From winged keel to bulb keel

Pitty that they didn't include a wing keel! Results confirm in some way the Mikko's CFD investigations. Unfortunately the bulb profile I used is not optimized as Paul B already mentioned. My first intention was to improve the keel perfomance with few changes even not knowing the wing keel theoretical performance.
After all the simulations I am convinced that The difficults to machine the keel and attache an appendix will be as costly as replacing it by a new one. Not to mention the risks involved on such a change.
The Master Thesis confirms a second finding of Mikkos study that the keel performance is tied up to the boat's hull design. Of course I don't have the means to test all type of keels in combination to my current hull so the idea was to build a model and make some experiments.
For the time being I plan to test the current wing keel and a L keel with a shape similar to J/111 which is actually also similar to the one investigated in the Thesis. Optimal draft seems to be 2,5 m but the shallow places I visit are forcing me to shorten it to 2 m.
As soon as I get some reliable results I ill pot it here for discussions.
Hope you follow my experience and keep contributing with comments.
Wayne

 

Running through the the references of the above thesis I have found another interesting material called "Shape optimisation on an integrated bulb-keel" from Kasper Ljungqvist also from Chalmers. I suggest you to have a looK ...

 

From the SLA 3D print I made a silicon rubber mold and poured a SnPb alloy with a similar density to iron. Weight matches to the expected scale although surface presents lots of bubbles. As soon as I define the new keel shapes I will print new protos and cast them with SnPb and Pb alloys. The model hull has a slot on the keel position that alllows to replace the keels. I expect to be able to test different keels in my "home made" tank both for hydrostatics and eventually L/D. In the meantime I follow reading about L and T-keels ...

 

What is the LWL of your model?
Looks like something ~0.6m.
If so, Re around foils will be so much out (in the order of 10'000 instead of 1000'000), that not much meaningfull results could be expected.

 

It's 1:20 scaled so lwl is around 0,6m ... What you thing shoulb be a minimun scale?

 

I will open a can of worms here .
LWL ~1.5m I find out to be an absolute minimum for qualitative similarity to full scale for free sailing model, but with increased chords and areas of underwater foils.
To keep the foils to scale, and at somewhat reasonable Re, something in the order of 3-4m LWL is necessary ... .
I know it is expensive and impractical.

 

To keep the foils to scale, and at somewhat reasonable Re, something in the order of 3-4m LWL is necessary ... .
I know it is expensive and impractical.[/QUOTE]

The target is not to obtain absolute measurements to validate the CFD results. Based on the calculated results from the actual wing keel the idea is to test the wing and L keels under the same conditions and compare the behaviour. From the L keel with lower CG and less wetted surface I expected significant improvements so don't you think would be worthwhile to carry out those tests? Will the lack of accuracy lead me to wrong conclusions? I am planning to test only few combinations of speed, leeway and heeling angle ...

 

The point is, at such small scale, wing speed trough the water will be so small, that fluid behaviour around it will be very remotely similar to that of full-size boat.
Due to the very nature of low Re flows (comparing X0'000 to X'000'000):
*about half the max CL
*about the half the stall angle
*X times less L/D achievable
*slow to reattach once separated
there is a risk your foils will be stalled most of the time.

 

*about half the max CL
*about the half the stall angle
*X times less L/D achievable
*slow to reattach once separated
there is a risk your foils will be stalled most of the time.[/QUOTE]

Understood! Let's see what happens when I test them ... You look very experienced in testing models so could you give me some hints as how to build a small towing tank? I am considering to build a 1x1 m section x 3m long box and use a pump and valves I have. Using the valve as by-pass I will control the flow trying to limit to equivalent 7 knots. What is still not clear is how should I sustain the boat and measure the forces. Can you give m some hints? Pictures or drawings would be great.

 

I have some experience in building free sailing scale models and in sailboat propulsion assessment in general. Like writing rudimentary VPP.
At one occasion, I have lost one model for sole reason: improper foils, and that prompted some investigation in the subject.
As to tank testing, I am one of the guys who see the results for particular project, and are curious enough to investigate how they are made, what are the limitations, what to expect and what to not expect.

In order to get meaningfull results, you will need the flow to be steady, free of eddies, etc., etc., ... . Neither of it is easy to achieve or cost effective in perspective to replace on own (not mega-) yacht.

In this case, towing the model in open water from suitable platform (from the bows of beach cat with small outboard?) could prove more economical.

BUT. Sailboat model towing tests are extremely time consuming:
you need to tow at 1deg of leeway and 5, 10, 15, 20, 25 degrees of heel at say, 5 different speeds, than repeat it for the 2deg of leeway, than for 3; repeat all of it for 2 or 3 displacements; plus some repeated runs to check the consistency and repeatability.
Do all the same for the next configuration. And for next.
Than all this matrix of data points is fed in to VPP where real evaluation of underwater or whatever mods could be really done.
Because you never know beforehand what particular combinations of speed/heel/leeway will be encountered in real sailing.

For all of this, if CFD like Mikko's and/or professional towing tank are not available, and America's Cup accuracy is not required, I would opt for free-sailing model, big enough to pull her foils through the water at reasonable Reynolds number.
Than either two otherwise identical models could be sailed together, or keels swapped on the single model. Results could be measured in variety of ways, but for single model, Mk1 eyeball and general impression are not so bad tools in the end.

 

Ok, vakuable advices, thanks. Looks like I will spend the rest of the summer dragginh my little boat... Actually in Brazil is not so difficult to find a small bay without wind and waves in the first hours of the day so towing the model in a very low speed should not be a problem. I am working on the new keel shape and expect to have a 3d print in 2 weeks so my first attempt will be to tow the 2 keels in a very low speed like 0,5 knots. Unfortunately still not sure how to design the system to fasten the hull and produce the heeling and leeway angle. Have a hint?

 

Consult a "Principles of yacht design" of Larsson and Eliasson.
There is a section about towing tanks and carriages.
(as stated by respected yacht designer elsewhere in this or other similar thread "... for small boats, it could be almost as cheap to experiment full size, as to run proper engineering investigations ..." or something like this)

Good luck with your efforts!

 

Consult a "Principles of yacht design" of Larsson and Eliasson.
There is a section about towing tanks and carriages.

The book is already ordered at Amazon but it´s a long way to Brazil ...
In the mean time I am finishing the new keel.

How do you like the changes? Think it´s going to work?

 

Mikko is in much better position to make a judgement on this.
Generally, I like this new keel. For details and accurate numbers -accurate tools are necessary.

 

Your new keel looks great to me.