What is a significance of a wing thickness

Lift is defined as the component of the total aerodynamic force that is at right angles to the apparent wind vector and at right angles to the spanwise direction. Drag is the component of the total aerodynamic force that is parallel to the apparent wind vector.

The camber line is the locus of points that are equidistant from the two sides of the section. The chord line is a straight line connecting the ends of the camber line. Chord is the length of the chord line. Camber is the maximum distance between the camber line and the chord line.

Normal force is the component of the total aerodynamic force that is at right angles to the plane of the wing (perpendicular to the chord line in 2D). Axial force is the component of the total aerodynamic force that is parallel to the chord line. (Note that while drag is always positive, axial force can be negative (forward) at high angles of attack. This caused the collapse of some wings in early aviation history because their designer thought the force would always point backwards along the chord.) It is actually this forward component of lift that propels a sailboat, when the forces are resolved into components perpendicular and parallel to the axis of the hull.

 

Tom,

Thanks,

That sounds like aerodynamics. At the first part of the thread a couple of posts seemed to confuse aerodynamic lift with filling a sail with wind, and those are two completely different designs. And methods to move a vessel forward in the water.

Lift (in this manner forward) is a more powerful overall method). It does have some inherent issues when applied to watercraft. Higher cost, fewer knowledgeable people, unexpected moments (inertia), and design issues to furl your sail.

But, lift is powerful.

When we get a lower cost semi-rigid wing designed, the world will forget about mouse traps. At least for a little while.

IMHO.

Wayne

 

Thanks Mr Speer for your analysis, and also for the separation criteria, I
use to focuse on transition criteria when playing with XFLR5, did not know how to address the separation, so thanks a lot, it will be put to full use.

By the way, I noticed your MSES simulation mentionned Ncrit=1, is it comparable to the XFOIL default value Ncrit=9 ?

Best Regards

EK

 

Ncrit in MSES and XFOIL are the same thing. e^Ncrit is the amplification factor at which free-stream disturbances reach a level at which they transition into turbulent structures in the boundary layer.

Ncrit = 9 corresponds to an amplification factor which means the freestream turbulence is 1/8000 as much as the level at transition. This is a very smooth, quiet freestream, which is not likely to be encountered by a sail. Ncrit = 1` means the freestream turbulence is 1/3 as much as at transition. I used this value arbitrarily to better represent the freestream conditions.

The effective value of Ncrit that represents the turbulent boundary layer close to the water surface at which sails operate is unknown. It probably depends on the venue and the conditions.

 

Thanks a lot Mr Speer,

For a newbie, it is always difficult to guess what could be a relevant assumption for Ncrit, consistent with free stream turbulences. I prefer to use your arbitrary assumptions than mines.

By the way, Do you know if there is significant difference, from an XFOIL perspective, between a Ncrit which triggers the transition at 7% chord (i.e), and a forced transition input @ 7% ?

My sailmaker partner in A-Cat rig is using Smar Azure software for sail design. This software provides loads infos, deformations under loads, and so on...and sail panels cut ..of course.

We plan to try to have some iterations between XFOIL and Smar Azure, in order to achieve a 2D wing section likely to meet our objectives.

The starting point is to achieve the best input datas for the existing mast section, in order to provides good inputs for XFOIL. That is not that trivial, I am afraid of the "Garbage in / Garbage out" pitfall.

As you can observe, your Teardrop Mast workpaper is put at full use!

Thanks for your enlighting comments.. as usual

Best regards

EK

 

I really need to revise that paper, based on what I've learned since. Particularly with regard to mast trim. I'm finding it's generally best to rotate the mast to put the stagnation point near the leading edge, and maintaining a smooth lee side contour is not really the best approach. I wish the tools I have were capable of handling a Bethwaite style mast section, as it would be really useful to compare the two approaches on a common basis.

As for the whether transition is the same forced or natural at the same location, I'd say try it both ways and see what the differences are. You really need to analyze for the extreme cases anyway. Make some runs with as high a value of Ncrit as you think could be possible, and some runs with transition tripped as far forward as possible. Then see if there's any undesirable impact to the characteristics, like premature stall or excessive drag. Then you'll be covered no matter where transition occurs in practice.

