Author Topic: Felonius Savonius (Read 4505 times)

disaray1 · « **on:** September 05, 2007, 07:10:34 PM »

I went to the county landfill and found a couple of nice new 55gal drums...sitting beside the big sign that reads "NO SCAVENGING, By County Ordinance". So I took em. heh heh.

They have now been cut in half, and I'm ready to weld them. I have studied this document

(WARNING! Very large PDF, 105 pages)

http://www.prod.sandia.gov/cgi-bin/techlib/access-control.pl/1976/760131.pdf

but have questions. The analysis concludes that best performance is achieved with a gap (s/d) of 0.1m to 0.15m, with the drum diam. of 1m.

Question: With a 22 inch diameter drum, would the s/d be 2.2in to 3.3in? I've seen other reference (not from Sandia) showing gaps up to 1/2 of the diameter of the drum, and others making a complete "S", no gap. 2.2"-3.3" sounds and looks awful skinny to me, I think it must need more flow between the halves. Can this be correct? I only want to weld this once.

Question: Lots of references to the Reynolds # in that doc, but I don't exactly know what to make of it. I understand that Reynolds is a ratio of viscous over laminar, however I'm a bit in the ruff on what it really effects and how the # is determined. Can someone either give me a laymans explanation on Reynolds#, or point me to reference material that's easy to follow?

This will be a two drum stack unit, one rotated by 90deg to the other. Thanks in advance for input and any suggestions are welcome.

David

SparWeb · « **Reply #1 on:** September 05, 2007, 01:31:18 PM »

I've seen the report, myself. If I were you, David, I wouldn't worry about the details. The Reynolds number difference between a 1 meter test article and a 22 inch drum diameter is not very important. Scaling things up or down by a factor of ten or so might make a noticeable difference. Besides, what are you going to do about it? Get a 82.7 gallon drum instead? :^)

The gap you measure when you scale the model is fine. You don't need a lot of flow "between" the halves. There is a point in each revolution where the airflow may still be attached to the first semi-circle and cross the gap to the next semi-circle. Three inches sounds like a good gap, too. If you compare the results on the charts, by laying all the curves together on one graph, you would probably realize that the data has a lot of scatter.

Your biggest problem is ensuring torsional stiffness of your structure. That, and even a savonius should be checked for balance. Will this big Savonius be supported at the top, or will it all be cantilevered at the bottom?

disaray1 · « **Reply #2 on:** September 05, 2007, 02:25:15 PM »

Hi Steve,

Thanks for the quick reply. I did consider the scaling differences and after some head scratching, asked myself the same "what are ya gonna do about it?" - I'll work with what I have. If three inches~ sounds about right, I'll go with it.

Yep, balance is on the mental to do list. I dont think I'll free stand it, and a upper bearing support along with a spreader to attach four guys will be necessary to keep things under control. I've seen a couple of others that have support structures built on the outside, but this look to me like it would cause considerable blockage, where 3/16 guys would be relatively invisible to the wind. Sound reasonable?

Rgds,

David

SparWeb · « **Reply #3 on:** September 05, 2007, 04:52:50 PM »

I've seen photos of wooden frames around these oil-drum rotors. They probably work very well. Dollar-for-dollar, it might make the most sense.

The energy from the wind is free - we get greedy and try to get that last 5%. If it comes at an increased cost it may or may not be worth it. At various times I've decided either to go for broke, or to rig up something cheap and replace it later.

A case in point was the tail vane on my first wind mill. I didn't know how big was too big or too small, so I just made something that looked right, as quickly and cheaply as possible. Twenty minutes with a scrap piece of particle-board. After 2 days, (and considerable teasing when I posted pictures), I tossed the board in the scrap and made a new one, because only after flying the mill could I tell that it needed to be bigger.

Good luck.

RP · « **Reply #4 on:** September 05, 2007, 08:43:35 PM »

I used that document a few years ago to build my savonius. It's less than half the size of yours but it was a fun project. About 10-15% overlap will be just fine for the size you're building.

One suggestion would be to build each pair independently and balance each "layer" before you stack them with the 90° offset. The unit will be tall enough for dynamic balance to be an issue.

Also, if you can, put a disc between the upper and lower sections. This will prevent "spillage" of air past the scoops.

RP · « **Reply #5 on:** September 05, 2007, 09:47:43 PM »

Whoops, actually add a disc at the top and bottom also.

disaray1 · « **Reply #6 on:** September 06, 2007, 10:37:11 AM »

Thanks RP for the advice- I'll look for some discs. I'd like to try to build this contraption completely from garbage/scrap. The search continues for bits. The alternator will be a different story, unless I happen upon some good magnets, somehow, somewhere...

David

vawtman · « **Reply #7 on:** September 07, 2007, 07:02:37 PM »

We are talking about landfill barrels and having fun right.

Why bring in the whitecoats for this?

TomW · « **Reply #8 on:** September 07, 2007, 10:34:47 PM »

wawtman;

Why bring in the whitecoats for this?

Mostly because they look cool with a pocket protector and a cluster of sharpies. =:-)

Cheers.

