Savonius wind turbine

Renewable energies except solar electric or thermal (seeforums dedicated below): wind turbines, energy from the sea, hydraulic and hydroelectricity, biomass, biogas, deep geothermal energy ...
Yamatai
I learn econologic
I learn econologic
posts: 12
Registration: 13/10/15, 10:48
Location: Huismes

Savonius wind turbine




by Yamatai » 13/10/15, 11:42

Hello everybody

I am a newcomer to the site and I have a few questions. I have been following the wanderings of Cristophe quite closely lately on his hydroelectric picocentrale and Yanyan 26 on his soapius.

I then pretty much look for docs on these two subjects on the net, namely hydro and wind.

The hydro I understand a lot how we can size and how it works.
Wind power is more difficult.

I am not lucky enough to have a stream near my home so I look at wind power.
The Savonius is the simplest to manufacture with limited construction knowledge and limited equipment.

I am looking first of all how to size and design this wind turbine.
So I found some formulas to have orders of magnitude.

The first is the theoretical maximum wind power.

P = 1/2 * ro * S * (V ^ 3)
No worries with this formula but it is obviously impossible and the famous law of Betz gives orders of magnitude according to the type of wind turbine.

For a Savonius I find a value of Cp between 0,15 and 0,3. This value is a function of lambda which declines as follows:

lambda = (rad.R) / V

No worries either with this formula.
The optimal lambda seems to be around 1 for a Savonius.

My concern is to calculate my theoretical maximum Cp. The formulas that I find require to have already carried out tests with a prototype.
To have the most efficient prototype possible, I would like to calculate it before completion.

Formulas found:

Cp = ((1 + k) * (1-k²)) / 2
k = V2 / V1 (v2 = v out; V1 = v inbound)

Or

Cp = (2 * C * lambda) / (ro * S * v ^ 3)

In both cases the realization of the wind turbine is essential. I would rather find it surprising that wind turbine manufacturers do not seek to more modeled before building a prototype.
That's why I tell myself that I miss something: p

For Yanyan 26.
I followed your project and I fear that your Savonius is running too slowly and a bit undersized.

Your lambda is approximately 25 m / s (90km / h) during your tests at this speed, your lambda must be less than 0,07.
Even if your Cp max is 0,25 I doubt that with this lambda you were more than very large max CP of 0,04.
Considering a surface of 2 m² your maximum mechanical power was 750W.
I wanted to know where your project was. You haven't updated your site long enough.

I wonder if for this type of wind turbine the most interesting is not an alternator with excitation controlled with for objective a direct current via a bridge of graetz.

Let me explain, to keep the CP closest to the max without modifying the sensing surface played on the mechanical resistive linked to the alternator by its excitation current.
This allows not to build a Savonius with resealable buckets at the speed.
It involves a measurement of the rotor speed by a small permanent magnet gene and a small servo.

I discussed it with a colleague electrician and it seems that the current of the excitation of the inductor plays rather little on the tension of the armature in exit once at saturation (fct of the no of turns) but much on the current and therefore the load.
This will tend to keep a fairly stable voltage but a variable current. Ideal for a solar inverter with an MPPT voltage value in the right range, right?

I do not know if I was very clear I tend to go a little all over the place when I think.
Let me know your comments.
Thank you and have a nice day.
0 x
ARMAND
I understand econologic
I understand econologic
posts: 92
Registration: 18/09/12, 21:14
x 2




by ARMAND » 13/10/15, 13:45

If it is to feed a household for the whole year take at the mini a surface "offered to the wind" of 12 m2, (assuming that there is wind in the place envisaged, and corresponding to the average : Cheesy: )
0 x
Yamatai
I learn econologic
I learn econologic
posts: 12
Registration: 13/10/15, 10:48
Location: Huismes




by Yamatai » 13/10/15, 16:41

No, it's not for my personal diet.

I am an EDF agent so I pay 10% + 15% in benefit in kind, ie 25%.
So it will never be profitable for me. It's for my technical culture, pleasure and if I get something drinkable for not too expensive my loved ones and info on the forum. :D
0 x
ARMAND
I understand econologic
I understand econologic
posts: 92
Registration: 18/09/12, 21:14
x 2




by ARMAND » 13/10/15, 17:32

The answer remains the same whatever the family concerned : Cheesy:
0 x
Yamatai
I learn econologic
I learn econologic
posts: 12
Registration: 13/10/15, 10:48
Location: Huismes




by Yamatai » 13/10/15, 22:39

It is not false.
The idea is to start with a scalable, scalable model for testing.
I have the idea for the first to do it with a PVC tube from 100 to 120 mm and 1 m high.
It will be 0,2 m2 and no more than 2 or 3 W mechanical at 5m / s.
Suddenly a pseudo blower test is possible and noted the characteristics at different speeds and overlap between buckets.
A simple dynamo to have the image of power.
A dynamo key for the starting torque and the upstream downstream wind speed dif to get an idea of ​​the dynamic torque.

