1. IntroductionLaminar flow control(LFC) has been known to have high potential to improve aircraftperformance. The aircrafts fuel consumption and hence the range and endurance can beimproved using LFC which is the only known single aeronautical technology that offers suchan improvement in performance. For transport airplanes the amount of fuel consumed could bereduced by 30% 1. Laminar flow over most part of the wing due to low drag BLS can leadto improvement in L/D ratio and when L/D ratio is improved especially in long range aircraftsthere is always a reduction in specific fuel consumption and inturn will increase the range ofthe aircraft at altitudes above 20,000 feet2, 3.
LFC can be categorized into two, passive LFCand active LFC, passive LFC is beyond the scope of this paper. A combination of passive LFCwhich is known as the Natural Laminar Flow Control (NLFC) and active LFC is called HybridLaminar Flow Control (HLFC). Currently two principles of active LFC are under research whichare surface cooling and boundary layer suction (BLS). It has been understood from previousresearch and literature that HLFC increases CL, decreases CD, reduces the boundary layerthickness and delays the transition of laminar to turbulent flow. BLS is an active type LFCwhere small amounts of energy input helps in keeping the boundary layer attached.
ThoughBLS is extremely successful in delaying transition, increasing lift and decreasing drag, it has notbeen implemented on aircrafts due to its high maintenance in keeping the BLS holes clean. Itis also true that once the flow is turbulent no combination of suction helps in reducing drag4.It was also observed that if BLS failed, then the max lift coefficient of the aerofoil drops belowthe max lift coefficient of the original airfoil without BLS5.BLS has also been very effective in reducing viscous drag. Two types of suction are generallyused, dicrete suction and distributed suction6. This paper surveys and reviews only BLS basedLFC.
LFC used in compressor blades would make it an exhaustive survey and hence has beenneglected. BLS on compressor blades are out of the scope of this paper.1.1.
Historical BackgroundThe very initial idea of suction being used as tool to control boundary layer dates back tothe inception of boundary layer theory itself, Pradtl in his article The Mechanics of ViscousFluids in Aerodynamic Theory, Vol. III in 1935 makes a mention of boundary layer controlusing suction7. Work on LFC began in the early 1930’s with United Kingdom, Germany, andthe United states using wind tunnels to test its effect. Initial tests using multiple suctions slotsresulted in laminar flow upto a Reynolds number of 7 million which was considered a phenomenonachievement during that period of time1.
The very first flight experimentation began in theearly part of 1940’s where researchers placed suction holes between 20 and 60 percent of thechord on a B18 airplane. Research gained importance and peaked in the 1940’s with the NACA6 series airfoils. With the introduction of sweep in the 1950’s due to high subsonic speed rangesthe problem of cross flow came into effect 8. High subsonic and supersonic speeds requiresweeping of wings and it was found that only suction can control the cross flow disturbancesthat helps in promoting boundary layer transition from laminar to turbulent flow 1 The workon suction based LFC further culminated in the 1960’s with experimentation on aircrafts builtby Britain and America. The application of suction based LFC had a new area of research inthe 1970’s which were related to supercritical wings. The present survey is to focus attention onthe suction based LFC and to understand the mechanism of boundary layer suction.
To gain adeeper insight into the developments of BLS during its early years of development the reader isreferred to references 1 and 8.1.2. Flight TestsIn the 1960’s Northrop X-21 which was an experimental aircraft with 300swept wings wasequipped with BLS slots along the entire span of the wing under the supervision of WernerPfenninger who was one of the pioneers in BLS. Experiments have been then performed onvarious aircrafts in the past which includes Dassault Aviation, Jetstar and Airbus to name afew, the results of these flight tests have been reviewed in this subsection.Flight tests done on the German made aircraft DO-27 have shown that by using BLS the netdrag during cruise can be reduced. The test results obtained on the DO-27 also accounted forthe power requirement for the BLS 9.
