Abstract:1. National Key Laboratory of Science and Technology on Aerodynamic Design and Research, Northwestern Polytechnical University, Xi'an 710072, China 2. School of Aerospace; Tsinghua University; Beijing 100084, China Non-linearities can be present in an aeroelastic system due to some aerodynamic factors that occur in transonic flight regime or at large angle of attack. The sources are motions of shock wave and separated flow. By using high-fidelity CFD codes, complex aeroelastic problem due to the aerodynamic nonlinearity can be studies. However, the computational cost of this method is very high. It is convenient to solve this kind of problem by constructing a proper unsteady aerodynamic model previously. Many research works are carried out in reduced order modeling (ROM) for aeroelastic analysis. Most of the reduced order aerodynamic models are dynamic linear models and in proportion to the structural motions. By using Radial Basis Function (RBF) neural network model, the nonlinear unsteady reduced order aerodynamic model is constructed. ROM is used to analyze LCOs behaviors for two linear structural models with large shock motion in transonic flow. Different from the traditional design method of the input signals, signals of self-excited vibration of the aeroelastic system are designed as the input signals in this paper. Coupling the structural equations of motion and nonlinear aerodynamic ROM, the system responses are determined by time marching of the governing equations using a kind of hybrid linear multi-step algorithm and the limit cycle behaviors changing with velocities (dynamic pressure) can be analyzed. Two transonic aeroelastic examples show that both the structural responses and the limit cycle oscillation (LCO) characteristics simulated by ROM are agree well with those obtained by direct CFD method while the computational efficiency of ROM based method is improved 1-2 orders of magnitude comparing with the direct CFD method.

Nonlinear Manifold Learning and Model Reduction for Transonic Flows

Nonlinear Reduced-Order Modeling of Compressible Flow Fields Using Deep Learning and Manifold Learning

Manifold learning-based reduced-order model for full speed flow field

Efficient Nonlinear Reduced-Order Model for Computational Fluid Dynamics-Based Aeroelastic Analysis

Data-driven reduced-order modeling for nonlinear aerodynamics using an autoencoder neural network

Reduced order modeling for flow and transport problems with Barlow Twins self-supervised learning

Learning physics-based reduced-order models from data using nonlinear manifolds

Registration-based nonlinear model reduction of parametrized aerodynamics problems with applications to transonic Euler and RANS flows

SNF-ROM: Projection-based nonlinear reduced order modeling with smooth neural fields

Efficient Method For Limit Cycle Flutter Analysis By Nonlinear Aerodynamic Reduced-Order Models

Nonlinear Reduced-Order Models for Transonic Aeroelastic and Aeroservoelastic Problems

Nonlinear Shape-Manifold Learning Approach: Concepts, Tools and Applications

Adjacency-based, Non-intrusive Reduced-order Modeling for Fluid-Structure Interactions

Machine-Learning for Nonintrusive Model Order Reduction of the Parametric Inviscid Transonic Flow past an airfoil

A Novel Model Reduction Method Based on Balanced Truncation

High Efficient Numerical Method for Limit Cycle Flutter Analysis Based on Nonlinear Aerodynamic Reduced Order Model Reduced Order Model

A manifold learning approach to nonlinear model order reduction of quasi-static problems in solid mechanics

Deep neural networks for nonlinear model order reduction of unsteady flows

An Approach to Enhance the Generalization Capability of Nonlinear Aerodynamic Reduced-Order Models

Enhancing high-fidelity nonlinear solver with reduced order model

Non-linear Manifold Reduced-Order Models with Convolutional Autoencoders and Reduced Over-Collocation Method