For transonic transports, the high-lift system is a critical part of the configuration design. To achieve “reasonable” field performance while also obtaining efficient transonic cruise, the design will require a fairly sophisticated high-lift system. Its critical importance is substantiated by the following excerpts (presumably referring to the B-777)from a paper by Boeing aerodynamicists:^[1]

• “A 0.10 increase in lift coefficient at constant angle of attack is equivalent to reducing the approach attitude by one degree. For a given aft body-to-ground clearance angle, the landing gear may be shortened for a savings of airplane empty weight of 1400 lb.”
• “A 1.5% increase in maximum lift coefficient, C_Lmax, is equivalent to a 6600 lb increase in payload at a fixed approach speed”
• “A 1% increase in take-off L/D is equivalent to a 2800 lb increase in payload or a 150 NM increase in range.”

For fighters, high-lift devices are also scheduled for efficient maneuver.

High-lift systems are critical for STOVL (short takeoff and vertical landing) and V/STOL (vertical and short takeoff and landing) aircraft. They also use the propulsion system to help generate the lift. It has always seemed peculiar to design fighter aircraft, or virtually any military aircraft, to operate from traditional runways; the one thing the adversary is going to know is the exact location of your runways. So a STOVL capability seems to be critical in a serious confrontation.

7.1 Characteristics of High-Lift Systems

High-lift systems employ various means of flow manipulation around a wing to increase its lift, enabling the airplane to perform specified missions. The systems exploit one or more of the following basic aerodynamic concepts: (i) increase effective camber of the airfoil, which increases C_l; (ii) increase wing area to increase lift; and (iii) boundary layer control (BLC) on the surface through active or passive means to delay or prevent flow separation.

Boundary layer control through active means requires additional energy to either remove low-momentum fluid from the boundary layer by suction or inject air into the boundary layer to increase momentum; both delay onset of separation and could increase C_Lmax. Although highly promising, practical implementation requires additional equipment to manage the distribution of energy.

Passive means of flow manipulation, on the other hand, are the well known leading-edge (LE) and trailing-edge (TE) devices that are found on most of the aircraft in service. For operating these devices, we only need a drive system for retracting and extending the flaps. Our primary focus in this chapter is on passive LE and TE high-lift devices as described in sections 7.2 and 7.3.

Table 1 shows typical values of C_Lmax. They come from papers by Brune and McMasters,^[2] Roskam and Lan,^[3] and Sanders.^[4] Clearly the 727 emphasized short fields and thus required a higher C_Lmax. Anyone who ever looked out the window while landing in a 727 noticed the elaborate high-lift system employed.

Table 7-1: Values of C_Lmax for some airplanes. (Note: There is a significant variation of values from different sources.)

Some key aspects:

• Compressibility can be important early.
• Reynolds number scaling from wind-tunnel to flight may be problematic.
• Today, simple high-lift systems are critical, and the high manufacturing cost for high-lift systems is important.

Classes of systems:

• High lift for a single element airfoil
• Multi-element airfoils
• Use of “blowing” in some form (powered lift)

Computational simulation:

• Computational simulation requires consideration of viscous effects immediately (unlike typical cruise airfoil analysis and design, where some insight can usually be gained using inviscid simulations that ignore viscous effects).
• Predicting high lift is done almost entirely with Reynolds-averaged Navier–Stokes (RANS) codes (one important exception is Mark Drela’s MSES ^[5] code). A summary of the computational capability is given by Rumsey and Ying.^[6]

Single element airfoils:

• Central to the story of how to obtain high lift on a single element airfoil is Liebeck’s high-lift airfoil^[7] and the Stratford “pressure recovery” shape of the pressure distribution. This introduces the classic paper by A. M. O. Smith.^[8] See section 7.4 below for some insights from this paper.

