Delta Wing - Natural Control Reversal

[RainmanAero]

Greetings mega-brains!

A delta wing configuration, because it does not have a separation between the main wing and the horizontal tail which holds the pitch control surface(s), has a natural aerodynamic, temporary control reversal when initiating a pitch maneuver from a steady-state, trimmed, unaccelerated condition. In controls terminology this is called a "non-minimum phase zero".

What happens is this:

1) The generation of pitch rate via pitching moment lags behind the production of lift and drag.
2) As such, when you deflect the elevons upwards in order to cause a pitch-up, you initially spoil the lift-generation of the main wing before you have a chance to build up enough pitching moment to overcome the pitching moment of inertia.
3) This results in the flight path vector dipping downward (i.e. the plane falls), until sufficient angle of attack is generated to overcome pitch inertia and begin pitching the nose upward.
4) The opposite effect occurs when trying to pitch down (you initially lift upwards before you go down).
5) This is counter-intuitive (and somewhat alarming) to a pilot who is not aware of this natural aerodynamic tendency.

My question (and where I am looking for help, suggestions, and references) is this:

Short of the typical solution of adding a canard, which is deflected in opposition to the elevons to counter the initial loss of lift, is anyone familiar with any control system techniques that can minimize this tendency (with the hope of avoiding it) over the entire flight envelope.

I've heard "tribal knowledge rumors" about manipulating the apparent aircraft instantaneous center of rotation, but no one can give me details.

Gurus? What do you say? Many thanks in advance,
Rainman

 

[Danger]

Rainman, you are so far ahead of me in this stuff that I barely dare speak in your presence. I haven't had a bird strapped on in over 30 years, and then I was only officially okay on 172's. The reality of what I could do was a bit different, but that's how it is on paper.
Since your question is in regard to whether or not anyone knows of a compensation method other than canards, I have to answer 'no'. My approach to the issue, however, would be to consider moderate thrust vectoring.

 

[Astronuc]

Hey RainmanAero - I figured you were the aero guru

I'm sure Fred will have something to say on this subject. Meanwhile, Danger has a good point - thrust vectoring - if one wishes to avoid external control surfaces.

Which brings me to wonder -

I imagine one wishes to avoid canards because they can add to the radar cross-section. Similarly, tails and other control surfaces/protrusions are similarly undesirable. I imagine that the jet engines are inboard (which could make thrust vectoring more difficult/challenging), also to avoid radar signature. Would a leading edge control element (e.g. 727, or B58) be undesirable?

Thinking of the B-58 Hustler, which is a really cool aircraft, I'm sure Convair must have enountered this issue. Perhaps Convair's or NACA's archives have something on Delta wing stability, and would have thought that Northrop's archives would have material on this issue.
http://en.wikipedia.org/wiki/Flying_wing

One way may be to link engine thrust with elevons, so the thrust increases with elevon up, and decreases with elevon down.

One question, are the elevons toward the roll (central) axis, or more outboard?

What are the constraints on this problem?

With regard to shift COR (CM), I've heard moving fuel or shifting mass, but that was long ago, and I don't think shifting mass is prudent. I think commercial airlines, like 747, can shift fuel between forward and aft tanks in the fuselage between the wings.


Just can't resist -

http://www.aviation-history.com/convair/b58.html

http://www.xs4all.nl/~mvburen/b-58/

http://www.b-58hustler.com/

http://home.att.net/~jbaugher2/b58.html

http://en.wikipedia.org/wiki/B-58_Hustler

http://www.b58hustler.net/

 

[Danger]

B-58 Hustler?! Man, I haven't even thought of that thing in decades. That was one of my favourite models that I built when I was about 12. Thanks for a good memory.
The main problem that I can see with mass-shifting, which doesn't matter to a commercial plane, is response time. I don't care what kind of pumps and lines are installed; I just can't see transferring a few hundred pounds of fuel in the fraction of a second that is required in a combat situation.
The leading-edge control sounds pretty promising, as does the auto-throttle/elevon link. That also brings to mind the concept of rear mounted speed brakes tied to the elevons. If you pop a top one when they go up, it might (and I'm not at all sure about this), impart a postive pitch. One on the underside, conversely, could hoist the tail a tad.
As to the thrust vectoring, I was thinking of just a simple louvre system behind the exhaust, which would still work with internal engines. I don't know what that would do to normal flight characteristics or afterburners, though, and I suspect that it would play havoc with the IR signature.

