Do vortex generators really work.

[stevron]

Are vortex generators worth fitting?

 

 

[Bob Llewellyn]

Are vortex generators worth fitting?

If you have a trailing edge stalling airfoil, or unwanted separation, and want a lower stall speed and/or possibly better stall handling (if VGs placed correctly!), or lower climb or cruise drag (depending when the parasitic separation occurs); YES. Otherwise, no. They are a very effective boundary layer control mechanism, so if your boundary layer needs more controlling... of course, you should probably wooltuft the whole aeroplane and get some vids of the airflows first...

 

 

[stevron]

Thanks I think, I do not know what your talking about.

 

 

[jetjr]

Mostly sold to lower stall or improve low speed handling, plenty of testimonials including tuft videos showing for this they do work

 

Check put Aussie maker Stolspeed on his website

 

 

[Old Koreelah]

They can lower your stall by about 3 knots. That means a lot less speed and a lower ground run if you have to get into a short paddock.

 

 

[Dafydd Llewellyn]

This has been covered in other threads. VGs need to be used with care; if they are mis-applied, they can result in an extremely vicious stall behaviour. So if you use them, do so ONLY in accordance with a set of instructions that have been developed by testing for YOUR aircraft model. Be especially wary of VG installations that run all the way in from the wingtips to the wing roots.

 

If they are correctly used, they can greatly improve the stall handling. They are quite powerful devices. On a certificated aircraft, an approved VG modification would normally require a re-run of the spin testing.

 

They can, correctly applied, reduce the stall speed approximately 6% - but ONLY if there is sufficient elevator power available at the forward centre of gravity limit, to take advantage of them. That is normally not the case; most aircraft are elevator-limited (i.e. the elevator does not have sufficient authority to take the wing all the way to its stalling angle of attack) at their forward CG limit. If that's the case, it may need VGs on the underside of the tailplane, before VGs on the wing will show any benefit.

 

The do not normally make a measurable difference to cruise speed or climb rate. They do make the aircraft more difficult to wash.

 

 

[facthunter]

It doesn't mean your plane won't develop some unwanted characteristics in the control sense. Airflow is what keeps your plane flying so you are fiddling with the basics. If you don't know what you are doing , don't be a pathfinder. If the mod is proven to work satisfactorily in the circumstances you are in THEN maybe go there. You have to have a problem to solve, to need them. It works by energising the airflow so it doesn't separate so early at high AoA's If there is a "fix-all" everybody would use it. I certainly wouldn't rule them out if there is a need to fix a problem, but you can create another problem by doing it.

 

This reply has crossed with Dafydds,,,,,,,, Nev

 

 

[Yenn]

They have been used on the horizontal stabiliser, to provide more elevator capability. They go on the underside of the airfoil to provide more down thrust. I believe there are some certified aircraft with this approved.

 

 

[Dafydd Llewellyn]

It doesn't mean your plane won't develop some unwanted characteristics in the control sense. Airflow is what keeps your plane flying so you are fiddling with the basics. If you don't know what you are doing , don't be a pathfinder. If the mod is proven to work satisfactorily in the circumstances you are in THEN maybe go there. You have to have a problem to solve, to need them. It works by energising the airflow so it doesn't separate so early at high AoA's If there is a "fix-all" everybody would use it. I certainly wouldn't rule them out if there is a need to fix a problem, but you can create another problem by doing it.This reply has crossed with Dafydds,,,,,,,, Nev

Nev, I used to see VGs as merely a "band-aid" too; but after developing the airflow kit for the Seabird Seeker, I've come to the conclusion that they are rather more than that; they're worth considering as part of the basic design of the aircraft - as Boeing obviously know. By judicious use of VGs and some other devices, it would be possible to virtually eliminate stall-spin accidents; I defy anybody to get the Seeker to drop a wing more than 15 degrees, if at all.

 

 

[facthunter]

If you want an example of where they are absolutely necessary look at the centre engine intake duct on the B727. If a couple of them detach you end up with the engine compressor stalling all the time.

