Materials Science and Engineering, 61 (1983) 173-184 173

Improvement of Creep Strength in a Nickel-base Single-crystal Superalloy by Heat Treatment

P. CARON and T. KHAN

Office National d'Etudes et de Recherches A~rospatiales, 92320 Chdtillon (France)

(Received February 5, 1983;in revised form April 21, 1983)

SUMMARY

The effect o f two heat treatments on the creep behaviour o f CMSX-2, a recently developed nickel-base single-crystal turbine blade alloy, was investigated in the tempera- ture range 760-1050 °C. It is shown that the heat treatments had no effect on the ~/ volume fraction, the composition of precipi- tates and the coherency strains. However, the heat treatment resulting in cuboidal aligned precipitates, 0.45 pm in size, leads to a two- fold improvement in creep lives over the corresponding values obtained with the other heat treatment which produces smaller and more odd-shaped particles. Transmission elec- tron microscopy on specimens crept to various strains at 760 °C showed that, in con- trast with the odd-shaped particles, the cuboidal precipitates promoted a very homo- geneous deformation which led to the lower creep rates. The precipitate shearing mecha- nism in the late stages o f primary creep is shown to be by {111}(112} slip regardless of the heat treatment. In secondary creep the shearing occurs by (a/2)(110 ) dislocation pairs but the difference in creep behaviour is thought to be associated with the higher stability of dislocation networks associated with cuboidal precipitates. A t 1050 °C the more perfect rafted structure produced by the cuboidal precipitates is responsible for lower creep rates. The ~" rafts at high stresses were sheared by (a/2)(11 O) dislocation pairs.

1. INTRODUCTION

After the development and application of columnar-grained directionally solidified blades and vanes, a further improvement in allowable metal temperatures of future jet

engines seems to be associated with the intro- duction of single-crystal superalloys. Some aircraft engine manufacturers are planning to power the critical stages of their engines with nickel-base single-crystal superalloys [1]. The durability of single-crystal blades is usually two to three times that of the columnar- grained material. Of prime importance for a blading material is its creep strength. The creep behaviour of ~/-'y' alloys strengthened by the coherent ~/' (Ni3(A1, Ti)) phase primarily de- pends on the following factors: (1) the volume fraction of ~', (2) the precipitate size, (3) the coherency strains due to the lattice misfit and (4) the compositions of the ~/and ~/' phases. This investigation is aimed at studying the effect of heat treatments on these factors and at understanding the creep behaviour of a nickel-base single-crystal alloy recently developed by the Cannon-Muskegon Corpora- tion and designated CMSX-2 [2]. A detailed study of the creep deformation mechanisms of this alloy was therefore undertaken at 760 and 1050 °C for two different heat treatments.

2. MATERIAL AND EXPERIMENTAL PROCEDURE

The composition of the master heat used for this investigation is shown in Table 1. Dendritic single crystals of this alloy 8 mm in diameter and 120 mm long were cast in the [001] direction in a temperature gradient of about 250 °C cm -1 and at a withdrawal rate of 15 cm h -1. The exact orientations were deter- mined by the Laue back-reflection technique; single crystals selected for creep tests were within 5 ° of the [001] orientation. The mean primary interdendritic spacing h was found to be 160 pm. The size of primary ~,' associated

0025-5416/83/$3.00 © Elsevier Sequoia/Printed in The Netherlands

174

[-,

L)

O

O

o~

O

O

V

V

V

,3

5

¢O

¢O

O 06

~e

¢D

¢)