The values of maximum lift by using BLS during flighttest were similar to that of using leading edge slats. At higher angle of attacks where the flowover the fuselage had already separated the aircraft was still maneuverable when BLS was on9.Flight tests early in 1941 were performed on a Douglas B-18 with BLS slots mounted on the leftwing with the NACA 35-215 airfoil, the BLS slots were placed on the upper surface between 20and 60 percent of the chord10. The Douglas B-18 was a twin engine mid wing monoplane withslot spacing of 5 percent of the chord.2. Study on the Effect of Boundary Layer Suction on PerformanceIn this section, the effects of BLS on curved surfaces(airfoils) will be reviewed and not flat plates.
As the fluid passes around the airfoil the flow is accelerated from the stagnation point onwardsand at the trailing edge the flow is decelerated, if the flow is decelerated over a large part of thechord the adverse pressure gradient is not so severe enough to separate the boudary layer, byusing BLS this limitation could be extended11. The total kinetic energy absorbed by the fluidduring acceleration is required to overcome the pressure gradient during decceleration in orderfor the fluid to come to rest at the trailing edge after accelerating from the stagnation point12.But in the case of air which is a viscous fluid the kinetic energy is lost due to friction, due tothis loss the kinetic energy remaining is insufficient to overcome the pressure gradient and hencethe flow does not come to rest at the trailing edge and there is present a velocity component inthe direction of the motion of the wing.
Initial theoretical studies of BLS were developed for flat plates where at very low Reynoldsnumber no suction is required to keep the flow attached and stable, but in the case of airfoils thepresence of adverse pressure gradient at the rear portion requires BLS at all Reynolds number tokeep the flow laminar until the trailing edge13. Suction could be discrete or distributed over thechord and span of the airfoil and wing respectively to obtain best reduction in drag, increase inlift and delay in transition. Anderson et al.14 studied suction distribution and defined optimumsuction distribution as the one in which minimum amount of suction is required at each suctionpoint to delay/prevent transition.Various factors affects the efficiency of boundary layer suction.
The suction pressure, suctionhole position, the size of the suction hole, Reynolds number, boundary layer thickness are afew of the factors than can influence the efficiency of the boundary layer suction in increasinglift reducing drag and delaying transition of laminar to turbulent flow. Reynolds number andsuction rate influence the boundary layer thickness which in turn could play a vital role inreducing drag15.2.1. Effect of Suction Coefficient and PressureIt was found that by increasing the suction coefficient(Cq) the CL increased and CD decreased,which also includes skin friction drag and pressure drag. But when Cq reaches its critical valuedrag remains unaffected and also increases slowly16. It was also noted by Glauert et al.
17that the airfoil with suction at low angle of attacks adhered to the quantity of suction requiredas per theoretical calculations but as the angle of attack increased and the flow separated thequantity of suction required to re-attach the flow was three to four times higher. Azim et al.18studied the effect of suction pressure on trailing edge separation, it was found that the separationpoint moves towards the trailing edge with the decrease in the suction pressure. There is drasticimprovement in the L/D ratio as the suction pressure is dropped. Kianoosh et al.19 studied theeffect of suction coefficient on stall angle increase, and observed an increase in stall angle whenthere is an increase in the suction coefficient. The increase in suction coefficient also caused andecrease in the vortex formation behind the airfoil.2.
2. Effect of Suction Hole Width and AngleSuction hole width also known and called as suction jet length has an effect on the performanceof BLS. Shi et al.16 studied the impact of hole width during BLS and its in influence onperformance and transition. It is a general understanding that suction hole width directlyrepresents the suction mass, and hence by increasing the suction hole width it was noted thatthe boundary layer thickness grows thinner and, as a result of the decreasing boundary layerthickness CL is increased while skin friction drag initially decreased but later began to increase.As the boundary layer is thin and Reynolds number increases, there could be a posibility thatthe critical Reynolds number for roughness could exceed and induce large disturbances at theboundary layer13. On the contrary pressure drag had a continuous decreasing effect withincrease in suction hole width16.
But a very large suction slot tends to have a adverse effect onperformance11. E. D.
