THE U N I VERSITY OF M I C H I GAN COLLEGE OF ENGINEERING Department of Materials and Metallurgical Engineering Progress Report TIME-DEPENDENT EDGE-NOTCH SENSITIVITY OF OXIDE AND GAMMIA PRIME DISPERSION STRENGTHENED SHEET MATERIALS AT 10000 TO 1800~F (5380-982~C) David J. Wilson Alberto Ferrari ORA Project 04368.0 supported by: NATIONAL AERONAUTICS AND SPACE ADMINISTRATION GRANT NO. NGL-23-005-005 WASHINGTON, D.C. administered through: OFFICE OF RESEARCH ADMINISTRATION ANN ARBOR September 1972

TABLE OF CONTENTS Page LIST OF TABLES iv LIST OF FIGURES INTRODUCTION1 EXPERIMENTAL DETAILS MECHANICAL CHARACTERISTICS5 Rupture Strengths Creep Resistance 5 Fracture Characteristics 6 Stress Relaxation and Creep Characteristics7 Negative Creep 8 MICROSTRUCTURAL FEATURES TD-Ni,Cr 9 As-received material ("original condition") g Tested specimens g Modified Waspaloy 10 As-heat treated 10 Tested specimens 11 DISCUSSION 12 TD-Ni,Cr 12 Modified Waspaloy15 Correlation of the Time-Dependent Edge-Notch Sensitivity with the Dislocation Mechanism 17 SUMMARY OF RESULTS 19 REFERENCES 21 iii

LIST OF TABLES Table 1. Smooth and Notched (Kt > 20) Specimen Tensile and Creep Rupture Properties at 1000~ to 1800~F (5538 -982~C) for O.015-Inch (0.38 mm) Thick TD-Ni,Cr Sheet 2. Smooth and Notched (Kt > 20) Specimen Tensile and Creep Rupture Properties at 1000~ to 1400~F (538~-760~C) for 0.025-Inch (0.64 mm) Thick Modified Waspaloy Sheet Solution Treated 1/2 Hour at 1975~F (10800C) and Aged 3. Comparison of Experimental and Theoretical Values of the Orowan Stress for TD-Ni,Cr iv

LIST OF FIGURES Figure 1. Larson-Miller parameter curves showing the time-dependent dependence of the rupture strengths of Waspaloy heat treated 1/2 hr at 1975~F (1080~C) and aged 16 hr at 1400~F (760~C), ref. 1. 2. Stress versus rupture time data at temperatures from 1000~ to 1800~F (538~-9820C) obtained from smooth and notched specimens of 0.015-in. (0.38 mm) thick stress-relieved, TD-Ni,Cr sheet. 3. Larson-Miller parameter curves showing the time-temperature dependence of the rupture strengths of 0.025-in. (0.64 mm) thick Modified Waspaloy sheet solution treated 1/2 hr at 1975~F (1080~C) and aged. 4. Stress versus minimum creep rate behavior at 10000 to 1800~F (538~982~C) for 0.015-in. (0.38 mm) thick, stress relieved, TD-Ni,Cr sheet. 5. Optical photomicrograph showing subsidiary intergranular cracks in the gauge section of a smooth specimen of 0.015-in. (0.38 mm) thick TD-Ni Cr sheet creep-rupture tested at 12 ksi (83 MN/m2) at 1800~F (982~C). 6. Iso-creep strain curves of life fraction versus stress at temperatures from 1000~ to 1800~F (538~-982~C) for O.015-in. (0.38 mm) thick, stress relieved, TD-Ni,Cr sheet. 7. Iso-creep strain curves of life fraction versus stress at temperatures from 1000~ to 1400~F (538~-982~C) for 0.025-in. (0.64 mm) thick, Modified Waspaloy sheet heat treated 1/2 hr at 1975~F (1080~C) plus 16 hr at 1400~F (760~C). 8. Deformation-time curve for a smooth specimen of TD-Ni,Cr sheet creeprupture tested at 10 ksi (69 MN/m2) at 1800F (982~C). 9. Optical and transmission photomicrographs of as-received 0.015-in. (0.28 mm) stress relieved, TD-Ni,Cr sheet. 10. Transmission electron micrographs of thin foils taken from the gauge sections of tested smooth specimens of TD-Ni,Cr sheet. 11. The variation of the dislocation mechanisms observed as a function of test stress and temperature for TD-NiCr. v

LIST OF FIGURES (Concluded) Figure 12. Optical photomicrographs of 0.025-in. (0.64 mm) thick Modified Waspaloy sheet heat treated 1/2 hr at 1975~F (1080~C) plus 16 hr at 1400~F (760~C). 13. Transmission electron micrographs of thin foils from the gauge sections of smooth specimens and the shoulder of a notched specimen of Modified Waspaloy solution treated 1/2 hr at 1975~F (1080~C), aged and creeprupture tested. 14,15. The variation of the dislocation mechanism as a function of test stress, compensated to 1000~F (5380C), and the particle size for Modified Waspaloy heat treated 1/2 hr at 1975~F (1080~C), aged and tested at 1000~ to 1400~F (538~-760~C). 16. The dislocation mechanisms, as a function of test stress and particle size, that can occur for a 7' strengthened alloy. 17. The effect of aging exposures at 1325~ and 1400~F (718~, 760~C) on the Diamond Pyramid Hardness of Modified Waspaloy solution treated 1/2 hr at 19750F (10800C). vi

INTRODUCTION Notch sensitivity of heat resistant sheet materials is being studied at The University of Michigan, Ann Arbor, Michigan. The research is sponsored by The National Aeronautics and Space Administration, Washington, D.C. Severe time-dependent edge-notch sensitivity (Figure 1) of Waspaloy and Inconel 718 has been shown to occur when notched specimens were loaded below the approximate 0.2 percent smooth specimen offset yield strength and when test data from smooth specimens indicated that small amounts of creep strain consumed large fractions of creep-rupture life [1,2]. The susceptibility to notch sensitivity has also been correlated with dislocation motion mechanisms [2,3]. Coherent precipitates smaller that a "critical size" were sheared by dislocations. This gave rise to localized deformation and time-dependent notch sensitive behavior. Larger particles were by-passed by dislocations and the deformation was homogeneous. Under these conditions, no time-dependent notch sensitivity was observed. Research is continuing to determine the generality, or scope, of the concepts that have been developed. Presently reported are results for two alloys: (1) TD-Ni,Cr, which is strengthened by an incoherent dispersion of ThO2 particles. This material was selected for study, because, dislocations would have to by-pass the oxide particles during creep. Therefore, the alloy was not expected to be susceptible to timedependent notch sensitivity. (2) An experimental nickel-base alloy, "Modified Waspaloy" having a relatively low volume fraction (0.095) of coherently precipitated y', Ni3(Al,Ti). This material was studied in order to permit extension of the results for Waspaloy (y' volume fraction 0.23) to alloys with less precipitate. The TD-Ni,Cr was evaluated in the stress-relieved condition while the Modified Waspaloy was studied after heat treatments that produced a range of y' particle sizes. Tensile and creep-rupture tests were carried out at 1000~1800~F (538~-982~C) for TD-Ni,Cr and at 1000~-1400~F (538~-760~C) for Modified

Waspaloy. The microstructural features, particularly the dislocation structures in the tested specimens, were evaluated. 2

