Dynamic Response Prediction And Reliability Studies Of A Wind Turbine Tower Under Near-field Ground Motions Excitation

Based on the national advocacy for green energy and sustainable development,the wind power generation is widely used as a mature wind energy utilization technology.With the gradual development of excellent wind farms,some wind farms have been established in near fault areas,such as the construction of large wind farms in the western areas of China and the coastal region of Fujian.The near-field ground motions usually contain velocity pulses and strong vertical components,which is easy to produce resonance-like effect with structures with long natural vibration period.Therefore,the impact of near-field effect on the construction of large-scale wind turbine tower can not be ignored.In this study,the effects of velocity pulse and strong vertical acceleration contained in near-field ground motions on the seismic performance of 2MW wind turbine tower are studied by means of experimental research,numerical simulation and theoretical analysis.The characteristics of near-field ground motions,the dynamic time history analysis,the dynamic stability,the collapse analysis and vulnerability analysis are carried out.This paper also carries out the random seismic response analysis and reliability analysis of wind turbine tower under the artificial synthetic pulse-type near-field ground motions.Finally,the dynamic response prediction analysis of wind turbine tower structure is carried out based on particle filter method.The main research work of this paper is as follows:1.The dynamic characteristics,acceleration response,displacement response,strain response,shear force and axial force of a 1/20 scale wind turbine tower model under the excitation of two pulse-type near-field ground motions and two ordinary ground motions are obtained through the shaking table test.The effects of peak ground acceleration amplitude(PGA),velocity pulse effect,vertical component of ground motions and the vertica to horizontal acceleration ratio(V/H)on the structural dynamic response are studied.The results show that the wind turbine remains in elastic stage during this test,and the structural dynamic response is mainly based on the structural basic mode.The PGA value and the velocity pulse have significant effects on the structural dynamic response.The horizontal dynamic response is not sensitive to the vertical component of ground motions,but the vertical dynamic response is particularly sensitive to the vertical component of near-field ground motions,and this sensitivity increases with the increasing of V/H ratio.2.Based on the ABAQUS software,the finite element model of the 2MW wind turbine tower prototype structure is established and then verified.The influence of velocity pulse is evaluated through the structural dynamic response under 30 typical ground motions.The influence of vertical component of near-field ground motions is analyzed through the structural dynamic response under 7 groups of near-field motions with significant V/H ratio.The results show that the frequency spectrum characteristics of pulse-type near-field ground motions are significantly different from those of non-pulse ground motion,which may have a great impact on structures with long natural vibration period.The mean value of structural dynamic response under pulse-type near-field ground motions is more than 1.5 times that under non-pulse ground motion,and the near-field ground motion with forward directional effect has a more significant impact on the structural dynamic response.Thus,during the seismic design of wind turbine tower in medium and high earthquake area,neglecting the influence of velocity pulse and vertical component of near-field ground motion may significantly underestimated the structural seismic responses.3.The structural ultimate buckling stress are calculated based on the GL and DIN 18800 codes.Combined with the dynamic instability criterion,the effects of velocity pulse and vertical acceleration of near-field ground motions on the structural dynamic stability are evaluated.The incremental dynamic analysis under 25 groups of ground motions is carried out to study the buckling and collapse characteristics of this tower.The results show that the bottom of this structure is prone to instability under near-field ground motions.Pulse-type near-field ground motions can obviously amplify the displacement at the tower top and reduce the seismic input amplitude required for the structure to produce plastic hinges.The failure mode of wind turbine tower structure is single.Once the plastic yield-line is formed,the tower will collapse.4.The pulse-type near-field ground motions is synthesized by the MP velocity pulse model and high-frequency component model.The effects of pulse parameters on the structural response are studied and compared.The dynamic responses subjected to 100 random near-field ground motions are analysed.The structural reliability analysis is carried out by using the improved subset simulation method based on Kriging model.The results show that the synthesized pulse-type ground motions can truly reflect the spectrum characteristics of pulsetype near-field ground motions.The peak value,pulse period and pulse number of the velocity pulse are positively correlated with structural dynamic response.The probability density curve of dynamic response has obvious random fluctuation.The probability of structural instability failure is slightly greater than the probability of strength failure.5.Based on particle filter method,the structural dynamic response prediction under the Bayesian framework is carried out.The rationality and effectiveness of this framework can be verified by the dynamic response observation results of wind turbine structure through the shaking table test.The results show that the combination of particle filter method and experimental values can effectively predict the dynamic response of structure.The particle filter method has strong prediction stability and adaptability,which has high research value and wide application perspective for the nonlinear response and unmeasurable state estimation.