Session Q23: Flow Control II

Boundary layer control using a dynamic free-slip surface

Recently we discovered that an oscillating free-slip surface attached to the solid surface in a turbulent boundary layer can reduce the wall shear stress for more than 40% (Wang & Gharib 2020). Evidence suggests that the significant drag reduction effect is linked to the streaming motion (e.g. a laminar jet) induced by the dynamic oscillation (Wang & Gharib 2021). To better understand the physical mechanism of the drag reduction effect, in this work, we investigate the dynamic interaction of a bounded free-slip surface with a laminar boundary layer. The velocity/vorticity field suggests that the near-wall transverse vorticity is gradually lifted up by the accumulative effect of the dynamic oscillation. In addition, well-organized counter-rotating streamwise vortices are created and sustained in the downstream regions. The lift-up of transverse vorticity and the formation of patterned streamwise vortices can explain the strong drag reduction effect of the dynamic free-slip surface.     

Underlying Drag Reduction Mechanisms of Slip Surfaces: Temporal Approach

Turbulent flow control is important in fundamentals and applications due to the potential benefits, particularly regarding drag reduction for energy savings. In this study, we will investigate the slip control strategy to better understand its underlying mechanisms of drag reduction in turbulent channel flows. Direct numerical simulations are performed at low Reynolds numbers with the inclusion of the slip surface. The temporal analysis is exploited to elucidate the underlying drag-reduction mechanisms. Two temporal phases are classified, depending on wall shear stress. Based upon the lifetime events of intermittent phases, periods of high and low wall shear stress are represented by super active and hibernating phases, respectively. The slip surface plays a distinct role in drag-reduction mechanisms with regard to these two phases. As the slip length increases, the slip flow causes the hibernating phases to occur more frequently with a decrease in the duration between each successive interval, which is intuitive for drag reduction. For the super active phases, however, both the duration of super active intervals and the time between these intervals increase. As a result, the total fraction of time spent in the super active phase increases, which is counter-intuitive for drag reduction. A detailed analysis will be given to elucidate this counter-intuitive behavior.   

3D-printed Models Inspired by Sinusoidal Dolphin Skin Geometry to Control Flow Separation

The dolphin is among the fastest marine animals, and therefore has attracted the research interest of both biologists and engineers. A unique feature of dolphin skin is the rows of transverse grooves that resemble a wave-like geometry circumferentially covering a large portion of their body. It is hypothesized that the subtle sinusoidal grooves create a partial slip condition over embedded vortices within cavities while inducing mixing that increases the flow momentum in the boundary near the surface thereby acting as a passive mechanism to delay flow separation to reduce pressure drag while swimming. Previous research showed that sinusoidal grooves, compared to rectangular, have a less intrusive streamlined shape to achieve the desired. This study investigates the effect of geometric amplitude for 3D printed sinusoidal grooved models dimensionally matched to the wavelength observed on dolphin skin.  DPIV experiments measured boundary layer profiles for the various surface geometries as well as a smooth plate within a separating turbulent boundary layer to determine the degree of separation control capability for each case.   

Wind tunnel testing of 3D printed shark scale surface for separation control

Low speed water tunnel studies have obtained favorable results of separation control with real shark skin and 3D printed models. However, the thinner boundary layers formed in wind tunnel studies have proved challenging for testing shark skin models in air. Sizing of shark scales within the boundary layer is an important parameter. Previously, the scales actuated to heights of over eight percent of the boundary layer, compared to within five percent in water tunnel studies and three percent as observed in nature. This experiment attempts to alleviate the sizing issue by growing a boundary layer over a long flat plate to a Reynolds number of 1,700,000 so that the smallest possible 3D printed scales (actuation height of 2 mm) will only reach within the bottom five percent of the boundary layer thickness. An array of over 5,000 scales is investigated on the suction side of a trailing edge flap at the end of the flat plate. Scales are locally actuated by reversing flow near the surface of the flap, impeding the development of separation. Force measurements are analyzed to determine if scales of this size are effective as a passive separation control mechanism.   

Experimental Study of Rigidity Effects of Shark Skin Denticle Models on a Turbulent Separation Control

Bio-inspired flow control mechanisms have allowed engineers to learn from nature.  One of these animals is the mako shark, whose scales bristle when met with reversing flow; this may be linked to a passive flow-actuated separation control mechanism.  A study of how the rigidity of these shark scales can affect the separation control mechanism is of interest to better understand the limitations and better design bio-inspired surfaces for separation control.  Using a rotating cylinder, an adverse pressure gradient is induced creating a separated region over several different bio-inspired models resembling fixed shark scales at different angles.  In this experiment, the boundary layer grows to sizes large enough that the scale of the flow is increased, making it more measurable to DPIV.  Additionally, the large boundary layer allows for models to be sized to fit within the bottom 5-10% of the boundary layer.  Plates with model shark scales fixed at 0, 15, 30, and 45 degrees are investigated at a Re in the range of 5 x 10e5.  This data is compared to a smooth flat plate as well as that obtained for passively bristling scales, free to move within the flow to verify how the rigidity of the scales interacts with reversing a flow region.   

