Foil Cavitation at Lower Speeds Than Expected

The new idea has another advantage, it makes it possible to fold the foil under the beam, when leaving the beach. Have an idea of using stainless universal joints to make it work.

Finding a way to adjust foil angle while sailing is on the list.

It will anyway be interesting to build a variation on the theme, as building a Broomstick clone would not contribute to development

Have at last found a source for thinner tubes in Sweden. Though not as thin as Broomsticks, 4”x0.065” and 3”x0.058. Prototype rig will be made of an A-class main and at first a 70x2,0mm round tube as mast using over rotation, together with Dyneema standing rigging.

About the foils, the RC-shop where I buy G10, suggested laminating 1+0.8mm to make the top of the foil less flexible, while using 1.0 at the bottom where the shape is more complicated with double bends.

About the amas, my idea is to make the main beam a little wider and place the amas(made of a split in two windsurfingboard) outside the foils. Eventually I will try to make them pivoting, to a certain degree, around the main beam axis.

 

I'll be interested to see what you come up with.

One other thing I found useful was to attach the amas with bungees (or at least lashings) so they could be removed quickly when turtled. Otherwise, it was impossible to recover without help.

 

Thanks for your interest and enthusiasm . As this is a little beyond this threads intensions, is it OK to pm you? Or shall I start a new thread with my preliminar plans? Anyway, the initial plans are using an A-class main, round tube(70x2mm) over rotated mast, together with the foils based on the www.windknife.com profiles we have discussed. Also have a 15sqm flat cut assymetric if needed ;-) . By the way, received the FF13 0009 today and immediately thougth it would need to be widened, to work as lifting foil on the rudder, 0010 and 0011 seems to be possible.

 

I think you've got enough interesting material to have your own thread, either here or in the Multihulls category. But I don't mind if you continue posting in this thread, or using PM's. It's up to you.

 

Thanks, have to make some drawings before starting a new thread in Multihulls. Will continue the foil discussion here, nice to have that part in one place.

 

About the 0009 strut it might be best to place it so it is above water level when foiling from ca 15 knots to high speed.

The projected area to start foiling at center hull bottom level, is according to my ballpark calculation, the one you have chosen with Cl=0,5, to start foiling at 12 knots of speed.

At 20 knots the foils should be 22cm below surface with the same ballpark calculation.

Are these figures far from your IRL figures?

 

Thanks for the warning, this speaks in favor of placing the amas outside of the foils, as my intention is.

The big problem in the lake I sail, is that it´s not very deep, so a complete turtle isn´t possible. The beam will even be close to touch bottom when rightning at some places.

 

There are several issues with analyzing a mast+sail section. The first is knowing what the sail shape is. You may be able to just assume a shape and get the difference the mast makes.

The next issue is the thin sail, which causes numerical issues. I just increase the thickness of the sail to, say, 0.2 % chord.

The biggest problem is the separated flow behind the mast. Javafoil might give you a result, but it is not capable of handling any separated flow, so the answer will not be accurate. Xfoil can handle modest amounts of separation, but the round mast is quite a stretch for it.

 

Tom, rounded the problem by realizing I had an old alu Sail Craft Tornado(70x135mm) in my backyard. Not perfect but probably a lot more aerodynamic than the round tube. Measured bend in both directions and found out, it had close to the same characteristics as the typical modern A-class carbon mast. Just ordered a used A-class sail from UK, to see if I can mate them! Big alu mast weight penalty though, 8kg!

Will anyway try to play around in JavaFoil by comparing round to T mast with sail at your recommende 0,2% chord.

 

My inititial idea of changing the Broomstick configuration to a 45 degree foil with a nonlifting 0009 strut is a waste of lifting area.

If we instead change the strut from 0009 to 0010, to have an 60 degree angle and also adding a 7,5 degree incidence, it would lower liftoff speed from 12 to 10 knots according to my ballpark calculations.

