what is cogging torque?

Joined
Apr 25, 2014
Messages
2,768
Location
Sausalito
its a loss of kinetic energy when powered and even more so when unpowered. Assuming this lost momentum coasting turns to heat what is the process by which it happens? it doesn't seem to be eddies or hysteresis as cogging is a loss at the slowest speeds while iron losses are high speed losses. an example of a motor designed for less cogging results in slightest iron loss increase here:
http://kpubs.org/article/articleMain.kpubs?articleANo=E1EEFQ_2016_v11n4_878

takes forever to open
 
From the Abstract & Introduction in your link:

Cogging torque has a negative impact on the operation of permanent magnet machines by increasing torque ripple, speed ripple, acoustic noise and vibration.

Synchronous motors...are negatively impacted by cogging torque [3 , 4] . Cogging torque reduces the torque quality by producing undesired vibration and mechanical noise, which seriously impact the machine performance especially in low speed and direct drive applications [5] . Large cogging torque has also caused difficulties during motor start-up [6] . Cogging torque is caused by the interaction between the magnets mounted on the rotor and the stator anisotropy due to stator slotting. Cogging torque produces zero net work, but it acts as a disturbance superimposed on the electromagnetic torque generated during machine operation.

That article does not appear to support the theory that cogging torque is any kind of loss except a tiny amount of energy converted to vibration (acoustic or otherwise).
 
Punx0r said:
From the Abstract & Introduction in your link:

Cogging torque has a negative impact on the operation of permanent magnet machines by increasing torque ripple, speed ripple, acoustic noise and vibration.

Synchronous motors...are negatively impacted by cogging torque [3 , 4] . Cogging torque reduces the torque quality by producing undesired vibration and mechanical noise, which seriously impact the machine performance especially in low speed and direct drive applications [5] . Large cogging torque has also caused difficulties during motor start-up [6] . Cogging torque is caused by the interaction between the magnets mounted on the rotor and the stator anisotropy due to stator slotting. Cogging torque produces zero net work, but it acts as a disturbance superimposed on the electromagnetic torque generated during machine operation.

That article does not appear to support the theory that cogging torque is any kind of loss except a tiny amount of energy converted to vibration (acoustic or otherwise).
It can be far from tiny. The idea that a 200 lbs rider will be slowed from coasting to a stop in 1/3 the distance of a regular wheel by vibrations isn’t satisfying. With that much energy being lost to vibrations I think coasting down a mountain the motor needs to be rattling apart or something. what exactly would be vibrating and I assume getting hot? The only thing that can move in relation to each other and isn’t solidly fixed are the bearings, rotor, and hub and the direction of the cogging torque isn’t axial.

For simplicity let’s say u had a motor w one magnet and tooth. The pull force of the magnet on the tooth should be the same when coming or going and maybe should even be net zero no?

On a hub motor iron losses are produced at the far end of the spectrum and even then generally a small loss and here we have an obvious fat guy who can’t coast 1/3 the distance.

I think it would be easy to rule out hysteresis by using a motor with all magnets having one polarity facing in.
 
Hummina Shadeeba said:
The idea that a 200 lbs rider will be slowed from coasting to a stop in 1/3 the distance of a regular wheel by vibrations isn’t satisfying.

I'm not agreeing that the 1/3 is correct, since on my ebikes it isn't, but you are completely leaving out the primary iron losses of hysteresis and eddy currents which increase in a generally linear manner with rpm. Those losses are what are responsible for the lions shares of the extra slowing. Plus in your example of coasting down a mountain, many controllers whether or not they have regen will end up sending some energy back to the battery once the voltage of the bemf gets high enough (above no load rpm).
 
John in CR said:
Hummina Shadeeba said:
The idea that a 200 lbs rider will be slowed from coasting to a stop in 1/3 the distance of a regular wheel by vibrations isn’t satisfying.

I'm not agreeing that the 1/3 is correct, since on my ebikes it isn't, but you are completely leaving out the primary iron losses of hysteresis and eddy currents which increase in a generally linear manner with rpm. Those losses are what are responsible for the lions shares of the extra slowing. Plus in your example of coasting down a mountain, many controllers whether or not they have regen will end up sending some energy back to the battery once the voltage of the bemf gets high enough (above no load rpm).

For sure a bad cogging motor could produce that much resistance

I’m not leaving out iron losses and have been writing them of because as u say they are rpm dependent and cogging happens at almost zero erpm. If the amount of current necessary to overcome cogging at 60rpm is let’s say .25amp that current would be rising to a huge amount at 10,000 rpm if we’re iron losses
But as I wrote a motor model could be made with all magnets with the same pole facing in and that should produce minimal hysteresis.


