Robust design optimisation of underplatform dampers for turbine applications using a surrogate model

Abstract

Underplatform dampers (UPD) represent an effective way to limit blade vibration in turbomachinery via frictional energy dissipation, leading to a wide range of applications. The design of an effective and reliable UPD is highly challenging, due to the inherently nonlinear nature of the contact forces, the associated computational cost for high fidelity simulation, and the manufacturing uncertainties in damper geometry. This paper presents a novel UPD optimisation approach that combines high-order, detailed nonlinear modelling of the damper interfaces with a surrogate model optimisation technique. The nonlinear dynamic behaviour of the UPD is predicted using the existing explicit damper model in combination with an ‘in-house’ multi-harmonic balance solvers, which enables capture of the damper kinematics and local contact conditions. A radial basis function based surrogate model will be used to address the computational requirement of the high fidelity simulations for alternative designs. The objective function takes into account the damping performance, resonance frequency stability and robustness due to possible uncertain variations of design parameters with manufacture tolerance. The feasibility of the proposed approach is demonstrated on a cottage roof UPD by comparing the proposed optimisation method with conventional parametric simulation method. A significantly improved solution with considerable reduction in computational effort is achieved by the current method.