Torque value on blade design

[Kyle]

I am working on an asynchronous induction generator. I have been researching blade design and have not found torque discussed. Perhaps I am not using the correct search terms. My question is. When considering design, don't you first need to know “how hard” your rotor turns? Then plug in desired rpms, tsr, cut in speed, etc and design from there? I have read about increasing the gap which reduces force required to turn. But, if that is fixed... I am assuming there is a guide that I have not found yet. Also, what is the best way to measure rotor torque - torque wrench?

[JW]

Do some research on cut-in speed

[SparWeb]

That's a very good question, Kyle, and no easy answers.

I recognized that if I went to the trouble of measuring torque while testing my generators, I'd have a power curve that I could count on to match up the blades better.  Never regretted it.
I don't have a guide, really, but I did show how I measured torque during my tests: 



http://www.sparweb.ca/3_Gen_MoCo/Baldy.html

This setup allowed me to measure, directly, the torque on the generator, because that board was the only thing preventing the generator body from spinning while the shaft was driven by the lathe.



No factoring efficiency, losses due to this or that, it's just the simple equation: Torque = Force X distance   and   Power = Torque X RPM

[SparWeb]

On the other side, the question is "how much torque do the blades deliver?"

There's a graph for that, too




The graph above shows three generator power curves (black lines) and the power curve for the blades of one wind turbine rotor, at different wind speeds.  You can see how the blades power curves grow very big with stronger wind, and that there is a theoretical perfect generator curve that will match.

In the example above, the generator power curve on the left is going to "stall" the blades, keeping them running very slow, while the generator power curve on the right is too weak and the rotor blades will run too fast.  The curve in the middle is, of course, Goldilocks.

Actual generator power curves don't quite follow exactly the ideal path, but making comparisons this way does allow you to either select the components you need to match the prop power to the generator power, or to adjust the parameters of the prop that you built so that it does match the generator you have.

[JW]

I am working on an asynchronous induction generator

http://www.fieldlines.com/index.php?topic=139034.0

this is the only link that works from the above topic

http://www.qsl.net/ns8o/Induction_Generator.html

[Kyle]

Thank you for the reply's. I have read some on cut in speed. I know mine is 181 rpm. SparWeb - thanks for that info. I don't have a lathe but my drill press may work.

As to the blade graph. Were those all the same length with different pitch, or same pitch with different lengths or various combinations? Was it all trial and error? What was your starting point and how did you arrive at it?

[SparWeb]

The drill press is good and has some advantages over a lathe.  Particularly, the space you can use to set up to either side of the drill chuck, ie, swinging the table to one side and running the generator with a belt from chuck to genny shaft.  Changing drill press pulleys to change speed gives you the points on a curve.



Anyway, I don't want to get ahead of myself.  And matching blades to generators doesn't have to be very complicated.  Certainly not as complicated as it may have seemed in my previous postings.  Really, it is plenty to focus on reasonable blade diameters to get the power range right, and reasonable TSR to get the blade speed right.  By "right" I mean a close match between the generator and blades you will use.  I went to the trouble of measuring the power curve with about as complicated a set-up imaginable - but I was after a very rigorous answer and I had the resources easily at hand.  I didn't do that for my first two wind turbines, and I won't likely do that again since I don't work at the same shop any more.

Here is a place to start finding out about the power that a wind turbine can collect:
http://drømstørre.dk/wp-content/wind/miller/windpower%20web/en/tour/wres/enerwind.htm

Yes, there was trial and error in my approach - it was by building several wind turbines and learning from them what I wanted to do better the next time. 
By the third WT, I wanted to know more, and started working through the math.  Eventually I had the formulas worked out so that I could generate curves, like the ones I posted above, for any theoretical combination of blades and generators.  I'm not saying you need to do that.  Taking the rough sizing and measurements that others have used is a much better way to get started.

Back to testing generators:  I had a conversation recently with another member that you might find interesting: 
http://www.fieldlines.com/index.php/topic,149388.msg1042987.html#msg1042987

I sketched out a rough set of blade-matching calculations to show what TSR means and how it's calculated, and DaveP68 has been testing generators he makes by converting motors that he gets out of washing machines.