Poppleton20 concluded based on his study of boundary layer suctionon a 40 degree swept back wing that the size of the slot is function of choking conditions, thebest slot size would be the one which is just not choking when Clmax is reached.Kianoosh et al.21 observed that when a suction perpendicular to the flow is used, not onlydoes the L/D ratio improve but the stall angle also increases, in the case of a NACA 0012 airfoilthe stall angle improved from 140to 220. The hole width also had a direct effect on the L/Dratio, where the increase in the hole width augmented the L/D ratio and delayed the separationfurther downstream21. It was also observed that the L/D ratio increased as the hole widthincreased to 2.5% of the chord and then insignificantly decreased. Kianoosh et al.
19 observedthat when the suction jet length increased till 2.5% of the chord length the lift increased anddrag decreased and as the suction jet length increased between 2.5% to 3% of the chord therewas insignificant improvement.BLS can be performed at various angle Huang et al.22 studied the effect of varying suctionangle and concluded that perpendicular suction (suction at 90 degrees to flow) has the largestimpact to increase lift.
2.3. Impact of Suction Hole Position2.3.
1. Impact of Suction Hole Position on the Chord It is important to understand that thereis a point on the airfoils chord which is the point of transition without any LFC. This point iscalled the natural transition point. Shi et al.
16 studied the effect of position of the BLS holeas a function of the chord length. It was noted that as the position of the hole moved towardsthe trailing edge CL increased and CD decreased. But a very important observation was madethat after the BLS point moved past the natural transition point towards the trailing edge, theCL decreased and CD increased16. There is no possibility of a benefit of transition delay ifthe point is placed after the natural transition point as the transition from laminar to turbulenthas already occurred and it has an adverse effect on aerodynamic properties16.
Azim et al.18found that on the NACA 4412 airfoil the trailing edge separation began at 0.7c from the leadingedge and a BLS slot when placed at 0.
68c moved the trailing separation to 0.88c. When aslot is placed nearer to the leading edge at 0.56c and 0.48c the BLS decreased performance bydecreasing lift and increased turbulence which causes an increase in drag 6 times to that ofwithout suction. Azim et al.18 concluded that placing the BLS slot only near to the separationpoint helps in reducing drag. Tutty et al.
23 also suggested that placing a BLS slot before thetransition point makes BLS ineffective and placing the suction point after the transition pointhas no effect in delaying the transition. It has also been documented by flight tests that leadingedge suction helps in delaying transition and is a very effective means of increasing maximumlift, it has a similar effect to that of a leading edge slat9, 22. Kianoosh and Reza24 observedthat when the suction slot is placed near the leading edge at 10% of the chord there is a negligibleaffect below stall angle (14 degrees).
But had a phenomenal impact on the L/D ratio beyondthe stall angle. A. T.
Piperas 25 observed that when the transition point is downstream thesuction location a delay in transition can be achieved but when the transition point has movedupstream of the suction point, the suction becomes ineffective. By placing the suction slot closeto the leading edge one can improve the pressure difference between the upper and lower surfacesand the aerodynamic behavior25. Millard J Bamber12 found that the best BLS slot locationwould depend on the angle of attack. After thorough wind tunnel test it was found that the bestslot position for small angle of attacks is near the trailing edge and as the angle of attacks keepsincreasing the BLS slot should move further upstream towards the leading edge12. Suctionslot spacing when multiple slots are present is another area of concern, Dale et al.
26 concludedthat when the slot spacing is very large the suction power required is very large and smallersuction slot spacing would cause a manufacturing challenge. It is a challenge to optimize theslot spacing when multiple BLS slots are used.The position of suction hole is a problem without a conclusion as different authors concludedifferently. There could be a possibility that different airfoils(t/c ratio), mach number, angle ofattack, suction pressure, the application and the conditions at which boundary layer suction isused to improve performance could have different positions. Clearly there seems to be a needfor further research in this area to sort out the discrepancies and to further understand themechanisms of boundary layer suction and the positioning of BLS slots.
