Solidworks 3D model of prop

[SparWeb]

Careful guys - you can have so much fun with the software that it leads you to think that executing a drafting command is equivalent to constructing something properly.  Take it from someone with over 5000 hours of CAD on various platforms.

The models you are producing will give limited results.  There are some aerodynamic principles that only require trigonometry and a piece of paper to figure out.  You will then see that a simple twisted extrusion in CAD is not what you want to build.

[BLACKBOARD=ON]

Looking at the tip of the blade, the leading edge must be straight overall, and the trailing edge must look "curved".  Here is a diagram below of the two different ways of making the trailing edge:

I drew the straight trailing edge in red.  It's not supposed to be this way.  The airfoils aligned to the red line will be at the wrong angle of attack.

The blade should actually have a "non-uniform" twist to account for the relative change in angle of attack at different stations along the radius.

There is a small range of angles of attack where the airfoil gives the most lift for the least proportional penalty in drag.  This is called L/D-max.  If it's 5 degrees or so, then you want all sections of the blade to be working at 5 degrees angle of attack when turning at the target TSR.

[BLACKBOARD=OFF]

Class over.  Go do your homework!  :^)  

I wouldn't be so picky on a small windmill project, but you guys are above the "toy" threshold, so you deserve to be warned.

[DanB]

Matt - thanks for the reply.  Im not going to get much deeper here, you could read what I would say all over the internet.

Millions of dollars and man hours have been spent figuring out what works and what doesn't.  There is a good reason most wind turbine blades look the way they do.  (actually several)

In my opinion you misunderstand how these things work - just my opinion and I mean no offence - but...  being that this is 'my' discussion board Ill always be quick to jump in when somebody puts fourth what I believe to be misinformation.

[MattM]

Nobody is disputing this.  The blade also acts as a lever and the center is the fulcrum.  The useful pitch towards the center is going to have to be much steeper in order to match the effective leverage.  But where is that in the magic formula?  I've seen a lot of formulas that like to ignore this point.

I also never see the coefficiency of fluid friction factored into the blade design.  This is another aspect of theory that has been shown a lot of research money.  But it too is ignored.

This argument is over granularity.  I've thought it out there, guy.  But modern science makes it possible to think outside the traditional box.  I'm only challenging people to look beyond the current set of beliefs and explore new, unchartered territory.  Get past the Dark Ages.

[MattM]

I was hoping you didn't take my post as an insult.  I recognize your hard work and like what you've done with your web site.  Surely I wouldn't be alone in saying thank you for putting this information out there.  Its very informative and you have a lot of gems buried in the archives.  I can only imagine how much hard work you've put into it.

[Boss]

"Can your friend Kevin create a basic airfoil profile sketch in one plane at the root, copy it to the tip (and scale it down), and then produce a blended surface between them?  That's not quite the correct way to do it, but if he can do it for the whole blade, then he can do so for small span-wise segments of blade at a time - to get the twist right."

Kevin says heck yeah he can do this.

I don't know what we are going to do with the models at this point. The college does have a 3D printer. The sample output of this device is a small working plastic monkey wrench. They also have a cute little milling machine which interfaces with SolidWorks. So far we have made small blocks which look like linoleum block printing blocks. Not sure what we can do with this yet either.

The idea at this point is to use the 3D model of the props for educational purposes.

When we have built a dozen or so turbines we intend to have workshops open to the public. We checked out the CNC plasma cutting machine the college has in the metal shop, very cool. We plan to try it on cutting out some rotors so we can get started next week at least making an AF alternator.

[TomW]

Brian;

Please follow board image rules. I think  Kurt mentioned it before.

This one is 800X600 much too large.

Thanks.

TomW

[ghurd]

His knowledge is from an unloaded 5' 100 RPM at 8 MPH AL sheet metal prop that spins in a very low wind, and it will be more cost effective per watt than any wood blades because AL is free.

And they won't break off because... I dunno... The MEs for the Comet and Boeing are complete idiots?  That's my best guess.

Last I recall, he was going to make a stator in a couple months.

