ANGLE DEPENDENCE OF TOWER WAKE DISTORTION: A PARAMETRIC STUDY

The controlling parameters (thrust coefficient (� ) and the tower leg lengths (� )) found in the centreline speed deficit � � expression provided in the International Electrotechnical Commission (IEC 61400-12-1) standard which derivation was based on the reference direction (� = �°) are angle dependent. It is therefore crucial to evaluate and incorporate values of these parameters into the centreline speed deficit expression to account for tower wake distortions at different incident wind angles. In this research, the variation of these parameters with the incident wind angles were investigated using two lattice equilateral triangular communication towers of different configurations, bel onging to the Mobile Telecommunication Limited (MTC) of Namibia, located at Amper-bo and Korabib. Physical modelling and computational approaches were utilized to establish the values of these parameters at incident wind angles other than the IEC reference direction. For both towers, deducting a range 2 % to 4 %, 9 % to 13 % and 15 % to 20 % out of the IEC reference value (� = �°) will accurately estimate � values for winds that arrive at the tower at 10°, 20° and 30°, respectively. An � expression of � for the tower at Amper-bo is � = �. ��(� − �)� and that at Korabib is � = �(� − �)�, provided winds � � � arrive at the towers at angles ≤ ± 70° with respect to the IEC reference direction (� = �°). For winds that arrive at the tower at angles ≥ ± 70°, a boom length of � ≥ �. �� and � ≥ �. �� at Amper-bo and Korabib respectively is enough to keep the speed sensors out of the tower wakes. Deducting approximately 1.5 %, 5 % and 12 % less of the IEC reference value will reliably estimate � values for winds that arrive at the tower at ±10°, ±20°, ±30° and ± 40°, respectively. Again, � physical modelling approach enables the computation of the solidity ratio (�) to an accuracy of 0.001 mm, showing an increase in the blockage to wind flow and corresponding increase in boom lengths due to the presence of secondary support structures, while the computational approach illustrates a more complex flow interference around the tower with no appreciable increase in boom lengths. Finally, the thrust coefficient (� ) and the leg length (� ) obtained at different �� � incident wind angles were incorporated into the IEC centreline velocity deficit expression to arrive at a modified speed deficit expression used to predict the booms lengths at incident wind angles other than the IEC reference direction(� = �°).