What is a significance of a wing thickness

These are two very different things.

Turbulators are tape, bumps, dimples, or grit, intended to force the boundary layer from laminar to turbulent. Vortex generators are little "sails" sticking up from the surface, whose tip vortices are intended to energize a boundary layer which is already turbulent.

You could use vortex generators as turbulators, but tape, bumps, dimples, or grit are a lot easier. Using turbulators on an already-turbulent boundary layer is detrimental, or useless at best.

 

You have to be clear just how thick the boundry layer is for turbulators to work efficiently. On high performance gliders that can fly at upto 120mph between thermals they are 1- 1.5mm thick zigzag tape because the boundry layer is so thin. They do work and create substantially less drag.

I beleive Vestas suffered from some interesting scrubbing actions near the trailing edge on it's underwater foil sections; maybe this cavitation/scrubbing action could be reduced by using turbulators or even a trailing edge like a birds wing - undulating on both axis to keep flow attached - it works for a bird through a wide speed range. Not sure how thick the boundry layer is at 52knts + though?

 

You also need to know the nature of the pressure distribution and boundary layer development to know where to place the turbulators, or if they will even be effective at all.

Turbulators are a cure for a very specific problem. That problem is laminar separation leading to either a long separation bubble or laminar separation with no reattachment at all. If your boundary layer is already turbulent, or if you place the turbulator past the point of laminar separation, it will do nothing for you. If you place the turbulator much farther forward than the separation point, then you will be creating more drag than necessary to solve the problem.

If you look at their application on sailplanes, they are typically applied on the lower surface just ahead of the cove in the trailing edge. The reason sailplanes use them is there is a favorable pressure gradient along most of the bottom but a steep adverse pressure gradient going into the cove. The favorable pressure gradient promotes laminar flow, but the increased pressure of the cove causes laminar separation. In the absence of a turbulator, the separated flow reattaches well back in the cove, making it less effective for the purpose of aft-loading the wing and increasing the drag. The turbulator is applied just ahead of the beginning of the increase in pressure, and the turbulent boundary layer is able to negotiate the adverse pressure gradient without separating.

The bottom of the wing is a good application for a turbulator because the separation point is driven by the shape of the cove and doesn't change much with angle of attack. On the top of the wing, the laminar separation point can move a great deal with angle of attack, from 70% of the chord or more to near the leading edge. So a turbulator is not a good solution there, and it's better to shape the upper surface so the pressure distribution will promote a short laminar separation bubble wherever the laminar separation occurs. If you wanted the turbulator to be effective over the whole angle of attack range, you'd have to mount it shortly behind the leading edge and you'd lose all the benefits of laminar flow on the upper surface. If you placed it well back, like on the bottom, then it wouldn't do much of anything for you because the laminar separation point would often move ahead of the turbulator.

On a bluff shape like a fixed mast, laminar separation can occur close to the shoulder of the mast. A turbulent boundary layer will stay attached for a greater distance around the mast before finally separating. A turbulator will use this effect to help to reduce the size of the separation bubble between mast and sail. But, again, you need to know that laminar flow is in fact present and where it is separating in order to apply the turbulator.

Vortex generators are usually a cure for a different problem - turbulent separation. They bring higher velocity air from outside the boundary layer down to the surface, and this higher velocity air is better able to oppose a region of increasing pressure to avoid separation. Like a turbulator, you need to know where the separation is occurring and place the vortex generators somewhat ahead of the separation point. The drag associated with vortex generators is higher than the drag of a turbulator, so you wouldn't want to use vane-like VGs if your problem is laminar separation that could be cured with a much smaller device. (There are small sub-boundary layer sized VGs that act more like turbulators, but they are not your typical VGs.)

Chances are, the designer didn't intend for there to be separation in the first place, so VGs are typically used to cure a problem that is discovered in testing. They need to be optimized using trial and error. That's why a friend of mine calls them, "the horns of ignorance."

 

Cavitation is a different problem. It is caused by the local flow velocity exceeding a threshold speed, which causes the pressure to drop below the boiling point of the water. Cavitation may cause separation, but it's not caused by separation. There would be a low pressure core to the vortices shed by a vortex generator, and cavitation could easily occur in those cores. So VGs may lead to cavitation where none existed before.

