In this study, we aimed to predict various impact failure behavior of cylindrical CFRPs with multiscale analyses using individual properties of fiber, interface, and resin including strain-rate dependence. We conducted compressive impact tests for the cylindrical CFRPs using split Hopkinson pressure bar (SHPB) method. Tests were carried out under four different compressive condition; radial direction and axial direction compressive tests for CFRP-H in which the fiber was aligned in the hoop direction and CFRP-L in which the fiber was aligned in the axial direction. Strengths in Hashin damage theory used in the macroscale analyses were determined by microscale analyses. Failure envelope in transverse direction and fiber-axial shear strength were obtained by 3D periodical unit cell analyses with considering the strain-rate dependence of resin strength. In addition, tensile and compressive strengths in fiber-axial direction were determined using the simultaneous fiber failure model and the fiber microbuckling model, respectively. Results of macroscale SHPB analyses were compared to test results. In terms of fracture process, analysis results did not match to test results because we conducted simulations without considering fracture energy. In terms of failure strength, failure load and failure occurred point, we could confirm rather good agreement between analysis results and test results.

在这项研究中，我们旨在使用纤维，界面和树脂的各个特性（包括应变率依赖性），通过多尺度分析来预测圆柱形CFRP的各种冲击破坏行为。我们使用霍普金森分体式压力棒（SHPB）方法对圆柱形CFRP进行了压缩冲击试验。试验是在四种不同的压缩条件下进行的。纤维在环向排列的CFRP-H和纤维在轴向排列的CFRP-L的径向和轴向压缩测试。通过微观分析确定了在宏观分析中使用的Hashin损伤理论的优势。考虑到树脂强度的应变率依赖性，通过3D周期晶胞分析获得横向破坏包络线和纤维轴向剪切强度。另外，分别使用同时纤维破坏模型和纤维微屈曲模型确定了纤维轴向拉伸强度和压缩强度。将大型SHPB分析的结果与测试结果进行比较。就压裂过程而言，分析结果与测试结果不符，因为我们进行了模拟而未考虑压裂能量。从破坏强度，破坏载荷和破坏发生点的角度，我们可以确定分析结果与测试结果之间的一致性很好。分别使用同时纤维破坏模型和纤维微屈曲模型确定纤维轴向拉伸强度和抗压强度。将大型SHPB分析的结果与测试结果进行比较。就压裂过程而言，分析结果与测试结果不符，因为我们进行了模拟而未考虑压裂能量。从破坏强度，破坏载荷和破坏发生点的角度，我们可以确定分析结果与测试结果之间的一致性很好。分别使用同时纤维破坏模型和纤维微屈曲模型确定纤维轴向拉伸强度和抗压强度。将大型SHPB分析的结果与测试结果进行比较。就压裂过程而言，分析结果与测试结果不符，因为我们进行了模拟而未考虑压裂能量。从破坏强度，破坏载荷和破坏发生点的角度，我们可以确定分析结果与测试结果之间的一致性很好。