Can saturation point be estimated by copper mass?

[madin88]

You left out the more pertinent information from Justin's post:

So there you have it folks! One thing that became clear to me in these tests though, was that even if there wasn't any saturation/demagnetization effects going on, just the I^2R consequence of pushing the motors this far means we are in a domain that is totally outside of any useful operating zone. The windings get hot, FAST, and that has huge effect on the motor performance. I had the current going through for just 10-15 seconds in order to get a peak torque reading, then would have to put the motor outside in the subZero weather and wait a good amount of time for it to cool off.

it is about when the IRON CORE starts to saturate and not when the windings start to melt.
edit: just because this 9C did overheat quick at that point, it doesn't obligatory mean every other motor will overheat also that quick when the core is saturated. the heat in the windings depends on the phase to phase resistance. a motor with only 28mm wide stator like that 9C also has quite high end turn losses compared to the losses of the copper inside the teeth. a motor with wider stator is better regarding total copper losses.

i would like to know if a high pole count stator can produce more torque as one with a lower pole count at same size.
copper fill should be better with a lower pole stator design as we see on the QS V3 or the Hubmonster..

 

[Alan B]

Justin measured saturation for the 9C motors years ago in a detailed thread here.

http://www.endless-sphere.com/forums/viewtopic.php?f=2&t=14494

Just map the torque vs current and look for the change in slope. Once the steel is saturated the field increases at the unity rate, same as an air core, there's just no benefit from the steel's magnetic properties.

An efficient motor should never be operated into saturation, but many folks exceed that point at low speed and high throttle, generating a little extra torque at the expense of very poor efficiency and significant heating. The 9C (of a few years ago) has saturation starting at about 450 amp-turns. On the standard 7x9 winding this is only 50 amps of phase (motor) current. Above this point Justin measured a 40% reduction in torque per amp increase.

So the motor phase current max limit for the 9C 7x9 winding should be set to 50 amps or less. The battery current can be whatever you want, that controls the maximum power, but there is no point in setting it above about 50 amps. At low speed the battery current will be quite low when the motor phase current reaches 50 amps.

 

[Miles]

i would like to know if a high pole count stator can produce more torque as one with a lower pole count at same size.
copper fill should be better with a lower pole stator design as we see on the QS V3 or the Hubmonster..

It's very difficult to separate the affect of pole count from all of the concomitant factors.... As a higher pole count will require less core material, one would expect this to allow a higher copper fill.........

 

[Punx0r]

Guys, excellent effort on pulling together existing saturation info on the 9C clone I've wanted to know about it for a while but just didn't know the testing had already been done. It turns out it's disappointingly little current... I'm baffled how Liveforphysics got (if I recall correctly) 10HP at 10kW input out of his flaming-dicks 9C on the dyno? I'll have to try and find the posts and see if the speed/torque is visible.

 

[madin88]

if the point with the linear relation has been exceeded, will the curve than flat out at some point so that more amps won't lead to any higher torque?

It turns out it's disappointingly little current...

unfortunately it seems so, thats why it would be quite useful to know

 

[Samd]

This is the point where the line starts to be non-linear, the point where good efficiency ends. it seems like saturation start at round about double that figure. But not double that torque figure, of course (as current and torque are no more linear related here). Saturation torque seems to look like maybe somewhere 20% above that 65Nm from here, say 80 Nm. That is ~2.3 Nm per mm of lamination
this would be my very rough rule of thumb

absolute maximum torque
65 Nm__28mm MXUS V2
80 Nm__35mm Leafbike 1500W
104 Nm_45mm MXUS 3000W V2
115 Nm_50mm QS 205/50H V3

max reasonable phase-amp-turns
280 At__28mm MXUS V2
350 At__35mm Leafbike 1500W
450 At_45mm MXUS 3000W V2
500 At_50mm QS 205/50H V3

I'm liking this. We do seem to bandy around watts a lot, like guys comparing size of something else. But this brings it home for me.
Not a big step from the Mxus to the Quanshun, but a big step down to the leaf from the Mxus.

John in CR, it would be good to compare your motors if the data exists.

 

[crossbreak]

John's data exists in the other thread:

Hubmonster:
Gap dia: 181mm, Stack: 58mm

Midmonster:
Gap dia: 164mm, Stack: 54mm

Material of these seems to be different, otherwise the low core loss cannot be explained. The other candidates are quite close together on this

What is the gap dia of the leaf, mxus and QS 205?

