Separated flows – such as those around bluff bodies – can be greatly improved by removing some of the fluid through bounding surfaces. This ‘suction flow control’ reinvigorates the boundary layer, delaying separation and altering the pressure field of the flow. In this research, suction and blowing control was optimised for two representative separated flows: the flow around the circular cylinder (external flow), and the flow through a conical diffuser (internal flow). The aim was to progressively develop from broadly-applied uniform suction to refined and passively generated (autogenous) non- uniform suction and blowing. These flows were investigated using numerical simulations employing the Finite Element Method.

It was found that non-uniform suction of the boundary layer was always more efficient, and usually more effective, than uniform suction in the representative flow-cases. Separation could be entirely eliminated around the cylinder when sufficient suction control is applied, and there is a compelling relationship between the optimal control parameters and the uncontrolled separation values. Drag on the cylinder could be reduced by over 30% and performance of the diffuser could be increased by over 50% in the investigated Reynolds number ranges. Combining suction with blowing control produced even better performance in most circumstances, especially if a non-zero mass flux was permitted. Constraining the control so that the flow-rates were balanced (Q-balanced) and that a positive pressure gradient from suction-to-blowing loci is present (P-Q-balanced) allowed for the design and testing of potentially autogenous suction control. For the 5◦ diffuser, this control arrangement was unable to improve performance due to the monotonically increasing pressure profile. For the cylinder at Re = 40 and Re = 120, however, it was capable of reducing the drag compared to the uncontrolled case (reduced by ∼ 5%. In unsteady simulations (at Re = 120) this also reduced fluctuations in the flow.

When considering a practical implementation, with ducting and porous materials to produce the connection between suction/blowing loci, additional losses are present that are not accounted for by assuming that dP > 0 is sufficient for autogenous control. Numerical tests of promising dual-loci control were performed with geometric design concepts applied to the cylinder. At Re = 40 the drag was reduced with these designs, but not at Re = 120 because the desired suction profile was not appropriately produced. Suggestions for overcoming this issue are detailed.

Overall, autogenous suction control has been proven to be a feasible method for reducing drag on bluff body flows, and can offer a modern tool for improving the efficiency of vehicles for a future “net-zero” world.