Understanding Wing Technology

[tspeer]

...What comes to mind is that we need a permanently stretchy film (latex, gold spandex?) vs a hard film so that we can stretch it more than is required for the aero loads. In this way it stays in tension when the panel warps. The skin "buckles" in a classic shear buckle (see AVI) which means the distance across the diagonals of the panels is getting shorter. Its about 3-4mm shorter under load than not. So all you twisted wing guys out there we need solutions for the film.

The load from heat-shrinking the film is also a problem. It is the design load for much of the structure, such as rib compression and trailing edge stiffness.

The hyperbolic sag produced negligible effects on the aerodynamics of USA 17's wing. That may have been helped by the fact that the main element section shape was flat for the aft 40% of the ME chord. But it doesn't seem to be a problem with rag-wing aircraft, either.

If you had a very stretchy film so it didn't wrinkle under shear, I don't think it would do a good job of holding its shape, either. A woven fabric might be a better choice, because the fibers are at 0/90 and shear load distorts it on the bias. So you can still have reasonable fiber tension to hold the shape, and still deform in shear.

 

[petereng]

Hi Tom,
Thats good information. Some years ago I was involved with temporary shelters for boat repairs (25m long and 8m high was a small one) using PVC tube hoops and shrink film and if the builders were a bit over zealous with the shrinking they would bring the structure down! I was thinking along the bias woven cloth path as well. I've been thinking of a net under the film which can be pulled quite tight, then have a light film over the top, but then its two things to acheive one outcome. Perhaps I can change the panel size to minimise this effect. Tom what would a typical panel pressure be? So I can model a descrete panel and optimise its size. Perhaps a long thin panel is better than a squarish panel in this case. Cheers Peter S

 

[tspeer]

You'd have to look up the specs for the material you are using. For example, Clysar HPG shrink-wrap film 50 micron thick has a shrink force of 1.853 N/15mm. If I interpret this correctly, bay that is 1 m high will apply a tension of 123 N/28 lb on the trailing edge. If the bay is 2 m wide, then the end rib will have a force of 246 N/56 lb trying to bend it. On each side. Even for ribs in the middle, you may have to design for an unbalanced force as the film is progressively shrunk along the span.

 

[petereng]

Hi Tom, Sorry I mean't aero pressure I'd like to not use shrink wrap. I'm working towards using 150gsm sail cloth and hang glider construction. The inside of this rig looks so good! Cheers Peter S

 

[Erwan]

Aero Pressure

Hi Peter,

Here are the results of my homeworks as I promised you a long time ago.
I had to do these calculs for myself, but it was not a priority, I just hope I didn't make any mistakes. That is why I provide a checking procedure.

Please take a look at the attached file page 41 figure 71 you will discover a single, asymetric wing section, not slot, and this wing section was elaborated for low reynolds applications, very consistent for sails wind speeds.

On the vertical axis you have the Velocity V/Vo
You can see the maximum is around 1.85 for the upper side and the minimum is around 0.4 on the lower side, for an average around 0.65

I used the following formulas

Cp=1-(V/Vo)^2 It is a dimensionless coefficient

P-Po= 1/2*rho*V^2* Cp It is a pressure in Pa or Newtons *sq meters

F= (%chord)*14sqm*(P-Po)

I used a stick on the computer screen to read average V along a part of the chord, and the above mentionned formula.

To check I used the Cl=1.95 provided with the graph.

For assumption I used A-Cat assumptions:14 sqmeters @ V=6 m/s around 12knts apparent wind speed.

According to the Cl = 1.95 the total lift should be (checking formula)
F=1/2*rho*V^2*S*Cl
with
rho=1.225 kg/m^3
V= 6m/s
S= 14 sq meters

So F= 602 Newtons

Now using the graph in order to have the pressures along the chord.

