Mini CNC Blades

[taylorp035]

I had the itch to make another set of blades, but this time on my mini CNC machine.  It can do about 11.5" on the long travel of the bed.  I decided to make a two blade design because it really is easy to balance it using the hole for the shaft and a lap joint.  I did this ~13 years ago in high school and they turned out fantastic.  The only issue is they had very little pitch, so they spun very fast.  At 30" in diameter, I verified one day they hit 3,000 rpm.

This time around, I decided to do a quadratic blade taper and a tangent based calculation for the pitch of the blade.  Do other people use the trig function tangent when finding how steep their airfoils are along their length?




For the CAD model in Inventor, I made a sketch that had all of the dimensions as a function of the width.  The thickness was 1/8th the length.   The widest part of the blade is 4" and the whole thing is 0.8" thick.

[MattM]

What is the second picture about?

[taylorp035]

What is the second picture about?

  It's the sketch at the 4" location.  I changed the angle and width every 1" down the blade.

[MattM]

The first picture is clear.

The second picture looks nothing like the first, so I'm struggling to understand how they relate.

[SparWeb]

Do other people use the trig function tangent when finding how steep their airfoils are along their length?

The simple answer is Yes.

Your math works just fine and produces an OK blade but I'll just let you know that there are actually 2 angles at play, here.  There's angle of attack, of course, but that's different from angle of inflow.  Above I was working out inflow.  You need that first, but then you adjust it by the angle of attack that you want.  Your calculations just use "5" degrees at the tip, and if that's just the inflow angle you calculate from your target TSR, then it's like assuming the blades are at zero angle of attack.  They'll actually take up a different angle of attack by running at a different TSR, probably slower.  So they'll run just fine, but not at the speed you were looking for.  All blades only run at the design target speed when loaded by the generator with exactly the same power as they're developing, so TSR is variable anyway.

Matt,
Picture 1 being the table probably isn't what you mean...  The "last" picture is a view looking face-on at the blade.  "Down" if it's lying on a table.  Taylor's trailing edge is swoopy-curvy.

[MattM]

The second picture... err third picture makes sense as a top view.  Thanks for the clarification.

[taylorp035]

Thanks Sparweb.  I considered the wind speed when I designed the blades for my Christmas windmill, but I didn't feel like thinking that hard for these mini blades.  As you said, the math is variable vs. TSR.

I actually didn't try to target any TSR.  I just picked 5 degrees and ran with it.  The slope was a 11:1 ratio, so that's probably a decent starting place to hit a real 6-8 TSR.

I haven't decided what I will do for the generator, but I'll probably take a run at making my first axial flux alternator with the leftover 17 gauge.  With minimal air gap and a few thousand RPM, I should be able to hit a high enough voltage.  I also have a 10A 1S BMS unit and 3 left over li-ion cells, so I can make a super mini version of what I have on my 7 footer and still charge my cell phone.  I could even mount it on top of the big windmill and skip the tower and tail...

[Adriaan Kragten]

The angle phi is the angle in between the direction of the relative wind and the rotor plane. The correct angle phi is found using formula 5.3 of my report KD 35. This formula takes into account that the wind speed in the rotor plane is 2/3 V and that the wake is rotating. If you neglect wake rotation, you get formula 5.9 which is easier to understand but which has the factor 2/3 at another place. Formula 5.9 gives too large blade angles at a low tip speed ratios (see figure 5.2 in which the graphs of both formulas are compared). The blade angle beta is calculated with formula 5.2 out of KD 35. So you need a certain angle alpha to calculate beta. The optimum angle alpha is the angle for which the drag/lift ratio is minimal and this angle depends on the chosen airfoil and Reynolds number. The chord c is calculated using formula 5.4 out of KD 35. The whole design procedure is illustrated in example 1 for a tapered blade and in example 2 for a constant chord blade.

