Tolerance Analysis in Manufacturing Using the MMP, Comparison and Evaluation of Three Different Approaches

Abstract

In previous works, the authors have developed the Model of Manufactured Part (MMP) [1], a method for modeling the different geometrical deviation impacts on the part produced (error stack-up) in a multi-stage machining process. They also proposed different solution techniques to identify the worst case for the purpose of tolerance analysis. The first proposed solution technique consists of optimization techniques as Sequential Quadratic Programming (SQP) or Genetic Algorithm (GA) [2]. The second one combines the MMP and the Jacobian-Torsor Model [3] [4] that benefits of the interval arithmetic advantages to solve the worst case searching. The last technique uses Monte Carlo simulation to generate a population of virtually manufactured parts representative of the real produced parts [5]. This paper first reminds the MMP model and the different solution techniques. The different strategies to simulate the deviation parameters of the model are then discussed. For each of the three proposed solution techniques, its convenience and inconvenience is explored in detail. The solution techniques performances are compared from different points of view (i.e. rapidity, convergence to the global minimum, analyzed case ...) and some perspectives are presented.