Development of New Multiscale Model to Evaluate Material Performance Using Phase-Field-Crystal Method

结果表明，相场-晶体模拟可以表达温度、应变速率和晶粒度对多晶金属变形的依赖关系。为了实现相场晶体模拟的高速计算，采用GPGPU进行了常规的相场模拟。结果表明，GPU相场模拟的性能加速比约为CPU的70倍。此外，利用多相场方法进行了静态再结晶模拟，结果表明晶界能和迁移率的偏位向关系对再结晶织构的形成具有重要作用。

It has been concluded that the phase-field-crystal simulation can express the dependencies of temperature, strain rate and grain size on the deformation of polycrystalline metal. To achieve the high speed computing of phase-field-crystal simulation, GPGPU has been employed for the conventional phase-field simulation. As a result, it has been confirmed that GPU phase-field simulations achieve about 70 times performance speedup for CPU. Furthermore, by performing the static recrystallization simulations using multi-phase-field method, it has been shown that the misorientation dependencies of grain boundary energy and mobility are important to form the recrystallization texture.