Optimization Efficiency of Taguchi-class Experimental Design Using Conjugate Arrays

• S.-C. TSAI ^[1] ; K. M. RAGSDELL ^[2]
  1. [1] General Motor Corporation (US)
  2. [2] University of Missouri-Rolla (US)
• Localización: Journal of Engineering Design, ISSN 0954-4828, Vol. 4, Nº. 3, 1993, págs. 167-188
• Idioma: inglés
• Texto completo no disponible (Saber más ...)
• Resumen

  • To ensure market share and profit, enterprises must constantly develop high quality products at low cost. Dr Genichi Taguchi's philosophy of quality engineering and experimental design methods been proven to be cost-effective tools for industries to meet this challenge. Taguchi-class experimental design, though very cost effective, may not work efficiently in cases of significant interactions between controllable factors. If significant interaction effects between controllable factors exist, one might make incorrect judgements on the optimal setting and relative significance of the controllable factors. Therefore, the optimization efficiency will be decreased. In this paper, a new experimental layout design method, i.e. conjugate arrays, is proposed to eliminate interaction effects between controllable factors; thus, the disadvantages caused by interaction effects will be reduced to a minimum and the optimization efficiency of Taguchi-class experimental design will be increased. A press-fit example shows that conjugate arrays can eliminate interaction effects of controllable factors. As a result, one can improve the optimization efficiency of Taguchi-class experimental design.