Foiler Design

Alan, I tried your URL on msn and google and got no results; can you check it and maybe post a "hot" link...

 

Search engines will not find the link ar this stage but if you copy and haste it into the address box of internet explorer it should come up.

 

It appears the designer of the "Kooee" and I have been exploring a similar concept. The use of foils to produce a torque to counter heeling is very intriguing.

I've posted some renderings to the "Sailing - Multihull" forum if anyone is interested.

 

auto or manually stabilized foilers

For what it's worth the Rave, designed by Dr. Sam Bradfield and the Hobie Trifoiler designed by Greg Ketterman use different systems to achieve automatic and virtually unlimited stability.
Some guys sailing Raves in Gainesville Florida have also eliminated the wands(surface sensors) and now fly the boats manually with a joystick!

 

Doug

For the record the relationship between Kooee and Rave is only slight, Kooee would be more akin to Sailrocket but would be much more useful as an every day sailer (given a suitable waterway or windward shore). However the guys with joysticks fitted to Raves will be having a ball. They must be getting close to the speed where the boat will get too “twitchy” to handle.
Regarding stability and limit conditions, stability is a very miss understood term; many people make the mistake of believing any balance of forces and moments establishes stability. In actual fact you need, as a minimum, to calculate the first derivatives of your balance equations and hence establish you are at a stable trough not an unstable pinnacle. The established aerospace way of doing this is to solve the differential equations in all six degrees of freedom.

In a 1956 book “Survey in Mechanics" K S M Davidson wrote, “Who can say what maximum speed will some day be attained in sheltered waters by following one or another of the various leads?” He was referring to a number of configurations covered earlier in the book. I believe we can see, maybe not that top speed, but a barrier. It is possible to design boats capable of two and a half possibly three times wind speed and some of these designs can maintain this relationship in relatively high winds. The limit occurs when the foils cavitate or when the four components of drag will simply not allow the apparent wind angle to be further reduced. For non-cavitating foils the limit is somewhere in just in excess of 60 knots. Foil design post cavitation is not my area expertise, I ask the question are there foil designs that will work above say 60 knots that have lift to drag ratios better that ten or there a bouts. Can Tom Speer enlighten me (us)?

Alan Smith

 

I think it's hard to get much above 40 kt without cavitation.

Here's a comparison of 3 sections, all around 12% thick. The grid of incipient cavitation speeds is universal.


The only way one can get a cavitation speed appreciably above 40 kt is to go to thin sections and a very narrow range of low lift coefficients. For 60 kt, the maximum velocity on the foil can only be 66 kt. Thin sections quickly develop leading edge suction peaks as angle of attack increases, and this drives one to small lift coefficients and low foil loadings.

Of course the actual cavitation speed will be above these incipient cavitation limits, but there will also be regions of locally accelerated flow, such as in junctions, that will cavitate at even lower speeds.

 

hi tom,

very intersting graph indeed.
i have one question though.

does the reynolds number have any influence on the cavitation properties of a foil?

i'm working on some windsurfer fins [symmetrical section] for a speed trial and those fins are very small in comparisson to most hydrofoils. 250mm span and 80mm chord at the base and around 40-50mm at the tip.

i have made fins around that size from tips of my course racing fins with a 9% modified Eppler836 section and they work realy well so far, but i hope to improve on that.

cheers
boogie

 

Tom

Thank you for the foil cavitation info. I had previously missed it in visiting your web site. The people chasing the WWSSR barrier at 50 knots are perhaps in more difficulty than I had realized. The estimated performance figures presented at [URL http://home.kooee.com.au/zach/hydrofoil.htm[/URL] are based on a 7% thick foil. The section has been developed using “Xfoil” operating at a Cl of 0.08. Xfoil gives a maximum Cp which at 25m/s which equates to 380lb/ft^2. At these high speeds the foil is intended to operate at zero angle of attach developing lift by use of flap angle only. Under these conditions less than one degree of flap is required and the section maintains a very even pressure distribution across its width. Careful attention to the strut – foil junction can minimize the junction problem but not eliminate it.

It maybe of interest, for the above performance point the drag components of the Kooee are:- air Cd0, Cdi water Cd0, Cdi 9.7, 25.9, 60.6 and 3.8% respectively.

Hence it can be seen that keeping the induced drag of the sail down is very important but induced drag and use of a high aspect ratio of foil is relatively unimportant.

Changing Kooee’s configuration and optimizing to stay within the cavitation limits of the Speers H105 section yields an estimated maximum speed of 40 knots in 12 knots of wind!


Further input will be appreciated.