with the 7-7 ' eutectic present in the inter- dendritic regions is between 10 and 30 pm. Specimens for metallographic examination and creep tests were solution treated in argon at 1315°C for 30 min and air cooled in order to solutionize the 7' phase and to homogenize the alloy partially. Microprobe analysis of segregations across the dendrites on the as- cast and homogenized single crystals showed that, except for tungsten, all the other ele- ments were homogenized fairly well. The specimens were divided into two batches and each batch was given the following precipita- tion heat treatments: batch 1 , 9 8 0 °C for 5 h, air cool + 850 °C for 48 h; batch 2, 1050 °C for 16 h, air cool + 850 °C for 48 h. These heat treatments will be later referred to as T 1 and T2 respectively. The T1 t reatment is the standard precipitation heat t reatment used by Cannon-Muskegon which was previously applied to a similar experimental alloy [3]. This heat t reatment produces particles of various shapes and sizes which depend on the area examined (Fig. 1). The extreme mean sizes of particles vary from 0.25 to 0.36 pm. The second heat t reatment T2, which is one of the various heat t reatments being studied at the Office National d 'Etudes et de Re- cherches A6rospatiales results in cuboidal particles having a mean size of 0.45 pm and with their (100) faces aligned parallel to the growth direction of the crystal (Fig. 2). This type of morphology and size can also be obtained by a precipitation heat t reatment at even higher temperatures for shorter periods, e.g. 1100 °C for 4 h. TEM did not show any 7-7 ' interfacial dislocations after the two types of heat treatment•

In order to determine the possible differ- ences between the 7 ~ volume fractions, the composit ions of the 7' phase and the 7-7 ' misfits on creep behaviour, the precipitates were extracted from the matrix after the heat treatments T 1 and T 2 using a standard tech- nique [4]. The composi t ion of the 7' phase was determined both by chemical analysis of the extracted precipitates and by electron microprobe analysis of the extracted and cold-compacted particles. The lattice param- eters of both the irregularly shaped and the cuboidal 7' were accurately determined at room temperature from the extracted par- ticles with an X-ray diffractometer using (111), (200), (311), (222) and (400) peaks•

175

Fig. 2. TEM micrograph of the ~/-~" structure of T 2 heat-treated CMSX-2. The mean particle size is 0.45 pm. The foil normal is [001].

Fig. 1. Transmission electron microscopy (TEM) micrographs of the T~7 r structure of T 1 heat-treated CMSX-2. The size of ~/' precipitates varies from o n e

region to another. In (a) the mean size is 0.25 pm and in (b) it is 0.36 pro. The foil normal is [001].

An a t t e m p t was also made to measure the pa rame te r s o f the 7 and 7 ' phases on bulk single-crystal specimens using the same X-ray d i f f r ac tome te r . No sa t i s fac tory measu remen t s could be made wi th this m e t h o d since the 7 and 7 ' peaks could no t be resolved. A n o t h e r a t t e m p t was made to measure these param- eters by grinding some specimens hea t t r ea t ed accord ing to the T 1 and T 2 modes and t h e n using the S e e m a n - B o h l i n camera. Pr ior to the measu remen t s the g round al loy powde r s were vacuum annealed in sealed quar tz tubes at 1000 °C for 1 h in o rde r to relieve the stresses.

I t was thus possible to r ecord reasonably well- separated peaks of 7 and 7 ' phases, t h e r eb y enabling the paramete rs to be de t e rmined f ro m (220) and (311) peaks by ex t rapola t ing to 0 = 90 °.

Tensile c reep tests were c o n d u c t e d in air in the t e m p e r a t u r e range 7 6 0 - 1 0 5 0 °C for b o t h hea t t r ea tmen t s T1 and T2. The speci- mens wi th a gauge length o f 20 m m and a d iamete r of 3 m m were used to s tudy the creep behaviour and d e f o r m a t i o n mechanisms. Most o f the specimens were c rep t to fai lure b u t a few tests at 760 and 1050 °C were in t e r rup ted at various d e f o r m a t i o n levels to s tudy the d is locat ion s t ruc tures f o rm ed dur ing creep. These specimens were fo rced air coo led u n d e r load to r o o m t e m p e r a t u r e to preserve the d is locat ion s t ructure . Th in slices were cu t pe rpend icu la r to the stress axis and parallel to the {111} planes. These slices were mechanica l ly th inned to 0 .15 m m and t h en e lec t ro ly t ica l ly pol ished at r o o m t e m p e r a t u r e in a 10% solu t ion of perchlor ic acid in acet ic acid. The th in foils were examined in a J E O L 200 CX transmiss ion e lec t ron mic roscope ope ra t ed at 200 kV.

