Microstructure and elevated temperature mechanical behavior of cast NiAl–Cr(Mo) alloyed with Hf

Journal of Alloys and Compounds 343 (2002) 142–150

L

www.elsevier.com / locate / jallcom

Microstructure and elevated temperature mechanical behavior of cast NiAl–Cr(Mo) alloyed with Hf J.T. Guo*, C.Y. Cui, Y.H. Qi, H.Q. Ye Institute of Metal Research, Chinese Academy of Sciences, Shenyang 110016, People’ s Republic of China Received 27 June 2001; accepted 13 December 2001

Abstract An NiAl–Cr(Mo) alloy modified with Hf was fabricated by vacuum induction, melted and drop cast and then was hot isostatically pressed (HIPed) at 1523 K, 200 MPa for 4.5 h. The alloy was mainly composed of NiAl matrix, Cr(Mo) and Hf-rich phase distributed on the NiAl and Cr(Mo) phase boundary. The NiAl–Cr(Mo) and NiAl–Ni 2 AlHf interfaces were also investigated. Its elevated temperature compressive behavior was studied and the deformation features of the alloy could be adequately described by standard power-law which is usually used to characterize creep behavior for various metallic materials. The stress exponent n as well as the activation energy Q was calculated by fitting the experimental data to the power-law and temperature-compensated power-law equation. The possible strengthening mechanism was also discussed. Because the alloy had relatively high tensile strength, the tensile creep was performed at 1373 K and, as concluded from the stress component of different testing methods, the creep behavior may not be influenced by the testing method.  2002 Published by Elsevier Science B.V. Keywords: Deformation; NiAl–28Cr–5Mo–1Hf; Hot isostatic press; Creep

1. Introduction In polycrystalline NiAl form, conventional NiAl alloys do not possess adequate strength to be useful at the elevated temperatures needed to compete with nickel-based superalloys [1,2] and there is little interest today in continuous fiber reinforced NiAl-based composites due to a number of presently irresolvable technique barriers [3]. However, there is an NiAl-based eutectic alloy system that has been actively investigated for potential utilization in niche high temperature applications [4]. One promising NiAl-eutectic alloy is the NiAl–Cr(Mo) system. Directionally solidified NiAl–Cr(Mo) eutectic has been studied for many years [5,6] because of its higher fracture toughness compared to many NiAl-based alloys. Ni–30.5Al–28Cr– 6Mo (unless otherwise stated, all compositions are in atomic percent) is regarded as the most logical choice of the multi-element system examined to date. Small addition of refractory metal from Group IVB, in *Corresponding author. Tel.: 186-24-238-43531 / 55493; fax: 186-24238-91320. E-mail address: [email protected] (J.T. Guo).

particular Hf, were found to be very effective in improving the high temperature creep strength of NiAl single crystal [7]. Extensive studies on mechanical properties of the single phase (Heusler) and the NiAl–Ni 2 AlHf two phase alloys have been reported [8–12], and the two-phase alloys were found to have excellent creep resistance, comparable to conventional nickel-based superalloys [8,9]. Directional solidification has been used to produce NiAl single crystal for use in aeroturbine engines [13–15]. However, the higher melting point of NiAl, close to 1912 K, which was about 300 K higher than the temperature used to process superalloys, pushes the stability of ceramic mold to their limit when in contact with molten NiAl [16,17]. Therefore the mold–metal reactions are serious issues, which led to Si contamination. In other words, the compressive creeps investigated by Whittenberger [18] and Zeumer [19] showed that the processing technique is not an important factor when NiAl–NiAlTa alloys are tested at 1373 K, which implies that use of the NiAl–X alloy at more elevated temperatures might not require the additional expense of directional solidification. Based on the above reason, the cast Ni–33Al–28Cr– 5Mo–1Hf (at.%) alloy was selected for research. The

0925-8388 / 02 / $ – see front matter  2002 Published by Elsevier Science B.V. PII: S0925-8388( 02 )00205-0

J.T. Guo et al. / Journal of Alloys and Compounds 343 (2002) 142 – 150

microstructure, interface and mechanical properties of cast NiAl–28Cr–5Mo–1Hf alloy in the HIPed condition were reported.

