Abstract
Blade strain measurements and load analysis have advanced significantly in wind turbine research with improvements in measurement durability, reliability, and accuracy employing fiber optic strain sensors (FOS), improved sensor configurations, temperature compensation methods to minimize errors, and application of finite element analysis (FEA) modeling to resolve the relationship between the strain field and load vector. Experience deploying advanced FOS sensor measurement techniques for marine turbine blades is relatively recent and rare. Published blade strain measurements for marine turbine blades are predominantly from conventional resistance strain gauges employed in scale-model axial-flow turbine studies in controlled laboratory testing. In this work, three turbine blades are tested for different loading configurations; to calibrate and validate FOS measurement and to validate the finite element analysis (FEA) model used to predict the blades response. Fiber optic-based sensors were selected as they have low noise and a relatively small footprint. In addition, based on the FEA simulations, the sensors should be capable of measuring the full range of strain that will be experienced during operations. The sensor layout for the blades is comprised of four rosettes at the blade root and four uniaxial strain gauges along the blade, as shown in Figure 1. The material and cuboid geometry of the blade root section was carefully designed so that placing the rosettes at this location meets the recommendations regarding the choice of sensor location.
Bending moments (flapwise and edgewise) and twist moments were predicted using OpenFAST model, which provides along-blade distributed loads, to mimic operational conditions. The maximum moments in the structural testing are replicated by applying a point force at specific locations on the blade that produces the same operational moment values. The locations of the point force were determined by comparing the deflections predicted using two FEA models; the first model used the distributed loads as input from OpenFAST, while the second model used a point force a selected locations on the blade. The objective of this modeling effort is to closely match the blade deflection profile between the two models for the same force magnitude. The overarching goal of this experimental testing is to perform a set of experimental testing to calibrate: 1) the bending moment values derived from the blade root fiber optic rosette strain sensors against a set of known loads along the flapwise, edgewise and combined flapwise and edgewise directions, and 2) the performance of the twist moment values derived from the rosettes. This is a critical step to ensure that accurate bending moment measurements can be derived from the rosette strain measurements. Results from the experimental program showed promising potential to predict the loading on the turbine blades using implemented FOS.