Multi-element airfoils:

• Understanding the physics: Section 6.3 of Smith’s paper^[9] is critical to understanding the physics. Review the paper to make sure you understand (i) the slat effect, (ii) the circulation effect, (iii) the dumping effect, (iv) off-the-surface pressure recovery, and (v) the fresh boundary layer effect. (Note: some people combine the circulation and dumping effects and call it the “vane effect”; see section 7.4 below for details.)
• Design: See a comprehensive survey by C. P. van Dam.^[10]
• Existing systems on commercial transports: Peter Rudolph has surveyed the high-lift systems on several subsonic transport aircraft.^[11]

7.2 Examples of High-Lift Devices

In this section, we include examples of high-lift devices based on the original hand-drawn sketches of Grumman’s Dick Kita.^[12] These are fairly realistic drawings, as opposed to many of the drawings in textbooks, which may at times be somewhat cartoonish. Kita worked in the high-lift aerodynamics area for many years. I am aware of his work on the Gulfstream II and F-14, but he worked on many other Grumman aircraft high-lift systems as well. Examples of trailing-edge devices are presented in section 7.2.1 and of leading-edge devices in section 7.2.2. Deployment of these devices changes the airfoil shape. Due to notable change in airfoil camber shape, the airfoil aerodynamic characteristics show substantial changes, as illustrated in section 7.3.

Of the large number of papers addressing high lift, several deserve mention. Van Dam, et al.^[13] provides a good summary of high-lift methodology. A summary of the Boeing 777 high-lift system development, as well as the overall design process, is available in the paper by Nield.^[14] A valuable description of high lift on transports in contained in Gratzer.^[15] The use of powered lift is covered in the survey by Korbacher.^[16] Somewhat dated but valuable resources include the book by Hoerner and Borst^[17] and the book by McCormick^[18] (recently reissued unchanged from the 1967 edition as a Dover paperback). Perhaps the best chapter on high lift in a basic textbook is in Shevell.^[19]

7.2.1 Types of Trailing-Edge Devices

Figure 7-1 shows examples of typical trailing-edge (TE) devices such as plain flap, slotted flap, and Fowler flap. In this figure, the flap deflection angle is denoted by δ_f. Note also that the standard convention for flap deflection angle is flap-down positive and flap-up negative. Deploying nonextending flaps, such as plain flap, split flap, or slotted flap, increases the camber of the airfoil. When an extending flap like the Fowler flap is deployed, it increases the planform area in addition to changing the camber.

More complicated TE devices, such as double-slotted flap or double slotted extensible flap, have been developed, as shown in figure 7-2. Their use results in greater increase in lift than one can obtain from the basic devices.

7.2.2 Types of Leading-Edge Devices

Figure 7-3 shows examples of typical leading-edge (LE) devices such as nose flap, Krueger flap, or slat.

7.2.3 Integration of Devices on the F-14 Wing

Figure 7-4 illustrates integration of LE and TE high-lift devices and control surfaces on the wing of the F-14 in their stowed and deployed positions. For low-speed combat maneuvering with the wing unswept, the outer two sections of trailing-edge flaps can be deployed at 10° and the nearly full-span leading-edge slats are drooped to 8.5°. This airplane has a fairly elaborate scheme, where the cove region is smoothed with a moving flap and the upper surface of the slat also has a movable piece to fair the flap. Many fighter airplanes use spoilers instead of ailerons for roll control, although their use varies from company to company.

7.3 Aerodynamics of Leading- and Trailing-Edge Devices

7.3.1 Trailing-Edge Devices

The typical effects of the trailing edge flap deflection on the lift of a cambered airfoil are illustrated in figure 7-5. For a given angle of attack, α, deflecting flaps down increases the lift coefficient due to the increased effective camber. With increasing flap deflection angle, δ_f, the zero-lift angle of attack becomes more negative and C_lmax increases, but the lift-curve slope does not change. However, the stall angle of attack for C_lmax actually decreases slightly with increasing flap-deflection angle. The lift characteristics in this figure represent deployment of trailing-edge devices only without deploying any leading-edge devices.

