help with chord width/pitch for blade w/ S822/S823

[cslarson]

I came across these profiles (NREL S822 and S823) when researching airfoils specific to wind power generation.  I think the article (http://www.ars.usda.gov/research/publications/Publications.htm?seq_no_115=212729) was highly relevant also.  The full pdf should also be available at some point from http://www.cprl.ars.usda.gov/REMM_Publications.htm.  These airfoil profiles were designed specifically for small wind turbines (1KW - 10KW) back in 1993.  My understanding is that now both Southwest Windpower Skystream (http://www.windenergy.com) and Endurance Windpower (http://www.endurancewindpower.com) are using them in their blade designs.  When made from wood, though, I can understand that they would significantly increase build complexity.  I'm hoping to only have to make one from wood which I will then use as a plug for a fiberglass mold.

I'm building wind turbines in Kabul, Afghanistan with the eventual goal of encouraging a small sustainable wind power industry (our project was instrumental 20 years ago in developing the micro-hydro industry here - arguably much better suited to the country than wind, sure).  We have a wind map now, though, and the improved power profile these airfoils offer may make a significant difference to the utilization of wind here.

Could anyone help me determine what would be an optimal chord width and pitch angle?  I am hoping to basically duplicate the blades from the article mentioned above.  They were 1.67m long and produced maximum power of 1.5KW.  The S823, thicker airfoil, is used for the root section, up to 40%, while the S822 is used for the remaining blade section to the tip.  According to the article they had a linear chord taper, and "a near Glauert twist distribution" (I haven't been able to find out what this means!).  Any help would be greatly appreciated and certainly help speed things up for us (I am admittedly a real novice with these aerodynamic issues).

Thanks in advance,

Carl Larson

RESAP Engineer, IAM

ps. I also posted as a comment to an old posting regarding blades incorporating an S822/S823 profile that can be found http://www.fieldlines.com/story/2004/7/5/113654/9380, but it is old and i'm not sure if anyone will read it!

[SparWeb]

I'm going to set aside my INTENSE curiosity regarding the logistics of procuring materials, acquiring skilled labour, building suitable facilities, and distributing parts, etc, considering the available infrastructure in Kabul!  You are ambitious!

I suppose you're involved with an NGO.

Hugh Piggot has already published manuals geared specifically for your task.  Several designs can be chosen, and scaled according to your needs, too.

To answer your specific question (briefly), the blades of a wind turbine need to be producing lift at a reasonable angle of attack.  Looking at the tip of a wind turbine blade, it should be traveling around the axis at, for example 6 m/s in a wind of 1 m/s.  But half-way down the span of the blade, that's only moving at 3 m/s in the same wind.  If the airfoil at the tip is inclined at its optimal angle of attack, then it won't be half-way down, unless there is some twist to make up for the slower rotational velocity as you get closer to the axis.

Glauert was an aerodynamicist from the early half of the 1900's who figured out an awful lot of the aerodynamics that we consider "obvious" today.  A "Glauert twist distribution" probably means a twist that conserves the angle of attack through the span of the blade, however Mr. Glauert was subtle enough to have also been aware of tip effects, and maybe adjusted twist at the tip to account for that, too.

I don't know why the researchers in the studies you posted would have recommended an S823 airfoil for such small wind turbines (as low as 1kW).  Say the wind turbine you propose to build has a diameter of 12 feet (3.6m).  The chord of each blade (3 blades) would be about 7.5 inches.  At the tip, the chord should taper down to about 5 inches.

At this scale, the difference between the Selig 822 airfoil and an "ancient" NACA is very small.  If you were then to compare the performance specs of these two airfoils, carefully considering scale effects, you would find very little difference in the end result.  It might matter in a bigger turbine, but not a small one.

Worse, the S823 has a nasty concave curvature at the trailing edge, that will be much more difficult to control than an airfoil with all convex curvatures.

