Data-driven parameter identification to simulate progressive damage in fiber reinforced laminates under low velocity impact

Abstract

Genetic algorithms are applied successively to identify input parameters for simulating intralaminar and interlaminar progressive damage in IM7/8552 carbon fiber reinforced polymers in finite element analyses. The simulation of various low velocity impact tests validates the data-driven parameter identification and the capability to account for the evolution of tensile, compressive and interface damage. It is shown that the efficient modeling technique can simulate impact tests in less than one hour on conventional computers with realistic representations of delamination. The objective and automated estimation of all damage input parameters enhances transparency and repeatability of finite element simulations of composites.