2. Experimental procedure The alloy used for this investigation was Ni–33Al– 28Cr–5Mo–1Hf in nominal. It was melted in a vacuum induction furnace and subsequently cast into ingots (120 mm in diameter by 170 mm in length). The mechanical properties were conducted on the HIPed (1523 K, 200 MPa, 4.5 h) condition. Compression tests with a specimen size of 43436 mm were conducted in air at 1273 and 1373 K and strain rates from 2310 23 to 2310 25 s 21 , using a Gleeble 1500 test machine. The tensile and creep specimens were also machined from the HIPed ingot with gauge dimensions of 232316 mm 3 . The tensile samples were tested on the AGN250 test machine. The creep tests were performed on DC-25 experimental machine and the creep deformations were recorded by dial indicator. The samples for transmission electron microscopy (TEM) were prepared for the HIPed state materials. After being cut into 3 mm discs with a thickness of 0.5 mm by the electron discharge machine (EDM), they were then thinned by mechanical polishing and dimple grinding to approximately 35 mm. TEM and high resolution electron microscopy (HREM) observations were performed with a JEM-2010 high resolution electron microscope. Scanning electron microscopy (SEM) techniques were used to examine the microstructure, the fracture surface of tensile and creep samples.

3. Results and discussion

3.1. Microstructure A typical microstructure of the as-cast NiAl–28Cr– 5Mo–1Hf alloy is shown in Fig. 1a. The alloy is composed of three phases, which are Cr(Mo) phase, NiAl matrix and semi-continuously distributed white phase. The EDXS

143

result is as follows: Al 23.40, Cr 3.25, Ni 50.69, Hf 22.66, in which Ni:Al:Hf52:1:1 indicates that the white phase is the Heusler phase (Ni 2 AlHf) with a low level of Cr. The X-ray diffraction analysis also confirmed the expected three phases. There is no phase change after heat treatment below 1573 K which is consistent with Oh-ishi’s result [12]. After the HIPed treatment, there is little change in the eutectic morphology of NiAl and Cr(Mo). The density of the intercellular white phase that is originally located on phase boundary is greatly reduced (Fig. 1b). The EDXS result of the white phase is as follows: Al 1.18, Ni 7.49, Cr 4.15, Hf 87.18, which indicates the white phase is Hf-rich phase, whereas the Heusler phase precipitates are found rather than Hf-rich phase in the DS NiAl–Cr(Mo,Hf) composite [15], which may be attributed to Si contaminant and aged treatment of the DS NiAl–Cr(Mo,Hf) composite. A large number of precipitate phases are observed in Cr(Mo) grain (Fig. 2a), which is identified as NiAl precipitate. Similar results have been reported by Chen [20]. In addition, Cr(Mo) precipitates with long strips are also observed in the NiAl matrix, as shown in Fig. 2b. Because of the addition of Hf element, Heusler phase precipitates are found in the HIPed alloy, as shown in Fig. 2c, Heusler phase precipitates ranging from 0.1 to 0.5 mm, are entangled with dislocation. TEM results confirm the Heusler phase as determined in Fig. 3c from the selected area diffraction patterns. TEM observations also confirm that there exist numerous dislocations in NiAl matrix, while fewer dislocations are observed in Cr(Mo) phases, as shown in Fig. 2d. Additionally, the dislocations existing in the NiAl matrix are found to interact with some precipitates which was suggested to be Cr(Mo) precipitate according to EDXS results. It is believed that these precipitates should be beneficial to the improvement of the strength by dislocation pinning mechanism. According to the NiAl–Cr(Mo) phase diagram [21] and NiAl–Ni 2 AlHf pseudobinary phase diagram [22], the maximum solubility of Hf in NiAl is less than 5% at the eutectic (NiAl–Ni 2 AlHf) temperature of approximately 1623 K, the solidification process of NiAl–28Cr–5Mo– 1Hf alloy could be deduced as follows: the NiAl and Cr(Mo) eutectic reaction occurs at about 1718 K firstly,

Fig. 1. The typical microstructure of the cast NiAl–28Cr–5Mo–1Hf alloy: (a) the as-cast state; and (b) the HIPed state.