3. EXPERIMENTAL RESULTS

3.1. Weight fraction and composition ofT' phase

The weight f rac t ion of 7 ' part icles was 68 + 2% in d ep en d en t of the hea t t r ea tmen t . The compos i t ions of b o t h the irregularly

176

TABLE 2

Chemical composition of the T' phase

Element Ni Co Cr Mo W A1 Ti Ta Amount (wt.%) 69.25 3.15 2.05 0.30 7.25 7.55 1.30 9.15

TABLE 3

Creep properties of CMSX-2 alloy after the two precipitation heat treatments

Test conditions Hea t t r e a t m e n t T 1

Temperature Stress Time for 1% Rupture Minimum (°C) (MPa) creep life creep rate

(h) (h) × 108 (s-~)

Heat t r e a t m e n t T 2

T ime f o r 1% R u p t u r e M i n i m u m creep life creep rate (h) (h) × 108

(s-1)

760 750 54 569 2.92

850 500 38 198 5.83 950 240 102 255 2.78

93 272 3.33 1050 120 323 468 0.56

330 461 0.56

120 1138 1.72 134 1117 1.53 140 450 1.81 142 370 1.31 144 380 1.31 870 1055 0.28 587 707 0.25

shaped and the cuboidal precipitates were also found to be independent of the heat treat- ment. Neither the chemical analysis of the extracted particles nor the electron micro- probe analysis of the extracted and compacted particles showed a significant difference in compositions. The composition of the 7' phase is shown in Table 2.

3.2. Lat t ice parameters o f 7 and ~f' phases The parameters of the extracted (uncon-

strained) 7' phase were found to be 3.5865 A for the irregularly shaped particles and 3.5864 A for the cuboidal precipitates. As expected, these results confirm that the composition of the 7" phase was not affected by the heat treatments. The in situ measure ments of parameters with the Seeman-Bohlin apparatus on the mechanically ground and subsequently annealed alloy powders gave the following results: a~, = 3.5848 A and a~ --- 3.5724 A for the T~ heat t reatment; a~, = 3.5849)~ and a~ = 3 .5798A for the T2 heat treatment.

Since the parameter of constrained ~/' is again independent of the heat t reatment it is evident that the parameter of the ~, phase should not be affected by the selection of T1

versus T 2. The discrepancy between the parameters of the ~, phase in fact arises from the unreliable value determined from T1 speci- mens due to rather broad diffraction peaks. The T2 specimens did permit an accurate determination of the misfit which was found to be 0.14%. Here, the misfit is defined as (a~ , - a~)/a~. The change in particle mor- phology from odd shaped to cuboidal is presumably related to a slightly higher misfit at 1050 °C compared with that at 980 °C as a result of the difference in thermal expansion coefficients between the two phases. How- ever, the lattice parameters determined at room temperature from the fully heat-treated specimens (T1 and T2) are almost identical.

3.3. Creep resutts In order to investigate the effect of heat

treatments T 1 and T2 which result in two distinct 7' morphologies and different sizes, tensile creep tests were run in the tempera- ture range 760-1050 °C. The creep results are shown in Table 3. These data include the time for 1% creep, the rupture life and the mini- mum creep rate. The heat t reatment T 2 which produces aligned cuboidal particles clearly leads to a significant improvement in the

creep behaviour of the alloy over the standard t reatment T 1. Both the time for 1% creep and the time to failure are increased by a factor of 1.5-2 depending on the temperature. Typical creep curves at 760 °C and 750 MPa corre- sponding to the two heat treatments are shown in Fig. 3. The creep curves exhibit the usual three stages at all temperatures but the duration of steady state creep is rather short at temperatures below 1000 °C.

At 760 °C the curves corresponding to the T1 heat t reatment show a very low creep rate (incubation period) for about 20 min, prior to the onset of primary creep (see Fig. 3, inset). The primary creep rate suddenly increases after the incubation period. For T2-type specimens the incubation period is absent and the primary creep rate is lower. A super- position of the curves corresponding to the

177

two heat treatments shows that they intersect at around 2 h. Subsequently, the creep rate of Tl-type specimens becomes much higher than that of T 2 specimens. After 15 h the primary creep strain of Tl-type specimens is about 0.55% whereas that of T2-type specimens is only 0.2%. The secondary creep rate also follows the same trend and is a factor of 2 lower for T2-type specimens. At 950 °C under a stress of 240 MPa, the primary creep period is very short and the steady state creep rate for the cuboidal precipitates is again half that of the irregularly shaped particles. However, at this temperature the creep properties of T2-type specimens showed only a 50% im- provement over the Tl-type specimens. Typical creep curves at 1050 °C and 120 MPa for the two heat treatments are shown in Fig. 4. The primary creep extends to about

