Airfoils

[Murlin]

These two airfoils seem to be a couple of the best ones I have found so far for HAWT's.

The NREL S822/S823 are for machines our size.

This one is designed to help with the leading edge debris problem.

But as you can see in the picture both chords are the same width.

The article suggests that you keep the thickness scaling the same or you will loose performance.  

What about the width scaling.

The blades that I have seen on this board are alot wider at the root than at the tip.  And the thickness between the tip and the root is very different.

Would the NREL design work with axle flux alternators?

I have not seen anyone using this airfoil here, I suspect that it is mainly a manufacturing issue as to keeping the design workable, since most of the carving is done by hand.

What does it mean in the picture at the bottom of the airfoil:

40% radius?  

Thanks for your time,

Murlin

[Flux]

Naca 4415 should work well.

Not sure about the others, I suspect they are intended for stall regulated machines.

Make sure you have your data for windmill blades, much of the aeroplane stuff doesn't work very well for windmills.

Model aircraft sections are usually right for Reynolds number but the fancy ones are too critical on surface finish.

I think Samoapower will likely be able to help you , he seems to be into this sort of thing.

Unless you are running at high tsr and are using mppt and don't mind a bit of noise I dont think you will see any improvement over ClarkY and freehand approximations to that seem good enough for most of us.

That was good enough for Jacobs. I don't find that the very wide roots with extreme angle provide any benefit, but a good aerofoil that has genuinely been proven for small wind power is likely worth the effort, Bergey and SWWP have spent money on blade profile research and their blades may be worth a look at.

Just picking aeroplane aerofoils on wind tunnel data doesn't generally give any benefit and some are definitely not very good. Air flow is different and Reynolds number is often way off.

Flux

[vawtman]

Flux i agree and still cant figure out why Naca profiles are even thought of for wind turbine blades.Vawt or Hawt.

Were dealing with a whole different animal.

[Murlin]

'I suspect they are intended for stall regulated machines'

Ahh...You are right. I didn't understand that those are different than Axel flux genny's.

Stall regulated.   Is that where you design a genny to where just the airfoil and the load do the braking,  no furling?  I notice that the very big genny's just turn slow and peaceful no matter how much wind there is.  There has to be something to that but I have not read enough about it yet.

I was Googling "wind turbine blades", but am having a hard time sifting through the data, there is so much for a noob.  

Trying to model up my blades... doing the R&D.  A small change of airfoil could go a long way. So I wanted to be really sure what I wanted to machine.

Any shape is possible and can be done very accurately.

The NACA 4415 is about what we have.  It is almost flat on the bottom.

So I will just take that one and scale it to the thickness, width's, and angle of attack for a 20' machine.

I think Danb's Dims for his 20' are.

10. ' long
    
3. 1/4" at the root
   
15. " wide at the root
    
8. " wide at the tip
   
6. degree angle of attack at the root
   
3. degree angle of attack at the tips
   
   

Unsure about the the scaling at the tips on the thickness

If a flat bottom is what is needed, then I can make it flat or give it a slight concave scallop.

It is just as easy to do either one, but do not have the data I need to draw them yet.  

I have read that the tips do all the work. Is the changing angle of attack, constant, from the root to the tip?

Murlin teh questions haven't found answers too yet....

[wdyasq]

The leading guru on HVAT airfoils is a fellow named Mike Selig at the University of Illinois at Urbana-Champaign.

http://www.ae.uiuc.edu/m-selig/

They claim there is 35% more energy over a long period using the proper airfoil over the Clark Y or NACA44XX series. UCIC calls a 10M turbine "small". s809 is one high recommedation by the UCIC group.

It doesn't matter what airfoil anyone who has researched this field recommends. The builder is going use WTF he choses with little or no regard to knowledge or research. Proper monitoring will not be maintained and in the end they will state,"It is all free anyway, what does it matter, I'll just build another windmill..."