144

J.T. Guo et al. / Journal of Alloys and Compounds 343 (2002) 142 – 150

Fig. 2. (a) TEM micrographs of the fine NiAl precipitates in the Cr(Mo) grain. (b) TEM micrographs of Heusler phase in the matrix. (c) Corresponding electron diffraction pattern taken from NiAl and Ni 2 AlHf phases.

L → NiAl 1 Cr(Mo) 1 L 1

(1)

The remaining liquid phase becomes enriched in Hf, where Heusler phases precipitate with decreasing temperature L 1 1 NiAl 1 Cr(Mo) → Ni 2 AlHf

(2)

Therefore large Heusler phases exist near the NiAl and Cr(Mo) interface. Since the content of Hf is higher and the solidification rate is also higher, Heusler phase can grow into large blocks by the diffusion of Ni and Al.

3.2. Interface Fig. 3a is a HREM image viewed along [1¯ 10] common direction of NiAl and Cr(Mo) after the HIPed treatment. The image shows an abrupt transition from NiAl to Cr(Mo) lattice with no transition layer. The misfit dislocation is clearly visible. Since the point-to-point resolution of JEM-2010 high resolution electron microscope is 0.19 nm, it is unable to resolve h200j spacing of NiAl and Cr(Mo). It can also be seen from Fig. 3a that the interfaces between

NiAl and Cr(Mo) are not smooth and straight. The NiAl– TiB 2 interface [23] was atomically flat and sharp. In some cases, thin amorphous layers existed at NiAl and TiB 2 interface, which is different from the present results. Due to the very small lattice mismatch between these two phases, a semicoherent interface between NiAl and Cr(Mo) is observed. The region in which misfit dislocations are located protrudes into the NiAl side of the interface indicating a high growth rate of these cusps during the solidification process of the alloy, since the segregating of Cr and Mo atoms on misfit dislocation line can result in a reduction in the system’s energy. The curved interface implies the following: (i) the interface was not in its equilibrium state; (ii) there should be a long range stress field at the interface; and (iii) stress concentration was generated in the vicinity of the misfit dislocation. All of which would enhance segregating of impurities to the interface and drive crack tip through the interface. There is no consistent orientation relationship between NiAl and Ni 2 AlHf. Fig. 3b and c are electron diffraction

J.T. Guo et al. / Journal of Alloys and Compounds 343 (2002) 142 – 150

145

Fig. 3. (a) HREM image viewed along [1¯ 10] common direction of NiAl and Cr(Mo) for the HIPed state. (b,c) Orientation relationship existing between NiAl and Ni 2 AlHf.

patterns taken along [011] and [111] zone axes of NiAl ¯ NiAl / and Heusler phase, i.e. [111] NiAl / / [111] H , (101) ¯ /(202) H , where subscript H denotes the Heusler phase Ni 2 AlHf. There is no amorphous layer free of any interfacial phase. The misfit dislocation is clearly visible with lattice dislocation elastically softer on the NiAl side and the elastic strain was largely limited to the NiAl side of the interface. The misfit dislocation is generated to reduce the long-range stress caused by the 5% difference in lattice parameter between the phases.

3.3. Elevated temperature compressive behavior The true stress–strain curves transformed from the load– time chart for the alloy are presented in Fig. 4. From this figure, it can be seen that the strength basically decreases with decreasing strain rate at all test temperatures. The deformation behavior at slower strain rates of 2310 25 s 21 is clearly different from that at higher strain rates 2310 24 and 2310 23 s 21 . That is to say, after reaching the highest yielding point, the alloy deformed at a slower strain rate

and displayed fast strain softening, followed by continued deformation at constant stress. On the other hand, those compressed at higher strain rates usually displayed gradual softening after the highest yield point. This result is very similar to the NiAl-eutectic with reactive element system [24,25]. In most materials, the strain softening during the elevated temperature deformation process generally corresponds to the occurrence of dynamic recovery and recrystallization, it appears that the elevated deformation process is associated with those processes in the NiAl matrix. True compressive flow stress–strain rate behavior of the alloy is presented in Fig. 5, where the flow stresses were taken at 3% strain from the true stress–strain diagrams. These data were fitted by linear regression techniques to the power-law equation and temperature-compensated power-law equation, respectively,

´~ 5 As n

(3)

´~ 5 Bs n exp(2Q /RT )

(4)

146

J.T. Guo et al. / Journal of Alloys and Compounds 343 (2002) 142 – 150

Fig. 4. True compressive stress–strain diagrams for NiAl–28Cr–5Mo–1Hf tested at (a) 1273 K; and (b) 1373 K.

