Foiler Design

[Stephen Ditmore]

Nice paper - thanks!

 

[markdrela]

Thoughts? I seem to recall decavitator having no bullets anywhere.

The drawing and photo show the Decavitator T-foil intersection. You have quite a bit more stagger -- probably more than is needed to alleviate any intersection drag problems. The only problem with too much stagger is that it leaves less room for structural material at the intersection.

 

[Karl Wittnebel]

The drawing and photo show the Decavitator T-foil intersection. You have quite a bit more stagger -- probably more than is needed to alleviate any intersection drag problems. The only problem with too much stagger is that it leaves less room for structural material at the intersection.

Thank you Dr. Drela - that configuration is close to what I remember.

I basically did what Hoerner told me to do; I had a minimum thickness in mind for the foot of the strut for structural reasons and I did not wish to change the section. So the trailing edge of the lifting foil got drawn out a bit. Bill Beaver did drag a mainfoil with a similar fairing down the tank for a midshipman project and it was very competitive.

We shall see if this rudder holds together; it is solid carbon L's with a solid carbon bottom plate. I might have been more conservative given that the last one failed at the corner, but that was two windsurfer fins glommed on with a secondary join. Quite a bit more material there now, even with the stagger. Nonetheless my next foil will be less extreme.

I infer from your photo that you saw no hydrodynamic reason for bullet fairing the intersection.

PS Stephen Ditmore - John Zseleczky built a foiler out of Bill's tooling also.

 

[yellowcat]

We will not pursue the foil design on our catamaran. Simply said, there are way too many obstacles (algeas, debris, ufos ) in the waters where we sail. We could use a "Ski" concept type of lifting addition, an hovercraft perhaps ... But our conclusion is to have a vessel that consumes no fossil at speeds up to 8 to 10 knots with no wind. Also the fact that even if foiling reduces the wave slaming at those speeds, it will be a problem for many other aspects. Like i will require that every body wear a life jacket (not really new) at all time, but also be bukled up at all times. Having a human body fly all over the place at 25-30 knots while cooking or other activity is not my definition of fun. A cruise ship can allow people to walk around and do all sorts of activities at 17-20 knots because it will not stop suddenly. Even if i saw some people fall for less. Not the case with a 60 ft catamaran , i cant garantee i will not hit an island of algeas . With all the abnormal high tides and floodings, we have all sorts of floating objects. We are very carefull when we follow ships for wind and waves protection for their plowing effect.
Nevertheless, i like the physics of the foils, cavitation studies and wave piercing . One (many) interesting parallel is the propulsion system. I had an idea to use a ring around the prop, that was when i was long distance swimming, to not only protect me but to increase the effectiveness of the prop. The ring liked the algeas ... so it was a problem, but after some reasearch, i found the Rice prop ring (derived from the Kort) . It is efficient at low speeds. Perhaps i would need to add a turning ring with a cutting edge.
Cavitation is in the picture. Food for thoughts ...

 

[tspeer]

A question for Tom Speer (if he is still listening on this channel):

OK so the latest Oracle rudders had a bit of bullet fairing on the front. but the strut was not staggered at ALL. The two-part moth foils tend to have a mortise and tenon at the junction, which requires a bulb/bullet to get enough material there. So the question is, if you go one piece at the T (so you don't need the mortise/tenon), is the bullet any faster than staggering the strut and putting a nice fillet on both the leading edge of the strut and the back of the intersection? I'm thinking not (a la Hoerner). I stretched the local chord of the lifting foil out to fit the foot of the strut, so about 2x baseline chord for the lifting foil. I will add a little fillet to the front of the strut, but then it's going to paint.

Bora said he would put a point in front of any intersection and thought Bieker would too in light of the Oracle rudder fairing, but I think something very similar can be accomplished without a bullet per se (see photos).

Thoughts? I seem to recall decavitator having no bullets anywhere.

I wasn't involved very much in the hydrodynamic or structural design of the rudder foils for 17, so I don't know exactly what Michel Kermarec's thinking was with regard to staggering the wing on the rudder. I know he was aware Decavitator used a staggered main foil.

With regard to whether there's a point in front or not, my approach to designing a bullet fairing is based on superposition of the local velocities. The problem is the velocities from the strut and the wing add in the junction to raise the local velocity there. That has two problems - it leads to cavitation and it exaggerates the velocity gradients, which causes problems for the boundary layer.

Since maximum velocity in the design lift range tends to occur near maximum thickness, the first thing to do is to keep from having everything get thick at the same place. That leads to stagger.

What I've tried to do with a bullet fairing is to superimpose a high velocity where the strut and wing have low local velocities and low velocity where the strut and wing have high velocities. This smooths out the velocity gradients. Putting a velocity peak near the leading edge stagnation pressure means the bullet fairing has to extend forward of the junction. The velocity distribution for the bullet fairing itself would be shaped like a dumbbell, leading the cross sectional area distribution to also look like a dumbbell.