 

What I don't like is fitting them without the necessary analysis of the effect. You may find the stall speed is reduced(it usually is) but the "docility" of the stall changes, markedly. It is a significant design change, sometimes .Nev

 

 

[Bob Llewellyn]

If you want an example of where they are absolutely necessary look at the centre engine intake duct on the B727. If a couple of them detach you end up with the engine compressor stalling all the time.What I don't like is fitting them without the necessary analysis of the effect. You may find the stall speed is reduced(it usually is) but the "docility" of the stall changes, markedly. It is a significant design change, sometimes .Nev

Could be exciting then... an unrecoverable flat spin IS exciting, ce ne pas?

 

 

[facthunter]

You won't get many answers to that question, by those who would know (for a short while). A flat spin is caused by different factors.(but you know that already) including weight distribution on the fore and aft axis. Nev

 

 

[Dafydd Llewellyn]

If you want an example of where they are absolutely necessary look at the centre engine intake duct on the B727. If a couple of them detach you end up with the engine compressor stalling all the time.What I don't like is fitting them without the necessary analysis of the effect. You may find the stall speed is reduced(it usually is) but the "docility" of the stall changes, markedly. It is a significant design change, sometimes .Nev

Nev, installing a significant number of VGs is a major modification, and would nowadays need to be handled via a Supplemental Type Certificate (or under CASR 21 Subpart D, if the mod. is being done by the holder of the TC). It's absolutely something that has to be properly approved, for any certificated aircraft. This would require full stall and spin handling tests at the very least. What's a significant number? Up to somebody's judgement, I suppose, but could be as low as two. There are a number of STC packages available for GA aircraft.

 

My warning was intended for homebuilders of -19 aircraft or VH Experimental aircraft.

 

 

[Old Koreelah]

Nev, installing a significant number of VGs is a major modification, and would nowadays need to be handled via a Supplemental Type Certificate ...My warning was intended for homebuilders of -19 aircraft or VH Experimental aircraft.

I'm in that category and my testing has shown Dafydd to be right on the money. I got a 3kt reduction in the stall, but it now happens with a disconcertingly rapid wing drop. Lethal if it happens near the ground. On the positive side, I have set it up so that it only stalls with max rear CoG.

 

 

[Dafydd Llewellyn]

OK, do your VGs run all the way in to the wing root? If they do, I may be able to suggest some things that may help.

 

 

[dazza 38]

I'm in that category and my testing has shown Dafydd to be right on the money. I got a 3kt reduction in the stall, but it now happens with a disconcertingly rapid wing drop. Lethal if it happens near the ground. On the positive side, I have set it up so that it only stalls with max rear CoG.

I found the same in a Savage cub cruiser , it was fine in a power off stall but in a power on stall, she wanted to flick over .

 

 

[Bob Llewellyn]

You won't get many answers to that question, by those who would know (for a short while). A flat spin is caused by different factors.(but you know that already) including weight distribution on the fore and aft axis. Nev

well, yes... I was hypothesising an aeroplane with inadequate horizontal stab, which inherently resisted entering a developed spin due to lack of elevator power, exacerbated by initial separation in the wing root area (so reducing downwash onto the HS); if VG'd full span, it could well be able to fully stall one wing, and - while the pilot sat in shock thinking "WTF?", it could quickly pass through the incipient phase into a developed spin; by which time the inadequate tailfeathers would not allow recovery nohow, and there's a reasonable probability that the prolonged rotation would begin to flatten before it hit the ground. Which would be bad.

 

 

[facthunter]

The question is why do some spins go flat and other aircraft don't. I wasn't so concerned with the wing drop aspect which may catch some people if they are relaxed or not aware it can happen. If it tends to do it on landing, that is not a good feature. If they are nor expecting it to happen it will be a surprise not welcomed. A lot of people believe their plane is docile and almost doesn't stall. I'm pretty sceptical about that because Cof G position will alter that. VG's will have a similar effect to a Cof G change will it not, in some situations. (Fully, on the wing ). I've been more inclined to worry about the horiz stab blanketing the rudder. If the horiz stab is not adequate it will make the plane supercritical with CofG and a smaller range would be tolerated and that would show in other areas besides the spin. Nev

