Usage of a Rate-Dependent, Elasto-Plastic Cohesive Zone Mixed-Mode Material Model for Spot-weld Modeling

Currently, there are several material models implemented in LS-DYNA which are more or less capable to simulate a spot weld behavior under crash conditions. The material parameters for crash simulations are determined by comparing the results of tensile, shear and peel tests specimen with an equivalent FE-model. By varying the spot weld material parameters, the simulation should reproduce the test and its results as accurate as possible. Important parameters that should be reproduced by the simulation are the displacement and the force at which the spot weld fails. With finely discretized models that divide the spot weld into several heat affected zones to consider as many deformation and failure phenomena as possible, the tests and their results can be rebuild quite precisely. The main disadvantage of these detailed models is that they require the use of approximately 50.000 solid elements for the spot weld only. This leads for explicit time integration schemes to very small time steps and therefore to computation times that are far too high for crash simulations. For this reason less detailed (i.e. coarse) models are needed to describe the behavior of spot welds during crashworthiness simulations. Spot welds are herein modeled with beam or solid elements that connect the two coarsely meshed flange partners modeled by shell elements. Usually *MAT_SPOTWELD_DAIMLERCHRYSLER (*MAT_100_DA) is used to describe the spot weld’s constitutive behavior. Using *MAT_100_DA, the parameters that define the onset of failure can be the effective plastic strain or a criteria incorporating a combination of axial, shear and bending stresses. Since spot welds in automotive structures not only show rate-dependent, elasto-plastic material behavior but also exhibit different stiffness properties in uniaxial and shear loading, the so called *MAT_COHESIVE_MIXED_MODE_ELASTOPLASTIC_RATE (*MAT_240) material card offers new possibilities to define the point of failure by using the energy release rates GIc and GIIc and the yield stresses in mode I and II fracture as failure criterion independently. Additionally, the rate-dependency in deformations can be considered using either a linear logarithmic or quadratic logarithmic model. In the present paper, the material behavior of a spot weld modeled with one solid element using *MAT_100 and a spot weld modeled with one, four or eight solid elements using *MAT_240 are compared with the results of a very detailed simulation describing the spot weld’s behavior during tensile, shear and peel test. To predict the spot weld’s behavior under various geometrical conditions, its alignment with regard to the flange mesh has been modified in steps of 30°, 45° and 60° from its original central position. Finally, an adaptive mesh is used for the flanges that are bonded by the spot weld to analyze the mesh size influence on the spot weld’s failure behavior.