Collapsible Flettner Rotor Project

You have suggested some excellent ideas. The wheel hub from a car was one of yours, I just downsized to trailer hub. You were the first to suggest the end plate be a belt driven pulley, and some good critique as well. Your input is appreciated. I'm just not yet resigned to a non collapsing rotor.

 

Alan, if you think the advertising inflatable is workable, I think I can afford one. Almost everything in Mexico is inexpensive with one NOTABLE exception; my WIFE!

 

Yes, i believe the advertising balloon will work, but some caveats so we understand the limitations. Im not trying to back out, merely make sure all difficulties we can think off are covered before construction. There will be plenty we didn't think of for later.

It will be held in column by vertical tension, not air pressure, so there needs to be a fabric flange around the circumference at each end. This is to hold/attach the centering plywood disk, an essential part of thr system. I hope one at least of these is already installed, but bonding alone may not be enough at the load i envision. i.e. they will need stitching i think, and this makes air proofing an issue. We can discuss this further when i see an image of the balloons configuration.

In my mind the scaffolding is the least attractive, but probably the best from a cost effective POV. The horizontal braces stabilize the otherwise too small in diameter poles. To do this without such braces, you might need 8” diameter, 1/4” wall thickness equivalent mast sections. Terminations at each end could be oak, or other tough wood.

Brace the poles/masts any way you like, but be aware you need a 6-7’ diameter disk about flush with the top of the advertising tube. The disk at the top of the tube should either run just inside this ‘end plate’ or just below it. Gap here is bad, but how bad, I'm not sure.

The downward pressure on the pole/mast bases, and the upward pressure on the spindle base will be large and opposed. I hope your deck molding can stand the loads, and not buckle upward. This will be the case for almost any mast, though placing the mast heels on the gunwale helps immensely. Use 4” X 4”’s to brace the fixed part of the hub to as near to the gunwale as you can get.

Use a 4:1 or 6:1 pully purchase on the 'halyard' to get enough tension on the rotor.

390 RPM is a very high speed, think of a large marine propeller at this speed in air. Are you sure its not 39 RPM with the decimal point moved? this is more like the RPM’s we worked with, and they were bad enough.

Im concerned that any such column, hard sided, or soft sided, will spin in column at this speed.

Adding vertical battens in the advertising tube would stiffen it quite a bit, but prevent getting it down in a hurry.

I am quite amazed the sail cloth tube would be so expensive, i.e. more so than the advertising balloon, though equally i would have thought the advertising balloon might be subsidized by the beer company. Perhaps it is.

 

Translates NOT expensive. or IS affordable.

 

No image turned up, but i hope i can guess your intent.

Yes, a series of plywood disks, concentric around a single pole would work extremely well. i do not know how we missed this option.

Make the disks about 1m diameter, and 1m apart, including one on the top, and one on the bottom. Make the cloth between them a cylinder, of low stretch, air-proof cloth, Blue/silver tarp, sailcloth, or etc. Attach the cloth to the circumference of each disk using tape lashings on the inside through holes in the rim of the disk, or by screwing a metal strip (galvanized steel strip with continuous holes in it?) to hold the sail cloth to the rim. http://www.homedepot.com/p/Simpson-Strong-Tie-ST2115-20-Gauge-16-5-16-in-Strap-Tie-ST2115/100375249

Make the tube and disks exactly the same size, so there is no ‘dishing’ or hungry dog effect between disks. Tension between bottom and top will keep it taught.

You could almost use your spinnaker pole as the mast, but spinnaker poles are not typically designed to take lateral loads, and wear from the disks might be an issue (but see below). Make a ‘bushing’ from 1” stock both sides around the hole in each plywood disk, a good fit on the pole, and use nylon strapping as a wear surface. Two pot paint might almost be enough for a 'tempory' bushing?

Brace the single mast with 3 or 4 guys, to taste. I suggest the upper end plate be fixed, and the guys go though slots or holes in it. Spreading the guys and missing the upper end-plate so it can rotate would require an inordinately long mast.

