Winglet Design

[WladEZ]

Hi,

 

This is my first post. I´m considering building a LE. I´ve already read plans a number of times and some doubt still remains. My first is:

 

Why are the winglets tilted inward in most LE. I´ve been studying winglet design for a while (I have them installed a pair on my SZD Std Jantar Glider) and ALL the times the winglets must be tilted at least 90 degrees to the wingplan, better still if a couple of degrees outward...

 

Also, the plans seems to not mention the toe angle. Should I follow the tradition and use 2 degrees toe out as the boeings 737?

 

wladimir

 

brasilia - df

brasil

[Waiter]

Wladimir, Welcome.

 

The winglets are 90 degrees to the wings.

 

The gear should be set at about 1 deg toe in.

 

Waiter

[Marc Zeitlin]

Why are the winglets tilted inward in most LE.

The winglets are 90 degrees to the wings.

Actually, they're not exactly 90 degrees (at least on the COZY MKIV, and the LE is, I believe the same as the COZY MKIV). There is a slight inward cant to the winglets in the nominal case. The tolerance on the dimension to the tip of the winglet is +/- 1 inch, so you will see some measurable and visible differences in the apparent angle of different aircraft's winglets. It makes no difference to performance, which is why the tolerance is so large.

 

I believe that you're thinking of Whitcomb winglets, which are used to maximize L/D when aircraft are cruising at or near the max. L/D point, which would be at a much higher Lift Coefficient than EZ's normally fly at (because we're flying lower and faster than would be optimal for efficiency). In that case, the winglets ARE more efficient when they're canted slightly outboard.

 

However, in the EZ family, that's not what the winglets are for - they're there as vertical stabilizers and rudders, not as effective aspect ratio enhancers. Whatever small effect they have at reducing spanwise flow isn't affected by tiny changes in cant angle.

 

They're vertical (or slightly inward canted) to minimize the yaw/roll coupling that would otherwise occur with rudder deflection with an outboard cant. Since airliner winglet don't have rudders on them, they don't need to worry about this issue.

 

Also, the plans seems to not mention the toe angle. Should I follow the tradition and use 2 degrees toe out as the boeings 737?

The gear should be set at about 1 deg toe in.

I think that he was referring to the toe-in/out angle of the winglets, not the main gear, yes?

 

If so, the dimensions on the plans explicitly determine the toe-in/out angle of the winglet, and in those cases, are toleranced to +/- 0.05". You don't get to pick your own aerodynamics for the winglets, and following the example of a 737 wouldn't be appropriate - it's an entirely different airplane, with an entirely different aerodynamic design, with an entirely different operating regime.

[Vacabrava]

Wladmir,

To inspire you, and if you don`t know that one yet, go to:

http://www.cozybuilders.org/Oshkosh_Presentations/2007_Blended_Winglet_Eracer.pdf

That is Marc`s Oshkosh presentation of Jack`s ERacer blended winglet.

Real cool stuff.

Cheers.

[ZUCZZ]

Welcome Wladimir!

 

My advice: You should follow the tradition of building as per plans

 

It really makes it simple and you end up with an aeroplane that does what everyone else says it does !

 

Now if you were building a Boeing, then I'd follow those traditions :)

 

Happy building!

[WladEZ]

Thanks all for the reply.

 

I'm definitively not wanting to change nothing in the well proven design. Just the winglet design that called out my attention.

 

Even operating with a low CL a well designed winglet can raise the efficiency of a wingplan (that is, without extending the wingspan).

 

Now I can see that in the LE case the main objective was to provide a vertical surface and yaw control. The only point I can see for the inward canted WLs is to provide a downforce and thus lower the loading on the canard. Even so the force is so minimal that I doubt to be of significance in pratical terms. And by placing a downforce at the wing, you are lowering your L/D and consequently the distance flown.

 

my 2c

[Marc Zeitlin]

Even operating with a low CL a well designed winglet can raise the efficiency of a wingplan (that is, without extending the wingspan).

The point of increasing the effective aspect ratio is to lower the induced drag. Since, when operating at low CL's, the induced drag is a very small part of the overall aircraft drag, lowering the induced drag by a few percentage points when the induced drag is only ~20% of the overall drag will change the overall drag by well under 1%.

 

Now, change the effectiveness of this induced drag reduction due to the cant of the winglets by a few percentage points, and the overall drag change due to the winglet cant will be so far deep into the noise (in the 0.01% - 0.1% range) that whether you had a large breakfast will have more effect on the speed of the aircraft than the winglet cant will.

 

Whitcomb winglets make a difference when at high CL cruise. Other operating regimes don't benefit much from them, or from higher aspect ratios.

 

... The only point I can see for the inward canted WLs is to provide a downforce and thus lower the loading on the canard.

I said pretty clearly:

 

"They're vertical (or slightly inward canted) to minimize the yaw/roll coupling that would otherwise occur with rudder deflection with an outboard cant."

 

Since the winglets are above the CG, when the rudders are deflected they will not only create a yawing torque but will also create a rolling torque, and the rolling torque will be OPPOSITE to the desired coordinated flight roll with yaw - i.e., as the right rudder is deflected, causing the nose to yaw to the right, a coordinated turn would have the right wing lowering. However, with a vertical winglet and a deflected rudder, the force vector from the deflected right rudder will cause a roll to the left to begin.

 

Canting the winglet inward points the rudder force vector more closely at the CG so that this opposite roll torque is minimized.

