Previous Page | Table of Contents | Next Page

Amateur Yacht Research Society, February 1997 Newsletter

Subject: Flettner rotors, (was Pumping)

From: Dave Culp Date: 26 November 1996 09:03

>Something like a Flettner type rotor will self rotate.

Peter: Thanks for the reply, but you're wrong (partially). A number of toys of this nature have been built, mostly of Savonnus cross section (cut a cylinder in half, lengthwise, and offset the halves one-half diameter. Screw on end caps and it will spin, and fly).

The problem is that a Flettner rotor needs to spin with a surface speed about 4-8 times the windspeed in order to realize high lift coef, and a Savonnus-type rotor can only deliver about 1.5 times windspeed. At this slow rotational speed, it's lift coeficient is lower than a corresponding ordinary airfoil, and yet it's drag is much greater. Also, the Flettner rotor wants to be a smooth cylinder (not an x, or s cross section) for minimal drag.

A nice oddity, but not for serious sailing.

AYRS members have proposed (several times, I believe) a pedal-powered Flettner rotored sailboat, which *would* make a good Winter project, but that needs it's own thread, don't you think?

From: FISHWICK Simon Date: 26 November 1996 09:49

One approach to analysing an autogyro is to consider one part of each blade as extracting energy from the airflow, and the rest of each blade as putting that energy back in, so changing the airflow and generating lift.

Maybe one can do the same thing with rotors and gearing. Use something like a Savonius rotor at one end of a cylinder to drive a Flettner rotor at the other, through a gearbox to handle the required rotational speed difference.

I don't _think_ it infringes the laws of thermodynamics, but I don't know either whether it can produce effective drive. Maybe it depends on finding an optimum distribution of areas.

From: Ian Clarke Date: 02 December 1996 12:35

Yes - A Savonius rotor made from a 200 litre drum will drive a water-well diaphragm pump quite well (through a cam linkage) - but of course a diaphragm pump likes a slow steady action anyway.

If you have a moment, please enlighten me about Mr Flettner's particular rotor?

From: FISHWICK Simon Date: 02 December 1996 13:25

Mr Flettner found that if you hold a spinning cylinder across an airstream it generates lift towards the side that moves downwind. The faster it spins, the greater the lift. In comparison with an aerofoil, which seldom generates a lift coefficient greater than 1.5 - 2.0, Flettner's rotors seem to be able to generate lift coefficients up to 4 or more. (OK it depends to some extent how you define the lift coefficient of a cylinder, but you get the general idea.) It also generates drag - quite a lot of it in comparison with a aerofoil.

I understand he built at least one boat that had a couple of cylinders sticking up, rather like very thick masts, that sailed quite well in comparison with the ships of his day. I have also heard that the principle was proposed for very short landing aircraft use. Whether that ever progressed beyond the model stage (if that far) I know not.

From: Dave Culp Date: 02 December 1996 21:01

>Maybe one can do the same thing with rotors and gearing. Use something like
>a Savonius rotor at one end of a cylinder to drive a Flettner rotor at the
>other, through a gearbox to handle the required rotational speed difference.

It would certainly be posssiible to mount the Savonnus rotor, of 1/3 the diameter of the Flettner rotor, *above* the Flettner. The 1.5 x windspeed rotational speed at the Savonnus would then become 4.5 x windspeed at the surface of the Flettner rotor. I expect that a rather large (read: long) Savonnus element would be needed, though, to generate the required torque.

From: Dave Culp Date: 02 December 1996 21:13

>>AYRS members have proposed (several times, I believe) a pedal-powered
>>Flettner rotored sailboat, which *would* make a good Winter project, but
>>that needs it's own thread, don't you think?

>No I don't! I'm helping a friend build a sailing cat with a solar-powered
>Flettner rotor, but it'll be a while yet...

Near as I can tell, this is perfectly legal, under WSSR (speedsailing) rules. It "bends" them quite a bit, but I don't think it's illegal. I doubt high speed was on your list of objectives, though, was it?

From: Claas VANDERLINDE Date: 03 December 1996 00:13

For what it's worth, here's a little quote from AYRS 87 by John Morwood:

"The Flettner rotor worked but has the disadvantage that its performance is hopeless in winds of less than 14mph (the fault of all thick wings on boats). This is due to _Greatly Increased Drag_ at low Reynolds Numbers. ... A yacht fitted with a Flettner rotor was a disaster. All the above applies to the Savonius rotor."

