This is the second part of a two-part contribution on modeling of the anisotropic elastic-plastic response of aluminum 7079 from an extruded tube. Part I focused on calibrating a suite of yield and hardening functions from tension test data; Part II concentrates on evaluating those calibrations. A rectangular validation specimen with a blind hole was designed to provide heterogeneous strain fields that exercise the material anisotropy, while at the same time avoiding strain concentrations near sample edges where Digital Image Correlation (DIC) measurements are difficult to make. Specimens were extracted from the tube in four different orientations and tested in tension with stereo-DIC measurements on both sides of the specimen. Corresponding FEA with calibrated isotropic (von Mises) and anisotropic (Yld2004-18p) yield functions were also conducted, and both global force-extension curves as well as full-field strains were compared between the experiments and simulations. Specifically, quantitative full-field strain error maps were computed using the DIC-leveling approach proposed by Lava et al. The specimens experienced small deviations from ideal boundary conditions in the experiments, which had a first-order effect on the results. Therefore, the actual experimental boundary conditions had to be applied to the FEA in order to make valid comparisons. The predicted global force-extension curves agreed well with the measurements overall, but were sensitive to the boundary conditions in the nonlinear regime and could not differentiate between the two yield functions. Interrogation of the strain fields both qualitatively and quantitatively showed that the Yld2004-18p model was clearly able to better describe the strain fields on the surface of the specimen compared to the von Mises model. These results justify the increased complexity of the calibration process required for the Yld2004-18p model in applications where capturing the strain field evolution accurately is important, but not if only the global force-extension response of the elastic-plastic region is of interest.

这是对由挤压管制成的铝7079的各向异性弹塑性响应进行建模的两部分贡献的第二部分。第一部分着重于从张力测试数据中校准一套屈服和硬化功能。第二部分集中于评估那些校准。设计一个带有盲孔的矩形验证样本，以提供可实现材料各向异性的异质应变场，同时避免在难以进行数字图像相关性（DIC）测量的样品边缘附近产生应变集中。从试管中以四个不同的方向提取样品，并在样品的两面用立体声DIC测量进行拉伸测试。还进行了校准的各向同性（von Mises）和各向异性（Yld2004-18p）屈服函数的相应FEA，并在实验和模拟之间比较了整体力-延伸曲线和全场应变。具体来说，使用Lava等人提出的DIC整平方法来计算定量全场应变误差图。在实验中，样本与理想边界条件之间存在很小的偏差，这对结果产生了一级影响。因此，必须将实际的实验边界条件应用于FEA，以便进行有效的比较。预测的总体力-伸长曲线与总体测量值非常吻合，但对非线性条件下的边界条件敏感，无法区分两个屈服函数。对应变场的定性和定量研究表明，与von Mises模型相比，Yld2004-18p模型显然能够更好地描述试样表面的应变场。这些结果证明了Yld2004-18p模型在要求精确捕获应变场演变非常重要的应用中增加校准过程的复杂性是合理的，但如果仅关注弹塑性区域的整体力-延伸响应，则不是这样。