Author Topic: Savonius gap clarification (Read 8932 times)

feral air · « **on:** October 16, 2007, 07:32:17 PM »

Because the Sandia paper wasn't real clear on what they meant by 'gap' I think a lot of people (myself included!) have been using the wrong measurement for determining the gap to use for their simple Savonius rotors.

If we have a barrel that's 22inches in diameter an 18% gap is about 4 inches, right? (2.2" is 10%, 4.4" is 20%) Maybe not...

If you cut the barrel in half and put it in the basic S-shape (no gap) the rotor diameter is double the barrel diameter, or 44inches. I think that's the number we're supposed to use...which means we should use a roughly 8inch gap in this case.

If I'm right (!) we've all been using a 9% gap...or half what we thought we were going with, anyway. That could easily explain (along with the central shafts!) why Savonius rotors are considered so poor.

Thoughts?

Stonebrain · « **Reply #1 on:** October 16, 2007, 02:15:42 PM »

I found 20% gap,compared to the diam.

So,you're right,make it 8-9 inches abit more if you take in acount the shaft.

That's my thoughts about it,though it's not much,it's a start.

cheers,

stonebrain

Norm · « **Reply #2 on:** October 16, 2007, 02:55:26 PM »

Easy answer go to....

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

scan down to where it says....

The Savonius rotor is easily made:

and all will be revealed !

a barrel is a barrel and wind is wind they've tried and it seems to works best the way they describe.

I've tried different air gaps with plastic cups

increasing or decreasing the air gap from this

recommended spacing gives unfavorable results

IMHO....

Use chain drive...more speed? use a higher ratio, just that simple....chain wears out get

some new chain.

So as I've said before

(Yay Savonius and 5-7mph turbulent winds)

http://www.fieldlines.com/story/2005/7/27/161357/070

I'd love to use barrels but I don't think the neighborhood would approve...

( :>) Norm

feral air · « **Reply #3 on:** October 16, 2007, 04:31:15 PM »

"The Savonius rotor is easily made: Just split a cylinder equally through its length, offset the halves by a distance equal to the radius of the original form and secure the segments to end plates the width of the new diameter. Then insert a rod through the center of the assembly, fix its ends in bearings, and the device will rotate when exposed to the wind."

That's more gap than necessary and will reduce the swept area. The larger gap may help offset the shaft-being-an-obstruction but, no...

They were trying to simplify it and I was trying to determine what the Sandia report really means when it talks about a rotor with an x% gap - I only mentioned a barrel and its dimensions to put my theory in context - x% of what?...

Probably the diameter of the s-rotor without a gap. If I were a scientist studying s-vawts, that's probably what I'd use as the baseline. If the first thing you test is a gapless s-rotor then a 10% gap would be 10% of the diameter of the gapless s-rotor...would it not?

So, if I'm right, double the diameter of your cylinder and 18% of that should be your gap. 20% if you intend to ruin the airflow with a shaft. Simple. :-)

disaray1 · « **Reply #4 on:** October 17, 2007, 06:33:57 AM »

I think what is at issue here is a misunderstanding of the terminology. Some reference "gap" and some "overlap", like they are the same thing. Are they the same thing? My current thinking is no. Overlap, the percentage of stagger between two half cylinders like this

(3D glasses required)

* * *

* *

*

*

----------------------

*

* *

* * *

and gap, the distance between the two halves, like this

* * *

*

* *

_____

______

*

*

* * * *

Different publications call the two terms differently, and that causes the confusion. I could be completely wrong about the above, but if drawn out on paper with proportional scaling, the model with gap looks like it would flow better. The barrel I have flying now has no gap, and 4" overlap. It'll start on a cricket fart, and spins very nicely up to about 75 rpm in 10mph wind, but has very little torque.

Is the above correct? Thanks all!

David

disaray1 · « **Reply #5 on:** October 17, 2007, 06:49:25 AM »

Actually, you'll need more than 3D glasses to see that. How bout this..

feral air · « **Reply #6 on:** October 17, 2007, 10:35:32 AM »

Here's how I always imagined it. I guess it helps if you mentally remove the straight lines.

