Simulation of multibody systems with the use of coupling techniques: a case study

Abstract

Simulation coupling (or cosimulation) techniques provide a framework for the analysis of decomposed dynamical systems with the use of independent numerical procedures for decomposed subsystems. These methods are often seen as very promising because they enable the utilization of the existing software for subsystem analysis and usually are easy to parallelize, and run in a distributed environment. For example, in the domain of multibody systems dynamics, a general setup for “Gluing Algorithms” was proposed by Wang et al. It was intended to provide a basis for multilevel distributed simulation environments. The authors presented an example where Newton’s method was used to synchronize the responses of subsystem simulators. In this paper, we discuss some properties of a simplified iterative coupling scheme, where subsystems’ responses are synchronized at discrete time points. We use a simple multibody model to investigate the influence of synchronization parameters on computations. We also try to provide explanation to the oscillatory behavior of the solutions obtained from this method.