Resumo

ARIAS-NAVA, Elías Heriberto; RIOS-LIRA, Armando Javier; VAZQUEZ-LOPEZ, José Antonio  e  PEREZ-GONZALEZ, Russell. Comparative Study between the Two Experimental Design Approaches Taguchi and Traditional in Presence of Control by Control Interactions. Ing. invest. y tecnol. [online]. 2015, vol.16, n.1, pp.131-142. ISSN 2594-0732.

Design of experiments plays an important role in the field of creating and innovating process and products directly in manufacturing and improving areas. There are several areas into designs of experiments; robust design is one of them. Robust parameter design is a principle that emphasize in products creation through a correct selection of values called "control" which make a product robust to the variability by the noise introducing by another factors known as "noise" factors. This article aims for a comparative study between two well-known robust design methodologies, making a special emphasis in the control by control interaction effects over optimal operating conditions. The results showed that Taguchi's crossed arrays are unable to estimate all significant terms in a model. The optimizations result concludes that the Taguchi's approach is less efficient than the traditional approach in both; maximization and minimization.

Palavras-chave : design of experiments; robust design; Taguchi's approach; traditional approach; crossed array; orthogonal array; mixed resolution.