Francis Donaldson's thoughts on Evans Volksplane VP-2

HomeBuiltAirplanes.com

Help Support HomeBuiltAirplanes.com:

Pops

Well-Known Member
Log Member
Joined
Jan 1, 2013
Messages
8,047
Location
USA.
Hard to beat the geodetic wing for strength and weight. Almost everyone thinks it a lot of labor, but its very quick and easy wing to build. The extra labor is in varnishing all the wood strips.
 

mcrae0104

Well-Known Member
HBA Supporter
Log Member
Joined
Oct 27, 2009
Messages
3,267
Location
BJC
The torque box.
No. A torque box out at the tip of a cantilevered element does nothing to prevent twist because it depends on all of the other inboard parts to transmit the torque back to the root.

Take this wing:
torsion.jpg
The blue part is a nice stiff plywood box. The pink part is foam rubber. Apply torque at the wingtip. Does the stiff blue part prevent the wing from twisting?
 

103

Well-Known Member
Joined
May 2, 2015
Messages
215
Location
Wauwatosa WI

FritzW

Well-Known Member
Log Member
Joined
Jan 30, 2011
Messages
3,889
Location
Las Cruces, NM
No. A torque box out at the tip of a cantilevered element does nothing to prevent twist because it depends on all of the other inboard parts to transmit the torque back to the root.
The wing isn't a cantilever element, it's two independent beams. One end of each beam is on a pivot, the other ends are *free to move. You have to keep the beams from moving relative to each other.

*if one moves up it twists the cockpit and causes the other one move down

Test.jpg

Test2.jpg

Test3.jpg
Think of a cereal box. When all 6 sides are still intact it's very rigid. Where would you glue the cereal box to keep the spars from moving independent of each other?
A: as far from the pivot as possible.
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
12,845
Location
Memphis, TN
That’s where a V strut would stop the twist. It would require a bigger bulkhead in the fuselage.
I would think boxing it in would be better at the struts. The cantilevered part. Is not very long at that point and the struts could react to it directly.

Cross brace at the struts.
 

Pops

Well-Known Member
Log Member
Joined
Jan 1, 2013
Messages
8,047
Location
USA.
Pops could the labor be cut by sealing all the wood wit thinned epoxy before bonding all the bits together with EPOXY?

REF Cygnet Geodetic Wing is very Stiff!
Matt
You would still have to leave areas where you will be gluing the wood bare. Would be nice to dip the wing :) Spraying would be a huge waste of varnish.
I brush on two coats of Poly-urethane varnish and then a top coat of Poly-Fiber Epoxy Varnish for wood. MEK in the covering glue will not attack it like regulator varnish.
 

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
12,845
Location
Memphis, TN
Even though it can be sprayed, I cannot see spraying the polyfiber varnish unless I was covering a Lockheed Vega. That stuff is rough with just a brush. A fog, wow.
 

Pops

Well-Known Member
Log Member
Joined
Jan 1, 2013
Messages
8,047
Location
USA.
Even though it can be sprayed, I cannot see spraying the polyfiber varnish unless I was covering a Lockheed Vega. That stuff is rough with just a brush. A fog, wow.
I agree, not much varnish would get on the small pieces of wood. Overkill.
 

Riggerrob

Well-Known Member
Joined
Sep 9, 2014
Messages
1,406
Location
Canada
That’s where a V strut would stop the twist. It would require a bigger bulkhead in the fuselage.
I would think boxing it in would be better at the struts. The cantilevered part. Is not very long at that point and the struts could react to it directly.

Cross brace at the struts.
V-strut (ala. Stitts Playmate) would stiffen the wing in torsion and allow for a wing-fold mechanism. And for goo measure, add a structural rib where struts meet the wing spars.
 

mcrae0104

Well-Known Member
HBA Supporter
Log Member
Joined
Oct 27, 2009
Messages
3,267
Location
BJC
The wing isn't a cantilever element, it's two independent beams. One end of each beam is on a pivot, the other ends are *free to move. You have to keep the beams from moving relative to each other.

*if one moves up it twists the cockpit and causes the other one move down

View attachment 97112

View attachment 97113

View attachment 97114
Think of a cereal box. When all 6 sides are still intact it's very rigid. Where would you glue the cereal box to keep the spars from moving independent of each other?
A: as far from the pivot as possible.
Fritz, I understand your way of thinking about this--levers and such--but your frame of reference doesn't take into account everything it needs to. I'm hoping to help you see this from the perspective of statics/mechanics.

