A Quick Explanation of Forward Swept Wings

[Collimatrix]

 

Every so often someone asks a question about the advantages of forward-swept wings, and usually they get a shitty half-assed answer about how they somehow improve maneuverability and stuff.  I will attempt to provide a fully-assed answer.

 

The short version is that forward swept wings do roughly the same thing as conventional aft swept wings; they increase critical mach number.  I found an excellent video explaining transonic effects, so watch that first if you don't already know what that is.

 

Typically, a straight wing starts experiencing shock wave buildup at around mach .7.  These effects are generally bad; control surfaces lose effectiveness, the aircraft's center of lift moves, stability can decrease, and drag greatly increases.

 

It's generally desirable to delay the onset of this badness.  The critical mach number is strongly affected by the thickness to chord ratio:

 

 

 

 

 

So, critical mach number could be increased by having really thin wings.  The F-104 does this, but at the expense of having ridiculously tiny wings that generate barely any lift and no internal volume for fuel storage.

 

Critical mach number could also be delayed by having wings with a normal thickness, but very long chord.  This would improve the supersonic performance of the wing, but subsonic drag would be negatively affected, because the wing would have a large amount of induced drag, and additional wetted area that would cause more drag.

 

Finally, the wing could be swept.  This would increase the chord length relative to the airflow, but would not give the wing undue surface area and thus subsonic drag.

 

In theory, the critical mach number could be increased by a factor equal to the inverse of the cosine of the sweep angle (much like calculating the LOS thickness of tank armor, and for the same reason), but secondary effects mean that it's less effective than this.  The practical effect of sweep on drag coefficient looks about like this:

 

(from Design for Air Combat)

 

This, incidentally, is why the ME-262 doesn't really have swept wings.  The change in M_cr is basically negligible for any leading edge sweep under thirty degrees.

 

Note that this logic applies whether the wings are swept forwards or backwards; as far as delaying and reducing the transonic effects, forward or rearward sweep should be equally effective.

 

There are some secondary effects that make forward-swept wings more desirable.  One of these is spanwise flow:

 

 

In any swept wing, the air isn't just flowing over the wing, it's flowing across them as well.  This means that while pulling Gs the tips of the wings will stall first.  Since the tips aren't producing lift anymore, but the rest of the wing is, the center of lift of the wing moves forward, which means that there's more pitch-up torque on the plane, which means that the nose goes up even more and the stall gets worse.  This is known as the "sabre dance," as the F-100 displayed this undesirable property.  With the wings forward swept, the root of the wings would stall first (although in practice, forward swept wing aircraft tend to have the wings attached well aft, so the CL still shifts forward during a stall)

 

To make matters worse, the air spilling out sideways and the early stall interfere with the effectiveness of the ailerons, which means that the aircraft can lose roll control effectiveness as it increases AOA.  This is a particularly alarming behavior during landing, as speed is low, AOA is high, and keeping the aircraft level is of paramount importance.

 

Additionally, the air spilling out outwards towards the wingtips reduces lift.  Reducing this bad behavior increases lift coefficient, therefore.

 

So, forward swept wings are a little more efficient, aerodynamically than aft swept wings.  Why aren't they more popular?

 

The problem is something called aeroelastic divergence; which is engineer-speak for "the goddamn wings try to tear themselves off."  I will attempt to illustrate with the finest MS pain diagrams:

 

 

The amount of lift that a wing generates is a function of the angle of attack.  The wing will generate more lift the more inclined it is relative to the airflow.

 

Wings in the real world are, of course, not perfectly rigid, so when they generate lift in order to pull the weight of the fuselage through the sky, they bend slightly.

 

In swept wings, the wings aren't just bending, they're twisting as well because the center of lift is not aligned with the structural connection between the fuselage.

 

In an aft-swept wing, the force of the lift tends to twist the wings downwards.  Increasing the angle of attack will increase the lift, which will increase this downward twist, which is a naturally self-limiting (negative feedback) arrangement.

 

In a forward-swept wing, it's exactly the opposite.  When the angle of attack increases, lift increases and the wings twist themselves upwards, which increases lift even more which increases the twisting...

 

This is why forward-swept wings had to wait until magical composites with magical properties were available.

[Toxn]

God that plane is pretty.

[Belesarius]

GI JOE! Knowing is half the battle!

[Priory_of_Sion]

Hmm, I never really questioned why FSW were not sought after until relatively recently. The wings trying to rip themselves off seems like a bit of a problem though. 

 

[Xlucine]

So the body of the aircraft becomes one huge-ass wing fence?

