Ambient vibration test of wind turbine of wind turbine blade in OWI-lab's Large Climate Chamber Selected data from the large scale icing experiment as conducted in OWI-lab's large climate chamber on 27/09/2022 and 15/11/2022. Results were presented during Eurodyn 2023 in :"Large scale test of vibration based icing detection for wind turbines", Weijtjens et.al. Data is (summarized, more details are given below): - 24 Ten minute acceleration data files collected (MO04_acceleration_YYYYmmdd_HHMMSS.csv) - Pictures during the experiment timestamped (local time: UTC+1)- Temperature measurements of the climate chamber's inflow temperatures- Modal parameter results for X and Z direction All times are in UTC unless mentioned otherwise. Version 2 : Addition of the powerpoint presentation held at Eurodyn, including additional information on the results and the methodology employed Version 3 : Addition of the data of the dry run on 27/09/2022 Measurement concept The measurement data is collected during an experiment as conducted as part of the COOCK fighting icing project led by Sirris. In the OWI-lab climate chamber a wind turbine blade was subjected to icing conditions. The main test on 15/11/2022 comprises the collection of ambient vibration data using three tri-axial accelerometers installed on the blade. During the day the blade is cooled and cold water is sprayed on the blade to simulate the growth of ice on the blade. The steps of the experiment are: 2022-11-15 10:04:00+00:00: Start cooling to -10°C 2022-11-15 11:23:00+00:00: Start spray 2022-11-15 12:13:00+00:00: Accelerate spray 2022-11-15 12:49:00+00:00: End of spray 2022-11-15 13:13:00+00:00: Start heating 2022-11-15 14:14:00+00:00: Start cooling to -10°C 2022-11-15 15:03:00+00:00: Start spray 2022-11-15 15:43:00+00:00: End of spray Ahead of the actual test of 15/11/2022 a Dry-run was done on 27/09/2022 as a preparation to the real test. As the dry-run were just to assure the functionality of the measurement hardware, conditions were not as controlled as during the final test on 15/11/2022 but do include an identical hardware setup. During the dry-run the chamber was also cooled down to -20°C but no icing was applied yet. The steps of the experiment are: 2022-09-27 12:19:00+00:00: Fans started 2022-09-27 13:05:00+00:00: Start cooling to -20°CThe dry-run only lasted for 169 minutes in total. Ten minute acceleration data Experiment 15/11/2022 Twentyfour ten minute samples of the 3 accelerometers on the blade sampled at 250Hz. The data is two 2-hour blocks, one at night before the testing, the second 2 hour block is during the spraying. The 10.1m long blade was instrumented with three tri-axial MEMS accelerometers (Micromega IAC-UHRS-Ud-03, ±3g) on the suction side of the blade. In which the X-direction corresponded to the edgewise motion of the blade, the Z-direction to the flapwise direction and the Y-direction to the less relevant lengthwise motion. The three sensors were installed at approximately 1/4, 5/8 of the blade length and 130cm from the tip of the blade. Dry-run 27/09/2022 A single dataset of the 3 accelerometers on the blade sampled at 250Hz is provided, with the same setup as previously described. Modal parameter data The resulting modal parameter data ( for the entire day of testing) , in both X and Z direction are provided in MO04_mpe_*_20221115.csv. The data has following shape: mean_frequency std_frequency mean_damping std_damping size algorithm timestamp 0 1.649219 0.001952 1.128411 0.103419 51 lscf 2022-11-15 00:00:00+00:00 1 2.028879 0.000927 0.432596 0.068354 60 lscf 2022-11-15 00:00:00+00:00 2 2.923999 0.009472 2.677156 0.615895 19 lscf 2022-11-15 00:00:00+00:00 3 3.779268 0.003649 3.462347 0.715628 7 lscf 2022-11-15 00:00:00+00:00 4 5.786029 0.005541 0.539010 0.261049 7 lscf 2022-11-15 00:00:00+00:00 In which `mean_frequency` and `std_frequency` are the cluster mean frequency, are the cluster std. frequency and cluster std. damping and the cluster size (size). The LCSF algorithm is used. The algorithm used is described in Source. Note: multiple rows share one timestamp, this is because the algorithm can detect multiple modes per timestamp. Temperature data The inflow air temperatures are shared in a separate .csv files: For the main experiment: ClimateChamber_20221115.csv, For the dry-run: ClimateChamber_20220927.csv Pictures Picture are collected during the test are shared. Each picture is timestamped in local time (UTC+1) Conference Presentation The presentation held during the Eurodyn 2023 conference is also made available.