My small quick vawt has a 1/2" overlap, one blade overlaps the other by that much, and a 1" air gap.

You're right though, it's mixed and matched depending on the source...no fun.

Bruce S · « **Reply #7 on:** October 17, 2007, 01:02:17 PM »

feral air;

You have it correct. Most all of the long term places show it 1/2 the radius of the diameter before cutting the barrel, that would be the overlap.

ALL the older books on this shows a 1/2 radii overlap as well.

There is now doubt that the term overlap and gap have been intermixed way too much.

A modified Savonius does show a true gap, but this goes further in that it actually modifies the overlapping section as well , where they are becomeing a flatten area.

Windstuff Ed and others a few years ago did a whole lot of design studies on the VAWT and the Sav. to the point where there was even a single 3 cup sectioned VAWT instead of the normal 2 cup designs.

The overlap is to "help" guide the air through the cup and the 1/2 radius seems the be the sweet spot. The Gap is more of a structual help than air flow.

Maybe Ed will chime in here about those long ago discussions , he's kinda the guru of the VAWTs.

Hope this doesn't add to the muck:-)

Cheers

Bruce S

feral air · « **Reply #8 on:** October 17, 2007, 09:07:10 PM »

Do I have the terms right or? I'm confused again. Anyway...

Half the radius isn't the same as 18% of the diameter of a gapless s-rotor. With a 22" diameter barrel there's a 2.5" difference between those methods. If we add the motherearthnews method then the spread becomes; 5.5, 8 and 11 inches.

I was trying to leave modified Savonius rotors out of it for simplicity - I was just trying to figure out how the % was determined in the Sandia research. "% of what?" was the question.

I'm not the guru Ed is but I like to think that I know my beans from my pork. Maybe I don't, I dunno.

feral air · « **Reply #9 on:** October 17, 2007, 11:46:03 PM »

I don't know how pork and beans entered into it but that's a terrible line. "I like to think I know my left from my right" is more appropriate.

windstuffnow · « **Reply #10 on:** October 18, 2007, 09:52:25 AM »

The above drawing is correct. The drum machines and sandia's machines are two different breeds. Sandia was leaning toward the lift portion as the wing crossed the 0 degree into the wind where the barrels tend to operate best as primarily drag machines but does achieve lift as the drum passes the 0 point.

The whole idea of the gap and/or overlap is to reduce the negative pressure or vacuum behind the upwind wing (negative drag). Also, by overlapping the drums the upwind side tends to feed the downwind while reducing Cd on the upwind side similar to a diffuser. Like all drag machines the power is extracted primarily by the downwind blade, the upwind blade simply reduces the amount of power extracted by the downwind blade.

I wouldn't read to much into the "perfect gap/offset" because this will vary based on windspeed. If it's to large then the turbine will suffer in low winds and visa versa. In lower winds ( 0-20mph ) your better off keeping the overlap small, 4" on a full size drum would be fine, and keeping the larger cup area of the downwind side. This will produce more torque.

You also need to keep in mind that your actually extracting energy from a wind speed that is less than the wind speed flowing across the machine. To follow the Betz theory and to extract the highest possible amount of power the blade shouldn't be running any faster than 1/3 the windspeed. So if your windspeed is around 10 mph then your actually extracting power from a 7 mph wind because the blade is moving away from the wind at the same time by 1/3. In the same instance the upwind is actually traveling faster by the same amount so upwind drag is calculated for a 13 mph wind.

If there is no load on the turbine and it's spinning at a TSR of 1 then the drag machine doesn't extract any energy. The closer to 1 you get the lower the extraction. Going below 1/3 does increase torque but you'll find the lower rpm also decreases the overall output.

Their fun machines and super easy to build unfortunately they'll never be super efficient overall. Great learning tools!

.

feral air · « **Reply #11 on:** October 18, 2007, 04:36:03 PM »

"I wouldn't read to much into the "perfect gap/offset" because this will vary based on windspeed."

You're right but missed the main point. Again, "% of what?" was the question...

Lets say I decide that for my situation/wind-region that a 10% gap really is the best choice, how do I figure out what 10% is? The more I think about it the more I think it's based on double the cylinder diameter...