 

 

[Dafydd Llewellyn]

The question is why do some spins go flat and other aircraft don't. I wasn't so concerned with the wing drop aspect which may catch some people if they are relaxed or not aware it can happen. If it tends to do it on landing, that is not a good feature. If they are nor expecting it to happen it will be a surprise not welcomed. A lot of people believe their plane is docile and almost doesn't stall. I'm pretty sceptical about that because Cof G position will alter that. VG's will have a similar effect to a Cof G change will it not, in some situations. (Fully, on the wing ). I've been more inclined to worry about the horiz stab blanketing the rudder. If the horiz stab is not adequate it will make the plane supercritical with CofG and a smaller range would be tolerated and that would show in other areas besides the spin. Nev

See http://ntrs.nasa.gov/archive/nasa/casi.ntrs.nasa.gov/19720005341.pdf

 

 

[Bob Llewellyn]

Thanks I think, I do not know what your talking about.

Ok Stevron, perhaps I can make it a bit more understandable.

In the case of a sphere, moving through the air at aircraft-like velocity, the air will only flow along the sides from the direct front of the sphere to a point ~92 degrees around - that is, only just past the fattest point. This is a problem, because once the air has separated, it creates a wake which is the drag.

 

Tapering the rear part of the sphere in a "teardrop" shape, with a total length of ~3 times the max. diameter, will persuade the air to stick until the wake represents 3~4% of the maximum cross section.

 

Most lumps of aeroplane - spats, fuselage, cabin, wing, etc - fall between these two extremes; the drag of most small aeroplanes can be reduced by reducing the wake size, which can be done by reducing the separation - where the flow breaks away from the body and eddies around - which can be done with VGs.

 

Back to our original sphere, if we put a ring of VGs a bit fowards of the fattest point, it is likely that the wake size can be reduced from ~95% of the maximum cross section (of the sphere), to 50~30% of that value (or perhaps less).

 

This is one form of "boundary layer" control. As several posters have alluded to, progressive separation - especially approaching stall - is often used by aeroplane designers to give desirable handling, even at a performance cost. Regaining the performance can lose the good handling, possibly in an extreme way.

 

On a poorly designed aeroplane - and there are a few out there - it is potentially possible, with VGs, to improve the performance and not impair the handling - if you get it exactly right. There is a fair bit of specialised knowledge involved.

 

If there exists an "approved" VG kit for an aeroplane you own, then it probably does what it claims, and probably has the bugs ironed out. Otherwise, do you really wish to become a test pilot?

 

 

[Dafydd Llewellyn]

I found the same in a Savage cub cruiser , it was fine in a power off stall but in a power on stall, she wanted to flick over .

I can't guarantee this will work on any aeroplane - however it did work a treat on the two that I have been able to try so far. I'm talking here in the context of somebody developing their experimental homebuilt:

First, remove the VGs on the wing from the wing root, to just past the same spanwise station as the tailplane tips. Leave them on the wing, outboard of that point.

 

Next, fit small leading-edge "fences" to the wing leading edge, in line with the tailplane tips - i.e. a couple of inches inboard of the first VG. See photo. The fence need only go to about 20% of the wing chord, on the top surface, and it need only be around 60 mm high. It can probably be stuck on with double-sided tape, for the initial test.

 

Then: Fit small stall-strips on the INBOARD side of the fence, as shown in the photo. Make sure you put them at the same height on each wing. The stall strips need to be about the size of a piece of triangular file, about 8mm wide on each side, and about 50 mm long. Butt the outer ends of them hard against the side of the fence.

 

Now, load your aircraft about 3/4 of the way to its most aft CG limit, go to a safe height (at least 3000 ft AGL) and try a gentle power-off stall, keeping the ball in the middle. If you get a wing-drop, go home and raise the location of the stall strips by 3 mm, and try again; repeat until you do not get a wing drop. If you do NOT get a wing-drop, go home and lower the location of the stall strips until you do get a drop, then raise them one increment.

