Thread Tools
Sep 10, 2019, 12:41 AM
G_T
G_T
Registered User
Thread OP
Discussion

An idea... for tweaking thermal turns


I'm going to suggest an unorthodox mix to optimize turning a bit more. I've thought about this idea for a long time, but essentially I don't fly any more. So I can't test it myself. So I'm passing it on, untested.

To be able to get the most out of this, you'll have to already have some skill using both thumbs to control aileron and rudder in thermal turns. If you haven't yet acquired this skill, it is better to be working on developing the skill then messing with an advanced mix such as what I'm proposing here. You won't have the skill to dial in the mix, or to use it effectively. Sorry.

If you have weakness in air reading or basic flight skills, work on those as you'll get much more benefit than you will from any mix.

This mix requires at least a 4 servo wing. The more, the better.

This mix should not be used along with an aileron->rudder mix as they'll fight each other. If you absolutely have to have an aileron->rudder mix, you probably aren't ready to try a mix like what I'm proposing here.

The idea is simple enough but the programming might not be.

--- The Issue ---

When a plane is in a thermal turn, the whole plane sinks at the same rate. But the inner tip travels a much shorter distance than the outer tip. The inner tip is slowest; the outer tip fastest. Due to the same sink rate, the inner tip sees a higher angle of attack, whereas the outer tip sees a lower angle of attack. This assumes no sideslip!

That's not the same situation as flying in a straight line...

I've optimized my wing designs to handle turns pretty well, but perhaps we can tweak any wing to possibly do a bit better...

--- The Idea, a THERMAL TWEAK mix ---

*** I'm going to assume everyone reading at this point uses one of your thumbs to control the rudder, and has the skill to do so independently of the other thumb controlling the ailerons. In a thermal turn (any turn, really), the aileron thumb controls the bank angle, while the rudder thumb controls the sideslip choice (from zero to a tiny bit of sideslip, depending...). You've been warned! Twice! ***

I also assume you can and do fly the sticks smoothly during turns. If you are a TWITCH, better to work that out than mess with an advanced mix. The mix won't fix pilot issues.

OK, enough warnings.

This mix is basically rudder->darned near everything. What we are going to do is make a non-uniform camber adjustment to the wing so that (1) the wing camber is optimized a bit better for position relative to the center of the turn, and (2) so that we reduce or eliminate rudder induced yaw->roll coupling.

(1) The outer wing tip is flying fast - too fast. The inner wing tip is flying slow - too slow. If bank angle is to be kept constant, then the outer tip needs to operate at a fair bit lower lift coefficient than the inner tip. It already does, in a steady-state turn, due to the effect I briefly explained in The Issue above.

If a wing is flying too fast, generally it is best if it has reduced camber setting. It is usually better optimized. This also applies to part of a wing. So the outer tip should have a bit less camber than the average for the wing. That will reduce it's drag. As that is the fastest flying part of the plane, it has the greatest effect there. This in turn (pun intended) slightly reduces the required rudder to coordinate the turn.

If a wing is flying too slowly, generally it is best if it has increased camber setting. It is usually better optimized. This also applies to part of a wing. So the inner tip should have a bit more camber than the average for the wing. Possibly. Testing required because it will depend on YOUR particular wing's design. Safest is to leave it initially at the average camber setting. But don't fear that adding a touch more camber there is going to increase your odds of an inner tip stall. In this speed and size range, for most airfoils, the flow is going to separate at the front not the back, so the camber setting won't matter much.

Now take the inner flaperon (4 servo wing) and increase its camber slightly.

So going from inner tip to outer tip, we have [inner tip] = average up to average plus about half a tweak, [inner flap] = average plus a tweak, [outer flap] = average, [outer tip] = average minus a tweak. We have rudder towards the inner tip a little, coordinating the turn.

The wing has the same average camber. We've just changed the distribution.

The trailing edge wave should be a bit subtle. If it looks like a lopsided crow, you've gone way overboard!

