Impact of strain gauge misalignment and model parameters on load estimation in wind turbines

Accurate load estimation in wind turbines is important for understanding the structural behaviour and optimising performance. Loads are commonly measured using strain gauges, but the process is strongly based on accurate placement and alignment. This work explores the influence of geometric, material, misplacement, and misalignment parameters on the performance of different strain gauge configurations for obtaining load estimates. An analytical model for a single gauge is presented which establishes the relationship between exerted loads and measured strain. Using second-order approximations, the sensitivities of the estimated loads are calculated in terms of the uncertainty of the model parameters. The model allows to calculate how the secondary loads affect the estimation of those of interest, which is particularly useful in wind turbines where all three forces and three moments occur simultaneously. When estimating a single load, analytic results show that sensitivities to geometric and material parameters remain consistent regardless of secondary loads, while sensitivities to gauge bonding parameters reveal a cross-talk effect. As these sensitivity results depend on the model parameters and the instantaneous values of the secondary loads (which vary with time) the sensitivities will strongly depend on the application cases. In this paper a horizontal-axis wind turbine is simulated, calculating the loads at the Low Speed Shaft and the Tower Base. The sensitivity results for these subsystems using Montecarlo simulations show that while some loads can be precisely estimated, other loads will be very poorly estimated even when the gauges are bonded with tiny misalignment errors.