The flow separation delay in the boundary layer by induced vortices
- 817 Downloads
A series of experiments involving the particle image velocimetry technique are carried out to analyse the quantitative effectiveness of the synthesized vortical structures towards actual flow separation control. The streamwise vortices are synthesized from the synthetic jet actuator and introduced into the attached and separating boundary layer developed on the flat plate surface. Two types of actuators with different geometrical set-ups are used to analyse the evolution of vortical structures in the near wall region and their impact towards achieving separation delay in the boundary layer. First, a single circular jet is synthesized by varying actuator operating parameters and issued into the boundary layer to evaluate the dynamics of the interaction between the vortical structures and the near wall low momentum fluid in the separated region. Second, an array of jets has been issued into the artificially separated region to assess the effectiveness of various vortical structures towards achieving the reattachment of the separated flow in the streamwise direction.
KeywordsBoundary layer Flow separation Streamwise vortices Synthetic jet actuator
List of symbols
Diameter of cavity or orifice, mm
Dimensionless stroke length
Characteristic velocity, m/s
Normal distance from wall, mm
Angle of attack
Molecular kinematic viscosity, m2 /s
Surface shear stress, N/m2
The streamwise inclined plate length
Diaphragm time period
Reynolds number based on dimensionless stroke length
Diaphragm oscillation frequency, Hz
Streamwise distance, mm
Spanwise distance from the middle orifice
Boundary-layer thickness, mm
Fluid density, kg/m3
Space averaged jet velocity, m/s
The streamwise distance on inclined plate
Peak-to-peak diaphragm displacement, mm
Synthetic jet actuator
Tilted vortex rings
Stretched vortex rings
Particle image velocimetry
On an aircraft wing, the flow separation occurs when the boundary-layer travels far enough against an adverse pressure gradient. The velocity of the particles nearest to the surface falls almost to zero. The boundary-layer flow becomes detached from the surface and instead takes the forms of eddies and vortices and results in enlarged drag, particularly pressure drag. This paper explains the novelty of the synthetic jet actuator (SJA) towards achieving the reattachment of boundary-layer separation and, hence, flow separation control on aircraft wings.
By employing the SJA, the jet is synthesized from the ambient fluid by a forced periodic excitation of the diaphragm to produce a train of consecutive streamwise vortices that interact with the boundary layer to exchange momentum with the relatively less energetic near wall fluid. The vortical structures are produced over a broader range of length and timescale, and their exclusive attributes make them attractive fluidic actuators for a number of flow control applications. Smith et al. (1999) and Mallinson et al. (1999) used piezoelectric diaphragm actuators, since they are easy to build and could be operated over a broader range of frequency. In this work, the piston type actuator is used, as the main focus is on the evolution of the vortical structures and their subsequent interaction with the boundary layer.
Amitay et al. (2001) demonstrated the suppression of separation on an unconventional symmetric aerofoil. The control jet is synthesized from two side-by-side rectangular actuators parallel along spanwise dimensions. The Reynolds number was calculated based on the chord profile and was kept in the range 3.1 × 105 to 7.25 × 105. For inactive actuators, the aerofoil stalled for α > 5°, and when the actuators were active, fully attached flow was achieved for up to α > 17.5°. As a result of flow reattachment, or delay of separation, a substantial 100 % increase in lift was achieved and up to 45 % decrease in pressure drag was observed. For α > 17.5°, the lift was increased, however, the enhancement in lift was also accompanied by an increase in drag. Tensi et al. (2002) used piston type actuator with a slot opening on a cylindrical surface and observed that the separation line had been considerably pushed back downstream when the actuator was active. In the smoke visualization experiments, Gilarranz and Rediniotis (2001) used similar slot-exit actuator deployed on the upper surface of NACA0015 aerofoil with varying angle of attacks. At 20° angle of attack and with the actuators inactive, the separation occurred at the leading edge, and when the actuators were active, the separation line was pushed back downstream up to 70 % of chord length.
Certainly the effectiveness of synthetic jet actuators towards flow separation control has long been recognized in laboratory experiments. However, the fluid dynamics involved in the interaction of vortical structures with the boundary layer ultimately causes the reattachment that is yet far from fully understood. The flow visualization by Ishtiaq and Zhong (2014) and the liquid crystal and PIV investigations by Jabbal and Zhong (2010) revealed different types of vortical structures formed under varying operating parameters when the actuator is deployed in the cross-flow zero-pressure gradient boundary layer. Jabbal and Zhong (2010) were able to come up with some quantitative analysis based on liquid crystal and PIV measurements and concluded that stretch vortex rings (SVR) were the most desirable structures to delay separation line. Zhong et al. (2005) proposed, however, without any quantitative analysis, that hairpin vortices (HP) were the most effective structures, since they produced two parallel streaks of enhanced shear stress as a result of counter rotating legs. On the other hand, the most recent PIV measurements (Ishtiaq and Zhong 2013) carried out on the spanwise plane revealed that the lateral wall shear enhancement was the maximum when the tilted vortical rings (TVR) were issued into the boundary layer.