Flaps that extend in a Fowler motion increase the wing planform area, which is shown in figure 7-6(a) by the hatched region in the planview of the wing. Therefore, the wing with extension generates more lift at any angle of attack. However, the original wing reference area, S_REF.—not the extended wing area, (S_REF. + ΔS_EXT.)—is used in computing the C_L for wing flaps down (with extension). This has the effect of increasing the effective lift-curve slope by approximately the ratio of the total extended wing area to the original wing reference area—i.e., (S_REF. + ΔS_EXT.)/ S_REF., as shown in the figure.

The lift curves, C_L versus α, for the clean wing (flaps up) and for the wing with its extending flaps down are illustrated in figure 7-6(b). Much like with other types of flaps, the flap’s downward curves show that the zero-lift angle of attack becomes more negative, C_Lmax increases, and stall angle is slightly reduced. The figure also shows that, when flaps with no extension are deployed, they do not change the lift-curve slope of the wing. However, the lift-curve slope for flaps with extension changes by the ratio of the total extended wing area to the original wing reference area.

In general, we expect C_lmax to increase with Reynolds number for a clean airfoil as shown in the top part of figure 7-7. However, sometimes the projection is not so straightforward, and C_lmax may even decrease.^[20] Considering the adverse effect of Mach number on C_lmax in the bottom part of figure 7-7, the low Mach numbers at which the effects take place is worth noticing. This result is rather surprising to the uninitiated.

7.3.2 Leading-Edge Devices

Leading-edge devices, such as a slotted flap or slat, work to delay flow separation near the leading edge. Therefore, they don’t really do anything until you reach the angle of attack where the leading-edge flow would “let go” if it didn’t have the slats for protection. Thus the slats allow the lift to continue to rise to higher angles of attack before C_lmax is reached, as illustrated in figure 7-8. The figure also shows that slats mitigate the effect of TE flaps on stall angle, which is somewhat reduced as flap deflection increases. Slats permit high-pressure air from the lower surface to blow over the top of the wing and delay separation by energizing the boundary layer.

Different types of leading-edge devices differ in their effectiveness. Figure 7-9 is Kita’s estimate of how each type of device increases the angle of attack and level of maximum lift. Due to the higher effectiveness of slats, they are more widely used.

Figure 7-10 shows Kita’s estimate of the typical “best” performance in terms of C_Lmax you can get from various types of high-lift systems. You can see that sweeping the trailing edge reduces the effectiveness of all high-lift systems. The curve labeled “advanced” is typical of projections made in advanced design or aerodynamics departments, where the assumption is that an advanced technology development effort can improve the performance of any system. This may or may not be true.

In addition to changing lift, flaps also cause a large change in drag, as shown in figure 7-11. Clearly, you don’t want the flaps deployed at low lift. This is why flaps aren’t deflected in cruise. As lift increases, there may be an optimum flap-deflection schedule; for example, the F-18’s flaps are scheduled with angle of attack and Mach number. This was also done on the Grumman X-29.

Flap deflection also produces a large change in pitching moment, as illustrated in figure 7-12. This is an important consideration, since you need to be able to trim this pitching moment. In the case of the Beechcraft Starship, the canards actually changed sweep to be able to generate the required force. So this effect cannot be ignored in developing high-lift systems.

In developing the high-lift system, two key parameters are the selection of the gap between the flap and main element and the selection of the overlap. Figure 7-13 provides Kita’s definition of these parameters. In trying to identify the values of these parameters that produce the highest lift, aerodynamicists require a lot of time for wind tunnel testing and considerable computer resources for simulations. At higher flap deflections, say 25° to 40°, flap tangency to the tangency line is secondary to gap and overlap criteria. Flap type and shape may preclude this occurrence.