Vestas Sailrocket uses base-ventilated foils. These are wedge-shaped with thick trailing edges. The flow separates at the trailing edges and the cavity behind the trailing edge is filled with air. This is intentional, because the air is at a higher pressure than the vapor pressure of the water. The pressure on the back side of the ventilated trailing edge is therefore higher than it would be if the cavity was filled with water vapor from cavitation, and has less drag.

The fine pitting they are experiencing may be due to cavitation bubbles collapsing near the trailing edge. When a bubble collapses next to a surface, there's a tiny high velocity jet that forms in the middle of the bubble and the jet hammers the surface as the water rushes into the collapsing bubble. These bubbles could be forming from a pressure peak at the leading edge as the foil is loaded at high speed.

I don't think vortex generators would be helpful in stopping the formation of these bubbles, or in preventing their collapse ahead of the trailing edge.

 

Hi Markmal,
Thanks for starting a very interesting thread.
I do not claim to know anything much at all about sailing or sails, but I do know a little bit about aircraft wings. It is my ambition to one day build a boat rather similar to Planesail. http://www.planesail.com/index.html

I have a few questions, and maybe a few answers to your question.
Q1. I would like to ask which is the computer program that you used to create your aerofoils?

Q2. Would you be able to label the axis of the graphs more clearly? I think we are missing some information to be able to understand them.

Q3. Did you invent the thick section (red) wing? I say this because; from looking at some reference material, it is very unusual for the lower surface (windward?) to rise above the cord line.

A1. I think you have answered your own question in your first post. A thin wing with large camber can perform equal to a thick wing. However, you computer program does not include the "parasite" drag of a cloth sail. There is a join from the mast to the sail, the fabric's skin friction, stiching, seams, battens etc. A thick section composite wing can be incredibly smooth.

A2. The Oracle wing sail has a symetrical section, and not only this it has a flap. It is usual for an airplane wing flap to be a continuation of the aerofoil, however it appears that they use one aerofoil behind another. Why I am not sure, but I guess that it is because it is constantly used, boats travel slowly relative to aircraft, and that aircraft only need these at take off and landing.
The flap alows the camber of the wing sail to be varied, and if it has a slot, then the air passing between wing and flap re energises the upper (leeward?) airflow. This alows an increase in Cl, and a delay of the stall.
I have no idea how to control a cloth sail, but I understand you pull ropes and try to stretch them into the shape that you want. The solid wing is obviously far more controlable, with levers for the flaps.

The oracles wing is completely impractical, and needs a crane and a huge support staff to hoist it and lower it every day. A sail can be reefed and completely furled in port, whilst a wing sail is always "flying".

I see that you are interested in a windsurfer. Just thinking about it, it would seem impractical. The weight to provide strength to withstand impact with the water would be huge. The sail would be several times heavier than the board.

Please excuse my nautical terminology, it is all swahili to me.
Best Wishes,
Adam

 

>Q1. I would like to ask which is the computer program that you used to create your aerofoils?

XFLR5

>Q2. Would you be able to label the axis of the graphs more clearly? I think we are missing some information to be able to understand them.

Unfortunately no. I am a newbe, and do not know the program well.

>Q3. Did you invent the thick section (red) wing? I say this because; from looking at some reference material, it is very unusual for the lower surface (windward?) to rise above the cord line.

Yes, I got a default section that is created when created the project, and just mowed the lower surface up to make this shape. I did not know it is very unusual. I thought it will add to camber and add lift. What is wrong with it?

>However, you computer program does not include the "parasite" drag of a cloth sail. There is a join from the mast to the sail, the fabric's skin friction, stiching, seams, battens etc. A thick section composite wing can be incredibly smooth.

Windsurfing sails nowadays made of film. They are very smooth. they have some stitching but not too much. Parasite drag is not very high, for example I can hold mast with 0 AoA sail on wind 10-15 knots without any significant effort.

>The flap alows the camber of the wing sail to be varied, and if it has a slot, then the air passing between wing and flap re energises the upper (leeward?) airflow. This alows an increase in Cl, and a delay of the stall.

Here I do not understand. You are saying the slot helps to increase Cl.
But why we do not see slots in airplane wings or bird's wings? Only animals that have doubled wings are dragonflies and butterflies. But I doubt they use their wings same way as birds, they create lift not by camber but just waving wings up and down.

