Assessment of Fatigue and Extreme Loading of Wind Turbines
Authors
Jóhannsson, Halldór Páll ; Søndergaard, Anthon de Luna
Term
4. term
Education
Publication year
2016
Submitted on
2016-06-09
Pages
81
Abstract
Dette speciale undersøger trætheds- og ekstremlastpåvirkninger af vindmøller i henhold til IEC 61400, med fokus på designlasttilfælde 1.2 og 1.3. Arbejdet kombinerer målte turbulensdata fra en specifik lokalitet med aeroelastiske simuleringer i FAST for at kvantificere, hvordan klimatiske parametre påvirker lastresponsen. De målte turbulensdata blev tilpasset både lognormal- og Weibull-fordelinger for at vurdere, hvilken beskrivelse der bedst repræsenterer forholdene; analysen viste, at Weibull var den bedste tilpasning. Der er gennemført verifikationer af, at FAST’s simulerede respons stemmer overens med inputparametrene, herunder konvergens- og stabilitetsstudier og en sammenligning mellem FAST version 6 og 8; overordnet set opførte FAST sig som forventet. For NREL 5 MW-møllen er der udført systematiske loadsweeps, hvor turbulensintensitet, vindskæring og lufttæthed blev varieret for at udvikle responsmodeller afhængige af vindhastighed. Resultaterne viser, at turbulensintensiteten har den største indflydelse på lastresponsen sammenlignet med vindskæring og lufttæthed; dette understøttes af en følsomhedsanalyse, som peger på, at turbulensintensitet bidrager mest til variansen i responsmodellerne. Til sammenligning er NREL 5 MW-møllen benchmarked mod DTU 10 MW-møllen, hvor responsens overordnede adfærd er ensartet, om end med visse uoverensstemmelser. Endelig gennemføres loadsweeps for DTU 10 MW i DLC 1.2 og der opbygges responsflader og -modeller, som kan understøtte vurdering af trætheds- og ekstreme laster.
This thesis assesses fatigue and extreme loading of wind turbines in accordance with IEC 61400, focusing on design load cases 1.2 and 1.3. The work combines measured site turbulence data with aeroelastic simulations in FAST to quantify how climatic parameters affect load response. The measured turbulence data were fitted with both lognormal and Weibull distributions to identify the most suitable representation; the analysis found that the Weibull distribution provided the better fit. Multiple checks were performed to verify that FAST’s simulated outputs correspond to the specified inputs, including convergence and stabilization studies and a comparison between FAST versions 6 and 8; overall, FAST behaved as expected. For the NREL 5 MW turbine, systematic load sweeps were conducted by varying turbulence intensity, wind shear, and air density to develop response models as functions of wind speed. Results show that turbulence intensity has the largest effect on load response compared to wind shear and air density; this is reinforced by a sensitivity analysis indicating that turbulence intensity contributes most to the variance of the response models. The NREL 5 MW turbine was benchmarked against the DTU 10 MW turbine, revealing broadly similar response behavior with some inconsistencies. Finally, load sweeps for the DTU 10 MW turbine under DLC 1.2 were performed and response surfaces and models were developed to support fatigue and extreme load assessment.
[This summary has been generated with the help of AI directly from the project (PDF)]
Documents