Make this proportional to rudder throw. And make the rudder throw reduced a little compared to what you are used to.

--- The benefits ---

TBD, but what you might find:

(1) Reduced coupling between rudder adjustment during the thermal turn to control sideslip, and the bank angle effects this induces due to dihedral. That should improve your ability to read the air while thermalling since there is less axis coupling.

You can test this in straight line flight. Go to the new thermal mode. Yaw the plane a little with the rudder. You should observe quite a bit of reduction in roll compared to a normal setup. You might even be able to eliminate the roll and have the plane fly in a straight sideslip. I would not recommend going beyond that!

(2) Possibly reduced sink rate and better energy retention during the sustained coordinated turn. The second part you can test by feel, switching back and forth between the old and new modes spending several turns in each. Sink rate you can test in dead air with a stopwatch. Go into a "standard" turn, and hold it from a fixed launch height until you have to land. The longer the plane stays up, the more efficient the setting for that wing design and that turn rate and speed.

---

I'd suggest an initial mix rate on the order of one degree of plus and minus per five degrees of rudder offset. That's just a starting point, to have one! It isn't a magic ratio and I haven't done analysis to determine a better starting point.

Again, the THERMAL TWEAK MIX is just based off some completely untested thinking of mine. If it doesn't work out, and you are skilled enough to tell the difference, blame me!

Gerald Taylor
Sign up now
to remove ads between posts
Sep 10, 2019, 01:19 AM
Have Fun and Just Fly!
l shems's Avatar
Wouldn't an aileron max deflection distribution gain the same?

So less aileron when you go more to the wing tips?

Or, if you would give counter roll to prevent banking more after having set in the turn, more aileron when going to the wing tips?

What are you assuming as stable turn aileron deflections? In the direction of the bank angle, or counter bank angle?

I would imagine that for most effective continuous roll in terms of energy retention, you would need to have less aileron deflection at the outside of the wing.
For fixed bank angle in a turn, with high dihedral, you would have to maintain a little aileron to maintain the bank angle. Then this aileron distribution would do the same as your camber distribution.
If you would need to give roll opposite to the bank angle, you would need to have the opposite aileron deflection distribution.

In short, aren't you just creating an extra roll effect, but indépendant of roll direction, and controlled by rudder instead, with the objective of optimising energy retention?

Perhaps it should be on a separate stick then instead of rudder?
Sep 10, 2019, 06:42 PM
Registered User
R.M. Gellart's Avatar
Interesting Gerald, I like mind benders.

Marc
Sep 10, 2019, 09:37 PM
G_T
G_T
Registered User
Thread OP
Quote:
Originally Posted by l shems
Wouldn't an aileron max deflection distribution gain the same?

So less aileron when you go more to the wing tips?

Or, if you would give counter roll to prevent banking more after having set in the turn, more aileron when going to the wing tips?

What are you assuming as stable turn aileron deflections? In the direction of the bank angle, or counter bank angle?

I would imagine that for most effective continuous roll in terms of energy retention, you would need to have less aileron deflection at the outside of the wing.
For fixed bank angle in a turn, with high dihedral, you would have to maintain a little aileron to maintain the bank angle. Then this aileron distribution would do the same as your camber distribution.
If you would need to give roll opposite to the bank angle, you would need to have the opposite aileron deflection distribution.

In short, aren't you just creating an extra roll effect, but indépendant of roll direction, and controlled by rudder instead, with the objective of optimising energy retention?

Perhaps it should be on a separate stick then instead of rudder?
I think you may have misunderstood some of it. It probably wasn't my best posting! Pictures might have helped, but I don't have any prepaired.

It is a mix keyed on rudder because it is proportional to rudder deflection which relates to the radius of the circle and bank angle when a coordinated turn is performed.

It is independent of the usual aileron progression across the wing from aileron stick input. The final camber setting of each control surface of the wing is the sum of the current thermal camber setting (set by flight mode etc) plus the tweak of those surfaces plus or minus a little bit from this new mix, plus or minus whatever the aileron input would provide.

