Best shape for surface piercing foil

[tspeer]

What I found in matching Hanno Smits' data was the importance of varying the surface piercing foil's trim with speed. The constant chord V foil works out just fine if you reduce the lift coefficient with speed to maintain the optimum immersion. The change in lift coefficient is not as great as it is with the fully submerged foil, but it's not constant like I was assuming in the generic trade study. You can get the same constant drag characteristic of the tapered V this way.

So I don't recommend the inverse taper V foil. It didn't work for the Decavitator team then they tried it, either (but for reasons that can't be predicted by my simple analysis).

As for the effect of forward sweep on ventilation, I think there are a number of reasons it might be benficial. One reason is that it will tend to reduce the loading on the forward (deeper) part of the foil, and this can reduce any tendency to separation there. If you stop the separation, you stop the ventilation, so even if the top ventilates it might not spread to the rest of the foil. The spanwise flow moving up the foil instead of downward can't hurt, either.

There's another reason to rake struts forward - to align them with the total force vector and reduce the torsion loads on the supporting structure. Sam Bradfield mounts his hydrofoil struts forward of the beam for this reason. Notice where the struts are relative to the beam on his latest Scat design:

Draw a line from the foil to the beam, and you can visualize the foil's expected lift/drag ratio. You could position the foil in the same place with a raked strut.

[markdrela]

Originally Posted by tspeer

If you stop the separation, you stop the ventilation, so even if the top ventilates it might not spread to the rest of the foil.

You sure about this? I don't think you NEED separation to get ventilation.

During the Decavitator project, I had a couple of students do some towing tank experiments with an angled surface-piercing lifting wing for their SB thesis. The goal was to characterize the onset of ventilation. The airfoil was some generic low-camber section, maybe NACA 2410 if I recall correctly.

The most interesting discovery was that the wing had two stable states at most speeds and alphas:
1) normal operation
2) fully ventilated operation

At the start of the tow run the wing would always be in normal operation. It then might or might not switch to the ventilated state. The higher the speed and the higher the alpha, the more likely the switch to the ventilated statem but stalled operation was not a requirement to get ventilation. At low speeds ventllation was unlikely to occur naturally, but it could still be forced to occurby tossing a cup of water at the wing as it ran by in the tank. In all cases, once ventilation started it would persist for the rest of the tow run. The load cell indicated that there was a dramatic reduction in lift from the ventilation. Total drag didn't change much, but this just meant that CDp increased significantly while CDi decreased (from the reduced CL). All as expected, I guess.

We ha a camera on the tow carriage and watched the ventilation-start process. It would always begin near the LE-surface intersection, possibly at the Cp minimum, and then propagate down and back, covering most of the wing surface.

[MalSmith]

Originally Posted by markdrela

We had a camera on the tow carriage and watched the ventilation-start process. It would always begin near the LE-surface intersection, possibly at the Cp minimum, and then propagate down and back, covering most of the wing surface.

Did you try using a fence near the surface?

Mal.

[tspeer]

Originally Posted by markdrela

You sure about this? I don't think you NEED separation to get ventilation..

The most comprehensive source I have on ventilation is a Hydronautics handbook ("Hydrodynamics of Subcavitating Hydrofoils", Technical Report 463-1), and they do list separation as a prerequisite for ventilation. In the report, Barr lists 4 conditions required for ventilation:
- a finite area of separated flow must exist
- the separated area must be of sufficient size to allow passage of the air
- the conditions must be right for the ventilated region to be stable
- ventilating air pressure must be greater than the ambient pressure

Barr then goes on to describe the role of short and long laminar separation bubbles and leading edge stall, and trailing edge stall. He also mentions laminar separation starting at the trailing edge and separation at blunt bases.

It seems to me this dependence of ventilation on separation explains a lot with regard to the difficulties people have had with sharp-edged foils, such as ogival sections.

...The most interesting discovery was that the wing had two stable states at most speeds and alphas:
1) normal operation
2) fully ventilated operation

...We had a camera on the tow carriage and watched the ventilation-start process. It would always begin near the LE-surface intersection, possibly at the Cp minimum, and then propagate down and back, covering most of the wing surface.

Fascinating! That's also where one would expect laminar separation to occur. Could it be that the splash from throwing water at it opened up a "hole" to the atmosphere that allowed air to get at the laminar separation bubble?

[markdrela]

Originally Posted by tspeer

Fascinating! That's also where one would expect laminar separation to occur.

I don't think so. For the moderate-alpha cases the separation bubble would be well back on the chord. But ventilation always started very close to the LE.