So, if I had a 20in diameter cylinder and wanted to use a 10% gap I would go with 4 inches (20x2/10=4). Using my theory and your advice for a standard barrel you're suggesting a gap that's less than 10% for 0-20mph winds. I wouldn't suggest 10% unless you're really windpoor...0-10mph.

I can say I used a 6.5inch gap on my barrel and that it seems to work really well in 0-20mph but that doesn't really tell you much. Trying to relate that to anything that's not a barrel or something that runs in 0-15mph is hard.

Saying I used a 14.5% gap though...that's different. Now you know exactly what I've got goin' on, you can relate it to yours and you get the why...I don't have to defend the exact amount since we can easily compare the percent I used to the wind speeds in my area to determine whether or not it was a good choice.

Now that we know (assuming I'm right is scary, I know) how to figure out the percent we can give sound advice and make informed decisions. That's why it matters. :-)

vawtman · « **Reply #12 on:** October 18, 2007, 05:32:17 PM »

Boy oh boy now im seeing why Volvo made his posting.

windstuffnow · « **Reply #13 on:** October 18, 2007, 05:37:47 PM »

I would imagine it's the % of diameter. I don't remember from one report to another,it's been to long ago. It's been probably 10 years or so since I studied the Sav. I built a multitude of them back then though and they all led me to the same conclusion of finding a better way to extract energy. I had a lot of fun with them, I actually still have a bunch of 2ft x 2ft sections using an airfoil shape.

Anyway, I don't see the gap/overlap as a major stumbling block - the power increase/decrease is so minor in comparison to the wing design or other changes that can be made to overcome the airflow problems.

In reality, the downwind blade, using a drum shape is one of the most efficient ways to extract energy. If you calculate the extraction from the reduced windspeed you'll find it's very close to Betz, unfortunately the blade needs to come back to it's starting position and in doing so it takes quite a bit of power. That of course led me to the Darrieus designs where the upwind wing was actually adding something to the output instead of taking from it. Another nice machine but there are way to many drawbacks to that design as well. For a quick simple machine the Sav is a great little machine.

.

feral air · « **Reply #14 on:** October 18, 2007, 07:26:06 PM »

vawtman, I'm just trying to get this nailed this down....you knew what this thread was about (and who started it) so if you didn't want to read it or get involved why click and post? Honestly, if you have something to add I'd love to hear it. :/

"I would imagine it's the % of diameter."

I thought so too but...

Lets take the most extreme advice (from motherearthnews); "offset the halves by a distance equal to the radius of the original form". If you're right then they've been getting away with murder suggesting a 50% gap. The other advice (I guess it's old-school), offset the halves by half the radius..that'd be 25%.

If we use my 'double the diameter' theory then neither of those sounds too bad. They're really suggesting 25% and 12.5%, repectively. See what I'm saying? It makes perfect sense. How could motherearthnews get away with suggesting a 50% gap?...they aren't, that's how.

"Anyway, I don't see the gap/overlap as a major stumbling block - the power increase/decrease is so minor in comparison to the wing design or other changes that can be made to overcome the airflow problems."

But it is a major stumbling block when you're trying to talk about or build a simple Savonius. No, it's not that hard to choose a more efficient design but that's not the point either. The fundamentals still matter...

The next time someone comes along and says they're building an s-vawt from barrels (or whatever) I want to know that I know what I'm talking about. I want to be able to explain it correctly and if there's an argument over the gap that I suggest it'll be a debate over a percentage point or two instead of several inches like it was last time when disaray1 asked.

It may not matter to you but it does to me...I guess. whatever.

windstuffnow · « **Reply #15 on:** October 18, 2007, 10:02:52 PM »

I understand, you might want to contact the inventor himself. I've found, alot of the times, looking at the original patent on a device will yield answers to questions not yet thought of. Here is a link to the original patent from 1928 ... patent # 1766765 http://www.google.com/patents?id=aN54AAAAEBAJ&dq=savonius

There are alot of follow up patents from those after him that have worked on improving the machine. I searched most of them way back when while trying to make my own improvements but the best ones I've built were in the mid to upper 20% efficient range, the best being 28%. The barrel type will yield in the range of 18-20% of frontal area.