[finnsawyer]

I don't like this talk about a vacuum.  There can be no vacuum in the atmosphere, only regions of different pressure.  Any blade or air foil reacts and responds to the pressure distribution around its surface.  There is nothing else in the atmosphere that can drive it.  I suggest you give this link a hard read and then let me know what you think:

     http://www.fieldlines.com/story/2007/12/5/16183/1323

You might also consider what the consequences would be of this paradigm.  It gets very interesting.

[MattM]

Funny you should bring up my experiments.  You are however wrong about what I am doing.  Just as you were wrong that aluminum is a bad material for blades.  If you are going to poke fun at a guy at least present the correct facts.

First off my 5' rig that spins 100rpm on wind in the 8mph range is a ten bladed 26 gauge sheet metal and that 5' diameter includes a 12" rim cut out of 16 gauge steel.  And it is a lot stronger than you gave it credit for, easily riding out wind gusts of 30mph.  The whole idea of using 26 gauge was to test the limits of the material.  For a 5' diameter the 26 gauge was good.  For its 15' diameter brother it was not such a good idea as it collapsed on one wing around 5' from center.  The aluminum blades I'm testing in free spinning rigs right now are a three bladed 10' rig and a 10' two-bladed one piece set.  I am looking at the effects of the elements on the materials.  I guess that makes me wrong in your mind.

Now, you insultingly chimed in on gotwind's forum the other day and said how aluminum - my choice of material - is such a poor choice of materials for a blade.  Your example was an extruded aluminum blade that broke.  And you got defensive when I pointed out that extruded material is weak by nature.  But ultimately the guy that supplied the pictures stated the same hub was responsible for destroying a set of fiberglass blades (and tail!) prior to the aluminum failure, and the problem was an imbalanced hub.

Now you went so far as to mock my claim that aircraft use largely aluminum subframes for good reason.  Somehow because I support aluminum you think that I am mocking the aerospace industry for using what I already consider the superior material?  Seriously, guy, this is completely opposite of rational judgment.  I don't understand the bone you have to pick with me.  I like the use of metal.  It is what I know best.  It is what I can get for next to nothing in cost.  It is my preferred choice.  Get over it already.

[fungus]

Have you ever heard of United Airlines Flight 232, Japan Airlines Flight 123, China Airlines Flight 611, and Aloha Airlines Flight 243 ? So dont be so confident in saying that aluminium is used with success on planes ..

[scorman]

Steven.

I think you are overly impressed with the equations.

"The blade should actually have a "non-uniform" twist to account for the relative change in angle of attack at different stations along the radius.

There is a small range of angles of attack where the airfoil gives the most lift for the least proportional penalty in drag.  This is called L/D-max.  If it's 5 degrees or so, then you want all sections of the blade to be working at 5 degrees angle of attack when turning at the target TSR."

Linear twist is chosen as the compromise in starting at the "ideal" tip pitch and then figuring for a given thickness of blank, the question:"What is the maxiumum twist I can get out of it"??.

In the hypothetical tapered model I presented, the tip was designed at 6 degrees and for a 1 3/4" blank, you can twist up to 14 degrees more or a root pitch of 20 degrees.

For a presumed TSR you think you will be running at, say 6, then the apparent wind angle is defined constant for any wind speed. The AOA is simply derived by subtracting the pitch at any station from the apparent wind angle at that station.

I have included a speadsheet for the 5 foot blade that Paradigmdesign is molding derived from the Sandia design:

http://www.otherpower.com/images/scimages/7526/Sandia.xls

Choose any TSR you want, any tip pitch you want, any WS you want

you will note that nobody makes a root at 72 degres to get the same AOA for the entire blade length!

if you read through the entire Sandia report, you will note that they were trying to use newest/greatest modeling to build a better 28 foot Jacobs blade ...

know what they concluded??? the uniform 6" cord width, non-twist Jacobs used for about 50 years couldn't be improved upon by more than a few % ...Bergy sells them that way too!

Me personally, have other reasons for building a tapered but non twist blade, and I'll hopefully have some data to share why I have implemented it that way.

