稀有金属(英文版) 2015,34(10),725-730

收稿日期：10 April 2013

基金：financially supported by the Inner Mongolia Science and Technology Reward Foundation(No.20101707);the Inner Mongolia Natural Science Foundation(No.2013MS0804);the Inner Mongolia High School Scientific Research Foundation(No.NJZZ14056);the Inner Mongolia University of Technology Foundation(No.ZD20120015);

Hot deformation behavior and microstructure evolution of TiC–Al₂O₃/Al composites

Rui-Ying Zhang Zhi-Ming Shi Xiu-Mei Zhang

School of Materials Science and Engineering,Inner Mongolia University of Technology

Key Laboratory of Light Metal Materials

Abstract：

Hot compression behavior of TiC–Al2O3/Al composites was studied using the Gleeble-1500 system at a temperature range of 300–550 °C and at strain rate range of 0.01–10.00 s-1. The associated structural changes were studied by TEM observations. The results show that stress level decreases with deformation temperature increasing and strain rate decreasing, which can be represented by a Zener–Hollomon parameter in an exponent-type equation with hot deformation activation energy Q of 172.56 kJ·mol-1.Dynamic recovery occurs easily when strain rates are less than 10.00 s-1. Dynamic recrystallization can occur at strain rate of 10.00 s-1.

Keyword：

TiC–Al2O3/Al composites; Hot deformation; Flow stress; Microstructure;

Author： Rui-Ying Zhang,e-mail:zhang_ruiying@126.com;

Received： 10 April 2013

1 Introduction

Particle-reinforced metal matrix composites (PRMMCs) elicited considerable attention in aerospace and automobile industries because of their advantages such as high specific strength, wear resistance, and excellent high temperature creep behaviors [1–4]. However, PRMMCs often need plastic materials for their structure. Hot deformation is used more frequently in follow-up processing. Thus, hot deformation behavior of the composites must be investigated.

The hot deformation of aluminum alloys was explored in several studies [5–8]. Previous research suggested that hot deformation behavior was greatly different due to addition of alloy elements [5, 6, 9]. As a product of aluminum alloys, PRMMCs possess more complex workability because of containing reinforcements. The deformation behaviors of PRMMCs are significantly affected by both the styles and fraction of reinforcements. It was reported that the flow stress of composites increases with volume fraction of the dispersoid particles increasing [10] and the particle size decreasing [11]. The flow stress behavior of metal matrix composites was governed by two main processes: the transfer of load from the ductile matrix to the hard particles and the microstructuretransformationwhichincludesthe evolvement of the microstructure for the matrix and the particles. At higher strain rates, the behavior of the composite is similar to that of the matrix and is controlled by the grain boundary self-diffusion. At lower strain rates, however, the composite exhibits much higher apparent activation energy than that for lattice self-diffusion unlike the matrix material [12]. Accurate prediction of flow stress is more important in the development of PRMMC deformation. Many mathematical models were proposed to describe the flow stresses of PRMMCs during hot deformation [13]. So hot compression testing is becoming increasingly popular.

In this paper, the isothermal compress test was conducted using the Gleeble1500 thermal simulator with a cylindrical sample. Linear regression method established a high temperature plastic deformation model for Ti C– Al₂O₃/Al composite. Its constitutive equation was defined,which would provide a theoretical basis for the hot compression process.

Fig.1 True stress–true strain curves of composite during hot compression deformation: a ε=0.01s-1,b ε=0.10s-1,c ε=1.00s-1,and d ε=10.00s-1

2 Experimental

Ti C–Al₂O₃/Al composite with 10 wt% reinforcements was prepared by in situ method. The composite was subjected to homogenizing annealing at 470 °C for 24 h and machined in a U8 mm 9 2 mm cylinder for compression testing. Compression tests were conducted using Gleeble1500 thermal simulator at different rates (0.01, 0.10, 1.00, 10.00 s^-1) and different temperatures (300, 350, 400, 550 °C). Graphite was added between samples, and indenter was placed to prevent sample bond. Tantalum sheets and graphite were placed between samples to decrease heat transfer and the axial temperature gradient between the sample center and surface. Each sample was heated to deformation temperature at a heating rate of 10 °C s^-1and held for 180 s, then compressed at different strain rates until deformed to half of its original height. Samples were immediately quenched in water. Flow curves were obtained using data from thermal simulator during hot compression tests.

