Speaker
Description
Remote thermographic inspection of wind turbines, either from the ground or from the air (drones), has great potential for condition monitoring throughout their entire service life However, the formation of measurable thermal contrasts on individual components depends heavily on dynamic weather conditions. A systematic and quantitative description of the observed effects is missing in literature, both because of the complexity of weather conditions and the lack of information about the internal structure of the rotor blades under investigation.
In order to study the evolution of thermal contrasts while reducing the variety of external influences (e.g. solar irradiation or strong wind gusts), a series of tests was carried out in a climate chamber. The object under investigation was a section of a wind turbine rotor blade (WTB) with a known internal structure.
In this study the climate chamber was used to heat and cool the test object as evenly as possible. To this end, air at a predetermined temperature was blown into the chamber. The IR camera used to observe the temperature distribution on the surface of the WTB-section was also located inside the climate chamber.
To gain a deeper understanding, an FE model of the WTB geometry was created and the temperature distribution under the influence of changing air temperature was simulated using COMSOL. Although all parameters for the thermal simulation are known in principle, it was not possible to achieve an optimal fit of the simulated temperature curve to the measured data by varying the parameters. Instead, a systematic deviation between both curves was observed, with a maximum difference of 0.4 K. Several underlying factors contributing to the observed deviation could be identified including:
1. non-uniform heating of the WTB-section (decreasing intensity of heating from up to down)
2. inadequate compensation of external temperature variations in the temperature calibration model of the infrared (IR) camera
3. Air infiltrates the specimen, which sides were only partially sealed.
4. Non-uniform initial temperature distribution within the WTB-section, despite the extended waiting times
In principle, all of the aforementioned effects could be incorporated into the simulation to enhance its precision. However, the original objective of a simulation without free parameters would demand a series of further investigations to experimentally quantify the influence of those disturbances
In summary it is evident that even in ostensibly rudimentary thermographic experiments, precise thermal simulation remains challenging, particularly when only minor thermal contrasts significantly below 0.1 K are to be considered.