Where A and B are constant in s 21 , n is the stress exponent, Q is the activation energy in J mol 21 , R is the universal gas constant, and T is the absolute temperature. The results of the linear regression analysis are given in Tables 1 and 2. The correlation coefficients (R 2d ) for these fits indicate that the elevated temperature compressive behavior of the alloy can be adequately described by the power-law equations. In comparison to binary NiAl, the introduction of Cr(Mo) phase and Ni 2 AlHf precipitate decreases the stress

exponent while maintaining the value of activation energy. It is noteworthy that the addition of Cr(Mo) to NiAl fabricated by hot pressing aided exothermic synthesis (HPES) and cast technique does not change the activation energy value, however the stress exponent decreases with increasing the Cr(Mo), which shows that the addition of Cr(Mo) can reduce the stress exponent of NiAl. From Table 2, it was found that the stress exponent of cast1HIP processed NiAl / Cr(Mo)–Hf is relatively less than that of DS NiAl–28Cr–6Mo [6], which may be induced by

Fig. 5. True compressive stress–strain rate plots for NiAl–28Cr–5Mo–1Hf alloy tested at (a) 1273 K; and (b) 1373 K.

J.T. Guo et al. / Journal of Alloys and Compounds 343 (2002) 142 – 150

147

Table 1 Power-law fits of the compressive flow stress–strain rate data for the HIPed NiAl–28Cr–5Mo–1Hf alloy Alloy

NiAl–Cr(Mo,Hf), HIP NiAl–Cr(Mo) [24] NiAl–Cr(Mo) [26]

1273 K

1373 K

A

n

R 2d

A

n

R 2d

3.06310 213 8.11310 216 3.89310 211

3.44 5.32 3.31

0.996 0.993 0.993

6.07310 211 1.92310 211 1.49310 28

2.87 3.73 2.46

0.983 0.996 0.999

different fabricated processing. The creep behavior of NiAl–(Ta,Nb) alloy with Laves phase investigated by Zeumer [29] showed that the stress exponent of the alloy fabricated by powder technique is smaller than that of casting. Such difference can be attributed to their different microstructure. Whittenberger [30,31] studied extruded and directionally solidified two phase Ni–45Al–10.5Nb with partially aligned single crystalline structure and fine grain size polycrystalline band structure, respectively. They found similar behavior. It was stated that a grain boundary sliding mechanism could contribute to plastic flow process in the fine grain structure and hence lower the observed stress exponent, while boundary dependent processes would not be possible in directional solidified NiAl– NiAlNb composite. Such mechanism may be applicable to the NiAl–Cr(Mo)–Hf system. Additionally, the introduction of Heusler phase (Ni 2 AlHf) may also exert influence on the compressive behavior. However, the activation energy for the present eutectic alloy is different from those measured from other Heusler strengthened nickel aluminides and directional solidification NiAl-based alloys. For example, NiAl–Ni 2 AlHf alloy [32] possesses activation energy of 435 kJ / mol, while DS NiAl–28Cr– 6Mo alloy [6] has an activation energy of 456.6 kJ / mol, which showed that the thermal activated deformation in present alloy is easier than that in NiAl–Ni 2 AlHf and DS NiAl–Cr(Mo) alloy. In order to compare the true stress–strain rate behavior of [001]-oriented single crystal Ni–50Al [27], the directionally solidified NiAl–28Cr–6Mo [6] and HPES NiAl– Cr(Mo) [24], as well as NiAl–0.7Zr [28] produced from prealloyed powder, were also included according to quantitative power-law equations. As clearly indicated in Fig. 5, NiAl–28Cr–5Mo–1Hf is much stronger than binary single crystal NiAl, HPES NiAl–Cr(Mo) and prealloyed powder NiAl–0.7Zr, but weaker than DS NiAl–28Cr–6Mo. Its structure produced by directional solidification may result

