How Leaders are Escaping AI Pilot PurgatoryRegister here
PINN Shaft Fault Detection
0
0
Statistics
0 Geometries
0 Meshes
0 Simulation setups
0 Results
0 Geometries 0 Meshes 0 Simulation setups 0 ResultsAbout this project
This project focuses on vibration-based fault detection in mechanical systems using Physics-Informed Neural Networks (PINNs). A simplified shaft model is created and analyzed under dynamic loading conditions to simulate vibration behavior. The system is modeled as a mass–spring–damper system, and stiffness variations are introduced to represent faults such as inner race, outer race, and ball defects. The simulation results are used to understand how changes in stiffness affect vibration response. These insights are then connected with a PINN model that estimates system parameters directly from vibration data while embedding governing physical laws. The objective is to combine physics-based simulation with machine learning for reliable and interpretable fault detection.