First, the struts, together with the inboard set of drag/anti-drag cables form a space truss that resists twisting. Let's assume for discussion that it's completely stiff. You see in this diagram how it resists letting the spars assume different angles by putting the the various cables & struts into tension or compression. I'm sure we see this the same way so far.

scan_1888.jpgscan_1889.jpg

Next, how does the VP2 wing resist torsion outboard of this space frame? A torsional moment applied at the wingtip is resolved as a force couple, one pointing down on the front spar and one pointing up on the rear spar (or vice versa). But the spars don't deflect as if they are pivoted at the root--they deflect outboard of the strut attach points as seen here (again, we are making the simplifying assumption that there is no deflection at the strut attach):

scan_1890.jpg

I believe the diagram above shows the problem you and Matthew want to address. Please remember this one very important thing as we move along: the spars are not perfectly stiff; 1) they deflect under bending and 2) they contribute little to torsional stiffness because their section is not suited well for torsion.

Now let's try fixing the problem by installing a "torque box" as you suggest (shown simply as a plane below). Let's assume it to be perfectly stiff. Because it won't twist, the leading edge and the trailing edge of the box always remain parallel, as do the chordwise sides. But notice that inboard of the box, the spars are still free to deflect and twist, just as they were before.

scan_1894.jpg

The net effect is that the bay containing the box remains unwarped, but it passes the torsion to the next bay, which deflects in nearly the same way as it did before. The diagrams below show the front view of the two spars, first without the torsion box, then with it.

scan_1893b.jpg
scan_1893.jpg



The last diagram illustrates the way that the box keeps the spars parallel, but only in that bay; it cannot prevent deflection of the beam at any point where it is not attached to the beam. (Remember, the spars are just as flexible and only as effective in resisting torsion as they were to begin with.) In addition to the excess weight, the bending stress in the spars in increased and new bending stresses are introduced about the weak axis of the spars.

A better solution would be either the lattice or VP-1 torque box location you showed in post #4. That location would wffectively reduce the cantilevered portion of the wing (with respect to torsion) to five bays instead of eight, reducing their deflection at the tips and the angle of twist of the tip rib.
 
Last edited:

mcrae0104

Well-Known Member
HBA Supporter
Log Member
Joined
Oct 27, 2009
Messages
3,267
Location
BJC
Also, others have mentioned V-struts or the need for a diagonal strut. The VP design already handles torsion between the fuselage and the strut attach points with a space frame as mentioned above. Mr. Evans was aware of the various solutions available and chose the one he did.
 

cluttonfred

Well-Known Member
HBA Supporter
Joined
Feb 13, 2010
Messages
7,292
Location
World traveler
OK, I now understand better what you are getting at, mcrae0104. Treating the wing as rigidly triangulated inboard of the strut attachments makes it clearer. At this point, though, I can see the merits of both arguments so I will admit to being a little confused as to which solution would be more effective in reducing the deflection that weakens aileron response.

On the lattice option, I have definitely seen wings that used the lattice over the ribs following the airfoil contour like the Sisler Cygnet. I wonder if a geodetic lattice replacing the drag wire bracing would be the best solution of all, though I am still intrigued by the simplicity of Fred Weick’s Ercoupe wing solution.
 
Last edited:

FritzW

Well-Known Member
Log Member
Joined
Jan 30, 2011
Messages
3,889
Location
Las Cruces, NM
Your argument assumes that the system is rigid inside first drag/anti-drag bay and the spars outboard of that are doing all of the deflection. That's not what's happening.

Your whole argument assumes the spars are the flexible part of the system. What's actually happening is the spars have very little deflection (bending) from the root to the tip. Almost all of the flexibility in the system is in the cockpit. If the cockpit didn't twist (like your assumption in post #31) the wings wouldn't twist. The spars are more than stiff enough to prevent to prevent twisting outboard of the struts. ...you just have to keep them from moving relative to each other.

For what we're talking about (the ailerons twisting the wings) the front spars and the front bulkhead are virtually stiff, same with the rear spars and rear bulkhead. In the VP-2 those two systems can rack (scissor) against each other (post #20), ...that's what causing the wing twist.

twist.jpg

You can either stiffen up the cockpit (no easy way to do that) or add a torque box to the wing. The most efficient place to put that torque box is out at the tip.
 
Last edited:

FritzW

Well-Known Member
Log Member
Joined
Jan 30, 2011
Messages
3,889
Location
Las Cruces, NM
Treating the wing as rigidly triangulated inboard of the strut attachments makes it clearer.
The wing is NOT rigidly triangulated inboard of the strut attachments, that's the problem. If it were the wings wouldn't twist. The ailerons don't bend the massive VP-2 spars outboard of the struts enough to notice.