I have read that the optimal theoretical recovery is 0.24.
It seems that this wind turbine draws its torque by drag and lift according to the orientation facing the wind and the air speed on the buckets.

In short there is the possibility of experimenting before making a model based on drums or and making its buckets.

On the internet I saw a nice mechanical realization but rotten on the generator part.
0 x
Yamatai
I learn econologic
I learn econologic
posts: 12
Registration: 13/10/15, 10:48
Location: Huismes




by Yamatai » 28/10/15, 14:31

Hello,

I ask myself a few design questions and I think that others have certainly had this question in mind.

For my wind turbine pico in order to do tests I will go on a wooden frame, the gutter descent in 100 or 125 depending on what I find for the blades, a bicycle dynamo for the generator and I have the question left of the axis of rotation.

Initially I thought of starting on two bearings with shoulder embedded in the chassis of small internal diameter (6mm). When I tell a mechanic colleague about it, he doesn't understand why I'm not getting bigger.
I explain to him that I want to limit the friction losses of the bearings so as not to have a high starting torque and unnecessary losses in operation.

For him the resistance is low but he reasons in my opinion compared to our electric or steam engines which have powers incommensurate with the few watts that I hope to capture.

So my question:

For small designs did you use bearings of what reference? And did you have the impression of a significant initial torque?

The soapius has a significant but good starting torque I prefer to have an idea of ​​the real before leaving on the basis of a chassis that can be used for a larger model with bearings and a larger motor axis.

Precision the manufacturers do not seem resistant torque release.
If some are interested a link to a manufacturer doc.

http://www.ntn-snr.com/portal/fr/fr-fr/ ... tID%3D7107

Thank you, a +
0 x
ARMAND
I understand econologic
I understand econologic
posts: 92
Registration: 18/09/12, 21:14
x 2




by ARMAND » 28/10/15, 18:09

Conrad stainless steel bearings, internal diameter 10 mm, the axis is a copper plumbing tube, the shoulders are formed by plumbing fittings ... very low resistance (not measured) below there are only the magnetic bearings ( good luck) : Cheesy:
0 x
Yamatai
I learn econologic
I learn econologic
posts: 12
Registration: 13/10/15, 10:48
Location: Huismes




by Yamatai » 29/10/15, 10:59

Thank you for this return :)

I think you took this hit suddenly:

http://www.conrad.fr/ce/fr/product/1984 ... g-6200-2RS

For those interested in a cheaper site than conrad and normally with better finish gear:

http://www.123roulement.com/roulement-6200-2RS-SKF.php

I tried to understand the bearing identification system and it is a bit obscure for me. What I understood is:
first digit the type of bearing.
the next 4 dimensions (when there are 4 numbers).
the reference following the type of sealing.
the type of lubricant if specified

On the SKF site there is the calculation of the resistant torque but it is unbearable for me. I think I will go on the simplest solution with say miniature bearings and integrated shoulder.

http://www.123roulement.com/roulement-F626-ZZ.php
0 x
ARMAND
I understand econologic
I understand econologic
posts: 92
Registration: 18/09/12, 21:14
x 2




by ARMAND » 29/10/15, 11:28

I order from Conrad because I tinker a lot and buy a little of everything from them. Amha it is better to take stainless steel rab. It is simply necessary to protect them (with a soda plug pierced on the axis) a minimum of dust and runoff, then it turns more than ten years without any maintenance ...

The resistance, on a Savonius, cannot be reliably calculated by an individual, in addition it is limited as information. If you don't want friction you have to mount on magnetic fields (several arrangements possible). This type of product, Savonius, has been studied extensively by engineering schools, it is more profitable to read their work than to want to do its own tests. I made a dozen models in situ before making my own (according to a model that did not exist)
0 x
Yamatai
I learn econologic
I learn econologic
posts: 12
Registration: 13/10/15, 10:48
Location: Huismes




by Yamatai » 29/10/15, 14:13

I am not looking for 0 friction but I was wondering about making a chassis that can be used for a larger one.

I have read a few of the subjects that hang around the internet the optimal overlap seems to be 0,24 radius. The mockup is just there to give me an idea of ​​the difference between a good and a bad recovery rate.

According to this study, the Kp max would be 25% for a lambda between 1 and 1,2. So the theoretical data which verifies very probably I have them in mind.
It does not prevent wanting to do some tests before carrying out a larger one;)

In this same study, the participants put forward the idea of ​​modifying the overlap of the blades with an angle that would not be 180 ° between the two blades.
It would improve the average torque but make it even less linear.
In their experimentation, a relative angle of 125 ° would be a good lead.

I do not intend to do much. A first with adjustable overlap. Greater according to this overlap which should be 0,24 radius.
A new experimental with an adjustable blade to modify the angle between the two blades for this second track.

It's more for the fun of making yourself :)
0 x

 


  • Similar topics
    Replies
    views
    Last message

Back to "hydraulic, wind, geothermal, marine energy, biogas ..."

Who is online ?

Users browsing this forum : No registered users and 242 guests