Radial Flux Generator Project

[ChrisOlson]

There's probably not too many people interested in this since it's not in the homebrew plans, and I obviously don't understand how PM generators work.  But I figured I'll post it anyway just in case somebody might like to look at it.

I wanted to build a iron stator core tonight but decided I'd better build the rotor completely first so I know what the dimensions have to be for the stator core.  This generator will run a .005" air gap so paying attention to the dimensions is critical.  This is the rotor for the generator:





It is built using a 2" wide section of 6" Schedule 40 pipe, a Weasler X-1 keyed hub and the center disc torch cut from 1/4" sheet steel.  The center disc is welded to the hub and to the inside of the pipe section.  The whole assembly was then chucked in the lathe and the ends of the rotor drum faced, and the drum surface machined to 6.500" OD and +/-.001" runout.  With the 12 magnets installed on the rotor, the rotor OD will be 7.5".  The center disc is welded off-center in the drum because the drum is going to be drilled, and the magnets both JB welded and bolted to the drum.  It will use 2 x 1 x .5 N42 bars with holes pre-drilled in the center of the magnet.

This will be a 2 kW capacity unit with a delta winding, 24 volt, driven by a 13 foot PowerMax rotor.  The stator will have 18 coils wound on an iron core using AWG 14 wire.  This generator will have a 360 degree plastic shroud around it with ducted forced-air cooling using a eight blade fan in the rear that draws air in the front, passing thru the generator, and exhausts it out the rear.  It will be mounted on the PTO shaft on the new rear-gen turbine I recently built.

That's the basics of the design - I'll post some more pictures here as I get time to work on it and build the stator and cooling system.
--
Chris

[TomW]

Chris;

As always, beautiful metal work!

Should be an interesting project.

I do not know if you know how alternators and turbines  work but it is obvious you can turn raw metal into useful objects!

Is it going to be a radial dual rotor air core which seems tricky or use an iron core like Hughs' brake drum unit?

Keep it coming.

Thanks for the share.

Tom

[Harold in CR]

 I'm watching. 

[dsmith1427]

Thanks for sharing and I am looking forward to seeing more.... And I am always interested in something different. 

[DragonFly III]

Wow I wish i could do that Kind of metal work...super impressive.  I gues It would be nice If i even had the tools.  Your alt is somewhat of a smaller version of mine but with oposite mounting and with much more output.  Mine is 16" rotor made from a caravan rim and the stator is a wood form with homemade laminates (if you plan on making homeade laminated make sure you have a lot of time on your hands).  My alt has 28 mags of unknow power rating but are strong enough to do serious damage to fingers (i found out the hard way)  21 coils of 20ga 140 turns each, 3 phase star.
How do you plan to mount your coils?

[ChrisOlson]

The purpose of my newest project is to built a compact, efficient, high output generator with an axial flow cooling system that works by design instead of building a contraption that makes electricity and hope it cools properly as an afterthought.  And to come up with a design that's a balance between outright crude and state-of-the-art designs used in commercially built generators,  that can be built by a person with reasonable skills and tooling.

So to meet my goals for this project, that means no waterjet or CNC cut or machined parts, no elaborate machining other than basic lathe work that can be done on a home workshop lathe with less than 8" swing, use materials for construction that are locally and commonly available, and still end up with a unit that matches the performance of commercially available generators in size, output, efficiency and duty cycle.
--
Chris

[SparWeb]

You have my attention, too!

How will you build/laminate the stator?  Laser-cut sheets?  Ductile iron or Carpenter HiPerCO?

Ah so many questions - I'm sure they'll be answered in time.

[DanB]

I made one of those once...



half of the stator even had iron/epoxy cores 

seriously... can't wait to see what you come up with Chris - again, I and lots of others are grateful for all that you contribute here (well.. maybe not all.. but most of it!)

[ChrisOlson]

How will you build/laminate the stator?  Laser-cut sheets?  Ductile iron or Carpenter HiPerCO?

