Horizontal savonius

[stop4stuff]

Hi All,

I've been playing with Lego again

After seeing some previous successes with the Sandia savonius VAWT ( US Patent 5,494,407 )

I had a go at a similar one in Lego, I went for a blade diameter of 150mm (approx 6"), the height ended up being 176mm.

The diagram is printed to scale (& size I want and the frame is made of axles and axle joiners, a few bushes and some short liftarms.

I covered the main surfaces with paper stuck on with insulation tape, I then covered and strengthened other parts of the blade with aluminium tape.

I live in a flat and when the wind is in the right direction it hits one outside wall, does a 90deg right turn against another wall. When my kitchen window and balcony door are open, the wind rushes in thru the kitchen window... sometimes hard enough to pull papers off the fridge (attached with magnets)

The image above shows a rotor and a coil, initial testing revealed peak rpm to be around 600... wind speed=a huge blast in the face

With a moderate breeze, the turbine was turning nicely at 300-450 rpm and picked up speed quickly with gusts.

The blade starts really easily... mmm...

I grabbed my 3 phase alt and hooked it up with a couple of universal joints and an axle.

The frame has a couple of deflectors to make up for the lack of end plates on the blade.

The turbine coped with turning the alt directly, and got the rotors to 600+rpm. I tried with 3:1 gearing, however, it was only with the strongest gusts that the alt would turn, and then only to a max of well below 200 rpm... (no freq reading

Directly driven and the alt on star output, open voltage peaks around 4.4 volts.

Pleased with the apparent ability of Lego, paper and tape, I tested voltage output and hooked up a 1.2v 1500mA NiCad.

Here the alt is running at 0.02kHz (400 rpm), and is putting 20mA into the cell.

The image below was taken as a big gust hit, 1.45v, 0.12A (0.174W) the ammeter peaked at 0.14A just after.



Another couple of windy days and that cell may be charged

On with the fun )

paul

[Ungrounded Lightning Rod]

My browser won't pull that diagram off the patent web site.

(Ancient version of Netscape and the scripting the patent office

uses doesn't get along with it.)

Could you please post it - or a link directly to an online copy?

Thanks.

[stop4stuff]

Hi,

I drafted the diagram for 'my own use', send an email to paul at stop4stuff dot com expressing that you will be using the diagram for 'your own use' and I can send a copy for free. I'm sure we won't be breaching any patent/copy laws

paul

[Ungrounded Lightning Rod]

I drafted the diagram for 'my own use', send an email to paul at stop4stuff dot com expressing that you will be using the diagram for 'your own use' and I can send a copy for free.

Thanks for the offer.  But while I have no current plans to use the diagram for other than personal use, I don't want to make a committment that would constrain me from using it in a commercial product at some future time if the patent law itself does not.

I'm sure we won't be breaching any patent/copy laws

Doubly so since, according to Micropatent Inc.'s posting here, patent 5,494,407 expired on March 1 2000 for lack of payment of fees.  B-)  (It's not an official patent office site so I wouldn't take it as gospel, though.)

I do note, however, that the text and diagrams of patents are public record.  (That is the POINT, after all.)  So there is no copyright on them and no IP violation in sending someone a copy of all or part of one.

Of course you said that you did that particular diagram yourself, rather than copying it from the patent.  I'm not sure whether YOU could claim a copyright on it if the patent didn't describe exactly how to draw it.  B-)

[Ungrounded Lightning Rod]

I hope that structure is really strong.

The G forces on the outer end of a Savonius (or Darrius) rotor at a given wind velocity go up with the inverse of the radius.

The Sandia Savonius design has the advantage that the airfoils are almost balanced, and loaded nearly in tension.  A counterweight at the inner end would just about balance the centripital loading of the blade, leaving the structure mostly loaded in compression to make it hold its shape.

(The Darrius puts essentially all the weight at the edges, and spins 5 times as fast for a given wind speed.  Ick!)

[stop4stuff]

you seem to know more than i do about patent laws etc... and i would'nt want to claim copyright for my rendering of the dimensions described in the patent.

do you want a copy or not?

paul

[stop4stuff]

Hi,

I hope that structure is really strong.

Seems to be

We had 2 days of good gusty wind, and the turbine ran at anything from 0rpm right upto 800+rpm (maxed 1.60v, 0.24A into the nicad).

In all the turbine had around 12 hours of flying time, and inspection shows no signs of stress (no ripped paper/sticky tape etc).

I think i understand what you're saying about loading and balance. Tho this turbine wasn't at all balanced, the imbalance was minimal and there was some vibration esp at high rpms... I also think the deflectors on the frame may have put uneven loading on the turbine... any thoughts?

cheers,

paul

[Ungrounded Lightning Rod]

I think i understand what you're saying about loading and balance. Tho this turbine wasn't at all balanced, the imbalance was minimal and there was some vibration esp at high rpms...

Actually I was referring to the centripital force / centrifugal pseudo-force on the individual airfoils.