You could have the spinnaker pole spinning, sit it on top of the trailer hub. In this case, you do not need the nylon surfaces or the bushings. This will require some thought to having a rotating bearing for the endplate and stays though.

Sounds like your easiest solution so far.

Check the spin speed, it still sounds too fast.

 

60 kts is 1nm per min. 20 kts, a nm every 3 min. If the rotor needs to have a circumference surface speed twice the 20kt wind, which is what I've been told, then a point on surface of rotor needs to travel TWO nautical miles in 3 min. 12,000 feet, or 4000 ft per min.
The circumference of a 1 metre dia cylinder is a tad over 10 ft, so I'll use 10 ft, a good round number.
A 10 ft circumference to spin 4000 ft per min, is 400 RPM.
Daiquiri figured more accurately and got 390 RPM.
That's the truth of it. Is that RPM impossible or dangerous?

The post with the drawing was #135 on page 9 0f this thread (at least on MY puter)

 

because its too simplistic and wont work for long ?

I would bet you are going to need a more precise method than this, involving more than wood to nylon sloppy tolerances.

I dont think you are going to get sufficient tension to hold the whole thing together for more than 5 minutes. I suspect driving the thing and expecting it to spin up uniformly along its length will be difficult.

Best to build a land based test rig to try it before you wreck your boat.

you keep saying that but it is correct. Do you want to try and get the Rotor Calculator to work on your computer like I offered ?

 

That's ANOTHER of your good ideas, you suggested it in earlier posts as well. Please stay engaged. Keep me straight and away from the deep end!

 

From http://www.boatdesign.net/forums/pr...ettner-rotor-project-50587-5.html#post691307:, posted on June 8th:

I didn't have enough time till now to check that spreadsheet. I will try to do it in the next days. The biggest issue seems to be the fact that no corrections for the aspect ratio (AR) are performed. The test data from Reid refer to the case of "quasi-inifinite" AR, where tested cylinder was extending all the way to the walls of the wind tunnel. Hence the tip vortex was not included into the measurements, thus giving a rather high Cl. So, in order to apply this data to a finite-AR cylinder, a corrective factor has to be applied to both Cl and Cd values before calculating the overall lift and drag of the cylinder.

I will see if I can find around some reliable and sufficiently accurate formula for this corrective factor.

Cheers

 

Thank you for the sanity check, this is exactly the sort of ‘rough’ reality check i was looking for. One of these days i will learn to write short replies.

Thank you Rwatson, i accept the result of the spreadsheet, just wanted a check. Nylon in any of its forms would be a disaster as a covering. This fabric must be very low stretch, like ‘blue’ tarps, or sailcloth.

Thank you Daiquiri. I do not think calculating end losses will help us. It will only make the numbers look worse. The end plate is a reasonably effective tip device, and in this case, almost the only practical one. We could add another 1m to the height, but that would increase heel angle, increase ‘whorling’ and some other issues to little improvement.

Yobarnical; There is no doubt this system being discussed will work, and be safe too, but there might be other reasons to review the whole project, see below. We are just a bit more restricted in what options we can use. First the pole/mast MUST spin with the rotor. The rotor will also need to accelerate to speed over perhaps 30s-60s, so no sudden changes of direction please.

I do not think the spinnaker pole will work properly. The bottom on the pole must fit around (or inside) a solid wood (oak perhaps) plug that it fits fairly exactly. This plug will be top-center of the lower disk. All disks above this must have the wood bushing, as we do not want them taking an angle against the central pole. It might crush the pole, a structural hazard. As the disks all need to fit the pole fairly neatly, it needs to be constant diameter, and not tapered.

You need each central hole in each disk to be a sliding fit, they should slide vertically for reefing. Though technically you could take the ‘fall’ of the halyard down the center of the pole/mast, and rotate with it, i suspect it might be a lot easier to take this fall down along a stay, and have it NOT rotate with the rotor, i.e. be above the upper bearing. Cable strap a piece of plastic conduit to a stay to hold the halyard fall away from the rotor etc.

Tying the skin to the disks is no longer viable, you will have to use the metal strips screwed on. I think screwing into the edge of plywood demands special screws and technique too.