[Kyle]

Thanks for that wealth of information. It is going to take a bit to process it all.

I do have a question for you on the height of anemometer. In a video from one of the links, you talk about it being different wind speed because it is mounted half way up the tower. If you mount it right on the generator, does the activity of the blades distort the readings or is that the ideal location?

[SparWeb]

Yes, read as much as you like!

The wind at high altitude usually flows smoothly.  Closer to the ground it gets stirred up by buildings, trees, fences - any obstacle messes up the flow.  Not only does it get stirred up, but it is slowed down.  Really even the presence of a perfectly smooth ground slows down the speed of the wind.  So the higher up, the faster the wind, as a general rule. This ignores local spots of gusts, accelerations in certain places between obstacles, etc. but it's a good rule for everyday use.  When setting an anemometer, its height determines how deep down in the mess it is.  You can expect the wind to be faster above the anemometer, and slower below it, most of the time.

There is also the case of setting an anemometer in the wake of a turbine, but that's a separate issue.  That's done when there's no better spot to put the anemometer, and since it's usually manufacturers of large commercial turbines who do this, it's trivial to compute the correction factor for the turbine wake since they have all that modeled by computers for the blades anyway. 

For us mortal folk, putting the anemometer a little lower on the pole and scaling the wind speed a little for tower height is fine.

[mbouwer]

To be able to build a completely reliable safe turbine
I think a team with various areas of knowledge is needed.
I would like to be a member of such a team.

[hiker]

Team...Fieldlines...would be a great start...😜

[DaveP68]

On the other side, the question is "how much torque do the blades deliver?"

There's a graph for that, too

Hi Steven

I'm impressed with the "Matching a Wind Rotor to a Generator" graph above. Can I ask how you collected the data points?

David

[SparWeb]

Hi Dave,

Warning:  the following material contains geekiness which may be shocking to some.

The graph is the result of a self-education I gave myself on a subject called blade-element theory.  If you read this page, you will get more of the context in which I created it:  http://www.sparweb.ca/1_Blades/Aero.html

Once I started making turbines, I found that I didn't really know what matched the power in of the rotor to the power out of the generator.  Threads on Fieldlines written by the late Flux were very valuable to get me started, and I wanted to know more.  I wasn't actually starting from scratch.  In my day job, I'm an aeronautical engineer.  I had been taught quite a bit of aerodynamics in school, so I had a lot to build upon.  What I didn't have was a way to get a hold of a rotor that takes power OUT of the wind.  All you get as an aero engineer is the mechanics of aircraft propellers that put power INTO the wind to create thrust.  There are some crucial differences that made it hard to just open an aircraft textbook and point at an equation or a graph.  If you take the graph of aircraft propeller performance, and try to plot a wind turbine rotor on it, usually you are below the x-axis and to the left of the y-axis, guessing how to extrapolate the lines in a region they weren't meant to.  So I kept looking.

I found a number of excellent technical papers on the NREL website, which were written to teach the theory of WT rotor blade performance.  Then the coin dropped.  I was able to use what I'd learned before and write out the equations of propeller performance start to finish.  I could take the input parameters of blade length, chord, twist, and so on and turn it into a performance curve for varying wind speeds and shaft speeds.  I could control the mathematical model like the TSR is fixed, or the RPM is fixed.

This was the most useful result of that work.  You can find this graph in lots of wind turbine textbooks, but I hadn't found any that did the full derivation before I did it for myself using those NREL papers.


It was RW Hamming who said "the purpose of computing is insight, not numbers".  When I started plotting this graph, I finally had my moment of insight.  The graph actually looks like the lift-curve of an aerofoil, which you might recognize, being a pilot yourself, but facing to the left not the right.  On it you can see where the peak power of the rotor is, where it stalls, where it spins so fast it generates no more power... the perfect graph for power-matching with generators.

Power matching is where the black lines on the upper graph come in.  The black line in the middle isn't a generator's power curve, it's the connection of peak power points of the rotor as the wind speed increases.  If you were to build a generator whose shaft input power follows that curve, you would have built the "perfect" matching generator for those rotor blades.

