Wing drop at the stall
Moderator
Thread Starter
Wing drop at the stall
I have a problem, and would be grateful if anybody can suggest a solution.
I'm in the middle of a certification test programme on a largely benign lightweight 2-seat, high wing, tractor monoplane with a mid-tail. Fairly torquey engine, which requires about 2/3 right rudder on t/o and a couple of pounds continuous rudder input in the cruise. 10° sweep, no dihedral to speak off, half span flaps.
At idle, the stall is marked when you run out of nose-up pitch authority. As we increase throttle however it shows an increasing tendency to drop a wing - reaching about 60° right wing drop at 60% MCP. It also rolls right about 60° from a 30° left turning stall with PLF and 30° from a 30° right turning stall with PLF. This all happens around min.weight, near MTOW it's just within certification limits.
Standard fixes for this are obvious: -
- droop the flaps
- reflex the ailerons
- increase washout.
All of which add up to roughly the same thing. But with 21° (take-off) flap, the stall characteristics are still only slightly more benign and still nowhere near the certification limit of 20°. To add to the fun, the aircraft is right on the certification limit for Vso, so although technically I can do so quite easily, increasing the washout is likely to increase Vso and stop the thing being certificated at-all.
I've considered Toblerone stips, which should help the lack of natural stall warning, but don't know if it's likely to help the wing drop much.
I've considered turbulators, but it's not a laminar flow wing, so I can't see that they're going to do much - but I could be wrong. Anybody know better?
Would defender style wing fences help?, I know they did on the Islander which is similarly marginal but are a bit of a ****** to fit.
I'm stumped. Help. Has anybody got any solutions that have worked for them. Sadly, I'm in charge, so haven't anybody else to dump the problem on.
G
I'm in the middle of a certification test programme on a largely benign lightweight 2-seat, high wing, tractor monoplane with a mid-tail. Fairly torquey engine, which requires about 2/3 right rudder on t/o and a couple of pounds continuous rudder input in the cruise. 10° sweep, no dihedral to speak off, half span flaps.
At idle, the stall is marked when you run out of nose-up pitch authority. As we increase throttle however it shows an increasing tendency to drop a wing - reaching about 60° right wing drop at 60% MCP. It also rolls right about 60° from a 30° left turning stall with PLF and 30° from a 30° right turning stall with PLF. This all happens around min.weight, near MTOW it's just within certification limits.
Standard fixes for this are obvious: -
- droop the flaps
- reflex the ailerons
- increase washout.
All of which add up to roughly the same thing. But with 21° (take-off) flap, the stall characteristics are still only slightly more benign and still nowhere near the certification limit of 20°. To add to the fun, the aircraft is right on the certification limit for Vso, so although technically I can do so quite easily, increasing the washout is likely to increase Vso and stop the thing being certificated at-all.
I've considered Toblerone stips, which should help the lack of natural stall warning, but don't know if it's likely to help the wing drop much.
I've considered turbulators, but it's not a laminar flow wing, so I can't see that they're going to do much - but I could be wrong. Anybody know better?
Would defender style wing fences help?, I know they did on the Islander which is similarly marginal but are a bit of a ****** to fit.
I'm stumped. Help. Has anybody got any solutions that have worked for them. Sadly, I'm in charge, so haven't anybody else to dump the problem on.
G
Guest
Posts: n/a
Not sure if the following is of any help, but I thought I'd contribute since no-one else has. I'm afraid it doesn't come from any certification flight test experience - any flight test experience I have is experimental in nature and on rotary-wing types.
Sharp stalls usually start at the leading edge. Gradual stalls start at the trailing edge. Thick blunt-nosed sections are generally good for trailing edge separations (ie gradual stall characteristics).
Wing-drop with power on I'd surmise happens because of propeller wake-induced changes to the local angle of attack on the wing, higher on one side than the other. Given the nature of the propeller wake, this is probably confined to the inboard sections, so that's perhaps where a solution should be focussed, rather than towards the tips - I'm not sure though.
Lower Vso could come from a higher Clmax. Drooping the leading edge can give this, but it can do nothing to improve sudden onset of stall, just give you a higher Clmax (I think this is effectively what you see on the underside of the Bulldog wing, there's a crease near the leading edge which looks as if it contributes a gentle 'droop' shape).
