Without considering the evolution of hardening parameter during ECAE, the predicted stress-strain curves of AZ31B ECAE processed samples up to four passes are shown in Figure 6.3. Those samples processed up to one and two passes have non-uniform bimodal microstructures consisting of large and fine DRX grains. Since, the dislocation hardening model used in the prediction does not account for such
non-uniform microstructures, the response of those samples are not predicted. Four-pass samples have more uniform microstructures as discussed in Chapter V.
It is important to note that the evolution of hardening parameter that is expected during ECAE, was not taken into account in the prediction of four-pass samples response. Exactly the same hardening parameter used for as received sample, including critical resolved shear stresses and dislocation densities, have been used for the modeling of stress-strain response of samples processed up to four ECAE passes. Despite this fact, the model was able to predict to some degree the hardening behavior of these samples along the three directions.
6.4 Discussion of Results
In order to find out the model parameters, tension and compression tests were carried out along the two directions of the hot rolled plate. These tests however, were performed only at room temperature. Some parameters in this model are temperature and strain dependent. Therefore, tests at several temperatures and strains are required to obtain the right parameters. These tests have not been done in this dissertation due to the time limitation. Despite this fact, the model was able to predict the hardening response.
111
Figure 6.3 Experimental (lines) and predicted (symbols) of AZ31B samples ECAE processed up to four passes following route (a) A (4A-I), (b) C (4C-I), (c) E (4E-I) and (d) BC (4BC-I).
6.4.1 Governing mechanisms during through-thickness and in-plane monotonic testing of hot rolled plate
As discussed in Chapter IV, the tension compression asymmetry of the AZ31B Mg hot rolled plate is high along the two plate directions: through-thickness and in-plane, where the stress differential (SD%) values were -78% and 73% along IP and TT directions, respectively. Figure 6.2 shows the predicted relative activities during tension and compression of the rolled plate. Tensile twinning is activated during stage II and III of deformation in TTT and IPC. The twin fraction reaches 90% of the matrix in both cases. In TTC and IPT, however, negligible amount of tensile twinning is observed in Figure 6.2.a and 6.2.d. Deformation during TTC is accommodated by basal and pyramidal slip systems. In IPT, deformation is accommodated by mainly prismatic and basal slip systems.
Figure 6.2.c and 6.2.f shows the predicted relative activities inside the twins. It is very obvious, that there is a delay on the start of activities of slip systems inside the twins. This delay governs the size of the plateau region observed at stage II. Klimanek and Pötzsch [36] pointed out that at strains below 8%, while profuse
{ }
1012 twinning takes place, there is little dislocation accumulation. In other word, twinning delays the onset of gross dislocation plasticity [114]. It is well known that the tensile twinning reorient the twinned region by 86.4°. This explains the difference in slip systems activated in the matrix and twins. During TTT, for example, plastic deformation inside twins is accommodated by prismatic slip similar to IPT. The same thing can be said113
about deformation inside twins during IPC, where only basal and pyramidal are active similar to deformation during TTC.
6.4.2 Deformation mechanisms of ECAE processed samples
Figure 6.4 shows the predicted relative activities during tension along the FD of the ECAE processed AZ31B Mg samples up to four passes. The response of these four samples along this direction is similar as shown in Figure 5.4 and 6.3. The samples of 4E-I and 4C-I are, however, softer than 4A-I and 4C-I samples. The measured and predicted basal pole figures of the four samples are shown in Figure 5.3. Basal poles of most grains in all four samples are oriented with the FD. Therefore, these samples are expected to deform mainly by slip systems in tension along FD similar to IPT of hot rolled plate. This deformation behavior is seen very clearly in 4A-I and 4C-I samples where plastic deformation is accommodated mostly by prismatic slip. In 4E-I and 4BC-I samples, the basal poles are more randomly distributed with respect to FD, although, most of them are oriented perpendicular to this direction. In fact, basal poles of significant amount of grains in these two samples have basal poles parallel to FD. This triggers some activity of tensile twinning during tension along FD as can be seen in Figure 6.4.c and 6.4.d.
1.0
Figure 6.4 The predicted relative activities of deformation modes during tension along the FD of the AZ31B ECAE processed up to four passes following route (a) A (4A-I), (b) C (4C-I), (c) E (4E-I) and (d) BC (4BC-I).
The flow direction of the processed samples is of great interest because it is the strongest direction in tension (see Table 5.1 and Figure 5.4) and these samples demonstrate the highest T/C asymmetry along this direction as can be seen in Figure 5.9.
As can be seen in Figure 6.4.a and 6.4.b, the activity of basal slip system is much lower than that of prismatic slip. Since basal is the easiest slip system, the material demonstrated higher tensile yield strength along FD due to the more restricted basal
115
orientation which is being perpendicular to the testing direction (FD). When the basal activity increases, as seen in Figure 6.4.c and 6.4.d, due to more favorable basal orientation, the tensile yield strength decreases.