7.4 Physics of High Lift: A. M. O. Smith’s Analysis of the High-Lift Aerodynamics

Now that we’ve surveyed the characteristics of high-lift systems, we need to examine the physical basis for the operation and limits of high-lift systems. A. M. O. Smith “wrote the book” on the physics of high-lift systems with his 1975 lecture “High-Lift Aerodynamics,”^[21] which is required reading. His message? You need to carry as much lift (load) as you can on the airfoil’s upper surface without separating the boundary layer. His classic paper describes the physics associated with the high-lift characteristics we described above and details how to achieve the available high-lift performance. We summarize his description here.

To obtain insight into the characteristics of pressure distributions that affect boundary layer separation, Smith introduced the use of a canonical pressure distribution. He felt strongly that this was necessary to understand and compare possible separation on different airfoils. It is essentially another type of dimensionless, or “scaled,” pressure distribution. For boundary layer investigations, it is found that the best scaling factor is the velocity u₀ just before the flow deceleration begins due to rising pressure. Smith defines the canonical pressure distribution as

$${\overline{C}}_p=1-{\left(\frac{u_e}{u_0}\right)}^2\text{.}\text{(7-1)}$$

Here, u_e is the velocity at the edge of the boundary layer. In this canonical system, the canonical C_p value of zero represents the start of the pressure rise and +1 is the maximum possible value, u_e = 0. The next step is to examine the best way to specify the pressure distribution to allow the pressure to recover to as close as possible to the canonical C_p of +1. Smith made a parametric study of various possibilities to gain insight into the best way to prescribe a pressure distribution to delay separation. However, our focus here is on his limiting case. It makes use of an analysis by Stratford ^[22] that was done to estimate separation before the days when boundary layer computer programs were available. Using Stratford’s analysis, it is possible to define a pressure distribution where the boundary layer is just on the verge of separation everywhere. To do this, we manipulate the Stratford criteria, as seen in equation (7-2).

$$\frac{{\overline{C}}_p{\left[x\left(d{\overline{C}}_p/dx\right)\right]}^{1/2}}{{\left({10}^{-6}R\right)}^{1/10}}=S\text{(7-2)}$$

Here, R is the unit Reynolds number (per unit length), defined as u₀/ν. Note that originally, Stratford used this relation to say that separation occurred when the quantity on the left-hand side of the equation reached the value of S (typically 0.35). However, we can define a canonical C_p distribution using this relation that is equal to S everywhere, just on the verge of separation, and this pressure distribution is the best way to achieve a very large pressure recovery without separation.

Figure 7-14 shows the resulting pressure distribution. Examining this pressure distribution, several key observations can be made. The initial slope is infinite and then decreases. Thus, when the boundary layer is thin, it can withstand a very large pressure gradient. As the boundary layer thickens (either when it starts to recover or as it recovers), it cannot sustain the large pressure gradient, and the pressure gradient to maintain attached flow decreases. This illustrates the idea that thick boundary layers are more likely to separate than thin boundary layers. Note also that the Reynolds number effect is relatively weak. Finally, the boundary layer could recover all the way to a canonical C_p of +1, but to attain this, x would need to go to infinity. These shapes are the best possible pressure distributions to use to recover the pressure without separating the boundary layer. As A. M. O. Smith notes, the only way to do better is to use some sort of active boundary layer control (suction or blowing).

7.4.1 Single-Element Airfoils

The key to obtaining high lift on a single-element airfoil is essentially the inspiration for Liebeck’s high-lift airfoil^[23] and the Stratford “pressure recovery” shape of the pressure distribution described above, as told by Smith.^[24] The question of how much lift you can obtain on a single-element airfoil involves how low the pressure can be on the upper surface and how the pressure can recover to a positive pressure coefficient at the trailing edge and keep the boundary layer attached. Smith describes two aspects of the problem. In the first case, he explains the limit of the pressure coefficient in terms of the vacuum C_p when a zero pressure is specified on the airfoil’s upper surface. Thus, using the definition of C_p,

$$C_p=\frac{p-p_{\mathrm{\infty }}}{\frac{1}{2}{\rho }_{\mathrm{\infty }}{U}_{\mathrm{\infty }}^2}\text{,}\text{(7-3)}$$

we can obtain an alternate form using q = 0.5ρ_∞U_∞² = 0.5γP_∞ M_∞² , giving us equation (7-4).