[kitestrings]

The angle in some of the common calculators (Hugh's, Alton's, Warock's) I believe is indicating the angle between the chord line and the plane in which the rotor spins, vs. the face of the blade and the plane in which the rotor spins; angle phi as Adriaan defines it.  IIRC Hugh has a "drop calculation, which is a Drop over the Chord Width expressed as a decimal, or rise over run.  Depending on the profile, and which station you are at, it has can be more of a difference, but I think it is good to know which you have.  One is more helpful for layout and carving.

Spar, you're much more informed, and current here - is the "in flow angle" you described different than either of these two?  I'm not familiar, or have forgotten, that term.

[Adriaan Kragten]

The angle in some of the common calculators (Hugh's, Alton's, Warock's) I believe is indicating the angle between the chord line and the plane in which the rotor spins, vs. the face of the blade and the plane in which the rotor spins; angle phi as Adriaan defines it.  IIRC Hugh has a "drop calculation, which is a Drop over the Chord Width expressed as a decimal, or rise over run.  Depending on the profile, and which station you are at, it has can be more of a difference, but I think it is good to know which you have.  One is more helpful for layout and carving.

Spar, you're much more informed, and current here - is the "in flow angle" you described different than either of these two?  I'm not familiar, or have forgotten, that term.

No this isn't true. The angles phi, beta and alpha are always defined in the same way in any literature which I have seen. A picture of these three angles is given in figure 3.2 of KD 35 (page 12). It can be seen that phi = alpha + beta. Beta is the blade angle, so the angle in between the zero line of the airfoil and the the rotor plane. Alpha is the angle of attack, so the angle in between the zero line of the airfoil and the direction of the relative wind W. Phi is the angle in between the direction of the relative wind W and the rotor plane.

The zero line or 0-line of the airfoil is not always the same as the line which connects the airfoil nose and the airfoil tail. For some airfoils, like most NACA airfoils, both lines are the same but for Göttingen airfoils with a flat lower side, the 0-line coincides with the flat lower side. The 0-line is the reference line for the Cl-alpha, the Cd-alpha and the Cm-alpha curves.

[kitestrings]

I've always admired how Hugh Piggott could present sometimes very technical information in very tangible terms.  I'm not trying to contradict norms for defining accepted terms.  What I was saying is that some of the spreadsheets that are available appear to me to have fields that are intended to help with the layout, and material to be removed when carving blades.  Alpha is the angle between the the chord line and the direction of relative air movement, but the chord line is not always the same as the flat of the blade.  And, if you are carving blades it is good to know which you have.

There is also a point as you move toward hub where the chord width becomes very wide at the root, and the angle very "coarse", as Hugh describes it.  At a certain point it becomes impractical, and you must choose some compromise between the theoretical results and what is practical for the material and properties of the blades.  So, with Hugh's calculator for example, there's a point where you are entering the actual wood width, and the drop at those station is calculated using this width.

I had to remind myself what the flow angle, phi is... but I can see now it is the angle between the relative wind and the rotor plane, alpha + beta.  It's interesting that in in looking this up, I found where beta is sometimes referred to as "setting angle", or simply "pitch", so nomenclature does vary some.

[SparWeb]

Warlock... I haven't visited that site in a lonnnnnng time.  And it's updated, too.  Funny how so much of the internet comes and goes, but there are still some things that are both useful and stick around.  Thanks for reminding me!

[MattM]

I didn't think there was much to debate in a foot long blade.

Apparently, I was wrong.

[SparWeb]

Welcome to the Internet, Matt! 

[joestue]

I didn't think there was much to debate in a foot long blade.

Apparently, I was wrong.

well doesn't that depend on whether or not its a foot long?

and why'd you have to measure it? 

[taylorp035]

I cut the first blade out of maple yesterday and plan on finishing the second blade today.  It took about 3 hours to cut the bottom and top with a 1/2" ball router bit.  I'm going to make a lap joint where the 1/2" hole is so I can glue the blades together without the need for a hub.