UPPER SIDE

% V/Vo Cp---- P-Po(N/sqm) F upper side
30 1.80 -2.24 49N/sqm---- 206 Newtons
10 1.65 -1.72 38N/sqm---- 53
10 1.55 -1.40 31N/sqm---- 43
10 1.47 -1.16 25N/sqm---- 35
10 1.43 -1.04 23N/sqm---- 32
30 1.37 -0.88 19N/sqm---- 80

100% 449 Newtons


LOWER SIDE

% Chord V/Vo Cp P-Po F lower side

100% 0.64 0.59 13N/ sq meter 182 Newtonss


F upper side + F lower side = 449N + 182N = 631 Newtons

I achieved a result which is within 5% of the theorical result (602 Newtons)
so if you "cook" a little bit the velocities you read on the graph, it is easy to match the theorical 602 Newtons lifting force.

If you double the velocity you double^2 the pressure in Newtons/ sq meter

So with 24 knts apparent wind, the pressure on the first 30% of the section upper side, can be assumed to be around 200N/sqm.

That is my KFD (Kitchen Fluid Dynamics) approach,Hope it can provide you with a rought idea.

Above all, not sure it is correct, that is the first time I process thiis calculations. So wait a little bit, if nobody sees any big problem ..

Cheers Mate

EK

PS: % chord are in fact like % A-Cat sail area
The first line of the upper side considers the first 30% of the chord where the velocity is on average at the highest level: 1.80

So the related area is 14sqm*0.30

 

[Erwan]

the attached file

don't know how to send the file just find it on the following link

http://smartsonde.org/Project/Vehicle/progs/AIAA-2137-802.pdf

Instead of page 41 figure 71, look at page 6 figure 13.

 

[petereng]

Wing Sail Twist Control

Hello Erwan - Thank you for your efforts. Is the pressure along the span or along the chord? eg is the first 30% at 39Pa from the foot up the sail? or form the mast backwards along the chord? I presume the 30% means 30% of the sail area? Attached is a plot from a CFD package as printed in "aero-hydrodynamics" by fabio fossati. Detailed structural analysis needs detailed aero loads. Especially around the edges of the sails. I'll work through your numbers so I understand them. Cheers Peter S

 

[Erwan]

Hi Peter,

It is only pressure along the chord wing section, not alo,g the span, I do not consider 3D effects.

I jusst make a rought "diiscretization" of the continuous velocity distribution presented in the document.

It is a kind of "hand made integral"if you prefer, I slip arbitrary the chord wing section 30%+10%+10%+10%+10%+30% for the upper side
and only one 100% segment for the lower side in order to make it simple.

So the first 30% of the upper side correspond to the rooftop of the velocity distribution(between leading edge and 30% of the wing section,nearly horizontal) then for the remaining part of the velocity distribution, the slope require to slice in 10% intervals, and at the end the slope is nearly horizontal so to make it simple I consider again a 30% interval.

When I compute the lift I consider 30% of the total aera so it is the first 30% of the wing section*span=partial area

The pressure results are independent from the lift calculation, I caalcultae the lift in order to be sure it was consistent.

the lift Coef provided in the pdf file: Cl=1.95 is the dimensioless result of the integration of the velocity distribution.

IE:The first 30% of the upper side exhibits a max pressure 49 Newtons/sqmeter but it is for 12 knts apparent wind !!

At the tail of the wing section, the last 30%, the pressure is 19 N/sqm

For the lower side I consider a flat line would make a good proxy of the velocity distribution that is why I consider only one V/Vo=0.64 for 100% of the chord wing section.

Cheers Mate

EK

 

[Mikko Brummer]

Attached is a plot from a CFD package as printed in "aero-hydrodynamics" by fabio fossati. Detailed structural analysis needs detailed aero loads. Especially around the edges of the sails. I'll work through your numbers so I understand them. Cheers Peter S

I realise I have modeled the A-cat sail with a small wingmast earlier. I include here screenshots of the mast+sail, showing relative pressure on leeward (negative numbers) and winward side. It's not the pressure on a wing (leeward side very much is), but maybe it is of some use.

This is downwind, doing the wild thing, at an apparent wind angle of 58 deg. There is a 20 deg twist in the inflow, and the sail has some twist but I cannot recall how much. Relative pressure in Pa is the load on surface in Pa, N/sqm, so you can use it directly as a loadcase.

