Homebrewed Electricity > Wind
Christmas Windmill Time
taylorp035:
It's been a while since I've posted and done anything with my windmills, but I wanted to get back to making some new blades over Christmas break. I think everyone has their talent or favorite part when making a windmill and mine is definitely the blades.
This time around I wanted to do a bit better and make a set of blades with something that closer resembles a high lift airfoil instead of having a flat wind facing face and a sharp front edge (school textbook airfoil). The rest of blades were have been historically pretty good with an exponential taper and twist, all done by hand. For my new blades, I would like to make a mini set on my CNC mini-mill (5"x6"x12" + rotary axis) as well a hand carved larger set for my treadmill motor. The mini set will probably be 10" long blades and the larger blades will be in the 25-30" range to hit a desired 455 RPM cut in. If the mini set goes well, I may CNC the large blades in several steps.
I found in the past my blades would struggle to get going and then would rocket up quick in speed and then seemingly not do good at high speeds. I was reading Sparweb's blade aerodynamics page and realized I made the same mistake he mentions in that my blades had way t0o high of an angle of attack.
http://www.sparweb.ca/1_Blades/Aero.html
So I did a bit of math to figure out the apparent wind angle on the blade based on the TSR and where you are along the length of the blade. Then I picked an angle of attack I wanted based on looking at a few dozen lift coefficient graphs. I picked an angle a bit below the peak lift such that the blade would provide more lift once it really ran into the load from the motor or the wind speed picked up... maybe this is the wrong way of thinking, but I liken to having more torque down low in the RPM's from a turbo charged diesel engine. If did this wrong, let me know. I have some time since I'm still trying to get my CAD software back up and running due to a Windows 7 update issue.
Hopefully I can break my old record of 13.3 amps going into a 12v lawnmower battery. I now have a ~5ah, 14.4 volt LiFe battery that used to be in my daily driver as a replacement for the 65 lb Pb battery to play around with as well as a pile of other 3S 5ah LiPo packs.
SparWeb:
Hi Taylor,
Happy to see that you are well and still experimenting!
That's a heck of a twist profile - the blades will look swoopy and fun. On the other hand, I'm looking at the "455 RPM" cut-in and it doesn't seem right. Can you confirm the open-circuit voltage per RPM of your generator?
When looking at CL, I always want to check the Reynold's number before judging what's realistic. I'm not sure you can get it to 1.4 at this scale. But XFOIL has provisions for dealing with RN, so just confirm for me that you can expect that. Also, let's look at the drag polar for this airfoil - it's pretty thick which means it will have a higher profile drag.
BTW, I've refined my work on blade twist by using a spreadsheet I got from Hugh Piggot and adding my own "twist". Let me know if you'd like to take a look and I'll post an attachment or find some way to get you a copy.
Adriaan Kragten:
The aerodynamic theory how to design a windmill rotor is given in chapter 5 of my public report KD 35 which can be copied for free from my website www.kdwindturbines.nl at the menu KD-reports. I give an example for a blade designed with the optimum lift coefficient and for a blade designed with a constant chord. Constant chord blades are much simpler to manufacture and the rotor can still have a high maximum Cp if the correct airfoil is used.
A problem with making small blades with small chords is that the Reynolds value becomes very small at low wind speeds. The Reynolds value can be calculated for every blade section using formula 5.5 of KD 35. Normal aerodynamic airfoils get very high drag coefficients if the Reynolds value becomes less than about 10^5. So it is useless to use a very advanced airfoil which may have perfect characteristics at high Reynolds values if it is used at low Reynolds values. For low Reynolds values, one can better use the 7.14 % cambered plate airfoil (see KD 398) as this airfoil has still rather low Cd values at low Reynolds values. This is because the sharp airfoil nose makes the boundary layer turbulent and this turbulence prevents stalling at moderate angles of attack.
taylorp035:
I took the windmill out a couple of days ago and decided my modified 46" diameter blade set doesn't want to get the treadmill motor spinning at all from a stop. Even spinning it by hand to get it moving didn't want to make it keep going in ~8 mph winds. Of course it wasn't very windy by the time I got it outside...
So today I did a generation study of the motor on my lathe while charging a 2ah 18650 lithium cell that was almost empty. I didn't have a RPM measurement under load, so it might push an extra amp or two at the listed RPM. I was able to get 7 amps into the cell at 614 RPM (no load RPM/volt used to calculate this speed)... there were a few losses through my alligator clips and ammeter setup, so I did some rough math to correct it based on resistance.
Doing some calculations with a TSR of 7 and a 6 foot diameter, you get the power curve shown. I hand drew the red line on there as the blades overpower the treadmill motor with a higher TSR. I might get 20 watts at 10 mph. To get the same power for a 8 foot diameter, I would need 13 mph. My wind site is really terrible since this is just a science project and I don't have it above the trees.
I have 14 of these 2ah cells that I harvested from a bad 6ah 20 / 60 volt Dewalt pack. I have a friend who might give me a few more to play with. If this works decently and I make some nice battery holders + have a battery management system, I may try purchasing some higher capacity cells for $2-3 a piece (~$200/kwh). Finding 1S battery management systems that handle more than 10 amps is proving to be elusive. The moment being, the BMS will have to be my eyeballs on a volt meter. I think it would be cool to get this to the point where I could easily charge two cell phones everyday.
SparWeb:
There are a variety of cheap/simple ways to set up a tach, especially if you're testing on a lathe. My method was to use a bicycle speedometer. There's a way to enter a wheel diameter value that it will multiply by the pulse count to give you a number equal to the RPM.
If you have a lathe to drive the generator for tests, then you can probably find a way to measure torque at the same time as power output. Torque X RPM = Power input.
Then you'd have a firm power input curve that you can use to match the blade TSR and diameter.
Navigation
[0] Message Index
[#] Next page
Go to full version