in better strengthening effect than the disorder structure where NiAl and Cr(Mo) phases are randomly dispersed in the present alloy. Although the present alloy can’t compete with DS NiAl–28Cr–6Mo alloy, they do show evidently higher strengths than binary NiAl. The addition of Hf and Cr(Mo) are good strengthening agents with respect to NiAl. At elevated temperature, the facts that dislocation were mainly generated in the NiAl instead of Cr(Mo) grain indicate that the Cr(Mo) constituent has a higher yield than the NiAl constituent. The NiAl precipitate may also contribute to the higher strength of Cr(Mo). The Cr(Mo) precipitates and Heusler phase precipitates in the matrix also contribute to inhibit the movement of dislocation and hence increase the strength of the composite through dispersion strengthening.

3.4. Tensile behavior The elevated temperature tensile properties of the alloy are given in Table 3. For comparison, the tensile property of in-situ NiAl–TiC alloy [33], DS NiAl–Cr(Mo,Hf) alloy [15] and a DS NiAl–28Cr–6Mo alloy [6] are also given. These results show that the present alloy exhibits a higher strength than that of HPES NiAl–TiC from room temperature to 1253 K. But it did not compete with DS NiAl– 28Cr–6Mo and DS NiAl–Cr(Mo,Hf) alloy. The stress– strain response indicates that the alloy exhibits a brittle characteristic at room temperature up to 1253 K at the test strain rate (fracture strain ,5%). However, the alloy exhibits ductility behavior at elevated temperature and the stress increases rapidly and reaches a peak at a relatively small strain of 1–4% followed by strain softening. The phenomenon also existed in NiAl–Cr(Mo,Hf) alloy [34]. The fracture surface of the present alloy is shown in Fig. 6. The tensile fracture behavior of the present alloy is brittle with cleavage of NiAl at a strain rate of 1.04310 24

Table 2 Temperature-compensated power-law fits of true compressive flow stress–strain rate data for the HIPed NiAl–28Cr–5Mo–1Hf alloy Alloy

Temperature range (K)

B (s 21 )

n

Q (kJ / mol)

R 2d

NiAl–Cr(Mo,Hf), HIP NiAl–Cr(Mo) [24] NiAl–Cr(Mo), HPES [26] NiAl–Cr(Mo), DS [6] NiAl [27] NiAl–Zr [28]

1273–1373 1273–1373 1273–1373 1200–1400 1100–1400 1100–1400

8.85 9.37 2518 0.0199 0.16 2.38310 8

3.10 4.01 2.76 6.38 5.75 4.57

307.4 322.1 308.8 456.6 314.2 555

0.982 0.986 0.986 – 0.99 0.993

148

J.T. Guo et al. / Journal of Alloys and Compounds 343 (2002) 142 – 150

Table 3 The elevated tensile property of NiAl–28Cr–5Mo–1Hf1HIP alloy Alloy

Strain rate (s 21 )

T (K)

s0.2 (MPa)

UTS (MPa)

d (%)

NiAl–28Cr–5Mo–1Hf, HIP NiAl–28Cr–5Mo–1Hf, HIP NiAl–28Cr–5Mo–1Hf, HIP NiAl–28Cr–5Mo–1Hf, HIP NiAl–28Cr–5Mo–1Hf, HIP NiAl–28Cr–5Mo–1Hf, HIP NiAl / TiC [33] NiAl / Cr(Mo) [6] NiAl / Cr(Mo,Hf), HIP [15]

1.04310 24 1.04310 23 3.12310 24 5.6310 25 1.67310 24 5.6310 25 1.67310 24 3.3310 25 1.04310 23

1223 1253 1273 1273 1373 1373 1253 1253 1253

– 362 284 253 161 96 156

412 415 322 292 193 116 173 405 459

– 2.4 42 15 14 18 16 .6 2

414

Fig. 6. (a) The fractographs of NiAl–28Cr–5Mo–1Hf alloy in tension at 1223 K; and (b) the ductile fractographs of NiAl–28Cr–5Mo–1Hf in tension at 1273 K.