That said, I've not actually finished the design of any bullet fairings myself. I've drawn up some candidates and did some initial panel code runs, but I had to put that project aside before I got to the stage of refining the fairings. Exploring fairing shapes is something I hope to do in the next year.

 

[MalSmith]

Tom,

Slightly off the current topic, are the section co-ordinates for the H105 freely available? I'm looking for a section suitable for a surface piercing foil. Is the H105 still the best option?

Mal.

 

[Karl Wittnebel]

I wasn't involved very much in the hydrodynamic or structural design of the rudder foils for 17, so I don't know exactly what Michel Kermarec's thinking was with regard to staggering the wing on the rudder. I know he was aware Decavitator used a staggered main foil.

With regard to whether there's a point in front or not, my approach to designing a bullet fairing is based on superposition of the local velocities. The problem is the velocities from the strut and the wing add in the junction to raise the local velocity there. That has two problems - it leads to cavitation and it exaggerates the velocity gradients, which causes problems for the boundary layer.

Since maximum velocity in the design lift range tends to occur near maximum thickness, the first thing to do is to keep from having everything get thick at the same place. That leads to stagger.

What I've tried to do with a bullet fairing is to superimpose a high velocity where the strut and wing have low local velocities and low velocity where the strut and wing have high velocities. This smooths out the velocity gradients. Putting a velocity peak near the leading edge stagnation pressure means the bullet fairing has to extend forward of the junction. The velocity distribution for the bullet fairing itself would be shaped like a dumbbell, leading the cross sectional area distribution to also look like a dumbbell.

That said, I've not actually finished the design of any bullet fairings myself. I've drawn up some candidates and did some initial panel code runs, but I had to put that project aside before I got to the stage of refining the fairings. Exploring fairing shapes is something I hope to do in the next year.

Thank You Mr./Dr. Speer.

I would think putting a point on the lifting foil and staggering the strut are to some extent accomplishing the same thing. Perhaps doing both is better though, to slow the max speed of the flow where it hits the strut.

To my eye, these bullets simply increase the local chord on the lifting foil, so the velocity gradients are less extreme where the strut intersects. It creates more of a wing/fuselage intersection, and less of a wing/wing intersection. But if I am already extending the local chord of the lifting foil at the intersection by 2x the base chord, with a large stagger, perhaps I am getting some of the same effect.

Thanks again for your comments.

 

[tspeer]

Tom,

Slightly off the current topic, are the section co-ordinates for the H105 freely available? I'm looking for a section suitable for a surface piercing foil. Is the H105 still the best option?

Mal.

The H105 was never intended to be the best option. It was intended to be a pretty good option. To get the best option, you'll need a section designed to your specific requirements.

The coordinates were publicly available on a Japanese airfoil database site, but it disappeared some time ago. I've provided the coordinates to everyone who's asked for them, with the proviso that they send me any data they collect related to its performance. The coordinates are attached.

 

[tspeer]

Thank You Mr./Dr. Speer.

I would think putting a point on the lifting foil and staggering the strut are to some extent accomplishing the same thing. Perhaps doing both is better though, to slow the max speed of the flow where it hits the strut.

To my eye, these bullets simply increase the local chord on the lifting foil, so the velocity gradients are less extreme where the strut intersects. It creates more of a wing/fuselage intersection, and less of a wing/wing intersection. But if I am already extending the local chord of the lifting foil at the intersection by 2x the base chord, with a large stagger, perhaps I am getting some of the same effect.

Thanks again for your comments.

I think you're probably right.

I think the approach taken to design the fairings on 17 was to look at the volumes that were cavitated using RANS CFD. Then they replaced the vapor cavities with solid fairing, on the theory that the velocity on the cavity boundary was the incipient cavitation speed, and the flow along the resultant solid surface would have the same local speeds. It was better to have solid structure of the same shape as the cavity because the cavitation undoubtedly represented a loss of energy in the flow.

That would probably work with a staggered configuration as well. And it wouldn't necessarily require RANS CFD to do it. I don't see any reason why even a panel code couldn't generate the surface in the flow of constant velocity at the incipient cavitation speed. That would allow you to use a combination of both approaches, with stagger used to reduce the volume of potential cavitation regions, and small fairings to take care of them.

 

[MalSmith]

The H105 was never intended to be the best option. It was intended to be a pretty good option. To get the best option, you'll need a section designed to your specific requirements.

The coordinates were publicly available on a Japanese airfoil database site, but it disappeared some time ago. I've provided the coordinates to everyone who's asked for them, with the proviso that they send me any data they collect related to its performance. The coordinates are attached.

Thanks Tom. At this stage I just require a realistic section to use for thinking about few different arrangements, but if I ever build anything, I'll let you know.

 

[Karl Wittnebel]

I think you're probably right.

I think the approach taken to design the fairings on 17 was to look at the volumes that were cavitated using RANS CFD. Then they replaced the vapor cavities with solid fairing, on the theory that the velocity on the cavity boundary was the incipient cavitation speed, and the flow along the resultant solid surface would have the same local speeds. It was better to have solid structure of the same shape as the cavity because the cavitation undoubtedly represented a loss of energy in the flow.

