Investigation of low-speed boundary-layer instability and transition using experiments, theory and dns

A collaborative study of boundary-layer instability and transition has been conducted using wind tunnel experiments, direct numerical simulations (DNS) and linear stability theory (LST). Experiments are performed in the Arizona Low Speed Wind Tunnel (ALSWT) at approximately 7 m s⁻¹ (unit Reynolds number of 4.2 × 10⁵ m⁻¹ ). The ALSWT test section velocity is adjustable through variation of the fan RPM and blade angle. A parametric study yielded minimum turbulence intensities of approximately 0.022% (1 Hz-10 kHz) over a range of RPM and blade angles. This low turbulence intensity was accompanied by benign spectral content primarily associated with blade passage frequency which could be moved outside of the range of interest for boundary-layer stability studies. Installation of a flat plate and corresponding support structure increased turbulence intensity to 0.028% (1 Hz-10 kHz) but had no negative influence on stability experiments. Precursor calculations of the test section flow field were performed with ANSYS Fluent to provide boundary conditions for subsequent high-resolution DNS in the region of interest on the flat plate downstream of the leading edge. A detailed comparison between experiments and DNS shows good agreement between boundary-layer velocity profiles. Subtle discrepancies between experiments and DNS are attributed to the existence of slight favorable pressure gradient (β_H = 0.015) in the experiment, which results in slightly fuller velocity profiles and a more stable boundary layer. Experiments investigating the (linear) disturbance development are performed using alternating current dielectric barrier discharge (ac-DBD) plasma actuators. A linearized Navier-Stokes solver was employed to map out the linear stability region which then guided the actuator location (Re_x = 10⁵) and forcing frequency (50 Hz) to allow observation of both growth and decay of the disturbances over the extent of the domain. The measured growth of the disturbances is in good agreement with and DNS results for both the first and second maxima of the u-velocity amplitude distribution and any discrepancies can be explained by consideration of the small pressure gradient and nonzero free-stream turbulence in the experiment which is not considered in DNS. Finally, branch one and branch two neutral curves for a range of ac-DBD plasma forcing frequencies (30 Hz-100 Hz) are identified from experimental data which compare well with LST. Overall, these results show the utility of the ALSWT as a test bed for boundary-layer stability and transition research.