron (but not titanium) showed little or no change in the IOO00hour rupture strength as compared with similar heats without boron when tested at 12000F after heat treatment for ~/hour at 2050'"F. In general, the ductility of these alloys was somewhat improved as the result of the boron addition. (b) The 1000-hour rupture strength was alt ered in the alloys to which titanium was added (no boron). The strength alteration is apparently related to the presence of carbon and/or nitrogen: (1) For the heat with very low carbon and nitrogen contents (Heat 1364A) 0. 03 percent titanium increased the rupture strength by a small amounts With a higher carbon content (Heat 1344A) the degree of increase in strength was greater. (2) The addition of 0 03 percent titanium to the heat containing about 0. 12 percent nitrogen (Heat 1344B) decreased the 1000-hour rupture strength. (3) The estimated ductility at rupture in 1000 hours of heats with the titanium additions was generally greater than in similar heats without titaJnium. 10

(c) For the heats with low carbon and nitrogen contents (<0, 02%), there was little or no difference in strength as the result of additions of 0, 03 percent tita.ni.um plus 15 pprn boron or additions of the titanium alone (d) In the heats containing carbon and/or nitrogen (>0O 05 percent), there was a significant improvement in the 1000-hoour rupture strength resulting from the combined addition of 0. 03 percent titaniium and 15 ppmr boron. The strength of a heat to which 30 pprm of boron was added (-eat 1339B), however, was not as great as the similar heat writh only 15 ppn boron added (Heat 1339A). The ductilities of these alloys were lower than those alloys with separate additions of titanium and boron, however, the ductilities of all the alloys were in excess of 10 percent. (e) Fromn Figure 6 it can be seen that the higher 1000-hour strengths at t1200F of the heats with both boron and titanium additions mn ay not be retained at times beyond 1000 hours. The stress-rupture timre curve for the heat with boron, titanium, nitrogen and carbon (Heat 1339A) underwent a change in slope just beyond 1000 hours, The slopes at longer timts have been constructed to be para.l.lel to the slope of the m.aterial tested at 1350~, which is significantly steeper than the shorter time slope noted at 1200~F\. This slope is noticeably steeper than the slope of other alloys tested at 1200" or 13500"F The rupture tests at 1200"F indicate that the heat with boron, titanium and nitrogen (PHeat 1338B) also underwent a change in slope; the heat with boron, titanium and carbon (Heat 1337A) apparently did not. (f) The stress dependence of the rninimunm creep rate for the heats containing boron and titanium appears to be different fromu that of the heats without either of these additions. This is seen in Figure 4. The reciprocal slope, n, of the log stress versu-s the 1200" or 1350~F minimumr creep rate curve (i, e the change in creep rate with change in stress) was between 6 and 7 for the heats without tita-nium and boron when heat treated eithher at 2050~ or 1.750 " F or the heat with titaniurn and II

boron (Heat 1339A) the value of n was also between 6 and 7 except when the alloy was heat treated at 2050"F and tested at 12007F under high stress where n was approximately 13, Influence of Manganese. The manganese level of the laboratory heats was intended to be about 1 50 percent, approximately the level of presentday production heats for seamless tube and pipe, One of the heats analyzed contained only 0. 81 percent manganese (Heat 1366). There was no difference in the properties of this heat which could be attributed to the lower manganese level, although this conclusion is clouded by the addition of 15 ppm of boron, Heats 1281 and 1282( 6) with no nitrogen added had 0 44 and 1 48 percent manganese respectively with no significant difference in their 1000-hour rupture strength (see Table IV), Two alloys (Heats 1340A and 1340B) were made to which no manganese was added, The properties were significantly different from. similar heats which contained manganese. The nickel contents of the very low manganese heats were increased in order to maintain a wholly austenitic structure. One ingot had a low carbon content (0 01 percent) while the other had a slightly higher carbon level (0, 04 percent), Both ingots had nitrogen additions. The 1000-hour rupture strengths of Heats 1340A and 1340B were 17 and 18 ksi respectively, This compares with values of 23 to 24 ksi for similar heats with 1. 50 percent manganese. These heats had relatively low ductility at rupture (see Table II or III). In addition the stress dependence of the minimum creep rate was considerably difference from the other heats, n being about 14 in contrast with values of about 6 to 7:for similar heats with approximately 1 50 percent manganese (see Figure 4). Influence of Aluminum, The data from the four heats to which 0, 02 to O0 03 percent aluminum was added indicate that under certain conditions aluminum can have a significant effect on the properties of 18Cr- 0ONi material. The data, however, were not extensive enough to clearly define the level of the effect. The rupture data for material heat treated at 2050"F and tested at 1200OF are shown~. in Figure 8.'Two. of t~he h.eats containin.-l small additions of titarnium. are also shown for cornparison. 3Based on these 12

(a) 3For the heat with the nxormal carbon and no nitrogen (Heat 1360A) there -was no influence of the alunmAinumri addition on the properties of the material. (b) The addition of aluminum to a heat colntaining normal carbon (0. 06 percent) and about 0. 1.3 percent nitrogen (HPeat 1358B) resulted in a decrease of the 1000-hour rupture strength from 25 ksi to 20, 3 ksi. The slope of the stress-rupture time curve of this heat wa.s slightly steeper than exhibited by the similar heat (Heat 1342B) without alur inunm (c) The addition of aluminum to a heat containing low carbon (<0. 02 percent) and 0. 12 percent nitrogen (Heat 1358-A) resulted in lower 1000-hour rupture strength than for a similar heat without aluminum (Heat 1343A). (d) The addition of aluminum to a heat with 0. 08 percent carbon and 0. 12 percent nitrogen and 15 ppm. boron (Heat 1360 B) did not significantly alter the short time strength. The slope of the stress-rupture time curve, however, was lower (it is comparable to that of heats with boron titanium and nitrogen, Heats 1338B and 1339A) with the result that the 1000-hour rupture strength of this alloy was higher than that of a similar aluminum-b earing alloy without the boron addition. Influence of Copper and Molybdenum. One heat was prepared containing 0.20 percent copper (Heat 1365 A), approximately the amount found in many cornTercial materials., Rupture tests at 1200"F showed an insignificant decrease in strength compared with heats without copper; compare, (8) for instance, Heat 1365A with Heat 1342B in Table IV. Nowak (has also reported that copper has no effect at levl els below 0 30 percent. A heat similar to the above was prepared but with the further addition of 0. 20 percent molybdenum (Heat 1365B). This is about the level of residual molybdentum found in many commnercially produced Type 304 heats. Rupture tests of this heat at 1200PF indica.ted no significant improvement in strength attributable to the molybdenum present, A higher mcolybdenumn level (2. 0 percent) was shown ( to increase 13

the 1000-hhour rupture strength at 1i200G of a low carbon - low nitroge-n heat (HF-eat 131 OB) to about 18, 000 psi from a. value of 12, 500 psi for a t9) similar h eat withot molybdenumr (Heat 13 10A). Nakagawa, a.nd Otogunro have recerntly reported on the influence of mrolybdenun in a.n 18 r-Qi 2Ni alloy., They indicated an increase in the creep and rupture str engths of the'mnaterial as the result of additions of unolybdentumrn in the range fromf 0, 85 to 3, 70 percent. Mi c r os truc tur a1 Exatni, nation.Samples of the materials were taken:for optli aI. and electronron:micro — scopic examination, The microstructural chaxracteristics of the grains ard of the grain boundaries are controlled by so-rlewhat different processes. Thiese characteristics -will be discussed in separate sections. G(rain Structure, The salient features of the grains of m aterials heat treated above 1800Q F and tested at 1200 "F appear to be pri.marily depe:ndent on the carbon and nitrogen content. The points to be enphasized are as follows: (a) With very low carbon or nitrogen contents, there was little evidenlce of precipitation within grains. The grains were inherently soft and deformned readily. This resulted in high rupture ductilities. Typical of this behavior -were I-eats 1310.A, 1338A, 1363A and 1.364A. IPhotomicrographs of rupture specimens of Heats 131.0A ad re shown in Figure 9, (b) With carbon present the structure showed carbide precipitates within the grains often along slipped planes. This precipitation was gener — ally not optically visible u ntil the testing tine exceeded about 100 EhoursThe extent of this precipitation depended on. the carbon lev el and upon the place within the specinmen where the precipitationl occurred. The precipitation was nm-ost intensive near tfhe fracture a-nd decreased 1towa-rd the shoulder of the speci-ne n. This is illustrated in. Fi gur res lOa and I /tA

1 Ob for t-eat 1309B3. In these figures the specimens were etched wvith chromic acid which revealed carbides and sigma phase. This stru.cture was typicalc of heats containing 0, 04 percent or more cuarbon..; The fitne carbide precipitates appeared to form on disl.ocations. Tfhis' can be seen by the'pile-ups on slipped planes near the bou ndaries in Fligu.re I0b, The variation. in precipitate density from the frot actutre to the threaded section of the specirnen (i, e, from a high stress to a lowstress region) was also indicative of precipitatin n dslocations. Finally, and in contrast to the heats with low carbon or -nitrogein, the grains of the heats containng carbon in solution before testing were not markedly deformred, (c) For heats containing high nitrogen and low carbon, the aroious etchants did not reveal. discrete precipitation on slipped planes as -was visible when carbon was present.,. However, slip and Multiplte slip were occasionally visible within the grains, These features are seen in the photonmicrographs of ruptured speciintens of Heats 13616A, I.343A and 1343B which are shown in Figure 11 It was further observed that the grains of specir ens fro:n these heats were not markedly defortmed, indicating that the grains had been strengthened even though no precipitation was visible. (d) Other compositional variations did not appreciablly alter the featvtu-res described above, althoug h ther microconstituents also:may have been present, In alloys to which titanium was added (0, 03-0, 05%), titanium carbonitrides or titanium nitrides were rsutally present in stringe rs. The titaniurm carbonitrides were generally fewer in xumber and more uniforn ly dispersed in the heats containing boron than in. those heats without Vboron. In the heats with carbon and nitrogen present, there were many more s-malll particles present in the grain boutndaries after testing when titanium was also present than when titaniurS as absent. These particles may have been titanium compoun:ds or possibly sigma. phas e, 15.

In specimens which were heat treated at low temperatures, i e. 1750'F or lower, the microstructural characteristics were somewhat different from those of specimens heat treated at higher temperatures. In general, alloys containing carbon did not develop extensive precipitation on slipped planes when. tested after a low temperature heat treatment. There was precipitation of carbides, but many of these carbides were located on the original cold worked structure and in the grain boundaries. Most of these carbides were formed during the heat treatment, n durring testing, These features can be seen in Figure 18d and 18e (also see Figure B-2), Grain Boundaries. The appearance of the grain boundaries in the alloys heat treated above 1800'F before testing was characterized by an apparent'thickening" The'thickening" was not found to correlate with the creeprupture properties of the materials tested, however, it is believed to be connected with some of the processes occurring during creep exposures, such as carbide precipitation and sigma phase formation, In various instances it is difficult to distinguish whether the'thickening'' is due to a new, continuous phase formed at the boundaries or to concentration gradients present at the boundaries. An example is shown in Figure 10c. The etching reagent used on this specimen was 60 percent HNO3> F.igure 10a shows the same specimens etched with chromic acid and illustrates that the boundary Ythickening" phenomenon was not revealed with equal clarity with all etchants, Further observations of the grain boundary characteristics are as follows: (a) The extent of the "thickening. varied within the specimen. It was most marked near the fracture and decreased toward the shoulders of the specimen. Figures 1Oc and 10d show a typical example of this, This characteristic implied that the phenomenon producing the'thickening" was stress and/or strain induced. (b) The extent of "thickening" increased with increasing testing time (at 1200 F). Figure 12 shows the development of the thickened boundaries in Heat 1338B over the time period from 33 to 4300 hours. 16

(c) A specimen, 38B- 1 which showed the largest anmo'unt of the boundary'phase'', was cold worked and recrystallized (at 1400'iF) in an attemipt to mr nove the grain boundaries away from the "phase", This was done primarily to differentiate between two possible cases:' (1) That this new "phase" -was recrystallized austenite of about the same composition as the mrnatrix, or (2) That " thickening" was simply the result of a concentration of one or unore elements in the alloy, The result of this experiment is shown in Figure 13a This mnicrograph shows new recrystallized grain boundaries and sharp concentration changes where the old grain boundaries had been. (d) Efforts to determine the nature of this "phase' by X-ray techniques were unsuccessful. An eletron microprobe examination did not reveal the nature of the'phase' but did imply that the areas around the boundaries were depleted of chromiumr (e) It was also observed (see Figure 13b) that the "thickened" boundaries were continuous with depleted areas at oxidized surfaces of the specimen. The oxides formed at 1Z00F on 18Cr -.0Ni type material were primarily chromiumn or possibl.y complex, chromium-rich, iron oxides. This implies that the metal areas adjacent to the oxides were depleted of chromium,, The continuity between these areas and the boxundary areas implies that the thickened boundary area was chromnium depleted austenite. The localization of the depletion at the boundaries is probably due to thIe greater diffusivity of substitutional atoms at bounda ries or interfaces than within the matrix. (f) The possible occurrence of sigma phase in the boundaries was veri(10, I1.) fied using several types of etching reagents. IONKOH used electrolytically was found to distinctly reveal sigma phase by a staining process. This etch showed particles of sigma phase to be located in the grain boundaries, Figure 14a shows an area of specimen 38B-1 etched with KOH,. This is followed by a photonm.icrograph of the same specimen re-etched using 60 percent HNO.3 in1 Figu.re 14b. The for17

mnation of sigma phase could account for a-il or part of the chromiunt. depletion adjacent to the boundaries. It should be noted that caroidei precipitation could also account for sonme ch.-trom.iunll depletion, ( - ) I; y 1 C specime G ns te d at 13 50 1 the...oudary k e i l.ie was i i s i - Lui.ntct'tlin in specimens tested at 12000F'. Specimens of Heat 1341A. tested at IZOO i- and 1350"Fi are comnpared in Figure 15, This feature can be accounted for boy considering that: (1)'There was less sigm a phase formation at 1. 350 F to cause cepiletio.0 than occurred at JZ1 000F Carbid ati:. e I.:ses: so,.... a chrom iunma depleti on, as is evident aro.und th e car bides withtini thie t. grains of various samples. (2) Diffusion of suostitutional atoms (Cr) at 1350~F from within the natrix to the b:)ou-taries was wasuff cl t I y r a1 id to comrpensatCe f)or. chromium depletion near the boundary, (h) Figure 16 shows various other heats having wide compositional variations which exhi.Ait ooundary t thickening" during testing at 200 0. These photon icrographs are fromD- rupture specimens which lasted between 500 and 1000 hours. Note that in some of th-e micrographs there appears to -be boundary migration within the nickel-rich area, (i) Electron micrographs of several samples etched to show the boundary phenomenon are shown in Figure 17, The distinctive feature visible was tha t the edges of thhe thic.ken.ed booundarv regionJis a-ppea.rdc;s - tra.du-a tra.nsitions rather than sharp phase boundaries Th e actual grain boundary was often. visible within the thickened regions a.s a sharp.line. eOccasionall y th.e.oun da.ry vrwas cointlcidetCal wi \iL on. e ed.-t. d o:[ tt.e thicke.tlllned region. In such a case the boundary is assumed to have mr. igrated away from its original position. In so ne cases the appearance of the grain boundaries in miate.rsial givena to tlo erw: ratur h.:at ter eat ment (i. e r 1750"'F or lower) prior to testi n't abi IOO; (' i was s gii * niia. n idif a ifemrnen ft forom that at.fi ler tfic er terne' orie atun"re..eat treat.n:ent;. t'i glure 1.8 sh.ows a. series of photomnic:ro. rauihts:from.'i l.i:Itu. red 18

specimens (600 to 800 hours duration) of a commercial tube, PT-9, cold reduced and heat treated at 1800~, 1700~ and 16000F before testing at 1200'~F The specimen heat treated at 1800 0F was very similar in its microstructure to a specimen solution treated at 2050~0 and tested at 12000F.~ The specimen heat treated at 1700~F, however, contained massive particles of sigma phase and exhibited considerably wider boundaries' These regions gave the appearance of being a distinct phase. This was further accentuated in the specimen heat treated at 1600~F1 As was noted with the specimens heat treated at 2050"F, the amount of the boundary "phase" progressively decreased as one moved away from the region of the fracture. This can also be seen in Figure 18. The amount of boundary "phase" in other materials heat treated at 1750~F appears to be less than that found in PT-9 heat treated at 1700"F but greater than the level found in this material when it was heat treated at 1800"F, Several other examples of the presence of this "phase" are shown in Figure 19. Heat 1357A with very low nitrogen showed the thickest boundary areas, Heat 1310A which contained low carbon as well as low nitrogen formred thick boundary regions and massive particles of sigma phase during testing at 1200~F after heat treatment at 1750~"F. This behavior was similar to the behavior of the alloy after heat treatment at 19500~F, The boundary thickness after testing at 1200~F of Heats 1341A, 1338B and 1339A, each having high nitrogen, was greater after 1750"F heat treatment than after heat treatment at 2050F, but it was not as great as in low nitrogen heats which had similar histories, The boundaries in the heat with high nitrogen and low carbon (Heat 1343A; 0 14 5%N and 0. 02%C) were only slightly thicker after testing following heat treatment at 1750 F than when tested after heat treatment at 2050 "F. These results imply that carbon was responsible for a greater variance in the grain boundary characteristics of the specimens after testing at 1200"F than was nitrogen, The variations related to the carbon content can be lessened, however, if nitrogen is simultaneously present, These microstructural 19

effects are very likely related to the tendency for sigma phase formation. This will be discussed further in a following section. Fracture Characteristics. As another part of the icroscopic examination of the alloys, the characteristics of the cracks found in the fractured specimens were studied. Two types of cracks could generally be found: (a) "Wedge" cracks, which usually initiated at a three grain junction and subsequently spread along the boundaries, and (b) "Cavity' type cracks, where numerous small cavities formed along the grain boundaries, (12) grew and linked up, to cause eventual fracture, The significant fracture characteristics of the specimens tested in this investigation can be summarized as follows: (a) Heats with very low carbon and nitrogen exhibited a ductile, transgranular fracture, (b) Most other heats had both wedge and cavitation type cracks after testing. "Wedge'cracks formed if carbon was the primary strengthener while "cavitation" type cracks formed if nitrogen was the primary strengthening element. In addition, the specimens heat treated at lower temperatures (1750~F) tended to have fewer cracks, The cracks which were present tended to be of the cavitation type. Numerous authors 14, 15, 16) have shown that thin, often continuous, films of carbide precipitate form at grain boundaries of Type 304 steel at 1200F. These films probably aided the formation of the wedge cracks. Decreasing the carbon content, or otherwise suppressing the formation of the grain boundary carbide films, should decrease the extent of the wedge cracking, (c) The alloy to which no manganese was added e xibited a significantly different type of fracture from these heats which contained manganese. This heat (Heat 1340) had relatively low ductility and exhibited extensive wedge cracks throughout the gage section. 20

Precipitate identification., Some further microstructural data were obtained using other techniques. Electron micrographs of some of the mrtinor microconstituents present in extraction residu.'es are shown in Fi ggur 2 e c2. ese constittuents were identified by selected area electron di f ir ac ti on The triangular particles in Figure 20a, b and c, identified as Ma3 C6, were found in all specimens except those with very lowv ca rbon content. This type of carbide precipitate has been commonly observed (' 1) and is generally believed'to form at twin and grain boundaries, The bead-like NMI C6 particles in Figu re Oe were formed together vwth the triangular particles P Figure 0Zd shows a grid-like set of particlec, fr.fom a specimen (13\ from Heat 1339A which may be M6(j rather than M23. C,6 Ki:nzel described these part icles as forming at twin ends. In addition, several smallt irregular particles were also found in this specimen and these particles appeared to be M6 C. It should be noted that the M23 C6 and M6 C carbides are both facecentered cubic and have lattice parameters which vary between 10 and 11. with the parameter of M6C being slightly larger. F'or this reason it was very difficult to distinguish between these two phases using ordinary selected area electron diffraction techniques. ihe only specinmen studied which had a very low carbon content also had a very high nitrogen content (0. 23%o)/ The particles shown in Figure 20f and 20g are from this material (IHea.t 1343B),., No pattern could be obtained fromn the particles shown in Figure 21g and, unfortunately the particles in Figure 20f were not positively identified,. The elctron diffraction pattern from the particles sh wn in i Figur-e Zf exh'bited two for ns, one pattern being diffracted normnal to the [Lo00 cube direction and one normal to the i0 l direction. The d values indicate that both patterns were from the same compound. This compound was cubic and had a lattice parameter of a - 5, 7X, No likely compound fitting this pattern was found in the ASTM Card File, 21

