This paper discusses the use of the continuous adjoint methodology for the design of transonic aircraft configurations. The objective of the present study is to illustrate that, despite the fact that continuous adjoint does not provide the numerically exact gradient of a discrete objective function, it is an excellent choice for large-scale optimization with complex geometries due to the accuracy of the computed gradients, the robustness of the method, and its efficiency in terms of memory and compute time. This is demonstrated through a series of detailed design studies on the NASA Common Research Model, which provides a challenging problem due to its geometric complexity and the large, unstructured meshes employed. All design work is completed with the SU2 software suite: an opensource, integrated analysis and design tool for solving complex, multi-disciplinary problems on unstructured computational grids.
No takes yet. Share an insight, caveat, or question.
Palacios et al. (2015) studied this question.
Synapse has enriched 5 related papers. Consider them for comparative context: