A method for drag reduction on bluff bodies

This thesis presents an experimental investigation on the drag of typical two-dimensional bluff bodies with a small circular cylinder (here called a "rod") placed upstream on the stagnation line. In most cases, the interaction was beneficial and the drag of the overall system was reduced. Fluctuating side force due to vortex shedding from the main body was also investigated for various rod positions and diameters. Two typical bluff bodies were investigated: a circular cylinder and a flat plate. The front rod diameter (d) varied from 0.17D to 0.5D where D is the frontal width of the main circular cylinder or flat plate. Reynolds numbers defined using dimension D were in the range of 1x10⁴ to 7x10⁴, and the longitudinal spacing L from the rod centre to the main body was 0.4 < L/D < 7.0. For both the flat plate and the circular cylinder, the measured overall drag coefficient at various L/D showed a discontinuous "jump" at some critical spacing. This change corresponds to the elimination of the usual single stagnation point on the bluff body centre line and the appearance of two stagnation points symmetrically placed, close to the lateral edges of the body. For the flat plate, the optimum configuration was found to be d/D = 0.33 at a spacing L/D of 1.81, and the overall drag reduction (based on the drag of the plate alone) was 36%. The drag reduction was due to a change in front pressure only, the base pressure remaining essentially constant and independent of L/D. No Reynolds number dependence was found, as expected. For the circular cylinder, the best overall drag reduction was found to be 58% for the configuration d/D = 0.33 and L/D = 1.73. Part of the drag reduction on the cylinder was due to the rod wake being turbulent which made the flow over the main cylinder critical, with a lower value of base pressure coefficient. The second effect was a sudden change in the front pressure distribution similar to that observed on the flat plate. Unlike the flat plate, however, the flow over the circular cylinder was sensitive to Reynolds number, with decreasing CD for increasing Re. The fluctuating side force (CI') on the cylinder due to vortex shedding was also investigated. The variation of CI' with spacing followed different trends depending on the rod size and Reynolds number but CI' never increased significantly with the use of the front rod and was reduced in most cases. The rod d/D = 0.17 seemed to be more suitable for minimum fluctuating side force. The position for Cl' minimum was found to depend upon the particular configuration.Applied Science, Faculty ofMechanical Engineering, Department ofGraduat