Ferrite4, During the course of the icrostru.ctural investigaion, the ferrite content of various cold worked mnaterials was neasured using a.ni Arnin co-Brenner Magne-Gage. The calibration curve ased wa.s essentially that developed by Simpkinson and Lavigne, ( The ferrite content was measured on samnples which had been heat treated at 2050(F and in some instances 1750 F, prior to cold working. These data are tabulated in Table VI, The results of this study are as follo.ws: (a) Increasing amounts of cold working resulted in the formration of increasing amounts of ferrite, (b) Increasing "purity" of the alloy resulted in increasing a.nlOints of ferrite at all levels of cold work.. With. simil.ar am. ounrts of cold -work the ferrite content of Heat 1310A with neither caorbon nor nittrog.en was. mmuch' greater than that of Heat 1318 wisith 09, 096 percent carbon or Heat 1343A with 0. 14 percent nitrogen or Heat 1341A with 0, 6 percent carbon and 0 86 percent nitrogen,, (c) The production tubes PT-8 and PT-9 showed essentially no ferrite, even after severe deformation^ The high austenite stability of the commnercial tube PT-9 is most probably due to the high nitrogen con-' tent of this heat * The high stability of tube PT-8 is also taken as being indicative of a high nitrogen level9 (The great austenite staabilizing power of nitrogen is well documented in the literature, see, for example, Ref, 4, 18, 19, 22, 23, 24, 25,) Tubes PT-'10 and PT-ll with lower nitrogen levels, but otherwise sinmilar compositiosns did become unstable under cold defornation and some ferrite was observed., to have formed. (d) Specimens that contained ferrite prior to testing, whether the result of cold working or not, contained no magnetic phase after creep-rupture testing at 1200 "F This implies that the ferrite was converted to sigma phase and/or austenite during testing. The slightly greater austenite stability of the commmer cial heats over lth aboratory heats of simila.r arogen ad c: arbon content is meot likely due to the sonmewhat greater homogeneity of the corn n"'erci"al mater ial. 22

(I) Recrystallization. In the investigation by White and Freeman it appeared that not all of the Type 304 naterial had recrystallized during heat treatmment for l bhour at 1750"Fi This, it was suggested, might be related to the low strength of: the material after 1750Y1 heat treatment. To check this possibility, the recrystallization characteristics of Type 304 steel were re-examined; a detailed description of this study is presented in Appendix B. The following observations were imade: (a) Cold reductions of 45 and 25 percent were sufficient to give a comnpletely recrystallized grain structure after L/ahour at 1600 F and 1750"F respectively. The low rupture strength at 1200 F after 1750"F heat treatmentls of material treated in this way cannot be att:ributed to lack of recrystallization since all material tested had recrystallized prior to testing(b) Precipitation of chromium carbides could pree de recrystallization at temperatures from 1400" to 1750 "F This precipitation occurs on the original cold worked structure and can mask the actual recrystallized structure (this, of course, is dependent on netallographic technique). This type of precipitation prior to recrystallization has (1) not been observed in Type 321 steel ) The reason for this difference between the two steels is not presently understood. (c) Carbide preipitation either before or after recrystallization ad a restrictive influence on grain growth. (d) It was further observed that specimens cold worked 45 percent could undergo recrystallization in times between 50 and 500 hours at 1200 Fo At 1200 F, recrystallization apparently aided the formation of sigma phase. This is similar to the phenom-enon reported by Lena and (26) Curry in Type 310 steel3 23

DISCUSSION In order to achieve the objectives of this program the experi.me-n.etal investigation focussed on the influence of variations in chemical composition and h eat treatment. The metallurgical implications of these results need to be discussed prior to general discussion of how the results relate to the objectives of the program., Carbon and Nitrogen The creep-rupture properties of Type 304 austenitic steel were found to be extremely dependent on the concentration of the interstitial. elements carbon and nitrogen, Increasing the carbon content of Type 304 steel resulted in increased creep-rupture strengthss. The strengthening effect gained by increased ca.rbon content was limited by the solubility of the carbide phases at the heat treating temperature, The carbon solubility of Type 304 steel is shown in (27) (Z) Figure 21, The carbon solubility curve for Type 321 steel is also shown for comparison. From Figure 21 it can be seen that the limit of carbon solubility at 1750WF is about 0, 05 percent while between 1950' and Z050 T the solubility of this element is 0, 09 and 0, 11 percent, These values correspond within experimental error to the changes in the 1000-hour rupture strength as a function of temperature of heat treatnent which were shown in Figure 1, The above statements indicate that there is a relation between carbion content and heat treatment temperature for Type 304 steel smilar to the 11minimumn temperature of heat treatment for maximum strength" relation that White') found for Type 321 steel. This relation appears to hold so long as no precipitation occurred during the heat treatment,. The influence of carbon in Type 304 steel was not nearly as strong as was observed in Type 321 steel. (where increasing carbon content in solution led to the formation of increasing amounts of TiC during testing) and can probably be accounted for by the weaker chromium-carbon interactions and/ (2) or solution effects. The correlation presented for Type 321 steel did not indicate a limit to the amount of strengthening with-in te range corn24

positions reported. Again this is probably related to the strengAth of te::iinteraction resulting in compound formation. The lack of a significant difference between the rupture strengths of the high carbon Type 304 material after 19500 and Z050~F heat treatments suggests that there may be sotme other limitation on the influence of carbon in this alloy, The strengthening due to nitrogen is somewhat greater on a weight basis that that due to carbon; on an atomic basis the relative effect is even greater, Other authors have also shown that nitrogen can have a potent strengthen(7 19, as, zg, 30) ing effect on the creep-rupture properties of Type 304 steel, The anount of nitrogen strengthening is probably limited by the solubility of nitrogen in the alloy. The solubility limit of nitrogen however, appears to be greater than that of carbon-" (this was also suggested by Huml and Grant ( and Brady ). The greater solubility of nitrogen compared with carbon is particularly evident after low temperature heat treatment. The rupture properties of the nitrogen strengthened alloy were only slightly lower after the 1750~F heat treatment than after the 2050'F treatment, This was in marked contrast to the behavior of alloys containinig carbon. The data suggest that the strengtlhening influeince of carbonl and n.it:rogsen is partly attributable to solution effects such as atnosphere:f orlt.ion. These types of phenomena have been described extensively in the litera.ture.' In the case of carbon, atmosphere formation is indicated by the strain aging observed in tensile tests at 1200~ (2) The observation that carbides do not appear within the grains until. extended exposure times have elapsed * Unlike the situation in the case of carbon, the survey of the literature failed to uncover the solubility data for nitrogen in 18Cr - ONi type alloys over the desired temperature range. However, there are data. available for related alloys and for various alloys in the nolten state (18, 31, 32, 33, 34). Studies of other alloys containing nitrogen shed some light on the extent of solubility of nitrogen at lower temperatures (Z0, Z4, 25, 28, 35, 36). F ron;- t.e v a arious (dtata it is suggested tlhat at 1900' F the solubility of nitrogen in Type 304 austenitic steel is at least 0. 20 perceint and increased with increasing temperature. At lower temperatures the solubility of nitrogen is less certainx but it may well be in excess of 0. 05 per cent at IZ00~F, 25

indicates that solution effects are operataive. This w e ven:torer evident in the case of nitrogen. Since no evidence of precipitation wasr foti( und, evetn. after very long exposures, the conclusion n.n ust be reac:lhed t...t the strengthening effect of nitrogen did nit rely oni a precipi.ita66tio. n mrecdhaniu.R Thus at 1200~F the strength of -single phased Tvr.ye 304 was apparently governed by the amo.unt of carbon and nitrogen in si s-lti-on.1 Tl. h. is behavior has also been found in Crn r-M-C-N steel,s The mrec ha. nisl.n is. probabily related to the dragging of atmospheres eby dlsloc at io-ns or the p llning:of dis - locations rathern than the microscopic compound prec:ipita.:i on,. Tihe behavior of the carbon ea ring heats during testing after I..lowx temi-y perature heat treatment has been. shown to be qluite differe1t1s: fr-iorin the behavior exhibited. after high temperature heat treatnlent. The differe nces in properties at 1200 pF after various hea t treatrlenttts -tpoint up the diffe.rences between the strengthening effects due to carbon a nd to -nitrogen.. The suggestions of the data as to the factors which control. the properties of the heats after low temperature heat treaitment ar a.s fo(.tows: (a) The low strengths resulting fromn low temperat ure hea.t treat-ment cannot be attributed to lack of recrystallization...All. the Type 304 speci. mi-nens treated at 1750 "F and the omn nercia l t ube treated a;t ibOO Fhad recrystallized prior to testing. (b) In materials containing carbon in. excess of the asoAl.abiityL. l it at lhe heat treating temperature, carbide precipitation o>ccu.trred to som-.e extent on the cold worked structure and at the grain boaundar>ies of the newly recrystallized grains., This had. the effect of restricting grain growth. (c) Th.e low temperature heat trea.tment red.uced the ca.rbon conte.nt of the matrix. This was made evident by carbide precipitation prior to testing and by the generally precipitate-free, deforrn.ed grains..(characteristic of low carbon, low nitrogen heats) observed after testing materials containing carbon and very low nitro gen, I[n the heats con- tai-ning nitrogenn the grains were gen efr. ally free.of arbide p recpi alte but t.hey were6 not greatly deformrred ind.ca't..ing that signifiat aounts.. 26

of nitrogen were not removed from the grains by heat treatment at 1750 F., (d) After 17500F heat treatment, the 1000 hour rupture strengths of the heats shown in Figure I with carbon contents above the solubility limit at that temperature and of Heat 1341A (0. 06%C, 0. 08%N) were reduced to a level characteristic of material with only about 0. 02 percent carbon, rather than that characteristic of material with 0. 05 percent carbon, which is the solubility limit of carbon at 1750 F. (e) Low temperature heat treatment did not appreciably alter the properties of Heat 1343A, which had very low carbon but high nitrogen. (f) The decrease in the number of wedge cracks in rupture specimens subjected to a 1750OF heat treatment is probably indicative of a change in the predominant type of carbide precipitate at the boundaries. (g) Low temperature heat treatment of carbon bearing heats resulted in the formation of more sigma phase during testing at 1200 F than did higher temperature heat treatments. (h) The more extensive sigma phase formation was accompanied by more extensive "boundary thickening". It would appear from the above statements that one or more distinct phenomena may be limiting the rupture strength at 1200F of the alloys with carbon contents above 0. 06 percent after heat treatment at 1750'"F, The following possibilities should be considered: (a) The carbide phases precipitated at the grain boundaries during the heat treatment had a deleterious effect on strength. This is suggested because such carbide precipitation occurred in all the cases vwhere the strength was adversely affected by heat treatments at the lower temnperatures. However, there appears to be no known mechanism whereby the results of both the carbon and carbon-nitrogen heats can be accounted for usinrg the amount and type of carbide precipitation at the boundaries as the strength -limiting factor. (b) The grain boundary depletion of chromiun rsuting from sigma pha~se 27

and/or carbide formation resulted in inherently weak material which limited the strength of the alloy.. First, this form of sigm a phase precipitate should not have an adverse influence on rupture properties. Secondly, the only basis for proposing that the chromiumn depleted region should be inherently weaker than the mnatrix would be that the level of carbon and particularly the level of nitrogen was lower in this region, This latter effect may be tenable, but little support for it can be offered. Furthernmore this would suggest that the deleterious effect would be decreased if the boundary thickening were decreased. Such a case is found in Heat 1341A, where the nitrogen apparently suppressed sigma phase formation and boundary thickening, but the loss of strength was still large, Thus there seems to be no correlation between the boundary depletion phenomenon and high temperature properties, (c) Restricted grain growth due to precipitation may be responsible for low strength. If this is the case, the actual mechansim would probably be something other than simply a grain size effect since grain size itself has been shown to have little influence on properties. (d) Carbon and nitrogen in solution may be the controlling factors as in the case of the high temperature heat treatments. The carbon content of the matrix, however, may be reduced below the solubility limit at the temperature of heat treatment by some other mechanism. This might be due to heterogeneous carbon distribution or to some precipitation phenomena. Such a reduction of the carbon level in the matrix to about 0, 02 percent in an alloy containing 0. 06 percent carbon or more would adequately account for the observed results' The factors limiting the strength of Type 304 steeAl after low temzperature heat treatment are not as clearly defined as those governing the behavior of the material after high temperature heat treatments The factor controlling the strength may be the carbon and/or nitrogen content of the solution and this may be adversely affected by a precipitation phen28

omenon during heat treatment. One further corrent is in order in regard to Type 304L steel. This material has low carbon content (0 0 35%0 roaximu ) in order to prevent carbide precipitation at rain boundaries which may lead to susceptibility to intergranular corrosion 13, 14) It has been observed that nitrogen in Type 304 steels does not contribute as significantly to the process of intergranular corrosion as does carbon. This may be due to the greater solubility of nitrogen or a difference in the nature of the compounds which precipitate. Thus it would appear that Type 304L steel with high nitrogen levels could have xnearly the same strength as Type 304 steel. This has been verified by Brady Titanium and Boron To this point the discussion has centered around eats with carbon and nitrogen additions. Some further discussion is in order relating to the influence of titanium and boron on the strength of the steel. Additions of small amountS of titaniun were made to different heats to investigate any influence which very "reactive" elements might have on the properties of Type 304 steel The results, stated previously, showed that in the absence of nitrogen the addition of about.0 03 percent titanium resulted in an increased 10OO-hour rupture strength, the extent of this increase being dependent on the carbon level of the alloy In heats containing nitrogen the 100 -hour rupture strength was decreased from 4, 000 to 22, 000 psi by the addition of 0 03 percent titanium Nowak has reported titanium to have a deleterious effect on th e rupture strength at 1200"F of Type 304L steel, He attributed the reduced properties to TiN formation, Kozlik in reporting on Type 304L has also shown an effect of titanium on the properties of the steel which can be attributed to titanium c ombining with nitrogen to form TiN. In the heats containing titanium and nitrogen, numerous TiN particles we re ob erved throughout the structure, These were not put into solution 29

during the heat treatment at 20500F, This confirms that the loss in strength of the steel is due to nitrogen tie-up by titaniumn, In addition, the remnoval of nitrogen (or carbon) by titanium would favor more sigma phase formation which in turn -would give rise to iimore depletion of chromium ar ound th e boundaries, This also was observed, Fron Figure 2 it can be seen that the loss in 1000-hour rupture strength of 2000 to 3000 psi would require the tying-up of about 0. 02 to 0 04 percent nitrogen. For this amount of nitrogen to be tied up as TiN, 0., 06 to 0, 14 percent titanium would be require. Thus it would appear that the loss in strength is greater than might be attributed to nitrogen tie-up by 0, 03 percent titanium. The change in the slope of the stress -rupture time curve of Heat 1344B may also imply that somrething morne is involved than simply nitrogen removal. The data, however, are not sufficiently clear to draw any conclusion other than that titanium can decrease the rupture strength of Type 304 by combining with and tying up nitrogen, When no nitrogen is present, or if the nitrogen is completely "tied up', titanium can strengthen Type 304 steel slightly through an interaction with carbon. Despite the efforts of many investigators the mechanisms of the effects which are attributed to boron are not understood. The results foutnd in this investigation indicate that small additions of boron (1b5 ppm) can be very beneficial if compound formation (other than borides) is actively involved, The addition of boron to heats containing carbon, nitrogen or titanium had only minor effect, however, similar additions to heats containing titanium and either carbon or nitrogen resulted in significant increases in the stress-rupture properties. The benefit derived fromn such additions was usually greatest in the shorter time (high stress) rupture tests with diminished effectiveness at long times (see Figure 6). The suggestion that the predominant effects of boron are associated with compound precipitation implies that boron alters the thermodynamic activities of some elements which are involved in compound forminng reactions. The activity change might produce a change in the nature of the compound precipitated,.the kilnetics of the precipitation, or the size 30

and/or distribution of the precipitate. The latter effect has recently been observed in Type 321 steel by Crussard et al ) and has previously (40, 41) (3) been reported for other alloy systems (4, An alternative proposal has been made in which it is suggested that boron markedly lowers grain boundary diffusion and cons equently alters the precipitate distribution within the alloy. Although these proposals are not entirely adequate, they do appear to show the indirect manner in which boron is effective, The rmodynamic Considerations In order to tie together some of the effects found, for the varioucs alloy additions made to the laboratory heats and to provide a basis for the prediction of the possible effects of other elements, it is necessary to introduce thermnodynamic data on compound formation., In general, thermodynamic considerations of compound precipitation must be based on free energy data and the thermodynamic activities of the species involved. The latter data, however, are not available. Some conclusions.may be drawn by considering similar cases and comparing the standard free energy of formation of various compounds. These data are available and several values are given below:' Standard Free Energy of Formation of some Carbides and Nitrides at 1200 0F, BTU/pound mole of C or N (Ref. 42) TiC -75, 000 TiN - 216, 000 Cr N - 48000 CbC - (50,000) AIN - 195,000 CrN - 38,000 VC ~ 35,000 bN - 135, 000 MozN 12,000 Cr3 C6 - 33 000 Si3 N - 95, 000 A14C3 - 22,000 VN 85,000 Moz C: - 17, 000: - In contrast with the other metallic carb ides and nitr ides, the standard free energy of nolybdenum carbides decreases with increasing temperatures (i. e. they become more stable), 31

From these considerations the following conclusions are suggested: (a) TiN should form preferentially to and be more stable than TiC,. (b) Small amounts of aluminum should not react with carbon, however, aluminum should behave about like titaniumn with nitrogen., (c) The free energies of the chromium carbides and chromium. ntrides are of the same order of magnitude and without further thermodynamic data it would be difficult to conclude which should form. fi rst, Even allowing for the high concentration of chromium as compared with the other elements which are present in "trace" amounts, the driving force for the formation of the chromium compounds may be significantly lower than that of many other compounds' (d) As vanadium is known to behave less ideally than silicon in solution it might be expected to form compounds preferentially to silicon. Judging from the free energies of formation listed previously, vanadium might be expected to behave with nitrogen as titanium does with carbon. (e) The free energy data also indicate that molybdenum. by itself should not be aactive compound former in the resdual amount such as are present in Type 304 steel. The above conclusions, despite their rather qualitative nature, show reasonable correspondence to the behavior of the elements added to the various heats., The following results for materials heat treated at 2050 F correspond closely with conclusions drawn from thermodynamic considerations: (1) The strengthening of the alloy realized by additions of carbon and titanium, (2) the decrease in strength noted when titanium or alunm inum and nitrogen were present simultaneously, and (3) the lack of any effect due to the presence of aluminuizi with carbon, In addition the frequency of the formation of chromium carbides and nitrides is adequately predicted by the free energy data. There is no data available on the effect of colurmb;iu and nitrogen on the strength of Type 304, but there is evidence that they combine readily to form bN Cb (43L 44) Columrbiurn 32

is known to behave with carbon simila rly t titani um. The creep arid rup - ture data and the rnetallographic information suggest that the influence of molybdenum is primarily due to solid solution effec'ts. This is consistent with the the rmodynamic consideration, As far a.s vanadiuni, i.s con(45) cerned, Nakagawa and Otoguro have recently reported on its behavi.or in a carbon bearing 18Cr - 12Ni alloy and their results showed it to beha.ve similarly to titaniumn, They further indicated that snmall anounts of va.ra.(46, 47) dium and nitrogen can be put into solid solution,, IHullt et al has recently observed that small precipitates of VN were formed. in Kromarc 58 steel. These precipitates were throught to be partly responsible for the strengthening of that steel. Kronarc 58 is not directly comrparable with Type 304 steel, the former being a 16Cr - 2ONi - IOMn - 2Moa base alloy, however, tlh precipitation of VNx during creep in Kron)a. rc 58 in a manner similar to that suggested by White ) for TiC in Type 321 steel, tends to confirm the suggested thermodynamric szimilarity betweevn the vanadium/ nitrogen and titanium/carbon interactions In addition, this cormbination of thermodynanmic-lmech.a:nistic rea soning suggests that strong interactions like that of titanium / carbon rt. ay be mnost potent in short til-re periods, buat that rmilder interactions rma- be more effective strengtheners over longer time periods. It appears possible th.at the stronger the interaction, the more rapidly its beneficial eff ects m:na-y ijecor rle mnit.igated due, to precipitation, This suggestion is to sone degree borne out in the discussion in the following section. Further mlechanistic speculation seems unwarranted at this point., There is in the literature a general discuss0ion of various aspects of such mtech-'isns (sGl:e Rets,. Z, 48, 49, 50). The primary conclusion to be drawn froum these con.siderations is simply that the'reactivity" of an added element can p-rovide some index of its behavior in Type 304 steel and how its behavior might subseqcuently influence the creep and rupture properties of the steel. This is based on the assumnption that the creep and rupture properties of Type 304 steel are 33

primnnarily affecte d by carbon and nitroen In solu tion, as w'ei as the int-. actions of these and other e'lemerints i.n solid solution and the precipita.tion of compounds at dislocationns. The predictions are necessarily qualitative Du th.ey provide the )l way pothe lthan intuition of ant. icipati ng- possible effects of other trace elements, Long Time Properties The 1000-hour rupture strength has been frequently used for comparison of the alloys studied in this investigation, For most tubing applications, however, the properties at much longer times are of primary interest. For many of the heats used in th]s investigation the rupture data'at 1200E' was sufficient to establish approximate long-time properties, For certain heats data were obtained at 1350~F which were also used in. the estimation of the long-time strength.'These data thave beern used to conc struct the Meiller -l.arson diagram shown in Figure 22, Thhere are two features of this figure that are of particular significance. First, the curves for the carbon and/or nitrogen-containing materials tested at 12 a00F aid 1350))' "'fitted' together smoothly and had saimlar slopes. This behavior was also observed in the data of Brady ) and suggests that a linear extrapolation of the 1200~F data to longer times is justified. Other iX.nv.e8ti aitors have confirmed the applicability of surch an extrapolation, (' Second, thh high strength of Heat 1339A (0. 07%C, 0. 124%N, 0. 032%Ti1, 1I ppm B)'n1 c(i"omparison with the other heats is not maintained at long times. The trend exhibited by H-eat 1339A is siuggestive of the findings reporttd fo::r'i.'Vp:'e sZ I an- d 347 -teels, T-he short time rupture strength of tlhese stabilized steels is significantly greater than that of Types 304 and 316 steels. The slopes of the rupture curves are greater, however, so that the 100, 000-hour rupture strength of the unstabilized steels can be as high or possibly higher than that of the stabilized types., 34

