January 2005

Message from the CEO

I personally have now been involved with numerical modeling for more than twenty-five years in both academic and professional environments. During this time I have often debated with my friend and colleague Dr. S. Lee Barbour at the University of Saskatchewan, the what, why and how of good geotechnical numerical modeling. Recently, we put some of our thoughts on paper, the result of which appears in an article in the December 2004 issue of Geotechnical News, titled Numerical Modelling – Prediction or Process? A copy of this article is available on our web site.

There seems to be considerable confusion in the geotechnical community as to what the objective of numerical modeling is, and what constitutes good modeling practice. We argue that modeling is more about process than prediction; and that the real benefit of numerical modeling is that it helps us understand the fundamental engineering process and behavior more than it helps us make quantitative predictions.

Immediately after we had submitted the article to Geotechnical News, a book came across my desk entitled Geotechnical Modelling, by David Muir Wood (Spon Press, 2004). I was excited to find he has a chapter on this very subject, and that he also advocates the same ideas we put forth in the Geotechnical News article. His chapter titled Envoi hits the nail right on the head in more simple and elegant prose than what we do in the Geotechnical News article.

Most engineers, when they first start doing numerical modeling, make their models far too complex. This is certainly something we at GEO-SLOPE observe continually through our software support. Intuitively, the idea is that more complexity will lead to more realistic results. However, this is usually not true. The best and most understandable results often come from the least complex models. As Muir Wood points out, “the art of successful modelling is to include just enough detail for the implied simplifications to be reasonable for a particular application.” I wholeheartedly concur. The best guideline in numerical modeling is still to keep the model as simple as possible.

The engineering books that accompany our software products also have a chapter on this subject, which is available for convenient reference for those who have acquired our latest software.

I believe that all of you who get involved in geotechnical numerical modeling will enjoy and greatly benefit from contemplating this published information on the “how-to” of numerical modeling. In the end, please remember that numerical modeling is more about process than about prediction.

John Krahn, Ph.D., P.Eng.
President and CEO, GEO-SLOPE International Ltd.

This message comes to you as you have shown past interest in GEO-SLOPE products. If you do not wish to receive this type of information from time to time, please click here and we will remove you from future communications.