Building the sickest ® VAWT ever. Brilliant minds unite please!!

[brandnewb]

sure some values have changed. of course they do. value taking, or in other words, sampling, is a very demanding highly accurate science.
That is not something I am renowend for (yet)

 

[Hedges]

ok, I found some time to read back. Please elaborate.. what is a 'pole piece'?

All I know is that I have magnetic field directions ( poles ) and I have coils. rotating poles over the coils at a given rpm gives a certain voltage. the lentz law if I am not mistaken.

By "Pole piece" I mean a piece of metal, ferrite, or other material having relative permeability much higher than air, used as a "conductor" in a magnetic "circuit" to guide magnetic field where you want it. In this case it is through the middle of a coil, around the back, through another coil. For a different system it could be to take the field of a permanent magnet and guide it to a different location or shaped surface.

If you have an air-core coil of wire and wave a magnet over it, only some of the field lines pass through the wire loops as desired. Other field lines curve out part way through the coil, only going through some wire loops.

https://www.chegg.com/homework-help/questions-and-answers/magnetic-circuit-fig1-two-windings-two-air-gaps-g1-g2--iron-core-assumed-infinite-permeabi-q15407907

I haven't watched the video, just glanced at the "circuit" analogy:

 

[Warpspeed]

Just to try to clarify things a bit, below is a horrible (near death looking) sketch showing a twelve pole machine with three phases, that has four widings in each phase.
Yours will be probably be different, but this is just an example to show how all the voltages combine.

Suppose we test one coil and measure exactly six volts at our test rpm.
This particular machine will have a final total of twelve coils (twelve poles)
But because we want to make three phases, each phase will have only four of the coils coils connected in series.
So the voltages in coils W, X, Y, and Z add directly. The combined output of the four 6v coils will be 24 volts between points A and B.

We then join together one end of each of our three phases to a centre "star" point.
The voltage between any two output phases will then be 24 volts multiplied by 1.73 which comes out to 48.4 volts.
That 48.4 volts is what we would see on a multimeter, and is the rms value.

However, when we connect our three ac phases up to a six diode bridge rectifier, the dc output voltage will be very close to the peak vaue of the ac waveform or 48.4 volts multiplied by 1.414.
Dc output should be about 58.7 volts minus a couple of volts drop across the diodes.
The dc output under full load will drop slightly below that.

 

[brandnewb]

Just to try to clarify things a bit, below is a horrible (near death looking) sketch showing a twelve pole machine with three phases, that has four widings in each phase.
Yours will be probably be different, but this is just an example to show how all the voltages combine.

Suppose we test one coil and measure exactly six volts at our test rpm.
This particular machine will have a final total of twelve coils (twelve poles)
But because we want to make three phases, each phase will have only four of the coils coils connected in series.
So the voltages in coils W, X, Y, and Z add directly. The combined output of the four 6v coils will be 24 volts between points A and B.

We then join together one end of each of our three phases to a centre "star" point.
The voltage between any two output phases will then be 24 volts multiplied by 1.73 which comes out to 48.4 volts.
That 48.4 volts is what we would see on a multimeter, and is the rms value.

However, when we connect our three ac phases up to a six diode bridge rectifier, the dc output voltage will be very close to the peak vaue of the ac waveform or 48.4 volts multiplied by 1.414.
Dc output should be about 58.7 volts minus a couple of volts drop across the diodes.
The dc output under full load will drop slightly below that.

hahah near death. let's continue this once I can focus again on the alternator.

Because now it's turbine blade time.

I've got excellent news.

Just tried pouring a 10cm high section of the 40cm rea 2822 airfoil with pourable foam.

Damned that thing is rock hard. I am not even going to put glassfiber sheets on the baldes during testing so hard it is.

Next up will be a 1 meter high section test.



I am not even sure we are going to need this aluminium tube in the center. maybe only where the sections are stacked. But not all the way through as I had thought.

Dude this thing is hard!!!

 

[brandnewb]

this is not to say that 3d print technology is no longer needed. on the contrary. I use the 3d printer to print a mold of sorts for the foam to be poured into.

 

[brandnewb]

here is another reason to never discard 3d print tech.

it looks flimsy... because it is.. but fear not. we can pour syco hard resin in it and it will become strong!!!!.

this is the central column connector. to which we will hook up the beams that connect to the blades.

fear not not gang. I have got this covered in my mind. I am on to something yet again

 

[brandnewb]

but why only half of a structure my dear most brand of newbs I hear you ask!