The main foil will have something like 1,5 degree incidence to get Cl=0,5. Ie the strut will thus have 6 degrees higher incidence than the mainfoil. On the downside is the mainfoil can not run above 3 degrees, where CL=0,7 as the the 0010 will stall above 9 degrees. Maybe the initial incidence of the lifting strut can be set lower if needed?

But would it work IRL? I guess strange things can happen where the strut and foil meet.

Anyway this will still be rather easy to build, using my method with Windknife Full Nose and G10 for the main foil and Windknife FF13 with an inserted 12mm alu tube for the strut.

Initial idea with Broomstick dihedral in black:




With lifting strut:

 

Yes, that looks right. You can take it a step further by adding lines of constant foil loading. Given the loading and speed, you can calculate the lift coefficient for a given foil loading (L/S): CL = (L/S)/(0.5*density*velocity^2). A fully submerged foil will operate along one of those lines. A surface piercing foil will behave differently, but if you have an estimate of how the area varies with speed, you can plot the locus of the design CL vs speed.

When you add the lines of constant loading, you'll see the cavitation envelope typically intersects a loading line at two places. One is the high speed/low lift intersection that corresponds to bubble cavitation well back on the foil. The other intersection is at low speed/high lift, corresponding to the takeoff condition. This intersection is due to the leading edge suction peak causing sheet cavitation. The operating regime of the section lies between these two intersections.

You can raise the high-speed incipient cavitation by making the section thinner, but that will also raise the takeoff speed. Shaping the section for both a low takeoff speed and a high cavitation speed, while providing enough thickness that the chord can be reduced to cut wetted area and still maintain structural integrity, is the balancing act for which Xfoil's inverse design modes are really useful.

 

You can get an initial idea of the feasibility by using a vortex lattice code like AVL. AVL will tell you what the local lift coefficients are and what the induced drag is for the combination. The interference effects can be counter-intuitive, so a program like this is really useful. I like to set up a spreadsheet to create input files for AVL. The spreadsheet makes it easy to change a few geometric parameters and ensure all the rest of the input to AVL is consistent.

However, AVL will not tell you what the local pressures are, so it can't be used to predict the incipient cavitation speed of the junction. For that, you need a panel method. CMARC from Aerologic is not free, but it's not expensive, either, compared to other aerodynamic codes. And the PSW comes with the pre- and post-processors (Loftsman and Postmarc) that you need to create models and analyze the results.

If you take the difference in the pressure distribution for a section in a 3D panel model compared to the pressure distribution from the 2D calculation, you get a 3D effect that you can subtract from the design pressures in the 2D inverse design method. That will allow you to design a distorted section that, when placed in the 3D context, will result in the desired cavitation speed. For example, near the apex of the V, you will probably need to flatten the inside contour and bulge out the outside contour compared to stations farther from the apex, resulting in negative camber in the middle of the section. This is to counter the acceleration of the flow on the inside of the V. The outside contours act more like an axisymmetric body, and the bulged bottom surface has lower velocities than you'd expect from a 2D calculation. It may take a few iterations of going between 3D and 2D to get just what you want, but the convergence is pretty quick.

 

Seems like my idea was unnecessarily complicated to realize, so I dump it.

The most complicated profile I manage to build is the untapered cambered 0012, so we will have to take it from there.

Will go back to the initial concept with the same profile on both parts. Have adjusted the angles to get more projected area at liftoff compared to Dougs Broomstick foils. Surely this is a compromise, but it is anyway buildable.

 

A bit of subject but has anyone seen this extra ordinairy video of
vortex bubbles propagating backwards on hydrofoil ( ventilation / cavitation)?

Gives a visual representation of how cavitation develops.
Interesting...

 

That's interesting, but I'm not seeing any backwards propagation. To me, it looks like the bubbles are created at the foil & are just being left behind as the foil goes forward.

Can you tell me specific frames (times) on the video that you think show backwards propagation?