 

I'm still not sure if I understand the idea behind it helping improve the ratio of chord to height - surely wing area (and so wetted surface) would be constant? Sure, the chord decreases (measured normal to the leading edge), but the length increases by the same amount so I don't see why it's different to a straight wing with the same chord measured along the airflow. I do know a guy who's meant to know a lot about this kind of stuff though, so I will talk to him and get back to you.

[SuperComrade]

This is why forward-swept wings had to wait until magical composites with magical properties were available.

 

Didn't stop zee Germans from trying

[LostCosmonaut]

Didn't stop zee Germans from trying

 

More than once, in fact;

 

[xthetenth]

So the body of the aircraft becomes one huge-ass wing fence?

 

I'm still not sure if I understand the idea behind it helping improve the ratio of chord to height - surely wing area (and so wetted surface) would be constant? Sure, the chord decreases (measured normal to the leading edge), but the length increases by the same amount so I don't see why it's different to a straight wing with the same chord measured along the airflow. I do know a guy who's meant to know a lot about this kind of stuff though, so I will talk to him and get back to you.

 

It's an alternative method to improving the aspect ratio, not a method of improving the aspect ratio.

 

 

Didn't stop zee Germans from trying

 

Proving yet again why we take manufacturer's claims with a salt mine until the damn thing flies and exhibits those selfsame characteristics because a design made by people who don't know the problems and a design made by people who have solved the problems can look similar and even have reasonably similar expected performance (if the means of making it actually work aren't too expensive).

[Collimatrix]

I'm still not sure if I understand the idea behind it helping improve the ratio of chord to height - surely wing area (and so wetted surface) would be constant? Sure, the chord decreases (measured normal to the leading edge), but the length increases by the same amount so I don't see why it's different to a straight wing with the same chord measured along the airflow. I do know a guy who's meant to know a lot about this kind of stuff though, so I will talk to him and get back to you.

 

I felt that I understood the issue quite well until you asked that question.

 

I'm really not sure what the advantage of a swept wing over a straight wing with the same trigonometric wing chord and span is WRT M_cr.  Apparently there are some secondary advantages to swept wings; they suffer a smaller transonic center of lift movement as a percentage of MAC, but surely that isn't the deciding factor?

 

Proving yet again why we take manufacturer's claims with a salt mine until the damn thing flies and exhibits those selfsame characteristics because a design made by people who don't know the problems and a design made by people who have solved the problems can look similar and even have reasonably similar expected performance (if the means of making it actually work aren't too expensive).

 

According to Design for Air Combat, you can get away with about fifteen degrees of forward sweep before the aeroelastic problems become too much for conventional metal structures to deal with.  As you can see from the chart above, this would do basically nothing for transonic peformance.

 

I believe that in the Hansa jet and JU-287, the forward sweep is intended to move the wing roots backwards so that there's more unobstructed cabin/bomb bay space.

[Belesarius]

_{Grumman X29 :}

[Donward]

Meh...

 

Planes. Planes. Planes. Planes.

 

Wait a tick...

 

 

 

 

 

Is that a redhead? Or a faded photo.

 

Doesn't matter to me.

 

Planes and redheads.

 

Don's interested is suddenly piqued.

[Priory_of_Sion]

Meh...

 

Planes. Planes. Planes. Planes.

 

Wait a tick...

 

 

 

Is that a redhead? Or a faded photo.

 

Doesn't matter to me.

 

Planes and redheads.

 

Don's interested is suddenly piqued.

How else was I suppose to get you to post in this thread? I don't know who she is, but she apparently was somehow involved with a FSW version of the F-16 which is pretty hot.

 

Spanwise airflow patterns on a FSW Mig-23 model. 

[Collimatrix]

Here you can see one of the issues with FSW; the area behind the wings is in some pretty hairy, turbulent airflow which reduces the effectiveness of any control surfaces that live there.  For this reason, most FSW designs feature canards.  I'm not entirely sure what the rationale behind having both canards and tailplanes on the berkut is.

[Belesarius]

Here you can see one of the issues with FSW; the area behind the wings is in some pretty hairy, turbulent airflow which reduces the effectiveness of any control surfaces that live there.  For this reason, most FSW designs feature canards.  I'm not entirely sure what the rationale behind having both canards and tailplanes on the berkut is.

 

I wonder what effect combining forward swept wings and thrust vectoring would have?

[Priory_of_Sion]

Here you can see one of the issues with FSW; the area behind the wings is in some pretty hairy, turbulent airflow which reduces the effectiveness of any control surfaces that live there.  For this reason, most FSW designs feature canards.  I'm not entirely sure what the rationale behind having both canards and tailplanes on the berkut is.

Here is a model of the F-18 using about the same colored smoke system for comparative purposes. 

