Abnormal Subgrain Growth by Monte Carlo Simulation Based on Hot-rolled AA5005 Aluminum Alloy Texture

The subgrain structure of hot rolled aluminum alloy AA 5005 has been characterized on as-received samples using Electron Backscatter Diffraction (EBSD). Based on the OIM scans of RD-ND and TD-ND, 3 dimensional microstructures of subgrains are built up using the 3D Microstructure Builder, which is a method for developing statistically representative digital representations of microstructures. Following the generation of microstructure, different textures were fit to these reconstructed 3D microstructures, based on individual components such as Brass and S textures. For this study, the Brass texture was chosen as an exemplary case. Monte Carlo simulation was used to model subgrain coarsening and visualization was a key to detecting abnormal grain growth. The main objective is to understand the circumstances under which we can expect abnormal (sub-)grain growth to lead to nucleation of recrystallization. Introduction Considerable effort has been expended to understand the initial stages of recrystallization since they dominate the final outcome of the process. The aim of this recrystallization nucleation study is to be able to predict both the size and orientation of the recrystallized grains. Mastering the parameters that control this process will allow us control recrystallization more efficiently. Typically, in an alloy of medium or high stacking fault energy, the dislocations are arranged after deformation in the form of a three dimensional cell structure [1]. During recovery, the dislocations cells rearrange (polygonize) into subgrains via annihilation of dislocations, which may have already taken place, especially at high deformation temperatures. Recovery leads to the formation of low angle boundaries. The subgrains are small regions which are delineated by these low angle grain boundaries. There are 3 factors that promote the formation of a subgrain structure during deformation, namely, high stacking fault energy, low solute content, large strain and high temperature of deformation [1]. Long range orientation gradients may be present in the material, which is also important for the occurrence of abnormal growth [2]. However, besides the interest in the subgrain structure itself, much attention also has been given to subgrain coarsening. Over 50 years ago, it was first proposed that subgrain coalescence is a nucleation mechanism of recrystallization by Cahn [3]. The nucleation mechanism of recrystallization has been an issue ever since. It is accepted that from a thermodynamic point of view, it is unlikely that the new recrystallized grains are created through random thermal fluctuations due to the low driving force and large interfacial energy of a high angle grain boundary, which is an essential factor in the recrystallization process [1]. Therefore, the nucleation of recrystallization involves the formation of a small nucleus that is a small volume of relatively perfect material pre-existing in the deformed microstructure [1]. This nucleus must be at least partly bounded by a high angle grain boundary [4]. A detailed analysis of the recrystallization process in an Fe-Ni alloy done by EBSD in the FEGSEM showed that the nuclei mainly develop from shear bands or near grain boundaries, or more frequently at the intersection between them [5]. In this study, they observed that the nuclei were highly misoriented from their neighborhood. However, this nucleation mechanism has not been thoroughly investigated. The most recent work on this issue is done by Holm [6] using Monte Carlo simulation, which concluded that abnormal growth needs to have relatively high mobility boundaries available to the system. Experimental Investigation The aluminum alloy used in present study is a hot rolled alloy 5005 that was supplied by the Alcoa Technical Center, Pittsburgh. The chemical composition of this alloy is given in table 1. The asreceived sheets are about 3.5mm thick. Small pieces were machined out of the sheet with 20mm × 10mm × 6mm. The samples were ground through a series of 400, 600, 800 and 1200 grit papers with water, followed by a 1 μm diamond polish on a nap cloth, and finally electropolished using an A2 Perchloric acid based electrolyte. The 2 orthogonal sections were scanned in an FEI XL-40 FEG Scanning Electron Microscope. EBSD maps were obtained using the TSL software (TSL-EDAX). Table 1. Chemical composition of AA 5005 in wt%. Element Si Fe Cu Mn Mg Zn Al Composition 0.30 0.70 0.20 0.20 1.10 0.25 97.00 Statistical Representation of 3D microstructure For simulating realistic microstructures, both the subgrain shape statistics and the texture need to be incorporated into the 3D digital microstructure as illustrated in the following figure. Figure 1. A schematic of the process to construct a 3D microstructure based on 2 orthogonal EBSD observation areas. The microstructure produced by fitting grain shapes to the observed subgrain structures is shown on the right using OpenDX. In order to make the 3D microstructure more accurate, subgrain shape statistics were obtained in the form of aspect ratios using the TSL software, figure 2. Both these sections have peaks at 0.5 and 0.6. The mean aspect ratio is 0.5. This means that the subgrains are half as long parallel to the ND in relation to the TD and RD directions. R e e2 e Orientation Imaging Microscopy scan areas