Here at Clemson, we have done a lot of work on the packaging problem as we call it. Back in 1995 I believe, or 1994, we presented a paper on GA packing of a
vat, considering the box envelop at the Dayton conference. Since then,
we have extended the work to look at general packing problems with multiple
objectives (not only dense packing, but also other objectives such as center of
gravity, vibrations, maintainability) and applied to nonconvex shapes.
This work is in the process of being published, and you can find information
on our web site.
Jon Cagan from CMU has also done work in this area with Simon Szyckman. Their
work based on simulated annealing has been published a couple of years ago at
the ASME design conferences.
Both our approaches deal with the more general packing problem such as the one
encountered when packing components such as the underhood of a car or an
The points that Marshall Burns raised are addressed in the dissertation of my
student Pierre Grignon. It is not out yet, he is still writing, but we do
consider several mathematically provable problems and show how our results
compare to them.
However, since we are more interested in general design problems, the solutions we find, especially when considering pareto problems (multiple objectives)
do not have to be the best possible, they have to be better with respect to a
number of criteria.
The approach we followed combines CAD data once tessellated with a genetic
--- Georges M. Fadel Associate Professor Mechanical Engineering Department Tel: (864) 656-5620 Clemson University Fax: (864) 656-4435 Clemson SC 29634-0921 USA e-mail: firstname.lastname@example.org www: http://design.eng.clemson.edu/dmg/people/faculty/fadel.html
This archive was generated by hypermail 2.1.2 : Tue Jun 05 2001 - 22:39:13 EEST