$$C_p=\frac{p-p_{\mathrm{\infty }}}{\frac{\gamma }{2}{p}_{\mathrm{\infty }}{M}_{\mathrm{\infty }}^2}\text{(7-4)}$$

Vacuum C_p occurs when the pressure is zero:

$$C_{P_{\text{vac}}}=\frac{-2}{\gamma M_{\mathrm{\infty }}^2}\text{(7-5)}$$

Smith points out that, at this point, only 70% of the vacuum C_p has been achieved in practice. This results in a C_p limit of M²C_p = −1 for γ of 1.4.

Next, Liebeck and Smith used the analytical analysis by Stratford illustrated above to specify a pressure distribution that allows the most lift to be obtained. Given this pressure distribution, an inverse method is used to obtain the associated airfoil shape. The result is the Liebeck family of high-lift airfoils.

7.4.2 Multi-Element Airfoils

Section 6.3 of A. M. O. Smith’s conference lecture (AIAA Paper No. 74-939, 6th Aircraft Design, Flight Test and Operations Meeting, Los Angeles, California, August 12-14, 1974) is an excellent source for anyone interested in understanding the flow physics of multi-element airfoils. The foundational paper discusses five main ideas underlying flow physics:

1. The slat effect. The slat protects the leading edge of the main element. That’s why its effect is only observed near C_lmax of the single element. Considered a point vortex, the slat velocity acts to reduce the velocity around the leading edge of the main element.
2. The circulation effect. The downstream element causes the upstream element to be in a high-velocity region, inclined to its mean line. To meet the Kutta condition, the circulation has to be increased. Instead of the airfoil deflecting as a plain flap, the trailing edge is placed in an inclined flow and something else (the downstream element) turns the flow.
3. The dumping effect. The trailing edge of the forward element is in a region of velocity appreciably higher than the freestream velocity. Thus, the boundary layer can come off the forward element at a higher velocity. You don’t have to recover back to C_p = +0.2 for attached flow, relieving the pressure rise on the boundary layer, alleviating separation problems, and permitting increased lift. The suction lift can be increased in proportion to U_TE² for the same margin against separation.
4. Off-the-surface pressure recovery. The boundary layer leaves the trailing edge faster than the freestream and now becomes a wake (a viscous phenomena). The recovery back to freestream velocity happens in an efficient manner since the deceleration of the wake occurs out of contact with the wall. Wakes withstand pressure rises that boundary layers cannot. The wake can actually “separate” out in the flow field. (Note: For a well designed high-lift system, the local boundary layers and wakes remain separate. If they merge, everything is more complicated.)
5. The fresh boundary layer effect. Thin boundary layers can sustain a greater pressure gradient than a thick boundary layer. Thus, three thin boundary layers (on three airfoil elements) are more effective than one thick boundary layer (on a single element). (Note: Some people combine the circulation and dumping effects and call it the “vane effect”.)

7.5 Computational Methods for High Lift

Significant effort has been devoted to improving prediction capabilities for high-lift systems. As stated in the introduction, one of the best surveys is by Rumsey and Ying.^[25] XFOIL^[26] can be used for low-speed predictions of the maximum lift for a single-element airfoil. Experience shows that the program’s predictions are slightly higher than experimental results. Figure 7-15 shows typical agreement between XFOIL and wind tunnel data for the NACA 0012 and 4412 airfoils. Although the prediction is not perfect at maximum lift, this is a remarkable capability of a code that can be run on a laptop PC.