[SparWeb]

PS:
While I was writing that, I meant to link to the NREL webpage that I used for sources.
It seems to have been moved behind a user-login page.  Well, it's been 10 years...

The key papers can still be easily found:
"Applied aerodynamics of wind power machines" by Wilson, Lissaman, Walker
http://ir.library.oregonstate.edu/jspui/handle/1957/8140

"Aerodynamic performance of wind turbines" also by Wilson and Lissaman
https://nwtc.nrel.gov/system/files/WilsonLissamanWalker_AerodynamicPerformanceOfWindTurbines(1976).pdf

(strangely I can still find with google the PDF on the NREL's site, but I can't follow links past the login to the page where these documents are listed & linked)

[DaveP68]

Hi Steven

Nice work and I like your very informative web page and other links useful resources.

Can I ask how measurements made at Altitude = 3000 feet compare with those made at sea level?

For my crude understanding on recovering energy from the wind the amount of energy that can be recovered at 3000 feet compared to sea level will be about 26% less?

David

[SparWeb]

26%? I don't think it's that bad. But density is lower... I can work it out...  stand by...

[SparWeb]

9% less at 3000 feet above sea level.

My "pressure altitude" is 91% of sea level, given the same temperature.  But it's often so friggin' cold here that maybe it comes out even!

[DaveP68]

Got my reference wrong relating to power. What I was really referring to was "Air Density - ρ - (10-1 kg/m3)" which is 12.25 at sea level and 9.093 at 3000 feet. That can make a big difference to a wind turbine performance. Most are located at sea level as they perform much better there than they would at 3000 feet.

As an aeronautical engineer you will understand the take off performance of an aircraft on a runway is better at sea level than it is at 3000 feet. Most major airport runways are located at sea level.

Yes the temperature is an important factor relating to air density, but is it cold at that 3000 foot location 365 days of the year?

[SparWeb]

I probably have the same ICAO charts as you.  But I think you are reading the data for 3000 meters.  I'm at 3000 feet ASL which is about 900 meters.

[DaveP68]

You are correct as I had typed into Dr Google readings for 3000 ft and forgot to look at results were in m, so yes I'll give you that one. Even Google can get it wrong try it yourself.

We all make mistakes from time to time. "A person who never made a mistake never tried anything new" Albert Einstein 

But even at 3000 feet there is still a large enough difference in power output over a year. That is the main point I was trying to make re 3000 feet vs sea level in wind turbine power production...

[kitestrings]

Most are located at sea level as they perform much better there than they would at 3000 feet.

That's not true in our area, or many areas I suspect.  Here (VT, NE) the better wind sites are mostly at higher elevations.  I understand this may not be the case in coastal sites, plains, etc.  It is also the reason there is much debate over utility-scale wind siting as issues of aesthetics/view-shed, access, habitat defragmentation, water quality, storm-water run-off, icing, etc. all come into play.

[Adriaan Kragten]

Matching in between rotor and generator is explained in chapter 8 of my report KD 35 which can be copied for free from my website www.kdwindturbines.nl. Matching is normally done in the P-n graph but it can also be done in the Q-n graph. The advantage of using the P-n graph is that mechanical and electrical power of the generator and mechanical power of the rotor can be put in the same graph. Any windmill rotor has an optimum cubic line and matching is optimal if the Pmech-n curve of the generator for the chosen load is lying as close as possible to the optimum cubic line of the rotor. To investigate the starting behaviour, one normally uses the Q-n graphs because at low rpm the power is almost zero but the torque can be rather high.

An example of checking of matching in between rotor and generator using the P-n graph is given in chapter 5 and 6 of report KD 484. An example of checking of the starting behaviour using the Q-n graph is given in chapter 7 of report KD 484. For checking of matching and starting behaviour you need measured characteristics of the generator for the given load and estimated or measured Cp-lambda and Cq-lambda curves of the rotor. Estimation of the Cp-lambda and Cq-lambda curves for a certain rotor is given in chapter 6 of KD 35.