EXPERIMENTAL DETAILS The commercially produced TD-Ni,Cr used in the investigation had the following reported composition (weight percent): Ni Cr ThC2 C Si 78.72 19.24 2.01 0.03 0.003 The material was received as 0.015 in. (0.38 mm) thick, stress relieved, sheet. The "Modified Waspaloy" was prepared as a vacuum induction melted ingot with aim composition similar to Waspaloy [1] except for lower amounts of titanium and aluminum (the Ti/Al ratios were essentially the same). The compositions were as follows (weight percent): Modified Modified Waspaloy Waspaloy Waspaloy Waspaloy Ni Bal Bal Zr 0.031 0.03 Cr 19.40 19.53 B NA* 0.005 Co 14.70 13.52 Fe NA 0.55 Mo 4.10 4.16 S NA 0.007 Ti 1.78 2.95 Mn NA <0.01 Al 0.67 1.35 Si NA 0.05 C 0.053 o.o6 Cu NA 0.03 *NA = Not Analyzed The ingot was radiographed to assure soundness, preheated at 2000~F (1093~C) and hot rolled in small reductions to a thickness of about 0.032 in. (0.81 mm) using reheats between passes. These sheets were pickled and cold reduced (20-25 percent) in several passes to a final thickness of approximately 0.025 in. (0.64 mm). Specimen blanks were individually solution treated 1/2 hr at 1975~F (1080~C) in an argon atmosphere and air cooled. To prevent warping, the blanks were aged in batches of 10 or 12 while clamped in a fixture. The aging treatments were: (a) 3 hr at 1325~F (718~C), (b) 1 hr at 1400~F (760~C), and (c) 16 hr at 1400~F (760~C). 3

For both alloys, smooth and edge-notched (Kt>20) specimens were machined in the longitudinal (rolling) direction. The specimen dimensions and test procedures were the same as previously reported [1]. For optical metallography, samples of Modified Waspaloy and TD-Ni,Cr were etched electrolytically in "G" etch [4] and a solution of 10 percent oxalic acid in water, respectively. The volume fraction of y' in Modified Waspaloy was determined by electrollytic extraction [5] using duplicate samples solution treated 1/2 hr at 1975~F (1080~C) and exposed 1000 hr at 1400~F (7600C). Techniques for the preparation of thin foils for study by transmission electron microscopy were reported previously [5]. 4

MECHANICAL CHARACTERISTICS The results of tensile and creep-rupture tests are included as Table 1 for TD-Ni,Cr and Table 2 for Modified Waspaloy. The principal features, particularly those related to notch sensitive behavior, are presented in the following sections. Rupture Strengths For TD-Ni,Cr, the notched specimen rupture curve at 1000~F (538~C) was somewhat below that for smooth specimens (Figure 2). At higher temperatures, the rupture strengths for notched specimens were similar or higher than for smooth specimens. At all test temperatures, the notched to smooth rupture strength ratios (N/S) were greater than determined by tensile tests (Table 1), i.e., the alloy did not exhibit time-dependent notch sensitivity. The heat treatment variations of Modified Waspaloy resulted in a range of smooth specimen strengths at the lower test temperatures. The strength increased as the aging treatment was changed from 3 hr at 1325~F (7180C) to 1 hr at 1400~F (760~C) to 16 hr at 1400~F (76o0~C). At the higher temperature and longer test times, the strengths were similar. (This resulted, at least in part, from y' growth during the test exposures.) For all heat treated materials the N/S rupture strength ratios decreased with increasing time and/or temperature (Figure 3), i.e., time-dependent notch sensitivity was not observed. Creep Resistance The minimum creep rates determined from the smooth specimen tests are included in Tables 1 and 2. The most striking feature was that the 5

stress-minimum creep rate curve (Figure 4) for TD-Ni,Cr at 1000~F (538~C) was considerably steeper than the curves at temperatures from 12000 to 1800~F (649~-982~C). This indicates that different mechanisms control creep in these two ranges of test conditions. For Modified Waspaloy, the creep rates and, hence, the creep resistance varied with heat treatment and test conditions in similar manner to the rupture strengths. Fracture Characteristics The fracture characteristics of Modified Waspaloy were similar to those previously reported for Waspaloy [1]. Both smooth and notched specimens failed by initiation and relatively slow growth of intergranular cracks followed by transgranular fracture. The latter fracture occurred when the increase in stress on the load bearing area, due to growth of the intergranular crack, exceeded that necessary to cause rapid shear. The intergranular and transgranular parts of the fracture were readily distinguishable by visual examination. (The lengths of the intergranular cracks, expressed as a percentage of the specimen width, are included in Table 2.) Optical metallography showed that the fractures of tensile and creeprupture tested specimens of TD-Ni,Cr were also both transgranular and intergranular. Intergranular cracks were distributed across the fractures. In notched specimens, the fracture adjacent to one, or both notches, was intergranular. The total intergranular crack length increased with decreasing stress and increasing temperature. Extensive intergranular cracking, perpendicular to the loading direction, occurred adjacent to the fractures and throughout the gauge sections of the smooth specimens (Figure 5). Intergranular cracks were also found in several test specimens discontinued before rupture. These results indicate that failure occurred by intergranular crack initiation and growth, the linking of microcracks, and finally transgranular fracture. 6

MECHANICAL CHARACTERISTICS The results of tensile and creep-rupture tests are included as Table 1 for TD-NiCr and Table 2 for Modified Waspaloy. The principal features, particularly those related to notch sensitive behavior, are presented in the following sections. Rupture Strengths For TD-Ni,Cr, the notched specimen rupture curve at 1000~F (558~C) was somewhat below that for smooth specimens (Figure 2). At higher temperatures, the rupture strengths for notched specimens were similar or higher than for smooth specimens. At all test temperatures, the notched to smooth rupture strength ratios (N/S) were greater than determined by tensile tests (Table 1), i.e., the alloy did not exhibit time-dependent notch sensitivity. The heat treatment variations of Modified Waspaloy resulted in a range of smooth specimen strengths at the lower test temperatures. The strength increased as the aging treatment was changed from 3 hr at 1325~F (718"C) to 1 hr at 1400~F (76o0C) to 16 hr at 1400~F (760~C). At the higher temperature and longer test times, the strengths were similar. (This resulted, at least in part, from y' growth during the test exposures.) For all heat treated materials the N/S rupture strength ratios decreased with increasing time and/or temperature (Figure 3), i.e., time-dependent notch sensitivity was not observed. Creep Resistance The minimum creep rates determined from the smooth specimen tests are included in Tables 1 and 2. The most striking feature was that the 5

stress-minimum creep rate curve (Figure 4) for TD-Ni,Cr at 1000~F (5580C) was considerably steeper than the curves at temperatures from 12000 to 1800~F (649~-982~C). This indicates that different mechanisms control creep in these two ranges of test conditions. For Modified Waspaloy, the creep rates and, hence, the creep resistance varied with heat treatment and test conditions in similar manner to the rupture strengths. Fracture Characteristics The fracture characteristics of Modified Waspaloy were similar to those previously reported for Waspaloy [1]. Both smooth and notched specimens failed by initiation and relatively slow growth of intergranular cracks followed by transgranular fracture. The latter fracture occurred when the increase in stress on the load bearing area, due to growth of the intergranular crack, exceeded that necessary to cause rapid shear. The intergranular and transgranular parts of the fracture were readily distinguishable by visual examination. (The lengths of the intergranular cracks, expressed as a percentage of the specimen width, are included in Table 2.) Optical metallography showed that the fractures of tensile and creeprupture tested specimens of TD-Ni,Cr were also both transgranular and intergranular. Intergranular cracks were distributed across the fractures. In notched specimens, the fracture adjacent to one, or both notches, was intergranular. The total intergranular crack length increased with decreasing stress and increasing temperature. Extensive intergranular cracking, perpendicular to the loading direction, occurred adjacent to the fractures and throughout the gauge sections of the smooth specimens (Figure 5). Intergranular cracks were also found in several test specimens discontinued before rupture. These results indicate that failure occurred by intergranular crack initiation and growth, the linking of microcracks, and finally transgranular fracture. 6