I don't see quite as much messiness behind the wing.

 

I would think the tailwing wouldn't be too effective on a FSW plane but could still be somewhat useful as a elevator. Then again, I'm just a little smarter than Donward on this subject. 

[Collimatrix]

Thrust vectoring works its magic best, as I understand it, when conventional control surfaces have lost their effectiveness.  When the plane is at very high altitude, or when it's not moving fast enough, or when the wings are mostly stalled, turning the thrust sideways will still produce pitch or roll moments or what have you.  FSW+thrust vectoring might be a good combination for short takeoff and landing.

 

The SU-47, as I understand it, began as a design study for a Soviet carrier bird.  The Soviets had just started doing carriers in the 1980s, and were discovering that landing an airplane on a boat is tricky.  Very tricky.  The SU-33 navalized flanker worked well enough, but if they could design something that was easier to land on a carrier, the Naval Aviation pilots would be much obliged.  The forward swept wing was to give better lift for a given AOA, which would either allow the pilot to slow down or allow the pilot to go the same speed, but have the nose closer to level (both good things), and the better roll authority at high AOA of the FSW would better allow them to keep the wings level while landing on a carrier, which is also kind of important.

 

Eventually the design evolved into a next-generation, land-based replacement for the SU-27 family.  There weren't really any takers, so the SU-47 ended its flight-worthy days as a technology tester for the PAK-FA program.

 

It is remarkable, given the state of the Russian economy, that Sukhoi has been able to debut two new fighter designs in the post-Soviet era, as well as continuing to update the SU-27 design.  I'm sure Lockheed Martin is thankful for Sukhoi's ingenuity, and ability to make scary-looking fighter planes that they might export to rogue nations on the limited R&D budget they have available.

 

Here is a model of the F-18 using about the same colored smoke system for comparative purposes. 

I don't see quite as much messiness behind the wing.

 

I would think the tailwing wouldn't be too effective on a FSW plane but could still be somewhat useful as a elevator. Then again, I'm just a little smarter than Donward on this subject. 

 

I think that in that picture, most of the turbulence is coming off of the LERX (the highly swept surfaces that blend between the wings and the fuselage) and not the wing.  That is actually what they're supposed to do; by some black magic, the vortex from the LERX adds energy to the flow over the wings, which helps delay the wings from stalling.

 

They don't seem to be mucking up the flow over the horizontal stabilizers that much, but I know that the baby hornet had issues with the turbulence from the LERX buffeting the vertical stabilizers, which was causing the vertical stabilizers to wear out faster than anticipated.  This was fixed on the super hornet by re-sizing everything a little.

[Collimatrix]

 

I'm still not sure if I understand the idea behind it helping improve the ratio of chord to height - surely wing area (and so wetted surface) would be constant? Sure, the chord decreases (measured normal to the leading edge), but the length increases by the same amount so I don't see why it's different to a straight wing with the same chord measured along the airflow. I do know a guy who's meant to know a lot about this kind of stuff though, so I will talk to him and get back to you.

 

OK, I think I found the answer to this.

 

A straight wing with an equal chord/height ratio would actually have a slightly better M_crit than a comparable swept wing.

 

However, in supersonic flight a straight wing would stick out the sides of the mach cone created by the nose of the aircraft, while swept wings could fit inside.  Having your wings stick outside the mach cone is a PITA for various reasons.

[Xlucine]

So it's safe to say that the swept wings on Me 262 resulted in worse performance?

[Collimatrix]

They probably marginally degraded the stall characteristics, and provided a very slight increase in critical mach number and drag-wise mach number that the engines were too weak to propel the airframe to anyway.  By the inverse cosine rule I get that under ideal conditions an 18.5 degree sweep would provide a 5.4% increase in M_crit, and in actuality it would have been less than that.

[xthetenth]

OK, I think I found the answer to this.

 

A straight wing with an equal chord/height ratio would actually have a slightly better M_crit than a comparable swept wing.

 

However, in supersonic flight a straight wing wound stick out the sides of the mach cone created by the nose of the aircraft, while swept wings could fit inside.  Having your wings stick outside the mach cone is a PITA for various reasons.

 

That makes sense and jives with what I've heard elsewhere.

[Collimatrix]

https://www.youtube.com/watch?v=o5vek4J-86E

 

GET HYPE.

 

The sweep angle looks very modest on this, so I suspect that the forward sweep may be for structural reasons (to place the wing spar in a more convenient location, for instance) rather than aerodynamic ones.

 

NASA also uploaded a very nice video of aeroelastic divergence in a FSW:

 

https://www.youtube.com/watch?v=I-tvVp1QT3U