(Never heard that about sending back to battery above no load speed when coasting.. wouldn’t the back emf be higher than the battery voltage at that point and no current flow?)
 
I don't understand why you're so interested in cogging. It's effect is only significant when you're trying to pedal through the cogging drag of a large DD hubmotor when you run out of juice. It feels about like a tire low on air, but with regard to only the cogging portion of the drag as you increase rpm it fades away and the tire inflates. It's quite an insignificant amount of power that doesn't increase with rpm, so only noticeable at quite low rpm where a little power is torque drag that can be felt.
 
The cogging torque doesn’t fade away though it’s just overlaid with the motor torque and not noticeable. Regardless of how much energy is lost to cogging when unpowered or powered I want to understand what’s going on. Where’s my easy roll go!? What’s heating up or possibly maybe even not heating, as lately been imagining the magnetic field like a gravity and gravitational fields don’t produce heat
 
The cogging itself does NOT create any losses. Hysteresis, eddy currents, and bearing friction do. There can be eddy currents in the magnets themselves, but this is generally very small compared to the iron.

If you use a magnetic coupling on a motor and load it until it breaks loose, the motor will spin like there is no load. There may be some eddy current losses from surrounding metal and the magnets, but there is no mechanism for loss just from the magnetic cogging.
 
Aha.
So where’s the kinetic energy going then? I dont see why it’s not net zero for starters.
Only analogy I can think of is like gravity but then there’s this craziness which throws the analogy out
https://m.youtube.com/watch?v=_GJujClGYJQ
 
Would i be correct to imagine it has a net zero total -100 to +100 and centre point being deadmiddle of the tooth so the initial energy taken to overcome the motor is gained back after mid movement ?
 
Haha ud think. for sure the one thing we can say is its not net zero regardless of the magnetic material, shape, distance or whatever. unless its a slotless motor and its not what we're talking about and it has no cogging. eddies and hysteresis the slotless will have and more of it I imagine since there's more material on the outside diameter of slotless motor, no slots.

There is a hum.
It would be nice to isolate and assure it isn’t coming from the bearings.
For sure some loss in the sound waves but wouldn’t that actually be super tiny and more so sound an indication of friction or pressure? How fast does cogging happen? Maybe the noise is from air pressure from very fast movement for milliseconds. Don’t think so.
 
Cogging losses don't net to zero, because it causes vibrations, some of which are audible. Tire friction, bearing losses, eddy currents, and hysteresis all increase with rpm. Since cogging losses don't increase above a certain point of rpm, my analogy of the tire inflating somewhat as rpm increases is a valid one.
 
Ah rite so if I had a 93% efficient motor cogging would be part of that remaining 7%, when I done my college it was mainly around induction motors so never came across any talk around it.

It can't be much I would have thought a good induction motor can be 90 plus so I would have guessed that cogging losses has to be less than 1% of the overall power consumption and the rest the common ones as John stated.

Took me a min to get my head round the tyre inflating analogy but I see now, flat tyre from start max drag as rotating starts the invisible tyre inflates allowing less rolling resistance giving less energy to overcome inertia.
 
Ianhill said:
Ah rite so if I had a 93% efficient motor cogging would be part of that remaining 7%, when I done my college it was mainly around induction motors so never came across any talk around it.

It can't be much I would have thought a good induction motor can be 90 plus so I would have guessed that cogging losses has to be less than 1% of the overall power consumption and the rest the common ones as John stated.

Took me a min to get my head round the tyre inflating analogy but I see now, flat tyre from start max drag as rotating starts the invisible tyre inflates allowing less rolling resistance giving less energy to overcome inertia.

Yes at normal operating speeds it's a tiny % of losses. At low speed though it is a higher percentage of losses...maybe even the majority of losses at very low rpm. It's not just the first cog point that you must get past if trying to pedal through the drag of big DD motor not being powered. At very low rpm the iron core losses along with the bearing and tire friction losses, all of which increase with rpm, are all still too low to be very significant, yet we can clearly feel the extra drag compared to a free wheeling rig with the same weight and tires. That's cogging losses.
 
Just for laughs, I will point out that switched reluctance motors have no cogging when they are unpowered. I believe the front motor on a Tesla Model-3 that has the AWD option is a switched reluctance.
 
