Skip to main content

Micro-textured Surfaces for Boundary Layer Control

Final Report Summary - MTSBLC (Micro-textured Surfaces for Boundary Layer Control)

In the bypass transition scenario, free-stream disturbances buffet the boundary layer. The response is the formation of Klebanoff streaks. Several mechanisms for spot formation due to these streaks have been identified. The secondary instability of low speed streaks via interaction with free-stream disturbances (Zaki and Durbin 2005) or local shearing between adjacent streaks (Nagarajan et al. 2007) The simulations of Nagarajan et al. (2007) included leading edge geometries and the above mechanisms were observed for sharp and blunt leading edges, respectively. These observations were explained by the secondary instability analysis of Vaughan and Zaki (2011).

Regardless of the mechanism which leads to breakdown, it is clear that not all streaks undergo secondary instability. In addition, not all potential breakdowns will contribute to a unique turbulent spot as breakdowns which occur upstream can quickly consume surrounding fluid. Since the number of turbulent spots is small compared to the number of streaks, what differentiates these rare events is masked by time averaged statistics. By performing conditional sampling on instantaneous velocity fields, the flow may be separated into its constituent laminar and turbulent components. Furthermore individual structures can be identified and tracked, and their role in the transition process elucidated.

Many transition intermittency-based models, such as the popular Dhawan and Narasimha (1958) model, assume a concentrated breakdown of the flow at a single streamwise position. Transition is therefore governed by the rate of spot generation per unit span and the spreading angle of the resulting spots until the boundary layer is saturated. This simple but effective model therefore only requires the location of transition onset and the spot propagation and generation rates. However experimental studies have shown that concentrated breakdown is an oversimplification and that prediction of the location of turbulent spot inception and production rate are important problems in understanding transition to turbulence.

In this work we exploit the wealth of data than can be extracted from direct numerical simulation (DNS). Conditional sampling is performed to identify regions of laminar and turbulent flow in a database of velocity fields from DNS. Within the laminar regions, individual streaks are identified and tracked in space and time. Those streaks which are observed to result in a localised breakdown to turbulence are contrasted with the full streak population. Finally the resulting turbulent spots are also tracked from inception and their growth rates are recorded. The same methodology has been applied to flows with various wall boundary conditions, some presenting smooth surfaces, others including the effect of wall heating and surface textures.