Ron

[richhagen]

Hi Ron, I have a number of 4415 blades that I have carved, and have a couple of testing rigs with matching alternators now, I have the s809 data, so some day soon, I will carve one up for comparison.  Rich

[Murlin]

These are the Windmill airfoils from the GURU. 2 of them are almost what everyone is doing now except that they are thinner.

One thing that keeps popping up from all the data from all the testing I have seen, is the little scallop 3/4 the way back that brings the bottom above center, then comes back either flush or below.

This is what caught my attention and it is the reason I am looking at airfoils.

35. % is alot.
    
10. % here, 10% there, adds up.
    
    

Murlin

[Murlin]

So you are saying you shouldn't go by Hugh Piggots book?

I did notice that the only difference from what everyone is using and the Naca, is  that the leading edge is sharp on the front side, instead of having a radius.

Murlin

[altosack]

Murlin,

Stall regulated means that past a given wind speed, the generator drag will exceed the driving force from the available lift, and the prop will go into stall, without enough driving force to maintain the nominal TSR, and will slow down dramatically. The airfoil itself will have an abrupt stall, meaning that past a certain angle of attack, the lift will decrease dramatically, and the drag will increase dramatically.

This has little to do with whether it's an axial flux machine or not; an axial flux machine could be designed to do it, however, many times it is a generator with a controlled flux in the stator, so it is an "active" stall, done by electronics.

This could also be done by a boost converter with an axial flux generator, even if you don't have an airfoil with an abrupt stall (not recommended since it plays hell with feedback). If it's designed well, and is in "soft" stall (maybe 85-90% of design TSR at the design maximum speed) without any boost, adding just a little boost will quickly stall the machine and idle the rotor; I think the Bergey XL does it this way. However, furling as a backup is a doggone good idea !

You wrote: "Is the changing angle of attack, constant, from the root to the tip?"

Actually, the blade angle (not the angle of attack) is changing from the root to the tip, in order to keep the angle of attack constant. The wind the airfoil sees is a combination of the real (axial) wind speed and the head wind, from the rotation of the rotor. Since the tip is traveling much faster than the root, the combined angle of the real wind and the head wind (called the apparent wind) changes angle from root to tip. This is explained quite well in a file called bladeDesign.pdf from Hugh Piggot's site (I don't have the URL handy).

Dave

[wdyasq]

Well Murlin,

As Selig and group probably has about 100 or so airfoils attributed to them I don't doubt it. I've long thought the DAE series has some good properties. I think one of the big things improperly done is designing for a high "Tip Speed Ratio". Props are more efficient running slower according to most literature I have read.

Well, there are many publications available on the subject. Many are online. Good luck.

Oh, the DAE series were developed for the Human Powered Vehicles - airplanes and high speed pedal boats.

Ron

[Flux]

Rich

This would be very interesting, at least it would be a reasonable comparison.

Even then it may not be a fair test unless you can alter the load characteristics of both alternators to achieve the best performance from both.

This aerofoil issue is not going to be solved until someone does it with the ability to take accurate measurements with quality data logging and has the ability to understand the results.

Data and blade sections from big machines may or may not work for small ones and until you can measure the difference it is pure guesswork.

If things are as critical as some believe, why is it with a blade with tsr of 8 you don't see any difference between a newly made blade and one with the leading edge chewed up and eroded from years of use.  I suspect that the difference is there but no one has the ability to measure the difference because of the cube law nature of wind.

With your 2 machine approach, if you can match both blades properly then the total power produced in a given time from each would be as good a comparison as we are likely to get.

I have the basics needed to plot power curves and this could be used to do this work but I don't have a good site with decent wind in all directions and it would take me years to get these results. It takes weeks sometimes to get one power curve and none is ever genuinely in wind free of turbulence.

Unless you can get your machine up 30 ft beyond anything for miles around this will have drastically more impact than any fancy aerofoil.

Flux

[SamoaPower]

Flux,

"Even then it may not be a fair test unless you can alter the load characteristics of both alternators to achieve the best performance from both."