Well, because then we can attach, detach it quite easily from the central column. That is a feature we are gonna need during prototyping.

 

[Hedges]

it looks flimsy... because it is.. but fear not. we can pour syco hard resin in it and it will become strong!!!!.

It will not be strong at all if the ribs are continuous (unless they themselves are strong and bond strongly to your syco hard resin.)
I would think you'd want most area of those ribs cut out, so it looks more like a truss bridge. Only full surfaces would be outside of the mold.

 

[Hedges]

... Then there is the art and science of mold flow; lines form when flow splits and rejoins. Matters aesthetically at least for finish surfaces; might affect performance in some way.

 

[brandnewb]

It will not be strong at all if the ribs are continuous (unless they themselves are strong and bond strongly to your syco hard resin.)
I would think you'd want most area of those ribs cut out, so it looks more like a truss bridge. Only full surfaces would be outside of the mold.

I will admit that I have not stress tested the column connector yet ;( I have not even started pouring the resin yet

But it's like I said earlier.

If it spins enough for usable data to be had before it breaks (or gets torn down), I am happy.

I am already seriously contemplating how to print structures to create molds with that can be used for aluminium casting

we'll get there gang!! I am on the case.

I still have not drawn that schematic of the coils so we have some more talking points. please give it some time as I am juggling quite a few things mid air now.

 

[brandnewb]

the mold of sorts in the making



i just hope the 3.2mm hard board won't buckle out when the foam starts pushing while it is rising. I do have a few holes for it to release pressure out of.

 

[brandnewb]

and yes, I do remember someone asking me cnc routers. My response back then was I had to start somewhere and additive felt more compelling than subtractive.

Well, truth be told that cost wise, if one already can predict having to make a lot of iterations of a blade. Then subtractive with off the shelf hard insulation makes more sense.

I am already looking what open source cnc routers exist.

 

[brandnewb]

you know what is even beyond uncool?

I magically got reinstated at fieldlines (thank you at who ever here made that be)

I presented my findings, I offered for a, on my cost, peer review.

But the gist is one needs to get paid for a peer review?

lol. I stand by my words. Sure I make mistakes but I don't do bullcr@p

 

[brandnewb]

Finally found some time to make a schematic drawing of one of the stators a.k.a. coil disk.


the dark red slab at the bottom is a 10mm high disk of iron powder to attract the magnetic field to pass through the coils (that are not shown yet)
the pinkish/purplish fans on top of that are coil separators printed using iron powder filled filament



here we can see the 3 phases of coils.

Regarding coging. if it was a electro motor it would be a huge problem as there are as many coils per phase as there are poles.

However, this is a generator and the wind charge controller does not put a load on the coils before the voltage has reached 48 volts and the controller will short the coils if the voltage becomes greater than 56 volts.


Should I worry about coging and if so can anyone suggest a way on how to reduce that?

Also please keep in mind that there will be 2 coil disks. 1 on each side of the magnet disk.

 

[brandnewb]

regarding pourable foam to make turbine blades out of.
It's too unwieldy and too expensive.

I will try making a hot wire foam cutter now and cut 1 meter sections out of off the shelf hard insulation board.
Far faster and affordable if I am successful.

 

[Hedges]

I think the windings around iron cores looks correct.
They will occupy more cross section where two coils cross each other. I typically see motor windings fanning out at the ends, so leave room for that.



I don't follow what the red/green distinction is for magnets.
I see they alternate between red/green and green/red where I expect opposite orientation. But I expect one to have North against the stator shown, next to have South in that orientation.

Fiberglass or other over a foam core could be one approach (like a surf board).
Got a CNC hot wire machine? (mill could work) Mechanical "blade" guide that changes position as workpiece fed through?

3-D printing a blade and dispensing foam into it does seem like an automated way to make many alike.

 

[justgary]

Safety wire (e.g. Monel) works very well for a hot wire cutter. Use your Variac (I think that I saw you bought one of those) to power the wire. Keep the wire slightly tensioned and increase the voltage (you might need only 10-15vac) until the wire just smokes off the oil, then fine tune from there to get good cutting without so much heat that you melt a blob out if you stop cutting.

Cut the wing shape out of thin wood and glue one to each end of the foam. Now just follow the wood forms with the wire to cut out your wing.

