VAWT: how to get more power out of a 4m diam x 1m height turbine?

[brandnewb]

Today with a rather high wind with anything between 3.3 m/s and 6.4 m/s depending on the measurement location I was unable to get any torque of note out of;
3 x 1m rae 2822 blades with 40 cm chord length and 1m height at a bit over 4m diam.



The torque it produced when measured at the blade location was a little over 100 grams of force as the turbine did spin continuously after it auto starts. Yet really slowly and nothing that can be used to drive an alternator of any size.

Are there any suggestions one can think of of actions I can take like decreasing chord length, increasing blade height. Adding 3 extra blades, Changing airfoil as to increase the potential power output of the turbine? What ever one can think of would help.

THe bearings are sub optimal yet I'd like to not focus on those as of now as I think that even when the bearings have 0 resistance then the blades will still not produce anything noteworthy to drive an alternator.

I am more than willing to change my thinking if I was set straight of course.

[MattM]

Looks too flimsy to catch the wind.  Just my uninformed humble opinion.

[SparWeb]

Can you clarify your torque measurement?  I would expect Newton-meters, or foot-pounds, or some other "force-distance".

That's an impressive machine, and it seems well exposed to wind, except for whatever building is casting shadows on it.

Are the blades given any "toe in" or "toe out" angles relative to the arms?

[brandnewb]

this is an early prototype and I will make it less flimsy over time. Dare I say I'll make it storm proof so much love and tender care it will get.
Can you please explain what you meant with 'too flimsy to catch wind'?

Torque was measured using a suite case weight measurement hand held unit. I placed it near one of the blades and pulled it to rotate the turbine. Then I read the value that was on the display. I am assuming the unit measures in steps of 50 grams as an earlier tests before some bearing improvements showed 150 grams and the turbine would not rotate and auto start and would even slow down to a halt after a good manual start. After the bearing improvements it showed 100 grams of force needed to rotate the turbine and it will auto start and spin continuously.

The wind was coming from an "unobtructed" direction at the time the photo was taken. Houses in that direction are +- 300 meters away with a lake in between.

The blades are positioned perpendicular to the central column so I think that is called having a pitch of 0. As the whole thing is designed modular it would be really easy to replace the blade connectors with variable pitch versions. I mean variable pitch during preparation time, not something as sophisticated as variable pitch at run time

What can I try next to make things spin with more force once the wind is picking up again?

[Adriaan Kragten]

In chapter 3 of my public report KD 601 (see website: www.kdwindturbines.nl), I give the design theory for a H-Darrieus rotor derived from the design theory of a HAWT. I also give the geometry for a 3-bladed rotor with a radius R = 1 m and a rotor height H =1.5 m. I found that the optimum tip speed ratio lambda = 4.2 and that the chord c must be about 0.2 m if a symmetrical airfoil NACA 0015 is used. If you scale this rotor up to a diameter of 4 m, so to R = 2 m, you need a chord of about 0.4 m and that is about what you have. The chord is independent of the height H, so your design with H = 1 m seems about optimal if you have used a similar symmetrical airfoil.

But there is a very good reason why you can get almost no torque out of this rotor and that is that a H-Darrieus rotor has a Cq-lamba curve which is negative for low values of lambda. You have no provision to increase the starting torque and so the curve "fixed blades" out of figure 4 of KD 601 is valid for your rotor. In this estimated curve, you can see that the torque coefficient is positieve for 0 < lambda < 0.3 because in this region the rotor is working as a drag machine. However, it is negative for 0.3 < lambda < 1.4 because for this lambda range, the airfoil is stalling at the front and the back side. A stalling airfoil has a lot of drag and this causes a negative torque. For a real positive Cq and Cp value, the lambda must be higher than 1.4 and this requires a starting mechanisme for a rotor with fixed blades. This is one of the disadvantages of a Darrieus rotor. Other disadvantages are given in my public report KD 215. It is never possible to solve all disadvantages of Darrieus rotors and therefore my advise is to forget this kind of wind turbines.

