MULE1.2 Axial flux test motor/bicycle specific
- Thread starter Thud
- Start date
Thud
1 MW
I am avoding all conductable materials in the stator san's windings. I started on the wrong foot in the OP :lol:
I am thinking about the mold itself & possible positioning jigs for coils. there will definatly be some blending of materials. There will also be some filler stuffs to take up any unavoidable dead spaces to reduce weight. Luke & I have talked breifly regarding "active air cooling" I can envision some duct work to allow air around heat generating sections. way to complex to attemp to describe verbaly. But quite simple to do in application.
If you can imagine vanes in the clam shell section of the rotors-we can make air flow thru the stator. water cooling is a cool subject
(bad pun sunday) but we certainly don't want it on a light platform like a bike.
Right now my vision of the stator:
A cast urathane foam core machined to fit all the requirments we want. It may be in several sections or just 1/2's
all the goods go into the final mold with glass fiber re-enforcment. Then its infused in a vacume to get a nice rigid assembly.
I am thinking about the mold itself & possible positioning jigs for coils. there will definatly be some blending of materials. There will also be some filler stuffs to take up any unavoidable dead spaces to reduce weight. Luke & I have talked breifly regarding "active air cooling" I can envision some duct work to allow air around heat generating sections. way to complex to attemp to describe verbaly. But quite simple to do in application.
If you can imagine vanes in the clam shell section of the rotors-we can make air flow thru the stator. water cooling is a cool subject
(bad pun sunday) but we certainly don't want it on a light platform like a bike.Right now my vision of the stator:
A cast urathane foam core machined to fit all the requirments we want. It may be in several sections or just 1/2's
all the goods go into the final mold with glass fiber re-enforcment. Then its infused in a vacume to get a nice rigid assembly.
liveforphysics
100 TW
I ordered them and sent paypal. 
I got you 40 of the 22.5deg section axial flux N42 4"x 1/4" thick magnets.
They are sent to the address I had the broach and freewheels sent. You only required 32 of them, but having worked with with magnets in the past, I don't think I've ever had a time when I didn't break at least 2 of them, and the next price break was at 40, so I sent you 40.
Enjoy my friend!
Order Invoice Number:
091122-125606-43916 Order Status: Order Submitted
Order Date:
Sunday November 22, 2009
Shipped To: Todd Lane Products: 1
Order Cost: $184.82
Displaying 1 to 1 (of 1 orders) Result Pages: 1
--------------------------------------------------------------------------------
My Purchasing History Quick Re-Order
Neodymium Magnets, 22.5 degree, 8 in OD x 4 in ID x 1/4 in thick Rare Earth Wedge Magnets for Wind Generators, N42
--------------------------------------------------------------------------------
I got you 40 of the 22.5deg section axial flux N42 4"x 1/4" thick magnets.
They are sent to the address I had the broach and freewheels sent. You only required 32 of them, but having worked with with magnets in the past, I don't think I've ever had a time when I didn't break at least 2 of them, and the next price break was at 40, so I sent you 40.
Enjoy my friend!
Order Invoice Number:
091122-125606-43916 Order Status: Order Submitted
Order Date:
Sunday November 22, 2009
Shipped To: Todd Lane Products: 1
Order Cost: $184.82
Displaying 1 to 1 (of 1 orders) Result Pages: 1
--------------------------------------------------------------------------------
My Purchasing History Quick Re-Order
Neodymium Magnets, 22.5 degree, 8 in OD x 4 in ID x 1/4 in thick Rare Earth Wedge Magnets for Wind Generators, N42
--------------------------------------------------------------------------------
liveforphysics
100 TW
Miles said:If the spirals were wound in opposite directions?
I need a coffee
Gotta wind in opposite directions
The whole double split coils thing seems like such an extra PITA. But ehh, it does solve the problem of using flat coils...
Thud
1 MW
hehe i do it all the time,
I am really anxious to get some finalisation on stator designs.