 

Like Tom Speer mentions, the problem with the Liebeck style high lift profiles is that when they stall, they stall with a crash: the point of separation moves with a jump from the leech to mid chord. Not good when you consider how sail boats are always moving around in waves etc.

Another problem that (usually) you want sails to produce a high lift at a small angle - Liebeck style requires a large angle of attack. If you are looking for high lift profiles, I suggest you look at Selig rear loading-profile mean lines, such as S1223, if I recall correctly the number. That said, we did use a Liebeck-style profile in the top of our 470 jib, world champion in 1997. Don't use nowadays anymore, though.

On the other hand, the sail on an A-cat is very lightly loaded as soon as there's wind (at least upwind). The righting moment limits so that the sail is lightly loaded, I would expect you rarely exceed CL=1,0, mostly sailing with CL around 0,8. What does your Smar Azure VLM predict?

 

Hi Mikko,

Thanks for pointing on S1223. I've found Selig book ( http://www.ae.illinois.edu/m-selig/uiuc_lsat/Low-Speed-Airfoil-Data-V1.pdf ). Reading ...

I am looking for windsurfing sail. I do not think it has that issue of righting moment. Sail can be highly loaded and tilted windward so it creates vertical lift helping a board go planing. Also rider can quickly change AoA. What would be a good profile?

 

The lack of importance of the bottom makes sense if you use the idea of the flow around the wing being the potential flow well away from the wing plus some sort of bound circulation, as per your linked article. At each point in the air, including along the wing top and bottom surface, the energy in the flow is constant, consisting of potential energy (proportional to the pressure) plus kinetic energy (proportional to the velocity^2). The key part is that "velocity^2" term.

The pressure term is what gives you lift. Since the energy is constant, the faster the speed of the air, the less the pressure. More speed over the top = less pressure = lift. Less speed over the bottom = more pressure = lift. But an equal change in speed faster/slower between top/bottom does not result in an equal change in pressure, due to the velocity^2 term.

If by adding the circulation and external flow you double the velocity on top the wing, in order to keep the total energy constant at that point along the wing you have to decrease the pressure by something proportional to 4 x the velocity. Meanwhile, on the bottom of the wing you've halved the velocity, and so you increase the pressure by something proportional to 1/4th of the velocity. So circulation had a much bigger impact in change in pressure up top than it does on the bottom.

Assuming inviscid incompressible flow, yada yada...

 

Ok, this is for Tom and/any other readers on the other side of the "foil section design" experience curve.

The above post got me thinking about Tom's paper on using a Clark Y to make a wing mast. This all combined to finally push me into trying Xfoil (then XFLR5) to play around with some stuff.

I have been posting over at "Understanding Wing Technology" in regard to my efforts to try out some "average Joe" construction methods. However, in terms of wing section design choices, this thread is probably a better place for the following.

First comes understanding what I am looking for.

The slotted Wings of the C Class cats are getting a lot of interest right now, but I am not convinced that they are the right answer for what I am looking for. Without the rules on sail area and without the need for max VMG on a windward/leeward course, do I really need the complexity of a slotted wing?

In terms of want it does for you, a wing can be designed for optimizing Max lift (Cl), Min drag (Cd), and/or best efficiency (Cl/Cd). The C Class tends to place a lot of priority on max lift where wing drag is a lesser concern (wing drag is a low fraction of total drag and/or they just seem to need more downwind "grunt"). Other applications can consider using the wing for upwind and/or strong wind conditions and rely on a soft foresail to add the extra lift when needed.

I was playing around with trying to get reasonable lift (say Cl = 1.3 which is high compared to most airplane wings) and good Cl/Cd with a non-slotted wing that is optimized for operation with "lots of flap deflection" to achieve something that could be described as a "reversible camber design". For structural reasons, I really wanted a thick front portion.

At first, I took a "high lift" airfoil and used Tom's "just mirror the forward top surface onto the bottom surface to make your wingmast section" approach, but instead of using a soft sail surface for the rear half, I also mirrored the rear half to make a flap. My "G3" foil was the result.