This is certainly true. But why complicate the issue by putting the alternator into the equation when we want to know about the rotor? Another variable doesn't make it easy.

"This aerofoil issue is not going to be solved until someone does it with the ability to take accurate measurements with quality data logging and has the ability to understand the results."

This has already been done and is continuing. Should we ignore the work of the professionals in the field, with the talent and resources to do the job right? Organizations such as NREL, UIUC, Sandia Labs and others have already done considerable work that relates directly to small machines (under 20kW). Are we arrogant enough, as DIYers, to say we don't believe the work of others unless we do it ourselves? I, for one, don't think so.

"If things are as critical as some believe, why is it with a blade with tsr of 8 you don't see any difference between a newly made blade and one with the leading edge chewed up and eroded from years of use.  I suspect that the difference is there but no one has the ability to measure the difference because of the cube law nature of wind."

One explanation could be that the alternator is mismatched to the rotor and is still overpowered by the rotor, even in its degraded condition. Also, different airfoils respond differently to leading edge degradation. One of the reasons the S822 and S823 airfoils were developed was to reduce sensitivity to this.

[SamoaPower]

 Congratulations murlin on starting at the beginning by examining airfoils. You may be one of the few.

"The article suggests that you keep the thickness scaling the same or you will loose performance.  

What about the width scaling."

When you scale an airfoil, you need to scale all dimensions proportionally using the same factor. Otherwise, the shape would change and you would have a different airfoil with different performance.

"Would the NREL design work with axle flux alternators?"

Rotor design is pretty much independent of alternator/generator type.

"I have not seen anyone using this airfoil here, I suspect that it is mainly a manufacturing issue as to keeping the design workable, since most of the carving is done by hand."

You're probably right. However, I think most airfoils can be reproduced using templates, work and patience. Choice of materials also has a bearing.

"What does it mean in the picture at the bottom of the airfoil:

40% radius?"

They're suggesting a two-airfoil rotor blade. The S823 for the first 40% of the blade radius and the S822 for the remainder.

The next step is to look at the performance data on potential airfoils. The main factors to be concerned with are the Lift to Drag ratio (L/D), the Angle of Attack (AOA) at which max L/D occurs and the Coefficient of Lift (CL). You will need these numbers to proceed with the design.

[SamoaPower]

And that difference completely changes the airfoil performance.

With all due respect to Mr. Piggott, I think there are better airfoils.

[SamoaPower]

Altosack takes care of your questions on 'stall regulated' and AOA just fine.

Yes, don't confuse AOA with blade angle, they're different animals.

I think you're under the impression that you can take a given airfoil and use it with the blade dimensions of someone's rotor. Sorry, you can't do that. Part of those dimensions come from the performance data of the airfoil. You need to start with the airfoil data (and other factors) and then calculate the dimensions. It would take several pages to explain the whole process which is probably why most just copy what someone else has done, right or wrong.

"If a flat bottom is what is needed, then I can make it flat or give it a slight concave scallop."

Whoops! Don't think you can do that. That 'scallop' is called undercamber. Adding it to a published airfoil that doesn't have it, completely changes it. You then no longer have known performance numbers to go by. ANY change in an airfoil's shape alters the performance.

[SamoaPower]

"These are the Windmill airfoils from the GURU. 2 of them are almost what everyone is doing now except that they are thinner."

ALMOST is the operative word in that statement. I'm not sure that you appreciate how little difference in shape it takes to significantly alter the performance. That's why there's hundreds of airfoils. If you want to emulate a given airfoil, it has to be reproduced exactly. ALMOST only counts in horse shoes.

"One thing that keeps popping up from all the data from all the testing I have seen, is the little scallop 3/4 the way back that brings the bottom above center, then comes back either flush or below."

Undercambered airfoils are generally considered to be high lift airfoils. L/D of 100+ is possible. Use one myself. Their thinness sometimes presents construction problems in making them rigid enough. Depends on materials and techniques.