To make the mix work, the rudder is going to have to work as a stick switch (to select a Left or a Right turn mix) and then the mix itself is proportional to rudder displacement. Like I said, the mix isn't simple.

The simplest variation of this mix that would be worth trying would deflect only the inside turn's flap, and the outside turn's aileron.

The mix is counter to the roll tendency from rudder input.

Normal scenario: Push right rudder... the plane yaws nose to the right. The angle of attack of the left wing increases, and the right wing decreases, due to dihedral and the sideslip. So, the plane rolls to the right.

With this mix in place: Push right rudder... the plane yaws nose to the right. The angle of attack of the left wing increases, and the right wing decreases, due to dihedral and the sideslip. But, the left aileron goes up slightly. And the right flaperon goes down slightly. This provides a weak "aileron-ish" roll to the left, partially or completely countering the roll to the right which would otherwise be induced by the sideslip.

Unlike just adding a rudder->aileron mix (which one could certainly use), this version is intended to also improve performance a little in thermal turns.

Gerald
Sep 10, 2019, 09:52 PM
G_T
G_T
Registered User
Thread OP
Perhaps consider this visualization:

You are doing a good job of flying a normal left-banked turn in a thermal. You are keeping the turn coordinated nicely. You are therefore holding some left rudder, and you are possibly holding a touch of right aileron (depends on plane).

In a normal thermal setup, you'll have the trailing edge nearly evenly down over its full span, except for a slight bit of right roll superimposed due to holding a touch of right aileron.

Now take that exact setup, and pretend you added a tiny bit of down, perhaps 2mm, to the left flaperon. And pretend you added perhaps 1mm of up to the right aileron. We've squiggled the trailing edge.

We've also reduced the aileron roll input a little, so the aileron stick will have to be held over a little bit more to compensate for the new mix. Or, the aileron rates adjusted to keep approximately the original feel on the aileron stick. I'd just let the mixes counter each other a little. So aileron will seem a little less sensitive as long as you are coordinating the turns. You'll probably fly smoother that way. The aileron stick will still dominate- it causes much greater motion of the control surfaces than this new mix does. The new mix really is just an optimization tweak.

Does this help?

Gerald
Sep 10, 2019, 11:05 PM
In F3J size does matter!
roydor's Avatar
In a full house models I have a flap to aileron mix which means that when turning tight and holding a bit of opposite aileron you get more deflection on the inner aileron due to the aileron input, more deflection on the inner flap due to the mix, slightly less deflection on the outer flap due to the mix and a lot less deflection on the outer aileron.
Effectively you get the distribution you were talking about from inner aileron to to outer aileron. And the more tightly you turn the more opposite aileron you need and more “twist” goes into the wing.
Sep 11, 2019, 12:39 AM
launch low, fly high
Sounds like a mix to generate a change in differential based on the absolute value of rudder deflection. That is, a left rudder deflection will reduce aileron differential just as the same as a right rudder deflection.

There are some interesting challenges on this, particularly for the very light weight F5J models. The inboard tip panel will be at very low Re. The amount of out aileron to get roll trim increases as the wing loading goes down.

The assumption that coordinated flight is the best answer gets far more uncertain at the very low weights that some F5J aircraft are flying at. Also, it is difficult to guess just how much rudder one needs to use to get coordinated flight. For some aircraft, the answer is "all of it". And that may not be enough due to the rather high yaw rate.

Assume a minimum weight 4 m span F5J plane flying with a CL of 0.9 and with an elliptical wing planform (assumed to make the numbers easier to calculate). In a fully coordinated 30 deg bank turn, the plane has to have a rather large CL differential between the inboard and outboard wing panel. The attached plot shows this. The green line is the local CL with respect to %span for straight flight (the elliptical planform results in constant local CL), the blue line is the required local CL if the spanwise lift distribution remained the same as the straight line flight. The tan or orange line is one possible set of aileron/flap deflections that result in the same total lift as the blue curve and also has zero rolling moment. The second plot is the lift distribution that comes from the local CL distribution on the tan and blue curves.

