Determination of Spot Weld Modeling Parameters from Test Data for Finite Element Crash Simulation 2004-01-0692

The authors have proposed a new formulation to characterize the mechanical properties of spot welds under dynamic loadings including separation. In this paper, the authors primarily discuss a systematic procedure to determine the parameters of the proposed spot weld model from test data using a Design of Experiment (DOE) approach and statistical analyses. All analysis pertaining to the spot weld modeling under impact loading has been performed using RADIOSS, a commercially available explicit FE crash solver.

In this study, the spot weld connection was modeled using a two-node beam-type spring element with 6 DOF at each node, and the sheet metal was modeled using a four-node shell element. The main objective was to develop a spot weld modeling methodology that is accurate and robust enough to be used in a full vehicle model which is composed of hundreds of different parts and will be crashed under different test conditions. Finite element models for spot-weld coupon geometry have been constructed for 190 different combinations of materials, gages, coatings, loading angles, and speeds. Using the simulation results of these models, response surfaces were developed for spot weld loading functions and separation criteria parameters in terms of material and geometric properties. Two different separation criteria, one based on peak force and another based on internal energy, were evaluated to determine suitability of these criteria for spot weld separation modeling. Two different approaches, interpolated and non-interpolated, were investigated in this study.

The results were compared with component test data. The non-interpolated spot weld modeling methodology was found to be better than interpolated approach for spot weld connections covering a wide range of materials, gages, loading conditions and speeds. The interpolated spot weld modeling methodology needs to be investigated further.