>I have no idea how to control a cloth sail, but I understand you pull ropes and try to stretch them into the shape that you want. The solid wing is obviously far more controlable, with levers for the flaps.

Windsurfing sails are kind of semi-soft semi-solid. They made of quite rigid plastic film with battens that hold their shape quite well. But I agree, solid wings hold shape much better.
But my question was not about materials - soft vs hard.
The question was what benefits thickness itself gives to a wing section.
And as I see - nothing.
Why I started this research - I read this doc http://users.soe.ucsc.edu/~elkaim/Documents/ElkaimDesignCat.pdf
It claims:
"A sloop rig sail can achieve a maximum lift coefficient of 0.8 if
the jib and sail are perfectly trimmed. Realistically, an operating
maximum lift coefficient is 0.6. The design goal of the Atlantis
wing is to achieve a maximum lift coefficient of 1.8. Since this
allows the wing to generate three times the force of an
equivalently sized sail, the wing area is reduced to one third of the
area of the original sails. Because the drag characteristics of the
wing are much improved, the performance of the wingsailed
catamaran should be superior to the original configuration."

And, the most important thing:
"First, in order to achieve the high lift coefficients at low Reynolds numbers, a very thick section is required,..."

That is why I've started my research. My idea was - if the above is true, I could make a thick sail from PE-foam, which will have 2-3 times more lift comparing to film sails. Also I looked on Oracle cats as an example.

Now I see - there are no benefits in thick wing sail. Better to get good flat sail.

 

I wouldn't say there are no benefits to a thick wing sail. Thickness will allow the structure to be lighter. This is important for a wing because it has a lot more structure to it than a typical mast.

And thickness allows the wing to be efficient over a wide range of operating conditions. A thin section can have less drag at a given operating condition, but it will perform poorly at conditions that differ substantially from the design condition. So a thin section needs to be adapted to the operating conditions. This is why a soft sail has so many controls for adjusting the draft, position of the draft, twist, rotation of the mast, etc.

When Gabe started his thesis, I'm not sure how much he really knew about wingsail design. The section shape he used is actually more reflective of high Reynolds number design philosophy, with its aggressive pressure recovery that leads to the concave contours of the section. And it wasn't necessary to have separated flow on his flap. He could have achieved higher performance if he'd used a typical two-element wingsail arrangement. However, the purpose of his project was really autonomous control, and he may have thought he needed to minimize the aerodynamic moment on the wing.

I don't have the coordinates for his section, but I reverse-engineered a similar shape, as can be seen in the first two attached figures. The first figure is from one of Gabe's papers and the second is my section analyzed for the same conditions.

The third figure shows a comparison between the Elkhaim-like single element section and a typical two-element wingsail section with a 40% chord slotted flap. You can see that the two-element section is much thinner. Flap deflections of 10, 15, 20 and 25 deg are shown.

The last figure shows the drag polars for all 4 flap deflections plus the Elkhaim-like section (blue), analyzed at a Reynolds number of 250,000. Except for the minimum drag at 0 lift (which is not of much interest to the wingsail designer), the two-element section out-performs the thick single element section in every way. The maximum section lift/drag ratio is higher and occurs at a higher lift. The maximum lift is almost twice that of the single element section. The plain flap on the Elkhaim section would raise the maximum lift somewhat, but nowhere near what the slotted section can achieve.

 

Hi Markmal,
As I said I doubt a windsurfer could be improved with a wing sail, however your question referred to Oracle.

Have a look at the first few minutes of this video, the camera work is shocking, but you can clearly see the guts of a variable camber rigid wing with a variable slot.
http://www.youtube.com/watch?v=_I7k2T2juxo

I cannot say from looking at your thick aerofoil what is wrong with it. Maybe nothing at all. At a guess I would say the windward surface rising above the chord might cause a moment pushing the leading edge to windward and the trailing edge to leeward. I believe very old aerofoils used to have concave lower surfaces, but most modern aerofoils are flat or slightly convex. Using results of a known high lift aerofoil, and a known symetrical high lift aerofoil may be more rewarding than a comparison using one you have invented as a beginner.