 

It has nothing to do with creating downforce (reducing lift) and thereby reducing canard lift requirements. This would make no sense whatsoever, especially given your belief that the winglets are substantially contributing to increased L/D as Whitcomb winglets - this would be in direct opposition to that effect.

[WladEZ]

I said pretty clearly:

 

"They're vertical (or slightly inward canted) to minimize the yaw/roll coupling that would otherwise occur with rudder deflection with an outboard cant."

 

Canting the winglet inward points the rudder force vector more closely at the CG so that this opposite roll torque is minimized.

I could not figure out how the LE rudders could adversely induce a yaw/roll coupling. Could you help me on this?

 

It has nothing to do with creating downforce (reducing lift) and thereby reducing canard lift requirements. This would make no sense whatsoever, especially given your belief that the winglets are substantially contributing to increased L/D as Whitcomb winglets - this would be in direct opposition to that effect.

But by canting them inward they do generate a (small) lift component pointed down. I do not know however if this is significant in practice.

[Marc Zeitlin]

I could not figure out how the LE rudders could adversely induce a yaw/roll coupling. Could you help me on this?

The CG of the aircraft is slightly below the longerons - actually probably about 9" to a foot below the longerons. Since the winglets are above this, if they were perpendicular to the wings (which have no geometric dihedral) then when the rudders were deflected, they would create a non-balanced (by the opposite winglet) force towards the fuselage, but above the CG by about 1/2 the height of the winglet. This force times the moment arm would create a moment around the CG of the aircraft, attempting to raise the wing with the deflected rudder.

 

With the winglet canted inward, the force vector from the rudder is pointed more closely at the CG of the aircraft, so the moment arm is smaller and the rolling moment is smaller, meaning that there's less yaw/roll coupling.

 

Think about the extremes - if the winglets were canted inward 90 degrees, then a rudder deflection would ONLY force the wing down - it would be like an upward aileron deflection. If the winglets were canted outward 90 degrees, essentially becoming wing extensions, then a rudder deflection woudl ONLY force the wing upward, like a downward aileron deflection.

 

While absurd conditions, analyzing these pathological cases indicates the sign of the effect, if not the magnitude.

 

But by canting them inward they do generate a (small) lift component pointed down. I do not know however if this is significant in practice.

It's not. With a one degree cant inward:

 

sin(1 degree) = .017

 

so the component of inward winglet lift that is directed downward is 1.7% of the total. If the inward lift is even a couple hundred pounds (I think that's way overestimated), then the downward force might be 3 - 5 lb. at an absolute maximum. This is hardly significant in the context of a 2000 lb. aircraft.

[TMann]

........... a one degree cant inward:

 

sin(1 degree) = .017

 

so the component of inward winglet lift .......

OWwwww, ow owwwwww! Marc's making my head hurt again!

[steve]

OWwwww, ow owwwwww! Marc's making my head hurt again!

mine too. i like the part where he mover the fin thingy to show the wing go up or down.

 

Degrees & Radians

1 radian = (180 ÷ pi) degrees

or approx. = 57.295 779 513 082 320 . . . .°

57.3 will usually do.

 

1 degree = 0.017 453 292 519 943 . . . . radians

90° = 1.570 796 326 794 896 619 . . . . rad

[longez360]

Gang, some Long EZ winglet dimensions...

 

Plans winglet locations [bL, FS, WL], (dimensions in inches)..

 

WRP [55.5, 149.5. 17.4] … I am assuming WL17.4 for the top of the aileron, given 0.6deg washin.. Delta in dimension has small consequence so I didn't bother with the calc.

 

Winglet Root chord, Leading edge [157.145, 160.5, 18.4], giving plans dimension A=102.15 …WL 18.4 is probably not appropriate, but again has small consequence. Excuse the decimal places for BL calcs.

 

Winglet Root chord, Trailing edge [157.259, 187.6, 18.4], giving plans dimension B=108.35 … WL 18.4 is good

Winglet Tip chord, Trailing edge [x, 196.6, 66.4], giving plans dimension C=118.35

 

That would be approx -0.241deg toe out :-) for the root chord. With plans tolerances assume 0 deg, hence angle of incidence of root and tip is 0. The tip has the same toe. I.e. Twist = 0

 

If the winglet chord line was 90 degrees to the root (x = BL157.2), Plans dimension C would equal approx 122.3". Ie. The winglet is canted in slightly as Marc states and is not 90 degrees.

 

The Berkut cant is 4" outboard measured from the relative trailing edge BL positions of root and tip sections. i.e. BL plumb bob from tip trailing edge to root trailing edge. This is a significantly greater value of C when transposed to a Long EZ. I believe the A and B numbers for the Berkut are same, becasue they use the same WRRP, not that which is moved by way of a deepr aileron chord... ;-)

 

These numbers might aid an individual who is blending at a 30.4" WL cut line like mine. :-)

 

I have completed a basic 2D aerodynamic investiagtion of revised airfoils and winglet planform, and am now doing some CFD work. All going well, I should be flying new winglets relatively soon (relative to my 18 month downdraft cooling mod! :-)) They will be 80% built off the aircraft.

[Jack Morrison]

Way to go on the winglets Wayne. You should see about 10 mph increase from the info I have. I showed a 10-11 kt increase on my E Racer but that intersection has a little more drag than the Long, no one knows how much, but there is a slight difference. Besides the increase in airspeed, it really looks neat.

 

Jack