 

Do NOT allow the aircraft to start an incipient spin; if the fence comes off on one side, it could get exciting.

 

Once you have found the correct location for the stall strips, progressively move the CG aft to the rear limit. If you get a wing drop in a straight power-off stall, raise the stall strips another increment, and try again.

 

Once you have the stall strips located in this way, you can put them on for keeps.

 

What this does, is to cause the central part of the wing - the bit between the fences - to stall just before the outer part, where the VGs are. That does three things:

 

Firstly, it creates a pair of powerful trailing vortices that turn the opposite way to the wingtip vortices. These vortices in effect magnify the effect of the fences so that the stall cannot spread outboard past the fence. The vortices are not there until the inboard part of the wing stalls, so there is no additional drag in normal flight.

 

Secondly, it reduces the wing downwash on the tailplane - which prevents the elevator from being able to pull the aeroplane any further into the stall.

 

Thirdly, it produces a strong "centring" effect on the vertical tail, which acts to prevent the aircraft from yawing.

 

Whether these effects will be sufficiently strong to overcome the effect of full power or full crossed controls, will vary from one aircraft to another. You just have to suck it and see.

 

If the setup works on your aircraft as well as it did on the Seeker and the Aeronca Champ, your aircraft will now be extremely reluctant to enter a spin. However, that is no guarantee that, if you do manage to get it into one, it will be recoverable. This is REAL test-pilot stuff, i.e. hazardous. Spin testing is a separate issue.

 

 

 

[Old Koreelah]

OK, do your VGs run all the way in to the wing root? If they do, I may be able to suggest some things that may help.

Pretty much so Dafydd. They stop about 1.5 bum-widths from the fuselage. Thus allows me to sit on the wing while tightening up the spar bolts when I assemble the aircraft. I also gave that inner section a slightly different leading-edge profile: the separation point is about 20mm higher. I had hoped this would generate enough buffeting to the elevator to warn of stall, but no luck. At normal CoG she just mushes thru the air losing about 700 fpm and won't stall.

When I move all luggage to the rear bay and fill the wing tanks (which are behind the spar) she mushes and finally lets go with a rapid wing drop.

 

As you know, inboard of the bent-up bits the Jodel wing has no dihedral. Wool tufting showed a pretty uniform reversal of airflow across the whole inner wing. I haven't got around to repeating that experiment since I changed the profile and tapered the inner wing.

 

I am interested in your suggestions.

 

 

[Dafydd Llewellyn]

On a high-wing aircraft such as the Seeker or the Aeronca, the effect of the fences and stall strips is to stall the whole centre part of the wing simultaneously - i.e. the stall seems to spread across the top of the fuselage as though the wing were continuous. You may find things a bit different on the low-wing setup on the Jodel; it may be necessary to position the stall strips slightly differently one side to the other, to get it to stall simultaneously; and power may alter that a bit. The effect on the Seeker is that you cannot get it to drop a wing, no matter what you do short of performing aerobatics in a non-aerobatic aeroplane. With the full-span VGs, it dropped a wing with extreme violence, such that I lost 600 feet in the recovery, despite being fully alert for the wing drop. With the fences & stall strips, and the VGs removed inboard of the fences, it became a pussycat; the change was astounding. The Aeronca was much the same, I understand, though I have not flown it myself. Ian McPhee rode in it for the testing, about eight years ago, he can tell you about it.

 

Whether you will get a perceptible buffet is determined by the relative position of the wake of the stalled wing centre-section and the tailplane, and that's something the designer either got right or not; in any case, it's almost impossible to get it right for all flap positions, if the aircraft has flaps. The Jodel tailplane being where it is, it has some unshielded rudder in a spin, and that takes precedence over locating it to get pre-stall buffet. Whether that was by accident or design, I cannot say; but the overall aerodynamic design is very clever, so it was probably no accident.

 

Note that the mechanism is NOT the same as a limited elevator power that prevents the wing from ever reaching its stalling angle of attack; that form of limitation is strongly affected by CG position. Instead, what I have described will work even if the elevator does have the power to take the wing to its stall A of A, because it uses the initial stall to inhibit any further elevator power. So the situation is not one of the aircraft "mushing" and not being able to use the full lifting capability of the wing; however the further aft the CG, the more powerful the elevator becomes - so you must set the stall strips for the aft limit situation.