Almost every aircraft from a tiny Cessna to an Airbus A380 has flaps which create a slot. Let us look at a brilliant aircraft the Boeing 747. The designers wanted a wing that would carry over 400 tonnes at over 500mph at over 40,000 feet. Yet the aircraft must be able to take off and land from airports all over the world at speeds much less than this. The wing has leading edge flaps, and a double slotted fowler flap. The leading edge flaps increase the wing camber. The fowler flaps extend the chord of the wing, thus increase the surface area. The angle of the fowler flaps increase the camber of the wing. The double slots re energise the airflow thus allowing a higher angle of attack. For take off 20' flap is used, this increases the Cl, and also Cd, however as it is travelling relatively slowly and has large thrust the increase in Cd is not a problem. For landing 40' flap is used, this increases the Cl, but much more the Cd. The lift to drag ratio decides an aircrafts glide angle, gliders have alot of trouble on approach, crop dusters can approach almost verticaly.
Leading and trailing edge devices can increase the Cl by 120%, and double the Angle of attack before a stall.
http://www.decodedscience.com/wing-flaps-for-lift-augmentation-in-aircraft/11831

The discussion of materials is important. If you made your thin rigid wing it would only be able to sail one way, and would have no strength at all unless it was heavy. On the computer it performs, yet in reality it is useless. That aerofoil can only exist as a soft sail. The thick section could be made much stronger and lighter.
A very thin symetrical section would in no way be better as a wing sail than a thick symetrical section.

Thickness to chord ratio is very important in aircaft wing design. 15% is considered thick and produces high lift as used on crop dusters. 10% is considered medium and general purpose. 7% or less is considered thin and high speed. Concorde had an incredible thickness of only 3% over a 30 metre chord.


"Now I see - there are no benefits in thick wing sail. Better to get good flat sail"
This statement is completely incorrect.
Perhaps you mean it will be better to stick to a cloth sail for windsurfing.

Best Wishes,
Adam

 

Hi Everybody,

Thank you for this interesting thread I've been scrutinizing since start. I would have a candid question regarding:

Thickness & Off-design performance,

As mentionned by Tom Speer thick wing sections have a larger spot (in term of AoA) than thinner sections.

A.O Smith, in his "High Lift" workpaper, shows a theorical high lift section (Liebeck style) with a max Cl around 3, and its thickness is minimun near leading-edge and close to zero elswhere, quite similar to a windsurf luff-pocket rig, but with some concave shape= Stratford-type recovery distribution in its aft part.

The same kind of section customized for low Reynolds (Thanks to M.Drela , T.Speer) exhibits a low theorical range for AoA , which seems unpracticable for sailing applications.

So the straightforward questions IMHO are:

Are these thin sections actually unpracticable for sailing applications ?

What could be the actual AoA range for similar thickness modern soft rig (A-Cat type) or windsurf rig ?

I feel there is a contradiction somewhere unless the average skill of an A-Cat or a windsurf crew enable them to address successfully a narrow AoA range ?

I think it is important to have more info on this issue in order to define some limits, in term of AoA minimum range for sailing applications.

Any comments are welcome


Regards Everybody

EK

 

High Lift Aerodynamics, the classic article by A. M. O. Smith from 1975, is available for free download at http://www.arvelgentry.com/amo/High-Lift_Aerodynamics.pdf

 

Whether they are impractical or not depends on
- whether the angle of attack can be kept within the design range
- whether the designed sail shape can be obtained

Maintaining the angle of attack range depends on how actively the sail is trimmed, and what can be done with mast rotation and sail twist. But regardless of how well the sail is trimmed, the turbulent fluctuations in the wind have to also be within the angle of attack range.

It's going to be very difficult to achieve a precise section shape, even for the case of a rigid wing. The critical leading edge shape can be built into the mast, so that's achievable. The inflected shape of the sail could be obtained by the use of spanwise tension.

Imagine a fully battened sail that is sheeted out until it contacts a shroud. The battens will bend around the shroud and generate an inflected shape. Now imagine that the shroud is built into the sail itself. This can be done by routing the mainsail leech tension inboard of the leech. High-modulus material that runs directly from clew to head would cause a flat spot in the sail and allow the roach to fall off behind the high-modulus band.

The roach can be reshaped to be much larger than is typical in the bottom of the sail, leading to a planform shape that has its maximum chord some distance up from the clew. This sail shape may also reduce the induced drag compared to having the maximum chord at the foot.