Therefore, it is essential to carry out further investigation into the fluid dynamics involved in the interaction of varying vortical structures with the near wall fluid. For more reliable and authentic results, the effectiveness of the various types of vortices needs to be evaluated in the actual artificially produced separation region. Furthermore, it has been observed that the SJA operating parameters seem to vary to produce similar vortical structures under similar free stream conditions.
2 Experimental set-up
Along the streamwise plane (x–y) normal to the plate surface, a laser light sheet is produced through the glass-bottomed floor of the test section, whilst positioning the camera, such that its optical axis is facing the side wall of the flume and the laser sheet. The suitable laser light sheet with desired thickness and width is formed using a combination of cylindrical and spherical lens and was directed to the required area on the test plate using a 50 mm-thick laser mirror mounted at 45° on a support. For safety purpose, the ends of the laser head and sides of the glass flume were enclosed and fully covered using black paper to avoid any unwanted laser leakage.
The smallest frequency used in the experiments is 1 Hz which is at least two times greater than Tollmien–Schlichting wave, so the effect of SJA does not influence the stability of the boundary layer. However, the stability is confirmed by calculating the indifference Reynolds number to be ≈383 which is not exceeding the minimum value of 520.
3 Results and discussion
3.1 Test conditions
3.2 Vortex impact on the boundary layer
For four phase points evenly distributed over the actuation cycle, the profile is shown in Fig. 4 for tilted vortex ring. Over the complete course of actuation cycle, the velocity increment near the wall region remains fairly enhanced suggesting a non-intermittent influence. At t/T = 0.25, where ‘t’ is the time from start of actuation cycle, the deficit is quite significant caused by the downstream branch of the vortex and the deficit region is far away from the wall, as the vortex has had travelled far away from the wall surface by virtue of larger stroke length. The abrupt decrement from y/δ = 0.5 to 0.78 is due to the upper side of the downstream branch having opposite sense of rotation relative to the local velocity. Similarly, the lower part of the downstream branch exhibits the same sense as the local velocity hence suggesting a far steeper velocity gradient from y/δ = 0.78 to 1.05. Such a gradient subsequently results in the velocity over shoot in the region outside the boundary layer.
At t/T = 0.50, the upstream branch approaches the measurement area and it displays clockwise vorticity having similar sense in the upper part as the local velocity and the opposite sense in the lower part. From y/δ = 0.4 to 0.65, the abrupt velocity increment is due to the upper part of the upstream branch. However, the lower part of the upstream branch decelerates the local velocity from y/δ = 0.65 to 0.85. Further deep towards the outermost part of the boundary layer, the velocity increment is due to the upwash of the fluid. At t/T = 1, the vortex of the previous cycle has passed the measurement plane, as the newer structure approaches. There appears a velocity increment near the wall region and then a decent decrement at about y/δ = 0.2 caused by the passing trailing secondary and tertiary vortices. Second, the velocity gradient near the wall region is larger in the case of tilted vortices than the stretched and hairpins, where the velocity gradient is generated by the upwash of the fluid resulted from the counter rotating legs. On the other hand in TVR’s, when the primary vortex has had moved away from the wall, the near wall velocity gradient could still be seen which is far significant and is generated by the trailing counter rotating tertiary vortices inducing a low momentum fluid downwash towards the wall. The tertiary vortices are, in fact, the trailing streamwise vortices with the opposite sense of rotation to secondary vortices thus result in downwash of the fluid. The appearance of secondary and tertiary vortices is more evident in the dye visualization technique (Ishtiaq and Zhong 2014).
3.3 Off-centre velocity distribution
For stretched vortex rings (SVR), the largest velocity deficit appears closer to the central plane in the outer boundary-layer region which is the direct consequence of the low momentum fluid lift up due to the interaction of the vortex head and the counter rotating legs with the local fluid. A stretched vortex clearly produces a stronger inflexion and larger deficit than the hairpins. Away from the centerline, such as z/D o = 0.4 and 0.6, there appears an obvious reduction in the velocity deficits and the deficit shifts farther away from the wall filling up the gap towards the undisturbed profile. Simultaneously, the gradient is increasing sharply in the near wall region, as the velocity increases abruptly from zero at the wall. Similar to hairpins, the maximum gradient occurs at z/D o = 0.4 and afterwards the gradient tend to decreases towards the farthest plane.