The first test is going to be a 6 coil 14gage wire coils. Since I am still obsessed by the LRK winding senario, I will terminat them so in the WYE configuration. I will have all leads coming out so delta is avalible for comparison.
Q: with a core in the coil, is there any more efficent EM shape than a true circle? given we are working with disc magnets?
I know there will be lots of discusion regarding lap winding's- I dont see that as a solution for this motor set up,given the scale we are working in.
-A small "bolt on" unit with several apps in addition to bikes-
My OPINION is :if we are going to get to that level of complexity we may as well make a true wheel motor & put a tire on it.
I am interested in Halbauch arrays, but I would rather some one present it completely thought through with a drawing, in a configuration applicable to the current prototype, to show us the simplicity. Right now it is a bit beyond my education level in that regard.
Lets see how varied the stator configurations become.
I can see more coils to get max copper fill. 12-15 in this small motor for sure.
I am anxious to see a corless version tested-There are several coil shapes to explore in that arena
I posted my concerns regarding the gage of the foil for "flat wound coils" .003 seems mighty thin even at .5 width
that =26 or 27 gage wire from my math.
John in CR
100 TW
Good one Miles, I wonder if any manufacturer has come up with winding coils that way.
About how many turns are we talking about on each coil? That's going to determine which is easier, wind separately flip one and connect at the center afterward OR wind one then the other from a single length of copper ribbon.
John
About how many turns are we talking about on each coil? That's going to determine which is easier, wind separately flip one and connect at the center afterward OR wind one then the other from a single length of copper ribbon.
John
John in CR
100 TW
Of course, but our torque machine in a wheel needs to be light, powerful, have the correct Kv, with torque to spare.
Let's do light, powerful and efficient with great cooling first, but yeah the perfect hub motor would be, just that.John
Thud
1 MW
LOL,
you'r absolutly correct John, I Just wonder how many forum pages would be filled finding that perfect set up? :lol:
you'r absolutly correct John, I Just wonder how many forum pages would be filled finding that perfect set up? :lol:
Jonathan in Hiram
1 kW
- Joined
- Oct 6, 2009
- Messages
- 446
Looking at this picture of the Etek motor it's clear that the magnets are separated to some degree, I posted a link to some FEMM animations a couple of days ago that show magnets right up against each other tend to leak magnetic flux to each other.
I think the wedge shaped magnets should be separated a bit from each other so they don't steal magnetic flux from their neighbors.
I think the wedge shaped magnets should be separated a bit from each other so they don't steal magnetic flux from their neighbors.
Thud
1 MW
I assume the trapizoidial would be max fill regarding cored units.
Corless is a lot of magic & trickery I assume. I am leaving room for a lucky guess or two in there also
I really am curious on the coil shapes & the effect on motor performance.
I wonder who has the modeling software for that? I hope Eric can illistraight it.
It may be something that only gets proofed in actual testing though.
Corless is a lot of magic & trickery I assume. I am leaving room for a lucky guess or two in there also
I really am curious on the coil shapes & the effect on motor performance.
I wonder who has the modeling software for that? I hope Eric can illistraight it.
It may be something that only gets proofed in actual testing though.
Thud
1 MW
Good observation JIH,
any starting points to suggest, I am drawing a motor with Lukes wedge magnets right now.
any starting points to suggest, I am drawing a motor with Lukes wedge magnets right now.
Jonathan in Hiram
1 kW
- Joined
- Oct 6, 2009
- Messages
- 446
Ah, I see the link worked correctly.
Something around 50% to 100% of the magnet thickness perhaps?
It seems fairly obvious that you reach a point of diminishing returns, just where that point might be is a matter of conjecture of course.
Something around 50% to 100% of the magnet thickness perhaps?
It seems fairly obvious that you reach a point of diminishing returns, just where that point might be is a matter of conjecture of course.
rhitee05
10 kW
I have some modeling complete for a few variations on the stator construction. As per Thud's plan, I assumed that the stator will use teeth made of ferrite or something similar shaped into teeth. I looked at a few configurations to see what the differences were. Model results at the end.
