LAUSR

Abstract The measurement of stress/strain on blade plays an essential role in the fatigue assessment of the hydroelectric turbine runner. However, a typical measurement campaign does not cover all the possible operating conditions due to the field measurement requirements and limitations, which leads to fatigue evaluation missing for unmeasured conditions. In this paper, kriging interpolation method is used to estimate the missing fatigue information based on the available measured operating conditions of the turbine. The rainflow counting matrix, which contains the fatigue cycles of strain signal, is considered as the input of the interpolation process. For validation, the fatigue cycles are randomly generated, using bivariate Gaussian distributions, from the interpolated matrix for comparison with the experimental values under similar operating conditions. The developed approach will help plan experimental campaign and might contribute to decreasing the measurement requirements for runner fatigue assessment.