Cogging just a loss to vibration? Weird hysteresis produced at super slow speeds?
What is cogging? I’ve got a couple resources showing the magnet as potential energy for things in its path so maybe it’s a resistance with no loss:

https://physics.stackexchange.com/questions/69400/does-a-magnet-contain-and-potentially-produce-energy/69404#69404

I go slower from cogging why?
 
As soon as you start thinking of magnets as an energy source you're in the compay of the perpetual motion crackpots...

Hummina Shadeeba said:
I go slower from cogging why?

How have you managed to eliminate all core losses in order to determine this?
 
So cogging torque takes initial energy to overcome then the resulting push we get back from passing the magnets midway point would be less than that put in due to losses.

But when we drive a motor there's power going through it so the result of that is it's hard to see this cogging effect happening ?
 
Maybe it’s due to core losses but so far in papers I’ve read that doesn’t seem the case and why would there be core losses at 0-10 erpm anyway as they would have to be some weird very powerful low speed losses. In the paper I posted here core losses as it maps them don’t correlate with cogging in fact it shows greater core losses w less cogging

It’s not net zero or hub motors would roll as far as regular wheels.

I’ve been reading how a magnet can be seen as an endless source of potential energy and opens the possibility there are no losses involved w cogging.

Wouldn’t be too hard to do an experiment to see if the temp increases from just cogging


How about a motor with all North polarity facing up, will it have cogging? Iron losses?
 
Hummina Shadeeba said:
Maybe it’s due to core losses but so far in papers I’ve read that doesn’t seem the case and why would there be core losses at 0-10 erpm anyway as they would have to be some weird very powerful low speed losses.

Assume a little four pole motor and that's 0-2.5 rpm, which is 0-0.017mph for a 60mm dia. skateboard wheel. Which prompts the question: what problem are you trying to solve exactly?

Hummina Shadeeba said:
In the paper I posted here core losses as it maps them don’t correlate with cogging in fact it shows greater core losses w less cogging

I skimmed the main section of the paper before and can't get the link to open now, but I believe they set out to produce a motor with less cogging and then characterised it to obtain the normal parameters. It's intuitive that if you change the design of a motor to optimise one characteristic (i.e. reduced cogging) then this may negatively impact other characteristics (i.e. iron losses).

Hummina Shadeeba said:
It’s not net zero or hub motors would roll as far as regular wheels.

But your statement directly contradicts the paper your are otherwise citing to support your argument.

I do not believe you can, with the information you have, separate out all the other losses (at low speeds) in a motor to conclude only cogging torque can be in play.

Hummina Shadeeba said:
I’ve been reading how a magnet can be seen as an endless source of potential energy and opens the possibility there are no losses involved w cogging.

With respect, you need to stop reading that stuff as it's nonsense. Magnets aren't magic and they aren't a source of energy. Any further reasoning or conclusions built on this base are inherently flawed...

Hummina Shadeeba said:
Wouldn’t be too hard to do an experiment to see if the temp increases from just cogging


How about a motor with all North polarity facing up, will it have cogging? Iron losses?

That would be an interesting experiment to try and tease out cogging torque. All magnets facing the right one is an interesting idea, but will you not still get discontinuities in the magnetic field where individual magnets meet? You might need a ring/tube magnet...
 
Punx0r said:
Hummina Shadeeba said:
Maybe it’s due to core losses but so far in papers I’ve read that doesn’t seem the case and why would there be core losses at 0-10 erpm anyway as they would have to be some weird very powerful low speed losses.

Assume a little four pole motor and that's 0-2.5 rpm, which is 0-0.017mph for a 60mm dia. skateboard wheel. Which prompts the question: what problem are you trying to solve exactly?
ok that speed is so slow it may be harder to visualize but if you get a high cogging hub motor and coast down the street at whatever slow speed it will not roll near as far as a regular wheel. what specifically is the lost moment going towards? eddies are exponentially produced and hysteresis at least being linear...maybe hysteresis but would like to see it mapped
Hummina Shadeeba said:
In the paper I posted here core losses as it maps them don’t correlate with cogging in fact it shows greater core losses w less cogging

I skimmed the main section of the paper before and can't get the link to open now, but I believe they set out to produce a motor with less cogging and then characterised it to obtain the normal parameters. It's intuitive that if you change the design of a motor to optimise one characteristic (i.e. reduced cogging) then this may negatively impact other characteristics (i.e. iron losses).
the overall point is the paper shows two motors with only one skewed being the difference between them...the skewed has very little cogging but more iron losses. if iron losses were related to cogging the motor with more cogging should have more iron losses no?

Hummina Shadeeba said:
It’s not net zero or hub motors would roll as far as regular wheels.