s 21 and below 1223 K (Fig. 6a). Beyond 1273 K, the fracture surface indicates the alloy exhibited typical ductile fracture with extensive necking showing dimples and evidence of microvoid coalescence (Fig. 6b). From the above analysis, the brittle-to-ductile transition temperature (BDTT) of the HIPed NiAl–28Cr–5Mo–1Hf alloy is higher than 1253 K and dependent on the strain rate. As to DS NiAl–28Cr–6Mo alloy, it exhibits a ductile behavior with extensive necking at a temperature higher than 1143 K and a fracture strain of over 6% [6]. The BDTT of NiAl–Ta eutectic alloy investigated by Zeumer [35] is about 1473 K at a strain rate of 10 24 s 21 and the BDTT of AFN-20 alloy [4] containing Hf, Ti is about 1423 K, which is higher than the present alloy. Whereas the BDTT of the present alloy is almost the same as the DS NiAl– Cr(Mo,Hf) composite (in the HIPed and aged condition) [15]. Therefore, the element Hf can affect the onset of climb by forming solute atmospheres around them and effectively pinning them (see Fig. 2), additional energy would be necessary to overcome this pinning effect, which leads to the higher BDTT of the present alloy. Without further studies, little can be said about the mechanism for the brittle-to-ductile transition.

onto the creep furnace, approximately half of the test sample failed. Three samples were successfully tensile creep rupture tested to failure and exhibited 70% or more strain upon failure. The time-dependence of creep strain at 1373 K is given in Fig. 7. It can be seen that these curves exhibit negligible primary creep, relatively little steady-

3.5. Tensile creep Sufficient HIPed NiAl–28Cr–5Mo–1Hf alloy was available to machine eight tensile creep samples while loading

Fig. 7. Creep curves of NiAl–28Cr–5Mo–1Hf alloy at 1373 K.

J.T. Guo et al. / Journal of Alloys and Compounds 343 (2002) 142 – 150

state creep and dominant accelerating creep stage where its time is more than 60% of total time to rupture and its strain is about 90% of total strain. The creep rates of 1373 K, 30.7 and 46.2 MPa which are calculated from the steady-state creep are 1.28310 26 and 4.1310 26 s 21 , then the stress component of the alloy at 1373 K is calculated to be about 2.7. For the HIPed NiAl–28Cr–5Mo–1Hf alloy at the experimental temperature, the SEM observation of the side surface near the fracture shows: The creep crack can initiate at the NiAl–Cr(Mo) phase boundary, where there exists a large amount of Hf-rich phase and propagate along the NiAl–Cr(Mo) phase boundary (Fig. 8a). When there is Cr, Mo, Hf solid strengthening element and dispersion Ni 2 AlHf precipitate in the NiAl matrix, the matrix has a higher strength. Sometimes Hf-rich phase can impede the crack propagation in Cr(Mo) phase (Fig. 8b). With the crack propagating along the weaker NiAl–Cr(Mo) phase boundary, Cr(Mo) phase would be dropped from the matrix, which lead to the failure of the sample in the end. There exist a lot of porosities observed in optical observation near the fracture point. The same fracture behavior is also observed in DS NiAl–Cr(Mo,Hf) [15] composite. The tensile fractographs of multiphase Ni–25Al–15Cr investigated by Chen [36] are characterized by complete debonding along b–g9 interface. So the bonding of the phase boundary in multiphase alloy is very weak, much work has to be done for the improvement of bonding of the interface. Up to now, there has been little investigation on the tensile creep of NiAl alloy because of its brittleness. Whittenberger [18] studied the tensile and compressive creep for the NiAl–Ta alloy and concluded that creep properties are reproducible and are not influenced by the testing mode. He also tested the SC NiAl–Hf alloy [37] and got the same result. In the present alloy, the stress exponent with compressive test is 2.87, while the stress component is 2.7 in tension at the same temperature. So the creep properties are not influenced by the testing method. The stress components n are believed to provide an insight into the creep mechanism. The values of 1, 3, 5 are