My interpretation is that using the offset structure we've been discussing doesn't seem to work on an iron-core stator with teeth. If you wanted to do lap-winding with overlapping coils, you'd have to do it old-school with with a lot of slots and multiple coils in each slot (PITA). Doesn't seem to matter if there's iron behind the teeth or not, there ends up being a lot of flux that leaks in between the coils and is wasted.
There don't seem to be any really significant differences between using two layers of coils with an iron ring between, two layers of coils with no magnetic material, or just a single thinner row of coils. Forming the teeth on the cores would be a little more involved. I think you could wind the coils onto the teeth individually and then assemble it all together, but doing the winding would still be painstaking.
Finally, an option to just have a 'plug' of ferrite in the center of the coils without the tooth part. This would be pretty easy to build, you could wind the coil on a jig and then just slide the plug inside and epoxy. It doesn't concentrate the flux as well as with the tooth. With the toothed designs you can use a smaller coil ID since the tooth tends to 'gather' the flux, but with this version I think you'd need to make the coil ID larger to encompass the entire height of the magnet. It would end up looking similar to an ironless motor, but having the plugs would give a smaller effective airgap length.
My interpretation is that using the offset structure we've been discussing doesn't seem to work on an iron-core stator with teeth. If you wanted to do lap-winding with overlapping coils, you'd have to do it old-school with with a lot of slots and multiple coils in each slot (PITA). Doesn't seem to matter if there's iron behind the teeth or not, there ends up being a lot of flux that leaks in between the coils and is wasted.
There don't seem to be any really significant differences between using two layers of coils with an iron ring between, two layers of coils with no magnetic material, or just a single thinner row of coils. Forming the teeth on the cores would be a little more involved. I think you could wind the coils onto the teeth individually and then assemble it all together, but doing the winding would still be painstaking.
Finally, an option to just have a 'plug' of ferrite in the center of the coils without the tooth part. This would be pretty easy to build, you could wind the coil on a jig and then just slide the plug inside and epoxy. It doesn't concentrate the flux as well as with the tooth. With the toothed designs you can use a smaller coil ID since the tooth tends to 'gather' the flux, but with this version I think you'd need to make the coil ID larger to encompass the entire height of the magnet. It would end up looking similar to an ironless motor, but having the plugs would give a smaller effective airgap length.
rhitee05
10 kW
Jonathan in Hiram said:
Jonathan in Hiram said:
The magnets definitely need some spacing between them. You probably want them at least a little further apart than the airgap is wide, so the flux will tend to prefer flowing through the airgap. The animation you posted is nice, so I won't bother re-creating it.
rhitee05
10 kW
Thud said:
That'd be my assumption. If memory serves correctly, the wedge magnets Luke bought are 8" OD and 4" ID. Assuming you're planning a fairly high number of poles for lower Kv, you'll probably end up with tall, narrow windings. That's more optimal anyway, since the inside/outside edges don't contribute anything but resistance.
I'd like to try some modeling to look at this, but that'll take some more time learning how to use the software package before I can model actual motor parameters like torque.
spinningmagnets
100 TW
Thud, I have a few thoughts to consider. Since you are using wedge magnets (rather than round) you have the option of leaving them only half-potted. If a small amount of magnet face is exposed, it can act in a way similar to the vanes of an automobiles front disc brake, and fling cooling air from the center to the outer edge. If the airgap to the stator is fairly close, this airflow can impart significant cooling to the stator.
Over at the DIY wind-gen sites, the best resin is expensive. Many add either talcum powder or marble dust (I will look for the reference of filler type and percentages). This acts as a filler to reduce costs, and also improves stator heat shedding.
One axial flux motor site I read recently had adjustable air-gap. The closest air-gap (between rotor magnets and stator) provided the highest torque, but didn't specify max airgap benefits. Perhaps less cogging for iron-cored coils, or less stator heat during partial load? I can't image a DIY AF-motor with air-gap adjustability on-the-fly, but it may be handy on the bench-mule to have the rotor standoffs be threaded for several different fixed air-gaps performance assessments.
