Qualtiy improvement and design of experiments

ABSTRACT Today, quality is vital for the companies that aim to improve its' competitive position. The cost of delivering a defective product to a costumer is very expensive. Thus, the quality characteristics of products and the production processes must be controlled continiously. The statistical process control techniques help determine whether the production process is under control. However, companies should not be content with the present quality level of their products. The continuous quality improvement concept is based on this idea. While the quality of products are monitored through process control, the techniques that improve quality characterictics is needed. The most important one is design of experiments. Design of experiments is used to improve the performances of products and processes. Especially, the diagnosis of causes of problems that are found by process control and the elimination of these problems are other two of the benefits of this technique. The purpose of the industrial experimentation is to determine the factors that influence a product or a production process and to find the appropriate levels of these factors to optimize a design criteria or performance measure. The most common experimental designs are two-level, multiple factors factorial designs and the robust designs that are found by Taguchi and consist of factorial designs. The aim of this study is to investigate the experimental design techniques and show how they are used in quality improvement. The widely-used design alternatives are explained and data analysis techniques are given. The construction of 2k factorial and fractional factorial designs, analysis of them and blocking concept are studied in detail. In the last chapter, an example is given to illustrate how experiments are conducted in quality improvement context. VI