Question re feasibility of casting a stator from...

Danny Mayes

Danny Mayes

1 kW
Joined
Feb 16, 2011
Messages
321
Location
Perth, Australia
...resin and iron powder.

Why? -

My understanding of the upper limit for the speed of an outrunner is iron losses (forgetting mechanical limitation for the minute).

If you reduce lamination thickness then you reduce the hysteresis and therefor the losses increasing the upper frequency limit. Is that accurate?

If you cast the stator with resin and iron powder would you not eliminate hysteresis altogether? Eliminating the upper limit all together? Leaving only the mechanical limit?

Is the problem that a cast stator like this would saturate earlier than a laminated one? And this is why it is not done?

Is the saturation of the stator a function of current? Is that what provides the current limit (forgetting heat for the minute)?

Any input form brains more equipped (or less :D ) than mine for this is much appreciated.

D
 
Hi Danny,

It's eddy current losses that go up as the fundamental frequency squared. Hysteresis losses go up in direct proportion to the fundamental frequency. It's eddy current losses that benefit from laminating the stator. Losses from eddy currents go up with the square of the lamination thickness.

For conventional motor designs, even the best commercial composite cores don't compare with laminated steel, until used at quite a high fundamental frequency.
 
Miles, Thanks for the swift reply and great data.

I thought that hysteresis and eddy current losses were the same thing so I have learnt that at least... :oops:


What I understand from those PDFs leaves me thinking that this is an ideal core for a variation of a RC type outrunners. It is the permeability that is its downfall right? If eddy current losses go up at the square, and this core material reduces them so much, can we not just spin the motor faster to make up for the losses in permeability? And theoretically, faster again for power gains? I am sure the answer is not so basic but from my understanding eddy current losses are the thing limiting the top speed of RC type motors and the faster we can spin them the more power we get? I am sure geometry is also a factor. I am envisaging a longer motor like the 80/100 but twice as long (80/200) rather than bigger diameter and shorter width like the CA120 for example.

D
 
It's always good to explore possibilities but if there were significant gains to be made in any direction, you can be pretty sure someone would be onto it....

A long inrunner would need bearing support at the distal end of the bell. If you intended to run it at a very high speed, the bearing would need to be a much smaller diameter than the can, so you'd end up with something like a sausage shaped hub motor, with the wires entering via a hollow shaft.

You can reduce the lamination thickness for steel cores, to ameliorate the eddy current losses, without affecting the flux capacity significantly.

SMC cores will have higher hysteresis losses than steel cores.

SMC cores are inherently fragile.

If you want to experiment, you could buy a few of the Hoganas blanks, machine them and then stack them to get the length you require.
 
Nothing came out of Hobby Kings' hype :mrgreen: :

http://www.hobbyking.com/hobbyking/forum/forum_posts.asp?TID=5212&PID=33371#33371

http://www.hobbyking.com/hobbycity/store/twit.asp?id=120
 
Thanks again Miles.

Yes, a sausage is what I am thinking (even 400mm long would be ok for my application). Small outer diameter for high rpm. No shaft, the bell has a tooth profile for extracting the power, enabling the core to be supported at both ends. I guess this speculation needs calculations to really work out the feasibility. I need to study more...

Is it sounding anymore feasible yet?

Btw, I am not attached to the powdered core, just exploring something, like you say.


D
 
I'm going to park this. I can easy enough model the mechanical aspects. I need to learn how to model electro magnetic...

Thanks again Miles for the input. Much appreciated.

Regards,

D
 
The iron losses you are thinking about are eddy currents, not hysteresis. In transformers eddy currents are reduced by making the iron sheets thinner. The thinner they are, insulated by the "glue", the smaller the eddy currents are in terms of actual physical size. The eddy currents are actual circular flows, loops, of electric current. They are induced by the magnetic fields "touching" the laminations of iron sheet.

In your plan the epoxy resin would be the insulator and indeed could theoretically eliminate eddy currents. You have the not small problems of getting enough physical strength and heat resistance from the resin.

Hysteresis is the property of a magnetic material to lose its magnetism as soon as the inducing magnetic field is removed. Some magnetic materials lose their induced magnetism slowly, some quickly, and of course some retain the magnetism "permanently".

I genuinely hope you have very good luck developing your idea!
 
Hey doctorGONZO,

Thanks for the input.

I need to get myself some modelling software. There are so many different factors and without calculations i am just guessing.

I think the strength and heat issues can be overcome. That area is my strength. The finer details of electro mechanical design is not :D

Theoretically eliminating eddy currents appears hugely advantageous for increasing power density but there are so many other factors too that i don't really understand yet.

Like Miles said, if it was that advantageous, people would be on to it. I am not so sure if it is worth perusing. I was interested in why it is not being done and I have some idea of that answer.

I think much higher speeds are feasible mechanically but then there is the problem of the controller. Are higher speeds possible there? I don't really understand the relationship of poles and fundamental frequency or e-rpm? I know that there is a limit in what controllers can do though. I know enough to be dangerous but more study is required :mrgreen:

But the time in exploring all of this is probably better spent just adapting existing motors to my application. :lol:

D
 
Danny Mayes said:
I don't really understand the relationship of poles and fundamental frequency or e-rpm?
Click to expand...
That's easy enough. You divide the rpm by 60 to get revolutions per second and then multiply that figure by the number of pole pairs (p/2) to get the fundamental frequency in cycles per second. erpm is just rpm times the number of pole pairs.
 
Danny Mayes said:
So what is the advantage to having more poles?
Click to expand...
That's not such an easy question.... :) There is a modest torque gain - compounded through better use of active materials. For the single simulated comparison that I did, with identical stators, it worked out as the cubic root of the pole ratio difference.
 
So is it fair to say, to increase the power density of a motor, reduce the number of poles, therefore reducing the fundamental frequency and then increase the Kv of winding to get a faster spinning motor? I have a suspicion that it is not that simple :lol:

I am trying to find what I can on google so i don't have to ask such noob questions, but i kinda don't know what to search for.

Would an RC outrunner be classified as an AC synchronous motor?

Can you point me to some literature that outlines this stuff?

Thank you.

D
 
There are lots of other factors to consider. It's all about finding the best compromise.

Best to refer to "Permanent Magnet Synchronous Motor" (PMSM). The AC DC can be confusing when applied to brushless motors.
 
You must log in or register to reply here.
Back
Top