Ross Poulton
Project Title
Flow Control (may be changed as project progresses).
Aims
With global warming and limited supply of fossil fuels being some of the greatest challenges facing humanity in the near future, the prominent feature of modern engineering design should be efficiency. Reducing drag provides the oppurtunity to reduce losses enormously and for a large number of applications the prominent type of drag is skin friction.
Drag along solid boundaries in fluid flow is proortional to the velocity gradient normal to the wall. Vortices in turbulent flow move high velocity fluid from far away closer to the wall, increasing the velocity gradient and thus the drag. Wall jets provide the ability to decrease skin friction by actively countering vortices near the solid boundary. Wall-normal velocity can be detected in a plane near to the wall and a jet at the wall acts to actively oppose the normal components of flow.
The project aims to explore these ideas by systematically investigating the skin friction reduction's dependance on the distance of the detection plane from the wall and the ratio of the detected velocity to the jet velocity.
Detection planes and wall jet velocity are impractical in all but theroetical investigations. Real sensors and actuators introduce random error (e.g. electronic noise) and cannot be placed in the same position as they could with CFD grid points. They must all be embedded into the wall, and as such correlations between theoretical detection plane velocities and either wall pressure or wall shear stress must be used to detect vortices. Looking at how practically plausible arrays of transducers affect the controllability of the flow is therfore the secondary objective of the project.
The code being used for the project is based on channel flow between two infinite plates using approximately 1 million grid points and a turbulent flow model. This restricts observations to fully developed flow, so it would also be beneficial to find how such results can be applied to external flow where the optimum detection plane would be permanently moving away from the wall.