The tilted vortices influence the boundary layer up to the very edge and beyond, as the deficit area widens up quite significantly. In the near wall region, the gradient seems rather enhanced especially on the closer locations to the central plane. The maximum gradient occurs at the central plane that corresponds to the fluid downwash induced by the tertiary vortices that remain closer to the wall. The gradient is steepest and is shifted to the central plane, where the maximum momentum redistribution and fluid mixing occur eventually. The change in velocity profiles is significant around the central plane. First, for central plane, there is an abrupt variation at y/δ = 0.4 caused by the secondary trailing vortex and second at y/δ = 0.9 due to the primary vortex. There appears a gradual reduction in the maximum gradient achieved corresponding to the movement away from the region of maximum downwash induced by the trailing tertiary vortex pair. Similarly, there is a spanwise velocity increment that coincides with the gradual movement from the inboard side of the trailing secondary vortex pair responsible to induce a flow away from the wall. Moreover, towards the boundary-layer edge, a localized region of velocity deceleration is noticed that is produced by the interaction of the primary tilted vortex that tends to protrude out of the boundary layer.
3.4 Actual control effect
For the second set of experiments, the design of flat plate is altered by attaching an inclined plate to the trailing edge. It is inclined upward at an angle of 5° to generate the artificial separation on the inclined surface, as shown in Fig. 2. The joint between the two plates is filled up with a thin rubber sheet to provide a smooth flat surface to avoid any premature separation at the joint. An array of jets have been generated from five orifices from a multiple hole orifice plate and issued into the separated region to evaluate the effectiveness of the vortical structures towards flow separation delay on the surface. At the end, a third plate 150 mm long is attached and deflected downwards appropriately to ensure stagnation point at the leading edge. The measurement area is illuminated with a laser sheet that forms the measurement plane. The laser sheet is kept parallel to the inclined plate and is kept 1 mm away from the surface to reduce any unwanted deflections. Although the SJA geometrical dimensions have been changed in terms of cavity volume and orifice diameters, the non-dimensional parameters like VR, Re L , and L are kept similar to those mentioned earlier.
The separation line is pushed back further (case b), as the hairpins are becoming stronger and more capable to withstand the near wall shear. Therefore, they interact strongly with the local low momentum fluid in the near wall region. The high velocity streaks are becoming more prominent, as the separation line is pushed further back. Finally, the vortices are turned to tilted vortex rings, as the ‘VR’ is increased (case c, d). Because of their tendency of protruding out of the boundary layer, the primary vortices do not affect the separation delay significantly. For tilted vortices case, the separation delay is caused by the trailing induced vortices that remain essentially within the boundary layer.
A series of particle image velocimetry (PIV) experiments have been conducted to evaluate the dynamics of vortical structures to confine the synthetic jet actuator operating parameters, where the best separation delay effect is achieved. The stretched vortices appear to influence the boundary layer in the similar manner as the hairpins in that the occurrence of the maximum velocity gradient in the near wall region is consistent with the fluid downwash towards the wall produced by the interaction of the counter rotating vortex legs. The stretched vortices are more likely to produce a rather fuller profile with a larger velocity deficit area across the boundary-layer thickness. Subsequently, this would appear to be capable of higher momentum redistribution of fluid from the outer part of the boundary layer towards the wall.
However, considering the quantitative evaluation, it appears that the tilted vortex rings seem to push the base line backwards to the maximum extent. It appears that secondary and tertiary vortices produced in the wake of the primary tilted vortex are the most effective structures towards flow separation delay. The tertiary vortices interact with the near wall fluid in the same way, as the primary hairpins in that they tend to force the fluid downwash towards the wall on the centre line. The effect seems more acute compared with the hairpins, as they interact with the fluid readily energized by the interaction of secondary vortices. Second, the fluid downwash on the single centre line is far more enhanced than the hairpins and stretch vortex rings (SVR) which tend to produce two streaks outboard of the vortex legs. In the artificially produced separation region, the single high-speed streaks seem to delay the separation to a maximum when compared with two streaks. The tilted vortices produced at low-velocity ratio are more effective than those produced at larger velocity ratio. At higher velocity ratio, the swirling flow around the separation line is achieved that inhibits the further delay of the baseline.
The first author would like to thank the University of Engineering and Technology Lahore, Pakistan, to provide funds for his Ph.D. studies and the moral support. Thanks are due to Dr. S. Zhang and Dr. F. Guo for their valuable assistance to build the rig for the experimental work.
- Gilarranz JL, Rediniotis OK (2001) Compact, high-power synthetic jet actuators for flow separation control. AIAA 2001-0737, 39th aerospace sciences meeting and exhibit, Reno, USA, 8–11 January 2001Google Scholar
- Ishtiaq AC, Zhong S (2013) The evolution of synthesized vortices in turbulent boundary layer. J Turbul 4(10):1–18Google Scholar
- Mallinson S, Hong G, Reizes J (1999) Some characteristics of synthetic jets. AIAA 3651:1999Google Scholar
- Smith BL, Trautman MA, Glezer A (1999) Controlled interactions of adjacent synthetic jets. AIAA 0669, 37th aerospace sciences meeting and exhibit, Reno, USA, 11–14 January 1999Google Scholar
Open AccessThis article is distributed under the terms of the Creative Commons Attribution 4.0 International License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.