Page 272 - Six Sigma Advanced Tools for Black Belts and Master Black Belts

Char Count= 0
3:5
JWBK119-17
August 31, 2006

17

Some Strategies for

Experimentation under

Operational Constraints

T. N. Goh

In quality and reliability engineering, design of experiments is well established as
an efficient empirical approach to characterization and optimization of products and
processes. However, its actual application in industry can at times be impeded by con-
straints related to physical arrangements, costs, time, or irrecoverable sample damage.
To enhance the feasibility and effectiveness of design of experiments, some practical
strategies for overcoming such difficulties are discussed in this chapter, with an ap-
plication in industrial practice as an illustration.

17.1 INTRODUCTION

Techniques of design of experiments have assumed an increasingly important role in
quality and reliability engineering in recent years owing to their efficiency in data gen-
erationandtherichnessofresultsobtainablefromthesubsequentstatisticalinference. 1
In particular, two-level experiments have formed the backbone of most studies where
the data requirement is greatly reduced via the use of fractional factorial designs.
Applications of fractional factorial designs are further stretched when those who
subscribe to Taguchi methods 2,3 routinely side-step issues of confounding of factor
interactions and use mostly saturated fractional factorial designs in an attempt to
include all relevant factors in a single experimental study.

This chapter is based on the article by T. N. Goh, ‘Some strategies for experimentation under operational
constraints’, Quality and Reliability Engineering International, 13(5), 1997, pp. 279--283, and is reproduced by
the permission of the publisher, John Wiley & Sons, Ltd
Six Sigma: Advanced Tools for Black Belts and Master Black Belts L. C. Tang, T. N. Goh, H. S. Yam and T. Yoap
C 2006 John Wiley & Sons, Ltd
257