"Old" versus "Modern" Material The metallurgical variables which have been shown to be primarily responsible for the fluctuations in the high temperature properties of Type 304 steel have been discussed in the previous sections. It is nsow appropriate to consider the implications of the findings in terms of comlmercial practice and to atteampt to use the findings to explain why Type 304 steel made during. andii prior to the early 1950's was generally wVeaker than currently produced material, (These materials will hereafter be referred to as the "old" and the "mnodern" steel,. ) Because of the lack of "old" Type 304 material for study in this investigation this latter explanation will necessarily be speculative., Assuming that the thermo-mechanical treatments of the nn aterial were "adequate" or at least similar for both "old" and "modern" Type 304 alloy, there appear to be two nmain areas which should be considered as possible sources of variations in. creep-rupture properties. These areas center around the influence of nitrogen and the influence of certain residual elements, Several mechanisms will be considered through which these elements can cause significant property variations in the alloy, Higher effective nitrogen contents of'fmodern" steel could very well be responsible for its superior properties as compared with those of the "old"' steel. The commercial tube material, PT-9, was studied in this investigation because of the high strength reported for it in the SP-6 investigation. ( This tube was found to have 0, 08 percent nitrogen. Its strength level corresponded very well with that predicted from Figure 2. Several other tubes from the same manufacturer were found to have nitrogen contents of the order of 0. 10 percent. Nitrogen analyses from ten heats of Type 304L steel from another manufacturer range from 0. 055 to 0. 10 percent, averaging 0,.078 percent nitrogen,': This level of nitrogen ^ These heats were quite simailar in base composition and averaged 0.. Z05%GC, 1. 32% I^n, 0. 54oSi., 18.. 57%Cr and 9: 31!Ni.fi U.'.

is considerably higher than the expected level osf 0~. 02-O( 05 percent. This lower nitrogen level is apparently characteri.s tic of the'old' material. In most cases the nitrogen contents of various commmercial heats are not given, so that further comparisons are not possible,. It does appear,: however, that high nitrogen levels are possible (and, indeed, probabl.e) as the result of "norral." present day commn ercial pra.ctice., There seem to be several ways in which a change in nitrogen. level. between'old" and "modern" Type 304 steel could have cosmne about: (a) Melting practice. It should be recognized that a melt with a nitrogen content in the range of 0. 02 to 0, 05 percent is considerably under18, 31, 32. 33, 34) saturated in nitrogen with respect to the air ( 3 3 3 Thus nitrogen pick-up fromn the atmosphere is a very real possibility and has been obse rved in this work and has been repo rted (5, 53, 54, 55) under various circumstances by other authors.'{ 3' Threre could be numerours ways in which melting pra.ctice could influence the nitrogen pick-up. (b) Deoxidation practice., It seems clear fron the results found here that the excessiv e us of deoxidizers, particularly alum inumn could have an adverse effect on the high temperature properities of the material by remnoving nitrogen from the melt, or tying it up as a nitride in the solidified ingots However, rather large quantities of a.luminum., above that required for deoxidation, would be required t ttie up substantial amounts of litrogen - one pound per ton (. 05/0oA1) could tie up 0,' 025 percent nitrogen. Furthermore, aluminum, is by no mneans a standard deoxidizer for stainless steels, Cormmon commnnercial practice also uses silicon and manganese as deoxidizers, if, indeed, any deoxidizer is used at all... For these reasons one is not justified in atri-' As an example, the nitride added to Heat 1343A should have resulted in a nitrogen content of 0. 067 mlax. This nmolten heat was exposed to an atmosphere of nitrogen under a pressure of L/ atnm for about ten minutes. The resulting analysis showed the prese nce of 0. 145 percent nitrogen - a pick-up of about 0. 08 perce}nt in ten minutes. 36

buting to deoxidizers sufficient nitrogen removal to account for th.e difference in rupture properties between the "old" and the m"1'odern." rmaterials. (c) Base composition. One of the more obvious differences between the'old" and the "modernl material is a change in the base compositi on. During the 1940's and early 19508s the composition of Type 304 steel was typically 18 percent chrormiurn anad 8 to 10 percent ntickel wtith thte manganese and silicon contents usually between 0. 10 and ) 0.60 percent., (Material considered for high temperature service, however, rarely had less than 9 percent nickel, ) Toward te 1960's the 9a nganese content was increased to about 1. 50 percent while the nickel was typically 100, to 10 5 percent, Various studies, particularly that of (6) Mionkman et al have found that variations in the nickel content have osnly~ smrnall effects on the creep-rupetere properties of Type 304 steel. Similarly, it has been shown ( that the change from 0, 50 to 2. 00 percent mnagane.se has no appreciable effect on creep-riupture properties,',, H*owever, a study of the available data indicates that the lower strength commercial heats of Type 304 steel. are invariably associated with nickel. and mnanganese on the low side of the allowable range. While this correlation may be fortuitous, at least one aspect suggested by it should be considered further. It is possible that the findings cited above "which relate to laboratory heats may not properly reflect the effect of simila.r comnpositiona.l changesin crommercial practice, Specifically, these changes may resu.lt in. cornsiderable variations in the nitrogen content of the steel, u First, the addi(54) tion of manganese to the melt may bring in nitrogen'or it may increase the solubility of nitrogen in the melt. (2 3 8) In this connection it was Note: "Eliminationt' of the manganese, i. e. reduction to l.ess than 0. 10 percent, was found to have a deleterious effect on the creep rupture properties of Type 304 steel at 1200~Fi This mnay also account for the differences in properties reported in this i ves stigat ion and th se cf an earlier study. (19) 37

observed that the ingots of Heat 1340 with very.low rma.nganese (.10%) and with high nitrogen added to the melt evolved considerable gas during solidification. This was not observed in any other laboratory heat and thus suggests that the solid solubility of nitrogen was indeed lower in the.low rmanganese heats than in those with approxim ately 1. 50 percent manganese, Second, the compositional changes could alter the solidification mode of the ingot. Inspection of the Fe-Cr-Ni ternary system ({Refs. 56 and 57) and estimates of the nitrogen solubility in the high tenlperature 6 -ferrite phase (from Refs, 18, 20, 31, 32, 58) suggests that solidification through the 6 -ferrite region, which would likely be in heats with low nickel, and rnanganese, could result in a heat low in nitrogen content beca.use of low solubility of nitrogen in 6-ferrite. A heat slightly richer in the austenite stabilizing elements could remain co7rmpletely a-ustenitic throughtout solidification and have a high nitrogen content, Haefner et al (18) and Whittenberger et al (0) have shown that increasing manganese and nickel contents resulted in increased solid solubility of nitrogen..A difference in the austenite stabilizing elenments (ni fckel and rmanganese) from. Z 0 to 4. 0 percent could conceivably change the nitrogen content of the rnaterial by as much as 0, 10 percent. This could easily account for the differences in rupture strength between the "old' and "'modern' mnateriali The results suggest that there may be other possible sources of differences in the properties Various combinations of trace amounts of boron and/or other residual umetallic elements such as titaniurm or columbiumna could affect the properties. The present data indicate that a substantial difference (probably greater than 0., 03 percent) in residual. content of titanium or columbiumn would be required to account for the difference in properties between the "old" and the nodern.' msnaterial, However, no evidence has been fcund that would suggest that the residual elemrent content is different in the "old" and the''modern" mmaterial The residual element content of -four "modern' Type 304 materials is given in Ta'ble I; for "older" mnaterials these data are not readily available. The few scattered data on various residual elerents that were foundt: i.ndicated that: Ref. 104 and elsewhere 38

the values in Table I are probably typical for the "old" mraterial also.. Considering that there have always been small variations between the practices of various manufacturers (both in regard to residual content and therrno-mechanical processing) it appears highly questionable whether these types of effects could be responsible for the consistent differences reported between the "old" and the':mnodern" materials, While none of the above possibilities have actually been proven, they illustrate several w ays in which differences in the creep-rupture proper ties could have come about. Changes in commercial practice which result in different nitrogen levels appear to be the most plausible explanation for the reported differences in the rupture properties of the "old" and the "modern"y Type 304 steel. CONCLUSIONS The high temperature creep-rupture strength of Type 304 austenitic stainless steel, as currently produced commercially, is considerably greater than that of material made in the early 1950's and often of m-aterial made in the laboratory. In order to determine the reasons for these diffe rence in strength level and to determine the factors controlWling the high temperature strength of the alloy, this investigation has focussed principally on the influence of chemical composition and heat treating variables on the properties of Type 304 steel. The conclusions of the investigation pertaining to the factors controlling the bhih termperature properties of this alloy can be sunmmarized as follovs: (a) The creep-rupture strength of Type 304 steel at 1200.F is primnar ly controlled by its carbon and nitrogen contents. Increasing a.cmnounts of these elemients in solution prior to testing (or service) can impa-rt increasing strength to the n aterial, Carbon is a less potent s:trengthener than nitrogen and increases in carbon content beyond 0. 1 0 pe:r39

cent had no influence on rupture strength at 1 00 "F. Nitrogen is a very potent strengthener in Type 304 steel at levels up to at least 0, 15 percent and possibly at higher levels, (b) The role of various trace elements in Type 304 steel is determnined by their "reactivity" with the carbon and nitrogen in solution in the alloy, Trace elements can have the following types of effects: (1) A deleterious effect if the element is so reactive as to remove carbon. or nitrogen from the solution prior to, or during, heat treatment (titanium has this effect with nitr.ogen), (2) A beneficial effect if the elenment is moderatel y reactive'with carbon and/or nitrogen, can be put into solid solution during heat treatment and can precipitate during subsequent testing (titanriumtl has this effect with carbon and it is likely that vanadium wou ld have this effect with nitrogen). (3) No appreciable effect if the element is not reactive or is only weakly reactive with carbon or nitrogen in comparison with other elements present in the steel (molybdenum appears to be such a:n element) In case (2) above, it was found that if the "reactivity't was moderately high the beneficial precipitation effect could be dissipated rather quickly so as to result in substantially improved short time properties but relatively unchanged long time properties (this appears to be the ca'ses with. t:ie-: titanium interaction with carbon)(c) The influence of varying heat treatment temperature appears to be related to the solubility of the carbides and nitrides, The presence of relatively stable compounds required that higher heat treating temperatures be used to achieve maximumn strength. xExcessively low heat treating temperatures (1750PF) had a particularly harmful effect on alloys in which the carbon content was considerably above the solubility limit of carbon at the heat treating temperature; the strength of these materials was lower than the strength expected in a material'whose carbon content -was equal to the solubility linmit at that termper40

ature. The reason for this is not understood but it mnay be related to precipitation during heat treatment. A minimum temperature of heat treatment of 1800.F for cold worked ma.terial appears to develop maximum or near maximum high te mperaturAe properties in material with normal carbon contents, (d) Heats with high nitrogen levels (around 0O 10%) and very low manga.nese levels (around 0. 1I0%) have very low rupture ductility and low creeprupture strength at IZOOF, These findings, together with available high temnperature property data and the known variations in chemical compossition and.n thernmomehammca;.. processing that are possible in commercial practice allow the follvoing conclusions relative to the difference in properties between t-he ^1old a.-d "r'odern"' Type 304 steel: (a) The difference in high temaperat'ure properties between the "old'" and "r'rnodern" mraterial cannot be attributed to the effects of trace eleinrents, (b) Differences in nitrogen content between the "old" a.nd the t":' odern"y material can account for the observed differen ces In the few c ommer ^cial heats for which nitrogen analyses are available, there was a marked difference in nitrogen content, ranging from about 0. 02 - 0, 04 percent for the "old" material (averaging about 0 03/o)t to approxiEma'tely'. 04- 04 10 percent for the "'lodern" zmaterial (averaging about 08%),/ (c) It is suggested that this difference in nitrogen has probably corne about through changes in melting practices and/or the small compositional change of the metallic austenite stabilizing elernents (mrangan-ese and nickel). The latter possibility would lead to increased nitrogen solu-twbility and/or-^ a change in the phase equilibriumu during solidification, which mright result in greater retention of nitrogen in the tmetal. 41

RECOMMENDATIONS The findings of this investigation suggest that consideration of nitrogen in Type 304 steel in needed, This conclusion follows regardless of whether or not nitrogen is responsible for the diferences between the "old" and the "modern" m aterial, since it has been shown that nitrogen has a substantial influence on the creep-rupture properties of Type 304 austenitic stee at 1200 F, The proper method of specifying nitrogen content is open to question., The approach which is suggested at the present time is to combine the carbon and nitrogen levels into a parameter, recognizing from Figures 1 and 2 that nitrogen is about 25 percent ore potent than carbon in strengthening this alloy, Subject to certain qualifications it is thought that the paramneter (%oC) + 1 25 (o0N), should be set equal to or greater than sone minimum value0 The primary qualification of the above statement is that the carbon and nitrogen must be in solution before the material is subjected to creep conditions. Figure 23 was prepared to aid in the determination of a minismum value for the combined carbon and nitrogen level. This figure presents the extrapolated 100, 000hour rupture strength of all the`'carbon-nitrogen" laboratory heats from this investigation and from that of White and Freeman (I) and data from the few commercially produced raterials for which nitrogen analyses are available, Specific information pertaining to the commercial materials used in Figure 23 is given in Appendix C. It should be noted that these materials are similar in that they are all compositionally balanced so as to be wholly austenitic, however, the prior histories of the various materials differ considerably If the'basis for specifying carbon and nitrogen contents is taken as a minimum 100, 000hour rupture strength of 8, 500 psi, then from Figure 23 a minimrum value for (oC) + 1.25 (%oN) of 0. 13 would appear to be appropriate. Because of the importance of a. criteria such as dicused above and because the correlation from which it is derived is b;ased on onlry limited4 42

commercially produced mataerial, it is strongly recormnn.enaded tha.at its acceptance be dependent upon further verification. Thi-s m. 1ight be a.ccom..plished by providing nitrogen analyses for materia.l.s -For which elevated tenp.aerature data presently exist, Based on the dlscussion of the differences between the''oldh a3nd the Tnmodern mlaterial, it is possible that low nickel (8, 0 9.. 0 percent) and manganese (0, 10 - 0 60 percent) levels mnay adversely affect th.e soit.ability of nitrogen and its retention on solidification. Although material. with these levels of manganese and nickel is not utsed for seamless superheate-. tubing applications, it is allowed by the present ASTik standards, (A-31 2, A-376, A-213) which cover this material. Therefore it is recomme. ene-.ded that Type 304 material selected for high temperature seamless tuibe aIppmie-n cations have nickel and manganese levels comparable to those of thle cusrrently produced material, iL e. 9, 0 - 11, 0 percent nickel, and 1. 0 -, 0 percent ranganese,. These levels of nickel and manganese should provide adequate nitrogen solid solubility in the steel so that the desired high strength level can be achieved. 43

REFERENCES 1. White, J E. and J. W' P Freernman: "Metallurgical Principles Governing The Creep-Rupture Strength of Type 321 Aiustenitic Steel Superheater Tubing with Limnited Extension to T'ype 304 and Type 316 Austenitic Steels", Trans, ASME,'Vol. 85, Series A, p. 119. 1963, 2, White J,. EP. and J\.'W Freeman: "A Study Designed to Explain the Creep-Rupture Strength of Type 321 Superheater Tubing", Trans, ASME, VoL. 85, Series A, p. 108, 1963. 3o Adcock, FI "The Effect of Nitrogen on some Chromium and some Iron-Chromium Alloys', Journal of the Iron and Steel Institute", p. 117, 1926. 4. Uhlig, H. H,: "The Role of Nitrogen in 18-8 Stainless Stee", Trans. ASM, Vol. 30, p 947, 1942. 5 GCluck, J. V'. and J, W. Freeman: "Effect of Creep- Exposure on Mechanical Properties of Rene' 41, Structural Studies, Surface Effects and Re-Heat Treatment"T ASD Technical Report 61-73 3Part II. 6. Timken Roller Bearing Company: Resume' of Investigations on Steels for High Temperature,.High Pressure Applications, 19601962. 7. Nakagawa, R, and Y. Otoguro: "The Effects of Nitrogen and Boron on Properties of 18 Chromium- 12 Nickel Stainless Steels'" Trans, NRIM, Vol. 5, p. 129, 1963. 8, Nowak, C - "Trace Elements in Type 304 Stainless Steel", Private communication, March 5, 1964,, 9. Nakagawa, R, and Y, Otoguro: "Effects of Molybdenum on Properties of 18 Chromium-12 Nickel Austenitic Stainless Steels", Trans, NRIM, Vol. 3, p. 99, 1961. 10, Dulis, E, J, and G. V. Smith, "Identification and Mode of Formation and Re-Solution of Sigma Phase in Austenitic ChroniumNickel Steels", ASTM STP No. 110, Symposium on Sigima Phase, 11, Gilman, J J,:, "Electrolytic Etching - The Sigma Phase Steels", Trans ASM, Vole 44, p. 566, 1952. 44

12. Low J. R., Jr. "'The Fracture of Metals", PProgress in Materials Science, Vol 12, 1963, Per gamon Press, 13, Kinzel, A. B: "Chromium Carbide in Stainless Steel'", Trans, AIME, Vol. 194, p.469. 1952, 14, Stickler, R, and A. Vinckier:'Morphology of Grain Boundary Carbides and Its Influence on Intergranular Corrosion of 304 Stainless Steel", Trans. ASM, Vol. 54, p 36Zs 1961. 15. Stickler, R. and A. Vinckier: "'Morphology of Grain Boundary Carbides in a 304 Stainless Steel", Acta., Met., Vol. 9., Letter to the Editor. P. 898, t961. 16, Stickler, R. and A. Vinckier: "Precipitation of Chrom.ium Carbide on Gain r Boundaries in a 302 Austenitic Stainles Steel", Trans,, AIME, Vol, 224, p. 1021 1962, 17. Simpkinson, J. V. and M. J. Lavigne: "Detection of Ferrite by its Magnetisnm", Metal Progress, Vol. 55, No. 2, p,, 164, Feb. 1949 18, Haefner, K,, A, F. Lahr,. W. L,, Meinhart and J, J, Kanter. "Property Relationships of Some Cast and Forges Cr-Mn-Ni-N Steels Containing 18 Per Cent Chromium", Trans. ASTM, Vol. 59, p. 804. 1959. 19* Hum, J, K. Y. and N, J. Grant.: "Austenite Stability and CreepRupture Properties of 18-8 Stainless Steel", Trans, ASM, Vol, 45, p.. 105, 1953, 20. Whittenberger, E, J., E, R. Rosenon and D. J. Carney: "Phase Relations in Cr-Ni-Mn-N Steels Above Z100'F", TransE. AIME, Vol. 209, p, 889, 1957. 21. Kasak, A,, C. M. Hsiao and E. J. Dulis: "Relationships Between Composition and Properties of Austenitic Chromium- Manganes eCarbon-Nitrogen Stainless Steels'', Trans. ASTM,'Vol. 59, p 786, 1959. 22g Franks, R., W. D, Binder and T, Thompson: "Austenitic ChromiumManganese-Nickel Steels Containing Nitrogen", Trans. ASM, Vol. 47, p.231, 1955e 23 Rudorff, D. W. "Nitrogen a asn Alloying Element in Cr-iNi Steels", Metallurgica, Vol. 27, p.68, 19 42. 45