Stew Corman from sunny Endicott

[MattM]

True, true, the lower pressure gradient is not a true vacuum.  And you could of pointed out how the flow of air is not like a fluid due to compressibility.  All in all it is just nitpicking.

Ironically I just read that same post about the Betz limit in the diary section this morning.

[MattM]

Is that Jacobs Blade from one of the old Jacob's wind mills you're talking about?  Seems like they used a pretty much uniform blade chord and pitch from hub to outer tip.  It also doesn't seem like they used much chord in the blade at all.

My little brother had an idea for a simple blade design that seems to work surprisingly well.  Keeping the blade chord thin relative to its length seemed to hold the key to how fast it could turn.  He built a test set of 3 blades out of 26 gauge for a 15' diameter turbine - admittedly it was too light of material - but they turned much sooner and faster than a comparable sets of tapered blades on 10' diameter turbines resembling the conventional wisdom around here.  His blade had no wider chord than 4" whereas the 10' diameter turbines had an interior chord width of 14 inches out to a tip of 5.5".  We've since used the same design on other turbines and they really seem to spin with some force, they can be massed produced in near-perfect uniform shapes, they are easy to balance, and - because its material we have in quantity - they don't cost us much cost-wise.  I call them the Bart blades in his honor.

Go ahead and blast us for using a shape that is not supposed to be efficient.  But its no worse for looks than what I've seen made out of pvc around here.

[finnsawyer]

Concerning the compressibility of air and fluid flow, I would like to quote H. C. "Skip" Smith from his excellent book "Aerodynamics":

"We have seen that anything moving through the air affects the air pressure in its vicinity.  This pressure change will also affect the density of the local air; however, at relatively low speeds such as 100 to 200 mph the density change is so slight that we can normally neglect it altogether.  In such low-speed flight regimes, the air is incompressible, meaning that the density of the air flowing over the aircraft does not change from that of the surrounding ambient air." - page 188.

Presumably, the statement would be true for most wind turbines since tip speeds fall into this regime.  So, don't waste your time trying to bring in compressibility.  Treat the flow as that of an incompressible fluid.  There are more important effects to consider.

[SparWeb]

Guilty as charged: I like the math too much.

You already knew all about this... I had no way to tell without asking.

I looked at the spreadsheet and it does show the variable twist required along the span.  You have a geometry that works with very little deviation from the "ideal" model.  So it does look like a sound compromise.

[MattM]

You may be right on a superficial sense, but I'd think it would be more difficult to explain Betz's Law without the idea of air compression.  Density cognates a thermal state of the air.  You can compress air without necessarily thinking of it changing the overall density because the change is dealt with as a wave propagation whereas a change of density is a state change.

[finnsawyer]

I haven't seen a treatment of Betz's Law that includes density.  And I don't think anyone can explain Betz's Law.  The effects don't add up.  I thought I made that clear in the diary.  The fatal flaw of Betz's Law as usually presented (I haven't read his paper in the German) is the deflection of the air around the rotor.  The power contained in the part of the flow due to that deflected air is lost forever according to the treatment.  It's gone, and yet somehow the reduced air flow through the rotor still could serve up 59% of the incident power.  The whole thing is weird.  Somehow they still come up with a pressure drop across the rotor in addition to the air flow through it.

The power of the model in the Diary is that the power carried by the diverted flow of air is not lost but shows up in the lower pressure behind the rotor, which is important as the pressure differential between the front and back of the rotor determines the power available to the rotor.  The process is subtle and beautiful.  All the known effects are accounted for in a simple and straight forward fashion.  

Another factor is that the Diary model allows for a change in the internal energy of the air.  There doesn't appear to be any reason to deny the possibility of such a change.  A change in density also could require a change in internal energy, so one could consider it.  Personally, I won't.  There are enough interesting effects to consider.  

