ES Motor Project?

[bigmoose]

Don't know if I shared this previously, these guys appear to have "stock" laminations. Inrunners here: http://www.rbourgeois.com/induction_motors_laminations.a234.en.html
and outrunners here: http://www.rbourgeois.com/dc_motors_laminations.a237.en.html

I never contacted them directly to inquire price/availability. They were showing drawings and stock numbers, so I assumed they already had tooling for the designs shown. Perhaps one or two are relevant? There are some outrunner diameters near 3.2 inches and some inrunners up to 5 inches rotor diameter.

[toolman2]

so yep, just saw your post, "What is the range of nominal speeds we need to cover?

I'd say that we need to make something that is optimised for speeds between 2000rpm and 4000rpm but is still usable at around 6000rpm.

With the 16 poles, flux frequency at 4000rpm is 533Hz - so iron losses will still be relatively low for 0.2mm lamination's. At 6000rpm it would be 800Hz.

these .3mm lams sit in that rpm range, running happily at 5-6000rpm, but with a 30mm stack of them and a 10w blower fan it can make 15.4NM continuous std and 20NM rewound. they weigh 2.4kg, 120mm can od, and will output 6kw cont and compare surprisingly well with a joby J1S.
is a 120mm can not quite close enough to the 127mm that your after,or are you just trying to be difficult?

it could be worth asking the factory about a lam deal..
they did write back about price of a bigger version, so i wonder if they would be happy to stamp us some thinner lamination's for more like 8000rpm.ca 120

[Miles]

yes, they appose each other like that so you would think it may go either way, but a comparatively large amount of copper is lost for a small increase in airgap radius.

That's true.

The copper area increases towards the outside. Better iron:copper ratio.

You are also increasing the airgap area.... I guess you'd need to reduce copper volume even more to take advantage of it, though...

I was curious....

Core OD 118mm (see illustrations below)

Airgap radius ------ Area for copper
59mm................164mm²
64.9mm (+10%).....153mm² (-6.7%)
70.8mm (+20%).....136mm² (-17%)

For this example, the loss of copper area is less than proportionate to the increase in airgap radius.

The proportions of the Joby motor are making more sense.

[Miles]

these .3mm lams sit in that rpm range, running happily at 5-6000rpm, but with a 30mm stack of them and a 10w blower fan it can make 15.4NM continuous std and 20NM rewound. they weigh 2.4kg, 120mm can od, and will output 6kw cont and compare surprisingly well with a joby J1S.
is a 120mm can not quite close enough to the 127mm that your after,or are you just trying to be difficult?

That looks interesting. What's it from?

[nieles]

looks like the stator form a ca120-70

[Lebowski]

Airgap radius ------ Area for copper
59mm................164mm²
64.9mm (+10%).....153mm² (-6.7%)
70.8mm (+20%).....136mm² (-17%)

For this example, the loss of copper area is less than proportionate to the increase in airgap radius.

The proportions of the Joby motor are making more sense.

I think you're looking at it wrong.
Let's say we change airgap radius and area for copper, like you
did, but we look at the change in power dissipation.

The following assumes an equal amount of windings (which
is a valid assumption as it has no influence).

First the airgap, let's assume an increase by N.
With the radius * N, for the same torque the force F must be divided by N.
Force is proportional to current so current is divided by N.

For the copper resistance, lets assume a copper area decrease by a factor M.
This means the winding resistance is increase by M.

P_loss_orig = i^2 * R

P_loss_new = (i/N)^2 * R*M

So to judge different configurations you need to compare the squared increase in
airgap radius (N^2) with the linear decrease in copper area (M)

[Miles]

So to judge different configurations you need to compare the squared increase in
airgap radius (N^2) with the linear decrease in copper area (M)

Thanks Leb,

I thought about this yesterday and wrote:
"Also, although torque and current are proportionate, losses go up by the square, so that creating the same torque, by using more force, with proportionately more copper, at a proportionately lesser mechanical advantage, will be less efficient. Isn't that right?"
But I wasn't sure, so I deleted it....

[jonescg]

Okay, so me chipping into this discussion will be about as productive as "Who wants lemonade?"

But silly question; are the laminations going to run radially? That is, a series of circular shapes like above will be cut which can be stacked up to make a motor as thick as you like?


Edit - I know my radial from my axial

[Miles]

Yes.

[jonescg]

I'm learnding!

