College of Materials Science and Engineering, Chongqing University, China

Speech Title: Simulation of Multi-system microstructure evolution by coupling phase field and tensor decomposition techniques
Abstract: Phase-field method is a versatile and powerful tool for simulating the microstructural evolution in a wide variety of material processes in multi-systems. However, in many cases, the input data for the multi-system is with large scale and the calculation of the thermodynamic data involves a lot of non-linear calculations. Hence, the normal simulation for the multi-system will require large memory storage and long processing time. To make the multi-system phase field simulation practicable, a tensor decomposition approach (Vervliet et al. 2014) has been imported to treat the thermodynamic data of the multicomponent system.
During our calculation, we found that the tensor fitting method can well reproduce the thermodynamic and kinetic data, as well as be further applied to phase field simulations.
Furthermore, the decomposed data were applied in phase field simulation of the homogenization processes. Two cases for the homogenization processes were studied in this work.
It is shown the results obtained from Dictra simulation and that from phase field simulation using original thermodynamic functions are consistent. The phase field simulations using tensor fitting functions are also identical with that obtained using original thermodynamic functions. We have found that the simulation errors using 5 rank tensor fitting functions are already acceptable in most industrial cases, while 7 and 10 rank tensor fitting functions lead to more accurate but similar results.
Therefore, the phase field simulation coupled with tensor fitting functions is quite practicable and can be extended to the research and industry application.

Keywords: Phase-field method, Multi-system, Tensor decomposition