I still don't see why separation is needed for ventilation. I haven't looked into ventilation modeling, but the process that I picture is:

Water depth is lowest at the min-Cp point on the piercing wing surface. The min-Cp value becomes more negative with depth because of the relief effect of the free surface. At sufficiently high q, the q*Cp gradient exceeds the hydrostatic pressure gradient, making the minimum-depth position unstable. The free surface then runs down along the wing surface, propagating the air pocket with it. A water surface disturbance can push this instability over the edge at some lower q, possibly because the ventilation itself aggravates the destabilizing Cp gradient in a positive self-feedback loop. Are there any alternative models?

[tspeer]

Originally Posted by markdrela

...Water depth is lowest at the min-Cp point on the piercing wing surface. The min-Cp value becomes more negative with depth because of the relief effect of the free surface. At sufficiently high q, the q*Cp gradient exceeds the hydrostatic pressure gradient, making the minimum-depth position unstable. The free surface then runs down along the wing surface, propagating the air pocket with it. ...

So what you're saying is the conditions were all set for ventilation on the submerged part of the foil, and were essentially pinched off at the surface because of the lift and minimum pressure dropping off there - so it started out in an unstable equilibrium. But once pushed over the edge, it was gone.

Evidently you were doing a force model test, because you say the lift decreased. If the model were free to trim in heave, I'd expect the instability to be even worse, since the full load would come onto the shrinking part of the foil that was still working.

I'm interested in how to engineer the foil system to prevent ventilation. I suppose the first step is to look at the threshold conditions where the minimum pressure drops below atmospheric - much like looking for incipient cavitation, but the pressure level would be different. I think the attached diagram captures tradeoffs.

The speed and foil loading to absolutely prevent ventilation are pretty low. It's going to be hard to find a foil that operates much higher than a depth Froude number of 2.0 without generating pressures low enough for ventilation. At 20 kt and a depth Froude number of 1.8, the section would have to be 11 ft (3.3 m) deep!

The loading factors are so low at shallow depths that any practical hydrofoil must be operating with pressures that can support ventilation. So there has to be some additional mechanism that determines whether or not the foil actually ventilates.

[boogie]

hi there,

very interesting thread. i have been working with windsurfer race fins for a while now and ventilation is a problem despite the bottom of the board working as an endplate. the air seems to be sucked in from the back of the board against the flow direction rather than with the flow from the front.

the best compromise to avoid ventilation we have found so far is at least 100mm of flat board bottom around the base of the fin.
if the fin is placed right at the tail then a little non structural plastic flap on the tail of the board that sits flat on the water surface does the same trick.


i am interested too in an answer to Tom's question to Mark Drela from earlier in this thread :

On a related subject, is Ncrit = 3 the recommended value in XFOIL for designing hydrofoil sections?

thnx for spreading your incredible knowledge
boogie

[MalSmith]

Originally Posted by boogie

hi there,

very interesting thread. i have been working with windsurfer race fins for a while now and ventilation is a problem despite the bottom of the board working as an endplate. the air seems to be sucked in from the back of the board against the flow direction rather than with the flow from the front.

It used to be common on sailboard fins to cut away the trailing edge of the fin near the root to minimise ventilation. There was also a "football fin" design, the profile of which bulged out away from the root. Are these designs used any more?

Mal.

[Kiteship]

"Modern" sailboard fins are very deep and have perfectly ordinary leading and trailing edges. Tapered rectangular profiles. Look a lot like... airplane wings.

There were a lot of "beliefs" throughout the development of sailboard fins, and before that, in the development of sailboard board shapes. It's started again with kiteboards, but it's still just a set of beliefs; rarely grounded in reality.

Dave

[sigurd]

Are fences not used any more in sailboard fins?

Is the sailboard speed limit (some 45kt's) considered to be due to cavitation or ventilation?

[boogie]

hi guys,

no more football or kangachook fins on windsurfers...
and lexan as a material for making them is long gone too....

outlines and sections have become a lot cleaner/simpler in recent years, but straight leading and trailing edges are not that common either. windsurfer fins vary a lot in shape and size depending on the gear they are used on. for light wind course racing [no centerboard] they are around 700mm long and for speed sailing they can be as small as 220mm with the same scaled outline. wave and freestyle fins are probably more what most people imagine when they hear windsurfer fin. like half dolphin tail or a dorsal fin.

i'm involved in windsurfing now for about 8 years and i have not seen a single fin with a fence. maybe it's time to spin the wheel again and try some out.