For my first experiment I'm going to build a powdered iron core.  Supposedly, the powdered iron has less magnetic efficiency than laminated electrical steel, but lower eddy losses.  I could build a laminated steel core, and I may for a continuing experiment, but I think if I throw enough magnet power at the powdered iron core it will work at the performance level I expect.
--
Chris

[ChrisOlson]

I made one of those once...
half of the stator even had iron/epoxy cores 

Holy crap, Dan!  Now I know why Louisiana Pacific ran overtime for 5 years to keep up with the demand for plywood   
--
Chris

[joestue]

This iron core you speak of, are you just taking an old motor core and removing the teeth from it on the lathe or are you getting new disks punched out?

Some time in 2012 i'm going to build a traditional iron cored axial, but using 15 teeth, should be zero cogging and very low harmonics, (less than 5%)
its mentioned in here https://oa.doria.fi/bitstream/handle/10024/31185/TMP.objres.74.pdf
or maybe this one http://alexandria.tue.nl/extra2/200111643.pdf that no one has ever built an odd toothed motor due to vibration issues, but wind is slow and by nature they have almost no cogging effect, so no need to waste magnet volume by skewing them.

[ChrisOlson]

This iron core you speak of, are you just taking an old motor core and removing the teeth from it on the lathe or are you getting new disks punched out?

I'm going to cast a powdered iron core.  I'm in the process of building the mold for it.
--
Chris

[ChrisOlson]

I spent two hours tonight making a fancy steel mold to cast a stator core in.  Then while walking to the wash rack in my shop to wash my hands I happened to spot a 5 gallon bucket sitting upside down on the floor.  I looked it, frowned, grabbed the tape measure - sure enough.  Perfect size.  Never did wash my hands.  Five minutes later, after cutting the bottom of the bucket off and putting a piece of PVC pipe in the middle I had the mold to cast my stator core:



You'll notice it's only half full.  I mixed up my iron powder paste then discovered I didn't have as much resin on hand as I thought I had.  So I ended up with only half a core, but I got to try out my new mold and get a little practice figuring out how to cast this iron powder paste.  Tomorrow when I can get some more resin I'll finishing casting it.  Maybe since I got half a ring I'll make another half a ring and laminate them together.
--
Chris

[DanG]

shame ya cant have at least a token rotating flux field imitating the final rotor wash aligning the particles while the amalgam is still gelling to maybe reduce losses...

[jeraklidis]

I'm making one as well (Its my senior project). It'd be great to compete/get advice for fun... You might want to choose 14 coils it gives better performance (winding factor) and has low cogging torque and harmonics... I used laser cut electrical steel laminations and used this program to model the flux http://www.infolytica.com/  its also uses a variable pitch method so I don't have to angle out of the wind in high winds the blades just pitch themselves... You are right about it not being homebrew at all. If I knew it'd be this much trouble/cost in the beginning I WOULD NOT have done it...

As far as the rotor goes I'd mill evenly placed slots and expoy the magnets to the rotor and not weld them... They will hold I've seen it in servo motors

I'd also use a solid Iron core stator using soft iron???...

You are a beast for working with those tolerances... be careful about vibration... your rotor can end up sticking to the stator and at 300 rpm say good night...

Someone said they wanted to do an axial flux laminated model... I looked into it... not to shatter your dreams but... first the laminations are hard to construct and they are NOT similar in form to a radial design and its very expensive $$$$$

Can anyone give me a break out of cost on their axial flux design... I want to use it in my paper to show that while my design is nice the cost per watt is best over 10 years with a homebrew design... please add the rated output as well thanks...

[scoraigwind]

Can anyone give me a break out of cost on their axial flux design... I want to use it in my paper to show that while my design is nice the cost per watt is best over 10 years with a homebrew design... please add the rated output as well thanks...

