SPEEDING TECHNICAL SOLUTIONS IN PARTICULATE TECHNOLOGY WITH DESIGN OFEXPERIMENTS AND RELATED STATISTICAL-METHODS

Technical challenges in particulate technology can range from troubles hooting and process and product improvement, to the development of new processes and products. Design of Experiments (DOE), and related stat istical modeling methods such as Sources of Variation Analysis and Mul tiple Correlation provide a systematic approach to speeding the needed technical solutions. Product cycles have been compressed by factors o f 3 to 4 with their use. Particulate technology examples, including ex perimental data on corrosion resistance of stainless steel parts and c rush strength of iron-copper-carbon alloys, will be used to illustrate the power of these methods and to compare them with the classic, Tagu chi, and Dorian Shainin approaches to experimental design. While much can be accomplished with simple designs and analysis, when the computi ng power of today's personal computer (PC) is coupled with an engineer or scientist knowledgeable in these tools, a powerful engine is creat ed for advancing particulate technology. Recently, highly efficient, P C generated, D-optimal experimental designs have become available. The flexibility of the D-optimal method and its capability to deal handil y with as many as 15 variables, opens an opportunity for radical impro vement in the strategy of experimentation in all three schools of expe rimental design.