A HIGH-FIDELITY NON-INTRUSIVE REDUCED-ORDER MODEL FOR THE STATIC AEROELASTIC BEHAVIOR OF LIFTING SURFACES

Date

2025-6

Author

Özkaya, Zarif Özge

Metadata

Show full item record

This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.

Item Usage Stats

36
views

0
downloads

Cite This

This study proposes an efficient, low-cost reduced-order method for analyzing the static aeroelastic behavior of a three-dimensional sweptback wing, focusing on the impact of wing elasticity on model efficiency. The primary method used is Proper Orthogonal Decomposition (POD) coupled with Radial Basis Function (RBF) interpolation. The results reveal that reduced-order modeling (ROM) is more effective for the aeroelastic wing, particularly when the pressure coefficient (Cp) is the output. This improvement is attributed to the bending-torsion coupling in the aeroelastic wing, which reduces the effective angle of attack and delays flow separation. For flexible wings, the study further shows that rigid node assumption is insufficient for accurately capturing Cp distributions; hence, separate ROMs must be generated for the deformed positions of the nodes. These node-specific ROMs also enable accurate predictions of the wing deflection. Additionally, the RBF method is applied to generate surrogate models by direct interpolation of the data ensemble. The accuracy, computational cost, and practicality of the POD-RBF coupled ROM and the RBF surrogate model are compared. The study also explores vi the use of Artificial Neural Networks (ANN) to predict Cp values, both in combination with POD to develop a reduced-order model and independently as a surrogate model. The findings show that the POD-ANN based ROM outperforms the ANN surrogate model in terms of predictive accuracy and computational efficiency. In conclusion, the methods proposed in this study provide a highly efficient and accurate alternative to high-fidelity analyses, offering substantial computational savings and making them valuable for future aeroelastic modeling.

Subject Keywords

Reduced Order Modelling, Static Aeroelasticity, Proper Orthogonal Decomposition, Radial Basis Functions, Artificial Neural Networks

URI

https://hdl.handle.net/11511/115112

Collections

Graduate School of Natural and Applied Sciences, Thesis