After cutting the first blade, I remembered on my original CNC blades that I only did the top part of the airfoil and then chose to use a band saw to cut the bottom flat side.  The very thin trailing edge ended up being a bit frayed, so there was a decent amount of sanding to make it look nice.  I think if someone wanted to make a larger blade, I would recommend drawing it such that the tail of the airfoil stopped at ~0.050" or 0.075" and then just sand it the rest of the way.

[kitestrings]

Nice!

Maple's a bit heavy (and hard), was that just you had on hand, or was there a reason you chose it?

[MattM]

Yes, very nice.  How will you mount it?

[JW]

https://www.fieldlines.com/index.php?topic=139256.0

[taylorp035]

Here's a time lapse video of the CNC machine cutting the second blade.
https://youtu.be/VL_uJNHBWFE

Maple's a bit heavy (and hard), was that just you had on hand, or was there a reason you chose it?

I had about a dozen different choices.  I decided to go with Maple because it was a bit harder and I had a board already roughly the right size.  The CNC machine doesn't care how hard it is.  My property has mostly maple, ash and poplar on it.   I once hand carved a set of blades out of maple and it was very difficult.  My latest 7' set was also made out of a harder wood (ash) and that was a good bit of work to carve.

Yes, very nice.  How will you mount it?

I'm not certain yet, but I was seriously considering just adding it to the box of my 7' footer (Christmas Windmill thread), since then I can house the electronics inside of it and have no tail.  I believe it should be fairly safe to use my 10 amp 1S BMS unit, since it would take a good 40-50 mph wind and the 7' would furl way before then.

[richhagen]

I made a bunch of cnc blades a while back.  Has it really been about 16 years now?  Time really flies.  https://www.fieldlines.com/index.php?topic=137906.0

[taylorp035]

I finished joining and gluing the blades together.  There are about as perfect as they could be without CNC cutting both blades at once from a solid piece of material.

In the mean time, I did the FEMM simulation to guesstimate my voltage based on different steel backing thicknesses and magnet options.  Because I'm an engineer, the biggest number attracted my attention of using a 1/4" steel plate (4" diameter) and some N52 1" x 1/4" circular magnets.  I see magnets have roughly doubled in price since I last shopped for them.... so this fun little project is going to cost a bit more than I was hoping for.  Even with the free copper, it's probably going to be $100.

My plan is to do at least 25 wraps of my 17 gauge wire with 6 coils in a single phase setup.  I could change my mind later, but I have a spare 25A single bridge rectifier to play with and I'm not too concerned about it being noisy since it's so tiny and will be spinning 1000+ RPM once she's loaded up.  The coils will be 3/8" thick and I simulated a 1/8" air gap on either side.   If my math is right (which the Christmas windmill alternator was pretty spot on when i tested it), this should cut in at 482 RPM with all 6 coils in series.  Since I think 1000 RPM will be a better cut in at about 9 mph @ TSR of 7, that means I can wire it in a 2x3 pattern.... I have a feeling I will be playing with the air gap to tone it down a bit.

I still have a 10A 1S BMS unit to control the charging if I want to make a mini version of my big system that will be perfect for charging a phone.  I also have 3 18650 cells left over.

I designed my first coil winder.  Should be 100x easier to wrap than my last alternator.  The base circle of the coil is 1" in diameter and those slots are 1/4" for scale.

[JW]

Yes, very nice.  How will you mount it?

[taylorp035]

Yes, very nice.  How will you mount it?

  I plan on having a ~1-2 foot tall wood post coming up off the back of my current windmill.  Probably right on top of the wood block where my tail is attached so I can still remove my roof to the box(my old snow sled).  It will be about 3 feet behind the 7' rotor.  It will rotate with the big windmill in terms of turning into and out of the wind.  This way I don't have to have another tower.   I can always change my mind later.