Stress Relaxation and Creep Characteristics Previous studies have shown that notch sensitive behavior is dependent on the manner which stress concentrations, introduced by the edge-notches, are relaxed [1,3]. When notched specimens are tested at stresses above the approximate 0.2 percent offset yield strength, yielding on loading reduces the stresses across the specimen at the base of the notch to the approximate nominal stress. Under these conditions, time-dependent notch sensitivity has not been observed. When notched specimens are loaded below the yield stress, the stress concentrations can only be fully relaxed by creep (from the approximate yield strength to the nominal stress). A material exhibits time-dependent notch sensitivity when the creep deformation necessary to relax stress concentrations causes excessive damage resulting in premature crack initiation. Experimentally, time-dependent notch sensitivity has been shown to occur when sharp edged-notched sheet specimens were loaded below the approximate 0.2 percent offset yield strength and when tests on smooth specimens showed that 0.1 and 0.2 percent creep consumed large fractions of rupture life [1,3]. Creep data for TD-Ni,Cr and Modified Waspaloy were evaluated to determine whether this same correlation applied: (1) Analysis of the data for TD-Ni,Cr showed that within the range of test data at 1000~ and 1200~F (538~,649~C) the life fractions for 0.1 and 0.2 percent creep were at low levels (Figure 6), Thus, the iso-creep strain characteristics were consistent with the fact that no time-dependent notch sensitivity occurred. At the higher temperatures and lower stresses, large life fractions were utilized for small creep strains. Despite this, no time-dependent notch sensitivity was observed. This apparent contradiction of the correlation was due to "negative creep" which is considered in the following section. (2) For Modified Waspaloy, the life fractions for small amounts of creep were generally at very low levels (Figure 7). Correspondingly, no time-dependent notch sensitivity was observed (Figure 3). (At 1300~F, somewhat larger life fractions were consumed for 0.1 and 0.2 percent creep due to "negative creep.") 7

Negative Creep When exposed under stress at elevated temperatures, many nickel-base superalloys shrink, i.e, exhibit "negative creep." Creep-rupture tests are usually conducted at test conditions for which the amount of positive creep deformation is larger than the negative creep. As a consequence, the measured deformation-time characteristics appear normal. In the present investigation, for both materials tested at the higher temperatures and lower stresses, the amount of positive deformation was relatively small compared with the negative creep. As a result, negative minimum creep rates were observed (Tables 1, 2, and Figure 8) and, large life fractions were utilized for small creep strains (Figure 6). There is evidence that indicates that negative creep is probably due to an ordering of the Ni-Cr lattice [6]. Being a volumetric charge, it should not act to relax stress concentrations or influence the notch sensitive behavior. Results for TD-Ni,Cr tests, for which the creep rate was expected to be low, showed that shrinkages of at least 0.2 percent can occur. It is therefore probable that the apparent contradiction between creep characteristics and notch sensitive behavior was due to the influence of negative creep on the observed deformation-time curves. 8

MICROSTRUCTURAL FEATURES TD-Ni,Cr and Modified Waspaloy were studied in the as-heat treated conditions in order to characterize the microstructural features present. Subsequently, tensile and creep-rupture tested specimens were examined primarily to establish the dislocation motion mechanisms. Samples were taken somewhat removed from the fractures of smooth specimens and from the shoulders of notched specimens. The latter specimens provided samples exposed at relatively low stresses (0.7 times the applied stress). TD-Ni, Cr As-Received Material ("Original Condition") The average grain diameters were 0. 16 mm and 0. 08 mm in the longitudinal and transverse directions, respectively (Figure 9a). The ThOC particles ranged in size from about 100 to 1000 X with a mean of approximately 285 A (Figure 9b). A negligible number of randomly distributed dislocations were present. Many annealing twins (about 0.24I wide) were observed (Figure 9b). Tested Specimens In specimens tensile tested at 1000~, 14000, and 1800~F, (538~, 760", and 982~C) many elongated dislocation loops were observed pinned to ThO2 particles (Figure lOa). These are prismatic loops which do not lie in the original slip lane and indicate that the dislocations by-passed the oxide particles by cross slip [7]. The dislocation density decreased with increasing test temperature (decreasing elongation at fracture). The microstructures (Figure lOb) of specimens creep-rupture tested at 1000~F (538~C) and at 1400~F (760~C) at 28.6 ksi (197 MN/m2) also indicated that cross-slip had occurred. 9

In specimens creep-rupture tested at the higher temperatures and lower stresses, prismatic dislocation loops were not observed (Figure lOc). In these cases the dislocations by-passed the ThO2 particles by climb. For a number of tests at high stresses at 1200~, 1400~, and 16000F (649~, 760~ and 871~C) only a limited number of prismatic dislocation loops were observed. It was concluded that dislocations by-passed the oxide particles by both climb and cross slip. The minimum creep rate characteristics previously reported (Figure 4), indicated that climb was the predominant mechanism. In all tensile and creep-rupture tests, the deformation was homogeneous. The dislocation mechanisms observed, as a function of test stress and temperature, are summarized in Figure 11. Modified Waspaloy As-Heat Treated Limited examination of the as-heat treated materials indicated that the microstructural features, other than the 7' morphology, were similar to those previously reported from similarly heat treated Waspaloy [3]. The average grain size was 0. 036 mm (Figure 12). The carbides present were identified, by selected area electron diffraction, to be Ti(C,N), and M23C6. The former carbide was a randomly distributed globular precipitate. The grain boundaries were partially filled with M23C6 carbide. A number of twin boundaries, particularly twin ends, also contained M23C6. As was the case for Waspaloy, v' was present as coherently precipitated spherical particles. Due to lower aluminum and titanium additions, the 7' volume fraction was less than for Waspaloy (0.095 compared to 0.235). The aging treatments, 3 hr at 1325~F (718~C), 1 hr at 1400~F (760~C), and 16 hr at 1400~F (760~C) resulted in 7' particles approximately 40, 75, and 140 X in diameter, respectively. No zones depleted of 7' occurred 10

adjacent to grain boundaries. Tested Specimens In tensile and creep-rupture tests at 1000~ and 1100~F (538~ and 593~C), limited y' growth occurred during the test exposures. Di. locations either sheared or cross slipped around the relatively small particles. In smooth specimens tested at the higher stresses, prismatic dislocation loops were observed indicating that dislocations by-passed the y' by cross-slip (Figure 13a). Under these conditions, the deformation was homogeneous. In samples (including these from the shoulders of notched specimens) exposed at lower stresses, superdislocations were observed (Figure 13b). A few dislocations were also present as extended stacking fault ribbons. These observations demonstrate that dislocations sheared the 7' particles. When this mechanism occurred, the deformation was localized in slip bands. At intermediate stresses, both dislocation mechanisms contributed to the deformation. During the higher temperature creep-rupture exposures, considerable y' growth occurred. In a number of test specimens for which the particles were large, dislocations were observed bowing between particles leaving pinched off concentric dislocation loops (Figure 13c). In other specimens, it was evident that the dislocations by-passed the y' by both cross slip and by looping. In all cases the deformation was homogeneous. The variations in dislocation mechanism observed with changes in test stress and y' particle size are presented in Figures 14 and 15. 11

