Predictor of Aircraft Structural Loads Due to Buffet

A hybrid simulation tool based on the immersogeometric compressible-flow fluid-structure interaction (FSI) modeling, data measured from in-flight sensor, deep learning, and multiaxial fatigue damage evaluation will be developed to predict fixed-wing aircraft’s buffet loads and their induced fatigue damage. The project will leverage the existing high-fidelity compressible-flow CFD code developed by the Iowa State University (ISU), a deep learning application code developed by Global Engineering & Materials, Inc. (GEM), and a multi-axial frequency domain fatigue damage evaluation code developed by GEM. The advanced FSI model will be validated thoroughly and used to predict the dynamic buffet load of the aircraft structure under compressible flow. After that, a hybrid deep learning approach is developed to predict the validated buffet load based on both simulation data and test-driven sensor data. The FSI simulation results and the deep neural network will be adjusted to minimize the prediction error. In addition, we will enhance the geometric representation flexibility by using a Nitsche’s method based non-matching multi-patch isogeometric description for complex aircraft geometries and develop a Kirchhoff-Love (KL)-continuum blended shell for better stress approximations and computational efficiency.