2.3.2. Impact of Suction Position on the Span of the Wing Kianoosh and Reza24 studied theeffect of suction area on a 3-D wing by placing the suction area along the span of the wing. Twoconfigurations were studied, tip suction and center suction where the width of the suction areawas 2.5% of the chord.
The location of the suction area was set to 10% of the chord from theleading edge.When considering the overall effect, center suction was considered to be a better choice, wherethe L/D ratio increased better with center suction than tip suction24. When the length of thesuction slot is greater than half the wing span center suction is better and when the suctionlength is less than half the wing span tip suction is better24. Leading edge contamination isoften seen in a 3-D swept wing when the swept wing is attached to the fuselage or to the wallwhile performing a wind tunnel test. By placing the BLS slot near the leading edge of a sweptwing along the attachment line it is possible to prevent the attachment line contamination27.H.
J. B. van de Wal28 as a part of his master thesis in Aerospace Engineering at TU Delftattempted to design a wing with boundary layer suction by redesigning the wing of the EuroENAEREE10 Eaglet, a research aircraft of TU Delft. The most critical region where mostsuction was needed was at the wing tip and the least needed was at the root. Also when flapswere deployed the suction required was less. H. J. B.
van de Wal28 also concluded that therewas no need of suction at the wing root due to the roughness which would trigger a turbulentboundary layer immediately and another reason for turbulent flow in that region is due to wakegenerated from the propellers. The flight parameters also had an improvement due to the Eagletsnew wing with BLS installed. The total drag reduction was very small as the wing profile dragis a relatively small portion of the aircrafts total drag. The aircraft had a steeper lift curve whencompared to the original28.
2.4. Effect of Suction Amplitude and Suction VelocityWhen a 2-D airfoil with BLS is investigated there is no cross-flow(CF) effect, flows with theabsence of CF leads to different flow physics. 3-D wings especially swept wings have the effectof cross flow which should be considered while analyzing boundary layer suction on wings29.Kianoosh et al.
19 defined suction amplitude as the ratio of suction velocity to free streamvelocity. Three suction amplitudes were considered, 0.1, 0.3 and 0.
5 and the increase of suctionamplitude from 0.3 to 0.5 had a greater effect in increases lift and reducing drag and hence animprovement in the L/D ratio.
The maximum L/D ratio was obtained at a suction amplitudeof 0.519. Below a suction amplitude of 0.01 there seemed to be no significant effect due toBLS, but above the suction amplitude of 0.01 lift increased as suction amplitude increased22.It was noted by Robert et al.30 that below the critical value of suction velocity which is theminimum suction velocity the drag increased with a decrease in the suction velocity and abovethe critical suction velocity the drag remained constant30.2.
5. Effect of Slats and FlapsThe effect of high lift devices on BLS have been rarely studied, a few researchers attemptedto understand the influence of high lift devices on BLS and this has been reviewed in thissubsection. The deflection of flaps also tends to influence the efficiency of BLS. It has beenfound experimentally that there is an increase in Clmax from 1.9 to 2.2 when BLS was appliedfor a 0.2 chord split flap which was deflected by 600.
But higher suction pressure may be requiredto be able to balance the addition pressure difference caused due to the deflection of the flaps31.Extension of leading edge slats tend to delay leading edge separation and it was found that asuction slot closer to the leading edge has a more favorable influence on the maximum lift of theairfoil without slat32.3.
Effect of Boundary Layer Suction on Aerodynamic ParametersBoundary layer suction as we know improves aerodynamic parameters like increase in Cl,decrease in Cd, delaying stall and delaying the onset of turbulent boundary. A few of theseparameters have been reviewed in the previous sections. In this section the effect of BLS on afew other aerodynamic parameters will be reviewed.Stall delay is one of the advantages of BLS and stall in a plain wing without suction appearsto be because of leading edge stall but whereas the stall in a wing with BLS present is aresult of boundary layer separation at the trailing edge30. Also the stagnation point movesforward further for a wing with BLS in comparison to a plain wing as the angle of attack isincreased. With the introduction of BLS airfoils with higher thickness to chord ratios(t/c) canbe used without the high drag expectation due to separation11. Though it has been foundexperimentally that no significant decrease in drag is possible for airfoils with normal thicknesson which separation does not occur33.Another possible feature of BLS is that it could improve the lateral control of the aircraft.