[scorman]

Matt,

The key to an airfoil which makes it different than a flat plank, is that there is a smooth transition .. BTW a flat plank angled into the wind will also turn, just not well.

have a look here ( note that L/D ratio is inverse of Cd/Cl shown in chart):

I won't go into detail about how you drew your conclusions, but I can add a small point on your construction technique:

If you can use a sheet metal roll former to contour your blade to a smooth curve rather than a sharp break from flat geometry, you'll see much better performance.

If you want to play and learn, open this simulator:

http://coloradowindpower.com/FoilSim/

hint: make input = shape/angle

thickness % = 1% (sheet metal)to start

camber: vary

output = lift meter : choose Cl or force

If you want to understand that there is some science behind the airfoil shapes, you might read through this from the beginning and quit when it gets too technical:

http://books.google.com/books?id=4UYm893y-34C&pg=PA173&lpg=PA173&dq=%22wind+turbine%22+%
22rotor+thrust%22&source=web&ots=2Maz5M06hz&sig=YRimxd4hN_IoZ1Vq47ZO9GQTibg#PPP1,M1

Stew Corman from sunny Endicott

[MattM]

I think you put too much mystery into Betz's Law.  Look at wave theory and it can explain the basis of his theory.  Again, don't confuse density with compression, they really are two different phenomena.

[MattM]

Those simple designs are from my little brother - btw he's not really little, he's an adult now.  They are simple and work very well for blade design.  They are not airfoils in the traditional sense, but they do treat wind like its an object moving along a simple slope.  Rounded curves just do not seem to work as well, especially in the narrow chord that we've been testing.  If I was to plot the force required to move an object along a concave curve or a straight slope I bet the straight slope wins every time.  After all this is about recouping energy from the wind, not about sliding effortlessly through the wind, so in my opinion people are putting too much effort into making airfoils.

Trust me, I've got overly complicated making sheet metal blades.  The problem is no matter how exact I tried to make the angles, the simple shape my little brother came up with worked better.  I made nice hollow and shapely ones.  I made strong and rigid ones.  I even made one set with twist.  But his design not only worked better than the overly complicated ones, they were much easier to fashion - mass production is important - and took a lot less material.  And this simple shape is quiet, another plus.  We first tried with two or three blades, but his overly simple design also works well in a multiple blade rig that has more than three blades.

I put together a lightweight 5' diameter 10 blade test prop that was better than using just a few blades.  I recently put together a 24 blade prop out of heavier material but don't have proper bearings set up to use with it yet; its weight penalty pretty well destroyed a simple 16 gauge guide I built for the axle this past weekend.  Multiple blades seem to pick up more inertia and startup under light breezes, but they are so much heavier overall the tradeoff may not be worthwhile.  So take my experiences with multiple blades with a grain of salt for now; until we mount pmg's to these rigs we won't get any real idea what kind of torque they are generating.

[ghurd]

Those designs are from your little brother?

He is confusing 'turning' with 'power'.

"so in my opinion people are putting too much effort into making airfoils."

At least they didn't waste time making something that works well.

"mass production is important" meaning GM couldn't make this at a gas cap factory in 1943? (M3A1 Grease gun)

Material?  $3 of AL?  That's out of my budget!

Quiet?  Under a load?  See 'wood'.

Multiple blades?  See Benz.

Inertia?  That stands alone.  Heavy goes in here too.

Startup under light breezes?  Like a heavy HF Garden Windmill?

"We first tried with two or three blades, but his overly simple design also works well in a multiple blade rig that has more than three blades."

Correct.  Crappy 1850 circa blades tend to do a lot better in quantity.

"cheap (from somewhere else)", still has no math of the PMA costs.

I am not arguing with you.

Passing gas for facts causes problems for years.

Some do-hicky AL blade being better than a real blade because its free, safe with hillbilly design parameters, and good because it spins in a breeze.

"Fatigue".  You might to look it up.

"until we mount pmg's to these rigs we won't get any real idea"

Correct.

Where is Ron?  I thought this was his job.

G-

[MattM]

Seriously, ghurd, you've gone from simply insulting me personally in previous posts across a pair of forums to what is now borderline harassment.  Your insults do not belong in a serious discussion.  Back off.