 

[Mech_Engineer]

While I am by no means an Aerospace expert, I do remeber the general idea of a project that was being worked on by several graduate students at the University of Arizona.

Their general design was flight control of an aircraft without the use of control surfaces. The concept they were working on was directly changing the pressure above/below the wing through the use of pumps and pinholes in the surface of the wing. By either pumping air from or adding air to a side of the surface (I'm not sure which it was), it was my understanding they were actually able to gain control of pitch, yaw and roll.

It would seem to me that if you had a system like this, which was able to actively increase or decrease the pressure in a specific area of the wing, you would be able to more actively control the lift of the wing, at least in theory.

Being a mechanical, I only heard about this project through my aero friends, but if you would like to know more about this project, I would recommend contacting the U of A AME department, it is possible the graduate students have written papers on this projects theories, prototypes, and applications.

 

[Danger]

Wow, Mech... that is seriously cool. Probably a bugger to implement, but elegant.

 

[Mech_Engineer]

Wow, Mech... that is seriously cool. Probably a bugger to implement, but elegant.

And a bugger to keep the holes from plugging up I would suspect. Of course, if you have enough air pressure that might not be an issue... Just throwing out ideas.

 

[Danger]

And a bugger to keep the holes from plugging up I would suspect.

Just make sure you stay above the birds.
By the way, is this system by any chance integrated with 'smart skin' technology?

 

[Mech_Engineer]

Just make sure you stay above the birds.
By the way, is this system by any chance integrated with 'smart skin' technology?

As far as I know, they implemented it on the wings of an RC plane, by making an aluminum skin with pinholes in it, and a series of ducts within the wing to supply air to the holes.

 

[LURCH]

Regarding thrust vectoring; NASA had an x-plane called the "X-31" that used thrust vectoring for most of its manuevering. They proved the concept, but more research and testing is required. Since aerodynamic temporary control reversal only occurs during pitch manuevers, this could prove to be a very viable solution, as the X-31 showed that thrust vectoring could possibly replace controll surfaces for pitch and yaw, leaving aerodynamic controlls for only roll.

 

[Danger]

an aluminum skin with pinholes in it, and a series of ducts within the wing to supply air to the holes.

So essentially they have a flying air-hockey table.

 

[Mech_Engineer]

So essentially they have a flying air-hockey table.

I'm thinking that's a pretty good analogy

 

[russ_watters]

Same idea is used for boundary layer control and blown flaps:
http://www.aerodyn.org/Drag/blc.html
http://en.wikipedia.org/wiki/Boundary_layer_control_system

The blown flap system provides extra downwards airflow which not only improves the flap, but also improves the airflow. A small amount of the compressed air produced by the jet engine is "bled" off of the compressor stage and piped to channels running along the rear of the wing. There it is forced through slots in the wing flaps of the aircraft when the flaps reach certain angles. This air follows the flap profile, aimed downward to provide more lift.

The bleed air prevents the boundary layer (slow-moving air that accumulates on the airframe surface) on the upper surface of the flap from stagnating, improving lift. At low speeds the amount of air being delivered by this system can be a significant fraction of the overall airflow, effectively "fooling" the plane into thinking it is flying at a higher speed. This costs little, during landing at least, as the engine power is significantly reduced anyway.

Boundary layer control systems usefully lower the stall speed of an aircraft, making them useful for STOL aircraft (like cargo transports intended for use on short fields) and high-performance fighter aircraft with poor low-speed characteristics. Their disadvantage is that they rob the engine of some thrust while in use, which can harm take-off performance, particularly in "hot and high" conditions.

In general, blown flaps can improve the lift of a wing by two to three times. Whereas a complex triple-slotted flap system on a Boeing 767 delivers a coefficient of lift of about 2.8, external blowing improves this to about 7, and internal blowing to 9.

 

[Danger]

Great links, Russ. Thanks.

 

[RainmanAero]

To all my Phyforum bretheren: THANKS for the fantastic replies in this thread!

Hey RainmanAero - I figured you were the aero guru

Thanks for the vote of confidence, Astro, but I am nowhere near the league of you and many others on this forum. I just feel honored to be able to contribute here and there on those few things I do know fairly well.