[toolman2]

So to judge different configurations you need to compare the squared increase in
airgap radius (N^2) with the linear decrease in copper area (M)

Thanks Leb,

I thought about this yesterday and wrote:
"Also, although torque and current are proportionate, losses go up by the square, so that creating the same torque, by using more force, with proportionately more copper, at a proportionately lesser mechanical advantage, will be less efficient. Isn't that right?"
But I wasn't sure, so I deleted it....

good question,
i spose no simple statement like more copper and less leverage gives higher efficiency or vice versa can be known to be true without knowing the exact proportions of each to begin with as well as a torque level that we are judging the efficiency at etc etc.

then theres optimizing the finer points of the new configuration, like say the really long teeth (with a huge wad of copper on it that can clearly generate more flux) would surely need to also be a thicker tooth to be able to carry this now much higher flux level that on top of that also needs to travel a greater distance, this thickening would keep all of that back in balance but now has higher rpm losses from eddy currents..

and the same goes lots of little details like that, and then all the other variables that might now want adjusting to 'fit better' with the new config and some of those next changes might have been beneficial anyway and others allready pushed further in the wrong direction.

in the end there has to be a point at which a further change in either direction is taken too far and gives higher losses, so have a look at the two extremes to see where its problems may lie.

more copper has the general habit of lowering ir losses and giving the motor a higher continuous torque output, and more flux radius (as well as greater leverage) has a higher peak saturation, so more ability for burst power.

what i mean is, i dunno..

but yes, its fair to say joby would have gotten it right, so thats got to be a good direction to head.
and that green stator pic is a ca120 (106mm stator od and 30mm thick) and with the very different layouts between this the 1.8kg joby jm1s its an interesting
comparison, i have good info on both, will post up.

[Miles]

I'm getting keener on the idea of an inrunner again......

126mm OD (ex 5"x 3/16"wall tube)
120mm stator OD
circa 80mm Rotor OD

[Thud]

Just recieved a reply for Arnold magnetics regarding their Arnon NGOES in 8 mil....they will roll me a sample & want to know what min I would need.
They havn't quoted a cost yet as there are a few undetermined variables....But I would think enogh for 400 laminations would be a good starting point. (for my pesonal project hehe, thats roughly 80 square feet or 170 feet of 5" wide materal)
6ish sq meters?

don't get your hopes up yet.........

[Thud]

I'm getting keener on the idea of an inrunner again......

126mm OD (ex 5"x 3/16"wall tube)
120mm stator OD
circa 80mm Rotor OD

Have you sketched a tooth/pole count yet?
your thinking a nice open architechure joby style?

[Miles]

Could be open or enclosed. I'm keen to explore the embedded magnets idea (like the Headline).

[TylerDurden]

FWIW, the shop also has a cnc pcb cutter. That might be able to cut lams?

[Miles]

.....and that green stator pic is a ca120 (106mm stator od and 30mm thick) and with the very different layouts between this the 1.8kg joby jm1s its an interesting
comparison, i have good info on both, will post up.

Great! Thanks!

[Thud]

Soo the rotor is a laminated part also eh?
very ambitious for a diy project.
here is my take-off on a 18t 16p inrunner with standard magnets & steel (I happen to have in the drop box already)

5inch 18t 16p inruner.dxf

quick take off

(69.07 KiB) Downloaded 35 times

TD,
Hmmnn...hand't considerd cnc cutting lams figuring run time would be prohibitive.....holding the thin material is the real challange, especialy on cut-thru parts like this.

I have done some very small & ornate wood moulding on the cnc (doll house scale stuff) I bonded wood to a sacraficial pc of .25"t mdf to be able to hold it for machining....then a quick dip in acetone to release the glue line...it is a possible senrio also. just need ot put a vacume clamp on the table top cnc....or even make a custom vacume spoil board for running the lams on my bench top.....need to time study the chemicle etch vs cnc run time for better idea of productivity.

keep the ideas coming.

[Miles]

I'll definitely have to increase the diameter of my rotor, so it's bigger than yours. 84mm it is then

[TylerDurden]

Hmmnn...hand't considerd cnc cutting lams figuring run time would be prohibitive.....holding the thin material is the real challange, especialy on cut-thru parts like this.

Ja, that's the nice thing about the laser, it doesn't push the work around.

[Miles]

holding the thin material is the real challange, especialy on cut-thru parts like this.

Magnetic chuck and sacrificial layer? You might be able to cut several layers together?

[Lebowski]

with the embedded magnets, how are you going to prevent the magnetic field from
shorting out via the 'magnet holder' ? If I was the magnetic field I would take the
shortcut instead of jumping the airgap and taking the long route through the stator.

[Miles]

If it doesn't look as if it would work well enough, I'll try something else. I'm intrigued by the idea and want to simulate it, though...



Ref: http://mrbill.homeip.net/albums/cyclone ... ndex4.html

[Lebowski]

If it doesn't look as if it would work well enough, I'll try something else. I'm intrigued by the idea and want to simulate it, though...

I think the clou is to have the bits next to the magnets very thin so they are oversaturated...

[liveforphysics]

The Chevy volt rotor is setup very similar to that, but the magnet bars are in pairs tipped at about 45deg angles to the face.