r o t t e d: fs2k2 flight dynamics: negative flap travel

Negative Flap Travel

Why negative flap settings are used on some aircraft is one of those issues which is frequently misunderstood, and deserves more explanation. Also, at the root of this thread is the question, "can we acheive the same effects in Flight Simulator?" A very good question and one to which I wanted to find an answer.

When an airfoil is designed or selected for a new aircraft, the engineer chooses a target point in the performance envelope for which the airfoil will be optimized, and also decides how much consideration must be given to the other ends of the envelope, knowing that the more compromises he makes, the lower the peak performance will be. Let's call this chosen point in the performance envelope the airfoil's optimum speed. The airfoil will be designed (or chosen) to provide it's best performance (lowest drag) at that speed.

For a sailplane, the airfoil must have very low drag at low speed (high lift coefficient) so that the plane will thermal well. The difficult part is that a sailplane must also have very low drag at as high a cruise speed as possible (low lift coefficient) for good penetration and cross-country performance. These two requirements are quite contradictory. An airfoil which performs well at low speed will necessarily have a high camber (camber refers to the upward curve of the airfoil's mean line; the line half way between the upper and lower surfaces of the airfoil connecting the leading and trailing edges). When flying considerably faster, the same airfoil will have to be flown at so negative an angle of attack that it will be operating well outside of its intended envelope and creating substantial drag. The way to "tune" an airfoil's optimum speed is to change it's camber. The camber is usually changed only by the airfoil designer and is "hard wired" into the airplane by the fact that the wing is of course rigid. If wings could be built completely flexible and could change cross-section freely in flight, then the engineer wouldn't have to choose an optimum airfoil at all. The pilot (or an on-board computer) could tune the cross-section of the wing to the precise requirements at each moment. In the real world, however, flaps offer us our only chance to alter the camber of the airfoil in flight. This method of changing the camber is obviously imperfect and quickly loses efficiency the further the flaps are deflected.

Generally with powered airplanes the optimization is for cruise speed, which leads to a thin, low-camber airfoil which doesn't perform very well at low speed. Flaps are then used to increase the camber (and incidence) for low-speed flight. The loss of efficiency associated with large flap deflection angles can be put to good use in the landing pattern where the increased drag is welcome. For a sailplane the optimization is more towards the middle of the speed range where having the absolute maximum L/D can mean making it home instead of having to land out. The airfoil itself is optimized for a middle speed and then flaps can be used to increase or decrease the camber a little to tune the optimization up or down the speed range. Flaps down for thermaling, flaps up (negative) for fast cruise, and flaps down much further for landing.

So now we know why it's done on real sailplanes. Can we do it in FS2002?

FS uses a single Flap Angle Limit value. Thus the other end of the flap travel is implied to be zero. There is also a single flap lift coefficient and a single flap drag coefficient. The coefficients are per radian, and thus are multiplied by the deflection which is the value of the flap control input (0-1) times the Flap Angle Limit. So we can have a wing with a standard (unflapped) lift and drag, and then add to those values more lift and drag in proportion to the deflection of the flap control. Ordinarily one would assume that this means that negative flaps (and the associated reduction of lift and drag) are not supported. But, let's look at it another way: assign lift and drag to the unflapped wing which represent the negative flap condition, then tune the flap lift and drag settings so that with the flap control at some intermediate setting representing neutral flap, the total lift and drag will actually be the wing lift and drag plus the amount contributed by the flap deflection from the full-negative starting point to neutral. If this is tuned correctly, we should be able to get a very believable variable-geometry wing, particularly for small deflections.

I did a study using xfoil to see what would be an appropriate change in lift, drag and moment over a small range of flap travel. Using a NACA 747A415 section, I simulated a 0.2c sealed plain flap and tested over the range of -5 to +10 degrees at Reynolds numbers appropriate for the Nimbus-4. This image is a snapshot of some of that testing. From these tests I learned that, over a small range of flap deflection, a linear change in the three coefficients is fairly realistic, and I measured coefficient deltas to use for the Nimbus-4.

home | flight dynamics