 

[Punx0r]

Madin, check out the first two graphs here: https://endless-sphere.com/forums/viewtopic.php?f=2&t=14494&start=25#p218312

You see the "knee" in the current-torque curve where it becomes non-linear. However, torque does keep increasing by a large amount, just not as much as before the knee. Alan B made a good point: Even when the steel is completely saturated, the motor will continue operate like an air-core motor and create more torque with increased current, but much less.

Having studied those graphs more closely, the 9C doesn't seem quite as weak as I feared. The 50A before the onset of saturation was on a 9-turn motor and made 70Nm. That's 450 amp-turns. The motor continued to 80A and 100Nm (720 amp-turns). 100Nm at ~700rpm (~60mph on a 26" wheel) would make ~10HP

 

[crossbreak]

just because this 9C did overheat quick at that point, it doesn't obligatory mean every other motor will overheat also that quick when the core is saturated.

yes, it will, since thermal design of all common motors suggests using them not near saturation. At least this is what all of em have in common

 

[John in CR]

John in CR, it would be good to compare your motors if the data exists.

All the data needed is in Miles spreadsheet. I wouldn't sell a motor to someone planning to run it near saturation, because it' s a recipe for failure. I also won't sell to someone wanting to run high power with the motor in too large a large wheel. Unfortunately around here people insist on gearing their hubmotors too steeply and treat current settings as a dick measuring contest.

Instead of listening to my recommendations, Hillsofvalp ran his HubMonster in the largest wheel he could fit with current settings as high as he dared (400A from the battery and 500A+ phase limit), and his dyno runs showed 69% efficiency at peak power, so he was probably flirting with the beginning of saturation. That's no way to run a motor. Instead shoot for good efficiency over a wide band, so you don't have to carry more battery just because you skimped on motor size for the power you want, and then heat issues become easy to solve.

 

[John in CR]

Material of these seems to be different, otherwise the low core loss cannot be explained. The other candidates are quite close together on this

While I'm sure the Japanese laminating steel used is higher quality than in lesser motors, the low core losses are primarily the result of the much lower pole count (20 magnets vs the all too common 48).

 

[John in CR]

Having studied those graphs more closely, the 9C doesn't seem quite as weak as I feared. The 50A before the onset of saturation was on a 9-turn motor and made 70Nm. That's 450 amp-turns. The motor continued to 80A and 100Nm (720 amp-turns). 100Nm at ~700rpm (~60mph on a 26" wheel) would make ~10HP

You're mixing and matching motors. No common size 9C can make 10hp in a 26" wheel. Luke and Methods dyno'd a Cromotor to 10hp on 24s. Luke's flaming motor was an H35xx rewound to high speed running in a 20" wheel. When you're ready to talk about real motors and forget the saturation nonsense, then we can talk about HubMonster's 28hp peak output on the dyno with a LiFePo4 battery sagging to 74V.

 

[spinningmagnets]

What is the gap dia of the leaf, mxus and QS 205?

I believe they are all 205mm stator diameter...even the MXUS 28mm 1000W

 

[crossbreak]

Material of these seems to be different, otherwise the low core loss cannot be explained. The other candidates are quite close together on this

While I'm sure the Japanese laminating steel used is higher quality than in lesser motors, the low core losses are primarily the result of the much lower pole count (20 magnets vs the all too common 48).

wasn't aware of that. now figures fit quite well. Half the pole pairs means half the polarity change for the iron. So half of hysteresis torque too. nice job they stuffed so much copper into it. As much per Weight as a HS4080, but into lesser teeth

 

[John in CR]

... nice job they stuffed so much copper into it. As much per Weight as a HS4080, but into lesser teeth

With the 2 motors being so drastically different in air gap radius, design, structure, efficiency, etc, how did you come up with that comparison? One is made as cheap ($0.02 halls, mystery metal covers, etc) and light as possible (without increasing cost), and the other overbuilt to the extent that kilos can be removed to get it down to it's basic form as a hubmotor, so copper/weight may very well be comparable. I don't understand how you are getting to the copper content to make the comparison.

 

[Punx0r]

You're mixing and matching motors. No common size 9C can make 10hp in a 26" wheel. Luke and Methods dyno'd a Cromotor to 10hp on 24s. Luke's flaming motor was an H35xx rewound to high speed running in a 20" wheel.