When it gets to the flap, or to the rear of many boat wings, real life make it hard to keep the design curvature if you are using a film covering over ribs. I was interested to see if flat for the flap would hurt the areo. When I tried it, I stumble on better results. I kept playing around an ended up with a design I designated OWC. It seems to have good Cl, good Cl/Cd and the shapes have advantages for my build methods.

I have included polars for the OWC that show better performance than a pretty good traditional thick foil, NACA 4418.

With a goal of thick forward section, high lift, good Cl/Cd and "reversible camber" are there any comments or suggestions?

 

Hi P Flados,
Those are some amazing figures for a reversible wing. Thank you for sharing.
I tried to search your posts. Can you elaborate on what it is you are building and how you are getting on?
Thanks,
Adam

 

I was around sailing as a kid, but never got hooked up with a club or a group or anything.

After I had kids, I re-acquired the old beat up sunfish I sailed as a kid & tried to get my boys interested, but that was a flop.

Then we got Laptops with wireless internet and I stumbled on all of the action as everyone was trying to "break 50 knots". The Sailrocket site led me to the site with Bernard Smith's stuff, and I was hooked on thinking of ways to sail while getting basic physics work for you instead of trying to use brute force to beat physics into submission (usually a bad approach).

I am by no means wanting to compete in area where the high dollar efforts (can you say America's Cup & Hydroptere) are pushing things. Instead, I would really like to figure out some stuff that would help an "average Joe" build something that he can throw together just for the fun of sailing faster than anyone else at his local lake/bay/pond. I do not really expect to achieve this, but it is probably comes close to my "ultimate goal" for my "personal hobby".

Wings are of course a very likely part of any such effort.

Next spring, I hope to play with a little Proa I have assembled. I will probably use the old sunfish rig for the trials with a goal of working out some ideas for the foils/rudders.

Eventually, the Proa could end up with a wing. Alternately, I have an old junk Hobie 14 and a very light 11' moth style wood strip hull that could be made into a one person Tri.

For the wing, I want to go lower tech on construction than many would think is needed. For now I am thinking: no molds, no vacuum infusion, no high dollar carbon fabric. Carbon tow can be had cheap (I got a 4 lb spool a while back). It is probably the lowest cost option if you can figure out how to use it. Also, recently Mark Drela turned me on to bulk pultruded carbon rods as a cost effective raw material (see the Understanding Wing Technology thread). With either of these, I am confident I could build the structure for a low cost, strong and light wing in either a free standing or in a stayed configuration.

I started out thinking of a multi element wing. Foil section from Steve Clark's Open Wing with a cord scaled down to 85%. I am sure I could do it, but I am not convinced that it would be that much better than a simple flapped wing.

My efforts with XFLR5 have been just plugging in different stuff to see how it affects the results. My background (mechanical engineer, son of an aeronautical engineer) is such that the simpler aspects (laminar vs. turbulent) are easy, but I really do not want to dive into the deeper theory stuff. I would rather focus on bigger picture things like combining cost, structural, aero and construction considerations together for a package that would be more useful for those that would like to build out of their garage.

 

Not much help in response to past 40, but I have kept pluggoing at it.

I was not real happy with my results for either max lift (Cl) or for low reynolds number performance. I started over & have made some progress on both of these with my GOE 679 based foil.

 

Hi P Flados,
It looks like a tadpole, I hope it will not metamorphosise into a frog.
Very interesting indeed, keep up the good work.
I look forward to your further revelations.
Cheers,
Adam

 

I'm not familiar with XFLR5 graphics. Are the green dots on the circles your actual coordinates? If so, and this is what XFLR5 feeds into XFOIL, then you have one panel and one boundary layer finite-difference interval over the entire flap! Imagine running Fluent with one grid cell over the flap. The resulting solution will be mostly nonsense.

You need to make sure you have enough panels on the airfoil to get a grid-converged solution. For this kind of geometry I wouldn't use fewer than about 200 panels.

Also, the blunt trailing edge base should be roughly perpendicular to the flap centerline. You have a steep bevel which influence the Kutta condition in some possibly weird way.