On the PA38, versions equipped with four 'flow strips' as opposed to two (they're toblerone-shaped, about 1cm high and are about 15-20cm long) have a 3 kt higher stall speed, and in my experience will still tend to drop a wing power-on, although not by 60 deg.
A good book (although quite difficult to get) is Fluid-Dynamic Lift by S. F. Hoerner. It was privately published in 1975, there's no ISBN number on the copy I have, but Ch. 4 deals with stall characteristics of airfoil sections.
Apologies in advance if this is of no help.
Sharp stalls usually start at the leading edge. Gradual stalls start at the trailing edge. Thick blunt-nosed sections are generally good for trailing edge separations (ie gradual stall characteristics).
Wing-drop with power on I'd surmise happens because of propeller wake-induced changes to the local angle of attack on the wing, higher on one side than the other. Given the nature of the propeller wake, this is probably confined to the inboard sections, so that's perhaps where a solution should be focussed, rather than towards the tips - I'm not sure though.
Lower Vso could come from a higher Clmax. Drooping the leading edge can give this, but it can do nothing to improve sudden onset of stall, just give you a higher Clmax (I think this is effectively what you see on the underside of the Bulldog wing, there's a crease near the leading edge which looks as if it contributes a gentle 'droop' shape).
On the PA38, versions equipped with four 'flow strips' as opposed to two (they're toblerone-shaped, about 1cm high and are about 15-20cm long) have a 3 kt higher stall speed, and in my experience will still tend to drop a wing power-on, although not by 60 deg.
A good book (although quite difficult to get) is Fluid-Dynamic Lift by S. F. Hoerner. It was privately published in 1975, there's no ISBN number on the copy I have, but Ch. 4 deals with stall characteristics of airfoil sections.
Apologies in advance if this is of no help.
Moderator
.. and if it all remains intractable .. why not tuft the wings and video the stall flows .. at least then you know where the problem originates and can play in a more directed manner .... ?
Join Date: Jun 2000
Location: Geriatrica, UK
Posts: 1,003
Likes: 0
Received 0 Likes
on
0 Posts
Agree with TM; before you can decide the fix, you ought to be sure what the problem is.
What sideslip detection mechanism are you using? Inertial (slip ball) or vane? The symptoms are typical of the category and a contributory factor can be the helical flow in the turn aggravated at high power. Ball-in-the-middle may not equal zero beta.
What sideslip detection mechanism are you using? Inertial (slip ball) or vane? The symptoms are typical of the category and a contributory factor can be the helical flow in the turn aggravated at high power. Ball-in-the-middle may not equal zero beta.
Moderator
Thread Starter
Gosh a treasury of advice.
I have to discount a few things. The LE is very rounded, and unsurprisingly the idle stall is very benign. So, I don't think that it's wing shape.
It's simply flown with a slip-ball, so empirical beta estimation only. But, it's got quite a torqey engine which requires a constant right rudder input (i.e. the thing keeps trying to turn left). I allowed myself slightly too little right rudder, so that it tended to turn left - and it still dropped the right wing at the powered stall.
Thanks for the confirmation that Toblerone strips aren't likely to work - I certainly couldn't stand anything like a 3kn increase in Vso.
I suspect that the problem may well as IM suggested be inboard LE rather than the tip, particularly since it only seems very slightly alleviated by the selection of flaps. I had a chat to a colleague at a certain manufacturer of high winged light twins who told me that they had the same problem, but it was solved by fitting a wing fence twixt aileron and flap to contain any inboard section stall. I'm going to see if this is possible to try out.
JT is quite right about tufting, it's just a bit of a headache when against a short timescale on a high wing - because the problem is almost certainly going to be on the wing upper surface not the lower. So I've talked to the "man who makes things" and think I am going to (carefully !) try a day or two of empirical fix-fly-fix, and if that fails, do the job properly as JT suggests.
I shall report back, in the meantime, thanks a lot for the advice chaps.
G
I have to discount a few things. The LE is very rounded, and unsurprisingly the idle stall is very benign. So, I don't think that it's wing shape.
It's simply flown with a slip-ball, so empirical beta estimation only. But, it's got quite a torqey engine which requires a constant right rudder input (i.e. the thing keeps trying to turn left). I allowed myself slightly too little right rudder, so that it tended to turn left - and it still dropped the right wing at the powered stall.
Thanks for the confirmation that Toblerone strips aren't likely to work - I certainly couldn't stand anything like a 3kn increase in Vso.
