Generalizability of Convolutional Encoder–Decoder Networks for Aerodynamic Flow-Field Prediction Across Geometric and Physical-Fluidic Variations

Abstract

The generalizability of a convolutional encoder-decoder based model in predicting aerodynamic flow field across various flow regimes and geometric variation is assessed. A rich master dataset consisting of 11,000+ simulations including cambered, uncambered, thin and thick airfoils simulated at varying angles of attack is generated. The various Mach and Reynolds number (Re) chosen allows analysis across compressible, incompressible, low and high Re flow regimes. Multiple studies are carried out with the model trained on datasets that are categorized based on the above parameters. In each study, the loss of prediction accuracy by training the model on a larger dataset (generalizability), versus a smaller categorically sorted dataset, is evaluated. Largely disparate flow features across the Re range lead to a 25.56% loss, while the generalization across Mach range led to an average of 23.95% loss. However, flow-field changes induced due to geometric variation exhibited a better generalization potential, through an increased accuracy of 12.4%. The encoder-decoder architecture allows extraction of relevant geometric features from largely different geometries (geometric generalization) providing a better out-of-sample prediction accuracy in comparison to physics-based generalization. It is shown that, through user-informed choice of training data (removal of geometrically similar samples), computational costs incurred in generating training data can be reduced. This is important for the application of such methods in the design optimization of platforms and components that require analysis of the fluid flows.

Discretization-independent surrogate modeling over complex geometries using hypernetworks and implicit representations

“…Researchers note that the accuracy is greatly increased by using the signed distance function (SDF -see appendix) as a network input instead of a simpler binary representation. Similar work extends these ideas to aerodynamic problems with parametric variation in flow condition and different geometries [7,8]. Pressure and velocity fields for parametric two-dimensional RANS flows around airfoils are predicted with the signed distance field as the input.…”

“…Convolutional neural networks (CNNs) have been used to construct surrogate models of partial differential equations as predictive autoencoders [6,7,8,9], which are sometimes used in conjunction with a time advance model for time varying problems [10,11]. The models may be thought of as image-to-image mappings, where input features are mapped to predicted quantities.…”

Prediction of Transonic Flow over Cascades via Graph Embedding Methods on Large-Scale Point Clouds

“…It is still necessary to enhance the generalization capacity of the existing loss function, which does not incorporate the constraints of the N-S equation. The introduction of the physics-informed neural network (PINN) may improve the model performance and effectively utilize the gradient information in graph neural network calculations [50][51][52]. In the future, the prediction and generalization performance of the model will be further improved by introducing N-S equation constraints, thus improving the interpretability of the model, and optimizing design will be developed based on the learned cascade flow channel characteristics.…”

Fast Aerodynamic Prediction of Airfoil with Trailing Edge Flap Based on Multi-Task Deep Learning

“…In addition, Thuerey et al [7] investigated the accuracy of flow field prediction by deep learning model, focusing on how training dataset size and the number of weights affect the prediction accuracy. Finally, Tangsali et al [8] explored the generalization ability of models based on encoder-decoder architecture in predicting aerodynamic flow fields with various geometric changes.…”