Dec. 2: Seminar – Design of Experiments

December 1, 2011 under CANR News

Recent Advances in Design of Experiments (DOE) Methods for Screening Factors

Attendees will learn about several exciting new Design of Experiments (DOE) methods for screening factors first published in 2011.  This presentation will introduce experimenters to the very latest and efficient DOE methods not yet found in textbooks.  Definitive Screening designs will be presented in detail.  These designs for all continuous factors, can – in fewer trials than classical fractional-factorial designs – not only detect main effects but also curvature in EACH factor.  When the number of significant factors is small, a Definitive Screening design can collapse into a “one-shot” design capable of supporting a response-surface model with which a process can be optimized or fully characterized.  A case study will be shown in which a 10-factor process is optimized in just 24 trials.  Moreover, when more than a few factors are significant and the design cannot collapse into a one shot design, the design can economically be augmented to support a response surface model in the important factors.  Additionally, “alias-optimal” designs (2011) useful when not all factors are continuous – i.e. when several factors are categorical at multiple levels, as well as a class of “supersaturated” designs (2008) for selecting dominant factors will be demonstrated. Graphical comparisons between these alternative methods and traditional designs will show the new ones to be superior or strong competitors.

Tom Donnelly has been using Design of Experiments (DOE) methods for over 25 years and has taught more than 250 short courses on the topic to scientists and engineers.  From 2005 through 2008 he worked for the US Army Edgewood Chemical Biological Center in the Modeling, Simulation and Analysis Branch supporting the use of DOE with real experiments as well as with experiments conducted using computers.  Now employed by the SAS Institute Inc., he continues to promote and support the use of DOE methods particularly for use with Modeling and Simulation (M&S) at Department of Defense sites, Department of Energy laboratories, and among government contractors.  Donnelly received his PhD in Physics from the University of Delaware and was a postdoctoral fellow at the Institute of Energy Conversion.  He was also a partner in the very first DOE software company, ECHIP Inc., based in Hockessin.

Submitted by Erma Wopert on behalf of Dr. Palaniappa