 

In the case of the Seeker, it was possible to set the CG range such that the elevator had just sufficient authority to stall the wing at the forward CG - and the effect of the full airflow kit was just sufficient to still control the situation at the full aft CG; and having thus set the CG limits, the rest of the handling and stability testing, and spin testing, was completed at those limits. You do not have that freedom except in a basic aircraft certification situation, so set it up for the aft CG condition and put up with a bit of mushing at forward CG if necessary; you will be getting the best performance the aircraft is capable of delivering, consistent with safety. In effect, the setup I have described gives the wing a two-stage stall; and provided you set it up so you cannot reach the second stage stall, it will not drop a wing. To get this control over a large CG range seems to need a three-stage stall setup, which is a bit more complicated; this is what the Seeker has.

 

Stall/spin accidents are entirely the result of inadequate control of the lateral spread of stall along the wing. The conventional approaches of planform and washout help, but are not sufficiently powerful to prevent assymetric stalling, leading to a spin entry, with crossed controls. The correct use of VGs and fences and stall strips is a much more effective means of control, from what I have seen of it so far. It's so effective that one would not want it on a trainer, because it would make proper training impossible.

 

 

[Head in the clouds]

I can't guarantee this will work on any aeroplane ...

 

On a high-wing aircraft such as the Seeker or the Aeronca ...

Your last couple of posts have been very interesting Dafydd, thank you.

 

I've often wondered - is the Seeker a particularly 'problem child'? And if so, why, given its relatively conventional high-wing layout? Do you think its being a pusher has a detrimental effect in general?

 

I realise that extra effort has gone into making the low speed handling of this aircraft very docile because of its primary mission requirements being low level ground-related airwork but it certainly carries more 'fixes' than just about any plane except some of the Migs perhaps ...

 

I remember chatting with Bill Whitney (designer of the Seeker, for those who might not know) about the Seeker's handling and IIRC he was quite disappointed about the stall characteristics in the early stages. He told me that for an untwisted parallel planform wing it would perform as expected 90% of the time i.e. predictable stall propagating from inboard, but every now and then it would quite unexpectedly demonstrate a vicious wing drop and tufting revealed that for no apparent reason the stall would occasionally break midspan or further outboard. He said, again IIRC, that he didn't have a simple explanation and so 3* of washout was applied and "that fixed the problem, it flew very well after that".

 

So, I've been curious why further work was done on it. Was there still a problem, or was it just a matter of making it even better? Did you do all that development work with the VGs, fences, stall strips etc? And - what airfoil does it use? Is/was the airfoil choice a contributory factor in the initial handling problems?

 

At some stage, from about mid-span it also grew a drooped leading edge with associated dogtooth. What happened with that? Was it abandoned, or just fitted on certain models? How effective was/is it, and what disadvantages, if any? Were slats ever considered, or tried?

 

And - why the winglets on the HS? And the bulge at the top of the fin?

 

Not asking you to violate any commercial-in-confidence, of course - just interested because it must be one of the most developed lighties in recent times.

 

.

 

 

[Dafydd Llewellyn]

Yes, I did all the work for the airflow kit - fences, VGs, stall strips etc.; the reason for that was that it was originally certificated to FAR 23 at amendment 34. Due to the passage of time, in order to get an FAA TC for it, to allow it to be marketed in the U.S.A., it needed - in the days before Australia had a bilateral airworthiness agreement with the FAA that really WAS bilateral - to comply with FAR 23 at Amendment 42, which brought in the requirement for dynamic seat testing - the aeronautical equivalent to the barrier crash test for cars.

 

That DID raise a problem at the time, because energy-absorbing seats need the whole cockpit designed to cater for them, and of course that wasn't on the agenda when Bill Whitney was involved. However there was, at the time, a loophole for a fixed undercarriage aircraft if its stall speed was less than 49 kts CAS. The Seeker as it was then stalled at 52 KCAS, so Seabird decided to try VGs to see if they could squeeze the extra three knots by their use.