149

considered to represent the diffusion creep (Coble Creep), viscous glide of mobile dislocation [38] and high temperature climb controlled creep due to lattice self-diffusion, respectively. Raj et al. [39] divided the creep behavior of polycrystalline NiAl into three regions. Region 1, which occurs at low stress, is characterized by n52 and Region 2 at intermediate stress with n56.5, while Region 3 occurs at high stress where n is stress dependent. The value of n obtained in the present work, n52.7, seems to be consistent with their value in Region 1 and controlled by the viscous glide of mobile dislocation. The reason for the low stress component in the present alloy is not entirely clear, since most creep resistant, two phase alloys exhibit significantly higher stress components than the matrix phase. One possible reason is that the microstructures of the present alloy were probably not optimized. For example, the precipitating phase is not fine enough, effective strengthening would not be expected and the coarsening of the precipitate phase may result in less strengthening in the low stress-high temperature region. Another reason may be attributed to a different process. Darolia [7] has recently reported promising tensile creeprupture properties for single crystals containing Heusler precipitates. A comparison of the creep response of NiAl– Hf in both polycrystalline and single crystal form is exhibited in Fig. 9 [1]. The polycrystalline material, which is strengthened by Heusler phase precipitates, show behavior similar to other ternary alloys like NiAl–Ti [8]. However, the single crystal version is not only stronger but displays a significantly different stress dependence, which implies that the present alloy is deforming by a superposition of dislocation and grain mechanisms.

4. Conclusions The HIPed NiAl–28Cr–5Mo–1Hf alloy was composed of NiAl matrix, Cr(Mo) and Hf-rich phase. Fine Ni 2 AlHf precipitate was found in the NiAl matrix after the HIPed processing and the interface between NiAl and Cr(Mo) is atomically flat. The elevated temperature compressive behavior of the

Fig. 8. (a) SEM observation of Hf-precipitate existing between NiAl and Cr(Mo) phase boundary; (b) Evidence of crack existing in Cr(Mo) phase.

150

J.T. Guo et al. / Journal of Alloys and Compounds 343 (2002) 142 – 150

Fig. 9. Comparison of creep response of NiAl1Hf in both single and polycrystalline form at 1300 K [1].

present alloy was studied. The strengthening effect of the alloy can be attributed to the precipitation strength by Ni 2 AlHf and solid solution strengthening by Cr, Mo, Hf elements. Additionally, the strengths of the present alloy did not exceed those of the DS NiAl–28Cr–6Mo composite. The creep failure may be induced by the Hf-rich phase existing in the NiAl and Cr(Mo) phase boundary. The stress component with compressive test was almost the same as that in tension. So the creep behavior may not be influenced by the testing method.

Acknowledgements The authors gratefully acknowledge the National Natural Science Foundation of China for its financial support under contract No. 59895152. This work is also supported by the National High Technology Committee of China through contract No. 863-715-005-0030, to whom our acknowledgment should be given.

References [1] R.D. Noebe, R.R. Bowman, M.V. Nathal, Int. Mater. Rev. 38 (1993) 193. [2] D.B. Miracle, Acta Mater. 41 (3) (1993) 649. [3] R.R. Bowman, A.K. Misra, S.M. Arnold, Metall. Mater. Trans. 26A (1994) 615. [4] R.D. Noebe, W.S. Walston, in: M.V. Nathal, R. Darolia, C.T. Liu, P.L. Martin, D.B. Miracle, R. Wagner, M. Yamaguchi (Eds.), Structural Intermetallics, 1997, The Mineral, Metals and Materials Society, Pennington, NJ, 1997, p. 585. [5] H.E. Cline, J.L. Walter, E.F. Koch, L.M. Osika, Acta Mater. 19 (5) (1971) 405.