You're right, 35% is definately worth going after.

[wdyasq]

Gee SamoaPower,

You have doen a great job of explaination, Thanks.

I will say even the GURUs of the design game don't agree on "what's best" and as they are spending "Public Money" - my tax dollars, I don't guess they want a real result as it would end funding.

One can get bogged down in the research on the best rotor airfoil, twists and cords. At some time it is required to lock the design and start cutting material.

It is quite probable a Variable Pitch mill might be more forgiving on design. The VP mill has the ability to change angles and stall is not as critical. But, a lot more expense and complication enters the equation.

Ron

[Murlin]

The 4 airfoils I have shown from the link you provided are for SLOW RPM Wind turbines, not fast racecars, airplanes, ect.

I am not sure what TSR they are designed around.  You guys use what a TSR of 5 to 7?

I do not think you can hurt anything by thinking about stuff before you do it.

In my business, you can waste thousands of dollars if you use the "cut first, ask questions later" approach.

Murlin

[Murlin]

Thanks buddy for taking the time.

I do see the small differences, and understand.  Sometimes I cannot put down on paper the right expressions.

Using the My CAD program makes what seems to be complicated stuff, easy by comparison.

I can take a picture from an airfoil on the internet and run it through a rastor to vector tanslator and duplicate the thing extremely accurate, very easily.

It is just as easy to machine a clark Y as one of the other airfoils.

One just needs to know which one to use.  

I guess that's the million dollar question...

From Hugh's book, you can use the chord calculations in the 5 stations, based on rotor dia.

One could just use the airfoil and scale it up or down and just use the width calculations and keep the same airfoil along the blade.

So Sammo, what TSR to use..I have read alot and find that people are using a TSR from 5 to 7.

What would your opinion be of the correct TSR to start with?

Murlin

[whatsnext]

"Are we arrogant enough, as DIYers, to say we don't believe the work of others unless we do it ourselves?"

Yes, many here are. Just do a search on 'engineers' here to see what I mean. The number of failed ideas that many here suggest we 'try for oursevles to see' really astounds me. I love this place but some of the things passed off as engineering are really quite frightening, or comical depending on my mood.

John...

[Victor]

"From Hugh's book, you can use the chord calculations in the 5 stations, based on rotor dia."

The chord width at each station is keyed to the coefficient of lift at that station which is affected by the relative angle of attack and the airfoil polar.

 In other words a turbine using an airfoil at an angle of attack to produce a higher coefficiet of lift, when properly designed, will have narrower blades for a given TSR than one designed to run at a lower coefficient of lift.

[SamoaPower]

Thanks Ron.

I certainly agree with your comment about variable pitch. That additional degree of freedom allows much better control of our machines. That's why I chose it for my 16' machine.

[finnsawyer]

The problem with wind tunnel data, as I understand it, is that it has been fudged to give the correct results for air plane wings.  What say?  At least in the early days the wind power studies were done with small dimension air foils, which more closly model a wind turbine blade than an airplane wing.  Seems airplane wings operate in a laminar fashion for six inches or so of the span, and then in a turbulent mode, whereas the wind mill blade and the test airfoil may operate completely in a laminar mode.  Does this mean one needs to evaluate windmill blades in a dynamic fashion, perhaps by building a circular wind tunnel where the windmill diameter is nearly the same as the tunnel, so that certain distracting effects are eliminated?  This could probably be done as accurately as any wind tunnel test.  I certainly would be interested in the results of such testing.  But from the stand point of these home brew wind mills, I agree that one shouldn't worry too much about the air foil shape.  There are other compromises that are made that affect performance.

One reason that damage to the leading edge of a blade doesn't have much effect is that most of the incident air never reaches the edge.  It is deflected before it gets there.  Basically, a cushion of high pressure air on the leading edge moves with the blade and smooths things out.