If you have a look at the table on this page, it gives very rough costs for materials and very rough ideas of energy production on different sites.
http://www.scoraigwind.com/axialplans/index.htm

The plans do not require you to use a lathe, nor rely on glue to hold the magnets onto a central rotor against centrifugal force.  Nor to work to 5 thou tolerances.  So overall a bit too crude to be worth looking at.  :-)

[jeraklidis]

I was bored and I jumped on CAD and made a visual of the rotor (12 magnets 30 degrees apart 1" thick 2" length) i have no clue how the magnetic flux would look but its just an idea... I hope you are using Iron and not stainless(I'm assuming)...

[jeraklidis]

Can anyone give me a break out of cost on their axial flux design... I want to use it in my paper to show that while my design is nice the cost per watt is best over 10 years with a homebrew design... please add the rated output as well thanks...

If you have a look at the table on this page, it gives very rough costs for materials and very rough ideas of energy production on different sites.
http://www.scoraigwind.com/axialplans/index.htm

The plans do not require you to use a lathe, nor rely on glue to hold the magnets onto a central rotor against centrifugal force.  Nor to work to 5 thou tolerances.  So overall a bit too crude to be worth looking at.  :-)

Thanks!!! How is the power rating derived? Peak or average? and do you use 30 days a month 24 hours a day?

[ChrisOlson]

shame ya cant have at least a token rotating flux field imitating the final rotor wash aligning the particles while the amalgam is still gelling to maybe reduce losses...

Hi DanG - I did that.  I used a horseshoe magnet with the same spacing as the rotor mags and went around the inside and outside of the mold with it.  I don't know that it did what I wanted it to but it was rather interesting watching the thick slurry move and arrange itself in there when I did that   

I used fiberglass resin for the binder in this stuff and it takes very little resin actually.  I only had 9 ounces on hand and it appears that's enough to make 5 lbs of iron powder slurry.  I used half the normal amount of catalyst in the resin so it wouldn't set up so fast because I wasn't sure how hard this stuff was going to be to work with.  As it turns out, it's a good thing I used less hardener.  This stuff does not pour at all - it has to scooped out of the mixing container in gobs with a putty knife and poked into mold and troweled out like pouring stiff concrete.

That being said, I took the ring out of the mold this morning and it turned out very nice.  I'm going to put it back in the mold and finish casting it to its full 2" thickness when I get some more fiberglass resin today.  It also appears to be very strong and machinable - about like working on a cast iron engine block for machining processes.  You can file it, grind it, chuck it in the lathe and machine it, mill it, drill it, whatever, and it's not brittle to work with.  I think if you gave it a sharp blow or dropped it you could probably crack or break it but I'm pretty happy with how this test core turned out.
--
Chris

[ChrisOlson]

The plans do not require you to use a lathe, nor rely on glue to hold the magnets onto a central rotor against centrifugal force.  Nor to work to 5 thou tolerances.  So overall a bit too crude to be worth looking at.  :-)

Hugh, now cut it out   

I'm actually going to bolt the mags to the rotor using the 2 x 1 x .5 bars with the holes pre-drilled.  I'm going to use expoxy (JB Weld) to hold them too.  I thought about putting the rotor in the mill and cutting flat spots in it for the magnets to sit in.  But to keep things simple I decided against that   
--
Chris

[scoraigwind]

How is the power rating derived? Peak or average? and do you use 30 days a month 24 hours a day?

Power rating is based on the surface heat dissipation rate and on past experience.  That's maiximum contiuous power rating.  It's not sustained 24 hours nor 30 days, no.

Energy production (which depends more on wind and rotor size than on power rating) iis based on typical performance for this type of machine.  YOu need to know the site mean wind speed.

Both a bit rough but better than nothing and precision is impossible in this area anyway.

