Tolerance optimization of a lower arm by using genetic algorithm and process capability index

• Published in: International journal of precision engineering and manufacturing Vol. 15; no. 6; pp. 1001 - 1007
• Main Authors: Lee, Kwang-Ki, Ro, Yun-Cheol, Han, Seung-Ho
• Format: Journal Article
• Language: English
• Published: Springer Korean Society for Precision Engineering 01.06.2014
• Subjects: Engineering; Industrial and Production Engineering; Materials Science; Process capability index; Lower arm; Genetic algorithm; 2nd Order Taylor Expansion; Response surface model; Tolerance optimization
• ISSN: 2234-7593; 2005-4602
• DOI: 10.1007/s12541-014-0428-4

Tolerance optimization that considers variances of design variables should be performed before beginning the manufacturing process from a cost-effective perspective in the design process. The authors used a genetic algorithm and the process capability index (Cpk) to solve the robust objectives and probability constraints and to formulate a constrained optimization problem into an unconstrained one. The design space provided by the Cpk-values of weight and stress on the lower arm of a vehicle’s suspension was explored by using the central composite design method and the 2 nd order Taylor series expansion. The optimal solutions were found via the genetic algorithm, in which the Cpk-values took into account the variances occurring in a design variable’s tolerances. The mean and standard deviation of Mass and Smax were predicted by using the 2 nd order Taylor series expansion and the 2 nd order polynomial response surface models generated from the central composite design method. The Cpk of Mass and Smax were calculated, where the Pareto set was generated by maximizing the Cpk-values via the MOGA (Multi-Objective Genetic Algorithm). From the Pareto set, optimal alternatives were selected and verified by simulated results from FE (Finite Element) analysis and Monte-Carlo simulation.