But your statement directly contradicts the paper your are otherwise citing to support your argument.
don't know what you mean but im saying, as the paper shows, a greater cogging motor doesn't have more iron losses making it seem cogging is not iron losses.
I do not believe you can, with the information you have, separate out all the other losses (at low speeds) in a motor to conclude only cogging torque can be in play. no but when all other parameters are the same except one skewed, and then the cogging torque drops substantially that gives a good indication that iron losses aren't causing cogging no?

Hummina Shadeeba said:
I’ve been reading how a magnet can be seen as an endless source of potential energy and opens the possibility there are no losses involved w cogging.

With respect, you need to stop reading that stuff as it's nonsense. Magnets aren't magic and they aren't a source of energy. Any further reasoning or conclusions built on this base are inherently flawed...
not talking about magic. theres a lot of info out there showing how a permanent magnet can be seen as a constant potential energy source in relation to anything ferromagnetic. kinda like a gravity field that's always there and all metal things are already raised and have a potential energy. nothing magical. it could also be used as a permanent damper where a magnetic material is slowed, losing it's kinetic energy yet not due to iron losses only due to the pull of the magnetic field.

Hummina Shadeeba said:
Wouldn’t be too hard to do an experiment to see if the temp increases from just cogging


How about a motor with all North polarity facing up, will it have cogging? Iron losses?

That would be an interesting experiment to try and tease out cogging torque. All magnets facing the right one is an interesting idea, but will you not still get discontinuities in the magnetic field where individual magnets meet? You might need a ring/tube magnet...

maybe as you say. i'll see if i can model it.
 
https://en.m.wikipedia.org/wiki/Cogging_torque

The wiki makes a bit of sence of it, from what I read it's undesirable but when the motor is powered it's this path that the field lines will strengthen so to remove cogging would reduce the available torque at start up.

Axial coreless machines need smaller air gaps it's to counter the lost torque from field lines not being focused by the steel into a tighter channel and getting to a higher tesla at the tooth face, the down side to this is the magnetic linkage that occurs between steel rotor/stator and magnetic stator/rotor.

It's hard to isolate the cogging to measure just that alone becuase soon as you spin the motor friction will still be part of the answer and soon as we add power to the motor we get more losses making it even harder to isolate, I guess you would need to take the efficiency loss of the motor of around 7% or so and break that down work out exactly how much becomes heat and from there noise and friction and so on till you have that cogging effort left.
 
As for magnet machines giving contious power that is smoke and mirrors not one to date has come close to running forever let alone adding power to it's own system, just think this where would it stop it would have to spin up to a point it's own inertia at that rpm is to much to overcome meaning if it could be done it would run away to a crazy speed an explode or like we see with motors it slows and stops.

I can't really give you the definitive answer on this one id love someone to chirp in and set the record straight on it but in the mean time I'll do a little research myself when I've got the time.
 
no one talking about free energy from magnets but they can be seen as an eternal potential energy sources. they cant do work alone though. similarly to in a motor when powered they greatly enhance the torque but they need to be oriented correctly to achieve that gain. the extra torque of the motor's pm magnets don't produce any losses in that increased torque production. the magnets do produce other losses but not related to that specific moment of torque produced when they are oriented in a way to do that.

a coreless motor isn't going to cog at all.

think easier if just talking about the cogging of an unpowered motor. you could drive it externally to get a measure of the torque needed to make it spin and get a number there....then skew the stator and do the same test... then just have to figure how to get a graphing of the iron losses. maybe just keep spinning the motor at higher speed and graph resistance? and a test of just the bearings with no magnets inside.

I don't know how they come up with their "core losses" in the paper though. assuming its accurate...where's the cogging coming from if a greatly decreased cogging motor has more iron losses? and then again it could be a powered motor they're measuring iron losses. im looking for just the unpowered losses.

actually looking at the paper above again im confusing things and this is about skewed magnets not stator but similar and same end result

!! looking at the paper in the core losses section it describes variations in eddy loses and its integrated into the equation for core losses!!
"Ke is the excess or anomalous eddy current coefficient due to magnetic domain,"
what is that?. that ive never heard of
"Typical conventional core loss models which are based on constant coefficients are not accurate. In order to accurately predict the core loss, variable loss coefficients should be incorporated in the model. "
ATTACH]

with the graph shifting in time I don't understand..maybe this graph is taking into account the anomalous eddies at the slowest ...time(speed)? why not speed here instead of time? or position or something and time doesn't make sense to me
 
Back
Top