Alan Smith

 

I don't think Reynolds number has an appreciable effect on cavitation, as long as the flow is attached. There'd be a small change in effective shape due to the different thickness of the boundary layer, but that would be in the noise compared to the other influences.

 

Yes, it's a tough problem!

I've not heard of anyone talking about base-ventilated foils - that might be a reasonable compromise between subcavitationg and fully cavitated approaches.

That's the way to go. Both the Eppler and NACA sections shown above are flat roof-top type sections, and the E817 was specifically designed to avoid cavitation. It shows what you can do with a 12% thick section.

The thinner 7% section would reduce the minimum velocity ratio, and the flap aft-loads the section. The change in lift due to angle of attack is concentrated near the leading edge, leading to higher velocities there and earlier cavitation. The flap would raise velocities all along the chord, reducing the peak velocity for a given lift increase. The flap is likely to form a pressure peak at the flap hinge that could cavitate, but the amount of foil affected downstream is just the flap and not the whole suction side.

Good thing, too! It'd be hard to make a thin high aspect foil stiff enough!

Of course, induced drag depends on span^2 and is inversely proportional to speed^2. It's going to be more of a factor in taking off in marginal conditions than it will be for breaking the speed record. So a good way to reduce the thickness ratio and foil loading is to simply extend the chord.

Bear in mind the H105 was designed as a low speed hydrofoil section, and cavitation wasn't a design consideration. It was intended to operate in the under-30 kt range, and also work well at the Reynolds numbers of a half-scale model. So it's a good match for a beach cat or Moth, but it's no record-breaker. By design.

 

I find angle of attack to be a somewhat arbitrary basis of comparison. For example, here is how cavitation buckets are usually presented:



Since Cp is proportional to velocity squared, this compresses the bottom of the bucket - which is really the region of interest. And the curves are shifted by the amount of the section's zero lift angle of attack. When you look at these curves, you don't realize the superiority fof the E817 for heavy foil loading.

By plotting the velocity ratio vs lift coefficient, I find it much easier to relate the section characteristics to the system design. It would be possible to put lines of limiting boat speed on the Cp plot, but it wouldn't be possible to include the lines of constant foil loading.

 

That sounds about right, although the foil loading may be on the low side. I scaled down the E817 from its normal thickness of 11% to 7% and 5%. Camber was scaled along with the thickness.


Here are the resulting velocity envelopes:


The E817 modified to 7% just gets you to 50 kt, and the E817 modified to 5% maxes out at 60 kt. But the thin sections have a very narrow range of angles of attack with which to work. The whole non-cavitated range of the 5% section at high speed is only two degrees of angle of attack wide. Although at 50 kt these two degrees of section angle of attack also represent a range of forces of 400 lb/ft^2 to 1600 lb/ft^2 - that's 0.5 to 2 g's at a mean loading of 800 lb/ft^2! It lends a new meaning to "twitchy" handling characteristics.

As you point out, induced drag of the foil is not a big player at high speed, so it would be a good idea to use a low aspect ratio foil to reduce the 3D lift curve slope and make the angle of attack less critical.

 

excuse my ignorance, but could you please explain this a bit more?
i understand that at high speeds or low loads the induced drag plays a lesser role, but what exactly do you mean by the second part?

keep up your excellent posts Tom. this forum is worth more than any book.

thnx
boogie

 

The lift curve slope (approx. 0.1/deg) you see in airfoil section data is the two-dimensional slope - what you would see if the foil had infinite span. But the three-dimensional planform has the downwash of the fluid displaced in the wake of the foil. So the foil operates in an adverse vertical current of its own making.

The adverse current is proportional to the lift on the foil, and it reduces the angle of attack actually experienced by the foil compared to what you would expect from the geometric angle of attack between the foil chord and the boat's velocity vector. The net effect is to make it appear as though the foil was not producing as much lift as it ought to for the same geometric angle of attack.

The shorter the span of the foil, the greater this adverse current is for the same amount of lift. So giving the foil less span reduces the three-dimensional lift curve slope. Adding sweep to the foil will also reduce the lift curve slope, as well as extend the cavitation speed.

A good approximation of the 3-dimensional slope is:

a = a0 / [1 + (180/pi) * a0 / (pi * AR * e)]

where a is the 3D lift curve slope of CL vs alpha, a0 is the two-dimensional lift curve slope (Cl vs alpha; per degree), AR is the aspect ratio, and e is the Oswald efficiency factor.

For more information, see
Figure 23, http://naca.larc.nasa.gov/reports/1948/naca-report-921/naca-report-921.pdf
http://naca.larc.nasa.gov/reports/1939/naca-report-651/naca-report-651.pdf

 

hi Tom,

thnx a lot. you have definitly given me some fat to chew over...

cheers
boogie