If you're going to be responsible for the detailed aerodynamic design of a wind turbine, that people will be investing money and livelihoods into, I strongly suggest you get some aerodynamics books and take a few seminars!

We like to help here, but remember:

Advice given on the internet is worth what you paid for it!

Good luck.

[Ungrounded Lightning Rod]

Great to hear that things are getting powered in Afghanistan!

Your links only give the summary, which doesn't describe the airfoil itself or identify what alternative it was being compared to.  Do you have that information, or pointers to it?

A little googling found the description of the profiles in question.  Stall regulated, so no furling required, simplifying the mill.  (Haven't gone deeper yet.)

[clflyguy]

Carl,

Here are the numbers for a set of blades that correspond to the 1.67m blade length

you mentioned. I assumed an 18" dia. hub- hope this helps.  Gus

[cslarson]

Hi Steven,

Thanks for the response!  About 6 months ago we installed a 500W wind turbine that I built based on Hugh Piggot's book "How to Build a Wind Turbine" (June 2005 edition).  There is now a German and also a (very inspiring - prior to meeting us he had already built his own!) Afghan working on producing more of these for distribution.  Carpenters are trained to produce the blades.  Neodymium magnets are imported from China.  There can be some fun involved finding good local supplies (i'm thinking of using schedule 8, 4" PVC we have here to stiffen up and supply the root shape of my blades).  

Taking build time, cost of materials, and cost of controller, batteries, and inverter into consideration I was generally unsatisfied with the resulting cost/watt.  I am hoping to improve this through the design of a larger, 1.5-2KW machine.  While not all do, some costs scale well.  Additionally, a larger machine might be designed that can be more generically suited to other applications, specifically pumping water.

So, I'm starting from scratch to really learn how these machines work.  As I mentioned, we have a wind map of Afghanistan now, which will give me much more confidence that any new design can be properly targeted for use here.

The researchers were comparing their new blades to those from a Bergey 1.5KW turbine.  Unlike what would be probably be suggested here, these incorporate no twist or taper.

Thanks again, it's nice to be welcomed to the board.

-Carl

[cslarson]

Hmm, through searching I realize this has been hashed over before... (sorry, but at least I have a fairly new reference which attempts to explain things...)

It seems as though one of the things I need to first determine is what tip speed ratio (tsr) I should incorporate in the turbine design.  I thought that maybe this was affected to some degree by the airfoil profile used, but maybe it isn't.  Numbers are thrown around all over the place, it seems, and for a three bladed rotor 5-7 is suggested (although that seems a large range to pick from).  This article (https://netfiles.uiuc.edu/mragheb/www/NPRE%20498WP%20Wind%20Power%20Systems/Aeorodynamics%20of%20Ro
tor%20Blades.pdf) was published 9/8/2007 and has some explanation of the math involved in optimizing tsr.  These calculations suggest tsr(opt)=4*pi/n where n is the number of blades.  So tsr(opt)=4.19 for n=3.  But the following sentence claims that if care is taken in airfoil design, then tsr(opt) may be 30% larger than these values!  Where does that come from?  He uses an approximation that for an n-bladed rotor s/r=1/2 where s is the "length of the strongly disturbed air-stream upwind and downwind of the rotor".

Anyway, according to that article, for a 3-bladed rotor, a tip speed ratio of 5.45 would be appropriate.  But I'm not very confident with that because it seems that the number 7 is thrown around much more!  And how does the particular airfoil chosen affect this?  Why is it suggested that a higher tip speed ratio is necessarily more efficient?  If it is too high then wouldn't the blades be traveling through turbulent air?

Hoping to one day understand.

-Carl

[cslarson]

hi,

thanks for the response.  the blades were compared to those from a Bergey 1.5KW which incorporated no twist or taper.

everywhere i have read seems to suggest that developing a stall regulated rotor is quite aerodynamically complicated.  the twist needs to be just so.  this might be more than i'm ready to tackle, but would be interesting to look into.  trusting something without a furling tail... i guess it depends how confident i become on learning about blade design.  there are certainly a lot of resources out there.

