Foil Cavitation at Lower Speeds Than Expected

Here are the inviscid pressure distributions for the NACA 63012, 0012 and 0009 sections at a lift coefficient of 0.15.

When you compare the 0009 and 0012 sections, you can see that the 0009 is better than the 0012 because of its smaller thickness. But if the lift coefficient were a little higher, the leading edge pressure peak on the 0009 would have a higher velocity than the 0012.

The NACA 63012 section has a lower maximum velocity than the 0009, despite being a thicker section. And there's room for additional lift before the leading edge velocity exceeds that of the 0009. If the leading edge were shaped so as to round off the front of the pressure distribution, there would be even more angle of attack range without cavitation.

So, with proper shaping, you can have both the thickness you'd want for structural stiffness and a high incipient cavitation speed. Programs like XFOIL are very useful for designing a section that will give you the right tradeoff of thickness, lift range, and cavitation speed for your application.

 

Will try to absorb all this interesting input ;-) . Have Java-foil installed, will check out XFOIL.

 

For what it's worth: Each of my V's weighs about 6lb = 2.7kg & each 1/2 of each V is slightly less than 1m in length, so 1.5kg/m for the aluminum wouldn't be a huge amount.

 

I was going to try almost exactly that, but I worry about the situation where the tips are the only thing in the water. If the flat portion is too big, then you would be throwing away much of the stability advantage of a steep dihedral. Of course, the boats with nicely rounded apexes have more-or-less the same problem, but they're not known for being easy to sail (I'm thinking of the NACRA 17's).

Another option is to just extend the outboard portion of the V (as I mentioned earlier). If extended enough, then the inboard portion of the V would be out of the water when flying high & a lot of the flow problems from the V would be avoided. I think this would be much simpler for a home builder.

Another thing that would help is to simply fly lower. My aft foil had adjustable incidence, so I could control it while sailing & usually avoid cavitating, but the mechanism wasn't as good as on the Moths' aft foils, so I sometimes couldn't adjust it fast enough.

 

God point and very doable.

I am doing a weight calculation and is desperately and without luck trying to get down to your stipulated 200lbs, anyway it seems to be below 220lbs. Are the 6lbs V´s including beam attachments? Also I see you have a nice platform of tubing with the shrouds attached to the tube between the cross beams, are the tubes all aluminium?

 

I have a breakdown of the weights of all the pieces somewhere, but can't seem to find it at the moment.

I do remember that the total was 210lbs with the amas, or 190 without (so 200 average). And your main hull is 3' longer, so you can be forgiven if yours is a little heavier.

You could probably save a bit by using carbon tubes. Yes, mine were all aluminum & weighed ~28 1/2 lbs altogether. You might want to consider how you can increase the torsional stiffness of the platform though. At high speeds (& small foil angles of attack), any twist runs the risk of getting the angle <0, which is a recipe for disaster.

 

Yes, the 6lbs includes the vertical stubs that bolt to the similar pieces on the crossbeam.

 

Oops, only my 100x2mm main and 75x2mm rear beam are 35lbs. How thin did you go, o.o65" ;-)?

 

The forward crossbeam is 17' long, 4" OD, 0.049" wall. The aft crossbeam is 17' long, 3" OD, 0.035" wall.

Each of the 4 connecting tubes are 5' long, 3" OD, 0.035" wall.

To get enough stiffness, I had to use waterstays (unfortunately).

 

Thanks Doug, the thinnest available tubes in Sweden are 2mm/0.08" wall, quite a difference. Think waterstays are unavoidable.

 

Did some more thinking about the aluminium section and modifying it to 630xx. Would really like to use this sturdy section from www.windknife.com. Did a spreadsheet to see much had to be milled or added to the corda.

We can rule out 63012 and 6309 but the other two would probably be acceptable candidates, 63A011 being close enough, without even in need of milling in the nose.

Thought of widening the section by adding some wood, but it might give a hump. Still the tail will need milling, but not if we go for a 63A0xx. Note the figures aren´t completely accurate, but good enough for comparing.

Theoretically modifying the section might be of benefit, question is how much it will do in reality?

Must order a sample of it to see if it is even possible to add the piece of wood.

 

Lars: This question is of great interest to me, both in terms of quantifying possible improvements to Broomstick & also for improving my overall understanding.

I have a lot of experience in using XFOIL for 2D sections & a V-Foil Vortex-Lattice code for analyzing the 3D effects, so I will see what I can do to combine the two & try to answer your question.​

 

Not being familiar with XFOIL but having JAVA-Foil installed, I tried that one instead, comparing 0009(line) and 63A011(dotted).

With 30 points and 63A011.3 this is the closests I can get to 0009´s nose. But I guess it doesn´t matter if you use 11,3 or 11.

 

The Windknife section is close to the NACA 0009.

The tail shape is not that important, although a sharply squared off trailing edge would be better than the round trailing edge of the extrusion. What matters most is the shape up to the maximum thickness, so I'd concentrate on making changes there.

You might consider adding material forward of maximum thickness, which could be just filler because the strength would be provided by the aluminum, and then milling that to the shape you want.

 

Thanks for copying and comparing it with the ideal 0009 section in XFOIL.

I began the discussion a few posts before just because I had found the Windknife 0009-specified profile. Then I followed your lead about a 5-digit like 63-0xx would work better and found that 63A011 was very close in the nose(needs a tad sharper tip) and also closest at the tail. And of course the trailing edge need a little final touch.

Waiting for a sample to see what inserting a wooden block to move max depth rearwards would do. Best of all would be if it didn´t need filler at all after that. But that is surely hoping to much ;-) .

Will anyway do like you did and copy the resulting section into JAVA-Foil to compare with 63A011 and 0009.