ventilation is more of a problem for the large lightwind race fins rather than the speed record ones. i personally have so far only managed 36.2kn on my GPS, so i think i'm still quite a long way from running into cavitation problems.
it is hard to tell when the top guys are wiping out at +40kn what exactly it was, but they all use fine entry sections on the foils of around the 8-9% thickness and quite a bit of rake in them [20-30deg]. the CL values are pretty small at those speeds despite the fins being very small in area.

would a cut out at the waterline or a reduction of chord length [with the same absolute thickness or greater %-age for structural reasons] actually be the best to prevent the ventilation?
if the suction peak close to the leading edge is the main contributor wouldn't an increase of the chord at the waterline drop the local CL and thus reduce the peak?

i'm still looking forward to an answer of the NCrit value question.

cheers
boogie

[sigurd]

"for light wind course racing [no centerboard] they are around 700mm long and for speed sailing they can be as small as 220mm with the same scaled outline."

As you say theyre CL might not be very high at those speeds, but from what little I learned I had guessed they would use larger fins still, to reduce AoA and thus the separation? But maybe it would cause too much friction that they couldn't get to the spinout speed?

What is kangachook?

Hydroptere, a rather new-ish foiling trimaran sailboat has several fences on its surface piercing foils.

I think it is odd that yellow pages claims to use superventilated or - cavitating sections, and still has the ~same speed limit as the subcavitating sailboard foils.

"if the suction peak close to the leading edge is the main contributor wouldn't an increase of the chord at the waterline drop the local CL and thus reduce the peak?"

I would have thought that the local CL would be about the same unless you changed its AoA? maybe play with especially low cP sections near the surface?

You say that the air is usually sucked in from behind, could this be because the trailing edge of the board is closer than any other edge?

Are there any vortices from the top TE of the fin to the surface behind the board?

Let me toss out an idea, what if there was a fence that was a downward pulling wing - it could have a positive pressure on the upper surface, right? It would create a high pressure channel between the board and the fence, would that stop the ventilation?

[boogie]

hi sigurd,

i nearly forgot about this thread....

it is a very fine balance on those course fins between area, section and planform in combination with the shape of the board.
the chordlength of the fin has a big influence on the "feel" of the board. a shorter chord makes the board more lively, but because the length is restricted you would have to increase the thickness to produce the same amount of lift, which makes the board creep a bit more sideways through the water [as the fin is fixed in position], which has all sorts of implications.

could someone give me an estimate how a proper fairing between the hull and the fin blade would affect the ventilation/separation close to the surface.
so far i've been using just a sharp corner as it is easiest to manufacture that way [file and chisel the rubberised resin which the base is moulded of after the blade is made].

if i remember correctly the kangachook was the same as a football fin.
but don't ask me what a female kanguru has to do with it...

we might see in october/november if the new YP [macquarie innovation or MI ] has the same speed limitations as windsurfers or other craft...
they have done calculations for YP and had a theoretical topspeed of just under 48kn, they managed 46.something
the same calculations for MI estimate a top speed of 58kn....
so far they just never had the right weather to show their potential.

reagrding the suction peak at the LE that causes ventialation.
i could think of several explanations, but certainly don't know.
a reduction in chord right at the waterline with the same absolute thickness [larger % thickness] could reduce the peak as the % thicker section has that peak at higher angles of attack.
doing the opposite of increasing the chord at the waterline, i thought would reduce the local CL due to the rapid change in taper in relation to the elliptical lift distribution [maybe it could be described as a non uniform downwash?].

wouldn't a downward pulling wing/fence have quite a bit of drag? wouldn't a larger fairing at the hull/foil juncton have a similar effect due to increase of pressure by its shape?

cheers
boogie

[tspeer]

Originally Posted by boogie

...i'm still looking forward to an answer of the NCrit value question.

IIRC from the XFOIL mailing list, Ncrit=3 seems to be a reasonable match to transition experiments in water.

Know of any high quality section data taken in water? We could do our own comparisons.

[Andrew Mason]

Regarding sailboard fins, cavitation is not the issue, ventilation will always occur first on a foil that operates close below a planing surface.

I have used fences on speed boards and the reduction in spinout due to ventilation was dramatic, however I felt that the additional drag was significant.

I found a better solution for speed sailing was to use assymetric fins that had a symmetrical section at the root. This meant that there was a non-lifting section of fin for approximately 1" below the bottom of the board and this then blended into a subtle assymetric section below that. The effect was also to make the board run straight through the water rather than adopting an angle of leeway. Immunity to ventilation on this particular fin was exceptional.

For return runs up the course the board was still sailable, although it would have quite a high leeway angle due to the section being used the wrong way around.