I think that your rotor will turn if you twist a rope around the shaft and pull the rope untill your rotor is turning fast enough to come in the lambda range for which the torque is positive. But once it is in this lamda range, it will accelerate and finally it will turn with an unloaded tip speed ratio of almost 7. This may result in a very dangerous situation at high wind speeds if there is no brake to stop the rotor! So don't try this for high wind speeds.

[Mary B]

You don't want to measure torque at the blade, measure it at the shaft! You have a LOT of mechanical advantage built into that machine with those long arms, and potential for self destruction!

this is an early prototype and I will make it less flimsy over time. Dare I say I'll make it storm proof so much love and tender care it will get.
Can you please explain what you meant with 'too flimsy to catch wind'?

Torque was measured using a suite case weight measurement hand held unit. I placed it near one of the blades and pulled it to rotate the turbine. Then I read the value that was on the display. I am assuming the unit measures in steps of 50 grams as an earlier tests before some bearing improvements showed 150 grams and the turbine would not rotate and auto start and would even slow down to a halt after a good manual start. After the bearing improvements it showed 100 grams of force needed to rotate the turbine and it will auto start and spin continuously.

The wind was coming from an "unobtructed" direction at the time the photo was taken. Houses in that direction are +- 300 meters away with a lake in between.

The blades are positioned perpendicular to the central column so I think that is called having a pitch of 0. As the whole thing is designed modular it would be really easy to replace the blade connectors with variable pitch versions. I mean variable pitch during preparation time, not something as sophisticated as variable pitch at run time

What can I try next to make things spin with more force once the wind is picking up again?

[MattM]

The top bar has no tension on it when going upwind, allowing it to change the angle of your blade.  You need the arms to be stiff to prevent that.  You might be able to do what you did on the bottom arm for the top. 

[joestue]

The top bar has no tension on it when going upwind, allowing it to change the angle of your blade.  You need the arms to be stiff to prevent that.  You might be able to do what you did on the bottom arm for the top.

That angle doesn't matter.

[SparWeb]

The top bar has no tension on it when going upwind, allowing it to change the angle of your blade.  You need the arms to be stiff to prevent that.  You might be able to do what you did on the bottom arm for the top.

That angle doesn't matter.

Not so much the angle, as the folding back and forth will be the problem.  That will start a nasty oscillation, so Matt's gut-reaction is a sound one.

[SparWeb]

With arms over 2m each and 100 grams to overcome shaft friction, your shaft friction torque is therefore 9.81*0.1kg*2m=20 Newton-meters or 15 lb-ft.
That's the torque to tighten a 3/8" nut. Way too much.

You are looking for "inch-pounds" of torque in your axle to make this work.  You will need better bearings. 
I believe most VAWTs need to have separate bearings for the axle, and drive their alternator from a belt, chain or other sort of transmission.

Once you build for yourself an alternator, it will also need good bearings and low iron-losses (Air-core Axial-flux) so that it doesn't compound the loss from the axle bearings.

[brandnewb]

thx all thus far.

I've added stiffners similar to those on the bottom arms to the top arms in the meantime and I tried to manually start the turbine by rotating it with a rope. Try as I might I can't get it over a certain speed which still falls below the required speed for it to start "working"

The orientation of the rae 2822 blades is with the more curved part of the bladed facing inwards like so.

I've measured both that length of the outer parts at 406.347210mm of the blade and the inner part at 406.013459mm.

Are the blades oriented correctly or should they be flipped? having the extra curvy side facing outwards?
I will also consider replacing the bearings. they really don't want to rotate all that easily. Perhaps that is why I can't get more than a certain speed out of it during manual starting with a rope.

[Mary B]

If your bearings have a grease seal remove it for now. Later on you can figure out how to make them water resistant. One step at a time! The seals create quite a bit of shaft friction.

[SparWeb]

Most VAWT's I've seen use a symmetrical airfoil, so top/bottom and inside/outside don't matter.

