The Data-Driven Surrogate Model-Based Dynamic Design of Aeroengine Fan Systems

Abstract High-cycle fatigue failures of fan blade systems due to vibrational loads are great concern in the design aeroengines, where energy dissipation by relative frictional motion dovetail joints provides main damping mitigate vibrations. The performance such a can be enhanced suitable coatings. However, analysis and coated joint roots gas turbine blades computationally expensive strong contact friction nonlinearities also complex physics involved dovetail. In this study, data-driven surrogate model, known as Nonlinear Parameter AutoRegressive with eXegenous input (NP-ARX) is introduced circumvent difficulties systems. NP-ARX model linear input–output coefficients nonlinear functions parameters interest, that Frequency Response Function (FRF) directly obtained used system design. A simplified fan-bladed disc considered test case. results show using an efficient accurate aeroengine achieved. approach expected extended solve problems many other