And for completeness sake, here's another quote, this time from Joe Norwood's "21st Century Sailing" (who did not specifically mention Savonius rotors):

"What about using a windmill to generate the needed power [for a Thom rotor, a more efficient Flettner rotor with added horizontal fences]? Using the windmill equations ..., we can show that the required windmill area AR, and the windmill drag is several times that of the rotor, Thus windmill power for rotors is not an attractive proposition."

Norwood does propose a pedal-driven proa with a Thom rotor that he predicts should reach sustained speeds of close to 15kn and maximum speeds pof 28kn on a beam reach. Any takers?

From: Dave Culp Date: 03 December 1996 07:41

Damn! Hoisted on my own petard (I *hate* it when that happens). I should have looked to AYRS pubs first, or at least considered efficiency, rather than just, "Can it be done?"

>Norwood _does_ propose a pedal-driven proa with a Thom rotor that
>he predicts should reach sustained speeds of close to 15kn and
>maximum speeds of 28kn on a beam reach. Any takers?

This has come around several times, over the years. It was, and is, an excellent idea. I'll bring the beer... ?


Response to Comments on Hagedoorn in AYRS 122

Roger Glencross, 34 Hollington Crescent, New Malden, Surrey, KT3 6RR


Dear Tony

With reference to your Conclusion to publication 122 (Ultimate Sailing IV), the reason why you have not received any critical review of Hagedoorn's original paper may be because such a critique would amount to speculation on a speculation. (I leave aside the unpleasant thought that it may be because there is no one out there!)

I agree with you on the aspect of Hagedoorn's paper which seem wrong but that does not lessen his genius. Even the Wright Brothers had blind spots. They were convinced that wing warping was the only way that an aeroplane could be controlled in a turn and that ailerons would never work. Hagedoorn's practical experiments were not as successful as his theory. If his experiments had followed the same path as Didier Costes he would have found the need for two hapa lines instead of one. Once he had accepted two (or more) hapa lines, the (falsely) apparent need for the hapa to be round-shaped would disappear. He believed that the hapa should be round so that it was symmetrical around an axis and free to turn around that axis so that it could never exert a couple around that axis. I believe that he despaired of any stabiliser being able to compensate for the turning couple without the stabiliser producing a prohibitive amount of drag. Having spent a long time trying to stabilise hapas efficiently without success I do not blame him his pessimism! Also, the round shape and single line were chosen in the context of an aquaviator jumping out of an aeroplane at the start of his hapa sailing. A single line could not tangle and a round hapa would always hit the water at a correct attitude.

Hagedoorn's paper improves as it progresses and the second half holds up better than the first half. He has plumbed the depths of the mathematics of hapa sailing and there is no point in regurgitating this definitive part of the work. Suffice to say that mathematics can only give an imperfect description of the infinite variables occurring in fluid flow. Only practical experimentation can add to our knowledge on this subject. Even if his figures prove to be 80% optimistic it will still be possible to hapa sail upwind (just).

That is why I do not agree with your hint that my last twenty years leisure activity has been a hollow mockery! I have never doubted that Hagedoorn's basic theory will one day be achieved. However, in my darker moments I sometimes think that the hapa may be superfluous! If one simply dragged a drogue anchor downwind, collapsed the drogue when the parafoil was at its zenith, glided upwind, and then redeployed the drogue when one had lost height, one might progress with greater ease that with a hapa.

Return to Table of Contents

Tri-Fly at Speedweek 1995-6

(Please Note: This Newsletter was published in February, 1997. Announcements, Dates and Times herein are appropriate to that time-frame, and may be completely inaccurate as of the time you are readingthis--Ed.)

Fred Ball

Tri-Fly was originally built by Tony Bloefeld as a tri-scafe with inclined foils on the forward floats and a "T" foil as the rudder: in this form he felt that he had developed her as far as possible and passed her to me as a starting point to develop a speed machine.

I felt that the first task was to assemble her and fit a simple rig and observe how the floats performed with a view to improving the floats and possibly using planning hulls rather than foils

The forward floats were fitted with centreboard slots to enable centreboards to be used or "J" shaped foils to be used (still to be made!).

The rig used was a simple sloop rig with a 3/4 height forestay, and a fully battened main set using a wishbone boom inclined downwards to provide outhaul and luff tension without use of a kicking strap; total sail area 9.5 square metres.

