4x less coasting resistance

Joined
Apr 25, 2014
Messages
2,782
Location
Sausalito
Maybe it’s old news and a thread somewhere but where is this used or why isn’t it? Less low speed control?

https://www.degruyter.com/view/j/phys.2018.16.issue-1/phys-2018-0008/phys-2018-0008.xml#j_phys-2018-0008_fig_001
 
Cogging torque is not what creates the drag when coasting. Cogging torque can be thought of like a ball rolling up and down a ramp as the rotor magnets move in relation to the stator. The ball rolling up and down a ramp has equal energy given back when it rolls down the ramp as it required to roll up the ramp. Many performance motor designs have almost undetectable cogging torque, like the motor on a Zero motorcycle, but this has no positive or negative performance impact (or coasting resistance impact) on the motor, but helps the bike to feel smoother when it rolls at low speed.

The drag when you're coasting is due to hysteresis iron core losses and eddy current losses.
 
Changing the shape of the stator is one way to do it.
Going with razor thin laminations is another.

Combine both and you have a DD motor that spins like a freewheel in it and achieves 90-95% peak efficiency..

Too bad mainstream motor makers aren't so interested in these things..
 
Ooo du. Thank you

Looking on emetor could I equate iron losses to this resistance while coasting


But motors with more poles have less resistance while coasting but would have more iron losses for a given rpm with all other parts of the design the same no?
 
neptronix said:
Changing the shape of the stator is one way to do it.
Going with razor thin laminations is another.

Combine both and you have a DD motor that spins like a freewheel in it and achieves 90-95% peak efficiency..

Too bad mainstream motor makers aren't so interested in these things..

I doubt it's as easy as that. Otherwise, some better-funded industry would be making motors like that for another application. Are there any?
 
C9DED185-DC4E-4500-8FEF-A77D306924C1.jpegYea are there any!!? :mrgreen: I’ve been looking for something to shoot for rr.. copy, and get some epdm stators made to try for skate hub motors.
 
Hummina Shadeeba said:
Maybe it’s old news and a thread somewhere but where is this used or why isn’t it? Less low speed control?
https://www.degruyter.com/view/j/phys.2018.16.issue-1/phys-2018-0008/phys-2018-0008.xml#j_phys-2018-0008_fig_001
IIRC that's exactly what the old Tidalforce motors did - some odd number of salient poles on the stator. (14? I think there were 7 individually commutated phases.) These did not quite line up with the 16 magnets. There was almost no cogging.
 
tidelforce may have a motor that doesn't have noticeable ratchetting of magnets at very slow speeds that you would feel, which is nice, but really im trying to find how to reduce the drag on the spinning motor at all speeds while unpowered. Is this enherent with a permanent magnet motor? what kind of stator design features, literally shapes, do what? i should really get back on fussion but got annoyed learning how to run it again importing shapes.
 
There are various ironless motor designs that eliminate the iron losses, but most of them have cooling issues due to low mass of the stator. Induction motors or SR motors would also have near zero coasting torque.
 
Chalo said:
I doubt it's as easy as that. Otherwise, some better-funded industry would be making motors like that for another application. Are there any?

That's exactly what happened with the 9C motors when they switched to 0.35mm lams from 0.5.
Rolling resistance dropped dramatically.
Efficiency went up >5%. Continuous power went from 500w rated to 750w rated.
You can see this on dyno graphs of motors before and after.

Similar thing with the 35mm sized bike hubs. An early crystalyte 35xx used to cog like a bitch and only hit 83% peak efficiency. Continuous wattage around 1200w.

Leafmotor comes out.. same size.. tighter windings.. 0.35mm lams.. way less cogging ( you can now pedal it with minimal penalty ), 1900w continuous capable, and 91% peak efficiency.

Difference in cost for better laminations is on the order of $20 for the manufacturer, so it's not an expensive change.

Bigger DDs would probably benefit from 0.27mm, but beyond that, diminishing gains in efficiency, power, and rolling resistance.

DD motors do not spin fast nor have high power density, so smaller motors spinning quickly benefit from thin laminations more, because saving another 50 watts of heat production is critical when your motor is the size of a fist.
 