I finished the blades with a few layers of spray lacquer.  I also ordered (12) 1"x1/4" round N52's, along with two 4" x 1/4" steel discs to mount them on.  I plan on having a friend 3D print some surrounds for the magnets so I don't have to epoxy them to the steel.  I also ordered some nice ball bearings so it spins up nicely.

My backup plan is to use this tiny axial flux alternator for my 7' footer for charging at 4 volts if I can't get my other setup to work well.  It would max out at ~150-200 watts anyways, so I think it should be able to do it.  For the mini blades, I think 40 watts is going to be the end of the world (~40-50 mph).  It's going to be a single phase alternator for the moment being.

[MattM]

Its good looking.  I hope you get more like 100W out of it at 50mph.

[Adriaan Kragten]

I don't understand why you give the blade such a large chord near the blade root. You get such a chord variation if the blade is designed with one lift coefficient for all stations but this isn't necessary. If you accept a certain variation of the lift coefficient around the optimum value, you can linearise the chord such that you have a tapered blade. So for a tapered blade you get a somewhat larger chord half way the blade but a smaller chord and a smaller blade angle at the blade root. This makes that the blade can be made from a much smaller wooden beam. The maximum Cp for a rotor with a linearised chord is almost as high as for a rotor which is designed with the optimum lift coefficient for all stations. An example of a 3-bladed rotor with blades with a linearised chord is given in chapter 5.4.1 of my public report KD 35.

[SparWeb]

I think he is taking some artistic license, AK.

[Adriaan Kragten]

An example of a 2-bladed rotor with tapered blades is given in my public report KD 532. This 3.1 m diameter rotor with a design tip speed ratio of 7, is made from Roofmate covered with glass fibre and epoxy. As Roofmate isn't strong, you need a rather large moment of resistance at the blade root and this is obtained if the blade chord and the blade thickness are increasing at decreasing radius. This rotor has a maximum Cp of about 0.42.

[taylorp035]

I received my new bearings for it today.  Here's a short video of it.  I would guess the wind speed coming out of the shop vac was fairly slow (6-8 mph).  It was spinning so fast you couldn't really tell what it was doing with your eyes, but you could see it the video.

https://youtu.be/_M88GnvuFEY

[topspeed]

Reminds me of the Paul Lipps props in aviation.

[taylorp035]

Here it is in some faster wind.  It started out at maybe 5 mph, but got up to a decent breeze (~20 mph).  I think it was spinning faster than 4000 rpm.  It sounded like the string on a weed wacker whizzing through the air.  It had quite a bit of torque on the shaft once it got moving, so I have no doubts it will make 5-10 watts in a medium wind and more at higher speeds.

I spun the blades in a 2000 rpm cordless drill and this was at least 2x as fast.

[taylorp035]

I wound my coils today.  Ended up finishing at 40 turns on each coil and 3/8" thick.  There is a Sharpie for scale.

17 gauge wire.  Should be about 0.071 ohms per coil.

[taylorp035]

I completed the magnet rotors and coils.  For the rotors, I CNC cut some 0.030" pockets for the magents into the 1/4" plate steel and then further super glued them in.  Hopefully they stay in place without having to resort to a 3D printed cage to contain them.

For potting the coils, I used JB-Weld 2 part epoxy.  It seemed like the best thing at the store for the job and it flowed pretty good.  The downside is the bubbles didn't float to the top very well.  Some urethane would of been nicer.  I made a plate to go on top to squeeze it all together while it dried.  I used my usual trick of coating everything in packaging tape so the glue wouldn't stick to things.

After wiring it up wrong the first time (I forgot the adjacent coils had opposite polarities), we were able to spin it up by hand to 1.4 volts.  I wired it with two coils in series and then the three sets in parallel going into the 25A single phase rectifier.

You won't be able to see it, but we used a 1/2" copper pipe as a spacer between the magnet plates to space them.  It's cheap and it fit over the 1/2" shaft.  If I want a closer spacing, I can cut a slightly shorter section of pipe.  It's looking like Sunday will be a good test day.