preview

Advanced Methodology With Response Surface Methodology Essay

Good Essays

Robust design (RD) based on the concept of building and improving quality into a design has been paid much attention by researchers and practitioners for years, and a number of approaches have been introduced in the literature. There exist many research attempts to integrate RD methodology with response surface methodology (RSM). The three-level full factorial design embedded higher-order polynomial models may be more useful in order to obtain better RD solutions than the second-order model. In this paper, we propose a dual response surface approach based on higher-order polynomial models in order to optimize the process bias and standard deviation responses at the same time while the process mean is restricted by the lower and upper specification limits. We also discuss why the three-level full factorial design is preferred over the other second-order designs, such as the traditional central composite design, with higher-polynomial models. In addition, the significance levels of the model fittings are sought out by the model selection procedure. A numerical example is given to illustrate the superiority of the proposed approach, as compared with the traditional methods.

In many quality engineering applications, the objective is to determine the relationship between the controllable input variables, xi, and a quality characteristic, y. Several designs are available in the literature for fitting first-, second- or higher-order polynomial models over spherical or cuboidal

Get Access