28

24

20

16 z < ~," 12 i- (/)

8

L TI"~

. , c ~ - HEAT / U) 0 2 4. 6 I TREATMENT /

TIME (h) / / ~" T 1 /

/ / J ~" HEAT / / J TREATMENT T 2

4 t . ~

I I

0 200 400 600 800 1000 1200

TIME (h)

Fig. 3. Creep curves for CMSX-2 at 760 °C and 750 MPa.

28

24

20

16 z < r,. 12 I - - u')

8

HEAT TREATMENT T 2

HEAT TREATMENT T| ~ "

/ I /

4 /

o~ 200 4o0 ~ o ~ lo~o TIME (h)

Fig. 4. C r e e p c u r v e s for CMSX-2 at 1050 °C and 120 MPa.

1200

178

24 h in both cases but the secondary creep rate for cuboidal precipitates is less than 3 × 10 -9 s -1.

3.4. Microstructural observations 3.4.1. Creep at 760 °C Typical dislocation configurations after 2 h

of creep are shown in Fig. 5 for both T1 and T 2 treatments. At this stage the creep curves intersect each other and the strain is 0.16%. Dense dislocation networks are formed at the ~/-7' interfaces by the glide of (a/2)<l10> dis- locations in the matrix. Stacking faults in a few precipitates are also visible on ( l l l } - t y p e planes. Such intrinsic-extrinsic stacking faults were also observed during the primary creep of MAR-M 200 single crystals at 760 °C [5] and analysed by Kear et al. [6]. The nature of the stacking faults was completely identified in the CMSX-2 alloy and our observations fully confirm the precipitate shearing mecha- nism by { 111}(112 > slip proposed by Kear et al. [6]. Careful examination of thin foils cut perpendicular to the crystal growth direction shows that the four { 111}< 112> slip systems are operative which is consistent with the [001] orientation. Although the same defor- mation mechanisms were observed for the two heat treatments, a detailed examination of many thin foils shows a striking difference in terms of their overall deformation struc- ture. In effect, the deformation in the matrix was found to be very homogeneous in all regions examined in cuboidal precipitates where all the ~/-~' interfaces were covered with dislocations (Fig. 5(c)). These networks are formed when the dislocations gliding between the precipitates leave dislocation loops at the ~/-~' interfaces. An example of such a configuration is shown in Fig. 6. In this case the foil was tilted at 40 ° and it was convenient to observe a homogeneous array of dislocations at the intersections of the 7' faces with the {111} glide planes. On the contrary, the deformation structure for T1- type specimens was clearly heterogeneous and the dislocations were wavy. The nature of the deformation structure was dependent on the region under examination. In areas where the precipitates were cuboidal and rather large in size the deformation was found to be homo- geneous whereas in other regions where the precipitates were less cuboidal (or odd shaped)

Fig. 5. Dislocation structures after 2 h of creep at 760 °C and 750 MPa (0.16% strain for both speci- mens). Two types of deformation structures are present in the T 1 spec!men: (a) heterogeneous; (b) homogeneous. (c) In the T 2 specimen the deformation is homogeneous regardless of the region. The foil normals are [001]; g]/[200].

and small the deformation was heterogeneous (Figs. 5(a) and 5(b)).

The deformation structure of a specimen with cuboidal particles at 0.2% strain after

179

Fig. 6. Arrangement of (a]2)(110) dislocations at the 7-7' interfaces after 2 h of creep at 760 °C and 750 MPa in a T 2 specimen. The foil normal is [001]; gZ[11i].

Fig. 8. Typical steady state substructure in a T2 speci- men strained to 1% at 760 °C and 750 MPa. The foil normal is [O01];g//[200].

