Diffusion Model Driven Airfoil Design: From Geometry Encoding to Practical Applications

The diffusion model, the state-of-the-art generative machine-learning architecture, has shown promising results in airfoil inverse designs. In this study, we implement and trained a series of diffusion models on three different airfoil geometry data encoding formats [principal component weights, ordered [Formula: see text] coordinates, and 2D signed distance functions (SDF)] to generate 2D airfoils. By systematically comparing the performance of diffusion models trained on different data structures, it is found that for 2D airfoil design problems the diffusion model performs the best when directly trained with coordinates. Training with latent space (principal component analysis weights in this study) limits the model’s design freedom and decreases the training effectiveness. Although the 2D SDF data appear to result in the least performing model, it proves its feasibility in aerodynamic shape generation, paving the way toward 3D problems where SDF is more favored. This study also investigated deploying the diffusion model in practical engineering applications. A multitarget optimization procedure is proposed based on the stochastic nature of the diffusion process, which drastically simplifies the procedure compared to conventional methods. The extrapolation performance of the model is also investigated by tasking the model with both aerodynamic and flow condition labels that are extrapolated beyond the training set boundaries.