Search papers, labs, and topics across Lattice.
This study explores the application of residual learning using LSTM neural networks to enhance the prediction of unsteady aerodynamic loads, specifically for aeroelastic scenarios involving the NLR 7301 airfoil. By training the model to capture the difference between high-fidelity CFD lift data and a physics-based Wagner function prediction, the residual approach outperforms a direct neural network model in generalization and consistency, particularly in leave-one-out tests. The findings suggest that integrating residual learning with classical aerodynamic theories can effectively reduce prediction errors and improve model robustness in complex flow conditions.
Residual learning with LSTMs can significantly enhance aerodynamic load predictions, outperforming traditional models in generalization and consistency.
This paper investigates the feasibility of using residual learning to improve unsteady aerodynamic load prediction for aeroelastic applications. The machine learning technique selected for the study is the long short-term memory (LSTM) neural network, which is used for its suitability for sequential data with aerodynamic memory effects. The approach is investigated for the NLR 7301 airfoil benchmark using high-fidelity CFD lift data for prescribed pitch and plunge motions in the transonic flow regime in the presence of shock motion. An analytical unsteady aerodynamic model based on the Wagner function is used as a physics-based baseline, and the neural network is trained to learn the difference between the CFD lift coefficient and the Wagner prediction. The residual model is compared with a direct neural-network model trained to predict the CFD lift coefficient. The comparison includes feature and normalization studies, external benchmark cases, and leave-one-out and leave-family-out generalization tests across a range of sinusoidal and non-sinusoidal motions. The residual model performs best when its inputs align with the Wagner formulation variables, generally giving lower error and more consistent performance across training runs, though the direct model remains more accurate for some high-frequency cases. The residual model also generalizes better in the leave-one-out and leave-family-out tests, with a smaller increase in error than the direct model when entire motion families are withheld from training. Overall, the results indicate that residual learning shows promise as a modular approach for augmenting classical low-order aerodynamic theories, especially when the physics baseline removes a structured part of the aerodynamic response and leaves a lower-variance correction for the neural network to learn.