7.6 Passive and Active Boundary Layer Control

7.6.1 Passive Boundary Layer Control

The boundary layer can be prevented from separating by the use of vortex generators, “snags,” and fences.^[27] Vortex generators (VGs) are small vanes mounted on the wing surface to produce mini-vortices that energize the boundary layer and delay separation. “Snag” is a discontinuity in the leading edge of a swept wing where an outer wing panel has an increased chord compared to the inner panel which creates a discontinuity at the junction of the inner and outer panels. A vortex is produced at the junction which delays flow separation. Fences are flat strips mounted to the surface of a swept wing at selected span stations, typically wrapping around the leading edge. The fences obstruct the spanwise airflow and thereby prevent boundary layer growth and flow separation.

7.6.2 Active Boundary Layer Control

If suction or blowing is used to suppress boundary layer separation, the blowing (which is generally preferred to suction) is known as boundary layer control (BLC); if the amount of blowing exceeds the value required for BLC, then the blowing is know as powered lift. The key parameter used to describe the amount of blowing is the blowing coefficient, which is defined as follows for 2D airfoils:

$$C_{\mu }=\frac{m_j{V}_j}{qc}\text{(7-6)}$$

where the subscript j refers to the jet, m is mass flow rate of the blowing device, V is the velocity of the blowing jet, q is the dynamic pressure, and c is the reference chord. Blowing-based BLC was used often in early fighters but is not used nearly as much today. The F-4 Phantom originally had blowing on both the leading edges and over the trailing-edge flap. However, to improve transonic maneuver characteristics and to improve resistance to departure, the leading-edge blowing was replaced by leading-edge slats.^[28]

7.7 Powered Lift

A large variety of concepts have been tried to increase maximum lift using high-pressure air from the engine. Some examples of powered lift concepts are:

• Propeller slipstream deflection (Brequet 941/McDonnell Model 188)
• Externally blown flaps (McDonnell DouglasYC-15/C-17)
• Internally blown flaps (Lockheed F-104, Blackburn Buccaneer)
• Upper-surface blowing (Boeing YC-14, NASA QSRA, Ball-Bartoe JetWing)
• Vectored thrust (AV-8 Harrier)
• Jet flaps (Hunting H.126)
• Jet augmentor wings (NASA-deHavilland Augmentor Wing Aircraft)
• Circulation control (advocated by Hokie Bob Englar,^[29] A-6 CCW)

7.8 Configuration Integration Issues

There are a few key considerations for integrating high-lift devices into your aircraft configuration:

• The best airplane C_Lmax you can achieve with a mechanical high-lift system is about 3 to 3.5.
• The military and civil air regulations require a margin between C_Lmax and the operating C_L of the airplane. This must be accounted for during design. For example, the approach speed must be 1.3 times the stall speed. This would suggest that the maximum approach C_L is only 59% of the C_Lmax. However, some relief is available because the measured stall speed, V_smin, is usually about 0.94 times the stall speed in 1 g steady flight (the wind tunnel case). This means that you can use a C_L of about 67% of the C_Lmax.^[30],[31]
• Increased span can be used to reduce the induced drag, so a bigger flap angle can be used before a climb limit is encountered (though this solution will increase the airplane weight).
• Sweep decreases max lift.
• In converting 2D to 3D, there are lots of losses. Don’t be mislead by 2D C_lmax values, as the actual 3D value will be much less.
• The maximum C_L available for takeoff and landing for many swept-wing airplanes is actually the limit on angle of attack to avoid tailscrape.
• Engine out is much more critical with V/STOL airplanes, typically leading to a complicated engine cross-shafting arrangement to ensure uniform thrust on each side of the plane if an engine fails. The ability to do this dictates whether the concept is practical.
• The best high-lift configuration integration description also involves powered lift. The YC-14 AIAA Case Study^[32] is highly recommended.

Finally, it is worth mentioning a couple more relevant issues, like the concept of the Gurney flap^[33] and the need for accuracy around the leading edge. Maintaining an accurate leading edge contour is critical at high-lift conditions. Once, after a Navy depot had repainted an F-14 wing, a small ridge was left on the leading edge where the upper and lower surface paint overlapped. This was enough to cause early stall. Ed Heinemann reported a similar experience on a Douglas airplane during World War II in his autobiography.

Figure References