I've always refered to mine as a percentage to diameter as a rule of thumb. It could have been from something I read about the sav at one time. That being the case a 24" diameter barrel with a 10% offset would be 2.4 inches. The cfm airflow through a gap of 2.4 x height of the barrel would be quite high since its under compression and has a low pressure drain.

I'd bet if you looked at alot of different sav designs you'll find their all built a little different. Is the Sav on mother earth built by ? really know anything about aerodynamics or was it something he put together in his back yard that someone found interesting. Sandia spend a bundle refining thier system but it can't easily be reproduced in the backyard

To be quite honest you'll get more power and higher efficiency if you move the barrels appart and refine the trailing edge removing the gap problem all together.

.

thefinis · « **Reply #16 on:** October 19, 2007, 06:08:32 AM »

Feral air I think I understand what you are wanting ie a good formula for future builds based on the best performance but you might be leaving out some of the factors involved. The best build for a set of materials may not be the same as the best build for a turbine of X diameter. If you are limited by size of turbine for zoning etc then you want the most efficient.

I think that it is still less gap than 50% of the original diameter or 50% of one side's diameter. I liked 25% of the original cylinder's diameter for an overlap(gap) as it seemed to do well both on paper drawing out the air flow and flying.

One problem you will encounter in trying to use the two halves added together as a figure for taking the percentage from is should it be the percentage of the turbine diameter before the overlap(gap) or after the overlap? If you are trying to make a turbine of X diameter then the formula would need the overlap as a figure of that diameter not double the starting material's diameter.

Dang it out of time more later.

Finis

feral air · « **Reply #17 on:** October 19, 2007, 02:26:10 PM »

"I think I understand what you are wanting ie a good formula for future builds based on the best performance but you might be leaving out some of the factors involved. The best build for a set of materials may not be the same as the best build for a turbine of X diameter. If you are limited by size of turbine for zoning etc then you want the most efficient."

I really just want to know how to get at the percent so that I've got that part right. I left out tons of factors but only for simplicity and to help keep the thread on topic. Materials, weight, bearings, build quality/tolerances, average windspeed, alt, etc...everything adds up to make a whole and the "best" gap is probably going to change some based on all those variables but it'd still be nice to know how to calculate the percent.

I'm not trying to say the gap percent is the be-all or the end of the line. Not at all....it's just this little thing that's been nagging at me the last few days. There just doesn't seem to be a very clear-cut answer.

"One problem you will encounter in trying to use the two halves added together as a figure for taking the percentage from is should it be the percentage of the turbine diameter before the overlap(gap) or after the overlap?"

Before is what I was thinking but then if you use my 'double the diameter' theory 50% makes it a cylinder again. That's the only problem I see, if it's a problem. It could've been done that way on purpose though - 0-49% could be clockwise rotation and 51-100% could be counter-clockwise or vice versa.

That's the best I can come up with and it seems to make sense but it's just a theory...I still want to know for sure.

vawtman · « **Reply #18 on:** October 19, 2007, 04:58:34 PM »

Feral

"im just trying to nail this down"

Well one thing you could do is design your rotor so that you could adjust your barrels to what works best for ya.

Sometimes you can study forever but until you try it you will never know.So many variables.

Maybe post on what you tried not what you read.

I love vawts of all shapes and sizes.

Its not easy and nobody has come with the perfect turbine and probably never will.

Take what you can get.

Get to work and have fun.

thefinis · « **Reply #19 on:** October 20, 2007, 08:04:35 AM »

I was never able to nail down any one way of figuring it as the best. I liked figuring it as percent of diameter of original cylinder as the overlap is mirrored. 6 inch overlap on one half is a 6 inch overlap on the other half. The middle support pipe was what I could never figure in for air flow as it increases the turbulence as well as blocks air flow. I tried to build one without the middle pipe but was never able to get it to stay straight over any length of time.

The big problem seems to be that the other type vawts were usually rated on a finished turbine diameter but the S rotors tended to be studied as (cylinder) halves and the best way to put them together. Most S rotor studies were done using two halves and trying to find the relationship to power produced vs overlap/spacing. That seems right but due to swept area differences it really did not address the real efficiency issue of most power produced in a certain diameter turbine. That means that the less overlap the greater the swept area so the more area to extract power from and it skews(screws) the figures.

