以原位生成法制备的TiB₂颗粒增强铁基复合材料为研究对象，通过纳米压痕试验及有限元反演分析确定基体的幂硬化模型参数，建立二维细观真实结构模型和颗粒随机分布的体胞模型，然后模拟单轴拉伸试验，用等效宏观方法计算真应力-真应变曲线，对2种模型的模拟结果进行对比，并探讨边界条件对模拟结果的影响。结果表明：边界条件对模拟单轴拉伸时的真应力-真应变曲线影响较小；2种模型模拟得到单轴拉伸的真应力-真应变曲线差异较小，且与试验结果吻合，相对误差小于5%；真实结构模型模拟得到的弹性模量与屈服强度的误差小于体胞模型；不同模型模拟得到基体与颗粒的局部微观等效应力场及应变场有明显差异。

Taking TiB₂ particle reinforced iron matrix composites prepared by in-situ formation method as research object, the parameters of power hardening model of matrix were determined by nano indentation test and finite element inverse analysis. The two-dimensional meso real structure model and body cell model with particle random distribution were established, and then the uniaxial tensile test was simulated. The true stress-true strain curve was calculated by equivalent macro method, and the simulation of the two models were compared. The effect of boundary condition on the simulation was discussed. The results show that boundary conditions had little effect on the simulated true stress-true strain curve during uniaxial tension. The true stress-true strain curves by simulating uniaxial tension with the two models had little difference, which were consistent with test results, and the relative error was less than 5%. The error of elastic modulus and yield strength by simulation with real structure model was smaller than that with body cell model. The local microscopic equivalent stress field and strain field of matrix and particle by simulation with different models were obviously different.