Now on to some of the thoughts and suggestions:

1) Thrust vectoring - Yep, we already have this in one of our configs. It is expensive and causes MTBF of the engine to decrease fairly significantly, so we'd like to avoid if we could. But the TVC is also ideal for lift compensation in bank maneuvers at low speed, where delta wings have a "falling off the edge of the cliff" lift loss once you get beyond a specific bank angle.

2) Canards - You're right on when it comes to RCS, Astronuc. You can imagine that what I am working on wants to be a stealthy bird.

3) B-58 Hustler - Actually, our human factors pilot that I work with is a former Hustler driver. He maintains "you live with it, and it is really just a training problem." But that was before the age of snazzy, non-linear, digital FBW flight control systems where you can make a rock fly like a sailplane with enough software! So I kind of feel it is my job to exhaust all possibilities.

4) High-pressure blowers - I can't talk about details, but suffice it to say this is being looked at by the "configurators" of the project I am working on.

5) Auto throttle "power leading pitch" - This was the first thing I added to the control system when I joined the team, mostly because I am trying to get our Cooper-Harper handling qualities scores UP, and the simplest way to do that is to take the throttle/speed task off the pilot's hands so he can focus on the stick. And in addition I have added "power lead" compensation by spooling the engine up in concert with the pilot's pitch-up stick demand (rather than waiting for the aerodynamic response and compensating with flight path angle). This helps a small amount, but not nearly enough.

One thing I forgot to relate is how my hands are tied. You see, I am only the flight control laws/software guy. The aero and control effector configurators are another group, and it's their charter to come up with the control power means to generate the forces and moments that I can use to make it do what we want it to do. And even these configurators are "subordinated" to the Low Observables gurus.

In any event, I tip my hat in appreciation to you gents. I wanted to make sure I left no stone unturned, and your replies have at least confirmed that I have looked at the right things (so far). And I do have one of my younger engineers out doing literature searches in our own company files, as well as NASA, AIAA, CalSpan, etc.

Thanks again!
Rainman

 

[Danger]

I'm just glad to see that I didn't make a fool of myself.
For the hell of it, though, I'm going to keep thinking about it. If anything radical crosses my mind, I'll get back to you.
And thank you for initiating a thread that really gave me something to chew on in an area that I do have some small knowledge of. I really enjoyed it.

 

[FredGarvin]

The first thing that comes to mind is what they used to fix the pitch authority problem on the Bell X-1 in the transonic range; they switched from a hinged elevator to a completely moveable horizontal stabilizer. The only problem there being that the entire surface may still be effected. Is resizing or moving the horizontal stabilizer an option?

 

[Danger]

Wouldn't that necessitate re-designing the Stealth surface angles?

 

[RainmanAero]

Hi Fred,

The first thing that comes to mind is what they used to fix the pitch authority problem on the Bell X-1 in the transonic range; they switched from a hinged elevator to a completely moveable horizontal stabilizer. The only problem there being that the entire surface may still be effected. Is resizing or moving the horizontal stabilizer an option?

I understand what you are saying, but perhaps you forgot that this is a delta wing configuration. So there is no horizontal stabilizer like a conventional "wing+tail" aircraft. In fact, this is the very reason why this initial control reversal happens. If we had a separate horizontal tail, moving the elevators would not have such a drastic effect on the lift produced by the main wing.

However, thanks for the contribution.
Rainman

 

[FredGarvin]

Details, details. Hmmmm. I thought the older delta wings had horizontal stabs. A quick peek shows you're correct. Hmmm... perhaps the you are limited to control system tweaks to try to corrall the problem. Without an extra surface, I don't see how you would fix it.

 

[Danger]

This might be totally impractical for any number of reasons, but I'm thinking back to a plane that I designed for a novel.
Would you be able to utilize positive and negative pitch jet ports in the nose? Not jet engines, but perhaps a compressed hot air source bled from the main engines or just regular compressed air from an onboard compressor. I have no idea what thrust might be required, and I guess that it would be determined by the leverage factor relative to the centre of gravity. Like Mech, I'm now just tossing ideas into the ring without worrying about the practicality.

 

[RainmanAero]

Hi Fred & Danger:

Details, details. Hmmmm. I thought the older delta wings had horizontal stabs. A quick peek shows you're correct. Hmmm... perhaps the you are limited to control system tweaks to try to corrall the problem. Without an extra surface, I don't see how you would fix it.