 

[petereng]

Hi Mikko,
Can your CFD analyse wing sails? Is the CFD in-house or contracted? Thanks for the plots. Regards Peter S

 

[Erwan]

Thanks Mikko

Hi Mikko,

Thank you very much to take time to post your research, at first glance your figures seem to be in the order of magnitude of my basic hand made calculations, are they ? If they are:

That is just the demonstration that paying attention to Tom Speer's comments and his countless efforts to disseminate correct fluid dynamic approach; combined with his great academic communication, can finally lift the knowledge of any "Beach-Billy", (more appropriate than Hill-Billy Isn't it?)

In addition, your CFD provide a 3D picture, that will be very useful for Peter I guess.

Thanks Mikko,

Cheers to all

Erwan

 

[Mikko Brummer]

Hi Mikko,
Can your CFD analyse wing sails? Is the CFD in-house or contracted? Thanks for the plots. Regards Peter S

It's inhouse, and surely could analyse wing sails, too. Biggest problem is I'm so busy with everything all the time.

 

[Doug Lord]

Understanding Wing Technology- Hounds System

Picture below of hounds system described here by Magnus Clarke:

It's a bent track, glued screwed and faired into the Leading edge of the wing. It has a single bent car inside the track with a number of holes for the T-terminals of the shrouds to connect into quickly. (Note, we broke a bunch of these terminals, getting them at the correct angle and seated properly is a good start to not dropping your wing overboard. that having been said, your loads would be smaller for the same fitting).

The whole show rotates around the front of the wing as the wing is trimmed from side to side. essentially your shrouds and froestay(s) stay in the same place relative to your hulls, the wing rotates behind / under the wires.

Good SA thread about Steve Clarks Open Source wing plans: http://forums.sailinganarchy.com/index.php?showtopic=118933&pid=3225410&st=25&#entry3225410
=========================
Steve Clark on the same subject:

Posted 30 March 2011 - 01:56 PM

There have been several solutions to the hounds on wing masts.
The problem is caused by the size of the airfoil section.
If you attach the shrouds and forestays to a single point at the leading edge, the mast hits the leeward shroud as it rotates. This happens on all rotating rigs, but gets worse the larger the wing. The wing is thereby limited in it's rotation, and the shroud can start to buckle the wing surface and ultimately damage the structural integrity of the wing.

So it is necessary to hold the shrouds away from the wing. We have done this with turntables, which work like Lazy Susans (a circular ball bearing race) where the shrouds are attached about 120 degrees from each other. The problem \here is that if you are using the skin of the D section for structure ( as we are on this wing) you end up cutting a fairly big slice out of it to accommodate the bearing races. This isn't ideal.
The PLVI wing has the first iteration of what is proposed for this wing.
There is circular shelf built off the leading edge of the wing that had a lip on it. The top surface of the shelf and the inside surface of the lip had a bearing material (like Teflon or UHMW) on it. There is a matching "slug" inside that shelf that the shrouds were attached to that could slide around the perimeter of the shelf. There is a "lid" that keeps everything from escaping. It looks too simple to work, but it does.
On the VI wing the slug is a pretty significant bit of stainless, on the Weta scale this can be something like UHMW.

The bottom picture is the Cogito version of the hounds. This follows a different concept because there is a structural mast inside the leading edge that takes all of the compression load. This was nicknamed "the garden gate" because it swings on the mast as if on a vertical hinge. The shrouds are again held apart at about 120 degrees so that they don't hit the wing surface as the wing rotates.

Many words, Hope they help.SHC



click on image:

 

[Doug Lord]

Understanding Wing Technology-Oracle+ Optimist

Check this out: http://www.boatdesign.net/forums/sailboats/new-rig-optimist-37734.html#post458053

click on image:

 

[Erwan]

Hi Peter,

With regards to the above-mentioned aerodynamic loads on the leading edge of a wing. Could you provide a little advice about the required stifness of the leading edge skin, in order to avoid "buckling" or deformations, especially for the leeward side.

For the leading edge, I would like to use a 170g/sqm kevlar fabric, (easily available) as the base material. Then I need to see if I must use it within a sandwich, and/or adding carbon UD verticaly between ribs, and also what would be the appropriate underlying ribs'grid?
I am not a structural engineer, I am just looking for basic guidelines, to avoid big mistakes or unconsistencies.

Thanks in advance

Cheers

Erwan