On a darrius most of the weight is in the blades.  On an H design, even a large one, you're talking about some horrendous number of Gs (Like a hundred on a machine a few yards across in a decent wind) trying to pull the arms off and throw them away.  The smaller the rotor the higer the force in any given wind.  Thus the catenary design and enormous size of the commercial devices.

On the Sandi Savonius the airfoils come back past the axis, so some of the force trying to pull the blade off the mount is canceled out - redicrected into stretching the metal and trying to uncurl the ends.  (The wind is also trying to do the latter, too.)  I was just commenting that counterweighting the inner end could get rid of the rest of that throw-it-away force from the spin, leaving the support to fight only with the uncurling and driving forces.  Probably unnecessary, though, since the force shouldn't be all that large.

I also think the deflectors on the frame may have put uneven loading on the turbine... any thoughts?

Could be.

But I'd suspect the major problem is that the rotor provides a different amount of drag and drive depening on where it is in its rotation.  I think that's why a typical mill consists of a stack of three in a three-phase configuration - to try to balance out the drive and drag forces.

I'd probably build a stack of six, with the outer, inner, and next-to-end pairs at identical angles.  This would minimize the net off-center forces on the support structure and leave only bending on the rotor.

[Ungrounded Lightning Rod]

do you want a copy or not?

I would still like a copy, if you're willing to send one without the pledge.

Assuming you are, email to rod at node dotcom will reach me.  Let me know if you need a snailing address.

    thanks

    rod

[stop4stuff]

Hi Rod

I know what you're saying about the Darrius turbine.

And stacking a savonious makes sense too.

Would a Sandia style turbine with a 180 deg twist thru it's height work?

I'm sure I could figure a way to make it

paul

[ricknplano]

Regarding twisting the vertical axis - It has been done and works well.  See www.windside.com and then click on models.  You will be impressed.

[stop4stuff]

Hi ricknplano,

windside.com's models are pretty cool... i like the power output/size, shame about their prices.

thanks,

paul

[ricknplano]

Also Paul, in the website I referred you to earlier you will notice the airfoil shape used by the Finnish manufacturer is not the same as the shape you copied from the patent.  If you remember, Savonius himself was from Finland and is a sort of hero over there.  So they have a history with Savonius generators and I tend to think they have a good grip on the best design.  The patent you looked at seems to suggest his shape is an improvement but I strongly doubt it.  He implies a 37% efficiency rating which I doubt is true, but even if it is true efficiencies of about 35% for other savonius type wind generators have been achieved at university tests (the highest Sandia ever measured with their design was about 25%) using traditional shapes without the twist that was added by the Fins.  The design you copied has a long straight axis on the blade that creates larger flat surfaces which probably will (in a large scale unit) create turbulence and balance issues.  The preferred airfoil shape is the same as your basic airplane wing, fine tuned for the size of your swept area, the primary wind design speed, the design rpm for your device, material stress limits, etc.  And as you suggested, twisted to allow for even and balanced spin no matter the wind direction - an important issue during thunderstorms, for example, when the wind may gust and change directions violently.

The twisting helps energy conversion efficiency and balance issues, but defeats one of the basic advantages of the Savonius, simplicity of fabrication.  I think the suggestion of using stacked simple but easily manufactured components may still have economic advantages that out weigh the design efficiency from fine tuning the airfoil.  As you mentioned, the price of the Finnish machine is ...ouch!

[Ungrounded Lightning Rod]

Hi, Paul.

(Didn't see your comment until you pointed it out in email.)

I'd guess that a spiral savonius as you described wouldn't work as

well unless there was a horizontal divider every so often to keep the

airflow from slipping down.  Absent such windwalls the hooked ends will

redirect the air vertically, not just through the tube to the opposite side.

That will give yo much of the momentum change, but not all of it.

Now that might not be all THAT bad.  If it goes far enough it will get

to a segment where the opposite side is pointed downwind and hook back

out, again giving the rotor a push.  But somehow I doubt it will be

as efficient as a bunch of linear segments of the same height.

(It might be interesting to try it all three ways (stack of untwisted,

twisted, and twisted-with-horizontal-baffles).  And it might also be

interesting to compare one that makes two or more twists to one that

makes a half-twist.  There might be gains, or losses, from air getting

windward-to-leeward by vertical motion versus hanging up at the end.)

But like I say, the above is a guess.  Perhaps the air entering the next

region down will resist the downward motion and force the air from

the segment of interest into the passage anyhow.  In a flat

rotor that would be it - you've gotten the momentum change from the

first hook, and nothing of major interest happens until it gets to

the opposite hook and gives up the other half of the momentum.  But

vertical air motion in a spiral structure might have other effects,

useful or harmful.

I don't know if this is much help.  B-)

[stop4stuff]

I see what you're saying

The air flow would be directed up and down as well as around and around with a hollow spiral... losing energy :0

...unless...

later

(thanks)

paul