Fig. 7. TEM micrograph showing stacking faults due to the shearing of ~/t precipitates by {111}(112) slip after 15 h of creep at 760 °C and 750 MPa in a T 2 specimen (0.2% strain). The foil normal is [001]; g // [ 200 ],

15 h co r re spond ing to the la ter stage o f p r imary creep is shown in Fig. 7. At this stage a ve ry large n u m b e r o f prec ip i ta tes have been sheared by {111}(112} slip. Similar observa- t ions were made fo r T l - t ype specimens at the end o f p r imary creep where the strain was a lmos t t h ree t imes higher (0 .55) a f te r 15 h of creep. However , at this c reep strain the defor- ma t ion has b e c o m e h o m o g e n e o u s which sharply cont ras t s wi th the he t e rogeneous charac te r o f d e f o r m a t i o n at the onse t o f p r imary creep.

The general m ic ros t ruc tu re o f T 1- and T 2- t ype t r ea t ed specimens dur ing secondary creep is main ly charac te r ized by a high dens i ty o f ( a / 2 ) ( l 1 0 ) dis locat ions arranged in

Fig. 9. Shearing of a ~,t precipitate by pairs of (a/2)(110) dislocations during the steady state creep of a T 2 specimen at 760 °C and 750 MPa. The foil normal is [O01];g//[111].

ne tworks at the 7 - 7 ' in terfaces {Fig. 8). Pairs o f in t r ins ic-ext r ins ic stacking faul ts are still visible bu t an o th e r m o d e o f shearing of the 7 ' precipi ta tes is also f r e q u e n t l y observed in secondary creep. Regardless of the hea t treat- m en t the prec ip i ta tes are cu t by cons t r ic ted ( a /2 ) ( l10~ dis locat ion pairs coup led by high energy ant iphase boundar ies (APBs) (Fig. 9).

3.4.2. Creep at 1 0 5 0 ° C The m o s t striking morpho log ica l f ea tu re

o f creep at 1050 °C is the d i rec t ional coarsen- ing o f the 7 ' phase. In fact , the t e m p e r a t u r e at which the 7 ' part icles s tar t raf t ing unde r stress is a round 900 °C. In the CMSX-2 al loy the 7 ' part icles coarsen as platelets perpen- dicular to the [001] stress axis. The thick- ness o f the platelets is close to the initial size

180

Fig. 10. SEM micrographs showing the morphology of the ~/' phase after 20 h of creep at 1050 °C and 120 MPa: (a) T 1 specimen; (b) T 2 specimen. The sections are parallel to the [001] stress axis.

Fig. 11. TEM micrographs of interfacial dislocation networks formed after 160 h of creep at 1050 °C and 120 MPa in a T2 specimen. In (a) the foil is parallel to the [001] stress axis and in (b) the section is normal to [O01];g//[200].

of 7' precipitates and their lateral extension is usually several tens of microns (20-50 pm). During the late stages of secondary creep it was not possible to differentiate clearly between the morphologies of rafts formed from the initially cuboidal or odd-shaped particles. Therefore, in order to observe the initial stages of raft formation, some creep tests were interrupted after 20 h of creep. The micrographs of Fig. 10 show that the rafts are already well formed in aligned cuboidal precipitates whereas the rafted morphology is less perfect for the initially non-aligned more irregularly shaped particles.

The dislocation microstructure at 1050 °C is characterized by a high density of (a/2)<l10) dislocations at the ~'-7' interfaces (Fig. 11). The hexagonal networks which are typical of coarsened structures are formed by inter- actions between the (a/2)<110) dislocations

Fig. 12. Shearing of a ~" raft by a pair of (a/2)(110) dislocations at 1050 °C and 120 MPa. The foil is parallel to the [001] stress axis in a T 2 specimen; g//[i11].

[7]. Another important feature of deforma- tion at 1050 °C which was also observed in secondary creep at 760 °C is the presence of

f (a/2)<110> dislocation pairs within the 7 particles (Fig. 12).

