Physics-Informed Neural Networks for Reduced-Order Modeling of Turbomachinery Blisks With Small and Large Mistuning

Abstract Comprehensively predicting the structural dynamics of turbomachinery blisks is of critical importance to the gas turbine industry. Located in the compressor and turbine stages, vibrations of these structures are amplified by their inherent small or sometimes large mistuning. Hence, it is paramount for the safe operation of gas turbines to predict mistuned blisk vibration responses. However, this requires significantly higher computational effort compared to computing cyclic system (i.e., tuned) responses. To address this issue, physics-based and data-driven reduced-order models (ROMs) have been developed. While many physics-based ROMs have been developed for predicting blisk responses with small and large mistuning, they require large finite element (FE) models to characterize the system dynamics, and they cannot be enhanced using experimental data. Thus, this paper proposes a novel physics-informed data-driven approach to compute blisk responses with both large and small mistuning. Similar to classical physics-based approaches (i.e., PRIME), this paper utilizes two physics-informed neural networks to compute the transfer function matrices of two systems: (1) a cyclic pristine blisk with small mistuning, and (2) a cyclic rogue blisk with small mistuning. These transfer functions are then introduced in a linear system of equations to compute the blade root responses of a blisk with small mistuning and rogue blades. Blade tip responses can then be computed using root responses through a third neural network. Results show highly accurate predictions, with absolute errors below 10% for all mistuned blisk configurations explored.