This PhD thesis presents a computational study of turbulent boundary layers and their interaction with wall-attached bodies, by means of high- idelity Direct Numerical Simulations (DNS) and Large Eddy Simulations (LES). The main objective is determining if the interaction with multiscale generated turbulence can potentially suppress, mask or modify the far-field signature originated by a wall-mounted obstacle.
The first part of the thesis is a DNS investigation on the fundamental physics of a zero-pressure gradient turbulent boundary layer at Reynolds numbers up to Re = 2130, based on the momentum thickness and free-stream velocity, which is the baseline flow for later stages of the project. This investigation produced new theoretical results in two main topics: the intermediate region scaling of the Reynolds stresses in wall-bounded flows and the relation between wall-shear stress fluctuations and outer fluctuating velocities in turbulent boundary layers.
Afterwards, the interaction between a solid wall- attached cube and a turbulent boundary layer
is investigated using DNS, at Re = 3000, and LES, at Re = 7900, with Reynolds numbers based on the cube height. The flow structures and frequency response in the near-wall region are in good agreement with reference data obtained in a channel flow. However, our simulations also predict the existence of a low-frequency peak in the far-field energy spectra, which has not previously been reported by other authors.
Finally, the effect of multiscale turbulence on the aforementioned far-field peak is studied with LES in two different configurations: the interaction between two parallel wall-mounted cubes and the interaction between the cube and a regular or multiscale grid. The first option does not offer great potential from an engineering point of view, but the second configuration produces promising results: the grids may reduce the frequency of the peak by around a factor 3. The flexible design of the grids and their inexpensive manufacture could make them excellent candidates for future industrial applications.