Automated Assembly Stability Evaluation by Theoretical and Physics Simulation Methods

One of the challenges in achieving an optimal assembly process plan via computational methods is the need to automatically estimate the stability of every subassembly in the process. A highly stable assembly plan can reduce the assembly time and the usage of fixtures. In this paper, to automatically estimate the assembly stability with high accuracy, we propose a novel approach, which divides the evaluation of assembly stability into two sections: theory-based evaluation and simulation-based evaluation. Theory-based stability is evaluated using degrees of freedom and tip-slide difficulty for each part in the assembly. For simulation-based stability evaluation, a physics engine is used to simulate the spatial kinematic behavior. Both approaches function with any given tessellated CAD model and can thus be used to optimize the stability factor of automatically created assembly plans. In this paper, the two methods are described along with a comparison on relatively realistic mechanical part shapes.