 

No, I wouldn't say the Seeker was a "problem child", apart from one minor issue that is common to all pusher aircraft having a single engine and a single vertical tail - e.g. the Sea Bee, the Trident Trigull etc - which is that in a power-off dive to the design diving speed - 111% of Vne - the windmilling propeller blankets the vertical tail; so what happens when you take your feet off the rudder pedals at that speed, is that the aeroplane gently yaws one way or the other, until the vertical tail comes out from behind the propeller. It's neither dangerous nor even exciting, but it's a technical non-compliance, and the finlets, & ventral fin were added to try to fix that. It finally took all those plus the geared vane under the fuselage, to tick that particular box. As a by-product of all that vertical tail area, the Seeker won't spin; it merely spirals.

 

The high-wing pusher layout had two very good handling features - the Seeker has zero trim change with flap deployment, and also zero trim change with power.

 

The wing airfoil is an NACA 64 series - I seem to recall, 642A215 - which was selected by Bill Whitney for reasons only Bill could tell you, but a low pitching moment was one of them, in order to reduce the download on the slender rear fuselage. It's an airfoil that stalls progressively from the trailing edge, which all the textbooks will tell you is a Good Thing; but what they do not mention, is that it makes the ailerons completely ineffective by the time the aircraft has slowed to about five knots above stall speed. In its original configuration, without the airflow kit, but with the washout, the situation with the ailerons was disguised, because what seems to happen - and this has been the case with the Skyfox and the Jabiru LSA 55 also, is that a "bubble" of separation forms in the centre of the wing; and if one wing starts to fall a bit, the resultant sideslip blows the bubble towards the opposite side, so the thing tends to recover - most of the time. The pilot reports that the lateral control is sloppy - but in reality, it is non-existent, apart from the secondary effect of rudder acting on the separation bubble. That was the reason for the somewhat disappointing (but not out-of-the-ordinary) stall handling with which the Seeker was originally certificated. I'd never have discovered all this if Seabird hadn't asked me to see what could be done to lower the stall speed, by the use of the VGs; I'd previously discovered the aileron issue, by a test with enormous stall strips that covered the whole wing leading edge for a distance equal to the tailplane span; that removed the "drifting stall bubble" and revealed that the ailerons were actually reversed by 5 knots above stall speed. However, it wasn't a solution that helped certification at the time, so merely remained as an item of knowledge - until Seabird decided to look into the possibilities of VGs.

 

The outboard leading-edge cuffs were originally tried on the Seeker's predecessor, the Sentinel, in an attempt to compensate for the absence of washout; they went into the rafters when the wings were re-built with washout; but I had them resurrected in the course of the testing of the airflow kit, in order to give the third-stage control of the lateral spread of stall, which they did very well - so they are now an integral part of the airflow kit, and all Seekers were retrofitted with the airflow kit, because whilst it did not quite provide the desired reduction in stall speed, it radically improved the stall handling, to what I suspect may be the most docile single-engine aircraft in existence. The Seeker is probably the most relaxing aircraft to fly I have ever experienced.

 

The bulge on the top of the fin is there to house the rudder mass-balance arm.

 

Overall, the pusher layout has its own set of compromises, which are a bit different to those of a tractor layout. They mainly affect the propeller - firstly, the propeller diameter is restricted, which limits its efficiency at low speed. Secondly, they tend to be noisy, because the propeller works in dirty air - so it's more difficult to achieve noise certification than with a tractor layout. Thirdly, the introduction into FAR 23 of a requirement for pusher propellers to be able to withstand chunks of ice falling into them from the airframe in front, poses a major obstacle to IFR certification. On the good side, you do not get propeller ground strike problems; and if some ham-handed twit manages to stand the aircraft on its nose, all you get is some grass-stains on the chin (I saw this happen with the prototype Seeker, with an overseas General in the pilot seat - the ground crew simply hauled the tail back down and told him to keep the stick back, and sent him on his way.)