As lift could be increased with BLS, a rolling moment could be produced by varying the wingpressure in the outermost region of the wing12.4. Mathematical Models Used in Numerical Analysis to Predict the Effects ofBoundary Layer SuctionFlow over an 2-D airfoil can exhibit different complex phenomenons such as wakes, flowseparation, boundary layers etc.
Computational Fluid Dynamics(CFD) developing at such arapid pace is being used to predict not the 2-D airfoil but also the finite wing which is a 3-D body. The 3-D effects such as downwash, induced drag, trailing edge vortex can also bepredicted using CFD. Experimental work is very important to produce data that is required toanalyze suction based flow control. Obtaining very fine and sensitive detailing requires repetitiveexperimentation which could be an expensive affair. Hence numerical methods/analysis couldbe used to capture and predict the effects. Various techniques and mathematical models areavailable to capture the physics. Different transition and turbulence models used to predict thetransition from laminar to turbulent flow have been reviewed in this section.
RANS(Reynolds-averaged NavierStokes) is an cost effective method compares to DNS(DirectNumerical Simulation) or LES(Large Eddy Simulation)34. Shi et al.16 took a numericalapproach to analyze hybrid laminar flow control (HFLC) using the Menter and Langtry’stransition model. Computational results which were obtained were validated using experimentalresults. It was found that the Menter and Langtry’s transition can predict the transition ofboundary layer from laminar to transition when boundary layer suction was present. Sun etal.35 studied boundary layer suction on a linear compressor cascade using RMS (ReynoldsStress Model) which was described in literature as the most suitable for complex threedimensionalseparations.
Azim et al.18 investigated the effect of BLS using the Spalart Allmarasturbulence model. The Spalart Allmaras turbulence model is a single equation linear eddyviscosity model which was initially designed for aerodynamic flows. Kianoosh and Reza24studied the effect of 3-D suction on a NACA 0012 wing using RANS with the K-? SST turbulencemodel which can predict flows with separation very accurately24. Kianoosh et al.
21 usedRANS equations in conjunction with the Menters shear stress turbulent model which is twoequation model(K-? SST) which is as mentioned capable of predicting flows with separation.A. T. Piperas 25 concluded that the SST Gamma Theta turbulence model predicts the effectsof transition better than the K-? SST model.Different turbulent and transition models were studied to verify the equations ability topredict turbulence and transition by Serdar et al.34 and it was found that the K-? SSTtransition and turbulent model tends to under-predict turbulence and the K- RNG overpredictedthe stall. The K ?Kl ? ? model was found to be relatively better in agreementwith available experimental data.
5. SummaryThere is a significant improvement in aerodynamic performance due to laminar flow control byBLS and hence an improvement in the overall performance of an aircraft with active BLS. Withthe increase in suction coefficient lift and stall angle increased with the decrease in the dragand vortex formation behind the airfoil. A conclusion on the most effective BLS slot positioncould not be drawn, further research into optimization of slot position is a requirement. Therewas a decrease in the boundary layer thickness and pressure drag reduction with the increase insuction hole width.
When suction is applied at center of the wing span it improves the L/D ratiobetter than when the suction is applied at the tip of the wing span. BLS near the leading edgeif the wing can prevent attachment line contamination. Higher suction amplitude influencesimprovement in lift and reduction in drag. Higher t/c ratio airfoils can be used when BLS isused without the increase in drag due to boundary layer separation. There is also a possibilityof better lateral control of an aircraft when BLS is present and active.
The K ? Kt ? ? modelwas considered to be the best turbulence model to predict the effect of BLS and flow separationwhile using computational fluid dynamics to analyze BLS.