Thread Tools
Apr 23, 2014, 09:21 AM
Detail Freak
target's Avatar
He bought a boat. I thought he was smarter than that....

Sign up now
to remove ads between posts
Apr 24, 2014, 02:08 AM
Registered User
Hello Mark,

do I understand your approximation correctly? In terms of the Supra and CL=0,6, the mesaured camber deflection at the TE (in length scale) should be about equal compared to the measured elevator deflection at TE (same length scale).

Originally Posted by markdrela
I use Xfoil polars and AVL to determine the best snap-flap mix as follows:

1) Using Xfoil polars with a range of flap deflection, determine the best camber deflection for a typical CL in a fast F3B or F3F turn. Typical values might be CL=0.4-0.6 . The best camber deflection is the one which gives minimum Cd at that CL. The Reynolds number must also be set appropriately for the high-speed case. For this you need the Type-1 polars, not the Type-2 polars typically used to compare airfoils.

2) In AVL, set the camber to the high-speed case (i.e. max reflex), set CL for the fast level-flight case using the C1 menu. This might be as low as CL=0.015 in F3B Speed. Also constrain elevator to seek zero pitching moment. Calculate the trimmed elevator deflection. make sure the reference X location (CG location) corresponds to the actual glider.

3) In AVL, set the optimum fast-turn camber found in step 1). In the C2 menu, set the speed equal to the high-speed case, and set the expected turn CL assumed in step 1).
Again calculate the trimmed elevator deflection.

The ratio of camber difference to elevator difference between steps 2 and 3 is the optimum camber to elevator (snap-flap) ratio. Below is the numerical example for a Supra flown in F3B Speed. The ratio is nearly 1:1, regardless of how tight the turn is. This ratio will depend somewhat on the CG position. The reason is that an aft CG requires less elevator to give a certain turn radius, and hence requires more snap-flap mix to give the same camber deflection as a forward CG.

Supra F3B trimmed-pitch elevator throws
m = 2.25 kg
V = 60 m/s
static margin = 10% MAC

CL Rturn G's Camber Alpha Elevator
----- ------ ---- ------ ----- --------
0.015 inf 1.0 -4.0 -0.98 0.59
0.400 13.8m 26.6 +1.0 -0.12 -3.84
0.600 9.2m 40.0 +3.0 +0.66 -6.16

CL=0.4: dCamber/dElevator = 5.0/4.43 = 1.13
CL=0.6: dCamber/dElevator = 7.0/6.75 = 1.04
Apr 24, 2014, 09:37 AM
launch low, fly high

The numbers shown from the AVL model should be degrees.
Apr 24, 2014, 11:55 AM
Registered User
thank you for clarification. Perhaps you might give us your recommendation wrt dead band of elevator to flap mixing. Some do it, other not, or do it even opposite. As I did you understand in former discussions, you would rather recommend high mixing level around elevator neutral position (pulling means I want more lift at once) and a degressive function towards larger deflections. Am I right?
Apr 24, 2014, 07:07 PM
launch low, fly high

Now that we are getting into more detail, time to differentiate the answer as based on intended usage... F3B and F3F optimize slightly differently than F3J and F3K (excepting the launch2 phase for F3J). The primary intent for F3B and F3J on snapflap is to put the trailing edge in the right location to minimize drag for turns that typically have a relatively high g loading. Neglecting the launch aspect, F3J and F3K has a primary intent for snapflap to adjust the drag bucket for short term elevator commands that tend to be above the optimal lift coefficient.

To get an idea of the linearity or non-linearity desired for the "elev to camber mix" (whoever named this mix "snapflap" needs to be dragged behind the woodshed!), plot the airfoil polars for various trailing edge deflections. For F3F/B, plot them at typical distance and speed flight Re. For F3J/K, plot them at typical thermal flight Re. One can correlate the optimal camber to lift coefficient, then convert that to elevator deflection via the methods outlined by Dr. Drela (well, kinda... the paradigm is there, just in reverse).

