Effect of strong disturbances on the evolution of turbulent boundary layers

Abstract

This thesis describes an experimental investigation into the evolution of artificially generated high Reynolds number turbulent boundary layers (TBL). Due to its large importance on TBL scaling, skin friction has to be accurately determined. With this purpose, a robust post-processing method is developed to extract the mean skin friction and the wall-probe relative position from the mean velocity profile with uncertainties better than 1% and 0.5 wall units respectively.

For disrupted TBLs, it is shown that, after a certain disturbance, TBLs evolve towards a canonical state after an adaptation region whose turbulent properties and length are strongly dependent on the trips' geometry. Two distinct mechanisms (so called wall-driven and wake-driven) are identified and associated with shorter and longer adaptation regions respectively. The latter is generated by disruptions exhibiting strong flow recirculation which enhances the influence of the obstacle's wake on the near-wall region and compromises the TBL properties in the adaptation region. Contrastingly, trips generating a wall-driven mechanism prevent this interaction from happening thus enabling a distinction between the trips' wakes and the near-wall region. Particle image velocimetry and low-order modelling of the flow in the close vicinity of the obstacles enable us to establish a three-way link between the main geometrical features of the trips, the length of the adaptation region and its turbulent properties.

Low-porosity wall-mounted single- and multi-scale fences are designed and tested to control the degree of interaction between their wake and the near-wall region. Turbulent properties in the near-wall region are shown to scale with the local thickness of that internal layer rather than with the thickness of the whole fence's wake. Further, an aero-acoustic characterization of the flow is conducted showing the spatial distribution and the velocity scalability of the noise sources. Finally, some topics for further work are proposed.