That would probably work with a staggered configuration as well. And it wouldn't necessarily require RANS CFD to do it. I don't see any reason why even a panel code couldn't generate the surface in the flow of constant velocity at the incipient cavitation speed. That would allow you to use a combination of both approaches, with stagger used to reduce the volume of potential cavitation regions, and small fairings to take care of them.

My limited foray into meshing has made me appreciate panel codes. I do not know how to get geometry this complicated into something like aVL though. Perhaps I need RANS after all.

 

[tspeer]

My limited foray into meshing has made me appreciate panel codes. I do not know how to get geometry this complicated into something like aVL though. Perhaps I need RANS after all.

AVL uses a vortex lattice method for wings and slender body theory for fuselages. That won't work for this kind of detail. You'd need at least a surface panel code like CMARC.

Once you'd calculated the flow for the baseline geometry, you'd need to output the velocities in a dense volume surrounding the junction. Then interpolate within the volume to get the constant velocity surface at the incipient cavitation speed. That would be your revised geometry.

The rest of the flowfield would be unchanged with the revised geometry - at that flow condition (speed, angle of attack, leeway angle). Naturally, you'd then need to look at what happens for off-design conditions and modify the geometry accordingly.

All this is speculation on my part, since I haven't actually tried this process. Yet.

 

[Karl Wittnebel]

Initially I was just trying to get a feel for how much anhedreal is optimal; aVL should do that.

But I spent a lot of time with OpenFOAM. Meshing definitely is definitely a challenging step with current free utilities.

 

[P Flados]

The recent AC results and the C Class results have left us lots to ponder.

I am working on a very small shunting/tacking proa be used as a test platform for some ideas. One specific idea I would like to try is operating in Atlantic proa mode (ama to leeward) with surface piercing foil(s) angled more horizontal than is typical. The primary purpose of the foils would be lift (to allow low volume ama even when pushing hard upwind) with all side force just being a nice side effect (reduced loads on the dagger/rudder foil pair).

This configuration led me to consider the following:

For a full foiler with T foils or L foils up front, the horizontal portion is typically relied on for supporting much if not all of the vessel weight. This means that you probably end up with bigger foils than you want when you get well above lift off speed. For this application you probably want a foil with a good lift at takeoff, but very low drag at operating conditions.

The design for a surface piercing foil does not really need to be anything like the above. Consider a foiler similar to Hydroptere, but designed to have only the leeward dagger foil and the T foil rudder in the water. The vertical extension of foil at the bottom could be sized for providing all require horizontal force with acceptable leeway. The inclined portion can be then angled such that the foil only stays submerged to the extend that the angled portion supports the entire weight of the boat. It does this while operating at an AOA that provides a good L/D. In this configuration, foil section design can be optimized with a focus on good L/D performance without being forced to try to get the more difficult objective of maintaining "very low drag at the low AOA needed during fast operation".

Although designing a boat that only requires one surface piercing foil up front sounds challenging, is there any reason it would not work?

If you can make it work, it would seem that the automatic reduction of total wetted foil surface area would be a big advantage as would the potential ability to use foil sections that are less of a compromise with respect to low lift off speed and good operating point at full speed.

 

[Doug Lord]

The recent AC results and the C Class results have left us lots to ponder.

I am working on a very small shunting/tacking proa be used as a test platform for some ideas. One specific idea I would like to try is operating in Atlantic proa mode (ama to leeward) with surface piercing foil(s) angled more horizontal than is typical. The primary purpose of the foils would be lift (to allow low volume ama even when pushing hard upwind) with all side force just being a nice side effect (reduced loads on the dagger/rudder foil pair).

This configuration led me to consider the following:

For a full foiler with T foils or L foils up front, the horizontal portion is typically relied on for supporting much if not all of the vessel weight. This means that you probably end up with bigger foils than you want when you get well above lift off speed. For this application you probably want a foil with a good lift at takeoff, but very low drag at operating conditions.

The design for a surface piercing foil does not really need to be anything like the above. Consider a foiler similar to Hydroptere, but designed to have only the leeward dagger foil and the T foil rudder in the water. The vertical extension of foil at the bottom could be sized for providing all require horizontal force with acceptable leeway. The inclined portion can be then angled such that the foil only stays submerged to the extend that the angled portion supports the entire weight of the boat. It does this while operating at an AOA that provides a good L/D. In this configuration, foil section design can be optimized with a focus on good L/D performance without being forced to try to get the more difficult objective of maintaining "very low drag at the low AOA needed during fast operation".

Although designing a boat that only requires one surface piercing foil up front sounds challenging, is there any reason it would not work?

If you can make it work, it would seem that the automatic reduction of total wetted foil surface area would be a big advantage as would the potential ability to use foil sections that are less of a compromise with respect to low lift off speed and good operating point at full speed.

============
Paul, do you think it might be diagonally unstable? Or wouldn't it matter?