Efficient simulation of solidification microstructures of a binary alloy under non‐isothermal conditions based on adaptive octree grids

Performing microstructure simulation of binary alloy in non‐isothermal condition always suffers from the decoupling problem of temperature field with concentration field and phase field due to different diffusion velocities of temperature and solute. Adaptive octree grids are introduced into phase field model to explore the microstructure simulation in non‐isothermal. The appropriate refining and coursing model based on octree system and phase field method is proposed. A solution is also put forward to work out the above decoupling problem by solving both diffusion equations on the different layer of the octree grids. Microstructure simulations are performed for the binary alloy Ni‐Cu, using phase field model, adaptive octree grids and parallel technique. The simulation results are discussed. Firstly, results of the phase field, temperature field and concentration field are discussed and are found to conform to the basic solidification principle. Secondly, the calculation time and resource consumption of uniform system and adaptive octree system are also compared. At last, a long‐time simulation experiment of large scale has been carried out. It illustrates that adaptive octree grids owns great applicability in microstructure simulation due to the computational efficiency on the premise of accuracy. The current research introduces adaptive octree grids into phase field model to explore the microstructure simulation under non‐isothermal conditions. The phase field model, the adaptive octree grids and a parallel technique are used to simulate the dendrite growth of the binary alloy, Ni‐Cu. The results of the phase field, temperature field and concentration field are discussed. The calculation time and the resource consumption of a uniform system and an adaptive octree system are also compared in this research.