24. Tisinai, G, F., J. K. Stanley and C, 1-I, Sa.mans:'"Austenitic Fe-Cr-C-N Stainless Steels"', Trans, AS/M, Vol 48, p 356, 1956. 25. Tisinai, G, F. and C. H.F Sarans: "Phase Relations in the Fe-CrNi-N Systemr", Trans. ASM, Vol, 51, p. 589, 1959, 26. Lena, A. J. and W, E, Curry: "T'he Effect of Cold Work and Recrystallization on the Formation of the Sigma Phase in Highly Stable Austenitic Stainless Steels", Trans, ASM, Vol, 47, p. 193, 1955. 27, Rosenberg, S, J, and C, R, Irish "tSolubility of Carbon in 18-Percent Ch.romium- 10-Percent Nickel Austenite"' Journal of Research of the National Bureau of Standards, Vol 48, p, 40, 1952Z 28. Brady, R, R. " AISI Type 304L Stainless Steel with Improved Strengthl, Trans,. ASTM, Vol. 59 p, 774, 1959, 29, Dulis, E, J., G, V. Smith and E. G, Houston: "Creep and Rupture of Chronium-Nickel Austenitic Stainless Steels','Trans. ASMVot, 45, p.42, 1953. 30. Monkman, F. C, E. Price and N, J, Grant:'^The effect of conposition and structure on the creep-rupture properties of 18-8 stainless steel, Transm ASM, Vol. 48, p 418, 1956. 31. S iithells, C. J.,"Metals Reference Book", Third Edition, 1962, Butterworths Publishers 32. Pehlke, R. D. and J. F, Elliott: "Solubility of Nitrogen in Liquid Iron Alloys", AIME, Vol. Z 18, p. 1038, 1960. 33. Humbert, J, C. and J. F. Elliott: "The Solubility of Nitrogen in Liquid Fe-Cr-Ni Alloys", Trans. AIME, Vol. 218, p. 1076, 1960. 34. Langenberg, F,. C, "Predicting the Solubility of Nitrogen in Miolten Steels", Trans. AIME, Vol. 206, p, 1099, 1956 35. Tisinai, G. F. and C. H-~ Samans: "Phase Relationships and Mechanical Properties of Some Fe-Cr-C- N Alloys", Trans, ASM, Vol. 49, p. 747, 1957, 36. Paranjpe, V' G., M. Cohen, M, B, Bever and C, F,, Floe: "The Iron-Nitrogen System", Trans. AIME, Vol. 188, p.261, 1950O 37. Freeman, J W, and C. L. Clark: "The Apparent Influence of Grain Size on the High Temperature Properties of Austenitic Steels", Trans. ASM, Vol. 38, p. 148, 1947.. 46

38 Kozlik, R, A.: "Developrnents in Alloys for High. Temr.perature Service: Stainless Steel", International Nickel Co., Research. Seminar, 196 1. 39, Crussard, C., J. Plateau and., HenrVy''Th.e Influence of Boron in Austenitic Alloys, Proceedings of the Joint International Con.ference on Creep, Paper 64, p, 1-91, 1963, 40. Decker, R. F. and J, W. Freernan: "The Mechanism of tBeneficial Effects of Boron and Zirconium on Creep Properties of a. Complex Heat-Resistant Alloy", Trans AIt E, Vol. 2 8, p. 277, 1960. 41, Levitin, V, V: "Investigation of the Influence of Boron on the Decomposition of a Supersaturated Solid Solution at rain Bo-undaries in Austenitic Steels", Physics of Metals and Metallography', Vol, 1I., 1No 3, p 67, 1961. 42. Elliot, J. F. and M. Gleiser: "Ther:mochemistry for Steelmaking", Vol, 1, 1960, Addison-Wesley. 43. Rosenberg, S. J. and J, H. Darr: "Stabilization of Asutenitic Stainless Steels", Trans, ASMI, Vol, 41, p. 1261, 1949. 44. Heeley, E, J., A. T, Little and D. F, Darbyshire: "Sonme observations on the niobium content required for stabilization of 18-8-Nb Steels", Journal of the Iron and Steel Institute, Vol. 200, p. 943, 1962 45. Nakagawa, RB and Y. Otoguro:'The Effects of Vanadium on Properties of 18 Chromium-12 Nickel Austenitic Stainless Steels", TransR NRIM, Vol. 4, p. 64, 1962. 46. Hull, F. C,: "A High-strength Weldable Stainless Steel for Elevated-Temperature Servic e ASTM, Sym.posiumn on Advances in the Technology of Stainless Steel - Atlantic City, June 1963, 47. Hull, F. C. and R. Stickler: "Effects of N, B, Zr and V on the Microstructure, Tensile and Creep-Rupture Properties of a CrNi-Mnn-Mo Stainless Steel", Proceedings of the Joint International Conference on Creep, Paper 43, p. 1-49, 1963. 48. Irvine, K. J,, J. D. MIurray and F. B. Pickering: "The Effect of lHeat Tre'atment and Microstructure on the High-Temperature Ductility of 18% oCr-1gNi-llNb Steels", Journal of the Iron and St ee Institute, Vol. 196, p. 166, 1960s 47

49, Garofalo, F, F. von Gerrlmin i gen and W.' F. Domis: "The Creep Behavior of an Austenitic Stainless Steel as Effected by Carbide Precipitation on Dislocations", Trans ASM, Vol. 54, p, 430, 1961, 50. Myres, J." "Hot Ductility of Three Austenitic Steels", British Welding Journal, Vol. 9, p, 106, 1962, 51. Krebs, T, M. and N. Soltys: "A Comparison of the Creep Rupture Strength of Austenitic Steels of the 18-8 Series", Joint International Conference on Creep, Paper 34, p.6-21, 1.963. 52, Murray, D, J. and R, J. Truman: "The High Temperature Properties of Cr-Ni-Nb and Cr-Ni-M 4o Austenitic Steels", Joint lnter national Conference on Creep, Paper 61, p. 5-55, 1963. 53, Smith, G, V., E, J, Dulis and E. G, Houston: "Creep and Rupture of Several Chromium-Nickel Stainles SteelsP", TIrans. ASM,'Vol. 42, p 935, 1950. 54. Carney, D. J, and E. J. Whittenberger": Electric Furnace Steelmaking, Vol 1, "'R aw Material-s p. -.1 75, AIM E, 1962, Interscience Publishe rs. 55. Feild, A, L. and P. R. Gouwens: Electric Furnace SteeInaking Vol. 1, "Melting of Stainless Steel", p 325, AIME, 1962, Interscience Publishers. 56. Amterican Society for Metals: "Metals Handbook"', 1948, 57, Pugh, J, W. and J, D. Nisbet: "The Iron-Chromium-Nickel Ternary System", Trans, AIME, Vol. 188, pZ269, 1950. 58, Flinn, R. A: "Fundamentals of Metal. Castings", 1963, AddisonWesley, 59F Simmons, W. F, and H. C, Cross: "The Elevated Temperature Properties of Stainless Steel", ASTM-STP, p. 124, 1952. 60, Grant,.N. J,, A. G. Buckling and W. Rowland: "Creep-Rupture Properties of Cold-Worked Type 347 Stainless Steel", Trans. ASM, Vol, 48, p 446, 1956. 48

TABLE I Chemical Composition of Four Type 304 Austenitic Steel, Seamless Tubes Tube Compositions weight percent Number C N Mn Si Ni Cr Mo PT-8 0.057 - 1. 19:060 10.44 18.35 0, 13 0 15 PT-9 0,054 0, 08 1.78 0.48 10,41 19 09 0, 03 0.07 PT-10 0,06 0. 028 1 35. 09 10. 50 8.45.13 0 1l PT-11 0. 04 0, 031 1,42 0 50 10, 00 18. 56 0,12 0 04 A^~ -~P S B Al Zr Ti Gb V PT-S 0, 027 0.010 <0. 001 <0. 00 <0. 01 <0. 01 0. 00 0. 0 PT-9 0, 019 0, 010 <0, 001 00 <0. 01 <0. 01 0019 0. 00 PT-10 0.024 0.01 0. 0013 <0, 005 <0 01 <0,,01 5 0, 029 PT-11 0, 027 0. 015 <0, 001 < 05 0 <0 01 0. 017 00 O02

TABLE II Chemical Composition of Laboratory Heats Heat Chemical Composition, Weight Percent Number C N Mn Si Ni Cr Ti B (ppm) Other 1336A 0. 05 0.74 0.54 10.75 19,0 15 1336B 0. 05 0 74a 0 54a 103 75a 19 15 15a 1337A 006 1 50 0.50 10.5 18,5 0 03 15 1338A <0 02 150 0.50 1.05 1 8,6 0 03 15 1338B <0. 02a 0, 13 1.50 0. 50 10. 63a 18. 70a () 03 15 1339A 07 017 0 124a 1,50,50 10 6 18,5 0 03 1.5 1339B 0 07a 0, 13 1.50 0. 50 10 59a 18,45a 0. 03 30 1340A 0 01 (0. 11) 0,08 0 50 tL 125 18.5 1340B 0, 04a (0, 12) 0 08 0. 50 11 25 18 5 1 341 A 0 0.6 0087a 1. 0 50 00 50 10 5 1 8 5 1341B 0. 13a 009 1 50 0 50 10.5 187.5 1342A 0 03 03a 0 10 1,60 0.50 10, 50 18 7 7 1342B 0,07a 0. 11 1, 59a 0 53a 10 51a 18, 77a 1343A 0,02 0. 145a 1 50 0 50 10.5 18 5 1343B 0,02a 0, 23 0. 150 0, 50 0 1085 18 5 1344A 0~ 06 - 1.50 0 50 10 5 18, 5 0,035 1344B 0, 06a 0, 12 1 50 0 50 10 5 18. 5 0 035 1344C 0 06 0,14 1.50 0, 50 10, 5 18 5 0 08 1357A 0. 05 - 150 0 50 10, 5 18.5 1357B1 0+05 - 1.50 0 50 10, 5 18, 5 0. 03 15 1358A <0,02a 0+12 1.50 0,50 10,5 18 5 - 03A1 1358B 0+ 06 0. 13 1 54 0. 39a 10, 17a 18. 18a - -.03 Al 1360A 0. 08 - 1.50 0. 50 10, 5 185 -. 03 A 1360B 0 08a 0 12 1 50 0. 50 10,5 18.5.03 Al 1361A 03 03a 09 051a L 50 0 50 10, 5 18. 5 1361B 0 03 0, 06 1 56a 0 44a 10 10a 18 17a 0 03 15 1362A 0 07a 0 09 1 50 0+ 50 10, 5 18. 5 13623 0 6a 0 13 1 50 0 50 100 105 18, 5 0 025a 15 1363A 0,02a - 50 0, 50 10. 5 18 5 15 1363B 0. 02 0, 12 1+.50 0 50 10, 5 18. 5 - 15 1364A 0 02 - 1.50 0+ 50 050 15 18 5 0 03 1364 3 0 0.2 0 12 1.50 0 50 10 5 18 5 0 030a 50

TA.BL E II conclud ted Nieat C hemrical Coml positio; n, Weig lt Percent Nu tLber ( lN M Si ( Ni C r i B.. B he 1365A. 0 06 0.12 1. 50 0, 50 0. 5 18. 5 1 365 B 0, O 06 0 13 1, 50 0, 50 1.0, 5 18, 5. 36G6A 0. 6 -. 08 50 10, 5 15 158 1. 366B 0, 06a 0 012.81ia 0 50 10, 5 18, 5. 5 -From2 Ref:'. 1 and 6: 1 310A 0, 009a 0, 0'1. 42 0., 46 10. 38 18, 46 1310B 0. 0(09 0, 008 1. 100. 46 1 2, 96 17. 2 20. 00 () 1311 002 2 1.47 0, 48 10, 36 18, 08 1. 281 0 06 0.44 0, 46 10. 59 18, 62 t 28 2 0, 064., 48 0, 46 10, 47 18< 78 1.229 0 083 1., 50 0. 71 12,63 17, 78 3:8 0, 096 - 1 59 048 10,54 1809 1312 0. 1 41 t1 60 0, 42 10, 54 18, 05 1309A 087 0, 049 0, 61 0 40 10 54 18 77 10 309B 00 (09 0.046 2, 06 0. 47 1 0 06 18. 46 a - Acttu al value from coinm. ercial analyst. Other val.ues are based on these values, the ai-m- analysis, and mnelting experience.. 51

TABLE III Summary of Creep-Rupture Tests Specimen Heat Test Rupture Min, Creep No. Treatment Temp. Stress Life R, Ak Elong. Rate " F OF psi hours % % %/ hr. Commercial tube, PT9: PT9-1 2050 1200 28, 000 133, 1 28, 0 20.4 0o 069 PT9-2 2050 1200 22, 000 1047,5 36. 2 23, 2 0 0087 PT 917 2050 1350 20,000 20, 1 43, 7 33. 1 PT9-8 2050 1350 10 000 1803.3 28. 5 24.6 Laboratory heats: 37A-1" 2050 1200 25, 000 1593.2 18.9 9 6 0. 00335 37A-2 2050 1200 30, 000 369.5 23,9 11.0 0 01300 37A- 3 2050 1200 35, 000 123,4 28.6 1 3, 0. 0.6380 38A.1 20 1 05 120 20,000 111 1 75.0 59 5 38A-2 2050 1200 24, 000 23 1 66.4 50 5 38A-3 2050 1200 15,000 1075.5 56.6 43.2 38B-1 2050 1200 20,000 4321.1 32.4 18.4 0. 00160 38B-2 2050 1200 30,000 486, 5 33. 3 22.6 0,01962 38B-3 2050 1200 40,000 33.7 45.0 24.8 0.4200 38B-4 1750 1200 30, 000 131. 3 47. 4 31. 7 0 1125 38B-5 1750 1200 24, 000 665. 0 46. 1 37,4 0. 01825 39A-1 2050 1200 25,000 1799.2 22.1 15,.4 0.0033 39A-2 2050 1200 30,000 721,.2 23 5 14.9 0 0091 39A-3 2050 1200 35, 000 189.7 27. 0 18.4 0. 0756 39A-4 2300 1200 35, 00 173 0 37.4 18,1 0.0656 39A-6 2050 1350 15,000 344.6 43, 8 39.1 0.0210 39A7 2050 1350 20, 000 94.9 46,7 33, 1 0. 1500 39A-8 2050 1350 12, 000 878. 7 38. 7 20.9 0.00405 39A-9 1750 1200 30,000 72. 5 48.6 40,6 0 2740 39A- 0 1750 1200 20,000 824.9 44. 8 35.4 0. 0187 39A-12 1850 1 200 30, 000 231. 5 25.1 17. 4 39A.-13 1950 1200 35,000 135.3 29.0 22.8 39B-1 2050 1200 25,000 1629,2 19.6 11. 7 0, 0041:39B-2 2050 1200 35,000 104. 3 26.0 16. 4 39B-3 2050 1200 30,000 443,3 21,9 14 5 Th'e specimrnmen numribers are based on the last digits of the heat rnumbers. 52

TABLE III continued Specimlen Heat Test Rupture Min. Creep No.. Treatmentt Te Stress Life R. A El. ong R ate ~F ~F psi hours o % / o/hr 40OA- 1 2050 1Z00 25,000 109, 0 12,7 5 1 (0. 080) 40A-2 2050 1ZO0 20, 000 381. 0 8,7 6,1 i0. 00586 40A-3 2050 1200 15, 000 2581.7 4,9 4. 7 0,00088 40B-1 2050 1200 30, 000 33 0 11, 3 9,6 0. 07375 40B-2 2050 120 25, 000 104.5 6 4 7.2 40B-3 2050 1200 17, 00 1490. 5. 2 5. 0. 0079 41A-1 2050 1200 30,000 215.7 17.5 12.9 0.0172 41A-2 2050 1200 25,000 656.6 21 6 17. 3 0. 0055 41A-4 2020 1350 15,000 258.3 20,2 19 4 0 0320 41A-5 2050 1350 12,000 1017,6 15. 8. 0 0.0066 41A-6 2050 1350 20, 000 44 0 23. 3 15,6 0 2250 41 A-7 1950 1200 25,000 677 1 15. 7 9.3 0. 0055 41A-8 1950 1200 41A-9 1750 1200 25,000 234. 266.4 21.9 0, 0500 41A- 10 1750 1200 20, 000 763.1 2. 6 12.2 0.0132 41A- 1 1850 1200 25,000 926 8 16,2 15. 41B-1 2050 1200 30,000 370.7 8. 5 5. 3 41B-2 2050 1200 35,000 126. 5 12.6 66. 1 41B-3 2050 1200 25,000 1000,2 3. 5 - 42A- 1 2050 1200 25,000 530.2 19. 6 9 7 0, 00928 42A-2 2050 1200 30,000 174,8 21.6 17.3 0 03780 4?A- 3 205200 20, 000 2760. 8 16, 0 10. 0 0 0, O0207 42B 1 2050 1200 25, 000 945. 0 12.8 42B-2 2050 1200 30,000 216. 2 15. 8 6. 6 42B-o3 2050 1200 22, 500 1743.9 13 4 11 8 43A-1 2050 1200 25, 000 810.4 19.5 10,4 0. 00758 43A-2 2050 1200 30, 000 228,0 24,9 13. 3 0 0330 43A-3 2050 1200 20,000 2884.9 14.2 7.8 0, 00177 43A-4 1950 1200 43A-5 1950 1200 25,000 766.0 21.,5 18 3 0.00998 43A-6 1750 1200 25,000 640.0 32,6 28Z 1 0. 01800 43A-7 1750 1200 30,000 206.7 36.3 22.7 0 05500 53

TABLE IiI continued Specinmen Heat Test Rupture Min. Creep No. Treatment Tenmp. Stress Life Re. A, Elong. Rate "F "F psi hours % %%/hr. 43B3-1 2050 1200 30,000 367. 7 26, 1 15 5 43B-2 2050 1200 35, 000 81.7 27 9 24, 4 43B-3 2050 1200 25,000 1137. 3 24, 4 14. 0 44A 1 2050 1200 25, 000 147 0 30 2 18 5 0. 02640 44A-2 2050 1200 20,000 819.9 32, 0 26.7 0 00350 44A-3 2050 1200 44B- 1 2050 1200 30,000 128, 0 33. 9 22 4 0. 09500 44B-2 2050 1200 25,000 517 3 28. 7 27 2 0, 01914 44B-3 2050 1200 20,000 1203 1 25 0 19. O 00759 44C-1 2050 1200 30,000 182 0 32,6 18~ 1 44G-2 2050 1200 25,000 755.7 26, 5 22 0 440-3 2050 1200 57A-1 2050 1200 20,000 135.5 30 9 28. 5 0 1380 57A-2 2050 1200 15,000 874.1 22.0 17.5 0,0110 57A-3 1750 1200 15, 000 746.4 48. 8 30.0 57A-4 1750 1200 Q - 571B- 2050 1200 35, 000 4.3 43, 8 34,2 57B-2 2050 1200 20,000 669.7 54 3 46 4 58A-1 2050 1200 30,000 76. 0 27.8 18.9 0, 1580 58A-2 2050 1200 20,000 1396. 7 21, 1 17.8 0. 00583 58B-1 2050 1200 30, 000 151. 0 19, 7 16. 8 0.03650 58B-2 2050 1200 25,000 370.4 12 5 11. 1 0.01002 60sA-. 2050 1200 20, 000 310 8 19 7 16, 7 60A-2 205 1200 17,000 815. 3 15.4 11 1 60B- 1 2050 1200 30,000 166. 3 18.6 15.5 60B-2 2050 1200 25,000 762 3 13,1 14.1 61A —1 2050 1200 25,000 48. 3 28.2 27,4 0 3200 61A-Z 2050 1200 17,000 856,2 18.0 17,8 0, 0099 613B- 2050 1200 120,000 161.4 54,0 39,4 0. 1620 61i:3 Z 2050 1200 25, 000 603.0 43,9 31 0 0.. 0258 54

TABLE III concluded Spe i- hnen IHeat Test Rupture Min. Creep No. Treatment Ternp. Stress Lif.e R A Elong. Rate ^F F ps, i hours %ho o %l./hr.. 62A-5 1 2050 1 200 25, 000 661.1 110 1 2. 5 0, 00480 62,A2 2050 1 200 30, 000 163. 4 17 2 1 5 0 62B-1 2050 1200 35, 000 194. 3 11.6 9 8 6 3AL1 050 1 200 20,000 60.3 3 52. 2 50 4 63-A-2 2050 1200 15, 000 29 1 3 44 8 41. 6 63B-1 2050 1200 25,000 687.3 23. 4 18. 5 0, 01260 63B-2 2050 1200 30, 000 131. 918 26. 1 22. 2 0 08 22 64A- 1 2050 1200 20, 000 7 6.8 52 0 48 9 64A2 2050 1200 15,00 0 0 770 1 46 0 44 9 64B-1 2050 1200 30,000 109. 6 488 8 46. 8 65A 1 2050 1200 30,000 199 3 1 2, 4 12, 0 65A-2 2050 1200 25,000 796 8 10, 3 11. 3 65B-1. Z050 1200 30,000 312 9 22 8 12. 0 0 01465 65B2 2050 1200 35, 000 89 7 24 4 163 0 0765 66A-1 2050 1200 25,000 180 1 73. 5 69. 5 66A-2 2050 1000 16,000 648, 2 63, 0 55, 7 66B-1 2050 1200 25,000 1759.9 32 2 26.2 66B-2 2050 1200 30,000 517,4 33 2 25. 4 i OA-7 1950 1200 20, 000 57. 5 70.0 91 0 0. 5 380 1OA-8 1950 1200 15,000 377. 5 62. 0 85 0 0, 1000 10OA-9 1950 1200 12, 000 1240, 1 56,0 63 0 0.0211 OA-10 1750 1 00 18, 000 82 2 80 0 84.0 0,4200 lOA1 I 1750 1200 14, 000 490. 6 69.0 76,0 0O 05630 1OA-12 17 50 1200 11,500 1074.0 73. 0 98.0 0 02880 55