Now practicality; You need at least a 1m diameter X 5m high rotor @ 390 RPM for 20kt winds. Anything less would tend to make the experiment too weak to validate the experiment. Racing, at 20kt wind, we are reefed, and very alert. At sea we are taking spray every wave, though on the lake (Washington) it might be smooth enough. This is not ‘cruising!!! The surface area of the rotor, 1m X Pie (3) X 5m equals 15 sq/m equals 140sq/ft, not far off the reefed sail appropriate to your 25’ boat in 20kts of wind. This is hardly super efficient, and does not ‘look right’ compared to more modern ships with rotors. I have to agree with the calculations, just wonder a bit under what conditions the ships rotors were supposed to contribute?

Now you have a rotor rotating at ~390 RPM just above you cabin top. The booming and rumbling (at 390 RPM, and harmonics) will be substantial and intense, exacerbated by the drum like effect of the fiberglass cabin top, and enclosed space of the cabin. Not acoustically comfortable at all. Defiantly not a comforting or relaxing form of propulsion.

We still do not have a very good handle on the drive system, though a “V” belt drive has been described, and a chain could also be utilized provided a large enough secondary (driven) sprocket can be found. Friction (small wheel) against the perimeter of the lower disk, pressure toward the pole/mast might work, but maintaining the appropriate pressure over time might prove an issue.

Mounting this rig on say a trailer, whilst a good idea in principle, will require a very big trailer to get appropriate staying angle for the pole/mast. Equally, the rolling resistance of a typical trailer is such that getting any kind of thrust angle measurements will be difficult.
A possible test base could be some plastic dock floats, foam or hollow, strapped temporally into a square, and free to travel in any direction. It would be wide enough, and stable enough to support the rig, but stiffness under the rigging loads will be an issue.

 

Once again, the input from all participants is pertinent, valuable and appreciated.
Both Alan and Rwatson posted concerns regarding the stability of the disks sliding on the center spit. In addition Rwatson has doubts that all disks would accelerate together and that the concept is over simplistic. I agree. I ALSO have a possible solution.

But first, the screwing into end grain plywood also has a solution, a simple one. Dowels. Anytime you need to screw in end grain, bore a perpendicular hole in the board a little back from the edge and inline with the intended screw. Glue a dowel section in the hole. Now the screw is biting cross grain in the dowel. See figure bottom.

Stability of internal disks and acceleration together:
If the center spit was assisted by 3 more tubes, angled and the tubes were made of incremental larger tubes partially inserted in each other and welded, so there are steps and the disks had keyslots in the center. Each higher disk had a larger keyslot, so it rose higher before reaching it's allotted "STEP". I think this would stabilize AND keep RPMs together those pesky disks.
Sur, it's a bit more complicated, but still within the skill level of a wood butcher like me. See figure top, I exaggerated the sizes for clarity.

 

The Albin requires very little power to make 4 kts. so whatever the wind speed actually is (above 6 kts) provides SOME propulsion and the speed attained is best determined through experimentation.
As to noise level, I will probably use the rotor ONLY on the pushed boat. We'll be aft in the pushing boat. The rotor is easily remotely controlled where sails are not. That was in the original prospectus.
I will be motorsailing almost always when underway. I want a bit of help from the other boat and the noise can be as remote as the control.

 

Regarding 390 rpm figure... Make it 400 for the ease of calculations. We are talking about approximate calcs based on approximate test data / spreadsheets, which are non-corrected for scale effects. I wouldn't be surprized if the full-size rotor showed +/- 25% to 30% difference respect to the calcs.
And, besides all that, you won't easily find an off-the-shelf motor/gearbox combination which will give you exactly 390 rpm. So 400 rpm is as valid a number as 390 is, for your purposes. And is also easier to memorize.
Cheers

 

I intuited as much. thanks for confirmation.

 

The stepped tubes are necessary because the semi-cone angle is less than 7 degrees. Any semi-cone angle 7 degrees or less (but greater than 0 )is a sticking taper, like a morse taper on a drill press. The step prevents the disk from jamming on the spit. Also 3 equidistant level stops force the disk perpendicular to the center shaft/spit.