My experiments are built around sound principles.  The Bart Blade concepts are similar to what has proven to be the best form in the sailing world, straight rigid wings.  They don't use curved, fat, and rigid sails to get maximum sail; no, they use flat, thin, and rigid sails.  There is absolutely no reason these blades do not work under load as well as they do free spinning.

And, guy, I have been working with metal since I was 13 years old.  You are making claims about the properties of metal and have completely dismissed them because of "fatigue".  Catchy word, that "fatigue".  But if you work with metal you don't let it intimidate you.  Wood fatigues, too.  Funny thing about pvc, its a brittle material far more prone to fatigue yet you still dismiss metal blades.  Go figure.

[wdyasq]

"And, guy, I have been working with metal since I was 13 years old.  You are making claims about the properties of metal and have completely dismissed them because of "fatigue".  Catchy word, that "fatigue".  But if you work with metal you don't let it intimidate you.  Wood fatigues, too.  Funny thing about pvc, its a brittle material far more prone to fatigue yet you still dismiss metal blades."

I've tried to stay out of this as everyone except me is allowed an opinion. MattM, I'd like to see your 'fatigue results' for wood. The Gougeon Brothers did tests a decade or two back and found it was minimal and about three quantum leaps lower than the next best material - 'carbon fiber'. Steel fatigues rather quickly and Aluminum is a lot more subject to fatigue than steel.

Starting to work with a material young will not change the physical properties of materials. A lot of airline passengers died because early designers of jet transport aircraft did not understand metal fatigue. They did not let intimidate them. They  are dismissed from this discussion.

UIUC "University Iowa Champlain Campus" is one of the most respected low-speed aerodynamics facilities in the world. They have done a lot of work on wind-turbines. If you want to learn a bit about wind-turbine blades you could go there and find out some facts. Or, you could choose to stay stupid.

Fluid dynamics is an involved field with a lot of mathematical variables. It takes a lot of time and education, self or school, to properly design the most efficient blades for a wind-turbine. The fact that many are using airfoils with characteristics not suited to the Reynolds numbers and are not using the proper twist just shows how forgiving  the aerodynamics are, not how good the design is.

Please feel free to have your won opinion, even if it is wrong. I would prefer folks and fools keep  ignorance of fatigue and fluid dynamics stated as conjecture rather than fact.

Thanks,

Ron

[TomW]

Ron;

Champlain is in Illinois not Iowa just to be accurate.

Thanks for taking this new "expert" to task. Facts are far more important than arrogant opinions.

I tried to warn him to look around and see who's house he was in before talking trash but he pretty much flipped me off. I don't give a flying fsck about egos I care about accuracy and just repeating it doesn't make it so.

Your serve.

TomW

[MattM]

I have a difficult time dismissing metal, especially steel, as a suitable material.  I can agree if cost was no object then carbon fibers would be my concentration of energy.  But because it is not cost-effective to me I am not personally prepared to work with that material.

I'd be glad to read up on the Gougeon Brothers' comparisons of materials.  I've googled their name 40 pages deep and mainly come up with references to boat rudders and epoxy, the occasion claim that they say wood is best for construction.  The main references to them concerning fatigue seem to be centered on the topic of boat building or epoxies.  I have an open mind and would love to see this information.  Help a guy find the pertinent information.

[wdyasq]

Mattt,

It is an older report. I read it in the late 80's or early 90's, IIIRC. I got the report from Gougeon. At that time I was using drums of epoxy to bond wood things together. I suspect one would need to use the telephone or maybe even write a letter to get a copy. The research was on several materials including the Douglas Fir/epoxy laminate they were proposing and used to build many large HVAT blades.

The conclusions and tests convinced me wood or ENGINEERED composites were the only safe way to build small turbine blades. The engineering for a set of prototype blades would be prohibitively expensive unless the mill was larger than those built here.

Ron

[wdyasq]

I get all those little "I" states mixed up.

Regrets,

Ron

[MattM]

http://www.hcs.harvard.edu/hejc/papers/wind_technology.pdf

Here's an interesting paper concerning wind turbines.  Unfortunately, even though its from Harvard's archive they do not give an author name or references to fact check.  But even without the proofs it still has interesting information.