 

[Stonebrain]

Hi,

Some newbielike intuitif thoughts about airffoils.

It might be that

1)Thick,basic nacashapelike,more symmetric airfoils will be moderately efffecient

  at a wider range of tsr.(Thinking that aoa changes with tsr).

2)Thin,sophisticated,cambered airfoils can have a very good efficiency,

  but for a smaller range off tsr.

Taking in acount the alternator limitations,you might want stallregulation

it's clear it can be better to have some sophistacation.(better performance over

a narrower range)

But iff you would have a alternator without limitations The overall poweryield

might even be better with a simple,fat naca-airfoil.

It's all a question of prop-alternator matching.

How much information is available (for free I mean)to figure this

kind of things out,I don't know.I think,the experience of some gurus

is very useffull.

Probably no one off the gurus did a expensive research program to find

these things out.

Private investigators won't give away their work for free.

The good thing about people working with taxmoney is that they do.

So people who can use their work will get more back then they ever

payd taxmoney.

ok,shoot me down now.

cheers,

stonebrain

[SamoaPower]

Sorry murlin, I may have been a bit heavy on the airfoil shape thing, but I see so many posts that indicate that many believe that an approximate generalized airfoil shape is good enough. Maybe it is if they're not very concerned or interested in rotor efficiency.

Rotor TSR is another one of those factors that fall out of an integrated system design. The various elements of a system, site, rotor, alternator, load and needs are all interrelated.

Ron had a good point earlier about lower TSR being more efficient. I have a curve that says that for a three=blade rotor, efficiency peaks at a TSR of about 4.5. This also results in a less frantic machine that stays together longer. Lower TSR is also more tolerant of blade imperfections.

Needed TSR depends on alternator requirements. The usual reference point is alternator cut-in RPM and windspeed which is a result of alternator design considerations such as poles, magnets, wire and the like and the chosen wind speed. A lower TSR requires a larger alternator for a given power output.

So, it becomes a chicken and egg game - who comes first, the rotor or the alternator? In reality, the site and needs actually come first in the system design, but for this discussion, I would choose the rotor because I want the efficiency.

To answer your question, I would choose a TSR of 4.5 and design the alternator and rotor around it.

[Flux]

In the end you have to make a choice. The lower tsr will likely get you better rotor efficiency but you pay a serious price in the cost of the alternator to keep the same efficiency. If you want to use a power matching scheme rather than the usual resistive matching it already makes the alternator larger and more expensive.

If you drop tsr to 4.5 and want good electrical efficiency then  you have a big and costly lump.

If rotor efficiency is your aim above all else then you live with it. I would have gone for about tsr6 and aimed for the smaller and cheaper alternator. Only you can make the final decision.

If cost and complexity counts for nothing then you can likely have the best of everything.

Flux

[ghurd]

I can not make 3 blades the same.  It just won't happen.  Surely I am not the only one.

Just because someone knows what is best, doesn't mean it is actually possible for them to do it that way.

Is a very high efficiency blade set more susceptible to imperfections, like the blades being 'a little different' from each other?

G-

[SmoggyTurnip]

What I think will happen if you design for 4.5 TSR.

You will build a big alternator that produces lots of power at low RPM.

Then you will build a set of blades to match it.

Then you will put it up in the air and gets good power and efficiency.

They you will say - but if I turn it faster I will get more power.

Then you will make a new set of blades with higher TSR.

Then you will get more power but less efficiency.

Then you will be happy (maybe).

.

[Flux]

Ghurd, it's all relative.

Some have the facilities to computer design these things and NC machines to produce them perfectly. They still have to decide what exactly to produce.

If you are investing a lot of time and effort into a 20 ft machine and have these facilities it makes a bit of sense to at least try for a tested aerofoil. If you have faith it will work better even if you have no means of proving it.

There is no doubt that perfect blades will work better than rough ones, if they are not all the same then they can't be the same aerofoil by definition.

