Stress state dependent damage modeling of self-pierce riveting process simulation using GISSMO damage model

Abstract

In this study, the damage model GISSMO (Generalized Incremental Stress State dependent damage MOdel) is used to describe the evolution of ductile damage and predict the onset of fracture during the self-pierce riveting of thin aluminum (EN AW-5182) sheets. By accurate prediction of the pre-damage and material separation in the simulation of joining process, the accuracy of crashworthiness simulations can be improved. Besides, over dimensioning will be reduced and more precise components will be designed. A simple approach to illustrate the separation of upper sheet in the simulation of the joining process is based on a geometric separation criterion. Such a criterion is not predictive und cannot be used in case of variations in tool configurations, sheet thickness, and material combinations. In this work, the material failure is described in dependence of the stress state. The stress state during the process simulation is studied and the variety of damage specimens is experimental examined to characterize the parameters of damage model. The inverse analysis is used to minimize the difference between the experimental and the numerical prediction concerning the load-stroke curves and the fracture zones. The numerical simulation of self-pierce riveting with semi tubular rivet using the damage model is performed and compared with experimental produced joints. It is shown that the model predicts the separation of material in the relevant region correctly. Furthermore, the results of simulation and experiment concerning the load-stroke diagram and the geometry of the joint are in good agreement.