Comparison of Semirecursive and Subsystem Synthesis Algorithms for the Efficient Simulation of Multibody Systems

A great variety of formulations exist for the numerical simulation of rigid-body systems, particularly of medium-large systems such as vehicles. Topological formulations, which are considered to be the most efficient ones, are often cumbersome and not necessarily easy to implement. As a consequence, there is a lack of comparative evidence to support the performance of these formulations. In this paper, we present and compare three state-of-the-art topological formulations for multibody dynamics: generalized semirecursive, double-step semirecursive, and subsystem synthesis methods. We analyze the background, underlying principles, numerical efficiency, and accuracy of these formulations in a systematic way. A 28-degree-of-freedom, open-loop rover model and a 16-degree-offreedom, closed-loop sedan car model are selected as study cases. Insight on the key aspects toward performance is provided.