Forum: F-22A Raptor

Right. I mentioned in an earlier post that "tail volume" is tail area x moment arm (tail center of pressure distance from the airplane cg). In determining required tail size, its "volume" is scaled according to the airplane pitch inertia (Iyy) and required pitch acceleration. And as PhillyGuy says, TVC affects all that too.

F-16.net Sponsor

I do wonder how the plane would handle if it couldn't use the thrust vectoring (say, if it conveniently got stuck in the neutral position).

johnwill wrote:

A few years ago, I did an analysis of T-50 (Korean trainer) flight test data which showed that every degree of flap deflection reduced tail effectiveness about 1/3 as much as each degree of tail deflection. So if the tail is 5 degrees trailing edge down and the flap is 15 degrees down. the tail is totally ineffective. That is a subsonic case, and supersonic would probably be different.

I'm a little late to this thread, but I find this curious. Was this a dedicated PID study, or just data gathered during normal testing? The reason I ask is that I'm painfully aware of how difficult it can be to extract individual control surface effectiveness (I assume you meant Cm_dh) from flight test for highly augmented aircraft. I'm not familiar with T-50, and don't know what kind of static margins it has, but it appears to be similar to an F-16. Assuming that the free airframe is unstable (like an F-16) such areas of ineffectiveness would make maintaining controlled flight effectively impossible.
Regardless of stability, such regions would, for instance, make PA flight (where the flaps would be down) unworkable. Ditto for AR mode, where the flaps are typically dropped some.

Tinito_16 wrote:

I do wonder how the plane would handle if it couldn't use the thrust vectoring (say, if it conveniently got stuck in the neutral position).

It's really not a problem. Remember, for most of the flight envelope, vectoring is either not used, or not used signficantly. It's really only at slow speeds, high AOA, and high high altitude that it plays a signficant role.
The aircraft is required to be perform all normal flight operations (including air refueling, landing, etc) with no vectoring (due to failures). It also must able to recover from any angle-of-attack and attitude positively, following any sudden loss of vectoring.

As for the effect of the trailing edge flaps on the horizontal tail, it's not a significant player for the vast majority of normal flying. At high AOA, the flaps and ailerons are scheduled "up" (for other reasons), but that likely has some beneficial effect on the tails as well.

Raptor claw,
I'm a structures guy as you know, so my interest was in determining what factors contributed to structural load on all major surfaces. Design loads are based on CFD, wind tunnel, mass properties, stiffness, and maneuver simulation, all of which have some uncertainty involved. Only in flight test are the real loads measured, and similar to PID, equations for real loads can be developed from the flight test data. These equations can be used for many different purposes as part of the structural integrity task.

Data used in the analysis comes from normal flight load test maneuvers, windup turns, rolls at several g levels, three kinds of sideslips, abrupt pitch, etc., all performed at four altitudes and about eight different mach numbers. For each specific mach/altitude, many flight parameters and structural loads are put into a regression analysis of all test maneuvers sampled at 40/sec. With maybe ten maneuvers lasting 3 to 5 seconds each, that's up to 2000 samples in the regression. Step by step, unimportant parameters are removed and the regression rerun until all significant variables have been identified. Throwing all maneuvers into the regression results in the necessary variable independence.

For the tail load equation, the variables were what you would expect, AoA, dh, df, p, pdot, q, qdot, dlef, nz, etc. No real surprises, but I was really pleased we could extract those coefficients with very high R^2 results, .98 being commonly found. And for the condition I studied, the flap coefficient was about -1/3 of the tail coefficient, in terms of tail load per degree deflection of the flap or tail.

About your concern that having an ineffective tail would make controlled flight unworkable, such is not the case at all. If you put the flap 15 degrees down and move the tail through its full range of motion, when the tail is full trailing edge down, there will be an up load on the tail. When the tail is full trailing edge up, there will be a down load on the tail. Therefore at some tail angle, there will be zero load on the tail. The tail is ineffective at that angle, but still works at all other angles. In my example of df=15 and dh=5, that was the point of zero tail load. True tail ineffectiveness is demonstrated by the F-16 at AoA=60, where deep stalls can occur. As you know, in that case, tail load is near zero at all tail angles.