Positioning strong N40 magnets of that size may prove difficult. You might consider making an aluminum/wood template with slots for 7 of the 14 magnets. And once the glue is dry, a second template with 14 wedge slots to position the remaining 7. It might be worthwhile to make a steel plate similar in shape to the 7-slot template to bolt on top of the glued magnets to short the flux of the first 7 during the positioning of the last 7 (just an idea, I don't know).
There are dozens of pics on the web of fingers severely damaged by magnet mishaps...I want your fingers to stay in working order, young man!
Over at the DIY wind-gen sites, the best resin is expensive. Many add either talcum powder or marble dust (I will look for the reference of filler type and percentages). This acts as a filler to reduce costs, and also improves stator heat shedding.
One axial flux motor site I read recently had adjustable air-gap. The closest air-gap (between rotor magnets and stator) provided the highest torque, but didn't specify max airgap benefits. Perhaps less cogging for iron-cored coils, or less stator heat during partial load? I can't image a DIY AF-motor with air-gap adjustability on-the-fly, but it may be handy on the bench-mule to have the rotor standoffs be threaded for several different fixed air-gaps performance assessments.
Positioning strong N40 magnets of that size may prove difficult. You might consider making an aluminum/wood template with slots for 7 of the 14 magnets. And once the glue is dry, a second template with 14 wedge slots to position the remaining 7. It might be worthwhile to make a steel plate similar in shape to the 7-slot template to bolt on top of the glued magnets to short the flux of the first 7 during the positioning of the last 7 (just an idea, I don't know).
There are dozens of pics on the web of fingers severely damaged by magnet mishaps...I want your fingers to stay in working order, young man!
rhitee05
10 kW
spinningmagnets said:
Increasing the airgap will decrease flux intensity and increase motor Kv. If it's adjustable on-the-fly, it gives you field weakening capability for a larger speed range.
I don't understand English language very well so this did not become very clear to me. Do you mean that if we, for example, have a one-stator-two-rotor yokeless axial machine (like ISIS), the stator shouldn't be formed of radially placed stacked iron strips ? The Lynch motor has such tooth in the rotor between the copper strips and the motor is known to have good efficiency.rhitee05 said:
And here is also one machine with very high efficiency:
http://www.ece.wisc.edu/~lipo/1993pubs/93-48_97.pdf
I could understand the leakage flux from one stator pole to another through the slot would cause high eddy currents in radial plates, not the main flux flowing through the strips from rotor to rotor.
John in CR
100 TW
I'd like to be able to contribute more here, but I need to understand some things in order to visualize them better. Specifically what is the actual timing of these motors and the shape of fields from different coil shapes? ie at what position is the coil relative to the magnet when the coil gets current if the timing is set to neutral?
I really want to try my hand at an overlapping winding like the advanced production motors, but using a thin carbon or fiberglass disk as a base to make the windings relatively easy though tedious.
Please just point me in the right direction to a good resource to catch up a bit.
John
I really want to try my hand at an overlapping winding like the advanced production motors, but using a thin carbon or fiberglass disk as a base to make the windings relatively easy though tedious.
Please just point me in the right direction to a good resource to catch up a bit.
John
kenkad
100 W
Hello,
Thanks to Miles for his (and anyone else that contributed) to the double wrap flat copper foil winding wrap drawing. This is the one thing I am/was missing (not anymore though) for my windings in the AFM motor I am designing. This makes connecting delta nodes for me much much easier. I am interested in using multiple 3phase groups in a AFM design. Again many thanks.
kenkad
Thanks to Miles for his (and anyone else that contributed) to the double wrap flat copper foil winding wrap drawing. This is the one thing I am/was missing (not anymore though) for my windings in the AFM motor I am designing. This makes connecting delta nodes for me much much easier. I am interested in using multiple 3phase groups in a AFM design. Again many thanks.
kenkad
Similar threads