4. DISCUSSION

4.1. Creep at 760 °C The T2-type heat t reatments producing

well-aligned cuboidal precipitates about 0.45 pm in size lead to a significant improve- ment in the creep strength of CMSX-2 single- crystal alloy compared with the Tl- type heat treatments resulting in the smaller irregularly shaped particles. A detailed experimental investigation has shown that the volume fraction, the chemical composi t ion and the lattice parameter of the 7' phase are inde- pendent of heat treatments. Consequently, the factors which should be taken into account in order to understand the difference between the creep behaviours of the T1 and T2 alloy specimens are the shape, the size and the distribution of 7' particles in the matrix. The major difference between the deforma- tion structures of T1 and T2 specimens during the early stages of primary creep at 760 °C lies in the more or less homogeneous character of deformation. In T2 specimens, primary creep occurs as soon as the load is applied and the creep rate decreases cont inuously until the steady state regime is attained. As illustrated by the homogeneous deformation structure after 2 h of creep (Fig. 5(c)), the 7' precipi- tates produced by the T 2 heat t reatment p romote a rapid and pronounced strain hardening which subsequent ly reduces the ampli tude of the primary creep strain. In contrast, the 7' precipitate morphologies obtained by the T1 heat t reatment p romote a distinctly heterogeneous deformat ion and result in a higher creep strain immediately after the so-called incubation period.

We shall now a t tempt to analyse the possible effects of size, shape and distribution of the 7' phase on the incubation period and on the homogeneous (or heterogeneous) character of deformation. The occurrence of a short incubation period in Tl- type specimens is presumably related to the higher Orowan stress as a result of the smaller particle spacing. Moreover, since the dislocations can glide

181

between the precipitates only in those regions where the 7' spacing is slightly higher, the deformation will tend to be heterogeneous after the incubation period. It is thought that the heterogeneous nature of deformation is enhanced by the variable strength of the 7-7 ' regions which arises from the hetero- geneous local distribution of precipitates. For T2-type heat-treated specimens the aligned and more loosely spaced configuration of precipitates permits an easier and more homo- geneous glide of the (a/2)(l10> dislocations immediately after the application of a load. The following line tension considerations and experimental evidence indicate that easy shearing at the corners of a homogeneous array of cuboidal particles will lead to a homogeneous deformation and rapid strain hardening during primary creep. Typical dis- location configurations observed in the {111} planes after 0.16% creep strain in both types of specimens are shown in Fig. 13. The (a/2)<110> dislocations gliding in the {111} planes are arrested at the 7 -7 ' interfaces. The cuboidal precipitates appear as triangles with rounded corners and the radius of curvature of the dislocation around the corner of the precipitate is obviously small. For irregularly shaped particles the cross section of the par- ticle in the { l l l } - t y p e plane is close to a sphere and the radius of curvature of the dis- location is significantly higher. According to Copley and Kear [8] a force due to the line tension of a curved dislocation acts normal to the dislocation and towards it centre of curvature. This force plays a key role in push- ing the dislocation into the precipitate and is given by

F = ~/r o

where ~ is the line tension and r0 the radius of curvature of the dislocation. Copley and Kear [8] considered this force for evaluating the critical resolved shear stress in MAR-M 200 single crystals by using the following expres- sion:

7 ¢ 1 T + - ( + Tp)

2b bro 2 -7"m

where Tm and r , are respectively the critical resolved shear stress of the matrix and of the precipitate, b the Burgers vector of (a/2)(110> dislocations which shear the precipitate and 7

182

Fig. 13. Dislocation configurations around (a) spherical and (b) cuboidal ~f' precipitates in primary creep at 760 °C and 750 MPa. The foil normals are [111];g//[220].

the APB energy. The force F will increase with decreasing radius of curvature and hence the dislocation penetrat ion in the {111} planes will become easier at the corners of the cuboidal precipitates than in the nearly spherical particles.

These considerations together with the regular precipitate morphology of T2-type specimens imply that there will be a rapid and uniform increase in the dislocation den- sity. In T1 specimens the only regions where the dislocation density is uniform are those containing rather large cuboidal particles. In other regions the shearing of odd-shaped particles is rendered more difficult as a result of the higher radius of curvature of precipi- tates. This gives rise to a heterogeneous deformation. Moreover, the wavy character of dislocations in areas where the deforma- tion is heterogeneous suggests that the 7' particles are being avoided by a combined

glide and climb process (and perhaps some cross-slip). This type of mechanism seems to be responsible for the higher creep rates.