[ChrisOlson]

I got tired of working on combines this morning so marked out and drilled my rotor for pins.  I decided to go with simply pinning the magnets and using JB Weld after figuring the centrifugal force on the mags.  At the rated speed of this generator, 330 rpm, with the small rotor, the magnets wouldn't need any glue to hold them on the rotor if there was flats milled into the drum for the magnets to sit in.  The forces on this small rotor are considerably less than that on the mags mounted on 10 or 11" rotors.  So going to the work to countersink the sides of the magnets that don't come with countersinks, for small bolts to hold them on, is not needed.  They'll be fine with JB Weld and pins up to and beyond 750 rpm.  If I got 13 foot blades spinning at 750 rpm I'm probably going to want the mags to come off the gen rotor and lock it up anyway, because something obviously went terribly wrong someplace.

I didn't "cheat" and use the rotary table to drill it.  I measured and marked it out with a vernier caliper, prick punched it, remeasured, adjusted the punch marks until it was perfect, center punched it and drilled it with a hand drill just to prove it can be done that way.  I double checked my work on the rotary table and I got the holes accurate +/- .003".  When you're good, you're good.  When you're not, you're not  


--
Chris

[ChrisOlson]

Finished the raw powdered iron core tonight.  I made a couple "dummy mags" out of 1" x .5" bar stock to fit the rotor the core so I didn't have to deal with the forces of real magnets.  Next, the core has to be drilled and slotted for the winding wire.

[scoraigwind]

You may get away with making the machine totally enclosed, because the core is probably a pretty good conductor of heat to the outer surface.  That's one of the advantages of using converted motors with windings on the outer body.  The disadvantage is that the magnets usually seem to fly off but of course that can't happen in your case :-)

[dlenox]

Chris,

Looks like you did a nice job on the iron/resin core, I only see small air bubbles.  It's tough when you are dealing with thick pastry like consistency like you were referring to.  Did you use any pressure on the mold to help eliminate bubbles?

I am a bit curious why you did not mill the rotor flats for the magnets though, you trying to limit the amount of machining that you are doing here?  Or is it that the pipe wall thickness would not allow this, ending up with too thin wall?

I certainly don't consider using a rotary table as cheating - heck I feel that if you have the tooling then use it!  I know that it takes time and money to build up shop equipment.  I also have been luck enough to have a knee mill (w/rotary table) and metal lathe, and making/getting new tooling whenever the opportunity allows to ultimately save me time - you can't make more time...

Dan Lenox

[ChrisOlson]

Looks like you did a nice job on the iron/resin core, I only see small air bubbles.  It's tough when you are dealing with thick pastry like consistency like you were referring to.  Did you use any pressure on the mold to help eliminate bubbles?

I am a bit curious why you did not mill the rotor flats for the magnets though, you trying to limit the amount of machining that you are doing here?  Or is it that the pipe wall thickness would not allow this, ending up with too thin wall?

The stator cored ended up very nice with very few defects.  It takes a little practice to mix up and put the iron powder paste into the mold.  I used approximately 5 lbs of iron for every 9 ounces of resin and layed it up in sort of layers, troweling it out like cement with a putty knife, plus poking at it to get any air out.  Then I would sprinkle iron powder on that layer in the mold until it was covered and poke at it until that got mixed in before troweling in another layer.

It remains to be seen what sort of permeability I got with this mix - I'll find that out when I wind a test coil for it and run it.

I didn't mill the drum with flats because it's not really necessary.  The mags are pinned so they can't tip sideways anyway and the void underneath will be JB Welded.  Plus JB Weld is actually stronger when it has some thickness instead of being just a thin layer between two pieces of metal, and the stuff prevents corrosion between the magnet and the steel with age.  JB Weld is good stuff - I've fixed broken cam bearings in a twin cylinder Kohler engine with it before and 3,000 hours later the engine is still running with the cam bearing glued in place with JB Weld.