There is an issue with having a very large local CL on the inboard wing tip, especially as it is at a very low local Re (at or below 30k). Because of that low Re, the optimum can be with increasing the center loading on the wing and reducing the loading on both tips. At the very light weights, having 6 servos instead of 4 can make a material difference in the circling efficiency as one can better tailor the local CL in turning flight. There is also an interesting trade where one increases induced drag via allowing the nose to point outwards in a circle (use less in-rudder) to get the velocity delta between the tips lower to reduce the magnitude of the local CL variance. I think most of the top F5J pilots have already found this solution via testing. Better to reduce the effective span and have some cross-flow than to have the very large out aileron with a reduced maximum total CL. The multiple dihedral breaks also really helps in this situation.
Sep 11, 2019, 02:31 AM
Registered User
Tuomo's Avatar
Quote:
Originally Posted by roydor
In a full house models I have a flap to aileron mix which means that when turning tight and holding a bit of opposite aileron you get more deflection on the inner aileron due to the aileron input, more deflection on the inner flap due to the mix, slightly less deflection on the outer flap due to the mix and a lot less deflection on the outer aileron.
Effectively you get the distribution you were talking about from inner aileron to to outer aileron. And the more tightly you turn the more opposite aileron you need and more “twist” goes into the wing.
You mean aileron to flap mix?

As I fly Explorer and mostly with 9 degree joiners, I do not really know how much I aileron (and to what direction!) I use in fully developed and stable thermal turn. It might be very close to neutral aileron with some rudder to the turn - must pay attention to this next time when flying. With less dihedral things get a lot different.
Last edited by Tuomo; Sep 11, 2019 at 03:24 AM.
Sep 11, 2019, 11:40 AM
In F3J size does matter!
roydor's Avatar
Yes, the flaps are mixed with the aileron stick. As you move the ailerons the flaps move just a tiny bit as well. I find this to be one of the more important mixes of a modern F5J. In thermal mode the flaps move very little but still needs to be set up properly to reduce overall drag and improve handling.
Sep 11, 2019, 04:40 PM
Have Fun and Just Fly!
l shems's Avatar
Quote:
Originally Posted by roydor
Yes, the flaps are mixed with the aileron stick. As you move the ailerons the flaps move just a tiny bit as well. I find this to be one of the more important mixes of a modern F5J. In thermal mode the flaps move very little but still needs to be set up properly to reduce overall drag and improve handling.
Which is exactly what I meant with aileron deflection distribution from the inside to the outside if the wing. I see the wing as one continuous object. All control surfaces act as ailerons and as camber adjusters.
So is this just an 'Ailerons to flaps' mix, controlled by rudder. Ailerons being the outer and flaps being the inner control surfaces?
Sep 11, 2019, 10:35 PM
G_T
G_T
Registered User
Thread OP
Joe,

I completely agree that flying the classic definition of "coordinated" may not be ideal for these planes as currently designed and flown, though I consider the light wing loading to have little to nothing to do with it. I consider relative turn radius to have everything to do with it.

Yawing a plane sporting dihedral nose-out in a turn will increase the angle of attack of the inner now forward wing tip, and slightly increase the airspeed of that tip due to the slightly larger circle traversed. The latter is a benefit; the former, not. A boost in Reynolds number tends to allow for a slightly greater Alpha and Cl range being supported before unacceptible drag is created. The inner tip is of course the lowest Reynolds number part of the plane!

The outer tip sees an angle of attack which is even less optimal than usual, due to dihedral and yaw. As the fastest part of the plane, a further increase in Cd there could potentially have greater impact in overall plane drag.