The deformation structures observed during the secondary creep are quite similar for the two types of heat treatments. There- fore, the effect of size, shape and distribution of 7' particles on creep rates is not quite clear. During secondary creep the rate of deforma- tion is controlled by strain-hardening and recovery mechanisms. The sizes of dislocation networks which contr ibute to the strain hardening are similar to those of the precipi- tates. Since the precipitate size is not too different after the two heat treatments, we think that the strain hardening cannot explain the lower secondary creep rates observed for the cuboidal precipitates. It is thought that the lower creep rates in T2 specimens are due to the higher stability of dislocation networks at the 7-'7' interfaces. It seems that both the size and the shape of particles play an im- portant role in the rate of recovery and hence on the steady state creep rate. However, the relative contributions of the size and shape of 7' particles to the recovery rate are not clearly understood.

4.2. Creep at 1050°C At high temperatures the rafting of 7'

precipitates perpendicular to the tensile stress axis is the major structural feature observed in CMSX-2 single crystals. We shall first briefly consider the factors which may promote a rafted 7' structure and then discuss the effect of this morphology on the creep strength of CMSX-2 alloy in terms of deformation mechanisms. The oriented coarsening of 7' phase was observed by various researchers during stress-annealing experiments or during creep tests at elevated temperatures. The coarsening of the 7' phase in the form of rafts perpendicular to the [001] tensile axis in single crystals was reported, for instance, in Udimet 700 [9] and MAR-M 200 [10]. On the basis of elastic energy considerations and assuming a low volume fraction of the 7' phase, Pineau [11] has a t tempted to predict equilibrium precipitate morphologies during stress annealing. According to Pineau, the major factors which affect the stress-coarsening behaviour are the direction and the absolute value of the applied stress, the ratio between the elastic moduli of the two phases as well as

183

the ~-~' misfit. The rafted structure of ~,' in CMSX-2, Udimet 700 and MAR-M 200 alloys is consistent with the theoretical predictions considering the slightly positive misfits in these alloys at room temperature (1.4 X 10 -3, 2 X 10 .4 and 5 X 10 -4 respectively). During the high temperature creep of T-~/' alloys with a rafted ~/' structure an important deformation mechanism is bypassing of the particles by a combined glide and climb process. However, if the stress is sufficiently high, the rafts can be sheared by (a /2)( l10) dislocation pairs as observed in CMSX-2 single crystals crept at 1050 °C and t 20 MPa. This mechanism in- volves some type of slip of (a/2)(110) disloca- tion pairs coupled by APBs. Pearson et al. [ 12] have suggested that the climb of the matrix dislocations around the ~/' particles becomes difficult if the aspect ratio (the ratio of the length to width) of the rafts is high. He proposed that under such conditions the slip process in ~' rafts becomes the rate-controlling creep mechanism. It is suggested that in such a situation the stress required to shear the ~/' phase depends on the magnitude of the APB energy and on the degree of misfit. In the particular case of the CMSX-2 alloy the com- position of the ~/' phase and the ~/-~" misfit remain unaltered by the heat t reatments T1 and T2. Experimental evidence shows that the lower creep rates in initially cuboidal precipitates are due to the higher perfection of rafted structure (Fig. 10). The aligned nature of the cuboidal precipitates promotes a rapid formation of more perfect platelets with a high aspect ratio. In contrast, the occurrence of a well-aligned rafted mor- phology is retarded for the initially irregularly shaped particles where the rate of recovery of dislocations by climb will be greater, thereby resulting in higher secondary creep rates.

However, the T2-type heat t reatments which lead to aligned cuboidal precipitates, about 0.45 pm in size, considerably improve the creep strength of CMSX-2 single crystals in the studied temperature range over the smaller and more irregularly shaped particles. A detailed analysis of deformation mecha- nisms and dislocation structures generated during creep at 760 and 1050 °C is presented in an a t tempt to explain the different creep behaviours associated with the two heat treat- ments. It is shown that at 760 °C the cuboidal precipitates promote a homogeneous deforma- tion during primary creep which leads to a uniform strain hardening. In contrast, the T 1- type heat treatments which give rise to pre- cipitates of various sizes and shapes promote heterogeneous deformation during the early stages of primary creep. The deformation behaviour in the early stages of creep prob- ably has a strong influence on the later stages of primary and secondary creep. Regardless of the heat treatment, shearing of precipi- tates by {111}(112) slip is a general feature of the deformation mechanism during the late stages of primary creep. In secondary creep the shearing of precipitates by constricted (a /2) ( l10) dislocation pairs is also observed. However, the lower creep rates for the cuboidal precipitates seem to be due to the higher stability of dislocation networks at the T-7' interfaces.

