Probabilistic Crack Growth Behavior in Turbine Discs Using 3D Finite Element With Design of Experiment and Monte Carlo Method

Abstract This study presents a numerical investigation into the 3D finite element based probabilistic crack growth through turbine discs in gas turbines, utilizing ANSYS SMART for simulation. This probabilistic analysis includes uncertainties regarding material properties, crack growth rate coefficients and fracture toughness, and crack geometries which are critical for predicting crack propagation life. Random variables of input quantities have specific distributions. An initial crack is located by regarding the maximum principal stress region. The effect of worst engine performance variations compared to the average engine due to product, development, and control uncertainty on the crack growth behavior is also considered. Performance scatters are based on the number of turbine disc rotations, one of the most critical boundary conditions. Design of Experiment (DoE) and Monte Carlo (MC) method are implemented to figure out the Probability of Failure (PoF) and probabilistic life prediction. Developing a methodology to conduct a sensitivity analysis to assess the effects of the inputs to the outputs, and finding out the main driving factors affecting crack propagation is one of the main objectives. By integrating the parameters into a gas turbine disc and calculating PoF, the study seeks to enhance inspection interval which is crucial for gas turbine rotor components to improve safety margins. Ultimately, this investigation aims for a more reliable risk management and cost-effectiveness method over traditional deterministic assessments.