[SparWeb]

Have you seen this thread?
https://www.fieldlines.com/index.php/topic,150186.29.html

[brandnewb]

Interesting thread indeed.

I am really worried about fast rpm in an urban setting. So I'll try decreasing bearing resistance by removing the ball covers like Mary suggested and adding 3 extra blades.
Shall I abandon this rae 2822 airfoil and go with naca 0015 or naca 0018? The rae 2822 was suggested here earlier and I just went with it, I never really chose it with a good understanding.

If the rpms are still too high with a diameter like this for use in an urban setting I will have to sacrifice efficiency by replacing the blades with cup like objects like an anemometer as I'd rather build 3 or more slow spinning turbines than a single fast spinning one.

[Mary B]

Your geometry is rather odd... normally see taller than wider... the taller blades provide more power

[brandnewb]

agreed. however this is just to see what has potential or not. I am aiming for a turbine of 4 diam x 5m height. And / or several smaller ones at 2m diam at 3m height.

The most important thing is that things don't spin too fast as to become noisy and a safety hazard when blades fling off.

[Crockel]

I suspect you'd need to get that spinning at a pretty decent speed to see much torque or power out of it which may be a bit scary for the neighbors. It does look pretty cool though. A lot better than my first Darrius attempt.
I'd suggest building a Savonius VAWT. You'll see more torque at start up especially with that diameter. Or maybe with Savonius blades towards the middle to get the Darrius blades up to speed.

[MattM]

By the time it gets up to adequate speed, tension should build nicely in the metal arms.  If the blades are well mounted it shouldn't fly apart unless you have a ground strike or oscillations break off a stiffener.  By then it will be safely out of balance and either self destruct or slow down on its own.  Much of the potential for your wind turbine depends on structural integrity.  They tend to not work so well once out of balance.

[SparWeb]

The most important thing is that things don't spin too fast

Put a generator on it and the speed will slow down. 

[brandnewb]

A Savonius type turbine, although much less efficient, somehow just feels better to me for my particular scenario.

EDIT: Also removing the ball covers on the bearings as Mary suggested did improve things a lot. much less friction.

I made a proof of concept and with 1m/s wind it auto started and rotated at roughly 6rpm. It took a force of 2.2KG to stop the rotation when measured with this suit case thingy at the and of an arm.

I am really happy with these results and will focus my efforts there. There are a million things that need improving here to make it combat ready but the proof of concept for the turbine is there.

Would anyone know a source for a prony break that includes a conversion from applied force to torque? I think i'll be needing those figures for the alternator design.

What we see here below is the proof of concept with cup like blades of 63cm width and 2.4m height.

[SparWeb]

Two fish scales and a leather belt will get you a passable prony brake.

[Adriaan Kragten]

A Savonius type turbine, although much less efficient, somehow just feels better to me for my particular scenario.

EDIT: Also removing the ball covers on the bearings as Mary suggested did improve things a lot. much less friction.

I made a proof of concept and with 1m/s wind it auto started and rotated at roughly 6rpm. It took a force of 2.2KG to stop the rotation when measured with this suit case thingy at the and of an arm.

I am really happy with these results and will focus my efforts there. There are a million things that need improving here to make it combat ready but the proof of concept for the turbine is there.

Would anyone know a source for a prony break that includes a conversion from applied force to torque? I think i'll be needing those figures for the alternator design.

What we see here below is the proof of concept with cup like blades of 63cm width and 2.4m height.
(Attachment Link)

This picture isn't showing a Savonius rotor but a pure drag machine. A pure drag machine is the worst choice you can make for a VAWT because it has a very low maximum Cp of about 0.05 (see my public report KD 416). A normal Savonius rotor has two buckets which overlap in the centre. A normal Savonious rotor works partly on drag and parly on lift and therefore it can have a much higher maximum Cp of about 0.2. But a well designed HAWT can have a maximum Cp of about 0.45 and the amount of needed material is also much less than for a Savonius rotor. A H-Darrieus rotor is a lift machine and can have a maximum Cp of about 0.3. So concerning the maximum Cp it is the best choice for a VAWT. But all disadvantges as mentioned in KD 215 make that it is very difficult to design a Darrieus rotor which starts at low wind speeds and which isn't shaken into pieces at high wind speeds.

