Robust design of friction-damped systems using multi-objective simultaneous optimization and PCE surrogate model

The friction interface in friction-damped systems exhibits pronounced nonlinearity. Designing the critical structural parameters requires simultaneous consideration of multiple objective functions and extensive iterative computations for optimization evaluation. Therefore, a novel robust design approach for friction-damped systems enabling multi-objective simultaneous optimization is proposed. This approach integrates adaptive Polynomial Chaos Expansion (PCE) with multi-objective optimization techniques to efficiently design the critical parameters of friction-damped systems. The objective function is defined to ensure that the damping performance of the system adheres to specified robustness criteria under parameter uncertainty. This approach is applied to optimize critical parameters of simulated and real turbine blades with Underplatform Dampers (UPD), targeting both damping performance and robustness to parameter uncertainty. The Pareto front is obtained, and experiments validate the optimization results against test data. Results show that the predicted peak frequency and amplitude agree well with experiments. Considering design and contact uncertainties significantly improves the damping performance and robustness of the friction damper. Additionally, the study explores the robust design of UPD damping effects in actual turbine blades under various excitation levels. This approach has proven to be highly effective in maintaining lower blade amplitudes across diverse operating conditions.