An Enhanced Static Reduction Algorithm for Predictive Modeling of Bolted Joints

Abstract

Bolted joints in assembled structures are responsible for much of the structure’s damping and gives rise to nonlinear behavior. While a bolted joint can presumably be modeled in commercial FE codes, even quasi-static analysis is extremely expensive due to the level of detail required to capture micro-slip. This work explores the application of a new static reduction method to reduce the computation costs associated with analyzing built-up structure. This novel method applies a Gauss-Seidel algorithm to a model that has been statically reduced to retain only the DOF at the contact interface. To validate the approach, a simple contact problem was first considered that examined contact between a deformable block and a rigid surface. Coulomb friction was used at the interface and the results were verified by examining the load dependence of contact area and model nonlinearities. Subsequently, a contact problem between two deformable bodies in 2D was considered which modeled two beams in a sandwich configuration. In all cases it was found the static reduction method’s resulted in dramatic computational savings while providing good accuracy as compared to simulations in commercial finite element software.