Finis

feral air · « **Reply #20 on:** October 31, 2007, 12:10:22 PM »

That there's this much ambiguity surrounding what's often called the "Simple Savonius" just seems strange. The terms and formulas should be concrete and anyone using them incorrectly should be easily corrected.

I always figured it by the diameter of the cylinder too but all I know for sure is that that's the prevailing theory.

"I tried to build one without the middle pipe but was never able to get it to stay straight over any length of time."

A design like my "quick vawt" makes that fairly easy. If you want any real power out you'd need to scale that up a ways but that shouldn't be too hard to do if you're determined. I've been considering ordering some custom "U" brackets so I could use larger diameter pipes (in the spirit of "go big or go home") but the only place I've found that'll do short-runs wants like $40ea. for 2 custom brackets. Ouch.

Savs are already less efficient to start with so I never even considered adding a shaft to mine. So far I've been lucky, my barrel-vawt has held its shape fine for the last few months and last week it saw a nice storm so I know it's good to go. My "quick vawts" are just toys (yard art) but they've held their shape just as well.

"... but due to swept area differences it really did not address the real efficiency issue of most power produced in a certain diameter turbine. That means that the less overlap the greater the swept area so the more area to extract power from and it skews(screws) the figures."

I get why they did it that way though; you have a cylinder with whatever diameter, now make the most efficient rotor you can. The diameter of the result doesn't matter so much when you look at it that way.

The most power for a given diameter turbine is a different question that you have to answer (at least in part) with modified blades, different designs, etc.

If you're trying for the most efficient (drag-based) rotor for a given diameter then I think the so-called "squirrel cage" vawts have the edge. The little one I made has a 9in diameter and sweeps 2.88sqft. The closest Sav. I have to that has a 10in diameter and if it was as tall as my squirrel cage it would only sweep 1.62sqft.

The squirrel cage wins by 1.26sqft (!!!) and the squirrel cage is an inch smaller in diameter. That's a fairly drastic difference at that scale. Of course that's not the efficiency difference but it should still give an idea. I don't know how that scales either but I'd imagine the squirrel cage keeps its edge.

And now that I've rambled off-topic, I'm done. I don't care that much to drag this out. Eyeballing the gap and overlap seems to work well enough for me and I probably won't be making many more Savonius rotors anyway. take it easy

mechamz · « **Reply #21 on:** September 18, 2008, 11:03:08 PM »

Hi, this is exactly what I wanted to know too

P
but based on this,

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

for the gap width ratio, s/d

bucket diameter,d = 2 r

My question is do we need to consider the diameter of the shaft?

spinningmagnets · « **Reply #22 on:** September 19, 2008, 11:23:28 AM »

I haven't built one yet, but from my reading...

If the shaft diameter is less than, say, 1/4 the width of the gap-hole its centered on, I wouldnt worry about it affecting output.

As I understand it, you are balancing the swept wind catching area on the retreating scoop side against the the wind pushing back against the back of the advancing scoop.

If there was zero gap, the wind resistance on the advancing side would be the same, but there would be turbulent drag behind the advancing side, and turbulent wind capture on the retreating scoop.

Once you introduce a gap so air can flow between the two sides everything changes. Wind "A" hits the outer portion of the retreating scoop. inside the scoop, air "B" gets pushed in a "relatively" non-turbulent fashion in an "S" shaped path to the other side. It pushes air "C" inside the back of the advancing scoop out, reducing drag on the advancing scoop.

I have read that the highest recommended 50% gap performs better than others in high winds, and the lowest recommended gap is 17% for low winds (the most common wind found near ground level)

These are not expensive, make two separate VAWT's with the scoops between a top and bottom solid disc. Make one with Gap "A" and the other with gap "B", and raise them the same height at the same time in the same wind. (or mount them to your front bumper)

Something two feet in diameter and about 3 feet tall should provide fairly accurate data, but then again, the most important thing to remember is, I never made one.

However, I insist you send me a check when you get rich off this!

News:

Author Topic: Savonius gap clarification (Read 8932 times)