Predicting Rain Erosion Damage in Wind Turbine Blade Coating Materials
Nick Hoksbergen is a PhD student in the department Production Technology. (Co)Promotors are prof.dr.ir. R. Akkerman and dr. I. Baran from the faculty Engineering Technology.
The offshore wind energy market is growing rapidly due to global climate agreements. This growing demand results in larger wind turbines with an increased annual energy production and lowered levelized cost of energy. These larger wind turbines however lead to tip speeds that exceed 100 ms-1. The high speed blade tips interact with airborne rain droplets which causes erosion damage over time lowering aerodynamic efficiency and with that annual energy production. Because of this, and increased maintenance cost, the levelized cost of energy increases.
In order to reduce leading edge erosion, coating systems are used. Typically, these systems are based on thermosetting gelcoats or elastomeric tapes/shells that are bonded to the blade structure through adhesives. Although these leading edge protection systems lead to an increased lifetime, regular maintenance is still required. Moreover, these systems are designed based on engineering expertise and long duration rain erosion tests.
The goal of the current work was to develop a leading edge protection system lifetime prediction method based on a numerical modeling strategy that conforms to the relevant physical phenomena. This approach should ideally be based on simply obtainable material parameters rather than expensive rain erosion tests.
For this purpose, first the currently used (analytical) Springer model was analyzed and a sensitivity study was performed. It was shown that the assumptions made by the Springer model led to overprediction of lifetimes for materials with a Poisson’s ratio close to 0.5 or similar ultimate tensile and fatigue strengths. Moreover, the loads used by the Springer model are based on the one dimensional waterhammer pressure and independent of droplet shape and time.
An experimental pulsating jet erosion test setup was developed that was able to replicate liquid droplet impact loading on material coupon level. The setup was successfully used to induce damage in thermoplastic material systems. Different damage mechanisms were found for different thermoplastic materials based on microscopy techniques and a newly developed volume loss algorithm for surface damage.
The numerical framework is described in three chapters:
1. Liquid droplet impact pressure on (elastic) targets
2. Coating-substrate system stress analysis
3. Fatigue lifetime prediction
First the relevant physical phenomena involved in the liquid droplet impact event were identified and a two phase flow fluid-structure interaction model was developed in COMSOL Multiphysics®. The model was used to study the effect of impact and material parameters on the resulting contact pressure. It was found that droplet diameter, impact velocity and target elasticity play an important role in the magnitude and shape of the contact pressure profiles.
The resulting contact pressure profiles were used as input to an axisymmetric finite element model to study the stress state in the coating-substrate system. It was found that, next to impact and material parameters, the geometric definition of the LEP system (thickness and interphase definition) plays an important role in the dynamic stress field development.
The resulting dynamic stress field was used in combination with the Rainflow counting algorithm and the Palmgren-Miner rule in order to predict fatigue lifetime of the system for single point impacts as well as distributed impacts. It was found that there is a significant difference between single point impact and distributed impacts due to overlapping stress history regions.
The current work is a next step towards physically representative modeling of liquid droplet impact fatigue of polymeric leading edge protection systems for wind turbine blades. Although progress has been made, validation of the developed models and further investigation of e.g. the effect of strain rate has to be done. Site-specific performance can be predicted when the model is related with meteorological parameters which was shown in the discussion. Ideally, the developed models form a first step towards an optimization framework for liquid droplet impact coating systems for wind turbine blades.