The requirements for higher speed flight turning are a bit different than the requirement for thermal style flying, and as Mark notes is dependent on cg. It is also dependent on the airfoil and planform and even to a small degree, the wing loading...

The simplest answer is to use a linear mix. It is a bit more optimal to use an inverse exponential style mix (more camber is added in the initial stick movement than in the large movement). For me, I am lazy and use linear typically...
Apr 26, 2014, 02:27 AM
Registered User
Tuomo's Avatar
Joe, do I read you right: you think that we should use snap flap more often in F3J?

I think I am a very typical F3J pilot here. I use snap flap in launch (zoom) and reflex mode. Otherwise I try to fly smoothly without overloading the wing and, if necessary, change flight phases. Actually the result of this is that at moments I have to change between flight phases very often. Or alternatively I opt for flying with little too much camber all the time just to keep plane flying in an effortless way.

My main criticism against snap flap has been that I have worried that extra camber liked to elevator will lead me to fly too slowly. Light F3J planes are draggy and have very little momentum... But the situation is almost reversed when flying with ballast. Overloading the wing in tight turns is then a very possible scenario - at least in moments when lift is weak or scattered.
Apr 26, 2014, 05:52 PM
launch low, fly high

I use elev>camber mix in all of my planes (F3B/F/J/K) and have it activated in all flight modes (I do not consider launch mode a flight mode). I probably should disable it when the glide path is in use (flap stick) but have not bothered to do this.

For me, I just want the control surfaces to be at the right position for minimum drag at all times, which includes any requests for maneuvering. If I am in thermal mode (or any flight mode for that matter), the intent is that the trailing edge preset is matched to the trimmed airspeed. Take the example of circling in a thermal in thermal mode. Presumably everything is optimized in the presets so that the elevator and camber is all near optimal. Then, the plane hits a strong bit of lift, and the pilot does a maneuver to recore in the stronger lift. If any up elevator is commanded, the wing will generate additional lift, above the lift that is optimal for the flight mode preset. Ideally, it would be helpful to have the camber adjust with elevator input such that the trailing edge will be in the right place after the elevator command. Hence the usage of elevator>camber mix.

Coincidentally, I found a handling quality delta yesterday when adjusting the elev>camber mix on an F3K plane. The plane was not handling well, and felt twitchy in pitch when thermalling in active and turbulent air where control inputs were being used significantly to counter the turbulence with a lot of adjustment in thermal circling. A pilot asked me about my elev>camber mix and I told him what it should be, then showed him what I had on my plane. Well, it turns out that sometime in the past, the amount of elev>camber mix had been reduced by about a factor of two as compared to what it should be. I changed the mix to be what it should have been, and the handling issue went away and the airplane "grooved" in thermals again instead of being a handful. F3J planes have significantly more inertia than F3K planes.

If I fly someones F3J plane, and it doesn't have elev>camber mix, the plane usually feels soggy to me when working thermals. When I ask for more lift from the plane while working a thermal the plane feels draggy. This may be confirmation bias, or it may be that I am feeling the drag that comes from the lift being too high for the trailing edge deflection. That said, I have had good success in the past of feeling whether the plane has the mix enabled or disabled.

A blind test would be something fun to do. Have someone put the elev>camb mix on a switch where the pilot does not know which position of the switch has the mix enabled or disabled. Then, go for some flights and have the pilot fly and experiment with the switch in both positions. Then ask the pilot as to which position on the switch feels more efficient. Ideally, this flying should be done at low level so that the pilot has the most accurate visual feedback as to the planes response to control inputs.

I have taken it for granted that elev>camb mix is required for a very long time... when I get back home from my current trip I'll put my elev>camb mix on a switch and do some evaluations to see whether it is still a good thing for my flying.
Apr 27, 2014, 02:55 AM
Registered User
Tuomo's Avatar
Thanks. This looks like worth a closer look. I will try snap flap in normal mode.
May 28, 2014, 09:20 AM
Needs to do 52 legs !!
jjmouris's Avatar
Joe, that is very interresting and directly opposite to what is suggested on the German F5B website.

