Session: 14-07 Modelling Methods for IAS

Paper Number: 128967

128967 - A Physics Informed Neural Network for Solving the Inverse Heat Transfer Problem in Gas Turbine Rotating Cavities 

Recently, Physics-Informed Neural Networks have displayed great potential in delivering swift and accurate solutions for inverse problems. In this work, a Physics-Informed Neural Network (PINN) was used to solve the inverse heat conduction problem in the rotating cavities of gas turbine high-pressure compressor internal air system. The neural network was designed to receive experimentally captured radially distributed temperature profiles as inputs and predict the associated surface heat fluxes. The correctness of these predicted heat fluxes is assessed by numerically solving the heat conduction equation using an FEM model, thereby recovering the original temperature profiles. A comparative analysis is conducted between the predicted temperature profiles and the initial inputs.

The physics informed neural network is trained using noise-free synthetic data, created from a range of radial temperature curve fit coefficients, and subsequently tested on noisy experimental data at engine representative conditions. The predicted temperature values exhibit good agreement with their respective actual counterparts.  Furthermore, the sensitivity of model hyperparameters are explored to showcase the capability of the proposed approach. The results validate that physics-informed neural networks exhibit reduced susceptibility to experimental uncertainties when addressing inverse problems, in contrast to traditional solution methods, and offer a new approach for analysis of experimental data in the field.

Presenting Author: Mark Puttock-Brown University of Sussex

Presenting Author Biography: Dr Mark Puttock-Brown is Senior Lecturer in Mechanical Engineering and a member of the Thermo Fluid Mechanics Research Centre at the University of Sussex.

Authors:

Mark Puttock-Brown University of Sussex
Goutham Krishna Mahesh Bindhu University of Sussex
Colin Ashby University of Sussex