Atlas home || Conferences | Abstracts | about Atlas

AD 2000 - From Simulation to Optimization
June 19-23, 2000
INRIA Sophia Antipolis
Sophia Antipolis, France

Organizers
George Corliss, Christele Faure, Andre Galligo, Andreas Griewank, Laurent Hascoet, Uwe Naumann

View Abstracts
Conference Homepage

Variational Analysis : A Parametric Structural Analysis Software
by
Bruno Reymond
CADOE S.A.
Coauthors: Jean Daniel BELEY, Pr. Mohamed MASMOUDI

Title :
Variational Analysis : A Parametric Structural Analysis Software

Authors :
Jean Daniel BELEY, Projects Director
CADOE - 69100 Villeurbanne - France
Pr. Mohamed MASMOUDI,
MIP, UMR 5640 CNRS, Paul Sabatier University, INSA - Toulouse - France

The researches on Automatic differentiation are mainly motivated by gradient computation and optimization.

However, in the Optimal Design area, it is quite difficult to use optimization tools : aesthetics constraints, manufacturing constraints, ..., are quite difficult to describe by mathematical models. And even when we succeed to write a correct optimization problem, the result is often disappointing.

In practice the optimal design process is a dialog between the designer and the analysis software (structural mechanics, electromagnetic, fluid mechanics, ...). One analysis may take a while. For this reason parameterization tools, like design of experiments (D.O.E.), neural networks, ... are used. The aim of those tools is to give to the designer real time responses.

Variational Analysis provides a parameterization of the structural mechanics responses using higher order derivatives. Those derivatives are obtained by automatic differentiation.

For statics and normal modes analysis, performances comparisons between Variational Analysis using AD and standard re-analysis will be shown on industrial examples.

Variational Analysis is integrated into the I-DEAS Master Series environment and is marketed and distributed by the Company SDRC.

Date received: March 8, 2000


Copyright © 2000 by the author(s). The author(s) of this document and the organizers of the conference have granted their consent to include this abstract in Atlas Conferences Inc. Document # caev-01.