DISCUSS ION TD-Ni,Cr Theories have been proposed to explain the manner in which edge and screw dislocations can by-pass particles by cross-slip [8]. Experimentally, crossslip has been shown to occur for a number of dispersion hardened systems F9], the majority of which have high stacking fault energies. Cross-slip has been shown to occur for TD-Ni [10]. The stacking fault energy of nickel is lowered by the addition of chromium. (30 weight percent chromium decreases the stacking 2 fault energy from about 225 to 70 erg/cm -ref. 11.) Consequently, cross slip could be expected to be relatively difficult in TD-Ni,Cr. Nevertheless, it does occur in low temperature, high stress, tensile and creep-rupture tests. At high temperatures and low stresses, dislocations by-pass ThO particles by climb. At each temperature there is a critical resolved shear stress, CRSS, above which the dislocations do not climb over the dispersed particles, but by-pass them by cross-slip (or looping). This concept was first considered theoretically by Orowan [12]. The following formulation for the "Orowan Stress" (the increase in CRSS due to the presence of non-deforming particles) was subsequently reported by Ashby [13]: Ab d, (1) AT = B 2(-d) 4b where p is the matrix shear modulus, b the Burger vector, y the mean planar center-to-center particle separation, d the mean planar particle size (0.82 x mean particle size, D) and B a statistical factor (about 0.85) related to the occurrence of a distribution of interparticle spacings. A is a constant equal to 1 for an edge dislocation and (l-v) for a screw dislocation, where v is Poisson's ratio. 12

Several relationships have been proposed to permit calculation of \ values from the precipitate volume fraction V and the mean particle size [14]. These are based on assumed arrangements of the particles in space. Simple cubic: k = d (r/4V)1/2 (2) Face centered cubic: \ = d (l/V) //2 (3) 1/2 Random: x = d (/16V )1/ (4) The results of the microstructural study of TD-NiCr established the "Orowan Stress" as a band of values (Figure 11). The minimum creep rate characteristics (Figure 4) indicated that the upper curve probably best represents the stresses where the creep controlling mechanism changes from climb to cross-slip. These experimentally determined values were compared with those calculated using equation (1). The constant A was assumed to be 1.2 (the mean value for edge and screw dislocations). Values for the shear modu7 lus, derived from tensile data were: 0.8, 0.7, and 0.5 ~ 10 psi (5.5, 4.8, and 3.4 * 10 MN/m ) at 1000~, 1400~ and 1800~F (558~, 760~, and 982~C), respectively. The Burger vector was 2.5 A and the ThO volume fraction 0.0174. Poisson's ratio was assumed to be 0.30. The CRSS values calculated were multiplied by the Taylor factor [15] so that they could be compared directly with the "applied stress" values determined experimentally. The factor of 3. 06 which was used can be considered to be an upper bound for this "stress conversion." The results (Table 3) show that the theoretical values vary by a factor of about 2, depending on the equation, (2)-(4), used to calculate the center-to-center particle separation. The experimentally determined stress values are within the range of those determined theoretically. The ThO2 dispersion should almost certainly be described as random, 13

Theoretical stresses based on this assumptionwere in error by a factor of about 2. Considering the approximations involved, this agreement is surprisingly good. The Th02 particles contribute to the creep resistance of TD-Ni by impeding dislocation motion. It has been proposed, however, that the good creep characteristics of this alloy are primarily attributable to the fine grain size and cellular substructure formed during fabrication and stabilized by the oxide particles [10]. The creep resistance at 1000~F for the TD-Ni,Cr studied was similar to that reported for TD-Ni [10]. However, the creep resistance of TD-Ni,Cr decreased much more rapidly with increasing temperature than was the case for TD-Ni. Correspondingly, the microstructures of the as-received TD-Ni,Cr did not reveal any substructure (consistent with the relatively low stacking fault energy of this material). In addition, the grain size of TD-Ni,Cr was about 10 times that of TD-Ni (80-l60it compared to 1-2i). Thus in contrast to TD-Ni, the ThO2 particles in TD-Ni,Cr increased the creep strength only by providing resistance to dislocation motion. Minimum creep rate characteristics are often represented by the familiar expression: n = A(t)an, (5) where n determines the overall stress dependence of the creep process. For TD-Ni,Cr, n was 10 at 1000~F (538~C) and about 30 at higher test temperatures. These values are significantly greater than predicted by theories for cross slip and climb (about 5). Anomalously high values for n, reported by other authors for dispersion hardened materials, have been explained in terms of "internal stress" and/or grain orientation effects [10,11,16,17]. "Negative creep" that occurs for TD-Ni,Cr reduced the minimum creep rates, particularly at the lower test stresses, and thereby must have acted to increase the measured n values. 14

Modified Waspaloy For Modified Waspaloy (Vf = 0.095), ' particles were by-passed by crossslip and by looping. For TD-Ni,Cr (Vf = 0.017) only cross-slip occurred, whereas, for Waspaloy (V = 0.23) only the looping mechanism was observed [3]. These results suggest that increasing the volume fractio: of precipitate tends to change the by-pass mechanism from cross-slip to looping. Other variables such as the stacking fault energy and test temperature, can also be expected to influence this transition. Both cross-slip and looping are dependent on the ability of dislocations to bow between precipitate particles. Consequently, they occur at stresses greater than the "Orowan stress." For Modified Waspaloy, both the average 7' size and the test temperature varied. To simplify evaluation of the experimental results, the data were compensated for temperature variation using the following relationship derived from equation (1): t1000 ~F 1l000F = T T (6) The dislocation mechanisms observed are presented in Figure 14 as a function of stress compensated to 1000~F (538~C) and the mean y' particle size. Orowan stresses at 1000~F were calculated using equation (1) in a similar manner described for TD-Ni,Cr. In this case, the following values were used: = 1.02 x 107 psi (7.03 x 10 MN/m ), b = 2.52 A and, Vf = 0.095. Periodic or "ordered" precipitate distribution are observed in many superalloys. These structures occur when the 7' particles are large and/or the 7'/y lattice mismatch is large [18]. Based on the composition of Modified Waspaloy and the observed spherical y' morphology, it can be concluded that the y'/7 mismatch was relatively small (similar to Waspaloy which has a mismatch of about 0.35). Also, the particles in the as-heat treated materials were relatively small. Consequently, the 7' precipitate particles were probably "randomly" 15