I have to admit, the idea for doing this analysis came from observing my S&C friends trying to extract PID from their data.

johnwill wrote:

.... when the tail is full trailing edge down, there will be an up load on the tail. When the tail is full trailing edge up, there will be a down load on the tail. Therefore at some tail angle, there will be zero load on the tail. The tail is ineffective at that angle, but still works at all other angles.

I think what I would disagree with is simply the use of the word "ineffective" as you correlate it with zero tail load. "Ineffective" indicates that the surface is somehow not functioning as desired/designed. Specifically it would imply that the tail has entered some kind of "dead-zone" where its local "lift-per-degree-of-deflection" slope has gone to zero. That's not at all the case here. The flap has simply moved the tail's crossover point - the deflection at which it changes from positive to negative lift. The effectiveness (how much the tail "lift" changes as it deflects) may not have changed at all (or very little).

I guess you could apply the word as it relates to the tail "effectiveness" at full deflection. If deflecting the flap moves the (local) CLo point of the tail from (say) Dh=0 to Dh=5 degrees, that would likely mean that full trailing-edge down tail (tail) would give less aircraft nose down moment. In other words, if Dh=25 is full deflection, the lift the tail is now producing at Dh=25 would be closer to what it produced at Dh=20, before the flap was deflected (assuming linearity, of course). Of course the opposite would likely apply in the other direction - full tail t.e. up deflection would give more nose-up aircraft moment than it did before.

I wouldn't have sweated this in the first place, but I think it (the use of the word "ineffective") led to some confusion. For instance, Tinito's comment: "Interesting fact there. Are there any measures/design changes that fix this, or is it something you just 'live with'?"
The fact is there's nothing there to fix, or live with, it's just the way it is.

You are entirely correct. But the answer to Tinito's question is we don't live with it. Here's why. On the YF-16, the original ratio of tail roll commands to flaperon roll commands was a constant 0.25. But supersonic roll performance was not good, so the ratio was increased to 0.5 between 0.95 mach to 1.05 mach. That same ratio change was applied to all F-16s. By determining the effects of the flaperon on the tail, we adjusted to tail ratio to get the desired performance.

On the Taiwan IDF, the flaperons are proportionally larger than on the F-16, so they really affect the tail capability to roll the airplane. Here, based on flight test loads, not only did we change the supersonic ratio to 0.6, we also reduced the commands to the flaperon. The wing was approaching aileron reversal, so little was lost in taking out flaperon and much was gained in adding tail deflection. An interesting thing (to me anyway) is that subsonic tail roll contribution on both airplanes (and the T-50) is negative, due to the flaperon effects.

So we don't just live with it, we use it to our advantage.

If you are interested in getting a handle on individual surface contribution to total airplane coefficients, as you said earlier, you might consider using measured loads from flight load test. For example, you can use vertical and horizontal tail loads to calculate individual and total tail roll coefficients, then subtract those from airplane total to get wing roll coefficients. Or you can use wing loads to calculate wing roll coefficients directly.

One of the problems we have is lack of understanding and communication between different disciplines. There is much to be learned from others.

The horizontal on the F-22 is advertised as being as large as the wing of an F-16.

arl8733 wrote:

The horizontal on the F-22 is advertised as being as large as the wing of an F-16.

The F-22 stabs are big, but not that big.

Here's a comparison of an F-16, F-35, and F-22:

arl8733 wrote:

The horizontal on the F-22 is advertised as being as large as the wing of an F-16.

Based on this, it's close, but no cigar.

Note that the stab spans of the Tomcat, YF-23 and Flanker all exceed (however slightly) the F-16's wingspan.

(Also note the difference in size between the JAS and the Flanker. Laughing

)

Tinito_16 wrote:

Here's a comparison of an F-16, F-35, and F-22:

The scaling in that image is off. The JSF and F-16 should be smaller.

Quote:

The scaling in that image is off. The JSF and F-16 should be smaller.

I thought it was pretty accurate, but thanks for pointing it out Smile

Forum: F-22A Raptor

Stabilators