Modeling the crashworthiness analysis of functional graded strength thin-walled structure with phenomenological GISSMO model

In this paper, the phenomenological generalized incremental stress–strain model (GISSMO) is investigated to predict the damage-mechanical behavior of the quenched boron steel. Four fracture strain vs . stress triaxiality hyperbolic curves of quenched boron steels with different hardness are established employing a hybrid experimental–numerical method, and the corresponding damage parameters m and n are further identified by the FE inverse method using a multi-objective particle swarm optimization algorithm (MOPSO). Subsequently, a tailored partial quenching process is employed to fabricate the two-segment functional graded strength (FGS) and the uniform strength (US) hat-shaped structures, and their experimental and simulated crashworthiness is further investigated to validate the effectiveness of the established GISSMO model. Compared to the simulation with the failure plastic strain and the simulation without damage model, simulated results with GISSMO model of US structure match the best with the experimental result in terms of both deformation mode and force vs. deformation curve. And the error of SEA is no more than 6%. Results of the FGS structure also demonstrated a good consistency between simulation with GISSMO model and the experiment. In terms of SEA and CFE , the predicted accuracy was improved from 13.437%, 23.404% to 9.140%, 6.687%, respectively by adding the GISSMO model in simulation. Thus, it indicates that modeling with the GISSMO model is highly adequate to predict the crashworthiness of the quenched boron steel for both US and tailored FGS structures. • A feasible process is developed to fabricate a tailored channel structure. • A 3D hardness/QT-plastic strain–stress surface for quenched 22MnB5 is established. • GISSMO model is successfully identified by a hybrid experimental–numerical method. • Numerical accuracy of the crushed channels was increased by using GISSMO model.