A numerical investigation on uniaxial compression tests is performed to highlight the heterogeneous nature of the deformation process. Resulting errors on the material behaviors are deduced from the obtained force versus displacement data with the assumption of homogeneous deformation. The numerical study considers a range of strain rates (from 0.01 to 0.5 \(s^{-1}\)), Coulomb friction coefficients (up to 0.3), and elasto-viscoplastic power law behaviors. The heterogeneous nature is characterized in terms of sample shape, strain, and strain rate heterogeneities. The results show that the final shape of the sample at a given macroscopic strain is influenced not only by the friction coefficient but also by the material properties. The levels of strain and strain rate heterogeneities in the samples can be significant in some conditions, leading to large errors when exploiting the force versus displacement data with the hypothesis of homogeneous strain. The estimations of the strain rate sensitivity parameters are not significantly affected by the strain heterogeneities, but the errors on the strain hardening parameters can be as large as 40 %. The apparent strain hardening parameter appears to be artificially strain rate sensitive. Being systematically lower than the material strain hardening parameter, when measured at lower strain rates, this underestimation will induce a systematic error in the determination of material properties and should be taken into consideration.

对单轴压缩试验进行了数值研究，以突出变形过程的非均质性。假设均匀变形，从获得的力与位移数据推断出材料行为的误差。数值研究考虑了一系列应变率（从 0.01 到 0.5 \(s^{-1}\)）、库仑摩擦系数（高达 0.3）和弹粘塑性幂律行为。异质性的特征在于样品形状、应变和应变率异质性。结果表明，给定宏观应变下样品的最终形状不仅受到摩擦系数的影响，还受到材料特性的影响。在某些条件下，样品中的应变和应变率异质性水平可能很大，在均匀应变假设下利用力与位移数据时会导致较大误差。应变率敏感参数的估计不受应变异质性的显着影响，但应变硬化参数的误差可高达40%。表观应变硬化参数似乎人为地对应变率敏感。当在较低应变率下测量时，系统地低于材料应变硬化参数，这种低估将在确定材料性能时引起系统误差，应予以考虑。