3 Results and discussion

3.1 Flow curves

Figure 1 presents a series of typical true stress-true strain curves obtained during hot compression of Ti C–Al₂O₃/Al composite at a strain rate of 0.001–10.000 s^-1and deformation temperature ranging from 300 to 550 °C. The flow stress values are dependent on temperature and strain rate. Given the same strain rate, the flow stress increases with the deformation temperature decreasing. Given the same deformation temperature, flow stress increases with strain rate increasing. However, flow stress stabilizes given increasing strain at low strain rate (0.01, 0.10, 1.00 s^-1) as seen in Fig. 1a–c. They indicate typical dynamic recovery curves. Given higher strain rates and higher deformation temperatures, flow stress curves are no longer smooth and fluctuant. When the strain rate and deformationtemperature are higher, the wave is more evident. The actual softening mechanisms will be discussed in detail in the following section by microstructural observation.

Fig.2 Evaluating value of a n1by ln ε-ln σ and b β by ln ε-σ

3.2 Constitutive equation for compression

Constitutive equations are commonly used to estimate flow stresses of materials during hot deformation. The stress– strain data obtained from compression tests given different strain rates and temperature conditions can determine the material constant of these equations. The correlation between the flow stress, temperature, and strain rate, especially at high temperatures, could be expressed by an Arrhenius-type equation. Furthermore, the effects of temperature and strain rate on material deformation behavior could be expressed by the Zener–Hollomon parameter in an exponent-type equation. The common base equations were applied as follows [14–17]:

where e is strain rate, r is flow stress, A₁, A₂, A₃, n₁, a, b, and n are the material constants. Constants (α, β, n₁) have the relation: a = b/n₁. T is absolute temperature in K. Q is deformation activation energy in k J mol^-1. R is gas constant. Z is Zener–Hollomon parameter. The power law (Eq. (1)) and the exponential law (Eq. (2)) break at high stress and low stress, respectively. The hyperbolic sine law (Eq. (3)) is a more general form suitable for stresses over a wide range. Using the natural logarithm of both sides of Eqs. (1) and (2), respectively, the following expressions could be derived:

Considering that the hot deformation process is conducted at constant temperature, n₁and β can be obtained from the slope of every single line in the ln ε-ln σ and ln ε-σ plots, respectively, by linear fitting method (Fig. 2). The correlation coefficients of the lines are more than 0.99. The value of n₁and β can be obtained for different deformation temperatures by linear fitting method. The mean values of n₁and β are 10.77 and 0.212 MPa^-1, respectively. Therefore, the value of α can be calculated using

Supposing that deformation activation energy is independent of temperature, partial differentiation of Eq. (3) at constant deformation temperature yields the following equation:

Supposingthatdeformationactivationenergyis independent of temperature, partial differentiation of Eq. (7) at constant deformation temperature can be expressed as follows:

The slope of the plot ln e_ ln sinh?ar? and ln sinh?ar? -1/T (Fig. 3) can be used to obtain the values of Q. The average slopes of these four straight lines are 7.705 and 2.71, respectively. Following Eq. (8), the value of Q can be calculated for 172.56 k J mol^-1. The hot deformation activation energy of Q is an important physical parameter serving as indicator of degree of difficulty in plastic deformation. The value of Q for 172.56 k J mol^-1is higher than that of pure aluminum (142 k J mol^-1). Thus, the existing particles increase barriers for further deformation.

By logarithmic transformation of the Eq. (4), the following expression can be written:

Fig.3 Evaluating value of a Q by fitting ln sinh?ar? -1/T and b n by fitting ln e_ ln sinh?ar?

Fig.4 Evaluating value of A3, n by ln Z–ln[sinh?ar?]

The relationship between ln Z and ln[sinh?ar?] is presented (Fig. 4). The value of A₃and n can be obtained by linear fit.