The real value of the ideal high-lift sections is to point in the direction that one might want to take the sail design. A practical solution will need to be iterated between the structural realities of the sail construction, atmospheric conditions, and revised theoretical section shapes.

Can you provide section coordinates? The angle of attack range is going to depend on the actual shape and sailing conditions.

Skill is definitely part of the equation. Some sailors will need a wide, shallow groove, while better sailors can stay within a deep, narrow groove. And thereby realize better performance.

 

Thank you Mr Speer,

Re-reading it, I realize how my former msg was fuzzy, in fact the clarity of my English is sometimes questionnable, depend on how tired I am.

the complete story is:
In the late 80', just after discovering the Liebeck wing section, I had the opportunity to talk with the french C-Class OTIP design team.(Not very successful)

Their comments were that these sections were unpracticable for sailing application as a consequence of too narrow AoA range.

As these guy were PhD in fluid dynamics, I did not dare to question their conclusions.

But some doubt remained in my mind:if thin Liebeck wing section would be unpracticable for sailing application, so it should be the same for similar ultra-thin wing section like A-Cat rig or windsurf rig.

Thank for taking time to explain the sail concept, it is already on the board, but as a second step. For an A-Cat, a first step is planned:

In cooperation with a sailmaker, it is planned to do first:a sail with a max chord 2.2 meters above trampoline. the foot sail will be sweeping the trampoline and much shorter in order to allow the crew to tack or gybe behind the sail.

Instead of a boom, it will be a light wishbone at 80 cm above trampoline, so at the clew, the chord section will be shorter than the max chord.

The short wishbone will be prolongated by a short pole, long enought to arrive above the circular track.

And if it works as expected, everything will be in place to try a "Liebeck style" wing section just like your description.

But it seems to me that optimizing induced drag, can lead to a significant gain, at a very reasonnable cost. That is why it will be the first step.

Coming back to the core subject, I gusee that 2 wing sectionswith similar thickness, similar leading edge radius, and similar camber, should achieve similar AoA range regardless of the shape, some section providing more power than others?

Thanks agin and best regards

EK

 

Definitions please

Before I correct anyone's definitions about aerodynamics and lift, I think I should ask you to define your terms.

Specifically, lift, drag, chord, and camber?

Wayne

 

I apologise if you already know it
Modern windsurfer sails are usually fully battened (many have camber inducers -"cams") and buttons are bent "automatically" when sail is rigged.
They can hold a desired shape quite well, without wind. With wind, though, shape can have some deformations, depending on wind force, and how well the sail is designed and rigged. For example camber can move from front side to middle.

Also I'd mention that windsurfing is very dynamic, a sailor can change angle of attack very quickly reacting to changes in wind strength and direction.
Sailor also can change camber to some degree pulling or releasing clew line.
But this is not that fast.

And what would be an ideal high-lift section for a windsurfer sail for winds 5-15 knots.

Also in previous post you mentioned that thick section works better in wider conditions. Could you elaborate what these conditions are.

 

The reflexed shape is due to the very aggressive pressure recovery. The problem with that is if you don't get it right, the whole region separates. As a practical matter, I've found that flat shapes to the trailing edge produce a less concave pressure recovery that is more forgiving. But has somewhat lower performance, of course

I don't really like the notion of leading edge radius. It was necessary to specify leading edge radius for the NACA thickness distributions, but the best leading edges are not circular. But given the same leading edge shape, you have a point.

It's certainly true that most sections can be fairly tolerant of distortion in the middle, but it really depends on what the most critical areas are for the section shape. If you use a Stratford-like pressure recovery and the slope is a little too steep at the beginning, then you will separate the whole region. If the pressure recovery is less aggressive so separation starts at the trailing edge and works its way forward with angle of attack, then you can probably get away with some distortion in the middle.

When analyzing a section using a tool like XFOIL, look at the boundary layer H values (OPER-->VPLO-->H command in XFOIL). When H is approximately > 3, separation will occur. You can also look to see where the skin friction coefficient is going to zero (OPER-->VPLO-->CF command in XFOIL).

So for a practical section, design it with some tolerance from separation, and then if you don't quite achieve it, you may still realize most of its intended performance.