distributed. Theoretical values for the Orowan stress, using this assumption, differ by a factor of about 2 from the experimental values (Figure 14). It should be considered coincidental that the experimental values agree well with the theoretical stresses based on face centered cubic and simple cubic distributions. Because the 7' is a coherent precipitate, it can be sheared by dislocations. Gleiter and Hornbogen [18] reported that the increase in CRSS due to ordered precipitate particles sheared by dislocations, AT h, is given by the following equation: sh 0.28 73/2 V1/3 (D/2)1/2 -1/2 b-2 (7) sh 0 A f where 7A is the antiphase boundary energy. AT values at 1000~F (535~C), 2 sh were calculated using a value of 260 erg/cm for 7A [19]. To permit comparison with the microstructural observations (Figure 15), the test stresses were compensated to 1000~F (538~C) using the following relationship from equation (7): 1/2 MT %1000OF T 1000/J F (8) All of the tests stresses (Figure 15) were greater then the calculated values (ATh x Taylor factor, 3.06). The specimens tested at stresses slightly sh greater than the calculated values had very low dislocation densities. (Below AT h the only mechanism that can occur is climb, which due to the relatively sh' low test temperatures, should contribute very little to the deformation.) Therefore, it can be concluded that the theoretical values for AT are in sh very good agreement with the experimental results. 1/2 The CRSS for shearing increases with D / (equation (7)) while for cross slip or looping it decreases as D increases (equation (1)). Therefore, for each 7' strengthened alloy (constant Vf), there is a critical particle size

at which the dislocation mechanism changes from shearing to by-passing the precipitate particles. This is shown for Modified Waspaloy in Figure 16. Correlation of the Time-Dependent Edge-Notch Sensitivity with the Dislocation Mechanism In previous studies, a correlation was established between the dislocation motion mechanism and the time-dependent notch sensitive behavior [2,3]. Precipitate particles, 7' in Waspaloy and y' and y"-Ni3Cb (bct) in Inconel 718, smaller than the critical size were sheared by dislocations. Under these conditions, the deformation was localized and time-dependent notch sensitivity occurred. Particles larger than the critical size were by-passed by dislocations, the deformation was homogeneous and no time-dependent notch sensitivity was observed. The results of the presently reported investigation are consistent with the above correlation. For TD-Ni,Cr, the dislocations by-passed the ThO2 particles for all test conditions. Correspondingly, time-dependent notch sensitivity did not occur. In a number of notched specimen tests for Modified Waspaloy at the lower test temperatures, dislocations sheared the y' particles (Figure 13b). However, no time-dependent notch sensitivity was observed. This occurred because the test stresses were above the yield stresses. For notched specimens tested below the yield strength, either the *est times were extremely long or the test temperatures were relatively high. As a result, considerable y' growth occurred during the test exposures. This caused the dislocation mechanism to change from shearing to by-passing which reduced the susceptibility to time-dependent notch sensitivity so that none was observed. Thus time-dependent notch sensitivity did not occur for Modified Waspaloy due to the yield strength and 7' growth rate characteristics. The yield strength, a, of a dispersion hardened alloy can be expressed by the following relationship: by the following relationship: 17

u = a + 3. o6 A - (9) y m p where a = yield stress of the matrix, AT is the increase in CRSS lue to the m p presence of the particles (AT or AT s). Consideration of the yield strength o sh in this manner (Figure 16) permits a regime of test stress and particle size to be defined for which time-dependent notch sensitivity can be expected. Rapid growth of 7' in Modified Waspaloy prevented time-dependent notch sensitivity from occurring at the low test stresses. The 7' ripening rate decreases considerably with increasing volume fraction [20]. For alloys with higher 7' volume fraction than Modified Waspaloy, particle growth cannot be relied upon to prevent notch sensitive behavior from occurring. Heat treatments should be selected to produce 7' particles larger than the critical size. Under these circumstances, it is actually desirable to limit the growth rate (by compositional control) so as to reduce strength degeneration that will accompany 7' growth. Hardness tests or yield strengths reflect the variations in CRSS with y' size and can, therefore, be used in the selection of heat treatments. For Waspaloy [3] and Inconel 718 [2] room temperature hardnesses were correlated with the notch sensitive behavior. For Modified Waspaloy (Figure 17), hardness tests correctly indicated that 7' particles in the as-heat treated materials to be smaller than the critical size (maximum hardness). It should be recognized that the critical size varies somewhat with temperature. This arises because of changes in the shear modulus, t, with temperature. From the equations presented ((1), (4), and (7)) it is evident that as t decreases with temperature the critical size decreases with increasing temperature. Thus if a heat treatment is selected, by room temperature hardness tests, so that the 7' particles are larger than the critical size they will also be larger than the critical size at elevated temperatures.

SUMMARY OF RESULTS A research program was carried out to determine whether an oxide dispersion hardened alloy, TD-Ni,Cr, and a low volume fraction y' strengthened nickel-base alloy, Modified Waspaloy, were susceptible to time-dependent edgenotch sensitivity. The results were evaluated in terms of the mechanical characteristics of the alloys and the dislocation motion mechanisms operative. Time-dependent notch sensitivity was not observed for 0. 015-in. (J. 38 mm) thick, stress relieved, TD-Ni,Cr sheet at temperatures from 10000 to 1800~F (538~-982~C). Dislocations by-passed ThO particles by cross-slip at high test stresses and low temperatures and by climb at low test stresses and high temperatures. The stresses at which the mechanism changed from climb to cross-slip (Orowan stresses) were in reasonable agreement with theoretical predictions. Time-dependent notch sensitivity was not evident from tests of 0. 025-in. (0.64 mm) thick Modified Waspaloy sheet at temperatures from 1000~ to 1400~F (5380-7600C). In tests at relatively low stresses, dislocations sheared 7' particles smaller than a critical size (about 180A); larger particles were bypassed by cross-slip or looping. At stresses higher than the Orowan stress, small y' particles were by-passed by cross-slip. Results indicated that the dislocation mechanisms observed can be usefully evaluated in terms of theoretical concepts. As far as could be determined, the results of the investigation were consistent with the following important concepts developed for Waspaloy [1,35] and Inconel 718 [2]: (a) Time-dependent edge-notch sensitivity occurs when notched specimens are loaded below the approximate 0.2 percent smooth speciment offset yield strength and when data from smooth specimens indicate that small amounts of creep consume large rupture life fractions. (b) When precipitate particles are sheared by dislocations, the deformation is localized and time-dependent notch sensitivity occurs. When dislocations by-pass precipitate particles the deformation is homogeneous. Under these conditions, 19

no time-dependent notch sensitivity has been observed. To date, alloys have been studied which have coherent and noncoherent precipitate volume fractions ranging from 0.017 to 0.23. The applicability of the above correlations for these alloys suggests that they can be used to characterize the notch sensitive behavior of an even wider range of superalloy sheet materials. It is apparent that hardness tests can be used to select heat treatments that produce 7' particles larger than the critical size and, therefore, avoid time-dependent notch sensitivity. 20