TABLE IV 1000-Hour Rupture Properties of the Laboratory Heats Temp. of 1000 hr. Rupture Curve Estimated Ductility Heat Test Rupture Slope at on Rupture Primary Comosition Variable Treatment Temp. Strength 1000 hours in 1000 hours Heat No, %C T%N Other, % ~F'F ksi ____ R.A.% Elong.. 1310A 0,009 0.01 - 1950 1200 12,5 175 57 65 1750 1200 11 8.175 73 98 1357A 0.05 - - 2050 Z00 14.7.155 22 7 1750 1200 14.2 50 30 1362A 0,07 0,09 2050 1200 23,5,130 10 11 1341A o006 0,087 w w 2050 1350 12.0.155 2050 1200 233,. 155 25 19 1950 1200 23.5 _ 17 11 1850 1200 24 5 16 15 1750 1200 19 0.190 21 t1 1341B 0.13 0.090 - - 2050 1200 25.2 5 4 1342A 0 03 0,10 - - 2050 1200 22, 5 146 18 10 1342B 0.07 0,11 - - 2050 1200 25.0 30 13 9 1361A 0.03 0.05 - - 2050 1200 16,8.130 18 17 1343A 0.02 0,145 - - 2050 1200 24,0,160 19 10 1950 1200 24,0 - 21 18 1750 1200 23,2.60 31 28 1343B 0.02 0,2 - - 2050 1200 25.5 160 24 i4 1340A 001 0, 12.08Mn 2050 1200 17.0. 150 5 5 1340B 0,04 0.13.08 Mn 2050 1200 18.0 150 5 4 1363A 0.02 - tSppmB 2050 1200 12.0.160 35 30 1363B 0.02 0. 12 15ppmB 2050 1200 24,0,110 22 17 1366A 0,06 - - 15ppm B 2050 1200 15,0 130 60 54 1366B 0.06 0. 12 15ppmB 2050 1200 275. 150 25 20 1364A 0.02 -.03 Ti - 2050 1200 14.5 123 45 44 1364B 0,02 0. 12.03 Ti - 2050 1200 22.0 - (35) (33) 1344A 0.06 -.035 Ti - 2050 1200 19,5 130 32 27 1344B 0,06 0. 12 035Ti - 2050 1200 23.0 ( l28 21 20 1.240 1344C 0,06 0, 14 08 Ti - 2050 1200 24.0.130 26 22 1338A <0.02 -.03 Ti 15ppn B 2050 1200 15.0.122 56 43 1357B 0,05 - 03 Ti 15ppmB 2050 1200 18.5 56 48 1337A 0.06 -,03Ti 15ppmB 2050 1200 26.5.122 20 10 1361B 0.03 0.07.03Ti 15ppm B 2050 1200 23.2.130 42 30 1338B <0.02 0.13 03 Ti 15ppmB 2050 1200 27.5 f.104 32 20 1.245 1750 1200 225. 140 45 36 1339A 0,07 0,124.03 15ppmB 2050 1350 11.5 37 22 2300 1200 29,.230 (25) (15) 2050 1200 29.0, 110 and (.245) 23 15 1950 1200 27 5 - (26) (19) 1850 1200 24.4 - (24) (16) 1750 1200 19 5.160 45 35 1339B 0.07 0.13 03 Ti 30ppmB 2050 1200 27. 5 20 12 1362B 0. 16 0.13,025Ti 15ppmB 2050 1200 28,0 - 8 6 1358A <0,02 0. 12.03AI 2050 1200 21.0.090 21 18 1358B 0.05 0,13,03A1 o 2050 1200 20,2.210 10 10 1360A 0,08 -.03AI - 2050 1200 16.6.150 15 11 1360B 0,08 0,12.03AI 5SppmB 2050 1200 24,3.110 13 13 1365A 0.06 0, 12.19Cu ~ 2050 1200 24,2.130 10 II 1365B 0, 06 0o 13 19 Cu 20 Mo 2050 1200 25.6.120 21 I1 56

TABLE V The Inflence of Trace Amou of aeof Titanium and Boron in Type 304 Laboratory Heats of Varying Carbon and Nitrogen Content Key: (a) 1000 hour rupture strength, ksi. (b) Estimated elongation for rupture in 1000 hours, percent (c) Heat numbe r B (ppm) None 15 None 15 Ti %) - None None.02-, 04 02. 0 04 | % C 0 Ti%N j.01-, 02 <.o01 (a) 12,5 t 12.0 14. 15,0 (b) 63 30 44 f 43 (c) 1310A 1363A 1364A 1338A,05-, 07 <.01 (a} 1 15 15 19.5 26.5 (b) 17 54 27 10 (c)} 1 357A 1366A 1344A 1337A 01 12-. 14 (a) 24. 0 24.0 0 27 5 (b) 10 17 (33) 15 (C) 1343A 1363B 1364B 1338 06.13 (a) 25 0 27. 0 I21. 290 0 (b)l 9 20 20 15 (c) ti342B i366B 1344B 1 339A ww~~~~~~~~tS_.ow_~~~~~~~~~~~~~~~ —_-A..-.. -............................................... ----- — W.' —W- -' —-_

TABLE VI F"errite Content of Type 304 Steels Before and After Cold Deform.nation ea Heat Heat Ferrite Content Ferrite Content as Number Treatment as Heat Treated Treated + 50% CW 1310A 2050"F 1 7 28. 1750 2. 3 35. 1310B 2050 0.8 1750.25 1338A 2050 3.1 13,0 1338B 2050 0 1 2 1309A 2050 4 0 1309B 2050 0 2.3 1318 2050 1. 3 1344A 2050 0 1.8 1344B 2050 0 0 1337A 2050. 3 0, 8 1339A 2050 0 0 65 1339B 2050 0 1.6 1343A 2050 0 5 1341A 2050 0. 1340B 2050 0 1.2 PT9 2050 0 0 1750 0 3 at 75% CW PT8 2050 0 0 1750 0 0 PT10 1950 0 2 0 PTl 1950 0 2, 3 58

( I TO' W 0 >) ua 2 e A'of AaaA OUTutxe4uoD Ta09S: 0~ edA jo p s4ea Ao:0exoqeI o Jo 007I 4 q4 T2ux s a9n4drna ano-oo0001 aX uo 43uaLu4ea~x reai jo nrecad a pud p 4uo4uoo uoqxo 7o oua nl"u! ilL I M na TeI 4uW3Jd l U 4UO9 uoqaxer) V I 0 7I 0 O 0 0 8 0' 90 "0 v0'0 zo 0 0 0 O0 -o (O ) " o 00 oa I F 11 \ I.v'.A \o -V - V' I Y, 09 o fmo~ 4 ^ \ I ap \\ o _ a~at~epo0 01 0 ~'Q -ogi ~ eM V I I C,-'8~~~~ I0 _..

30 28 26 ^ Q. r............. - /,. l^. ~,9 / 0-f / 5 24 2 / 110U/ 1 / _ 0 1 00~~~~~~~~18 / mq -'/ 1 /'l ~ 0 2 2 I // I t i (1 ___ C _ 4_____ ______ ___ UL 2 0 0 03Gabo 04 0 11 2 0 Code 1 / 0 1. 03 O/o Carbon X v e0a 0o3Caerbono17 o ^ ^o^ ~ 0. 04 f/o Carbon, Ref. 7 stel 0. T n s 007n % Carybo a 16 — ~ V 0. 14 % Carbon 0 63 0 oQ0 04. 08 0, z o0 l6 0. 20 0. 24 Nitrogen Content, percent Figure 2. The influence of nitrogen on the l'000-hour rupture strength of Type 304 steel at various carbon levels. The numbers near the symbols are the estimated percent elongation on rupture at 1000 hours. do60

50 420 ~~___ i__ J~~~_ ~t ^ - 0]~i~~Heat 1341A' 20 ____od_______________ ______ I___ 30 0. ___006%C 0 0Q8%N Code 1 Heat Treatment I I o0 /?hr. at 2050"F / D hr at 19 _ 50 "F _T- _ ___~ ~ ~ C1 I n' >A ~/hr. at 1850F 40 V /2 hr. at 1750~F Heat 131043A I Open points, tested at 1200 F I ___ _ H 1__eatl_343A ____ ___ 0 009%o C, 0, 014%N 5 heants tested as indicated.350 I0 Z0 30 40 60 80 I00 200 300 40000 0 800 1000 2000 4000 Rupture Time, hours Figure 3. Stress-rupture time curves at 12000 and 1350~F of laboratory heats 1341A, 1343A and 1310A, heat treated as indicated.

_______________________________^ ___ ____ _________S___ _________________ Code ~~ ~ ~ ~~l-~-~~- ~~ ~ ~-~"-~~~~tt ""- Heat C N 60 ~~ ~._ __.~ _. No % % 50 I i __ _I 4 1 HI I I i0 1341A 0.6 0. _50 H f ii....___L+-_I __...................... A 1342A 0,03 0.10 40 - ---- -------- - - -------- -- 1 1 343A o0. 0 14 I ~40 ~~ I Q 1343A (00Z 0..14 $ g l _________ _I_ I H T |1340A 0O01 0O1Z o 30 ~ B "1340B 004 0 0< | I gg-%:L __ G C - | ^-| - ) 1310A 0.009 001 a 0 ~~ -t-~I-<~ o- " 1 I Symbol Heat Treat, Test Temp. 10Q - n.: Code open 2050QF 1 O0 F 1/2 open 1750 1200 l Heat C N Ti B I si_________d 2050 1350 No. % %o % ppm 6 60 0C 1338B <0.02 0,13 0.03 15 _______ ____ _ 61 1339A 0. 07 0 124 Q0. 03 15 50 -~ V 1337A 0,06 - 0.03 15 -- 40 1361B,OQ3 0,07 0,03 15 j 1________ n 3.6 10 atd. Heat 3 131A with vry ow carbon, ni and rsidua nt coten is how 0. 01.002: r6 00802 04 06 08nce20 Minimum Creep Rate, %/hour Figure 4. Stress versus minimum creep rate curves for several alloys containing ca-rbon and/or nitrogen, with and without boron and titanium additiorls. eat treatments and test temperatures are aLs indicated. Heat 1310A with very low carbon, nitrogen arid residual element, contLent, is show.for reference

N~o^^ ~~ ~ _ o4000 s saCode No. 2000 +\ap a %, % 20z ooo X, vi { { XC0 1341A 0 06 0 08 \ 1342A O003 0.10 __ - <Q 1343A 0.02 0 1l45 1000 ____ ^\~\ _____ v t 1310A 0. 009 001 800 60 oo0 0 -, o60 6000 A___ \ ____ _ 0 |......... N... 4 400 4000 o \1 \|] }|} 0 2~~ 0\ _ 00 P100 1000 __ _____ I 08 00 - ---- - - 60 a \ O^ X X1 o i 600 1 600 ~ ~ ~!\\ X\\ x__x ___ 40 I I I\ s a to X > I ~ ~ 400 V"``' H II. 11 X ^ ^\l I ~~ % av g { | { l } \\<S\ 200 04 200 C Gode t < 00 l eat C N Ti B " __ i00 10 0 1338B <0.02 0. 13 0.03 15 \ 80 ~ 1339A 0~07 0. 124 0. 03 15 I~ ~\ ~ 60 ~ V 1337A 0 06 - 0.03 15 _ 1310A 0o 009 o 01 - 40 a open 1750 1200 solid 2050 1350 0..... 0 0- 1 7 0.001.00Z 004 006.01.02.04.06.08.10.20.40.60.80 1 Minimum Creep Rate, 7%/hour Figure 5, Rupture time versus minimum creep rate for several laboratory heats containing carbon and/or nitrogen, with and without additions of boron and titanium. 63

5 0 - ----- EiI - | Heat 1338B: 40 ~ -0 ~~ 0 ——. 02%C, 0. 13oN, 0.03%Ti, 15ppmB 2 30 _=v=' 1 I20 I0 Co Code 1 ___ _i 1o 0 Heat T reatment 0.. hr. at 2050~F _______ _ i Heat 1339A 40 /2hr. at 1950~F 0. 07%C, 0.01240%N 0.03%Ti, 15ppmB0 30 i/hr. at 1950~F ~~O; _, -, o i A A ~hr. at 1850"F 30 1. 0 ^^ V, hr- atl1750'F Xi B B /2?hr. at 1700~F I i i i i I B'/-zhr. at 1600 F 0 1_ -- vi 2z |Open points, tested at 1200~F B Solid points, tested at 1350~Ft 0h io _.__....__.... I _ _ __..........I 40 l -h'I 3 ___0 — I' iij j _ Commercial Tube, PT-9 ~." 30 I ~~o 0 054%C. 082%N'^~ ~ ~~. ~ ~ Q l il lt iIo iI (0 I' I 6 I 1 i I I I i400 I I 0 10 Z0 30 40 60 80 100 200 300 400 600 800 1000 200 4000 6000 Rupture Time, hours Figure 6 Stress-rupture time curves at 1200F and 13500F for laboratory heats 1338B and 339A and the commercial tube PT-9, heat treated as indicated.

30 28 _ 0 A 26 O 24 22 o~ t t I f t~ Code 0 0 Heat C N Ti B No. 7 % ppm O 1339A 0. 07 0. 124 0. 03 15 A 1338B <0. 02 0. 13 o. 03 is V 1343A 0. 0g 0. 145 - 1 8 - 1341A 0.06 0 08. ~ PT-9 0.054 OA082 16 1750 1850 1950 2050 2150 2250 Heat Treatment Temperature, ~F Figure 7. The influence of the temperature of heat treatment on the 1000-hour rupture strength at 1200'F of several laboratory heats and a commercial tube of Type 304 steel, 65

I ~~~~~~~CodeI O ~ ~____ _eat C _TA No. 70 % % 250 13430 02 02 000 0 at ~~ ^ ~ ~~~~~ ~ ~ -~ ~ d9 358 0 02 20 21 -Q --— ~~~~~~~~~~~~~ ~ ------ 40 A 1 —-~ 358B O* 06'O. I 0 0 1360A OS~, 08 P 0 1344A9 0. 06 - 0 3 30 ~1344B 0.06 0. IZ - I I 0 80 100 200 300 400 500 600 800 1000 2 0 "Rupture time, hour s. Figure 8, Stress -rupture time'. curves at I 200' F for laboratory heats with aluminum a itardum additions. Heat 1343A Is shown for comparison. All material was heat treated t12-hour at 2050OF, WQ, Q

I OA-9 12, 000 psi, 1240 rs ~S8BS ~ ~ 38- ll 20"I0B psiIII hS s p,, i~~~~~~~~~~~~~~~~~~~~~~~~~ iaCiaaKiK iIi-.C ^^~~~~~~~~~~~~~~S~ i. ~' 610 38A1~ <0 000-s-:.~i }k 1A "hu" SSgare 9 S ki''(OicrrS p of p i fe 1 a 3 Sfter ruptureItstinaIISIgth ran characteriist~aic of i ateriss~BSw very low carbon and nitrogen BIBIKS~cl~fl881'A" 4' SS'ffll81 B81BS8SIB~~~~~ftlSSIS;%ISIISS%^^^~~(A" ~7BSBS^^-I~t~tll^ C,';5.~.9-" ~~., ~:.;-.-4 "-~ chaaceriti of-.:rnateriai wit.b^.. xr"':.w Ww carbon and nitroge coteit h aeire:nv-skan upue ie r'~~:^-ho-nuc.rpet whd lctriytciiy: h give~.~^?^'-"-.. o utf th u::" _-....~ __- ~.'.:^~^ ^;'-">.60%; 1O..M;:o..uon ^ia^^^^^'^^ ~ ~ 6

~ ~.....I'.~ 1.,' e-~? ~ ~-~-~l'~": I- ~ ~!.... -.......' k, --!;j ~rq~.rr~..`,-,,5~~p ~ ~:'..-/" ~.. I.v5..: A.,,..,.-.;... I~~~ ~ ~~~~~~~ ~ ~ ~ ~ ~ 2 ~'< ~'',"'....,,,...,.~.......? ~..!:o,-~ ~ t / 2....*/,' X.. ~ -.~~~~~- -~~~1..~,..~...... "~, —,;g~, I,~j,:~;L —,',~.*~,->~~m,~,_,,'~ll""..,..-I.-..-..-.!~~~~~~~~~~~~~~~~~~~~~~~~~~~~~,!iw-~! ~'-1 ~, ~ -, I,; -, —.-~,~.r, ~ ~,% *',~:?;l~ "I.~.. ~.-.. I,.-1: -~~~~~ ~ 1,~ d'2'I?-'....C~,.: — -~~~~~~~~~!~ ~-'...:.::: j.2:.,'.ii~1...'k. Itr i~ I~ ~,,.,I "'Ill.~l 1, - " ~? c{, 1.,10.1~~~.1:1-11-4-,:,1v -,,. -~~~~~~-, ~~I ~~~.~~.,I-,.,'a-!~!~ " ~ ~ ~.:!!.:,.. -..?~'~,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~.,, ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~.~0',..B, -lb. e a I ~ ~ ~ ~ t'~a,ed. " Wt 19('~F',~ te'e el:'l:; —0-~'::;" ~;,.,..17~1.1 " sec';~i. e n e ~ "ie ~"..... - i-..,. ~~~~,,,% I.N:.~..........5.,.................... 68

61A-2 17, 000 psi, 856 hrs. 43A-1 25 000 ps1i 810 rs 3 30000 ps 38 0X;0400Lw00 /; 4x'tt::;;.: / 4; 4:....-.._I..-'-...i- _:::.... - $,-IS ^^: - -:::..-^:.,1 -: - i..:-^-f IF A.I? t0; s-' f:.a -~'' -*"- -.. p —-— ":~-.i-l -'.- *^ 1 * -;:-^ 1:,~: %^ ~~~'~^:~-';:-:~:'-::-:: _-::\^:^:-~ -.::;. - - -;~:;!::.:.::::':., -1:.::;:.,:.. 7. 2:i ~/:: Fi _:. 0 6 ag; re t. 0 -^^-t'":"-":-' 1 Photo.icro;arps;':';l -spei esX f heas".:' -.:, of ar. n tr c te.nt- with: 0 20; P: are etched ele trlyticaly ith 60'% H Magni fication - 50 X. ~-x~-.l ~ —:::':-ai -^^ ~-: - or e s9 caron4 p cirnr;eiehea at 2050' F wzro totctbg Th >e en nn~be s St% p s n rupture bfe ar.e Jxen wthe oeU oAollrogr phs. IL the s rrens

3 3 7 hour s 486 hours 43 1 hour s Figure 12. Photomicrographs of specimens from Heat 13383, tested at 1200~F, showing the 7thickening" of the boundaries with increasing time to rupture. The specimens were heat treated at 2050FP prior to testing. Stresses were 40, 000, 30, 000 and 20, 000 psi respectively. Specimens were etched electrolytically with 60 %, HN 0 agnification 500X.

