Assignment 4: Design of Experiments-Taguchi Methods

Introduction

Pertain to the modern quality management, the proactive approach gradually takes over the predominant position from the reactive approach that is about detecting and correcting problems that are already exist, which means the reactive focuses on the past events. Moreover, the reactive approach is proved to be with high cost, low efficiency, difficult to achieve continuous quality improvement as well as very impracticable to solve the causes of quality problems. On the contrary, the proactive approach places emphasis on measurement, analysis, prediction and prevention,[1] which is cost effective, high efficient and easier to troubleshoot and problem solve. In associated with the proactive approaches that are employed by quality management, a greater emphasis is also needed at the design stage in order to prevent defects and errors appearing later in the product’s life cycle. To this end, quality should be designed into products and processes rather than inspected into them.[2] An important and well-recognized tool to achieve this goal is the Taguchi method.

Taguchi method, termed the robust parameter design that was popularized by Japanese engineer Genichi Taguchi in1980’s, is known as the offline quality control and engineering optimization by using design of experiment. The method derives from the Taguchi philosophy that regards the quality as the loss imparted by the product to the society from the time the product is shipped[3]. As long as the product does not comply with the target specification, the loss will occur, no matter the product is outside the specification or fall within the quality specifications. To minimize the loss, the management is faced with the task of producing the product at optimal levels with minimal variation in its functional characteristics. Generally, the product’s functional characteristics are affected by several certain factors/parameters, which can be categorized into two types: the controllable factors and the noise factors. The controllable factors are those factors that can be easily controlled, whereas, the noise factors, on the other hand, are those factors that are either difficult, or impossible, or expensive to control. The objective of Taguchi method is to identify optimum levels for these controllable factors so that the product is least sensitive to the effect of those noise factors.[4]

Taguchi not only called attention to the philosophy of robust parameter design, but he also proposed techniques of experimental design to identify the settings of those factors that would achieve robust performance as well as techniques for the analysis of these studies.[5]. For the purpose of identifying robust factors efficiently, the Taguchi method proposed the orthogonal array involving the controllable factors array and the noise factors array to optimize the design. Meanwhile, some mathematical methods including mean response and the S/N ratio are utilized to facilitate the analysis on these factors.

What worthy of being taken in account of is that in term of parameter design, there are three widely accepted methods utilized: the trial and error method, the full factorial method and the fractional factorial method. Amongst these methods, the fractional factorial method is the most cost effective, most efficient method that is widely used by Taguchi method. The effectiveness of fractional factorial design has already been proved by statistical method