REFERENCES 1. Wilson, D. J. "Sensitivity of the Creep-Rupture properties of Waspaloy Sheet to Sharp-Edged Notches in the Temperature Range 1000~-1400~F, Trans. ASME, J. Basic Eng. 94 (1972), No. 1, pp. 13-21. 2. Wilson, D. J. "Relationship of Mechanical Characteristics and Microstructural Features to the Time-Dependent Edge Notch Sensitivity of Inconel 718 Sheet" (accepted for publication to ASME). 3. Wilson, D. J. "The Dependence of the Notch Sensitivity of Waspaloy at 1000~-1400~F on the Gamma Prime Phase," Trans. ASME, J. Basic Eng. (In print). 4. Bigelow, W. C., J. A. Amy, and L. O. Brockway. "Electron Microscope Identification of the Gamma Prime Phase of Nickel-Base Alloys," Proc. ASTM 56 (1956), pp. 945-953. 5. Kriege, O. H., and J. M. Barris. "The Chemical Partitioning of Elements in Gamma Prime Separated from Precipitation Hardened Nickel-Base Alloys," Trans. ASM 62 (1969), pp. 195-200. 6. Nordheim, R., and N J. Grant. "Resistivity Anomalies in the NickelChromium System as Evidence of Ordering Reactions," JISI 21 (1954), pp. 440-444. 7. Humphreys, F. J., and J. W. Martin. "The Deformation of Single Crystals of Copper and Copper-Zinc Alloys Containing Alumina Particles (II) Microstructural and Dislocation-Particle Interactions," Proc. Roy. Soc. (London) A318 (1970), pp. 73-92. 8. Hirsch, P. B., and F. J. Humphreys. "Physics of Strength and Plasticity." Mass. Inst. Tech. Press. (1969), pp. 189-216. 9. Duesbery, M. S., and P. B. Hirsch. "Fundamental Aspects of Dislocation Theory." N.B.S. Special Publication 317 (1969), pp. 1115-1134. 10. Wilcox, B. A., and A. H, Claurer. "Creep of Thoriated Nickel Above and Below 0. 5 Tm," Trans. AIME 2X6 (1966), pp. 570-580. 11. Wilcox, B. A., A. M. Claurer, and W. S. McCain. "Creep and Creep Fracture of a Ni-20Cr-2ThO2 Alloy," Trans. AIME 29 (1967), pp. 1791-1795. 12. Orowan, E. "Symposium on Internal Stresses in Metals and Alloys," JISI (1948), pp. 451-465. 21

REFERENCES (Concluded) 13, Ashby, M. F. "Physics of Strength and Plasticity," Mass. Inst. Tech. Press (1969), pp. 113-131. 14. Ashby, M. F. "The Hardening of Metals by Non-Deforming Particles," Z. Metallkunde 55 (1964), pp. 5-17. 15. Kocks, U. F. "The Relation Between Polycrystal Deformation and Single Crystal Deformation," Met. Trans. 1, No. 5 (1970), pp. 1121-1159. 16. Li, J.C.M. "Physics of Strength and Plasticity," Mass. Inst. Tech. Press (1969), pp. 245-253. 17. Claurer, A. H., and B. A. Wilcox. "Steady-State Creep of Dispersion Strengthened Nickel," Met. Sci. J. 1 (1967), pp. 86-90. 18. Gleiter, H., and E. Hornbogen. "Precipitation Hardening by Coherent Particles," Matls. Sci. Eng. 2 (1967-68), pp. 28-5302. 19. Merrick, M. F. "The Structure of Nickel-Base Creep Resistant Alloys," Doctoral Thesis, University of Cambridge (1963). 20. Decker, R. F. "Strengthening Mechanisms in Nickel-Base Superalloys," presented at Steel Strengthening Mechanisms Symposium, Zurich, Switzerland (1964). 22

TABLE 1 SMOOTH AND NOTCHED (Kt > 20) SPECIMEN TENSILE AND CREEP RUPTURE PROPERTIES AT 10000 TO 1800'F (538~- 982~C) FOR 0.015-INCH (0.38 mm) THICK TD-Ni,Cr SHEET Smooth Specimens Notched Specimens Test Rupture Minimum Test Rupture N/S Temperature Time Creep Rate Temperature Time Strength Ratio ( ~F) ( C) (ksi) (MN/m2) (hr) (). (%) (%/hr) ( F) (~C) (ksi) (MN/m2) (hr) 1000 538 84.6 50 40 36 34 30 583 tensile 345 49.8 276 628.8 248 2570.0 234 2473.0 207 9311.7 1200 649 26 179 54.2 24 165 351.1 22 152 9009 1400 760 39.5 28.6 22 20 16 1600 871 18 16 14 13 12 272 tensilea 197 0.5 152 74.5 138 973.7 110 11000 124 35.9 110 132.1 97 3611.6 90 695.3 83 1944.0 9.8 4.5 4.4 4.4 4.2 3.5 2.7 2.4 In Progress 3.4 2.7 2.8 2.3 Discontinued 1.7 2.5 1.1 1.1 1.6 16 8 6 6 8 6 4 0.040 6 o0.0054 0.00008 0.035 0.0036 0.0010 0.0010 0.00029 1200 649 26 179 24 165 1400 760 38.0 262 20 138 16 110 1000 538 65.6 40 34 30 452 276 234 207 tensile 31.1 803.6 2022.0 59.3 708.0 tensile 839.8 9062.78.78.87.88 1.00 1.02 0.96 1.0 Discontinued 1.0 1.0 In Progress 1.08 1.08 1.16 6 6 5 4 <0.032 0.00014 - ve 1600 871 16 14 12 1800 982 23.3 12 10 110 151.6 97 501.7 83 11000 4 0.022 2 0.0016 2 0.000050 2 0.000004 2 - veb 161 83 69 tensile 62.0 2147.8 1800 982 21.6 149 12 83 10 69 8 55 a0.2% offset yield strengths: 1000~F (5380c) - 55.5 ksi (383 1400~F (760~C) - 34.1 ksi (235 1800 F (9820C) - 21.2 ksi (146 tensilea 47.7 196.2 5183.6 1.4 4 2 6 0.18 b - veb - veb MN/m2) MN/m2) MN/m2) -ve = negative minimum creep

TABLE 2 SMOOTH AND NOTCHED (Kt > 20) SPECIMEN TENSILE AND CREEP RUPTURE PROPERTIES AT 1000~ TO 1400~F (538~-760~C) FOR 0.025-INCH (0.64 mm) THICK MODIFIED WASPALOY SHEET SOLUTION TREATED AT 1/2 HOUR AT 1975~F (1080~C) AND AGED Smooth Specimens Notched Specimens Test Rupture Minimum Intergranular Test Rupture Intergranular N/S Temperature Stress Elong. R.A:. Stress Temperature ress Time Creep Rate Crack Length Temperature Time Crack Length Strength (~F).(~C) (ksi) (MN/my2 (hr) i _).(_ ) - (%o/hr) (W) ~.(F) (~C) (ksi) (MN/ (hr) Ratio Aged 3 Hours at 1325 F (718 C) 1000 538 115.8 798 1100 593 80 552 70 o '483 1200 649 70 483 60 414 tensile* 54.0 40 171.9 15.8 25 0.020 2309.6 10.2 14 0.00034 135.4 - 9.2 16 0.015 894.0 5.9 10 0.0012 0 1000 538 80.9 558 tensile 19 1100 593 60 414 6743 30 1200 649 55 379 198.2 23 45 310 5806 28 40 276 2637 1400 760 30 207 1155.8 0 0.70 In Progress >0.85 25 0.82 In Progress >0.87 Discontinued > 0.74 63 -1.0 Aged 1 Hour at 1400 F (760 C) 1000 538 116.5 804 tensile* 48.0 41 95 655 4355.6 18.3 16 0.000090 1100 593 90 620 277.2 14.7 17 0.0018 80 552 1124.5 9.7 15 0.00044 70 483 5843.0 6.2 10 0.000080 1200 649 80 552 59.0 8.4 15 0.024 0 1000 538. 80 1100 593 80 10 60 41 1200 649 60 40 1300 704 50 16 552 1.0 552 0.6 414 9745 414 1357.0 345 386.4 0.5 -6.70 1 - 0.70 Discontinued >0.88 48 -1.0 50 -1.0 Aged 16 Hours at 1400 F (760 C) 1000 538 122.0 841 110 758 100 689 1100 593 100 689 80 552 tensile* 50 41 310.5 25.8 25 0.015 4635.4 12.5 13 0.00040 '205.0 16.5 19 0.029 3542.5 4.7 12 0.00025 0 4 13 7 32 18 42 1000 538 91.7 632 tensile *85. 586 5893.9 70.483 4271 1100 593 80 552 2679.5 70 483 11184 1200 649 60 414 770.3 ph 55 379 7528.2 45 310 8354 1400 760 30 207 1170.8 0 0.75 19 0.85 Discontinued > 0.70 24 o.98 Discontinued > 0.96 -- >0.86 51 1.06 Discontinued > 0.88 47 1.15 1200 649 80 60 1300 704 50 1400 760 25 552 259.2 414 2688.0 345 351.4 172 1592.1 3.55 14 0.013 0.00029 6.2 13 - ve 15.5 28 0.00045 Notes: 'ph = failed at pin hole. -ve - negative minimum creep rate. *0.2% offs.et yield strengths at 1000OF: 3 hr at 1325~F (718~C) - 54.0 ksi (372 MN/m2) 1 hr at 1400~F (760~C) - 57.5 ksi (396 MN/m2) 16 hr at 1400~F (760~C) - 68.2 ksi (470 MN/m2)