Be very careful since the ends of the wire are at opposite voltages and the wire is very hot.

But why did you give up on just printing wing shapes and stacking them? A small overlap as a mortise with the next section as a tenon works very well.

 

[brandnewb]

I think the windings around iron cores looks correct.
They will occupy more cross section where two coils cross each other. I typically see motor windings fanning out at the ends, so leave room for that.

yes sir, will make the serpentine coils long enough to be able to overlap.

I don't follow what the red/green distinction is for magnets.
I see they alternate between red/green and green/red where I expect opposite orientation. But I expect one to have North against the stator shown, next to have South in that orientation.

that is how the magnets are oriented in a repelling orientation. a traditional orientation will have a north or south magnet pole facing the coils.
After extensive testing I am going with an iron powder filled magnet disk with a repelling orientation as shown in the schematic.

Fiberglass or other over a foam core could be one approach (like a surf board).

indeed, that is the route I am going to try. But while testing the blades I am not going to bother putting glass fiber sheets yet. Only when happy with the correct size and shape of the air foil I will start battle hardening them with glass fiber sheets and resin.

Got a CNC hot wire machine? (mill could work) Mechanical "blade" guide that changes position as workpiece fed through?

hahah I which nah I found a vid on ytube, several actually, which I am going to try and replaicate

3-D printing a blade and dispensing foam into it does seem like an automated way to make many alike.

yes indeed. But it will take a few tweaks to my setup I am not willing to make just yet in fear of grinding to a halt again.

Safety wire (e.g. Monel) works very well for a hot wire cutter. Use your Variac (I think that I saw you bought one of those) to power the wire. Keep the wire slightly tensioned and increase the voltage (you might need only 10-15vac) until the wire just smokes off the oil, then fine tune from there to get good cutting without so much heat that you melt a blob out if you stop cutting.

niochrome wire is what I ordered in the meantime. Should that fail I will try Monel. thanks for the suggestion.
You are correct I have a variac in the meantime bought for the DIY transformer project.
But i've also got a switching power supply cc/cv that has a 10 amp capacity.

So we can choose between AC or DC. What would be best?
I think with the cc/cv I can configure both the voltage and the amps so that might give for better fine tuning than only playing with AC voltage.
But I could be mistaken here and suggestions welcome.

Cut the wing shape out of thin wood and glue one to each end of the foam. Now just follow the wood forms with the wire to cut out your wing.

yes sir. I'll try that!!

But why did you give up on just printing wing shapes and stacking them? A small overlap as a mortise with the next section as a tenon works very well.

well, since I am now making blades with a 40cm chord length and a 5 meter height the amount of print time and filament needed is astronomic. Even with a 1mm nozzle like I have now. For most other intricate parts I will still use 3d printing and for mold making and what not but for the blades it stopped making sense at this size. Also because it's difficult to exactly know when the filament will run out with these large prints.
Oh yeah I have also seen layer shifts for the first time ever. And I am worried that they will become more pronounced the taller the prints I am making.

 

[brandnewb]

btw regarding to 3d print or not for blades. Once I've settled on a final air foil shape and size I will revisit 3d printing again.
I'd like to know where the tipping point for time and cost lies in terms of blade size.

 

[justgary]

niochrome wire is what I ordered in the meantime. Should that fail I will try Monel. thanks for the suggestion.
You are correct I have a variac in the meantime bought for the DIY transformer project.
But i've also got a switching power supply cc/cv that has a 10 amp capacity.

So we can choose between AC or DC. What would be best?
I think with the cc/cv I can configure both the voltage and the amps so that might give for better fine tuning than only playing with AC voltage.
But I could be mistaken here and suggestions welcome.

Nichrome is just fine. The melted foam is caustic and will stick to the wire in places and make hot spots, which will allow the wire to stretch there, which will thin it, which will change its resistance, which will make a hot spot, which will.... Tension is important. Don't make it more than it needs to be to keep the wire straight when it is hot. Enough cutting and you will eventually eat through the wire and have to change it.

Use whichever supply gives you the most heat control (probably the CC/CV) and isn't taxed too much by the power used. When you get the heat right, the wire will glide through the foam but not make a huge gap if you stop for a second. When you get everything right it will be quite satisfying to be able to cut foam to the exact form you desire. It will also go quickly once you are all set up.