The intercept of straight line is ln A₃, and the slope of straight line is n. The values of A₃and n are 1.54 9 10¹²and 7.622, respectively. The correlation coefficient is 0.99482. The flow stress for Ti C–Al₂O₃/Al composite can be represented by Zener–Hollomon parameter Z in an exponent-type equation. Therefore, the constitutive equation that relates flow stress to Z is predicted from the following equation:

Flow stress can also be described by the Zener–Hollomon parameter Z as follows:

Fig.5 OM image a and SEM image b of Ti C–Al2O3/Al composites

3.3 Microstructural evolution

The microstructures of Ti C–Al₂O₃/Al composite in the asreceived condition are shown in Fig. 5, and from Fig. 5, it is concluded that the structure is characterized by the particles with well-proportioned distributing. The sizes of reinforcement of the prepared composites possess broad range from 0.05 to 2.00 lm.

Transmission electron microscopy (TEM) images of specimens deformed given different conditions are shown in Fig. 6. TEM observations in Fig. 6a indicate that the grains remain elongated at lower temperatures and lower strain rates. To enlarge the elongated grain, a large number of the subgrains with sub-boundaries constituted by dislocation net are present inside (Fig. 6d). The dynamic recovery microstructure is typical, based on flow curve analyses. With the deformation temperature increasing, density of dislocation in sub-boundaries decreases because of dislocation glide or climbing (Fig. 6e). During the early stage of deformation, flow stress increases rapidly due to dominant work hardening which is controlled by reinforcements blocking dislocations motion, then dynamic softening. After reaching peak flow stress, curves exhibit remarkable stability due to equilibrium of work hardening and dynamic softening (Fig. 1a–c). The deformation at elevated temperature is a competing process of dynamic softening and workhardening. Aluminum alloys possess high stacking fault energy. Recrystallization is difficult to conduct. Thus, their softening mechanisms often undergo dynamic recovery same as Al alloys. However, there is still a large amount of small size reinforcements in composite as shown in Fig. 5b. Dislocation is hindered by these small particles and is difficult to execute at higher strain rate (10.00 s^-1) and higher deformation temperature (450, 550 °C). Therefore, major dislocation accumulates around particles (Fig. 6f) and leads to storage energy increase more rapidly compared with that of pure aluminum. The required critical energy level for crystallization can be easily achieved. Typically, in the mechanisms of dynamic recrystallization nucleation, it appears that some subgrains are coarsening through the merger of other adjacent subgrains to produce high-angle boundaries [18, 19]. Elevated temperature may provide added driving pressure for boundary migration. Subgrain size increases via the migration of dislocations and the decrease of dislocation density. The subgrains merger and high-angle sub-boundaries with a certain amount of dislocations are observed in the specimen deformed at temperature of 450 °C and strain rate of 10.00 s^-1(Fig. 6b, e). The selected area diffraction patterns (SAED) show that the diffraction spot is only slightly elongated. It indicates that dynamic recrystallization (DRX) occurs during hot compression deformation. Recrystallization nucleation is also observed at temperature of 550 °C and strain rate of 10.00 s^-1(Fig. 6c). However, dynamic recrystallization consumes a large amount of energy. Continuous recrystallization is difficult to maintain using residual energy. Further plastic deformation produces hardening to store energy. Materials are hardened, and the stress increases. Thus, recrystallization occurs again. Materials are softened again, and stress decreases. Correspondingly, flow stress curves appear as waves (Fig. 1d).

Fig.6 TEM images of specimens deformed under different conditions: a, d 300 °C, 0.10 s-1; b, e 450 °C, 10.00 s-1; c, f 550 °C, 10.00 s-1

4 Conclusion

The flow stress behavior of Ti C–Al₂O₃/Al composites strongly depends on deformation temperature and strain rates. The flow stress values decrease with deformation temperature increasing and strain rate decreasing, which can be represented using the Zener–Hollomon parameter with the hot deformation activation energy of 172.56 k J mol^-1. The softening mechanism is different from that of pure aluminum due to existing reinforcements. At lower strain rates (0.01, 0.10, 1.00 s^-1), the softening mechanism indicates dynamic recovery. In contrast, the softening mechanism indicates dynamic recrystallization at strain rates of 10.00 s^-1.