>>>

Lattice would be great but your trading 4 pieces of plywood for 128 pieces of lattice and 16 glue joints for +/- 850 glue joints (literally) . You'd add the hassle of getting those 128 sticks through the ply ribs. And the wing has to be built dead nuts the first time, no twists and the spar spacing has to be set just right when the lattice goes on. On a lattice wing there's no adjusting the drag cables to trammel the wing square or tweaking the inboard compression strut to get the spars to line up with the bulkheads. ...it's very doable and has it's advantages, just know what your getting into.

I built a set of wings for a Fisher Youngster (four panels) and loved it. But by the time I was finished I was more than ready for it to be over.
 

mcrae0104

Well-Known Member
HBA Supporter
Log Member
Joined
Oct 27, 2009
Messages
3,267
Location
BJC
Your argument assumes that the system is rigid inside first drag/anti-drag bay and the spars outboard of that are doing all of the deflection. That's not what's happening.

Your whole argument assumes the spars are the flexible part of the system. What's actually happening is the spars have very little deflection (bending) from the root to the tip. Almost all of the flexibility in the system is in the cockpit. If the cockpit didn't twist (like your assumption in post #31) the wings wouldn't twist. The spars are more than stiff enough to prevent to prevent twisting outboard of the struts. ...you just have to keep them from moving relative to each other.

For what we're talking about (the ailerons twisting the wings) the front spars and the front bulkhead are virtually stiff, same with the rear spars and rear bulkhead. In the VP-2 those two systems can rack (scissor) against each other (post #20), ...that's what causing the wing twist.

View attachment 97163

You can either stiffen up the cockpit (no easy way to do that) or add a torque box to the wing. The most efficient place to put that torque box is out at the tip.
It sounds like you are saying that the cockpit lacks torsional stiffness between the two bulkheads that the struts attach to--understandable since that's right where the cockpit opening is. If so, a torque box at the wingtip will still not solve the problem; the fuselage needs to be stiffened.
 
Last edited:

FritzW

Well-Known Member
Log Member
Joined
Jan 30, 2011
Messages
3,889
Location
Las Cruces, NM
It sounds like you are saying that the cockpit lacks torsional stiffness...
Yes, me and the entire Volksplane community have been saying it since 1968. It's not a new discovery.


If so, a torque box at the wingtip will still not solve the problem; the fuselage needs to be stiffened.
Of course a torque box will solve the problem. The VP-1 had the same problem and the problem was solved by putting a torque box in the wing.

-

This discussion is just going around in circles, I think I'm done.
You can either stiffen up the cockpit (no easy way to do that) or add a torque box to the wing.
 

mcrae0104

Well-Known Member
HBA Supporter
Log Member
Joined
Oct 27, 2009
Messages
3,267
Location
BJC
[deleted by author - analyzing rear spar assuming rotating rear bulkhead]
 
Last edited:

TFF

Well-Known Member
Joined
Apr 28, 2010
Messages
12,845
Location
Memphis, TN
On a plane with the performance of a VP2, the torque box may be enough. It’s not an Extra. An Extra would need more. Just like flutter, all it has to do is not flutter for its job. If it went 200 it would be way under built. If you keep pushing the engineering, the question becomes more and mor scab fixes or fixing it. Two different approaches trying for the same result.
 

mcrae0104

Well-Known Member
HBA Supporter
Log Member
Joined
Oct 27, 2009
Messages
3,267
Location
BJC
Fritz, thanks for clarifying that the lack of torsional stiffness in the fuselage is the real culprit. This does change things somewhat, but the conclusion is the same: a torque box at the tips is still inadvisable as a way to prevent wing twist. I did an analysis of the rear spar with the assumption that the assembly of rear spars, struts, and bulkhead is not restrained from rotation by the fuselage, but rather by the proposed wingtip torque boxes. Using these torque boxes instead of the fuselage to react out the twist has several negative consequences. If the torque box transmits enough torsion to keep the rear spar in plane with the front spar when the ailerons are deflected, then it is introducing a significant additional bending moment at the tip of the spar (the worst place), and:
  • Depending on the load case (aileron up or down; I did not consider gust loads), the rear spar would be overstressed by +/-20%. This might be within the margin of safety, but the spar would need to be beefed up unless limit load is reduced significantly.
  • The wing attach fittings and possibly the bolts would be overstressed by +/-25% (the reaction at these fittings increases).
  • The front spar may be even more problematic (although I did not analyze it). The problem case would be aileron down. As the rear spar wants to rise, the torque box exerts a positive moment on the front spar in order to keep the rear spar constrained. This is in addition to the already-higher bending moment on the front spar of the ascending wing. The WAF & bolt would be even more stressed than the one at the rear spar, and also the forward strut could likely be overstressed or subject to buckling in this load case. Again, gust loading would exacerbate the situation.
 
2
Top