I tend to think the same as you on this issue, an extra control surface (canard would be the simplest) would be the easiest way to correct the problem. But Danger's idea is also interesting, in that he is pointing out that the "real" requirement is simply to be able to generate a force that restores the lost lift (temporarily) at the start of the maneuver.

Would you be able to utilize positive and negative pitch jet ports in the nose? Not jet engines, but perhaps a compressed hot air source bled from the main engines or just regular compressed air from an onboard compressor. I have no idea what thrust might be required, and I guess that it would be determined by the leverage factor relative to the centre of gravity. Like Mech, I'm now just tossing ideas into the ring without worrying about the practicality.

Where I think this is headed is to the fact that what is really needed is a total IMPULSE... A specific magnitude of Force for a specific duration of Time. And when one starts talking impulse control a discussion of rocket engine (mono or bi-propellants) is not far behind. High-pressure bleed air from an engine might be able to produce such impulse jets, but at the expense of robbing too much power from the turbine flow. For higher specific impulses I would think you could easily generate such temporary forces at the beginning of each pitch maneuver with reaction thrusters.

Does it add a huge complexity to the overall aircraft? Hell yes. Does it solve the problem? Again the answer is yes. Would it be cheaper to just train the pilot to expect it and deal with it? Probably. And so I guess part of my job is showing the cost vs. benefit trade-off of trying to remedy the problem vs. living with it.

Rainman

 

[Cyrus]

I am suprised that this is not in literature somewhere. I am sure you are not the first person to come across this problem and it should be in a controls paper.

**With all due respect to everyone in here, we are not qualified to answer your question. This is a very complicated question, that frankly, we cannot give a serious answer to.

Have you looked for any AIAA papers? It would be interesting if you could share any results you find from published papers.

 

[FredGarvin]

Engine bleed is used for a lot of other items on the aircraft. IF you could produce the required impulse from the regulated bleed flow, then it's usage would not have an adverse effect on engine performance. Chances are that the engine is not dumping enough into the bleeds to provide enough to overcome aerodynamic loads. Somone from your airframe and powerplant groups will need to run the numbers.

 

[Astronuc]

Like Mech, I'm now just tossing ideas into the ring without worrying about the practicality.

This is what the experts do during brainstorming sessions. First, consider anything and everything, then worry about making it practical later.

Some additional thrusting (either propulsive or vectored/directional) would seem desirable, since use of any additional protrusion (e.g. canard) costs fuel (increased drag) and stealth.

Keep in mind that the maneuver is short term, so the solution additional thrusting is also short term. The ideal solution should not add much mass or change the aircraft's surface area.

 

[Danger]

If this was a once-or-twice-per-mission manoeuvre, I would even seriously consider detonating a concussion missile just below and in front of the plane and riding the shockwave up. Of course, that's not overly Stealthy.

Hey now... I just thought of something else, but it's way, way outside of my arena. What would be the effect of injecting ram-air into the exhaust stream? I'm thinking of a modification of the speed-brake idea a few posts back. Would it do any good to pop a scoop on the top or bottom of the tail and channel the air inward through the exit flow from the engines? (Maybe even inject a bit of fuel with it to approximate a vectored afterburner?)

 

[RainmanAero]

Playing around with the concept

Greetings and thanks to the PF folks who contributed to this thread.

I thought it would be a good thing to give you an update on this "problem".

I think I am beginning to see/agree that the short term reversal is a fairly predictable effect. Once you understand that the airplane does this, you learn to compensate for the "dips" as you fly around and change the flight path angle. I have also found that there are computational compensations that can be applied to "shape" how the pilot feels the eventual aircraft response at his/her seat position. These came from the results of our literature searches in our company files.

This is a great place to be able to share such thoughts and get such great discussion feedback. Thanks a lot folks.

Rainman

 

[Danger]

Awww... does this mean that we don't get to design a new birdie?

I can see, from a pilot's perspective, that it's something that you just anticipate and compensate for. From a tinkering perspective, I'd rather eliminate the problem. From a cost-effectiveness perspective, then the pilot anticipation is of course the correct option... as long as it doesn't result in the loss of said pilot. I suppose that given the hundreds or thousands of hours that these folks put in on simulator work and excercise flights, it should be second nature by the time they're in a critical situation.

Still, though, it bugs the hell out of me to leave a problem unresolved.