(a) (b) Figure 13. Photomicrographs of specimen 38B-1 fro Heat 1338B. Heat treated at 20500F, tested at 1200Z F and 20, 000 psi. Rupture time - 4321 hours. (a) Recrystallized structure produced by cold working and heating for 10 hours at 1400~F, showing'thickening" (b) Edge of specimen as teted howing continuity of "thickened" boundaries with the depleted surface area 71

I I I I 1. II -.., 1. 1. II;: ,I11,1II 11,IIII.I I... III.,.,,,IIIII IIIIII.I.,IIII II.V,,IIII11 I11I I,..II,.I11-IIo,,IIII.I11.11I-II.II.II*1II.II1.. II.I,,. IIIIIIIIIII II.I. II.IIi,,. 1, 1. 1.III III. II11..I.II-II I11.1...I I IIIII. , IIIIIIIII IIII,.11III,..II0III, I-11:'...,II.11I I III.,% 1;.II,", IIII,I..III II.III...I II,I. IIIIII.I III %.II...IIIII.111I IIII".1;4.I'.I11I,.IIII%...II,III1,II II5 I..I'...I..II..IIII1,III II.,II-I,,f..III11IIIIII II11II. I11::. ,,,;I.IIII,III11 IIIIIIII,I.IIII.I1.I11 I..IIII., I0111I,, I I..'11.'.III,,II1,.I.,, 11I-IIII:1: ., _-,,, -,, ,0II., 11III1. 11- I ". 11 1. 1.I 1I,I II"III:IIIIIIIIIIII- I 1111I11-II11-1. 1.I II.IIt I",II., I ,:II1IIIII.IIIIII1,II11''. 1. .II.I..I1. 11IIIII.I1.I,, I,-.1.`11IIIII.IIIIII I.II,,:.II.1.II..IIII. IIIIII '. I.kI I: II 0 IIIIIIII II. 1-1III1II11I I11: - II11II41I.11II.. 1. II.-I.II.II1.IIIII 1:, I II.1,IIII11 1. 11 1,I1II". IIIIII,, I1. I11 1.. III'., .IIIIIII1,I I.I11 1.I IIII1 III',1. 11III:1II,.11II,::t.1I"IIIII.IIIIII11II1.II-,I-1.I ..I.III. —I1.,I ::::A,-::::.11 4:: , ,,sI11III I11III.,I::*:1 41.-III.. ,:,::...,,,,, *, ". - - IIIII"" II.,11.<Y-IIII, I-..IIII1I11II,.: - .: ,;,I"M,I1,I ,V"..-:.*-,I1,I.I1.II- -,III-I11 1.III-,',,-v I1.-II,:,,,I11I1. 1. 1.I;.w., 11.11I.III.IIII1III1.III1,-II1. 1.. , 1.III *,:-:-,11 11IIIIIII. I.I.I- 1.-I, 1. 1. III,,I II11-. 1..II I 1.I.IIIIIIIIII-II11I,.,III!, I1.IIII,1.., 1.-I-,..I "-. 1.III11I1,IIIII,. 1.,1. ,. I1.IIIIII.I-III1.IIIIIIII1.III.- II,,1.II- I,.1.IIA1.IIIII IIIIIII1.-,,1. 1. - 11 ,1. 1.. I.I-11I0!,III11I::II-..,,,,. II.IIIII11II. II. 1. 1.,-II..IIII-IIIIIIII.- I,1...,I1.,II11II.:1 ". II - I1.9,'.IIII.II"IIIIIII,. -.I I11, 1'., 1..I,-.1.I1.III. -11.1IIIII..IIII I1II.:, 111.1-II"I 1. IIIIIIII1. 1. 11,1. 1.,,1.1IIIIIIIII II-,I11II,.I.IIII.1.I1.III1.,' 1....I1.I1. 1. 1. 1.,,IIIIIIIIIII11..II..1I" ..I1...a _k.1 1,II IIII I1I II'-' 1,,. 11..:I111II11-II1. 11.111.1, I:..I1. -'.IIIIII.,II.I1.-1. -''.III, IIIII-IIIIII III,-I1.II1.. 1..I ,, I:,W-.. I.0,,IIIIII. IIII-I. IIIIII,1.11,II, I..1III,.IIIIII11IIII.I11 - I.. -I-I.II.III.,,11-.1.::::,1.I11 ".-,1I1,II1k-IIIIIIII1.IIII.,I1,I-I1-III1, 1,,..IIIII,. 11.I1.11, I:::,,,II.IIIII1.IIIIII1,I,I':!III1. II..11II11 11,.III: 1.,' I —-.,IIIIIIIIIIIIIII'' 1.,, I,,IIII I, 1.IIIIII..I11 1. -111,.IIIIII.I,. 1.,,I.II,1. 1. 1. 1. II.IIIIIIIIIIIII.," —. IIIIII. -IIIII1.III11I,I.11.11.1..11 1.IIIII,,III1..-I —..I.'%.IIIIIIIIII.11IIII., .:,''.II.I-.III.. ,, , I I1.IIII.II11III1.III,.-, 11.II.,I IIII 1II11 I II..:.,:: :.:1- x , I11IIIIIIIIIIIIIIII". 1.II ': . 1: IIII.-.,,IIIIIIII.IIIIII11I.11.11. 11.. III,.II.I-III". I..,I.II.I.I-I *:::III.::: IIIIII.I,.-II:I:::, : — I-I11I11IIII-IIII-"IIZ-111 11II.I11II.IIIII11II11II.1 1.I II, 11 - 1. ... 11 I... IIII.I,11.II..1.IIIIIIIII.1... 1.,I11, 11-II-1. IIIII.IIIIIIII.IIIII.,I.IIIIII11,.,1..1.. — IIIII,,I1,.II-.::.: -'1.44 III1.IIIIIIIIIIIII.I,.I11 1, I.,.IIIIII-.-I.II.1.I.II1.II.-1. 1. II.,I,.IIIIIIIII.II IIII..,I,:!,_.,IIIII.,I,.I,IIIII': : 1..1.:; I11II,-IIII11 A :II:IIII,.;:::, ii:: IIII.,I:,I.I.... I..III II.I II.II.-III II1.IIII.I1. 11III,,.I.I III- -I 11., —, ., III.I,...II1: ,I:* - : I11 1. 11 11,II, 1.III.IIII1-II11II,,.11 11..,II Ii:i, -, 11...I.III.11III-I..II11.11I-.1111.I! : : , ": i:%,r.I,II- 1.IIIIIII,1.II-,I-II,. IIIII-IIII,11III''.,1. I. 11.1 -.. IIIIII11.11II. II11III, I1.II II - 11.1.,.''.II I, II.III I.,I.III..- - I1... 11IIIIIIIIII III,, I-I IIII..,II, 1. 1, -11,-I, II".I.11.I::: : : .: 1. 11.I11-II II1,II1. I I.,II - ". I.. :: , :." - 1-1.-.I11411"1II11: 4:,,.1...II,:, ::: I: ".,,,.1.".., I*,. ::,II.IIIIII.. IIIIII III II I' II. -I..I: IIIIIII1.II 1..III".I11II11.IIII. I- I..,,.I-I,..IIIIIIIIIIII..IIII.. I-II.IIIII.IIII.I-.,-''I,1.IIII1, I...*III1.I.I,..1 ".., I1.I1,II1.I-IIII-I-II III.-I.II.II11.III. K.--,II,..1.11I11,111.:: . . - I-.1. 1. II-II- -..''I" I11 I'IIII I,::IIII..I1. "I A I1,II__ I.1.-11.1I III...I-,II,::*!':!,::::JII1,III- 1. "I IIII., 11,I —1. 11-,: -:::::,,11I.I, :".II1. : 1 -III.I..-III,,. 1. I::. 1-1.1 I.I,. -11 -, -": ".. 1'1,.1.1. 1.,, 1, I:,.II X, I I-:". ".III.I., 1. II,,II1. ""I1. — 1I11.II..- ". -I1. I'::0:,I -.. I' IIII,,:,:II',' 1I.. "I. 1... I.. I11 I I I1. 11 I'. - . ". 1 11, .: 1 ,: ::: :,_ - 1. I ".11- I 1. -I 11 1.I1.4.......,..1 ". 1..,...,IIIIIIIIII,,..,II,., I-I1.I1111-I ,,-II11 11.I —I1.IIII.1I.1I..II,''.. ,I., ,. I', I 1 I. 1. II- II111.I-III,III.I11II.1...,I.'',-1.II IIII,I,IIIII1.II1. I..,1.I1. 11 1,:. . 1. 1- I,",,. 11 1. 1.I -I 1, : ,,, _0 I,- 7 ,_.I11 1.I,, : 1, . : : .I.I II,... I I11,,.II1.II II.II,_ - "..._, -,.., "... -4,:,I-.,1..IA: : "..I 1.IAV. 0..".IIIIII".II,.II,., II':::,O 16&I1I.1- - I.I-II1. 1. . I 1,.._. .. III- 1. 1. 1. I 1 11 i,,:,::.II. 1. —IIIII1, - )_ I.. . I11 1:.,IIII,- -:': : W - 1 1, 11,"..,I...,II11I1.I11 I-1, 1, ". I',,::.-I-.-I.II I1. -I-I11 -.1 I.. 1. IIIIIIII I1..-III X. :". -11.I,II-.- I- - - I 1.1111, I.,-1.I-`011 *;:, :,..:.::":,. ,I1I.1 11I, I, I,. 11 1.,1I.II: A.I ".. I. I.1III..I. 1. 11 11II1.II. 1.. I..' 11 1. 1. 1. I 1,.'' I,1..I1. 1.IIII 11,II'.1,.1..1.I :::V 1:::,:,:., *.q X:X,,,,I1.I.1,,I11.. 1.-II-1. 1. I.II1.-:I . , 11 1,I'"'.I.1, 11, , 11IIIII11III".,.:.I.,. 1. 1. 1. ,:, I., III.1.. I_,11I1.1I.1. I I1..III I 11I11II',11:: ,': '-. -:. :I I:...,I-,,:.,".II11III,:,., -.111.,,:I1. - I11.II.1. 11II1. I'III1% 1. 1.11.1: 11 1.,-_:III: I,.'. - I , : ,,: -1I11,,.." — ,:.A,"I-1.,'t .I..: <V,, 1.IIIIII1,,..11II III-I. I-, :,. _ II I I11 1,: X:, ,II. :. : A-,: I, 1. I:::, II1. 1..IN.,.I..,,I,::.::I11.1,I11I.I- --'..I I. 11I,..,I1,1.II1. 1.II 1. 1..:::_::::,.,'*'t.... "-. , ,,,,.,.I1.III1, 1.,. : I.III11I11I-I ,.: I::II "...,1.,,,.,I11IIII11,,,, .::::., I.% I'I1. I I ". I I. - 1 ", :.: : . 1 -,: X . I I . ". I 11 I' 1, I I I 4 - 1. "., I' 1 I I., 1I I I - I. I1I-I,.. 1. I., I'-II1,II1.I- -I-I,.1III1". '.. 11. 1, .-III.II.I III.,1.11,,I II.IIIIIIII.I.IIII,-..I ...I II.I:,II.III.11.. I..1. 1.,.I,1.III,1. 1. I..I11-:......1-1.III,11 11111,1.. -I —1. 11 1.I-.,.. 1.IIII-II,-,.: . :: .: ,, 1,IIII —-,..I.IIII 11 -1. 11. I.. . :I11, 11 ,.*o ::.. II.I1 - - 1. 11,- 111111 I:1. - I ::001II-",.-1, :: ikv, -II III -, II1. 11 1.I, ,1. 1.X "II1.I -I.I -1.II, ", I ,., I.. "I 1,.I II1I1,..:: .IIII.1,: :: : 0. :,11II I11II111.11 I.,I- 11 11 1. I'll.. - . :,,.11 1..I.,..11.-... 1.:1'. : %-'.: ,:: .:::x:::: -,.I1. I11 1..IIIIII I,.IIII 1.11.I. -.I:! :..:*: : _."III,.,II.. III... 1. 1.I1..11.I 1. 1.,. '::, . 1.: -::-II1, - ". 1,..,:, 1. ,III'II....1 1,.I... "'. .,,' 0..,. -1.IIII1.II-...11II —.I-II1. 1,,II.I I..I1-1 - 1. 1. 1. -11I.I11.1. 11 1.I- I.,, II1. 1. 1. 1 1.I,,', - 1. ".. - 11-.1.-:.: I ". I 11I 11 I.. I' 11 I., - I...... 1. I I 1. . I.. I I I. I I I.. I 1. I --.1.. - I... I I I I,..1 11 - .... - -.:: -1.III1. -I. 1.,,. 1.,. 11II1. ..:::,.I.. I, I...11. "I III -.III1. , " — :1. 1. -I I,,. 1. 1. ". 1. I -::::I,1. 1. I I1. 1. 11. I..: -,-.-.. I : : I ,. II111. 1, .I, , -1 ",I, ., -.11. I1.. 1. III. 11 11 -, 1::III I1.IIII- I". 1 ..I1 - 11 I' : ::-v- 1. -.11II'll, 11 —.. I: ,.'' 1.I.11I1.,I.".,-I, 7 :.' . :, : ::: I : I I I I. 4 1. I. 11: : - '' I' -..I I 1 1, I.1, - I.I'- ". I I, I I ".,. 1. I.. I 1 _::: -::,:: " .,. I.II..'',III.. II.II-,I.,I. —I1: 11...II1.I11II.-. 1. 11, I ,'..I, I". 11. 11 11 I 1.11,11..7".11-1I- 1. "I 11,1.11,11, 11, - -,.I11II-I, I., -,.,,., -,. -, -.1 1. -,- 1...::: :, I.,,, -1, 1. 1,. -1 I..1 1 111. 1. 1. ". :: . I.." -- - I"..:::, :.,'I.:, I1.I1.7: . ,II :,:. M-::, .. ". 1.. I.,.II,. ,, .: : 1: -"...., 1.I".I.II1,.I,,.-:i:i: , :: '.,.I .. - 1. I' 11-I.0 *I., .1 - I1, I.. 1. III1... 11: I ,:: . 1. -1,I: -III.11 11, I-1. ". I.., II I ".1.:.": I I I. I,. 11'. I.. 1..:: -,11.II,,I1.. III 11I,:: I- I., I". I., 1.I- I.1, - I11 1. 1. I II.-:::: I...''. -,... ,I1.I-. —,.:I.1.".11.::,.. 1,-!!' *`,, ::,., IIIIII'll 11 t : '.. I11,. 1.-,.:x-., .I,11 ".I., 1,.. :: !:::!,,., "..,. 1.-rl 1,::...,II.I.1.,.. 1., 1. 1.-I I-,I: .::,;. 1. "."I. 1;.1-,II-...,1. "..I, 11 I.. 1..1, :.I-.. —II-I ".. "... III11111.. 11 :,,' III III.,I I1 ::.; -,,_,1. -,.-1. I I11 ".,. 1. 1. I1.I * 1. -.1''...,1. 1.II-A 11II.III-1. 1. - - — III::....1..I1. 11,I,11,I .:,;.,. -11.- I-::_-, II.,11. 11,-.II1. 1...",.- IIII,,-,-,- I1II1.I11-., 1. II....-II,.1.1,:*oII'::X..".I II11I1.I11 I I1.I1 —,il 1::.':.:: .::,: :: :, 1. I:. : -Ow: .I0II1,I1. II'I,I1. I.., 1., II-,-,II1. 1. ". 11 111. I XI-.II1-1. I - -I.- I.. - 1. : 7.: II.1 1, I.., I: ::V-I11,, 1. - 11.-1. — 1.-I I I1. "I... -- I'.1. I I'., - . 1. I, 1 I...: III,I,.IIII I,.:.. ,,,,. I-1. I' . O::.1 - 11.. II11.1. 1.-I I "I.,..W o I-. _- 1. 1: 11.11. '. I -.- I I11 1.II,-,.,11i,,..,'',.''."'. _:.:II-1.IIIII-", I...1 I.. 11II1. 1.II-1. ..,:.:: :.::% ".. ".,-.....,1.I1.I11I1.., 11,.,,I.II-.I 0 ,I1. 1. —- 11-I,. 1. 1.I:,1.-II1. 11 1.11 11 1.I.,-I.. -11. 1. 1.II11I- -1. I I —.. .. 11 11 "I 1.I11..,I1.-I11.IIII1, 11I: 1. -.,.1.1.1,.., 11 I,. -I-'' .I1.-I.II.11 1. 1..1 I -1. 1..1. -II,: -:::::::x..: -,1. 11III1,11,::::-II 1. II..,.III,. 11 :_., IIIIII,I:,:I..-1. I,11,:, I.. -1 -I -- - 1. - I- -..1. I -,.''. .-,. .,,- -1.11.1."I :..,*,,1. - .: :::. 7: . I,.., -IIII.1.II:11-1. 1.I,11 1..I11 1. 1..,I,,I11, ,, I: .1, ,I ,. —I1I -. -:II, ,: . ,,!: ,o*:,:. 1.I.I -1.-I.1IIII-1 1.,. I.:. 1. "I I 11.1, ,.-1. 1..I,. ".:: . ::... : I II. 1,.."'I,:I-I11 1. 1. 1. .I.. I' 1. I:, -,.'... I..., I.:,:::, i: 1.-I I-1.I.I..1...I- - "..1 - I —- 1. 1. -1I-1. ' —.- 1. 1. 1: :.''.. I -.... Y 1. II,. I I". I.I I,, I,1., I..q..I.I.I-I .1 1. 1. 1. 1.-I", ". ", 11,. -, p:I:..I..':: - .:, ::... ... III11,,II,-,.I-1.-1. 11. I::::. ., 1.II I —..'. I.. 1,I11,. -.:.I'' 1..II1..1.,,. I.,.. .,.-.''.III. III- -1.I- 1: :::: - I -- ,IIII.III-I,.11 ". -.1. 11I,. 1. 1.I,.,.-1.,I.. 11II,1.I-,III.:I1. 1. 1..-..,- I-I. IIII...I. I-: 1. 1.1.I.II...II11 1.-,1... I *:.:,,1- II%, "...,, I,-.. -. , 1. 1., II,- 1. .,I,1. 1.II, 11..I I-II11 11I..I11 1. 1.II1.-III1. 11I,-: II11.I...I,.I —III,-..1. ,-,. —II.I' 1. ::: ::,: .::,: 1: .:::X:::-.,:-..I1. 1.I.11 1.I.11.IIII-, I,,1.-I11,,1..,,1. 11II 1. 1.I'.., 1. 1.-- Al.1. I.:1. ". "I 1,I1. 1. II,, I 1..1.IIIIII,''-I-1.. x..I.1I- —.I-,.I.,I1..I,1.I11I11,I- 1.-...I 11. ..II-I-..III-.I I::::!::.-:::. -.,, I;.11 I", 11 11II1.III- 1:.1..., IIII. I".III II.I'% - —1.,.1..1.III - I., ".. 1. ::,::.1 1. -I1. — II.. 1.I1.II11I,.1 - I, I,I. I.,I,.11 —I...I,.I,:,..,.. I.I I.I.I.I II.I II-.1 -I — 1. 1. I I11 11..1 11 1. 1.I1. I.''. —11 4.I". 11 -, 1,,..,,.,. 1.- ",1..II,.,1. 1.-I-,-..,!::..III.-I-1. 1. 1. -.II.1 1.. I II-1.II — I-.. -.1 1. 1, 11I,1. ". 1.::,11 ". I'll,1. 1. I.,: :t,,.1' 1. I.",I - ". - 1.11 -. 1. 1. I I ".. 1". , I.''. I II.1. , I 11 1. I'll. I.,. 1.:: ::::,. I'.,, 1.... 1. 1- 1. 1'1..- I.,I1.., -11 1.I.II-.. 1.-I-I- I11 1. 11 1. 11II11 -III.: ,II II11 1.I11III,,.,.,.I.I,,..,I, 1.,,11.;. 111.., —III1. I —,,1.I 1.I1.I-. 1. - —: ". I', I' 1,I- ". .. 1,III1. 1. 1. -:C.:.7 111,II..1.I,11, 1.I1:,I 11- II- 11. 1.1I11 1. 1.,, . I 11 I.. I11. 1%,. . 1..1... 1.II11. —1 II.1 1. I- - I.. 1. :::,,. ..:, I,,....I1..I1. 1.. 11 ,. -,,.1.1 - 1. III,- I1. 1. I.I1..,,1.11 - 1. -II, 1. 11 1. 11,1. 1.,-.- 11 - - 1.. 1. - I..I,:: I,. 1. 1. ,I11 I III,.,....I.II.,:,.. I I 11. 11I- I1. - -:1. 11 ". : -.1 I 1.. 1.I,11 ". 1. 11 I'.., 1. 1.. 1-1 , ".. -. II.1. 1. 11 -,. 1..11 11: : : 1".., I'll,.1, " 1.1, I',.1 - -I' — 1.... , I:,::;:::,- - I-..-I I1..11I:: , , ,... - 1.II1.I ,I1.11.I:I-.1.1 II- - 1. -1 - 1. I -.. 1. "I..1IIIII-II1. -1 ,- 1. 1.,- 1. 1.. - 11. 111..: :,X!:!Iql . 1. - 11 I - I 1. - I I. 11 I I: I' 1. - I '. I1_ - -,,,,x :: .:.:. I. I., I'll 111 I 1... . ::: :: : . ,:::: ::l,,;::: -.1 1. - 11.,:, :,. ". I'- 1, I1. - - I'., I.. "..,. 1. II/11. - 1. - -:..., -.. 1, 1.II IIII',.- I - ". -::.,:1 I.11,I 1.: ,.. .-I11:::1.. 1. -1 11.. - 1. ". I I-1. 1. ".-I I.1 - 1., 11 :..I1,. - - - 1. 1, I.. -,.. - 1. 1. 1.:III', I- 1..I:-:1.1 - 1. I I. 1. 11 I'., I',.1-1-1 I-!- 1, I.. I'. - I,. -;: III.I I. I . IIII11 1...1 11..1: , ::. ,:: : -- 1. I11-,111.#::::::: A... 1. ". I I,1..,,,. 1. I.., —,. ". I I,,..I I,'' -.1-1 - -., 1% 11 1. 11,.. 1. :,. /::.-: . ,..IIII11-I1..I -.1.11I- -, 1. 1. .. I_: .: .:: :::'1_1.1-.11 - 1.,.,. 1. - I.I.- 1.,. ".'-.II,.,!,:..,',.II-I,, -II1.., - 1.III..I,, 1. 1. I,- I- -* : .: !: !-..I I —-II,-::,::: .,,, - A_ -:,,-::: 1.I1. 1. - II111....-I,..1.I,,1. 11 -," 1. 11 -'' 1. 1. 1. I11,1..I11 I IIII 1. : , :-''., I.1 - 1. -I..1 .. 1. I., - -.1!!.11,.:,,.'. ".,1......-I.11 -....1- --:. -,'. 1. 1.. 11II- 1. 1. II.: :. . ''-: " :::-".C: , 1" : - _-,. II -III.111.. II ".. 1. 1. - -I..I.I- - I ,,,I:I. 111-:_': -", I'...'".- X1,-1.1. I-, - - I I - - I 1- 1 , -.,.. - - I .1 - : 11.1, - I II -.II. 1.,,. I1. ".I- 1. I.., I.., I.. 1 ::::- ,1.I,1. 1. I 11. 1. I-,.I.1.,..I.., I' I ,.111.1,.'I,.11 11 -. 1..11.I —-'' - 1. . I,,,. 1. 1. I.1.I:, ,A, 1I,-.. 1:,:. —.1.I., I-1. 11:: I -.,I1..11.11,-.1 - I - I I IIII-1I:.... ...1- —II-I -, . :: X : ..'.::::. 1.. II,.., 1.: 1. ". ".., I:,,..11,.,:: - - 1. 1.,,. I,..1,I- 11 ''I'll.,I,. - 1... I - I,,-:,,::IIII.,. -1.., -.11 - II-I". 1. 11 11 I...1 11 '',.. ,, 11 11: I'..: I II,1.1111,III1. I.. I'. ". 1,I1. 11 I', 1". 1 , 11 1. "..I1.,:::1, 1-11. 1-1-I, I111.1 11 I, 1.,..-I 11:.::::::::::::: — -.I, 1,."I I .I 11,. 11. I-. 11, 11 -1. I'-.. ".. 1. 1. 1. -.II 1- 1: -1-I- "I.." 1, 1, 'I',.-_ I I — II'll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— " ,..,. 1. 11, - -.1.1 I.. 1. I I -I, ". 11. A 1. ".. 11.1.1I:, — 1. 1.,.x:,. -...'':.: X ,:.;:,-,,..." 1., -.I.: . ..::-:- "..1 I'% - 1". I 1. 1.,.-I I-. 11 - 1. 1.. - 1. _: .,'_ ". 1. 1. 1. - I - 1. 1.,.. 11:::,.]_..-,.,,.I." 1, ". 1, I1. 1. . II! - 1. 1. 1-1:I.,.".. I:." , 1 I -:: 1.I,I-'... I.,. ::,..,.,.,:.:,:::, 1. "..,.,..,_ ,:: : ". ".,1 I I. I 1.1 ". I.. II,,, I... I I.,...1., 1. - -- "I...,11... -. -1. ". -. - -.1 - - I. -I.,11 1. - - 1,11::._:-::,:::::... 1..I, I.. 1. III 1,.,..I ,::--.... I I ,.;-, ., - 1. I. ", 1. , - 1. I. -.1-,.,. :'. -11..1 a..1. 1.-%:: : 0. . ::'.. 1. I-.1.111.11. I., -, 1. ,.., 1". I.-,".. - I.. 11.1 11. I'll, - 11 I.. "-!:: " -'',- I1. I -. — 11 -. I. - 1.. 1.-I11 1. 1. -"'' -. -- 1. 1...II.-. -III.1.I,...I-I.. I —..., -.1. I I.''.. - ..''. - I1..I — .- I" 11I,I,:.,.- 1.. — -I..,:: ., "'?. -'II1...,:11. 1..,,- 1. I.I..".. 11-11 -,.. I I I11 —.: 1.... 1.,::.. I' I. 1. 1,... I I% -.., -... .: :::::- :, - 1. 1. 1. 11 I I..11. 11.1..,,-,..,, - 1. 1.. I...I.1 1:.. I -1. I %- I, - "..""I.I. I.. II,1.,,I,,.,. - --,1.II. I.- I 1.. . --.III.. II.1'., I 11.,. ". I'll 1.,..kI.1 yt.:::::.::l::.- I.,- I... 1. I I I... 1.I.,,1. -,-11I. -I - I -.. -,., -: 1. 1. 1. I,,. I'% I.. I I. :'. ".,1.,,a,.''. 0 ". 1:.::I1. -I1. ".I. I11:: -.::::: -.,,. -...1 1.:I'. - 1.. I.".., :.-.:: '::::::::, - -I... , 1. ...:_ ,,?-:, -,,,...'''.,.__ -. __.III I1.-,..1 :: ,:: 1::,- ... ".... 1...I . -. 1. I — "..., I -: : i: :::: -111.1.I1. 1. I.., ",.- 1. I.. -I 1. I., ".. 1::::::: ::-,.,..,..,'' , -,.I,- 1. 11. 11.1 -I,,. -- -.. "..II..II::.,,.1. 11:::11.. 1,J:.. II.I11 1. 1..: __ ". ,_ I, —- 1. ,.: - II'll:I."...-1,,...,.. II.., 1. I ,. 1. I I11., . 1. I'., I.. II I'% I - 11 - ". 1 1. ". ".. 1. 1. ,' 1. 1. 1,,I. I -- - I I.I -...:,.1. 1. 1. I,: —.I .:: :I,... 11 -11,::: :::-:Z_` 1.-:,. 1:,..::,:..,,,., :I.., " II,, I'-.. I - 1,1. !_,:::::: :::.: '::.:: :: :.: :.,.:: I - - - - - I.,.1 ". 1. 1. -I - 1. 1. 11 11. I.- 1: :::.!: .,:.:: -I Y -— 2 I: - - 1.. I I- - ::-v,` -1-1...-.- 1: ...k:. r:!!!:!:,,!. 1. -- — 1. -—..-,.::,.1`111-1. I'-:., I -: I — I - I.- ,.. 1-1-:1- - :: ".-.1". , I-.. 1. :::: ..:,-:: 11 ". I-. I'. 1.,, 1.,. 1.. I, I -, I.,, -- 'i:,'. I- ". I I.. -.1II I.. -. 11: I - ::: - -'. I 1:::. I I1..". I- 1.. I %::,:: 1. 1-1-11''-I.1 11 -1,I I ". "I I', - I- 1. ". '.- I.. 1. -::.: ,. I. 1.. I I I - I'., 1. -I. X.- :::::::- -:_ . ::.::: ::.:. — I' - I I, - ". "..%1.1 1 I - I I 11 - - - -::.. - -.11-1 ''..:. - .....,.I'll,,."...".I.."":...".,.i.:,::.-,. ::::::,:: :: :: . - - - 1. -.1 ,. - "..,. ''..11. —.. 11.11- I'll.. —_- — , -, 1.1j-1.,..- I.. I -I :.::: _. -.1, 1., —.1-11::: :: : .:-,.:,.:: : .. I,-::.11. I .:: : : -, 1. 1, 1,:I.:- I I.1 -, I-,,,I III I-,,. II - .- 1 I, 1, I' I 1. :, . -:. -1.I.I1.,. II'% I I. I1..,I:::..:.1 ..I —-,-: :-,. ... 1,.,:,... : 1.II II..1.I.....1..! -.,::,:::,: :: . :: ::.::::,.., . —.,, .: ,.,7 ", .... 1. - 11 -.-I.,.,. 1.,". 1.,. II . - 1.'1 1.. -- I::,,,.. 1. 1. 1.:: .. - 1, 1. 1..I I — I- - -.: :::,:., I-...-.: ..:".,'', ".1-11 ::: ,1111'% 11-". I ". 1. 11.1'.1,.. . -:I-. - 11 I...- - -.1. 1. 1. I --:::::: -,II 1. I1.11 I:: :,,,4.II-I4..I.. -... 1. - - 1. I I 11. I 1. II11 11 1, 11 ".1 I I-1, ". : - -.:::..,,.: I. I I - 1. -.:. -:: I. . -.1.. 1. I - - -..,.,. 1.... I - ,. I... 11.1 - 1. 1. I.1 — I.,.1,. 1. -.. - -.. 1.:::: ., :::: :::: :::.... I,,1. 1. 1. - - 1., ". - 1. 1.I- - - -I- 1. 1. -1:I- 1. -.-:..1-:1171:_!: 1::.. ",: - -,.-,- 1.I. -.'. I.. ---. - 1. .,. ::::: : I.., ,.. I I: X:::: - _ : .:: I - % 1. — I.1.,,... 1..,.I I, 1. I I- -.. 1. -::.-',.;:,.,..,. I.., - I1. I 1. I...1. - I1. 1. 11 -I11I- 1. I -11 - - I.. 1. I I -- -- - W. I,. I,-... -:: -__-, —. -1.1. -,X:....-..,:'a. 11. --::,,, — " 1.I-.:::::: :--,:*,: -, I.- II.:-: I-,,:......:::.,,. I.. 1.'-: I.:,,.,:. b. ::.:::, .: -, - "4I I - ".:: . .,% -.''.-,,''....

