Design Optimization

 

Design optimization, some call predictive analysis, is the engineering process that delivers the greatest part, or assembly, performance with the least material weight, the smallest net volumes, or the greatest strength from a material, etc. This process provides for designs that can operate at extremes of performance or cost. For example, what material web thickness in an area of stress concentration is required for the part to exhibit an “infinite” fatigue life? What shape minimizes the material volume, while maximizing the reactive sealing force? What arrangement of electronic components minimizes the cooling fan size while keeping the peak temperature below some maximum? These questions and more can be answered using the finite element method.

Design optimization questions like these involve optimizing the design in some deterministic way. The FEA method is ideally suited for this purpose as it can easily reconcile contradictory effects, or complimentary effects, that magnify problems if left undiscovered and therefore not addressed. These are physical effects that are always material and/or process driven, and/or load and geometry dependent.

Complex Made Tractable
It is often very difficult to analyze loads applied to single parts in complex assemblies without some guessing. We have the capability to analyze the part, while assembled, to remove this variable.

DOE and SPC
We employ a very useful tool called Design of Experiments (DOE) for the purposes of design optimization, which was developed as part of statistical process control (SPC) that exposes hidden effects in complex systems that when brought under control provide leaps in performance. This process simplifies complex interactive possibilities to statistically manageable proportions that leads to very quick and useful implementations of the FEA method. This translates directly to lower costs from start to finish in your product development life cycle.

If you are considering design optimization in the development of your products, please contact Designit4u, it’s what we do.