My mistake, it wasn't a 9C, but it did (just!) make 10 HP: https://endless-sphere.com/forums/viewtopic.php?f=2&t=27020&p=390221

Probably a very clear demonstration of saturation: the motor made only a fraction more power at 40kW input compared to 10kW

 

[jmz]

There's a few confused sounding posts here.

Core hysteresis loss (from exercising the stator steels BH curve, dependent on the magnitude of phase current MMF and external magnet forcing, multiplied by the frequency of the BH curve-cycling, independent of lamination width) != core eddy current loss (induced currents and associated I^2R losses in the stator steel due to the *rate of change* of perpedicular magnetic fields (and hence mechanical frequency), strongly dependent on lamination width) !!!= copper loss (shouldn't have to explain this one).

 

[crossbreak]

people who read threads like this do know the difference between hysteresis, eddy and copper loss, otherwise they wouldn't read it.

With the 2 motors being so drastically different in air gap radius, design, structure, efficiency, etc, how did you come up with that comparison?

specific Km² is the about the same, same copper loss per weight. While they are totally different as you objected, there are different ways to archive high Km² figures. Lower pole count does not automatically mean higher copper loss (due less space for copper). But this is what most people think. Or how do you explain why most hubbies have these 48 poles? Sry this is OT. We should continue the discussion there https://endless-sphere.com/forums/viewtopic.php?f=30&t=65757

 

[madin88]

yes, it will, since thermal design of all common motors suggests using them not near saturation. At least this is what all of em have in common

there is no doubt that the motor will overheat sooner or later, but it must not mean every motor will overheat AS QUICK as the 9C justin used for the test when the core starts to saturate.

about the different stator design:
lets say we compare motor A that has 16pole pairs with motor B that has 23pole pairs (like the QS V2 and QS V3).
the air gap diameter, stator width, kind of steel and magnet strength etc is identical.

will both motors saturate at exactly the same Nm (or ampturns), or will it be different?

 

[fechter]

about the different stator design:
lets say we compare motor A that has 16pole pairs with motor B that has 23pole pairs (like the QS V2 and QS V3).
the air gap diameter, stator width, kind of steel and magnet strength etc is identical.

will both motors saturate at exactly the same Nm (or ampturns), or will it be different?

I think it will be nearly the same.

 

[crossbreak]

i'm not so sure about that. on the one hand, stator tooth width may grow equally, on the other hand, the amount of flux in those teeth may (maybe) or not :? confusing
would be great to see some sim data/image here for comparison

 

[sendler2112]

Why not put this Saturation data in the motor comparison sheet ? it already has all the motor details in their. Maybe just add a column for material ?

Does the saturation point only have a relation to current or also voltage related?

Eg Crossbreak for the cyclone motor in that thread we were discussing the efficiency at 48v 100 amps vs 72v 68 amps is the same power, but maybe higher efficiency at 72v, is this related to the motor saturating and hence lower current is better ? Or is the saturation point of the motor above 100 amps and not relevant to this?

This is a good question that hasn't been addressed in coming up with a concept like listing phase amps/ turn for different motors. Core saturation point should be based on power, not simply amps. Right?

 

[spinningmagnets]

I'm sure volts have some effect, but amps seem to be the major culprit affecting motor overheat. In theory, 36V @ 40A = 1440W should be the same power as 72V X 20A = 1440W. Even adjusting the Kv so that both motors spin at roughly the same RPM....what would the "likely" heat differences be?

Back in the days of low-amp batteries, there was no choice. You had to volt-up/amp-down (and lower Kv if possible?) or suffer heat and abused battery packs. Now, theres a wider selection of motors and high-amp batteries, so...now we have a choice.

I don't think it's a horrible thing to go into saturation a little bit at the beginning of an acceleration run (with 98% of the cruise being in an efficient zone and cooling off), but...if we do so, it should be with our eyes open and also being informed of where the line is.

 

[Punx0r]

No, amp-turns only...

You can probably get away with an occasional brief excursion into saturation, but it's getting away with it rather than acceptable as the efficiency will be very bad and in reality the motor is undersized.

 

[sendler2112]

No, amp-turns only...

You can probably get away with an occasional brief excursion into saturation, but it's getting away with it rather than acceptable as the efficiency will be very bad and in reality the motor is undersized.

So field strength and saturation have nothing to do with voltage? 50A at 36V will have the same saturation as 50A at 72V?