CAE-Based simulation and Design of Experiments (DoE) are becoming mature and increasingly effective in development of complex industrial products such as automobiles. We present in this paper a CAE mesh-modeling paradigm that ultimately led to fast, automatic generation of a family of meshes based on a base design. This paradigm is hinged on the so-called mesh features to achieve productivity for modeling CAE meshes. Mesh features are self-contained mesh deformation operations that are context-free, stored separately from the base model, and can be applied to the model in a proper mix at any time. Libraries of mesh features can also be established to archive useful features for future use. Furthermore, by assigning mesh features for DoE factors, one can specify for the system the proper way to assemble features and apply them automatically to the base model to generate input meshes for a DoE study. Automatic generation of a family of DoE input meshes results in maximum time savings and minimum chances for errors, especially for applications involving large-scale CAE models.
Goal-oriented optimal design of experiments for large-scale Bayesian linear inverse problems