The main structure was around a central space from welded from large (2.5") aluminium tube with the central stern float firmly attached at its bow and mid length, the forward floats are on aluminium tube extensions stayed to the centre frame to reduce flex ion and torsional effects.

During the week a maximum of9.5 knots was recorded; the limiting factors being the lee bow burying aggravated by slack forestay necessitating over sheeting of the jib, this was observed by Ian Hannay and graphically shown in one of the photos which came to hand after the event.

The shape of the floats (deep inclined "V") was unable to develop lift; at the end of the week I added some extra buoyancy and a flatter bottom to the aft part of the starboard float and this helped improve the speed but introduced some obvious twisting in the longitudinal plane reducing the effect of the added planing surface.


I hope to improve the forestay tension possibly by using running back stays, to try more buoyant floats with a step allowing planing lift at a point avoiding bow burying, to improve the stern section to allow this to plane more readily and to rejuvenate the present forward floats with lifting foils.


No lifting foils but new floats; one twice as wide as the other and both with a step half way between the supporting beams, and a moderate "v" (5 degrees) and an angle of attack also 5 degrees.

In practice I found that although the step appeared to work under favourable circumstances, the angle of attack was probably excessive reducing bouyancy at the front of the float; when travelling fast the float bottom appeared to lie on the back of the wave induced by its movement, and was unable to climb onto the crest and permit planning.

The narrow float appeared to have enough bouyancy to support me when climbing over the boat and to permit the windward float to fly. The other alteration was a bowsprit and running backstays which did appear to improve the set of the jib. What happened.

Tuesday my first day afloat in the new form showed that more sail area was needed but I felt that she was moving better.

Wednesday there was even less wind so Margaret and I had a day as tourists!

Thursday there was plenty of wind and sailing out to the course she was going well* flying the windward hull at times until the clew of the jib blew out! A quick trip to the sailmaker in Hope Square soon had this repaired and was back on the water after lunch making the decision to be towed to near the start to prevent more breakages before a run. Almost instant disillusionment as I capsized her; too keen to get the windward hull up and ? the wind under the trampoline. As this was in shallow water she was easily righted and was sailing along the course but without the excitement and sensations of the morning: a best speed of 9.1 knots was recorded.

Friday morning I went to drain out the water from the floats and found that I could see the beach through the access hatch! the lack of performance was explained. A simple patch soon had her ready for sailing and some more runs were made in a 10-12 knot breeze so no real speed, tried various dagger plate positions and felt that their size was probably about right for the conditions.


I think that Ians thoughts about biplane rig are a logical way of increasing sail area without raising the centre of effort.

Using a T foil rudder should remove the need for the stern float (and its weight) probably reducing drag; this means that she will become a wide beam catamaran and hopefully the supporting structure can also be reduced with weight and aerodynamic drag improvements.

I plan on developing the rig aspects during the winter so that October 97 Tryfly is a practical and reasonably fast performer.

Then serious efforts will be needed to reduce the total weight so the hydrodynamic drag can be significantly reduced. While shape and planing area, aspect ratio, angle of attack etc are important reducing weight must be the simplest way until those design factors can be better understood. The present floats although constructed simply of 3mm ply weigh Kg and while I think my structural failure was related to a thump on the bottom when I righted her after capsize at least one of the cracks suggests some panel flexion was taking place.

Previous Page | Table of Contents | Next Page

The AYRS Newsletter is published quarterly by the Amateur Yacht Research Society Ltd, BCM AYRS, London WC1N 3XX, UK. The AYRS as a body is not however responsible for opinions expressed by individual authors.

This publication is copyright under the Berne Convention and the Universal Copyright Convention. All rights reserved. Apart from any copying under the UK Copyright, Designs and Patents Act 1988, Part I, Section 38, whereby a single copy of an article may be supplied under certain conditions for the purposes of research or private study, by a library of a class prescribed by the Copyright (Librarians and Archivists) Regulations 1989:SI1989/12, no part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means without the prior permission of the copyright owners.

Multiple copying of the content of this publication without permission is always illegal.

Authorisation from AYRS to photocopy items for personal use, or for the personal use of specific clients, is granted ONLY to libraries and other users registered with the Copyright Clearance Center (CCC) Transactional Reporting Service, CCC, 21 Congress St, Salem, MA 01970, USA. This authorisation does not extend to other kinds of copying, such as for resale, or to other unregistered organisations. Permission to copy articles for these and other purposes should be sought on each occasion from AYRS. Permission for non-commercial purposes will not normally be refused.