[6] D.R. Johnson, X.F. Chen, B.F. Oliver, R.D. Noebe, J.D. Whittenberger, Intermetallics 3 (1995) 99. [7] R. Darolia, J. Organic Metall. 43 (1991) 44. [8] P.R. Strutt, R.S. Polvani, J.C. Ingram, Metall. Trans. 7A (1976) 23. [9] R.S. Polvani, W.-S. Tzeng, P.R. Strutt, Metall. Trans. 7A (1976) 33. [10] M. Yamaguchi, Y. Umakoshi, T. Yamane, Philos. Mag. A 50 (1984) 205. [11] Y. Umakoshi, M. Yamaguchi, T. Yamane, Philos. Mag. A 52 (1985) 357. [12] K. Oh-ishi, Z. Horita, M. Nemoto, Mater. Sci. Eng. A239 / 240 (1997) 472. [13] R. Darolia, in: R. Darolia, J.J. Lewansdowski, C.T. Liu, P.L. Martin, D.B. Miracle, M.V. Nathal (Eds.), Structural Intermetallics, The Minerals, Metals and Minerals Society, Pennington, NJ, 1993, p. 495. [14] I.E. Locci, R.M. Dickerson, A. Garg, R.D. Noebe, J.D. Whittenberger, M.V. Nathal, R. Darolia, J. Mater. Res. 11 (12) (1996) 3024. [15] J.T. Guo, C.Y. Cui, Y.X. Chen, D.X. Li, H.Q. Ye, Intermetallics 9 (4) (2001) 287. [16] K.O. Yu, J.A. Oti, W.S. Walston, J. Metals 45 (5) (1993) 49. [17] J.A. Oti, K.O. Yu, in: R. Darolia, J.J. Lewansdowski, C.T. Liu, P.L. Martin, D.B. Miracle, M.V. Nathal (Eds.), Structural Intermetallics, The Minerals, Metals and Materials Society, Pennington, NJ, 1993, p. 501. [18] J.D. Whittenberger, R.D. Noebe, D.R. Johnson, B.F. Oliver, Intermetallics 5 (1997) 173. ¨ [19] B. Zeumer, Inr entwicklung von Laves-phasen-verstarkten NiAl ¨ andendnngen bei hohen temperaturn, Ph.D basis legierunger fur thesis, Reihe 5: Groud und werkstoffe, Nr. 374, VDI, Verlay, Dusseldorf, 1994. [20] X.F. Chen, D.R. Johnson, R.D. Noebe, B.F. Oliver, J. Mater. Res. 10 (1995) 1159. [21] H.E. Cline, J.L. Walter, Metall. Trans. 1A (1970) 2907. [22] M. Takeyama, C.T. Liu, J. Mater. Res. 5 (1990) 1189. [23] J.T. Guo, Z.P. Xing, J. Mater. Res. 12 (1997) 1083. [24] D.T. Jiang, J.T. Guo, Mater. Sci. Eng. A255 (1998) 1154. [25] C.Y. Cui, J.T. Guo, G.S. Li, Y.H. Qi, H.Q. Ye, Acta Metall. Sin. 36 (2000) 1144. [26] D.T. Jiang, J.T. Guo, G.S. Li, C.X. Shi, Acta Metall. Sin. 34 (1998) 1143. [27] C.T. Lu, J. Yang, Z. Suo, A.G. Evans, R. Hecht, R. Mehrabian, Acta Mater. 39 (1991) 1883. [28] J.D. Whittenberger, R.D. Noebe, Metall. Trans. 27A (1996) 2628. [29] B. Zeumer, W. Wunnike-Sanders, G. Sauthoff, Mater. Sci. Eng. A192 / 193 (1995) 817. [30] J.D. Whittenberger, R. Reviere, R.D. Noebe, B.F. Oliver, Scr. Metall. Mater. 26 (1992) 987. [31] J.D. Whittenberger, L.J. Westfall, M.V. Nathal, Scr. Metall. Mater. 23 (1989) 2127. [32] J.D. Whittenberger, M.V. Nathal, S.V. Raj, M. Pathare, Mater. Lett. 11 (1991) 267. [33] Z.P. Xing, J.T. Guo, Y.F. Han, L.G. Yu, Metall. Trans. 28A (1997) 1079. [34] C.Y. Cui, J.T. Guo, Y.H. Qi, H.Q. Ye, Mater. Sci Eng. A325 (2002) 187. [35] B. Zeumer, G. Sauthoff, Intermetallics 6 (1998) 451. [36] R.S. Chen, J.T. Guo, W.L. Zhou, J.Y. Zhou, Intermetallics 8 (2000) 663. [37] J.D. Whittenberger, I.E. Locci, R. Darolia, R. Bowman, Mater. Sci. Eng. A268 (1999) 165. [38] F.A. Mohamed, T.G. Langdon, Acta Metall. Mater. 22 (1974) 779. [39] S.V. Raj, S.C. Farmer, Metall. Mater. Trans. 26A (1995) 343.