TABLE 3 COMPARISON OF EXPERIMENTAL AND THEORETICAL VALUES OF THE OROWAN STRESS FOR TD-Ni,Cr Temperature ------ Experimental Theoretical Values* (ATo x 3.06) Temperate Applied Stress fcc Simple Cubic Random (~F) (~C) (ksi) (MN/m2) (ksi) (MN/m2) (ksi) (MN/m2) (ksi) (MN/m2) 1000 538 30 207,20 141 25 162 57 392 1400 760 23 159 18 123 21 142 50 343 1800 982 16 110 13 88 15 101 36 245 *Theoretical values were calculated using three particle distributions: face centered cubic (k = 1764 '), simple cubic (X = 1563 ~), and random (X = 782 A).

20 TEST TEMPERATURE OF ~C 200 0 1000 538 D 1100 593 o 1200 649 10 7 MOSMOOT H 1300 704 *^^ ______ D ^^^^Ds~~~ — ~ -0 1400 760 ~ ~ ~ ~ ~ ~ ~ ~ ~ ~832 34 36 38 40 42m P = T ( 20 + Log t) $ 1003, T = ~R = /s ~K 6 80 wu _ z 60 NOTCHED 40 32 34 36 38 40 42 P = T(20 + Logt) Y 10 -3 T = R = 9/5 OK Figure 1. Larson-Miller parameter curves showing the time-temperature dependence of the rupture strengths of Waspaloy heat treated 1/2 hour at 1975~F (1080~C) and aged 16 hours at 1400~F (760~C). Reference 1

600 SMOOTH -Open Points NOTCHED - Solid Points NOtrCC 1'D" ol 400 80 60 40 a. cr - 20 vat u t/) tvl -1;0 60 100 i 80 3 Z lsanoF (982~C) A A &. -61 A- a #'W- -- A a 60 40 ) 8 6 10000 1000 L — J —^100 0 ours obtained from \00 |.Hours 10 100 RUPTURE TIME, Hours ligueoStrsse utu e tm d ata at tnperatres from 000 to 1800F (5398ZC obtaine d from Fi~Stress vrss.rupt. t. -....- n-inch (0, - notched specimens f 0 15h ( 3mm) thick, stess sNotim dndent notch sensitivity was observed No time-dependent no

AGED 3 HOURS AT 1325~F(718C ) 6 80 60 -- 8 40 -- 2 10 AGED 1 HOUR AT 1400~F(760~C) _.z.1 1400 _760 u, Q I 1 I I I I a — 66 60 — 4 8 31 33 35 37 39 41 43 40 |P =T(20 + LogE) RA 10- T =~R _K10 AGED 16HOURS AT 1400~F(7600C) 100 6u 3 60 - 37 344 8 0 1000 538 z X 40 -- 20 P T(20 + LogES ) 10-3 T =OR ATR0K Figure 3. Larson-Miller parameter curves showing the time-temperature dependence of the rupture strengths of 0. 025-inch (0. 64mm) thick Modified Waspaloy sheet solution treated 1/2 hour at 1975~F (1080~C) and aged. Time-dependent notch sensitivity was not observed at test temperatures from 1000~ to 1400~F (538-760~C).

80 - - 600 60 400 40 000~F1538C M 812DD 120F(649~C) D 200 Z ~S;~~~ 20 - -------- ~~ --- —-----------------— r, ^"3.~~~~~~~200 1400~F(7608C),~ 20 H 100 1800~F (982~C) 80 10 - 8 -60 10-5 10-4 10 -3 10-2 10 -1 MINIMUM CREEP RATE, Percent per Hour Figure 4. Stress versus minimum creep rate behavior at 1000~ to 1800~F (538-982~C) for 0. 015-inch (0. 38mm) thick, stress relieved, TD-Ni, Cr sheet. The curve gradients indicate that the mechanism controlling creep changes between 1000~and 1200~F (538-649~C).

lOOx Figure 5. Optical Photomicrograph showing subsidiary intergranular cracks in the gauge section of a smooth specimen of 0, 015-inch (0. 38mm) thick TD-Ni, Cr sheet creep-rupture tested at 12ksi. (82. 7MN/m2) at 18000F (982~C).

STRESS, MN/m2 1000~F (538~C 1400~F (760 ~C) A. z 0 IIL u. UL C m 4 -4~ r,. 0 0 0 8 6 1600~ F (871~C) 20 50 40 30 20 10 STRESS, 1,OOOpsi. Figure 6. Iso-creep strain curves of life fraction versus stress at temperatures from 1000~ to 1800~F (538-982~C) for 0. 015-inch (0. 38mm) thick, stress relieved, TD-Ni, Cr sheet. The apparant large amounts of rupture life consumed for small creep strain at the higher temperatures and lower stresses was due to "Negative Creep".

STRESS, 100MN/m2 4 6 2 1,000 F (5380C) 12000 F (649 C) 1400~F (760~C) C 0 0. 0 UU U uL. -I c a0 U u - - 80 60 40 20 0 ' 2.0 2.0 1.0 1.0.. s_0-4.2 1100~F (5930C) 1300 F (704 ~C) 2.0 11.0 1.0 60 — 0. --- 0.2 -------- 0.1 20 110 90 70 50 30 STRESS, 1,000psi. Figure 7. Iso-creep strain curves of life fraction versus stress at temperatures from 1000~ to 1400~F (538-760~C) for 0. 025-inch (0. 64mm) thick Modified Waspaloy sheet heat treated 1/2hour at 1975~F (1080~C) plus 16 hours at 1400~F (760~C).Relatively little rupture life was consumed for small creep strains for all test conditions, Correspondingly, no time -dependent notch sensitivity was observed.