[brandnewb]

Thank you Adriaan,

All I care about is to have a turbine and alternator that is easy to construct for anyone. I really do not care about maximum efficiency.
As long as It generates power then we are golden.
This all in fear of armagedon.

Now if dark days do not come then it would be nice of course to have the DIY turbine cost efficient and produces more power than a poorly designed one. That is certainly an aspect that needs more scrutiny.
But most of the general public, the ones that need cheap/free power now more than ever, I for once certainly, can't really make heads or tails from any scientific publications.

@Sparweb, please elaborate? what is a fish scale? and why do we need 2 of those?

[MattM]

A simple drum brake using a lever arm.

https://gfycat.com/anguisheddimwittedamericangoldfinch-brake

[SparWeb]

You could also call it a spring scale.

https://duckduckgo.com/?q=fish+scale+weight&iax=images&ia=images

How to use 2 of them:

https://enginemechanics.tpub.com/14037/css/Figure-8-3-A-Prony-Brake-53.htm

[Adriaan Kragten]

Thank you Adriaan,

All I care about is to have a turbine and alternator that is easy to construct for anyone. I really do not care about maximum efficiency.
As long as It generates power then we are golden.

This is just the point. You want that it generates some power and you build something which should do that. This needs at a certain investment and you want a certain amount of energy in kWh for a certain investment. If you use a drag machine you have a very large investment and you get only a very little power if you get any. Realise that the optimum tip speed ratio of a pure drag machine is about 0.15 (see report KD 416 figure 2). So the rotational speed for a reasonable wind speed will be extremely low and you need a transmission with a very high accelerating gear ratio to get a sufficiently high rotational speed for the generator. Only this transmission makes the design very expensive. A H-Darrieus rotor has an optimum tip speed ratio of about 4.2 so a factor 4.2 / 0.15 = 28 higher than for a pure drag machine. So for the same rotor diameter, the rotational speed at a certain wind speed is also a factor 28 higher. But for a well designed 3-bladed HAWT, the design tip speed ratio is about 6 and you can use a direct drive generator. The ratio in between the generated energy and the required investment is much and much better for a HAWT than for a pure drag machine.

[brandnewb]

Your geometry is rather odd... normally see taller than wider... the taller blades provide more power

Mary, would you please direct me to a resource that goes a tiny bit deeper into that? based on
https://www.omnicalculator.com/ecology/wind-turbine
for VAWT the height is equally important as the diameter. i.g. 4d x 5h results in the same output as 5d x 4h

It could be of course i am using that calculator incorrectly or maybe the calculator it self is broken somehow.

[MattM]

The surface area impacted by the air is calculated, creating the confusion.  Your torque is better on the short and wide turbine.  Your RPMs will be higher on the tall and skinny turbine.  Under load YMMV.

[Mary B]

Your geometry is rather odd... normally see taller than wider... the taller blades provide more power

Mary, would you please direct me to a resource that goes a tiny bit deeper into that? based on
https://www.omnicalculator.com/ecology/wind-turbine
for VAWT the height is equally important as the diameter. i.g. 4d x 5h results in the same output as 5d x 4h

It could be of course i am using that calculator incorrectly or maybe the calculator it self is broken somehow.

Just a comment from seeing commercial and home builds that are generally taller than they are wide... swept area = more power available in any given wind speed. But as mentioned, narrower = less torque to overcome any cogging.

[Bruce S]

brandnewb
Here's a link to a preeminent fellow forum member that basically wrote the book on VAWTs, his knowledge is vast enough that his foil profile is named after him (LENZ2). Take a deep dive into his website and why he built them the way he did. I still go to his website for information.
 
https://www.windstuffnow.com/main/

Cheers
Bruce S