Using Parametric Effectiveness for Efficient CAD-Based Adjoint Optimization

Parametric effectiveness is a measure of the ability of the parameters defining a CAD model to be used for optimization. It compares the optimum change in performance that can be achieved using a CAD model’s parameterization, to the maximum performance improvement that could be obtained if the model is free to move. The aim of this paper is to present an automated approach to efficiently compute the parametric effectiveness for the parameters defined within a CAD modelling software CATIA V5. The approach is further developed to automatically identify a subset of CAD parameters which provides the greatest potential for performance improvement. The rationale for selecting such a subset is to reduce the time required to update a parametric CAD model during the optimization, which is an important factor to be considered in an industrial workflow. The approach is applied to the shape optimization of an S-Bend duct for minimizing the power-loss and an automotive car mirror for minimizing the noise perceived by the driver of the car. The flow sensitivities are computed with a continuous adjoint method