[cslarson]

more literature...

Assessment of Optimum Tip Speed Ratio of Wind Turbines:  http://www.asr.org.tr/pdf/vol10no1p147.pdf

(i had problems downloading that article so i used the google cache of it's html version found here: http://64.233.183.104/search?q=cache:6B54na1rXCYJ:www.asr.org.tr/pdf/vol10no1p147.pdf+optimal+tip+sp
eed+ratio&hl=en&ct=clnk&cd=2&client=firefox-a)

In the conclusion, the author states "Having analyzed the findings of this study, it can easily be said that the optimum design speed ratio for every profile is different rather than being 7 as pointed out in related literature."  The following profiles were analyzed in that report:  NACA 4415, LS-1, Clark Y, Gottingen-398, C-80, M-6, Raf-15 and NACA 2212.  Unfortunately, not the S822 and S823.  It looks like the idea though, is to optimize the power factor by choosing a tip speed ratio optimized for the following losses:  profile losses, end losses, whirlpool losses, and blade number losses.  Because of the google translation of the pdf, the graphs do not appear, some text is lost, and the equations are a little difficult to read.  The following seems to be the driving equation.

Cp = Cpschmitz(λA).ηprofil(λA,ε).ηend(λA,z)

where

ηprofil(λA,ε) = 1 - (λA/ε) with ε = CL/CD (lift coefficient over drag)

ηend(λA,z) = ?  (google cuts off text)

Cpschmitz(λA) = look at table  (google doesn't show diagram, but table is available)

The S823 and S822 are supposed to achieve quite low CD.  It might be safe to assume that profile efficiency (ηprofil) approaches 1).  How much does end efficiency (ηend) affect the choice of tip speed ratio (λA)?  Otherwise, according to the table for Cpschmitz, choosing a larger tip speed ratio would, to a lessening degree (improvement tapers off) offer an improved power factor.  Then you need to consider blade wear, vibration, noise, etc.  Hmm.  I'm beginning to think that picking a number for tsr is to some degree, fairly arbitrary...

[wdyasq]

Carl,

The operation of blade selection goes something like this:

Build turbine base, guess size and number of wire turns for test coil to determine coil size, build stator and test, make blades fit machine.

In the end, TSR, the Tip to relative wind Speed Ratio "Falls Out". Personally, I feel quoting the TSR has about as much to so with design of a set of blades as car color has to do with the seat number selection in an airplane. But, I'm hard-headed. The TSR does give a twist for a blade-set. That twist depends on the carving program chosen, in most cases. But, folks normally carve a twist to fit the 'blank' thickness as opposed to building to the real twist specified by a blade carving program.

The UIUC aerodynamics program has a lot of information and a very large database on airfoils for wind turbines. They also have a computer program 'PROPID' that should help with blade design.

I am of two thoughts on your selection of the Selig 822/823 airfoils. I do believe they should give a better power production for the swept area. They are more difficult to duplicate. The same power may be obtained by using a slightly longer blade-set with an easier to carve profile. Blade-sets are very dependent on alternator power profile. Unless one is building near identical machines, or one is satisfied with mediocre  performance, molded blades may be more trouble than help. my opinion only.

You are correct in selecting a 'low speed airfoil'. Airplane airfoil development has  almost always been into developing lift at higher speeds, the practical limit of airfoils in wind machines is limited by wind speed and rotational velocity. For an easy to carve airfoil in the speed range you may want to look at airfoils like the USA35, used on the Piper Cub or some of Harry Ribblets low speed/high-lift airfoils.

Good luck,

Ron

[cslarson]

Hi Ron,

Thanks very much for the information about PROPID.  I'm keen to check that out.  You certainly make a good point about achieving the same power profile from slightly longer, easier to carve blades.  There are a few other factors, though, that make me want to attempt a blade with the S822/S823 profile.  They are thick profiles that should enable me to make lighter, stiffer composite fiberglass blades.  Being shorter while retaining the same power profile should also help this.