I suspect that the problem may well as IM suggested be inboard LE rather than the tip, particularly since it only seems very slightly alleviated by the selection of flaps. I had a chat to a colleague at a certain manufacturer of high winged light twins who told me that they had the same problem, but it was solved by fitting a wing fence twixt aileron and flap to contain any inboard section stall. I'm going to see if this is possible to try out.
JT is quite right about tufting, it's just a bit of a headache when against a short timescale on a high wing - because the problem is almost certainly going to be on the wing upper surface not the lower. So I've talked to the "man who makes things" and think I am going to (carefully !) try a day or two of empirical fix-fly-fix, and if that fails, do the job properly as JT suggests.
I shall report back, in the meantime, thanks a lot for the advice chaps.
G
Moderator
Over the years we've down quite a bit of video data recording... the current pot of toys available includes some very small units which would suit fin installation of multiple cameras appropriately directed ... lenses are just about tailored to the requirement.
Main problem is low resolution so one often needs to go for multiple cameras with inconvenient focal length lenses. Overall a very easy, cheap and nasty way around the problem ....
Main problem is low resolution so one often needs to go for multiple cameras with inconvenient focal length lenses. Overall a very easy, cheap and nasty way around the problem ....
Join Date: Oct 2000
Posts: 53
Likes: 0
Received 0 Likes
on
0 Posts
Just keep in mind that tufts probably will work as mini VG's and probably improve the overall stall characteristics and lower the stall speed. They however should give you a good indication of how the stall breaks. I suggest you videotape both wings simultaneously to see how and when the flow separates.
If a rudder force is required in cruise it sounds like you have a fundamental design problem!
If the wings are adjustable you could also contemplate slightly different incidence on one wing relative to the other. This might also help alleviate currently required rudder force in cruise.
The key to stall recoveries is to ensure flow over the ailerons. When the stall breaks make an effort to stay off the rudder since rudder inputs might aggravate the recovery.
I would start with a full wing fence approximately 2-3 inches high in combination with either VG's ahead of the ailerons and/or boundary layer energizers on the leading edge. Just remember that the BLE edges must be sharp to shed a vortex. Use silver speed tape and try multiple configurations. When something works then start reducing or removing items. The key is to document what you do for later reference.
If your problem is positively identified as asymmetric only one wing might need treatment. Keep an open mind!
Small inboard stall strips might trigger the stalls evenly on both sides, but it does not sound you have much margin speed wise since strips likely will increase your stall speed. On the other hand a fence and VG's might lower your stall speed. The goal however must be to have the minimum amount of junk on the wings. Mounting the engine on a slightly offsett angle could also be considered if you already haven't done so.
Good luck!
If a rudder force is required in cruise it sounds like you have a fundamental design problem!
If the wings are adjustable you could also contemplate slightly different incidence on one wing relative to the other. This might also help alleviate currently required rudder force in cruise.
The key to stall recoveries is to ensure flow over the ailerons. When the stall breaks make an effort to stay off the rudder since rudder inputs might aggravate the recovery.
I would start with a full wing fence approximately 2-3 inches high in combination with either VG's ahead of the ailerons and/or boundary layer energizers on the leading edge. Just remember that the BLE edges must be sharp to shed a vortex. Use silver speed tape and try multiple configurations. When something works then start reducing or removing items. The key is to document what you do for later reference.
If your problem is positively identified as asymmetric only one wing might need treatment. Keep an open mind!
Small inboard stall strips might trigger the stalls evenly on both sides, but it does not sound you have much margin speed wise since strips likely will increase your stall speed. On the other hand a fence and VG's might lower your stall speed. The goal however must be to have the minimum amount of junk on the wings. Mounting the engine on a slightly offsett angle could also be considered if you already haven't done so.
Good luck!
Join Date: Jun 2000
Location: Geriatrica, UK
Posts: 1,003
Likes: 0
Received 0 Likes
on
0 Posts
Your beast is lightweight; if it's side-by-side, don't forget to take into account the very significant lateral CG shift between one- and two-crew. The change in AUM will probably be of the order of 15% between one and two crew.
If it's tandem... Well, I'm sure you know all this, Genghis.
Once had a serious blockage (mental) during a RW, ground observed, project with very inconsistent results between successive reciprocal runs along the RW during the same sortie. Turned out it was the enormous shift in lateral CG caused by the on-board FTE changing from side to side so that he could be on the same side as the ground observer. "Ball-in-the-middle" meant nothing to the static ports that were in different flow each time the FTE shifted. He was quite a normal sized chap really! Got him to sit in the middle and the trial continued...