OA w w I 0.2 \ / 0. 4 c -0.2 w TIME, Hours Figure 8. Deformation-time curve for a smooth specimen of TD-Ni, Cr sheet creep-rupture tested at lOksi. (69 MN/m2) at 1800OF (982'C). At these test conditions a large amount of "Negative Creep" occured resulting in a negative minimum cree p rate,. 0 ^0 00 50 150 200 TIME, Hours Figure 8. Deformation-time curve for a smooth specimen of TD-Ni, Cr sheet creep-rupture tested at l0ksi. (69 MN/m2) at 1800~F (982~C). At these test conditions a large amount of "Negative Creep" occured resulting in a negative minimum creep rate,. }

(a) IOOx (b) 50, 000x Figure 9, Opticeal and transmagission ptlotomicrogra phs of as-received 0O 015-inch (0Q 38rnan) thick, stress relieved,r TDNi, Cr sheet The grains, 0. 080- 0. 16rmx int dianieter, were somnewhat elongated in the rolling direction. Many narrow twins were present. Relatively few, randomuily distributed, dislocations were ob.lserved.

7f. n00n (b) 100,000x (c) 80, OOx Figure 10. Transmission electron nAicrographts of thin foils taken fromn the gauge sections of tested smooth specienens of 0. 01%5.inch (0, 38mrm) thick TD-NNi, Cr sheet. (a)'tensile tested at 1400I0F (760CC). (b) creep.-rupture tested at 34ksi (234. 4 MN/mn-2) at 10000 (1760~C). (c) creep-rupture tested at 12ksi. (82. 7 MN/n ) at 18000F (982ZC). In (a) and (b) prismati c dislocation loops can be observed which indicate that dislocations by.*passed the ThO2 particles by cross.slip, Oxide particles in (c) were by.passed by c f i mb,

TEMPERATURE C TEMPERATURE C 500 600 700 800 900 1000 1100 1200 100 80 60 I I I I TENSILE CREEP 6 5 4 CROSS-SLIP CLIMB I 0 MIXTURE 0 u, 4/ Iv) 40 * 0 I-011-.,.. -. 1-1. "I-, I.. I... -, I.. I.. '..,."4 3 4 -4 m u, ~, 2 0 0 3 z 3 1 0- "." I 20 - -_ 0 %-% 0 % 0 'N NON " N ' O " N 0 0N % N 10 0 800 1000 1200 1400 TEMPERATURE OF 1600 1800 2000 2200 Figure 11. The variation of the dislocation mechanism observed as a function of test stress and temperature for 0. 015-inch (0. 38mm) thick, stress relieved, TD-Ni, Cr sheet. The mechanism changes from climb to cross-slip as the stress is increased beyond the "Orowan Stress".

250x Figure 12. Optical photornicrograph of 0. 025.-inch (0. 64mnlm) thick Modified Waspaloy sheet heat treated 1/2 hour at 1975"F (1080 C) plus 16 hours at 1400~F (760oC).

(a) 70, 000x (b) 55, 000x (c) 110, 000x Figure 13 Transmission electron micrographs of thin foils fro-t the gauge sections of smooth specimnens and the shoulder of a notched specimen of Modified Waspaloy solution treated 1/2 hour at 19750F (1080"C), aged, and creep-rupture tested. (a) smooth specimen aged 1 hour at.1400' (7600C), tested at 90ksi. (620 MN/nm2) at 1100~1F4 (59 3C). (b)notclhed specizlien aged 1 hour at 14000F (760~C) tested at 60ksi. (414 MNl/m2) at 11004F (5930C)-the stress in the shoulder section was 42ksi. (290 MN/n2). (c)snmooth specimnen aged 16 hours at 1400F) (760C), tested at 50ksi. (345 MN/im2) at i300~ F (704 ~C) Prismatic dislocation loops in (a) indicate that the dislocations by-passed the y' by cross-slip, ' particles in (b) were sheared by super dislocation-s. The concentric dislocation loops in (c) were formed by dislocations bowing between y' particles pinching off the loops,

Fa ce - Centered Cubic N N N 120 k Experimental N % Simple Cubic. \ Random \ m 100 1 - --- \ ( - 80 - IA I. n a-! A il 8 N 6 j I % 60 - S \ \% \ S m 4 -I m u, u, z N fI 4A 3 N,., N. 40 - SMOOTH NOTCHED N. CROSS-SLIP and LOOPING O _ SHEARING 0 MIXTURE D 0 20 O I I I I II N.. N -14 N N%. %-IN - ' _ _ N_ __ I 2 I II I I I I I I I -I 0 I I I a I I 0 10 50 100 500 1,000 PARTICLE SIZE, A Figure 14. The variation of the dislocation mechanism as a function of test stress, compensated to 1000~F (538~C), and y' particle size for 0. 025-inch (0. 64mm) thick Modified Waspaloy heat treated 1/2 hour at 1975~F (1080~C) and aged and tested at 10000 to 1400~F (538 -760~C). Cross-slip and looping occurred at stresses greater than the "Orowan Stress". Theoretical values for the "Orowan Stress" varied according to the y' particle distribution assumed.

120 8 6 100 1 - 80 _ -0 I tn QA u, ad 4e 60 - -4 C, -4 m v8 z 3 -3 M S I~~~~~~~ 40 1 - 1CA L cl..Y L:)itr ' I; -- 2 20 [ bC ( /d CROSS -SLI F SHEARING MIXTURE SMOOTH NOTCHED 0 0 S U Ii / I I I 0 O-1 200 0 50 100 150 0 PARTICLE SIZE, A Figure 15. The variation of the dislocation mechanism as a function of test stress, compensated to 1000'F (538~C), and the y' particle size for 0,025-inch (0. 64mm) thick Modified Waspaloy heat treated 1/2 hour at 1975~F (1080~C) and aged and tested at 1000~ to 1400~F (538-760~C). Dislocations can shear y' particles when the stress is greater than a ''Critical Stress".

100 \ CROSS-SLIP or LOOPING 80 rL, SHEARING _ --- 60 STTENGO T 4 __ a~ 00 ' $< TIME-DEPENDENT 40 NOTCH SENSITIVITY -- I^ 0 OROWA-N STRESST '^.. _i (At "high" temperatures) 0 ii 0 100 200 300 400 PARTICLE SIZE, A Figure 16. The dislocation mechanisms, as a. function of test stress and particle size, that can occur for a y' strengthened alloy. The curves shown are for Modified Waspaloy at 1000~F (538~C). Dislocations can shemr y' particles snaller than a critical size. Time-dependent edge-notch sensitivity has only been observed when the test stress is below the yield stress and when dislocations shear y' particles.

3CO - 1400~F (760 Z Z is ' i-3250F(718~C) - 200 IQ Z 0,00 5 I 1 I 100..i I i i I I I IIi11 1l l. I 1 10 100 AGING TIME, Hours Figure 17. The effect of aging exposures at 1325~ and 1400~F (718, 760~F) on the Diamond Pyramid Hardness of Modified Waspaloy solution treated 1/2 hour at 1975~F (1080~C).All of the aging treatments used in the study, 3 hours at 1325~F (718~C), 1 and 16 hours at 1400~F (760~C), resulted in y' particles smaller than the critical size ie. maximum hardness.

UNIVERSITY OF MICHIGAN 901 1111111111111103 ll30 1111111111111111 3 9015 03530 0147