I'm really trying to figure out how we can make many wind turbines as cheaply as possible.  My experience with carving the 1.2m wood ones for the 500W Hugh Piggot design made me immediately want to tackle this first.  Once I have the fiberglass mold, for which I should only need one wood plug, making the blades from fiberglass will hopefully be quicker, need fewer tools, be easier to train someone to do, have more easily repeatable results, be cheaper, and may also result in lighter, longer lasting blades (I think the wood we have available here is a little softer than would be ideal).  If I only need to make one plug from wood, then I can take a lot of care and time in making it.  Using templates it shouldn't be too difficult to eventually get a more complicated shape like the S822 and S823 employ.

I am aware of a few of the blade carving programs available.  One is an excel spreadsheet that is available from Hugh Piggot's site (i believe).  I want to understand some of the assumptions being made in these programs in order to be more confident that they are appropriate to use for the two profile blade I am attempting.  Basically, if a number of these are going to come out of a mold, and be installed around the country, I want to have some confidence that they are half-way decent.  So, sure, there's a lot for me to learn

[cslarson]

The UUID aerodynamics website also had wind tunnel data for the S822 airfoil.  That should help with determining appropriate pitch.  Thanks very much for the useful reference.

[SparWeb]

Carl,

I don't want to sound negative, but your focus on the Selig series of airfoils is the wrong approach.  If your purpose is to justify the use of this airfoil at all costs, then you can probably concoct something, but the analysis will disregard the performance of the final product: a wind turbine.

Start with the basics:

You want to produce "X" kW in a wind "Y" mph.

Power production must start at "Z" mph wind.

I want said wind turbine to survive winds up to "A" mph.

The turbine must cost "B" dollars per unit.

The turbine will provide power to a) local grid, b) national grid, c)local battery, d) some other local consumption, eg, heated water.

...et cetera...

There are a dozen more questions that an ENGINEER should be asking BEFORE picking some detail bit like airfoil shape.  You don't even know how much torque must be overcome in the theoretical generator to start it as the wind speed increases.

Note that the questions I asked above didn't even cover the basic specifications like diameter, voltage, current, frequency, mass, material selection - these cannot be selected until you have studied the most basic economic questions.

Sorry for the lesson, but there are a lot of things to consider.  This site is great for learning about the details.  I love that part, too.  But few here are concerned with economics.

[Ungrounded Lightning Rod]

Build turbine base, guess size and number of wire turns for test coil to determine coil size, build stator and test, make blades fit machine.

In the end, TSR, the Tip to relative wind Speed Ratio "Falls Out".

Expanding on this:  TSR and choice of blade profile are details, making a modest change to efficiency.

But for a wind turbine, efficiency is not a major issue.  Unlike consumable fuel, which must be paid for, wind is free for the collecting.  So turbine efficiency, like horsepower in a Rolls Royce engine, merely has to be "adequate".  If it's a little low, make the mill a little bigger.

There is a broad variety of easy to calculate, easy to fabricate, blade designs that will get you to (or above), say, 80% of the Betz limit for the turbine proper.  It's impossible for ANY blade design to go above Betz.  So a "perfectly efficient" blade would only get you another 25%.  You could get more additional power by sticking with the simpler blade profile and increasing the blade length by 12%.

Similarly, extreme precision when constructing the blades is also not an issue.  The blades just have to be strong enough to hold together, balanced well enough that you can finish that job with some counterweights, and kinda close to a decent profile.  If it's a little off a good profile the main effect will be a minor reduction in efficiency.  See above.

[Countryboy]

I'm really trying to figure out how we can make many wind turbines as cheaply as possible.

If that is your goal, then you are asking the wrong group of people.  You should be following the proven business model provided by the Chinese.  They pretty much have the market cornered on cheap junk.

You also need to remember that cheap in equals cheap out.  If you want a quality product, it isn't going to be cheap.