It is more likely that the deliberate yaw apparently providing improvement is substantially due to the IMHO overlength tailbooms that are currently popular. Yawing the nose out reduces the fuselage drag by reducing the FUSELAGE's sideslip, and lowers the angle of attack induced into the vertical fin. A fuselage flying even moderately sideways has rather poor aerodynamics! At least, compared to any other part of the plane! I look at pronounced outside yaw in a turn as a bit of a bandaid - though considering what is being flown and the turn radius relative to span and tailboom length that is often attempted, a probably legit bandaid. But frankly I'd rather see plane designs altered from where they are now to be more accomodating of how people seem to want to fly the planes. If a plane is supposed to be spending a lot of its time in really tight turns, it should be designed accordingly! I look around and see epic fail.

All IMHO, of course!

If I really wanted to make it easier on just the wing to make a very tight turn, I'd yaw seriously nose-in, and camber up the inner wing quite a bit, while taking camber out of the outer wing tip. Then the inner tip could actually handle a tighter circle due to the nose down aft loaded profile it presents to the freestream in this flight geometry. And I'd still get the inner tip airspeed and Reynolds numbers boost. But fuselage drag would probably be on the order of all the rest of the plane! So sink rate would be high.

IF one wanted to get freaky, one could design separate tip foils for the inner tip and the outer tip, and thermal turn in only one direction to benefit from the optimization. Trimming, of course, would be an interesting challange!

Or as a much more conventional alternative, bias the dihedral more towards the center of the wing and have reduced tip panel angle - as I did with the SynerJ reference wing design. It's a minor tweak. Then the effects of minor wing yaw are reduced in tight turns.

I've had way too little sleep in too long due to work, so I'm out of here.

Gerald
Sep 11, 2019, 11:02 PM
Registered User
exf3bguy's Avatar
What I have discovered while doing tight Thermal turns with my Supra is that once the bank angle gets established I can come off the ailerons and maintain the circles with rudder but find myself going opposite aileron a fair amount of the time. That in affect increases the camber on the inboard wing while reducing it on the outboard wing. Unless I'm misunderstanding what you were describing, it seems to be the same thing.
Sep 12, 2019, 01:22 AM
Registered User
Tuomo's Avatar
Quote:
Originally Posted by G_T
All IMHO, of course!
I have a feeling that this topic will heat up when summer heat fades away. Something to think about in the cold darkness of January

My personal IMHO is that sensible camber setting combined with air speed control and good most crucial aspect of thermal turns. Also flying reasonably coordinated turning rarely hurts. Looking at people flying in competitions, it is easy to see that these are the aspects of flying that suffer most when ground gets closer and thermal gets weaker. Or the other way round: when turning with good energy and gaining altitude, the theoretical problems or flying are not very relevant.
Sep 12, 2019, 02:50 AM
launch low, fly high
Quote:
Originally Posted by G_T
Joe,

I completely agree that flying the classic definition of "coordinated" may not be ideal for these planes as currently designed and flown, though I consider the light wing loading to have little to nothing to do with it. I consider relative turn radius to have everything to do with it.
...

Gerald
Turn radius is rather strongly associated with wing loading. That is unless you decide that one will circle at 10 deg bank when light and 45 deg bank when heavy. Then, the turn radius may be similar... For a fixed bank angle and lift coefficient, the turn radius is directly associated with wing loading. Increase the wing loading by 10% and the turn radius increases by 10%.
Sep 12, 2019, 05:32 AM
http://f5j.eu
jpecar's Avatar
Quote:
Originally Posted by G_T
and camber up the inner wing quite a bit, while taking camber out of the outer wing tip.
I actually have something along these lines in my prototyping pipeline, hope to have it in the air still in this year. I do 99% of my circling to the left only so it's fine for me to be a bit asymmetric


Quick Reply
Message:

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion which way birds turn into thermal? phil alvirez Thermal 52 Jul 18, 2016 06:47 AM
Cool Champing, an interesting idea for a holiday. Punkie Life, The Universe, and Politics 0 Jul 14, 2016 11:44 AM
Discussion Idea for an airblower george84 Power Systems 15 Feb 28, 2016 10:01 PM
Discussion Thermaling , direction of turns N vs S hemispheres soarbuff Thermal 14 Dec 23, 2015 12:37 PM