Modelling of active flow control devices using hybrid RANS/LES techniques

The focus of the present thesis is on the effects of two active flow control devices on the periodic components of the turbulent shear layers and the Reynolds stresses. One of the main aims is to demonstrate the capability to control individual structures that are larger in scale and lower in frequency against the richness of the time and spatial scales in a turbulent boundary layer. In order to carry out this investigation, computational fluid dynamics CFD simulations are performed. The turbulence modelling approach for the two dimensional initial cases is RANS and URANS and with regards to 3D simulations IDDES, a hybrid RANS/LES technique, is applied. The geometry for the studies is taken from experimental configurations for each case; both cases comprise a turbulent flow over a backward facing step (BFS), where separation is induced after the step edge. The results from the simulations are compared to the experimental data for both cases with and without control. The first active flow control device is a single DBD plasma actuator located upstream of the step. The effects of quasi-steady and unsteady – or pulsated- plasma actuation using two different phenomenological models are studied. The resulting turbulent structures, Reynolds stresses, skin friction and velocity profiles are analysed applying the aforementioned models to simulate the plasma actuation. The results for quasi-steady plasma mode show very good agreement with the available experimental data and a reduction of the reattachment length which matches the experimental data is observed. Regarding modulated actuation of the DBD plasma device, three dimensional simulations were carried out and the results also showed excellent agreement of the overall behaviour flow when compared to the experimental data. The second flow control device is a novel device known as spanwise vortex generators. It consists of a strip of magnets placed along the span of the BFS upstream of step and the device oscillates at a given frequency and amplitude. Like for the first control device, turbulent structures, Reynolds stresses, skin friction distributions and velocities are analysed and compared to the experimental measurements. A remarkable effect of the device is observed especially in the reattachment length which is considerably reduced. Experimental measurements for the baseline case were available and a comparison with such data is performed.