[finnsawyer]

Perhaps you'd like to put that Wave Theory idea in a Diary.  I'm not going to pursue that.

I'm not sure what you mean by "compression" either.  Usually compression implies a change in the internal state of the gas, such as one finds in an internal combustion engine or an air compressor.  The pressure dome that exists in front of a wind turbine does not imply compression in that sense.  The real issue with a change in density is whether it is also accompanied by a change in the internal energy of the fluid.  Bernoulli's Equation can potentially be satisfied with a change in density.  But you need to be aware that for Bernoulli's Equation to hold the internal energy of the fluid must remain constant.  In fact you can find a derivation of the equation based on that requirement.

[JW]

Hi All,

-True, true, the lower pressure gradient is not a true vacuum. And you could of pointed out how the flow of air is not like a fluid due to compressibility. All in all it is just nitpicking. MattM"

*

-"Concerning the compressibility of air and fluid flow, I would like to quote H. C. "Skip" Smith from his excellent book "Aerodynamics" GeoM"

~

-"We have seen that anything moving through the air affects the air pressure in its vicinity. This pressure change will also affect the density of the local air; however, at relatively low speeds such as 100 to 200 mph the density change is so slight that we can normally neglect it altogether. In such low-speed flight regimes, the air is incompressible, meaning that the density of the air flowing over the aircraft does not change from that of the surrounding ambient air." - page 188. GeoM"

   -"Presumably, the statement would be true for most wind turbines since tip speeds fall into this regime. So, don't waste your time trying to bring in compressibility. Treat the flow as that of an incompressible fluid. There are more important effects to consider.

GeoM"

Not so fast Fin,

 Your touching on hypersonic con-trails from a scram-jet aircraft wing-tips. In relation to blade tips.

-"In such low-speed flight regimes, "

-"the air is incompressible, meaning that the density of the air flowing over the aircraft does not change from that of the surrounding ambient air."

Yes, but there is a cooling effect, this can be observed thru relative humidity de-pressurizing, at sea-level. Assuming humidity is atleast 70%. Such a thing would not be observable at the north pole or at anytype of high-altitude, yet, con-trails are most easily observed from sub-sonic aircraft, passing at a very high altitude.

JW

[finnsawyer]

Actually, "they" say that some of that CO2 ended up as CaCO3 directly as the acid rain reacted directly with the calcium in the rocks.  There are deposits of CaCO3 that do not contain any fossils.  This event occurred at the end of the ice-ball Earth stage.  More importantly it seems, is the question of what happens when the level of CO2 reaches 4 to 5 times current levels.  It's not the global warming per se, but the CO2 poisoning that might be the real problem.  The oceans will stagnate and die.  There will be a major die off organic material on land.  The conditions will be the same as those which layed down the oil and coal deposits in the first place.  It will last for 250,000 years.  Or so "they" say.

[finnsawyer]

I'd like to some proof of this effect.  If the water is already in a gaseous state and everything is at the same temperature then the water vapor should behave like any fluid.  That is, its internal energy should remain constant and Bernoulli's Equation should apply.  We are not talking about a steam turbine here.  

[finnsawyer]

You want a magic formula?  Here it is.  The component of lift force per unit length in the direction of rotation, Fl, at radius R times the speed of the blade section at radius R, Vr, times dr is equal to the power available from the wind in the annular area cut out by the blades at radius R having width dr.  To recap:

     FlxVrxdr = 1/2xEfxpx(Vi^3)x2xPixRxdr,

where p is the density of the air, Vi is the speed of the incident air and Ef is the maximum allowable efficiency.  We may further reduce the equation to:

     FlxwxR = PixEfxpx(Vi^3)xR, or

     Flxw = PixEfxpx(Vi^3).

The term w is the angular velocity.  You can use this equation to find the total width of the blades at radius R.  Once you have that you can divide by the number of blades and can then calculate the drag force per unit length at radius R.  Note that Fl will also contain a Vi^3 term.