With small things such as you make, other factors will have more effect on the output than the aerofoil section. If it is terrible then it may not work too well but if it is half reasonable then it will be ok.

It is perfectly true that some of the profiles need to made with extreme accuracy and need to be kept in near perfect condition. They are fine for gliders, model planes and man powered flight.

Some of these also seem good for wind power if you can sort the tolerant ones from the fussy ones.

While looking at one of the links to Selig's site I found that one of these highly prized sections is used on the Air 403. It does indeed put up an excellent performance at 30 mph, but the thing doesn't go round at 10 mph.

This is no criticism of the blade, just a wrong design philosophy for a land based machine, no doubt excellent for a boat at sea with good wind.

One thing that has surprised me in all this discussion is that after all the scientific discussion on the perfect shape we are advised to keep to tsr 4.5 .

I can believe there are real differences at tsr 8 and above between hand made things and proper sections. If we have to drop to low tsr to see any advantage then I expect your three sticks all different will still work quite well.

If DanB had taken a lot of trouble to make ideal blades for his 20 ft machine i don't think anyone would have seen a difference. How you extract the power and load the blades will make more difference than anything else. Siting and tower height comes next, then when these are all right then you can worry about blades.

If you are absolutely desperate for the maximum power in a 30 mph wind and can use it all or if you have a grid tie system then it makes better sense.

Where your interest and skill lies is largely a factor in what aspects of wind power you concentrate on.

No final answer, bound to offend someone, adopt the Zubbly approach, just do what you find fun.

Flux

[ghurd]

Good call.  

I can, and do, add a large percentage of power with another 2" of crappy blade.

"sort the tolerant ones from the fussy ones"

I assumed all high efficiency blades were fussy ^2.

"do what you find fun" is the best advice I have seen in quite a while!

[SparWeb]

I got a feel for how airfoils work in college aero engineering classes, and put the experience to use at work afterwards.  Concepts like Reynold's number, boundary layers, and induced drag are hard enough to handle on helicopter rotor blades, so trying to use the same aerodynamic tests for a windmill drives the ignorance factor up another factor of ten.  Even though you CAN work it out, (I've tried) it's not worth doing so, because the number of realy delicate variables is too high.

Putting it into real terms: if you have a 12 foot span, then the chord of your blade is going to be about 12 inches.  A 15% thickness at the root means your blade is 1.8 inches thick.  An error of 3% (which would totally throw off your aerodynamic prediction) equals 1/16".  That's pretty hard to maintain with wood and hand tools.

In my opinion, it's most important to get the length and taper of the blades to match (easy enough, right?),

Second in importance is getting the angle of attack exactly the same on all blades. (not so easy...),

Third in importance is to buff and polish until it's smooth as glass. (that's all elbow grease),

Fourth is having a consistent airfoil shape.

I'm with Stonebrain.  Build what you can build, make it as smooth as possible, with no sharp edges (except the trailing edge), and it'll fly just fine.

[richhagen]

Ron, there is a lot of truth in what you said there.  I find that I become stretched between obligations and things I would like to do.  I only have had so much time for trying to build a better (at least for me) moustrap.  I have thus far built relatively simple designs because those are the only ones I have finished thus far.  

The difference between a good blade and a perfect blade is probably often less than a 5% difference in blade diameter, if one adjusts the alternator to match.  I have just been trying to build better smaller blades than I would otherwise be able to.

Rich

[wdyasq]

"Rotor TSR is another one of those factors that fall out of an integrated system design. The various elements of a system, site, rotor, alternator, load and needs are all interrelated."

A mouthful of correctness. The TSR term gets thrown around like rice at a wedding and means about the same. What are you going to do with all the damn rice. What you are trying to do is match speed to power and a few other variables. AFTER the fact we get a "Tip Speed Ratio", maybe.

My opinion is design first - calculate after it is flying. Want to talk - talk TSR. It don't mean a damn thing but make one appear as if they do. like a politician.

Ron