Tested at 12g00F Tested at 1350~F Figure 15. Photomicrographs of specimens from Heat 1341A, heat treated at 2050~ F and tested at 1200F and 25, 000 psi and at 1350~F and 12,000 psi. The rupture life of these specimens was 657 and 1018 hours respectively: Etched electrolytically with 60% HNO. Magnification - 500X. 73

:T., v ,X009 - uoT4,eoTjTUB-e. AjjRZ)T4AjOl43aT9 P;a-qo4a -qd-e-TSo...;au.TT4 gin4dua pue,Issaa4s I.. 1,e pq4sa4 pu-e j.OSOZ Iv p94 I'lqVfCj IVLSC-l sle9H XUO.Tj SU;aXUT:) .: .",.-, —-.:,.,: -. 1. :: - 1. - , - --- - :- : 1.'A .I:,. I -.-, I I -,; .' -., ::.,,. ,., II. I.I. ,-I — ,I.II.,,.,, ;, - -, - - :!; ;,,.,:: :!,. - 4%&, i:" -, , i:, - " - .,'. I, - . I,,, -j! " -., I'I`1.jli-,?. ?: ?, - - -Ii.-l, ", -..,.,. - -!'': :.':!- I". - - - I I-.1 I " I I - - !' —C,-,J!::. ::',.-.,-,'! - - -. ism —,-.-,,,..! --— , —i,.,.,-.a.,. -I -,.., -. !,:. ii..,..,..-k...,, _,,, -- li %_ ,&t , _,,''. , I, -- -....., - P ".. - ",. F........ ,-W 4 —,i.- I -N., ! -.: 0. , "':!,,-,, "., -.... I.,'. "'. t 1,4,.:::!.-, ..7', -,:x:: -,,.`I::.,.,...., . ... I, I.., ,. I -'1'1.11 ":,I '. I I -.,,.,:. .., ,",. I i, -,. - I. &,..:,, 1. . ,:!:;,.i — I,- -., I..:.",./I.1 -.., -,.::::, —I ..,II--....1,. I:.:I` I,, ---- Ale';11,.`P"7!!-., ., !7K, 11 -. ".1 10.!!, I.... l. , -I Z 3::x :i:::,,""',. -- " I -.:o ... " -W:: I - ,.".: -,:! !.:::.-..I -4 ".. I, , l .I .,:,-, 11 - .. ,., - -., !': : ....-. ': "..., I - - i'l.1- : . . .:I.I,...: -: I.- 11 - - -. i ii::,,.-P.-,.`. ::: :: : -"...,,:,,:; -,,,, --u.. b- " - , , ,'..1 I .., ': ,. :. .:., ,.. "II 1, I7t: .,.:. ,, ,I,.,- , ", I:'...,- '..,, _:.,,., --..1,, I'll" -..11-11 -, :,, I, ::: . - — rol I -, ,!'': 7 j - I'..... I. -i:::: ,::., I" k:.,.,.'::,-.:: :_ ,. - I.- " 1., ., .. v.;i,- 1, - ., ,'-.-.. , ...., 11.::: : :!.1.11-.... -, . _.. ,.,i"! -',. "I I... -..1. ii: ' T -.:/,:N -,. , ,.:. : -::::.11,- "...', - . I ::,x!:::,:,'::,.:, - I -.. " - I -.11 : ! :.1 " "., 7::..,,, 1'e;, x:: "- -- , ".-.. i`1,'! " j -,.1 -: N:i:IN 7.-,,,, -. I: , lo:14 -1. - i,', :.1 1::::: , I,..:': ,:: i, 1 - ,, . -.- - ,- I.l.. -.., -:. q -.:, - r,- -, , -,,. -.1.- -.-`:; !,".. I-', ,::.I.. :::: : z:::: r::'-it::.: :-::.,. ". I, "., — l I"' -,:: .. -,I.. 1,... 1.,,;:, - I -I. - ,,, I %. I11I,:: :..-l. -.. -..I,:: . ::`: 4::::. , -F ,. I. 1.:,k.,. -:: - ',. 11-1-1,,,; "I,.. I..: : ::::: ,V:: :: :'....:!.::]'* . : "",I ,..:-IN. .... - - - -: -: :, - II., .,. -1 - j... I 11 11- 1-1.., -" .:- ::' - '', I, -. II:....I.e-,I II- .,w,::i - -.-I II 1-1::' : X ::,...1 1, ". 1-:::, 1. ::: ,.:.4,,... :,::.'.: . :: !: ::::- .:,::..... :,,,:...,'.,::,.w..1 111T :1.1.:- X. , j;,,,,, —,:, :- :: : ::::: 1.1- : - - - -,- - I.":::] -.. -,111-1..-, -.,::,Il.1.,-!JT "!"...- :, .: i1!7-:::!:..,. - .,,.., -::::: :. ,:::,2,1-1 1. I, " ".. ". -, I -.:::: -::, :.::: :::::::: :,:,I, I'1 I:;, .,.:-,",I :.,4 — ,.i,., ..1 - 1, , ,':-':.xl-,:. 11.,.-...-. - - - - . m ,:.. —".::: I,::111- -::.,.: ".. 11 ...11,'.11.,., ..., ",,-I-ll ,-. 1-11'. -1, - I::: 11 ::: -....,, -..::: I I:: 1::-::", ::, I, 1,.. -,,!:4-,'-' :' —!!!.-. -. I ,.. :: r,: -:,-If,. .I.,,.,I11. " -, i,,:,::,.::-: ::,.' , —.., I, : .,, I I. I "I,. x,., -:,;.,.4 4 4 iM ',I,,`,.P. :,:.:::::...:;,. I --.-...-,".."",..-'...., ...- - l::::, ' -, -. - -,. -l".., j ". ,.,:, —. ,:::::, , , 1, i.s.-I -,:: I 1, ,:.:. -,- 1.11, . 11.1.. , -.. I - ::, 7. -::::, 7. 1-1:1 : -. ,-:- ,.,- ::::: :::::.,::: ,.,.., I I, II.I,-: II - : - -::... ,; I.11, - -.1 -, `,,,,!:,,% :._,, A I I. :,, I.. - - -.,.,-;.. .:. .,- ,-,... . - ,.-!t I".- ::X-: I. I ::. ,,,*,`>:.-',:, I.,.' I,... I ..,.: :. — - --,;,, p ,'zer.. .:- - - - -, .,,.1. : I:: .:.,, - . 1 - I., 11......... . mIII:.:: II.I.-. 1...., -..:.: —. k ". ::& - I ,: .;:,O-,:,b, " I `: I,, I..'. II I - -.. I -', n,,, ...., .-: - I.. - ::j:,::! :,-:,..,:" . II." ],::-.:: -, I... -..4- '-.':-',., .:.- I.. I. I.. ,. I... I I I: ]:. : x.,;,:,.,,',-I,..:::.: . -.- !: :,:- -' - - 1- l*,.-:.l.'.....`., f . I. .. . I - ... ,.: !:.... - ., -. 7 —,.'.,,:. .::. I.,. 1.I..,, I.I. x,,, .. o'',Iw l ..:,,.:,: 11;I I, ::,. ,:!:::: :: : x: gk: :i: % 7'.;,!,.,i:,-.-.;i ',,- ! ". ....,... I:,,, ., .I -,, >.,.::,.::`,:: 1:..11 l :.:.j, .' ,, ",: `: " -iii'-". I-.:,w,.'-, , - .:.::7:..::,:: -,,;-,:,,:...:,:4:::,,:-I1'1uI1. II ,. I., : " f!! -.. - -:.....: .I I - I -,I X.T,.k".. - - -.. -. I:::!::!,. i,.-..,.,. -. - ,::: x,::::: ::.:!: -': .,,, .11.; II:::,::: -*,. I -- o sl..::.::.:::. :: ::::. ]: ir`:.: ,.: 1,;-.1".. 7, 1. , I.,.''. ll, i,,,:.,-...,..... ',,.I, .l -.... —. .,:.1..' I %...k, -:.,., ,, I : x:... A..::., ,:" — .,..- --

39A-3 39A-3 41B-1 43B-1 HNO 2, 500X HN03 2, 500X H PO4 3, 700X HNO 2,500X 39A-3 39A-l 4I:B-1 43A-1 HNO3 6,300X H3 PO4 3,700X H POO 3,700X H~=PO4 3, 700X Figure 17. Electron micrographs of several specimens showing various aspects of the microstructure. The figures above the pictures are the specimen code (see TableHI), those below are the etchant and the magnification.

1800~F 1700~F 1600OF 1600~F l600~F Figure 18. Photomicrographs of specimens from PcomercialtubePT-9, testedat 2l00~F. Photomicrographs (a), (b)^ and (c) were takne near Whe fractures (i, e ro a high stress region) of specimens heat treated at 1800~, 1700", and 1600~ F respectively. Photomicrographs (c), (d), and (e} of the specimen heat treated at 1l600^ t F before te sting, illustrate mic Xe rostructural variation between different areas of the specien. The stress and rupture life of the specimens are as follows: (a), 21,000psi, 1147hrs ~ (b), 17,000psi, 682 hrs. (c) 16 000 psi, 788 hrs Etched electroltically with 60% 1N. Magnification f" f-" """4 4 4 "'"' ~':''4'' ~~~e ciey ooiror s(),a e o h pcmnettet t: n n

57A-3: 15, 000 psi, 746 hrs. 43A-6- 25, 000 psi, 640 hrs. l41A-10: 20,000 psi, 763 hrs. 39A-10: 20, 000 psi, 825 hrs. i!: S Figure 19. Photoinicrographs of specimens from Heats 1357A, 1343A, 1341A anrd 1339Aheat treated at 1750~F and tested at l2 00'F The specmlennum'ber, stress and rupture time are given above each photornicrograph Etched electrolytically with 60%HN Magification 50 77 41-1. 0,> 00 sl 6.r,3A 0- 0 0 s',85h above eah photoicro rah. tche elcroyical wIthA

41A-1 39A-3 44B-2 (a) 2, 5oox (b) 1z, 3oox (c) 5o, ooox 0( 39A-3 PT9-2 43B-1 43B1:M,~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~'ZiK'~~~~~~~~~~f:::I~ ~ ~~~~M M.a ~~~~~12OR 0~~~~~~~~~~~~~~~~~n:~l: I (d) I,OOOX (e 5ZOOX (f) 5, ZQOX Figure 20. Electron micrographs of particles extracted with HCI-picric mixe acids, from variou specimens. The figures above the pictures are the specimen numbers. (See TableI.)

2300 2200 2100 olubility of Carbon in Type 321 Steel (Ref. 2) 2000..... 1900..... ^ \ / 1700 1600 1500 - /.... 002 0. 04 0. 06 0.08 0. 10 0. 12 0. 14 Content, percent Figure 21. Carbon solubility curve for Type 304, 18%Cr - 10%Ni, austenitic steel -- from the data of Rosenberg and Irish(27) The carbon solubility curve for Type 321 steel, as derived by White and Freeman( 2 ), is shown for comparison. 79

Time at 1200~F, hrs. - 10 0 100000 000 4 0 30 co A A PT 9- 9 10 -------- -. —.20 - 9.-. _ A. 0133 138A 002 01 0 0| g 139 07 01 2 1 0 1 -o ---- 1 00 0.01 --'6 Soi point - T a 0 34 35 36 37 38 39 40 41 42 43 jCode_____ I~Heat C N Ti B 10 0 1341A 0.06 0.08 - - DT 1343A 0.02 0. 145t 8 V 1338laboratory heats ad of te ci t, Q 1337A 0.06 0.03 15 0 1338A <0.02 0.03 15 <> 1310A 0.009 0.01 Open points- Tested at 1Z00F 6 ~t- Solid points- Tested at 1350'F.. 34 35 36 37 38 39 4041 42 43 T(20 + log t) x 103 Figure 22. Miller-Larson parameter representation of the stress-rupture time curves of several laboratory heats and of the commercial tube, PT-9.

13 I ~ I ~1 z I I ~~- ~~__ 2 10 ~~__ ____ co) 6 101 _________ o _______________ ________ G Corduercial Material 0 5......... ---. ___ — 0.-04 008 0.12 0 o. 16 0.20 0.24 0.28 - C) l. 25x( %N) Figure 23. The influence of carbon and nitrogen on the 100, 000-hour rupture strength of Type 304 austenitic steel at 1200QF, including data from laboratory an commercially produced material. The base composition of these materials was balanced so as to be wholly austenitic. All material was heat treated above 180.0F, but the history of the various materials prior to heat treatment is diverse. For further details, see Appendix C.