If it's tandem... Well, I'm sure you know all this, Genghis.
Once had a serious blockage (mental) during a RW, ground observed, project with very inconsistent results between successive reciprocal runs along the RW during the same sortie. Turned out it was the enormous shift in lateral CG caused by the on-board FTE changing from side to side so that he could be on the same side as the ground observer. "Ball-in-the-middle" meant nothing to the static ports that were in different flow each time the FTE shifted. He was quite a normal sized chap really! Got him to sit in the middle and the trial continued...
Join Date: Oct 2000
Location: Wichita, KS, USA
Posts: 7
Likes: 0
Received 0 Likes
on
0 Posts
Should on-board camera installation become too problematic, perhaps photo chase would be helpful. If the speeds are as low as you imply, a helicopter could be used. In particular, the ability to look down seems to address your tuft visualization problem. There are obvious safety issues with this approach but with proper crewing it should work well.
Moderator
Thread Starter
Well, I'm delighted to say that in 1 working day and 2:15 flying hours, the problem has been solved. In the process, I've certainly learned something.
We started by re-visiting the wing dropping stall. It was discovered that the wing drop was greater with more nose-down pitch trim. Looking back, in previous tests, if keeping height and power the same, inevitably the aircraft was trimmed further forward to maintain level flight at a lighter weight. I confess that this wasn't an expected effect, it was spotted entirely by accident.
On the ground we examined the pitch trimmer, which was through a single tab on the port elevator. With nose-down trim, the trimmer was of-course more trailing edge up. This initially made no sense, since one would imagine that a trailing edge-up tab on the port side would cause the aircraft to roll to port at the stall, not the starboard we were getting.
Then we looked at the elevator mechanism, and took some measurements. The port and starboard elevators were each linked through a pushrod, arm and pivot pin to a single pushrod in the fuselage. Through bending in the pushrod and a small amount of freeplay at the pivot pin, there was a remarkable (and previously unnoticed) 7° of available freeplay between the two elevators.
We theorised that the trim tab on the port side could push against the freeplay at the stall and create up to 7° of differential elevator. This was likely to be more marked at higher power, since the tail is right in the propwash.
Some rapid redesign ensued, and several washers inserted in the elevator lever arm joint, reducing the freeplay between elevators to 2°.
We flew it again, and although the wing drop was still there, so long as stall recovery was rapid, it was down to about 15° - which is within certification limits.
So, I end my day with the problem solved, I am a little wiser, and we might actually get the beastie certified on schedule. The company is inevitably very happy, since two extra washers is an awful lot cheaper than a wing fence or vortex generators.
Many thanks everybody for your help.
G
We started by re-visiting the wing dropping stall. It was discovered that the wing drop was greater with more nose-down pitch trim. Looking back, in previous tests, if keeping height and power the same, inevitably the aircraft was trimmed further forward to maintain level flight at a lighter weight. I confess that this wasn't an expected effect, it was spotted entirely by accident.
On the ground we examined the pitch trimmer, which was through a single tab on the port elevator. With nose-down trim, the trimmer was of-course more trailing edge up. This initially made no sense, since one would imagine that a trailing edge-up tab on the port side would cause the aircraft to roll to port at the stall, not the starboard we were getting.
Then we looked at the elevator mechanism, and took some measurements. The port and starboard elevators were each linked through a pushrod, arm and pivot pin to a single pushrod in the fuselage. Through bending in the pushrod and a small amount of freeplay at the pivot pin, there was a remarkable (and previously unnoticed) 7° of available freeplay between the two elevators.
We theorised that the trim tab on the port side could push against the freeplay at the stall and create up to 7° of differential elevator. This was likely to be more marked at higher power, since the tail is right in the propwash.
Some rapid redesign ensued, and several washers inserted in the elevator lever arm joint, reducing the freeplay between elevators to 2°.
We flew it again, and although the wing drop was still there, so long as stall recovery was rapid, it was down to about 15° - which is within certification limits.
So, I end my day with the problem solved, I am a little wiser, and we might actually get the beastie certified on schedule. The company is inevitably very happy, since two extra washers is an awful lot cheaper than a wing fence or vortex generators.
Many thanks everybody for your help.
G