At 1050 °C the precipitates coarsen perpen- dicular to the [001] tensile stress axis. At this temperature the lower secondary creep rates observed for cuboidal precipitates arise from an early formation of nearly perfect rafts because of the initially aligned nature of precipitates. Climb around ~' rafts becomes difficult and at high stresses these rafts are sheared by (a/2)(110) dislocation pairs.

5. SUMMARIZING REMARKS ACKNOWLEDGMENTS

The present investigation provides some interesting results concerning the effect of two different heat t reatments on the creep strength of a single-crystal superalloy, in the temperature range 760-1050 °C. It is shown that the composi t ion and volume fraction of precipitates as well as the ~-~/' coherency strains are not affected by heat treatments.

This work was supported by the Direction des Recherches Etudes et Techniques under contract. The master heat of the CMSX-2 alloy was provided by the Cannon-Muskegon Corpol t ion, MI. The authors wish to express their sincere appreciation to Mrs. Diot for the lattice parameter measurements, to Dr. Lasal- monie for stimulating discussions regarding

184

t h e d e f o r m a t i o n m e c h a n i s m s a n d t o P r o f e s s o r A n t o l o v i c h , U n i v e r s i t y o f C i n c i n n a t i , O H , f o r v a l u a b l e c o m m e n t s .

REFERENCES

1 M. Gell, D. N. Duhl and A. F. Giamei, in J. K. Tien, S. T. Wlodek, H. Morrow III, M. Gell and G. E. Maurer (eds.), Proc. 4th Int. Syrup. on Superalloys, Seven Springs, PA, American Society for Metals, Metals Park, OH, 1980, p. 205.

2 K. Harris, G. L. Erickson and R. E. Schwer, Metal- lurgical Society o f AIME Fall Meet., S t. Louis, MO, October 27, 1982; Publ., 1982, p. 36 (Cannon-Muskegon Corporation, Muskegon, MI).

3 T. E. Strangman, G. S. Hoppin III, C. M. Phipps, K. Harris and R. E. Schwer, in J. K. Tien, S. T. Wlodek, H. Morrow III, M. Gell and G. E. Maurer (eds.), Proc. 4th Int. Syrup. on Superalloys, Seven

Springs, PA, American Society for Metals, Metals Park, OH, 1980, p. 215.

4 O. H. Kriege and J. M. Baris, Trans. Am. Soc. Met., 62 (1969) 195.

5 G. R. Leverant and B. H. Kear, Metall. Trans., 1 (1970) 491.

6 B. H. Kear, G. R. Leverant and J. M. Oblak, Trans. Am. Soc. Met., 62 (1969) 639.

7 A. Lasalmonie and J. L. Strudel, Philos. Mag., 32 (1975) 1.

8 S. M. Copley and B. H. Kear, Trans. AIME, 239 (1967) 984.

9 J. K. Tien and S. M. Copley, Metall. Trans., 2 (1971) 543.

10 C. Carry, C. Houis and J. L. Strudel, Mdm. Sci. Rev. M~tall., 78 (1981) 139.

11 A. Pineau, Acta Metall., 24 (1976) 559. 12 D. D. Pearson, B. H. Kear and F. D. Lemkey, in

B. Wilshire and D. J. R. Owen (eds.), Proc. Int. Conf. on the Creep and Fracture o f Engineering Materials and Structures, Swansea, March 24-27, 1981, Pineridge, Swansea, 1981, p. 213.

本文献由“学霸图书馆-文献云下载”收集自网络，仅供学习交流使用。

学霸图书馆（www.xuebalib.com）是一个“整合众多图书馆数据库资源，

提供一站式文献检索和下载服务”的24 小时在线不限IP

图书馆。

图书馆致力于便利、促进学习与科研，提供最强文献下载服务。

图书馆导航：

图书馆首页 文献云下载 图书馆入口 外文数据库大全 疑难文献辅助工具