"I saw the new pics this morning. It must be exciting to see it that way at last. In one photo I can even see the "Kiribati dimple". Isn't it time for you to release this dirty little secret into the blogosphere?" Gary Dierking
O.K. Gary you asked for it!
I've been asked this question many times, over many years and after much practice, research and reading I think I'm ready to answer.
The so called Kiribati dimple is a feature of some of the variants of proa characteristics found around the Pacific past and present and is most pronounced in the proas of Kiribati.
In the late nineties I was fortunate enough to have been able to converse with several of the great aero-hydro dynamics experts of our time. In Auckland these were professor Peter Jackson and Tom Schnackenberg. These conversations lead to hours of reading in the engineering library at the University of Auckland. I was preparing research material for my thesis to contribute to an advanced degree in design.
Kuchemann was a German professor of aerodynamics invited to continue his research in the United States after the Second World War. During the 1950's he was experimenting with supersonic aircraft design for McConnell Douglas. One of his papers talked about delta wing flow dynamics. This was interesting enough to me however I then came across some experimental wind tunnel work on supersonic fuselage design. The "Coke bottle" profile of several jet fighter bombers solved the problem of shock wave development at the wing root where the wings joined the fuselage. This is an area of dramatic increase in fluid displacement. It was found that where this secondary drag inducing shock wave developed, the fuselage could be "waisted" or reduced in circumference to improve the overall aerodynamic efficiency of the airframe at sub and supersonic speed. It was explained to me that supersonic air has similar flow characteristics to water when compared to fluid flow over a hull.
I made the assumption that a hull with significant rocker and combined proximity to the widest section of the hull of an asymmetric proa (or any displacement sailboat hull for that matter) would benefit from a waist on the high pressure (or leeward) side in order to reduce drag.
I had already had my curiosity aroused on this subject by J.S. Taylor many years before when Mike and I were building Takapu. Taylor published a series of articles in the NZ Seaspray Magazine in which he wrote about his experiments on the hull of a proa with just such a dimple. I recall him relating his experiments which involved the filling and fairing of the dimple on his trial hull. The hull was slower and lifted less after the treatment than before!
Mike and I have since that time incorporated the dimple in our proa design work with good and satisfying results. Both Takapu and Toroa are faster than their predicted maximum hull speed, based on prismatic coefficient calculations, by a significant factor. Furthermore we have found that our hulls are not as susceptible to broaching when sailing down wind in a following sea.
Of course the artists right side of my brain also has a say. Intuitively I know that in nature pure symmetry is anathema in living organisms. Where symmetry appears to present itself there, on closer detailed inspection (on the other side of the coin), you will find variation.
Both the United States Navy and Jaques Cousteau carried out experiments on the form drag of fish and marine mammals. When towed (dead) at their known respective speeds, the from drag of their streamlined bodies was far greater than their musculature could possibly propel them.
I believe that we have been looking at the dynamics wrongly.
No fish dolphin, shark or seal is ever symmetrical when swimming at speed. They are always at any given moment either asymmetric one way, or the other. Whilst asymmetric at the extreme extent of flex the body is curved on one side and hollow on the other (in the case of a dolphin the same is true though of course the asymmetry is in the horizontal plane). Whilst in the fully curved stage of motion, at speed the lift generated is slightly forward of lateral in a similar way that a sail generates lift. The body is inclined to generate a zone of low pressure on each oscillation which reduces drag and effort.
I believe that the canoe builders of Oceania either observed this phenomenon or generations of builders stumbled on the benefits of asymmetry thousands of years ago. Either way the fact remains that it works and it works very well. I would not consider building a shunting proa without this feature when there is so much to gain and so little to loose.
Harmen R Hielkema.
Fascinating! I'm glad Gary ratted you out on this dimple ;-) I would not have noticed it on the photo, or if I had, I would have disbelieved my eyes and thought that the dimple was an optical illusion. Interesting that the little concepts I learned when I was a technical writer for Pratt & Whitney writing about turbines and rocket engines is often applicable to boats -- but of course, incompressible water is always "in shock". -- Wade
ReplyDeleteHarmen, thanks for the lesson! I remember reading a cryptic reference to the dimple in an old J.S. Taylor article, and was always curious about it.
ReplyDeleteCheers,
Michael
Been reading up on rudders, and just saw the 'Schilling rudder'. It has a hollow mid section and flared tip (in profile). Is this the same principle I wonder? And the 'Gurney flap'. Both are said to increase lift - thats sounds useful!
ReplyDeleteI also look at Garys latest blog post: Austonesian Sail Types and wonder if the various 'crab claw' sails get some Gurney effect from the spar at the trailing edge, especially say type c?
Hello Harmen,
ReplyDeleteI've been looking at the Kiribati hull shape on the computer lately, for a new project. When you draw the shape without the dimple and look at the area curve (rather than the hull lines), you see a very quick turnover at the centerline from increasing area (front half) to decreasing area.
This is a consequence of the "long V" profile of the Kiribati-style underbody.
There is considerable drag involved in making such a quick transition, because you are forcing the water to reverse lateral and flow and then accelerate in the opposite direction. The sharper the change in direction of the area curve, the higher the acceleration.
On the lee side (high pressure side) of the Kiribati-type underbody, the effect is to force water to flow downwards along the hull, then under the keel- exactly what you don't want.
When you use the dimple, this effect is much smaller. The area curve transition is much flatter, so lateral acceleration of the water through the CL transition is greatly reduced, with a consequent reduction in drag, and the waterflow loses much of its vertical component. It stands to reason that better lift to windward would result from less water flow across the keel- which is why the dimple goes on the lee side.
This is pretty much the same thing that the "coke-bottle" shape accomplishes for jet aircraft. The effect of the volume of the wings on airflow along the fuselage is to create a similar quick rise and turnover in the effective volume curve of the aircraft, creating a premature shock wave at transonic speeds.
John Dalziel
Very impressive discussion. What's the right amount of dimple? Is there a guideline rule of thumb % based on the mid point lee hull offset or some other dimension?
ReplyDeletetimpeever123@yahoo.com
We bought a fat catboat with no centerboard. The seller assured us that a taught line from the bow to a stick held below the transom would provide ample lift.
ReplyDeleteWe were too gulible.
Wish you would gove us more objective info.
Hopefully,
Hi Illwind
ReplyDeleteSorry you didn't leave me an email address or a name to reply to.
Your catboat would be fine with an asymmetric, pivoting lee board on each side if you want a shallow draft foil.
look at the traditional sail boats of the Netherlands for ideas. Lee boards will be relatively easy to build and retrofit.
I always admired your catboat designs they remain a favourite of mine.
Please don't give up too easily.
Hope you read this message.
Best
Harmen
Hi Guys:
ReplyDeletevery interesting post, indeed. I wonder how you would incorporate the dimple in a stripe planked canoe like Garys T2 and if the dimple would also be effective on the Amas leeside?
Any help?
Thank you very much for this highly interesting blog.
Best
Max
max@der-reporter.de