Stochastic Analysis of the Influence of Tower Shadow on Fatigue Life of Wind Turbine Blade

Fatigue damage accumulation in upwind turbine blades is primarily influenced by turbulence in the inflow. However, the stress reversals during blade passages through the stagnating and deflected mean wind field in front of the tower also contributes significantly. In the paper the lower order statistical moments of the fatigue life of a blade are estimated and compared for a turbine with a tripod tower and a standard mono-tower, respectively. The stagnation zones for each of the legs of the tripod are narrower than for the mono-tower, and hence the stress reversals will be comparable smaller. The blade stresses are calculated from a dynamic mechanical model based on a two dynamic degree of freedom with quasi-static correction for higher modes. The self-induced aero-elastic loading and the turbulence loading are modeled by means of a quasi-static model linearized around the operational point, ignoring any memory effects on the load coefficients. However, such memory effects are taken into consideration at the calculation of the aero-dynamic load during tower passage by the use of a rational approximation to the relevant indicial function. Based on Monte Carlo simulations it is demonstrated that the expected damage accumulation per unit of time in the turbine blades are reduced significantly for the tripod when compared to the damage in a comparable mono-tower design.Fatigue damage accumulation in upwind turbine blades is primarily influenced by turbulence in the inflow. However, the stress reversals during blade passages through the stagnating and deflected mean wind field in front of the tower also contributes significantly. In the paper the lower order statistical moments of the fatigue life of a blade are estimated and compared for a turbine with a tripod tower and a standard mono-tower, respectively. The stagnation zones for each of the legs of the tripod are narrower than for the mono-tower, and hence the stress reversals will be comparable smaller. The blade stresses are calculated from a dynamic mechanical model based on a two dynamic degree of freedom with quasi-static correction for higher modes. The self-induced aero-elastic loading and the turbulence loading are modeled by means of a quasi-static model linearized around the operational point, ignoring any memory effects on the load coefficients. However, such memory effects are taken into consideration at the calculation of the aero-dynamic load during tower passage by the use of a rational approximation to the relevant indicial function. Based on Monte Carlo simulations it is demonstrated that the expected damage accumulation per unit of time in the turbine blades are reduced significantly for the tripod when compared to the damage in a comparable mono-tower design