Perhaps an F5B would also benefit from some elev>camber mix in the duration phase. Normally i have had this turned off going by the advice above but will now experiment and see if i notice a difference.

On a side note, with the high ish wingloadings in F5B (also F3B/F) it is very possible to induce a tip stall at low ish speeds when to much elevator is pulled without enough elev>camber mixed, yet when pulling full up at speed nothing happens. May be a function of how efficient the elevator is at producing lift at the various speeds. The whole thing is rather complex but would love some input from your self and Mark on the subject.
May 28, 2014, 01:34 PM
Stable genius
vespa's Avatar
jj, I don't see where you're getting that interpretation from the F5B site. It all looked Greek to me.

Note that he is plotting 6deg flaps across the range of CL as if the flaps "snap" downward at the slightest touch of elevator and permanently remain in that position. In practice however, you have essentially zero flap deflection at low CL and they deflect proportionally at CL is increased by the elevator. A more informative plot would show a small section of 0deg polar near CL=0.2, then 1deg polar around CL=0.3, ... up to 6deg around CL=0.8 or such.
May 30, 2014, 12:54 AM
Needs to do 52 legs !!
jjmouris's Avatar
From right down the bottom of the page.
Snap-Flap im Segelflug einzusetzen macht übrigens keinen Sinn, da hier die Querruder und Wölbklappen sowieso bereits weit nach unten stehen. Würden nun im Kurvenflug die Klappen noch weiter nach unten ausschlagen, hätte man im besten Fall eine bremsende Wirkung, im Schlechtesten einen Strömungsabriss.
Snap-flap in a glider to use way makes no sense, since the ailerons and flaps are already here anyway far down. Would now in the turn the flaps further deflect downward, would have, at best, a braking effect, at worst a stall.
I know, not the best translation from Google but trust me it says that using snap flap for the duration task makes "no sense". Suggesting more camber would translate to a braking effect at best, worst case a stall.

Thinking about it, you would only pull up when you think there is excess speed. So I don't think it should lead to a stall. The primary reason for that would be pulling to much up.

And yes, I agree that the two situation (0deg and 6deg only) plot is a bit misleading. Probably done to show the biggest difference in the polars. In reality the polar moves linearly with the deflection of the elev>camber mix as you say.

May 30, 2014, 03:12 AM
launch low, fly high

It definitely depends on the airfoils as to whether snap flap benefits the planes performance. For the type of planes that I fly (B/J/K) having the mix shows a benefit. I am not familiar with f5b designs. It may be that it does not work well for that design style although I would find that surprising.

As to the German link disparaging snap flap I can only hope that the euro competitors believe that! The NZ team needs all potential advantages!
May 30, 2014, 04:06 AM
Registered User
Originally Posted by Joe W

It definitely depends on the airfoils as to whether snap flap benefits the planes performance. For the type of planes that I fly (B/J/K) having the mix shows a benefit. I am not familiar with f5b designs. It may be that it does not work well for that design style although I would find that surprising.

As to the German link disparaging snap flap I can only hope that the euro competitors believe that! The NZ team needs all potential advantages!

We hope too...

Not all....
May 30, 2014, 05:33 AM
launch low, fly high

Snap-flap is horrible! Trust me...
May 30, 2014, 06:27 AM
satinet's Avatar
Snap flap is badly named as it give the impression of a violent, sudden movement, which is misleading.

CL mix anyone?

Quick Reply

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Question EVO9---snap flap on a slider-how to do it? Mike_Y Radios 0 Jul 30, 2006 05:56 PM
How much elevator to flap mixing is a good starting point? 4MistakesHi Beginner Training Area (Aircraft-Electric) 4 May 16, 2005 01:41 AM
Flap to aileron mix--how much? Far4nugn Electric Sailplanes 4 Mar 12, 2002 11:16 PM
How Much Motor? How Many Batteries? hoppy Electric Plane Talk 3 Apr 29, 2001 05:16 PM