APPENDIX A Special Creep-Rupture Tests On the basis of the type of strengthening mechanism proposed by White (2) and Freeman a cursory effort was made to determine whether increased dislocation density, produced by cold working, could markedly alter the creep-rupture properties of Type 304 steel. In addition, it was intended to correlate the recrystallization behavior during testing with creep-rupture properties. In order to accomplish these objectives bars' of Heat 1310A were solution treated at 2050F, W, Q. and cold reduced by rolling 15, 25, 50 and 75 percent; bars from Heat 1282 were heat treated at 2050 F and cold reduced 25, 55 and 85 percent; a bar from the commercial tube PT-9 was heat treated at 2050"F and cold reduced 45%. These bars were sectioned and the resulting blanks machined into creep-rupture specimensi The results of tests on these materials are given in Table A-I and can be summarized as follows: (a) The specimens of Heat 1310A were appreciably strengthened at 1200~F by the 15 percent cold reduction. The rupture ductilities of these specimens were very low in comparison with material heat treated after cold reduction. These data are plotted along with that of the original material in Figure A-1, (b) A specimen which had received 25 percent cold reduction also showed an increase in rupture strength at 1200"F over the solution treated material. The increase was not as great as resulted from 15 percent cold reduction, Microstructural examination showed that a considerable portion of this specimen had recrystallized during testing. The specimens which were cold worked 50 and 75 percent completely recrystallized during testing and as a result had drastically reduced rupture strengths and very high rupture ductilities. (c) In the other materials (Heats 1282 and PT-9) considerable recrystallization occurred during testing at 12000F when the amount of cold 82

reduction was 25 percent or more. The rupture life of each of these specimens was reduced, compared with the solution treated specimens, (d) Microstructural examination revealed that rather large amounts of sigma phase had formed in those specimens which recrystallized during testing, (1, 51, 60) It would be expected, based on other studies, that cold reduction would increase the short time rupture strength. At some time period which is determined by the test temperature, a pronounced increase in slope of the stress-rupture curve should be noted which corresponds to the period in which recrystallization occurs. At time periods beyond that required for the completion of recrystallization, a marked decrease in the slope of the stress-rupture curve should occur. It would appear that the specimen of Heat 1310A, cold reduced 25 percent, ruptured in the transition stage between high and low strengths. The 15 percent cold reduced specimens, however, showed high strength. It is not known whether longer time tests would yield evidence of recrystallization with a corresponding loss of strength in the alloy cold reduced 15 percent. At the longer times sufficient recovery might be induced so that recrystallization will not occur. This is a matter for conjecture. These experiments have confirmed that small amounts of cold reduction, presumably by increasing the dislocation density, can increase the creep resistance of Type 304 steel, In connection with the theories proposed by White and Freernanl for Type 321 steel, two tests were conducted on Type 304 steel to determine if the effect of the formation of a particular dislocation array (substructure) by prior creep exposure at one stress level could have any influence on the creep and rupture properties of the specimen at other stress levels' In Type 321 steel White and Freeman showed that such an exposure can have a marked effect. These tests were conducted on a commercial material. The initial 83

exposure was for 50 hours with a stress of 25, 000 psi at 1200F, The results are listed in Table A-I (compare specimens PT-9 - 1, 2, 3 and 4). To compare the rupture properties, the data should be corrected for the rupture life "used up" during the prior exposure. This can be accomplished using the "addibility-of-life-fraction" rule. The corrected rupture data and the creep rates are as follows: Stress Prior Rupture Life Minimum Creep (psi) Exposure (Hours) Rate (%/ 1 hour) 28,000 None 133 0.069 28,000 50 hours at 25, 000 psi 122 0. 108 22,000 None 1,047 0 0087 22,000 50 hours at 25, 000 psi 1,091 0.0072 There was a slight change in the minimum creep rate due to prior exposure; at the high stress level the creep rate was increased by prior exposured and at the low stress level it was decreased. There were corresponding changes in the rupture properties. These changes, however, are insignificant. The conclusion resulting from these data is that in the absence of a strong precipitation reaction to "stabilize" a particular dislocation array, prior creep exposure does not markedly influence subsequent properties. 84

TABLE A-I Summary of Special Creep-Rupture Tests Specimen Heat Test Bupture Min. Creep No. Treatment Temp. Stress Life R. A Elong. Rate ~F ~F psi hours % to %/hr. PT9-1 2050 1200 28, 000 0 133. 1 28 0 20. 4 0. 069 PT9-2 2050 1200 22,000 1047 5 36.2 23 2 0 0087 25,000a 0, 0240 PT9-3 2050 1200 2000 157,,8 31 5 210 0180 28, 000 108 25, 000 0.., 0245 PT9-4 2050 1200 1020.0 34,7 26. 6 22,000 0. 0072 PT9-5 2050+45oCW 1200 22,00 0 196 1 28,3 14 7 PT9 -6 As received 1200 22,000 454 2 16 0 8 0 (37%CW) PT9-7 2050 1200 20,000 20. t,- 43 7 33 1 PT9-8 2050 1200 10,000 1803 3 28.5 24.6 82-1 2050+25%CW 1200 15,000 508.6 8.7 8.2 82-2 2050+55%CW 100 15,000 71.2 59 4 46 7 82-3 2050+55%CW 1200 20,000 28,0 59.2 42 6 82-4 2050+75%oCW 1200 15,000 25.9 80.7 72.6 10A-' 2050+25%CW 1200 15,000 1100 9 57, 2 33, 1 10A-2 2050+50%CW 1200 15,000 50.2 91 0 91.2 IOA-3 2050+75%CW 1200 15 000 12.9 95.3 76.7 1OA-4 2050+15oCW 1200 20,000 607.8 24,4 9 0 1OA-5 2050+15%CW 1200 15,000 3191 1 15.6 4,6 10A-6 2050+15%oCW 1200 25,000 4.8 48.5 20.4 a - 50 hour exposure at 25,000 psi 85

50 ~~~e 40 30 0 ~ ] ^ 20.4 Q9, 0. 10 Q 0Code i' -~ 844 10277 0 400 600 800 1 4000 Tm, 6hor 8 0 All solution treatments ts D i /2 hour at 205 O~F + 15% CW 6 A 1 /2 hour at 2050~F+ 25% CW_ 1A /2 hour at 2050~F + 50% CW | 1/2 hour at 2050 F + 75% CW 10 20 40 60 80 100 200 400 600 800 1000 4000 Time, hours Figure A-1. Stress-rupture time curves at 1200I F for heat 1310A showing the effect of varyirg amounts of cold work. (The numbers after the symbols are elongations. )

APPENDIX B Recrystallization The recrystallization characteristics of several types of 18Cr - IONi base alloys have been determined. This study was undertaken to observe possible differences in recrystallization kinetics among related alloys, to aid in the understanding of the response of the alloys to various heat treatments and to provide a basis of comparison for materials believed to have recrystallized during testing or high temperature service. Seven alloys were investigated, each with a specific compositional characteristic (a) Base alloy - low carbon and nitrogen - Heat 1310A. (b) Normal carbon content, low nitrogen - Heat 1282. (c) High nitrogen, low carbon - Heat 1343A. (d) High nitrogen, normal carbon - Heat 1341A. (e) Nitrogen, carbon, boron and titanium present - Heat 1339A. (f) Very low manganese - Heat 1340B. (g) Commercial tubing material- PT-9. The alloys were solution treated at 2050 F and water quenched prior to receiving cold reductions of 15, 25 and 45 percent. Aging was carried out at 1200, 1400, and 1600~F for times from one hour to 500 hours. Rockwell B hardness determinations were made on the specimens after aging and these measurements were plotted versus aging tinme. A curve of this type is shown in Figure B-l for the commercial material (PT-9). The time of the start of recrystallization (which was also determined metallographically) and the "time to half-hardness" were determined from the hardness versus time curves. The "time to half-hardness" is defined as the time at which the hardness is midway between the value at the start of recrystallization and the "final" hardness value after prolonged 87

aging (completion of recrystallization and growth). These values are tabulated in Table B-I. Measurements made on a commercial Type 321 alloy are also shown for comparison. Due to the difficulty of obtaining curves such as are shown in Figure B-I, the parameters in Table B-I should only be considered to be semi-quantitative. Considering first the data for the comm ercial material PT-9, which are presented in Figures B-1 and B-2 and Table B-l, the following conclusions can be drawn: (a) In comparison with Type 321 steel, the commercial Type 304 steel recrystallized rather rapidly. For instance, when cold worked 45 percent and aged at 1400QF, the time to half-hardness for tube PT-9 was less than l/ahour as compared with an estimated value of about 40 hours for the Type 321 alloy. (b) The cold reduced commercial tube of Type 304 underwent some recrystallization at 1Z000F. iRecrystallization started between 10 and 50 hours at 1200"F in a sample cold reduced 45 percent. In a sample cold reduced 25 percent, some recrystallization was evident after 500 hours at 1200 0F. (c) The microstructural changes occurring at 1200 TF in the cold reduced samples are characterized by the following sequence of events; (1) Carbide precipitation at grain boundaries and slipped planes which ~is accompanied or followed by general'thickening' of the boundaries'This latter phenomenon was discussed in the section on Structural Examination, (2) Recrystallization initiated at triple points and at grain boundaries in the form of irregularly shaped areas or patches. Some of the': This value, i, e "the time to half-hardness" is simply a convenience parameter and does not necessarily represent a pbysic:allysignificant point such as the time at which "recrystallization" is half completed. Indeed, by the time the "half-hardness" point is reached, in most of the present work at temperatures of 1400"F or higher, there is a complete set of new grains (often irregularly shaped rather than equiaxed) throughout the structure. Whether or not this is to be referred to as completely recrystallized is a debatable point. 88

original carbide precipitates were visible within these areas under high magnification. Within the patches, concentration variations were evident and were apparently related to the concentrations gradients formed at the grain boundaries and at precipitate particles prior to recrystallization (3) At somewhat longer times the boundaries of the small new irregular grains became visible and a new phase, probably sigma, formed in the recrystallized sections. The sigma phase, which was generally etched out by the 60 percent HN03 etch, appeared in a "worm hole" like pattern. This can be seen in Figure B-2b. The sigma phase was most frequently observed at the new grain (26) boundaries. Lena and Curry have also reported observing sigma phase in this form. (4) At still longer times, the sigma phase was less apparent, the new grains grew somewhat and the boundaries of the new grains "thickened" slightly. (d) The microstructural changes which occurred in the cold worked samples during aging at 1400 F were considerably different from those which occurred at 12000F. The following sequence of observations was made: (1) At 14000F carbide precipitation on the cold worked structure vas very intensive, (2) Recrystallization at this temperature was characterized by the form'ation of new gra i:s,;'first near th.e original boundaries' (which,.unlike the case at 12Z00 F, rem.ained distinctly visible because of the precipitated carbide"') and then throughout the old grains. This appeared to occur uniformly throughout the sample, not in distinct patches as occurred at 12000F. (3) No sigma phase was apparent in this material after recrystallization at 1400~F. ^ It should be noted that at low magnifications and when certain etchants were used, specimens such as those shown in Figure B-2 which were aged at 14000F will show only the carbide precipitate in the pattern of the original cold worked structure. Recrystallization will not be evident under these conditions. 89

(e) At 1600OF the behavior during aging after cold working was very sinmilar to that occurring at 1400~F, except that fewer carbides precipitated on the original cold worked structure prior to recrystallization. Some carbides also precipitated at the boundaries of the newly recrystallized grains. For the other materials the recrystallization characteristics varie.+. considerably? as can be seen in Table B-I. The values for the recrystallization parameters (i. e. the time to half-hardness and the time at which recrystallization started) for the material with low nitrogen levels were less than those for the material with high (>0. 10 percent) nitrogen content. All of these values, however, were less than the values characteristic of Type 321 steel, Where possible, the temperature dependence of the recrystallization parameters (the so-called recrystallization'activation energy) was estimrated. Although this quantity is not very accurate there was no indication of any significant difference in this quantity between the different compositions. Furthermore, this quantity facilitated estimation of the recrystallization characteristics at higher temperatures.. These estimates indicated that for all seven materials studied, the time to half-hardness was less than 1/hour at 17500F after about 25 percent or more cold reduction. Indeed, it appears that material with as much as about 45 percent cold reduction (which is common in the production of seamless tubing) will have reached the "half-hardness" point in less than l/hour at 1600F (this can be seen in Figure B-1 for tube PT-9). Careful re-examination of the microstructure of the commercial tubes of Type 304 steel used in the SP-6 investigation indicated that they had recrystallized during heat treatment at 1600 F. (In low magnification photonicrographs this fact can be easily missed. ) Some further information about the recrystallization and growth characteristics can be' arrived at by comparing the variations in the grain size of the laboratory heats. These are given in Table B-tIl The follow90

ing points are of particular interest: (a) After solution trea.tment at 2050~F the heats which only had carbon additions showed an ASTM grail. size number of approximately 2 - 3 reltardless of lthe car)bo: concat'.., (:,) After the 1750)F t:reatment the grain size tended to decreaset wit. inc reasing carbonr content, indicating that grain growth had been rest'rict'eu..)v car ide'p recipitation(c) i- t i.t;rogen c:onte: ts (a out (, 10%) Qapparelln.id a. rstrictive eiect on grain size. Lower nitrogen conte: ts (about 0. 05%), however, had Lit'tle nflu",cn.ce on:tra.in si;e, (d) h.:re. was -no difference i-nr the grain size of specimnens of {T.eat 1 4. 3'({0 145iN, 0, 020tC) heat treated at 2050~F and those heat treated at. 1 750 oF, (e) t'itanliurim, at thle 0. 03 percent level. had only a slightly restrictive e:ffect on the grain size in theeasece of nitrogen during the 2050 3 heat treatment, however, it was markedly restrictive when nitrooen was present, t'hins inforiymation, in the light of tih f-i inlings, sbecom-es o: only seconidary interest to the objectives of this investigation,. T'the relevant co:nclusions will be summarized at this point, without further discussion. These conclusions are intended to apply specifically to material subject to the con(lit:ion-s which have oeen studied herein - this beiing that the material is in a well solution-treated condition. prior to the cold w/orking and aging. The conclusions are as follows: (a.) For all the Type 304 heats studied (the laboratory heats of tilis investigation and ihe production tubing material from the SP-6 investigation.), it was found that heat treatment of 25 percent cold. reduced nmaterial for,'hour at 1750~F was sufficient to result in a'completely":'C recrystallized m:icrostructure. The heat treatment of material cold. reducedl * That is, an entirely new grain structure. This does not necessarily mnean that grain growth was completed or that the hardness had reacihed a minimum value, 91

approximately 40 percent for Vahour at l 600C3F had the same result, (b) The combination of conditions leading to the recrystallization of Type 304 steel was somewhat less severe that those necessary to produce a similar tesult in Type 321 steel. (c) Carbide precipitation, under certain conditions, could occur before recrystallization had annihilated the cold worked structure, This precipitate could (depending on metallographic technique) mask the recrystallized structure. This has not been found to happen in Type 321 steel. ) The cause for this difference between the two steels is not presently understood. (d) Carbide precipitation either before or shortly after recry-stallization had a restrictive effect on grain growth. (e) At 1200OF recrystallization apparently aided the formation of siginta phas e 92

TABLE B-I Recrystallization Characteristics of Commercial Tube PT-9 and Several Laboratory Heats of Type 304 Steel and of a Commercial Type 321 Steel Tube Reduction Temp. Heat Number Type %___ __ F J1 31 A 1282 PT9 1343A 1341A 1339A 1340B 321 Time to half-hardness, hours 15 1200 >1000 >1000 1400 25 1000 1600 <.3.5 25 1200 800 350 >1000 1400 1. 5 1 4-10 20 100 9 0 0 1100 1600 <.2 <.3.3 2.5 3 4 4 45 1200 80 >500 1400 <.4 40. Time to start of Recrystallization, hours 25 1200 50-100 50-100 500 >500 1400.3-.. 5 5.5-1 10 40 50 30 50 1600. 2 2. 1-1, 5 1, 5 1-2 2 These parameters are derived from aging curves, and have the following meaning: I I \ Log Time ---:^^ Time to half-hardness "Time to start of recrystallization 93

TABLE B- I ASTM Grain Size of Some Laboratory Heats after Heat Treatment at 2050"F and 1750 F Composition Variables, % Grain Size Heat No. C N Ti B, ppnm 20501F 1750 1310A 0. 009 3 6 1338A <0. 02 0 03 15 5 1282 0. 06 2-3 8 1357A 0.05 - 1-2 5-7 1312 0.4 - - 3 9 1361A 0 03 0.06 - - 1-2 1343A 002 0. 145 - - 6 5-6 1362A 0,07 0.06 - - 2-3 1341A 0 06 0,08 - 5 7-8 1344A 0.06 - 0. 035 - 4-5 1344B 0.06 0. 12 0.035 8 1337A 0.06 - 0.03 15 3-4 1361B 0.03 0o 07 0 03 15 3-4 1338B <0.02 0. 13 0.03 15 7 6-7 1339A 0.07 0. 124 0.03 15 7-8 6-7 1362B 0. 16 0, 13 0. 025 15 4-5 PT9 0.05 0.09 3-4 8 1340A 0 0. 12 02 8 Mn 3 -4 1340B 0 04 0, 13 0, 08 Mn 5-6 94

1B~~~~~~05 i — 1~~~~~~~0 --- _ \ —a _ _-_ _i - i - 80 H ~ d~eC j~t>' l \;-" - O 15 % Open points - 1200,F I 25% Oen poits 1400 _\ \ I m 7 5 - - t.~2.4.6.8 1 Z 4 6 8 1Q 20 40 60 8_0 100 Z00 406~~ Worked Figure B- 1Aging curves at 12000, 1400, and 1600F for commercial tube PT-9, 80As~ C ol d worked 15, 25 and 45 percent in cross-sectioal area F 0igure B-1- en Aging curve s at 1200~, 1400~, and 1600 ~F for com mercia tube PT-9, cold worked 15, 25, and 45 percent in cross-sectional area.

. Cold;~~~~c~^A^~~~": -4...-.. Reduced N -- — ii a) 502hrs. b) 100hrs. c) 500hrs. 5 10hoso e) 50hrs. ) 500hrs. W~Wf.4~~~~ Figure B-2? Photomicrographs of a commercial material, PT-9. at various stages of recrystallization at 1200~F after 45 percent cold reduction and at 1400F after 25% reduction. Etched electrolytically with 60o HN03, Magnification - 500X

APPENDIX C The influence of carbon and nitrogen on the estimated 100, 000-hour rupture strength of Type 304 steel. e i..l —~.19- - k4- > i b. -_ 0. 139 83 D~ata Hreat lt 00,00o0-hr-.t ~ ~.-" ~ ~ ~ i - 4 g 0o0 4 6 18 004 H 0.2 ( 0. 0024 02S (sC)+ 1 S^x sN) Data Heat 100.000-hr. Point Chemical Co position,,psrcent (%C) + Type of Treatment Rupt. Strength No., C N Mn Ni Cr Si 1,25x(%N) Working OFgpsi eft 1.03.036.07 10.92 18.68 251a.073 H.R 2000 (7. 3 59 2.067.042.16 111 3 18,28 2) 193 H. 9 ( 8.0) 59 3.104.039 (.10) 10.97 18.8 25).153 H.. 000 (10. 0) 30 4.036.034 (<.1) 102 40 185 47 (<. ). 07 H.R. 2050 (810.0 8 19 5.030.10 1.47 9.72 1913 3.64. 1554 1., 1975 815 28 6.025.13 1.63 9.87 18,87.55.188 H.1R. 1975 9.5 28 7.070.04 1.62 9.51 18.88.79.12 H. R. 1975 9.( 5 28 8.05.028.53 10.7 18. 5.61.84 Ho R. 1900 7. 3 59 9.018.14 11.3 19.2.193 - 1950 (10.5) 59 10 018.032 113 t 18 9.058 2 1950 (8. 5 59 11.052.10 1.66 10.40 18.47.5 177 HR. 1950 10.8 6 12.059.10 16 1,9 1 9 33 4I.67.47.184 C.R. 1950 t 1o5 6 13 050.10 1.64 10 45 18.86. 37.170 C.R. 1950 11.0) 6 14.057.096 1.63 10.61 19.15.48.170 CH.R. 1950 (115) 6 15.02.045 1,10 10,56 18,55.32-.076 - 1950 8,2 16.026.13 1.12 9.48 18.,19.34.188 H.,R. 10.2 17.054.082 1.78 10.,41 19.09.48.160 C.1R. 2050 11.2 1 18.06.028 1.35 10.50 18,56.69.095 C0R. 2050 8.2 1 19 4.0 31 1.42 10,00 18.45.50.079 C.R., 2050 9.5 I 0.022 (.01) 1. 47 10. 36 18.08.48.034 C R. 1950 6.7 d 21.083 (.01) 1.50 12 63 17 78.71.095 C. 2050 8.9 22.096 ( 01) 1.59 10,54 18.09.48 o108 C.R. 2050 9.3 d 23.087.049.61 10.54 18.77.40.148 C,. 1950 9.7 td 24.090.046 2,06 10.06 18.46,47.147 CoR. 1950 9.4 Laboratory heats from this investigation: Point No. Heat No. Point No. Heat No. 5S 1310A 31 1362A 26 1357A 32 1343A 2? 1361A 33 1342B 28 1360A 34 1341B 29 1343A 35 1343B 30 1341A a - Composition values in parentheses are estimatedo b - Type of working prior to heat t reatment: H. W. - hot reduced C. R. - cold reduced c - Extrapolated 100,000ohour rupture strength. Values in parentheaes are derived from data. of less than 500 hours duration. d - Small scale laboratory heat 97

a — - - -- I G r ": —----------;E - L.O @##w — - ooaaaa~