Steels are the most commonly used multi-phase materials in the industry, and their mechanical behaviors depend on the microstructure, composition, and phase fractions. Generally, the material behaviors need to be measured by experiments like a tensile test or split Hopkinson bar test, which is very time-consuming and expensive. Once the heat treatment and phase fractions are changed, it needs to be tested again, and, to avoid this, a better method is required to obtain the material behavior quickly and easily. In this study, a novel multi-scale approach is described to predict the material behaviors of multi-phase steels based on the phase fractions. A crystal plasticity finite element method is used to obtain the material behavior of each phase at a micro-scale with elevated strain rates, which is validated with experimental data or theoretical models at static or quasi-static conditions. Then a homogenization procedure with the rule of mixture method, which is based on the phase fractions measured from the microstructure characterization, is used to get the macro-scale constitutive behavior, and it is then implemented into the commercial software Abaqus/Standard to simulate the process of tensile test and compared with the experimental data. Good agreements are obtained between simulation and experimental results.

钢是工业上最常用的多相材料，其机械性能取决于微观结构，组成和相分数。通常，需要通过诸如拉力试验或分裂霍普金森棒试验的实验来测量材料的性能，这是非常耗时且昂贵的。一旦改变了热处理和相分数，就需要再次测试，为避免这种情况，需要一种更好的方法来快速，轻松地获得材料性能。在这项研究中，描述了一种新颖的多尺度方法，以基于相分数预测多相钢的材料性能。使用晶体可塑性有限元方法来获得微观相中应变率提高的每相材料的行为，在静态或准静态条件下已通过实验数据或理论模型进行了验证。然后使用基于混合法则的均匀化程序，该程序基于从微观结构表征中测得的相分数，获得宏观的本构行为，然后将其实施到商业软件Abaqus / Standard中，以模拟拉伸试验过程并与实验数据进行比较。在仿真和实验结果之间取得了良好的一致性。然后将其实施到商业软件Abaqus / Standard中，以模拟拉伸测试过程并与实验数据进行比较。在仿真和实验结果之间取得了良好的一致性。然后将其实施到商业软件Abaqus / Standard中，以模拟拉伸测试过程并与实验数据进行比较。在仿真和实验结果之间取得了良好的一致性。