ABSTRACT
The current work numerically investigates commercial polycrystalline Ag/17vol.%SnO2 composite tensile deformation behavior with available experimental data. Such composites are useful for electric contacts and have a highly textured initial material status after hot extrusion. Experimentally, the initial sharp fiber texture and the number of Σ3-twins were reduced due to tensile loading. The local inhomogeneous distribution of hardness and Young's modulus gradually decreased from nanoindentation tests, approaching global homogeneity. Many-scale simulations, including micro-macro simultaneous finite element (FE) and discrete dislocation dynamics (DDD) simulations, were performed. Deformation mechanisms on the microscale are fundamental since they link those on the macro- and nanoscale. This work emphasizes micromechanical deformation behavior. Such FE calculations applied with crystal plasticity can predict local feature evolutions in detail, such as texture, morphology, and stress flow in individual grains. To avoid the negative influence of boundary conditions (BCs) on the result accuracy, BCs are given on the macrostructure, i.e., the microstructure is free of BCs. The particular type of 3D simulation, axisymmetry, is preferred, in which a 2D real microstructural cutout with 513 Ag grains is applied. From FE results, Σ3-twins strongly rotated to the loading direction (twins disappear), which, possibly, caused other grains to rotate away from the loading direction. The DDD simulation treats the dislocations as discrete lines and can predict the resolved shear stress (RSS) inside one grain with dependence on various features as dislocation density and lattice orientation. The RSS can act as the link between the FE and DDD predictions.
ABSTRACT
Our work investigates the polycrystalline composite deformation behavior through multiscale simulations with experimental data at hand. Since deformation mechanisms on the micro-level link the ones on the macro-level and the nanoscale, it is preferable to perform micromechanical finite element simulations based on real microstructures. The image segmentation is a necessary step for the meshing. Our 2D EBSD images contain at least a few hundred grains. Machine learning (ML) was adopted to automatically identify subregions, i.e., individual grains, to improve local feature extraction efficiency and accuracy. Denoising in preprocessing and postprocessing before and after ML, respectively, is beneficial in high quality feature identification. The ML algorithms used were self-developed with the usage of inherent code packages (Python). The performances of the three supervised ML models-decision tree, random forest, and support vector machine-are compared herein; the latter two achieved accuracies of up to 99.8%. Calculations took about 0.5 h from the original input dataset (EBSD image) to the final output (segmented image) running on a personal computer (CPU: 3.6 GHz). For a realizable manual pixel sortation, the original image was firstly scaled from the initial resolution 10802 pixels down to 3002. After ML, some manual work was necessary due to the remaining noises to achieve the final image status ready for meshing. The ML process, including this manual work time, improved efficiency by a factor of about 24 compared to a purely manual process. Simultaneously, ML minimized the geometrical deviation between the identified and original features, since it used the original resolution. For serial work, the time efficiency would be enhanced multiplicatively.