I'm going to spin test the rotor but I'm thinking it will take considerable rpm with split pins and JB Weld before those magnets will come off.  I've spun 10" rotors over 1,000 rpm with wedge magnets glued on with JB weld and they have held.  The rated speed of this thing is only 330 rpm and the drum is only 6.5" OD.
--
Chris

[fabricator]

I wonder what would happen if you mix the resin and iron powder with a thinner like MEK, mix it thin enough to make it liquid enough to pour, mix it in a bucket with a paddle on a drill motor so you get good thorough and even mixing, then pour it and put it in a container and pull a vac on it, and get all the air out.

[ChrisOlson]

I wonder what would happen if you mix the resin and iron powder with a thinner like MEK, mix it thin enough to make it liquid enough to pour, mix it in a bucket with a paddle on a drill motor so you get good thorough and even mixing, then pour it and put it in a container and pull a vac on it, and get all the air out.

I don't know how that would work.  It's not really all that bad to work with - it wasn't any harder than casting a traditional fiberglass stator really - just different.  The iron powder and resin makes a very thick paste sort of material that can be scooped out of the mixing container with the putty knife and put in the mold.

I only used half the normal amount of hardener in the resin (6 drops per ounce).  I ran out of resin the first day I started casting it and had to finish it the next day.  It warmed up a bit when it started to "cure" but within one hour after putting the final iron powder in the mold I pulled it out of the mold, had it chucked in the lathe, faced it and started cutting the center before it even cooled off.  It's pretty strong but I suspect like most powdered iron castings it's probably brittle and would break if you dropped it on the floor.

It transfers heat VERY well.  I got a wild hair today and blew some flame on it with the rosebud on the torch around one side of the ring.  The ring warmed up all the way around within a few minutes, and about 5 minutes after that test it had already cooled off to room temperature again.

I think the permeability of it is good too - I tried to magnetize the thing with a big horseshoe magnet and it won't hold any magnetism.  But if I place the horseshoe magnet on the inside of the ring it'll hold a nail to the outside of it.  Wiggle the magnet and I can make the nail move.  That same nail won't move if I put it on the floor and hold the magnet the same distance from it thru the air.  I assume that means the magnetic flux is traveling thru it quite well.

I also did another test on it - electrical conductivity.  Can't get an ohmmeter reading thru it anywhere.  So I assume all the iron particles are well enough insulated from one another so that it won't have significant eddy currents in it.

So far it's not a heavy generator - the rotor with no magnets on it weighs 5.2 lbs.  The stator core weighs 8.4 lbs.
--
Chris

[jeraklidis]

Are you gonna mill a negative mold something like this

and then pour in the mold?

[ChrisOlson]

Are you gonna mill a negative mold something like this

It'll be similar to that but it'll be a 18 tooth instead of 36.  And I'm going to use air-dry electric motor varnish to seal all the voids when I get the coils in it.
--
Chris

[ChrisOlson]

Stator core is ready for a test wind to see if I got enough room for the coil legs.  If it's too tight I'll have to machine the holes bigger.  This thing should only require 35 turns with 18 coils so I think it will fit.

[jeraklidis]

Stator core is ready for a test wind to see if I got enough room for the coil legs.  If it's too tight I'll have to machine the holes bigger.  This thing should only require 35 turns with 18 coils so I think it will fit.

(Attachment Link)

Interesting.... You know Electromagnetic Plastics are the next big thing...

Have you thought about making a heatsink for it?

Have thought of just winding one phase then attaching the rotor to a drill or a motor with a known rpm and calculating the voltage? Kinda like a pretest for the number of turns...

I'd really like to see how it preforms... It'll be interesting to see if and how the eddy currents will form... Do you have a temperature gauge to test it out?

[jeraklidis]

Stator core is ready for a test wind to see if I got enough room for the coil legs.  If it's too tight I'll have to machine the holes bigger.  This thing should only require 35 turns with 18 coils so I think it will fit.

(Attachment Link)

If it works you should increase the stator stack length and get longer magnets... it'll probably shave a few turns off each tooth....