Prediction of hydroelectric turbine runner strain signal via cyclostationary decomposition and kriging interpolation

Strain measurements by gauges can play an important role in analyzing the fatigue damage of the hydroelectric turbine runner. However, these measurements cannot cover every steady-state operating condition due to the experimental limitations. Thus, the aim of this research is to predict the strain signal on runner over every possible steady operating condition using available experimental measurements. The strain signal measured during steady state involves several components (such as periodicity, vortex rope component and stochastic components) which can generate difficulties during the prediction. This paper proposes a solution to predict the runner strain signals by independently interpolating each physical phenomenon over different turbine operating conditions. These components are extracted using cyclostationary decomposition operators. A case study is performed on a Francis hydroelectric turbine to verify the interpolation performance. The proposed methodology can contribute to the fatigue assessment and help to reduce the requirements of infield measurements.