Abstract We developed a parallel adaptive computational domain method for accelerating the simulation of solidification in two dimensions using the phase-field model. The algorithm is paralled with the MPI library based on the domain decomposition approach. We discuss the implementation details of the parallel adaptive computational domain method and the domain decomposition based strongly implicit linear solver. We examined the performance of our algorithm by comparing with the algorithm based on the fixed computational domain and show that it greatly reduces the computational cost of solving the phase-field model. We also simulate the competitive growth process of the second and ternary dendritic arms successfully.

中文翻译：

---

使用并行自适应计算域方法加速凝固的相场建模

摘要 我们开发了一种并行自适应计算域方法，用于使用相场模型加速二维凝固模拟。该算法与基于域分解方法的 MPI 库并行。我们讨论了并行自适应计算域方法和基于域分解的强隐式线性求解器的实现细节。我们通过与基于固定计算域的算法进行比较来检验我们的算法的性能，并表明它大大降低了求解相场模型的计算成本。我们还成功模拟了第二和三元树突臂的竞争增长过程。