TomW

hiker · « **Reply #9 on:** September 08, 2007, 01:10:56 AM »

try a stock car alt--should work with a flat belt wraped around a plywood disk pully.

should have tons of tork with that rig--so try as big of a plywood disk as possible.

or even a cheap tread mill motor for some low watts-100-200?..any more would most likly

burn out the brushes...check my diary-pedgens-or-pedgen stock car alt..

disaray1 · « **Reply #10 on:** September 08, 2007, 05:00:58 AM »

Hiker- I found a Baldor 1/3hp PM dc motor, 0-90v....looks to be brand new...at the same landfill. It makes 8-10v with just a finger spin. I might just save this one for another project. I do have a couple of Delco alts laying around....

Thx

David

disaray1 · « **Reply #11 on:** September 08, 2007, 05:20:12 AM »

Mark- The whitecoats at Sandia spent many hours testing this design(s) in the tunnel. Why not start with a optimized design and work from there? R&D already consumes too much of my time. Besides, I dont mind the whitecoats, unless they're coming to take me away. Ha Ha. Too the funny farm, where life is beautiful.

David

vawtman · « **Reply #12 on:** September 08, 2007, 08:43:43 AM »

David,

Why not bolt the barrels so you can change the angle of attack to suit your liking?

Who knows you might come up with a better design.

In the tunnel the airflow doesnt swirl like in the real world.

However this comes from guy whos been in la la land a few times.I wonder how they can see through those glasses.

Back to the wheel.

feral air · « **Reply #13 on:** September 09, 2007, 01:05:53 PM »

From what I've read an 18% overlap is supposed to be the best. Being a tad off in either direction shouldn't make much difference though...

The gap on my plastic barrel vawt is about 6 and a half inches and that seems to spin up well. I think it's the same dimensions, give or take a smidge, as a standard steel barrel. The way the plastic barrel was made determined the gap for me...not math...I just put the halves together in the most convenient way.

If you plan on having a center rod run through it then maybe an 8" gap would be better...?

I'd say more than 4 inches but less than 10. Eye-balling it seems to work well enough for me that wasting time on exact measurements isn't worth it.

Felonius...too funny!

electrondady1 · « **Reply #14 on:** September 09, 2007, 06:51:22 PM »

it's important to remember that by increasing the overlap

one is also reducing the swept area.

sandia tested at between 10% and 15%

but i have also heard of as much as 20%

feral air · « **Reply #15 on:** September 09, 2007, 09:10:04 PM »

From the report;

"A very simple means of fabricating a Savonius rotor is to cut an oil drum in

half, invert one half, and weld the two pieces together. For this type of configuration

where the bucket size is fixed, it would be desirable to know what gap width will pro-

duce maximum power, P = QSI, and torque, Q. "

Translation; We don't want to come out and just state what the best gap would be, that'd be too easy. If you can't follow our numbers and/or you fall asleep trying, too bad.

Yeah, it would be desirable to know. Grrr.

I can't find where I got the 18% number from but I think it was in harvard's abstract site...thing. Here's a link to a list of related abstracts.

One thing I found there that I hadn't seen before;

"High aspect ratios are favored for high wind velocity regions, while low aspect ratios are preferable in regions with low winds."

That's good to know I guess.

thefinis · « **Reply #16 on:** September 13, 2007, 08:19:59 PM »

On the one I built I think it used a 4 in overlap. I used 42 in disks cut from plywood and the sections bolted on leaving extra space on the edges of the disk. L clips would have been better than drilling holes and running bolts through the barrel tops, bottoms and plywood disks. Caution the bolts need spacers between the wood disks and any gaps to the metal barrel halves otherwise it will warp and want to spin off center. The one with an overlap spun better than the later version I built that used halves with the edges touching with no overlap. Your results may vary and I never tested power output.

Finis

feral air · « **Reply #17 on:** September 14, 2007, 02:08:52 PM »

electrondady1 is right, increasing the gap decreases the swept area but I think the trade-off is worth it in this size/weight class...

4 inches isn't enough, imo, not when each barrel weighs 60lbs+. The more the rotor weighs, the more gap you want...up to 20%...and then it probably depends on your average wind speed as well. The larger the gap, the more air transfer and the sooner it starts spinning.

The most inefficient windmill is the one that's not in the air yet.

My advice is to forget the gap and make each vane opening 3.75sqft - the gap will come out close enough.

disaray1 · « **Reply #18 on:** September 15, 2007, 07:10:12 AM »

Thanks Feral, and everyone else for your imput- been scrounging for bearings and mounts...almost there...

David

electrondady1 · « **Reply #19 on:** September 15, 2007, 08:10:34 AM »

i think it's great your trying to build the mill from recycled stuff

but don't feel bad if you need to spend a bit on new bearings.

thats the part that will take all the guff, dictate the support structure design

and give your mill it potential life span.

once you get the thing up and spinning,

you can enter into the real debate, which is,

just how much can i overdrive the alternator !!

Norm · « **Reply #20 on:** September 21, 2007, 06:25:11 PM »

This has always been encouraging to me...just cut

up some55gal drums and fasten them together like they did here...you could be collecting a lot of

free energy while trying to make it 5% more

efficient ....not worth the hassel !

http://www.motherearthnews.com/library/1974_March_April/The_Savonius_Super_Rotor

( :>) Norm.

News:

Author Topic: Felonius Savonius (Read 4505 times)