Inertia-based identification of elastic anisotropic properties for materials undergoing dynamic loadings using the virtual fields method and heterogeneous impact tests

The characterization of material anisotropic properties at high strain rates remains a challenging work due to the limitations of the conventional dynamic testing methods. In this study, a novel inverse identification strategy is proposed based on the dynamic virtual fields method (VFM) to enable a simultaneous identification of the anisotropic constitutive parameters from a single high speed impact test of a non-uniform specimen. This is implemented by taking advantage of the heterogeneous full-field strain and inertial acceleration data, in which the restrictions of one-dimensional stress wave propagation and homogeneous deformation states for the conventional methods are released. Specifically, the dynamic VFM-based identification algorithm is developed first. Then, several virtual impact tests with different specimen configurations are designed to provide distinct stress/strain distributions. Next, the simulated full-field strain and acceleration data are utilized to extract the input target orthotropic stiffness components. The results show that the identification accuracy of the anisotropic parameters is highly dependent on the heterogeneity of the stress/strain distributions. Also, the identification results are unstable for different time steps but can be significantly improved using a minimization algorithm on multiple time steps. Finally, the influences of noise, impact speed and loading mode are analyzed in the sensitivity study. [Display omitted] •The inertia-based identification algorithm for the orthotropic elasticity is developed using the dynamic virtual fields method.•Heterogeneous impact tests are designed for a simultaneous identification of the multiple anisotropic constitutive parameters.•The identification results are unstable for different time steps but can be significantly improved using a minimization algorithm on multiple time steps.•The proposed identification method remains robust when the factors such as noise, impact speed and loading mode are changed reasonably.