I already use .2mm lams. I forget the silicone percent.
You think it would be worthwhile doing a powder core or will the hysteresis be the dominating loss anyway? Or a hallbach array seems an easy plus

As it is now I’ll get maybe one amp for a no-load test at 7000erpm. (1000rpm) but things depend greatly on the bearings used and getting them installled with the races aligned
 

Attachments

  • BD9D92B6-1162-43AF-BEB8-DB3E38586A44.jpeg
    BD9D92B6-1162-43AF-BEB8-DB3E38586A44.jpeg
    213.3 KB · Views: 3,071
Hummina Shadeeba said:
tidelforce may have a motor that doesn't have noticeable ratchetting of magnets at very slow speeds that you would feel, which is nice, but really im trying to find how to reduce the drag on the spinning motor at all speeds while unpowered.
Out of the box idea -

Make a motor where the stator + windings can be slid sideways off the rotor via a brake cable or something. The more you slide it, the lower the freewheeling resistance (and the higher the base speed, if that's useful.)
 
This might be worth reading:
https://metglas.com/wp-content/uploads/2016/12/Advanced-Materials-for-Motor-Laminations.pdf

Traditional iron-based, soft magnetic materials have been used for laminations in electric motors for over 100 years. Such materials provide excellent manufacturability and adequate magnetic performance, but exhibit more losses than are ideal when greater efficiency is also a design goal. The demand for higher motor efficiency is therefore a catalyst for research into better performing lamination materials. A number of material choices exist, such as Nickel-Irons and Cobalt-Irons, but these materials are generally expensive, compared with traditional silicon iron materials. However, there are other candidate materials with costs more in line with standard motor lamination materials. Two of these are amorphous iron and nano-crystalline iron formulations. These materials exhibit superior magnetic performance combined with reasonable cost. While cost-effective in their “as cast” form, these materials are challenging to manufacture into traditional motor structures because currently they are only readily available in a thin (25 micron) ribbon format and exhibit very high hardness. This creates an issue of how to design and construct a cost-effective motor utilizing these advanced materials. This paper will outline past and current efforts to build commercial motors with these materials and will also project potential future paths for developing cost-effective motor structures that utilize these high performance magnetic materials.

My guess is if powdered iron or ferrite was a better core material, somebody would be using it.
 
I’ve searched the crap out of powdered cores and it’s not for me although with the right 3D shaping it could be good it shows
https://pdfs.semanticscholar.org/c950/ba1a6213782d95258d58d90cd6ad2da7ab7a.pdf

Need that high mag field and high torque. Thanks for all your advice guys
Last question: with the same goal of easy coasting how well u think the hallbach would do? That’s a hard one to answer without details I know but given that I’m using N48sh at 3mm thick and something like 95% fill, and this weird shaped back iron of 1020 steel still will grab iron from the outside even when I position the magnets in front of the gap which is optimal. I assumed the 1mm thick gap with 3mm thick “teeth” between the gaps would do better as a flux ring. Surprised I never see hallbach arrangements, I guess because it’s work to make.

I’d assumed 1mm thick mildest steel would reflect almost everything going off this

https://www.kjmagnetics.com/thickness.calculator.asp
But w the hallbach wouldn’t even be a field in the steel and less iron loss and that easy coast hopefully
 
A hallbach array will just allow the back iron to be thinner. It won't reduce the iron losses in the stator.
 
Hummina Shadeeba said:
By reducing eddies and hysties in the rotor then will be an easier coast no?

Yes.

By looking at the shape of the losses curve, you can see which type of loss is eating your coasting energy. If it's a squared relationship with motor speed, it's eddy related drag, if it's linearly increasing with speed it's iron hysteresis loss.

Thin laminations help, but it's greatly diminishing returns below 0.2mm. Exotic 3% Silicon laminations help a ton for higher electrical frequency motors, but costs $$$. Cobalt/iron lams can let a stator be lighter and smaller and lower losses and make up for being smaller with higher core iron fields. Costs a lot in materials and mfg to make cobalt-iron laminations though.

Ironless motors may have very low core losses, but also very low torque density. Soft magnetic (ferrite) core motors exist and can be electrically efficient. They suffer from low torque density due to iron laminations offering greatly higher field density.
 
How much loss is in the rotor back iron vs the stator? looking at this first animation I don’t see the field changing in the backiron at all. Strong but not changing. Is this sim with low amps?


https://m.youtube.com/watch?v=d7wTI2V3gb0

But looking at this it’s way out past the back iron
https://m.youtube.com/watch?v=WsyumkL1Tnw


I always thought eddy current n hysteresis were just rpm dependent barring saturation but here is showing potential switching of field in different places (These are just random motor shots off the web and I have no idea about the motors).
 
