This paper on "Preparation of CAD Geometry for Analysis and Optimization" was presented at the NAFEMS World Congress on Effective Engineering Analysis - 25-28 April 1999, Newport, Rhode Island, USA.

Abstract

Integrated design and simulation can only be deemed successful if it ultimately reduces product cost, improves product quality, or shortens time-to-market. However, the big picture is often obscured by the direction most companies are taking. It is the nature of the design engineer that when given two options, he/she will typically choose the path that most quickly leads to tangible, visible results. Hence, the propagation of "push-button", CAD-embedded mesher/solvers. These tools promise immediate results from CAD solids, which were going to be created as part of the design process anyway, with no regard to geometric applicability or boundary uncertainty. Unfortunately, with FEA, the path to instant gratification will more likely result in bad results than shortened design cycles.
An alternate approach to assuming that a CAD solid is the only launch point for analysis is to better educate the entire design team on the geometry needs of a proper simulation program. To that end, we have developed guidelines and workshops for product development organizations to maximize simulation efficiency by providing the right geometry at the right point in the design process, thereby minimizing waste in model clean up, redundancy in modeling effort, and the limitations on true part and system optimization.
These guidelines as they pertain to the actual CAD users, or Geometry Providers, include definition of the goal of the project; pre-mature design commitment; communication with the analyst(s); identification of the proper idealizations; and choosing the best geometry option for the task at hand. When CAD solids are deemed appropriate for the analysis, the guidelines present optimal practices for feature creation taking into account dirty geometry, fragile model construction, geometry simplification, and optimization.
The author has seen that FEA success is more than a mesh. In the product development process, it requires an orchestrated effort based on a clear implementation plan driven by the goal(s) and an effective measurement/evaluation process. The organization must support new work flows directed at seamless processes, not stand-alone functions [1]. Expecting that the analysis process should bend to the convenience and availability of CAD geometry will inevitably present roadblocks to improving the cost, timeliness, and quality of the product development process with FEA.