Control of Turbulent Boundary Layer Separation by a 3D Printed Shark Skin Model with Passive Bristling

Turbulent boundary layer separation can be problematic in many engineering applications. However, nature may have a solution in the form of flexible shark scales found on the shortfin mako which have proven to passively bristle under reversing flow conditions and control flow separation in past experiments. An investigation of how these shark scales interact with reversing flow in the near-wall regions of the boundary layer is of interest to better understand the fluid-shark scale interactions. Enlarging the geometry and constructing 3D printed models of shark skin is the best route forward to developing a bio-inspired surface for real applications. Using a rotating cylinder above a flat plate in a water tunnel setup, an adverse pressure gradient was induced creating a separated region over a tripped turbulent boundary layer with approximate Reynolds numbers ranging from 4.9×10⁵ to 8.1×10⁵. 3D printed shark scales were mounted into a plate to  replicate low-resistance passive bristling angles of 50 degrees.  Models of shark scales were constructed with crown lengths of 2 mm, ten times greater than those observed on a real shark.  This low speed flow study makes the boundary layer dynamics and shark scale motions more measurable while allowing for actuations heights of the scales to be within the bottom 10% of the boundary layer. Baseline studies document flow separation and reversing flow development in the presence of an adverse pressure gradient over a smooth plate. The same experiments are then repeated with the shark skin models to document control of the flow separation and observe how reversing flow induces scale bristling.   

Harnessing Phononic materials for Unsteady Aerodynamic flow control

Challenges involved with active actuation have motivated the need for flow-control paradigms that can passively adapt to the unsteady flow dynamics. We focus on harnessing phononic materials towards this aim. Phononic materials are architected materials with frequency-dependent characteristics that arise from their periodic and/or resonant microstructure. These frequency-dependent dynamics offer significant potential in the passive control of unsteady aerodynamic processes which inherently contain flow structures with characteristic frequency content. The phononic material is modeled as a nonuniform bi-layer flat plate (i.e., having two periodically repeating materials) using linear Euler Bernoulli beam theory. We perform high-fidelity 2D numerical simulations of flow past the phononic flat plate at an angle of attack of 15 degrees and a Reynolds number of 1000. We identify phononic material parameters that lead to lift/drag benefits compared with a rigid flat plate and characterize the FSI interplay between the vortex shedding phenomena and the plate vibrations. Where relevant, connections between the performance benefits and the phononic material regime (e.g., the bandgap) will be drawn.   

Structural Dynamic Analysis of a Flexible Wing Through Distributed Bleed Flow Control

Controlled flow-structure interactions on a flexible 3-D wing model are employed to induce desired adaptive aeroelastic characteristics. The coupling between regulated aerodynamic loads and aeroelastic properties is explored using controlled distributed bleed actuation that is effected by the pressure differences between the model's pressure and suction surfaces. Wind tunnel experiments are conducted to evaluate the transient response of the wing to controlled load perturbations when bleed ports are opened and closed according to a square-wave schedule. The recorded displacements are employed for the estimation of natural frequencies and damping ratios associated with the first fundamental modes, which are estimated by estimating the logarithmic decrement during transients. Furthermore, periodic excitation to the wing is investigated when the bleed actuation is applied for fundamental periods of actuation that are below, at, or above the first fundamental bending mode and the aeroelastic properties obtained from the transient response are compared with those corresponding to the time-stationary response in the absence and presence of actuation. The present investigations enable estimation of the apparent, bleed-induced changes in the wing's characteristic stiffness, damping, and natural frequency, which provide insights into the aeroelastic performance of the wing with bleed.   

A bio-inspired membrane wing for passive aerodynamic control

We present a novel, lightweight design of an adaptive membrane wing for flapping wing systems. The wing holder is equipped with variable leading and trailing edges which allow for passive cambering of the wing and to align both edges with the surrounding flow. To evaluate the performance of the new design, we conduct experimental optimizations to find the wing kinematics yielding highest aerodynamic lift and efficiency in hovering flight, both for the adaptive membrane wings and for a rigid reference case. The results of the optimization show that the flexible wings reaches higher lift coefficients without increasing the aerodynamic power of the system. This leads to an overall better efficiency of the flexible wings compared to the rigid reference case. The adaptive holder is able to produce the same amount of lift at lower average angles of attack due to the passive cambering of the membrane. We perform additional deformation and velocity flow field measurements for the optimal kinematics and find that the adaptive cambering and lower angles of attack keep the flow attached to the wing longer for the membrane wings promoting the efficiency of the system relative to stiff wings.   

Fractal/multiscale trippings for passive control of turbulent flow separation.

An experimental investigation was performed on the control of turbulent flow separation by fractal/multiscale tripping at the LMFL boundary layer (BL) wind-tunnel. The passive devices have a height equal to 0.125 δ, where δ is the BL thickness. They are located at approximately x = -3.85h upstream of a rounded backward-facing step where h is the step height. The Reynolds number based on h is Reh= 100,000 and the expansion ratio is 1.15, based on the step height h. The recirculation region is studied by 2D2C particle image velocimetry (PIV) in the streamwise-wall-normal (x–y) plane. Time-resolved velocity fields are acquired at 192 Hz allowing the characterization of the unsteady nature of separated flows at high Reynolds number. The baseline configuration (without devices) is compared to three fractal/multiscale trippings, all having the same properties except for the number of fractal/multiscale iteration. All obstacles have the same frontal area. The recirculation region is smaller in size for the fractal/multiscale trippings than for the straight baseline configuration, and the size of the recirculation region decreases as the fractal iteration number increases.