Magnet back iron and is a small fraction of the total iron core losses (eddy + hysteresis).

Stator iron has a major change in flux density everytime a magnet passes by if you're powering the coil or not powering the coil.
 
The simplest thing would be to convince a manufacturer to do 0.27mm lams on a DD hub.

MAC does have these in their 1000W geared hub, so these kind of laminations are certainly floating around in the manufacturer-sphere.

MAC said that this only added $20 to the cost of their motors. Not expensive for all the benefits less iron losses provide.
 
Thanks for all ur thoughts and info it’s been educational. Given the iron losses in the rotor are said to be relatively low I’m just going to try some wider magnets instead of the hallbach as so far bigger has shown better.


Im already using .2mm lam stator.
How do mellow boards get a totally free coast? Maybe it’s programming as the stator and magnets are pretty ordinary other than the stator is separate for better winding and The teeth end up with no slots. that should produce no cogging. But my motors don’t coast nearly as long. Maybe inherent in a bigger motor w bigger bearings. Was surprised to see such narrow magnets inside.
https://youtu.be/GcLAvYbbnzM



https://www.mellowboards.com/en/In-Wheel-Motors/
 
Emetor’s motoranalytic program is down with a bug and no knowing when will be back up. I sent him the bug report, he’d asked for it after my writing saying it was down. Hard to push someone to fix their free program. Anyone have a lead on how to get a good program for only maybe two weeks? Motorsolver i think I have like two hours left.

21n22p motor has a huge amount of cogging steps at 462, amounting to low cogging torque. I think not the only indicator of cogging torque but a big one.
http://www.bavaria-direct.co.za/scheme/calculator/

But it’s got 21 teeth and asymmetric. Unbalanced magnetic force/pull. In my inexperienced looking it’s a trade off for either higher winding factor with it wound someway (I think I got to .85 wf with some configuration at best) or the imbalanced torque produced with all teeth of a phase wound together.

I’m going to get one or two 21 tooth stators cut to test once I can get designing it on something

The risks of the imbalance seem to relate largely to the motor’s construction and it could cause noise or cause bearings to die sooner. Seems worth a shot for such a potential ability to coast

With the no load losses being maybe 1/15 of the losses generally, even at highest speed I imagine the cogging torque is even a greater loss in speed and than iron losses no?. ..what kind of loss is cogging torque?

Eddy currents produce both a mechanical resistance and also a heat as it’s loss right, how much of each? Similarly w hysteresis?
 
not out of the box in the least but rather ancient.

first i ever heard of pulling the stator was the late 90's solar challenge that several solar car in-wheel motors sometimes incorporated.
https://endless-sphere.com/forums/viewtopic.php?f=30&t=16345
[youtube]6_41btVawMc[/youtube]

billvon said:
IIRC that's exactly what the old Tidalforce motors did - some odd number of salient poles on the stator. (14? I think there were 7 individually commutated phases.) These did not quite line up with the 16 magnets. There was almost no cogging.

from the engineer who built & designed it.
FalcoeMotors said:
When you run off battery, You could not pedal the Tidalforce or E+. It was a killer on your knees.
 
A difficult method to implement in that vid. Field weakening like that is more suited for going higher speeds than reducing cogging torque at all speeds though.

Got motoranalytic to work so will design a 21n22p and get some made n tested and report back.

Strangely I never hear of cogging torque as a loss and just copper or iron losses. It contributes to torque ripple too. I bet overcoming cogging torque on a lot of motors while riding through stop n go stuff is a bigger loss than iron losses in a lot of motors. No? And what exactly would be the loss due to cogging torque?


If I’m going 30mph let off the throttle I’ll drop back down to zero in way less time than would happen without a motor. I bet the loss from even this higher speed is mostly comprised of cogging torque
According to wiki:
“at high speed the motor moment of inertia filters out the effect of cogging torque.”

Wouldn’t higher speed reduce the ability to notice the ratcheting/pulsing of the torque but it would still be there slowing u just the same? Can think of no reason the attraction of a magnet to iron would be actually reduced by speed.
 
Back
Top