BLDC Flux ring thickness

[Tench]

Is there a basic rule governing the thickness of the magnet backing ring?

I wish to lighten an HT3525 motor, it has 6.85mm thickness to the steel ring in the thinnest area, in the centre between the spoke flanges, the magnets are about 3mm thick, there is no detectable magnetism present on the outside of the ring. Can I reduce this thickness without having a significant effect on the motors performance? and any ideas on what I can reduce the thickness to?

 

[Miles]

I guess you can reduce it until it actually starts to affect the performance...... If the airgap flux falls, the motor will run faster for a given voltage.

What is the circumferential width of the magnets?

 

[Tench]

Hi Miles, reducing the thickness until it affects performance is not an option as I will have to build it into the wheel, yes this one is going to be used as intended, modified but still a frocker!

The magnet ring measure 35mm wide, each one is 14mm x 3mm, the backing iron over hangs at least 6mm each side. the dia inside the magnets is 199mm with a 1mm air gap.

https://lh5.googleusercontent.com/-1vNAY1zFOrs/UCr1wmrWaNI/AAAAAAAAJ6o/ZbrzcMhkP1E/s864/20120814_210358.jpg

that's not my pic, borrowed it, but it shows you the magnets.

and the outer case profile;
http://static.electricbike.com/wp-content/uploads/2012/05/crysalyte-1024x765.jpg

 

[Miles]

Why not run it unloaded, out of the wheel? You just need a way of accurately measuring the speed.

 

[Tench]

Yes I could go through a machining and testing routine but I was hoping someone would drop by and say yes its safe to go to 1.5x magnet thickness for example, allowing me to build the motor once.

 

[Miles]

You're not going to get a simple rule of thumb like that. You'd need to take account of the pole width and magnet strength, too. If you knew, or could measure, the magnet strength, you could simulate the flux density in FEMM.

The flux density in the back-iron varies. It's maximum at the point between the poles and minimum at the centre of the pole. If you really want to use the least material possible, you need a wave shaped back-iron...which is what I've done here:

 

[zappy]

Also have a quick read of this post by me.

http://endless-sphere.com/forums/viewtopic.php?f=2&t=55211

I have done several cro motors and magic pies and take them down to around 3 to 3.2 mm thick and the KV stays the same. Some flux leakage can be felt after the flux ring is installed on the stator after modifying it to be this thin. flux ring weight can nearly be halved which can help a lot off road etc. This seems like some of the best bang for buck/time for removing weight from our bikes. Especially as it goes round and round and up and down it has a proportionally larger effect on our bike dynamics than say a carbon fiber frame would have. My aluminium down hill frame only weighs around 2.5 kg a carbon equivalent may save 700 grams?
All the standard flux rings i have machined down seem to be mild steel hollow bar so it machines quite easily and is way strong enough for any wheel hub loads even when thinned way down. Masking tape up the magnets so the lathe swarf doesn't stick annoyingly to the magnets and do not grip the magnets in the chuck, they are brittle and will probably crack. Let us all know how it goes and the approximate weight you have saved.
Good luck
Zappy

 

[fechter]

Sort of depends on the permeability of the steel. If you take off enough, you'll start getting saturation and flux leakage. This will change the Kv of the motor (will spin faster).

If you don't have significant attraction to steel objects on the back side, then it's thick enough. Any thicker won't help much.

Under very high load conditions, you have a higher flux than you have with no power, so you need a little extra thickness to avoid saturation under high load.

 

[Tench]

thanks guys, Zappy I have turned these before to remove spoke flanges etc so not new to mounting these in the lathe but thank you for your concern. The Crystalyte HT/S ring from the 35mm motors weighs 2.900kg with the magnets still in. I would be over the moon if I could get it down to 2kg! I am going to take it down to 3.5 mm. the edges will have to remain a little thicker to accommodate the side cover screws and the flanges add thickness in some areas and also strength to the overall structure. I have saved 500g elsewhere in the motor too.

Miles I had seen your motor design, I see the backing ring although modelled to be exactly what is required is less than the magnet thickness. Maybe some of these hub motors are under developed and over engineered?

if a slight rise in the Kv occurs then that would not be a bad thing as this is going in quite a small wheel.

 

[Miles]

Miles I had seen your motor design, I see the backing ring although modelled to be exactly what is required is less than the magnet thickness. Maybe some of these hub motors are under developed and over engineered?

if a slight rise in the Kv occurs then that would not be a bad thing as this is going in quite a small wheel.

You can't really use the the magnet thickness as a guide. The flux from half the magnet face area is channeled through the back-iron. For my motor half the magnet width is 3.5mm which is a bit more than the maximum depth of the back-iron. For your motor, half the magnet width is 7.5mm.... My magnets are 4mm thick yours are 3mm.


The problem is that if Kv rises Kt falls, so you need more current to achieve the same torque whilst the resistance stays the same. Copper losses go up with the square of the current............

I knew I'd seen a reference to someone safely removing a substantial amount from the back-iron but I had forgotten it was zappy!

 

[Tench]

Miles, thank you, I am becoming more enlightened all the time, I can see now how the magnet width has more relevance than the thickness. I will give this some more thought before diving in! but I still want to get some weight of it!

 

[John in CR]

Does thinning the backing ring with no change in Kv really tell us that we haven't changed the motor other than its weight? ie does the flux leakage change as we apply more current and make the electromagnets of the stator stronger?

Why would the Chinese use so much more steel than is necessary? That doesn't make sense to me unless the Chinese associate higher quality to more weight in the same erroneous manner that most in the west perceive higher quality with higher price.

 

[Teh Stork]

The "flux ring" is in reality steel pipe of appropriate thickness and diameter - if they need 3,25mm and only find 5mm thick steel tubing, that is what you're going to get.

 

[Miles]

Under very high load conditions, you have a higher flux than you have with no power, so you need a little extra thickness to avoid saturation under high load.

Does thinning the backing ring with no change in Kv really tell us that we haven't changed the motor other than its weight? ie does the flux leakage change as we apply more current and make the electromagnets of the stator stronger?

Looking at the results of my virtual testing, the rotor flux density varies from 1.6T at 20 Amps to 1.7t at 200Amps. 20 Amps being peak Eta and 200 Amps being the point where the stator core is saturating.

 

[Tench]

I took the motor outer ring to the machinist the other day, we have come up with a plan to get some weight of it without spending too much time on it. I brought it away with me to do some drawings then it is going back to them to go on a diet. We came up with a plan where one simple cnc program can be easily modded to do the 3 opps. The thinnest the backing iron will be after machining is 5mm so not a drastic reduction in thickness, but it is going to be flower pettled around the spoke holes too. When it is back together I will measure the KV.