AC motors, multipoles, torque

[coulomb]

Something tells me that three of the magnetic fields are reversed in the parallel Star Dahlander arrangement and this could cause the rotor to abruptly spin in reverse if these were switched while in motion...

I believe that this is intentional. The fields are sized so that one subtracts from but does not reverse the other.

The pole switching is even more dodgy; the rotor sees a horrible approximation to sine wave.

As has been mentioned elsewhere, I think that the star/delta switch is the only kind of coil subterfuge that we should be attempting for EVs.

[Richo]

The motor can be in motion.
ACIM's don't produce/consume current unless the VFD is actually switching.
So a properly designed VFD could pole switch or swtich between star/delta no problem while at speed.
But really it's not worth the effort or cost.
Just buy the proper size motor and controller.
It's cheaper and more reliable.

[woody]

I think series/parallel has promise - easy if you have 2 motors
Star / Delta is tantalizingly there for the taking though.

[weber]

Hi weber. Do you know how Ross overcame the problem of lubricating the rear oil seal (around the tail-shaft slip-yoke) without swamping the front motor bearing with oil? This is my current concern with using my tail-shaft housing. Sorry - a bit off topic here.

Not too off topic since one aim of many AC conversions is to eliminate the gearbox. But you can email me direct on d dot keenan at bigpond dot net dot au.

I have emailed him the question and will let you know. But I suspect he didn't solve it. I note, for others who may not be familiar, that the tail end of the gearbox of a front-engine rear-wheel-drive usually has a plain bearing that supports the front end of the propshaft (also called confusingly tailshaft or driveshaft in various countries) which is driven internally by a spline that allows for some misalignment. This plain bearing, and the spline, need lubrication, but you don't want your electric motor filling up with gear oil. 90 grade oil between rotor and stator could be a serious drag.

My own thoughts are to put gear oil into the tail casing, while on level ground, with the propshaft out, until oil just starts to run out the back, and rely on it moving about, as the car accelerates and goes up hills, to get it to the spline and plain bearing. And rely on the motors own oil seal to keep it out on the downhills and decelleration. Sounds a bit risky I admit.

If it did get into the motor, could it be flushed out with a solvent without disassembly?

[Johny]

I'm impressed that you knew what I was on about. I agree it's too risky. One solution is to place a blanking plate halfway along the tail shaft housing and put yet another oil seal there. The problem may be trying to assemble it all.

My alternative is to NOT use the tail shaft housing and instead use an external slip-yoke as per a4x4kiwi's Electrolux (even my wife laughed at Mal's nickname for the Hilux).

Since AC motors are less expensive with only foot mount (as against foot and face mount) this is also cheaper (at least for the motor). It does require I get the motor as rear-ward as possible to not have an excessive length drive shaft - which is why I seek a 132 frame motor so that it fits in the transmission tunnel.

Anyway - the tail-shaft housing still has appeal if problems can be solved. I just happen to have a spare gearbox...

[woody]

I was talking to an electrical engineer friend on the weekend, about Coulomb's and my AC EV project and he mentioned that he'd worked on a lathe that had used an off-the-shelf high frequency (400 Hz) induction motor driven from a stock standard Control Techniques VF drive. So I googled "high frequency induction motor" and amongst a lot of sites with sub kilowatt motors intended for use in aircraft I found this monster:

150 kW continuous at 400 Vac, 400 Hz and it only has a mass of 39 kg!
http://www.400hertz.net/Products/ME-400-200-416.htm

The bad news is it's a four pole, so that's 12,000 rpm and you'd need a gear reduction of around 2 to 2.5 for it to be useful in driving a diff.

If you ran it at 50 Hz you'd presumably get a continuous 18.75 kW (150*50/400), but I note that that would be at 50 Vac and 300 A. The ABB 400 V, 50 Hz, 18.5 kW motor has a mass of 92 kg -- more than twice as much.

You could presumably use the Azure Dynamics DMOC445 VF drive (220 Vac, 400? A) and only run it up to 220 Hz (220 V) and still get 82 kW _continuous_ out of it at 6000 rpm. A 39 kilogram motor! Can this be for real?

Of course I have no idea of the cost.

-- weber

This sounds similar to the Tesla's motor (on top gear last night) which revs to 13,000 rpm or so. Tesla use a one speed gearbox.

Does anyone know how much industrial gear reduction stuff costs and if it spins to 12,000 rpm?

[weber]

The 150kW is a peak rating, normal motors are nominal.

Hi Woody,

That would certainly explain the low weight, but what makes you think that's a peak rating, _apart_ from the low weight. But I shouldn't have written that it was a continuous rating either, since it doesn't say either way. It's just that when you look at their other offerings they are industrial, not EV and so I assumed the usual continuous rating of industrial motors.
http://www.400hertz.net/

I have contacted them and will let you know if I get any further info.

[weber]

Hi Johny,



To continue this discussion about using the gearbox tail casing, and similar issues, I've started a new thread:

Direct Drive: Mechanical issues

[Edit: Fixed yet another link broken by a dumb domain-name/directory-structure change by the forum administrators.]

[weber]

Here's what I emailed to <info@400hertz.net>:

Dear 400Hz,

There is some interest at present in your http://www.400hertz.net/Products/ME-400-200-416.htm on the Australian Electric Vehicle Association forum. viewtopic.php?p=9229&t=585#p9229

Can you tell us if the 150 kW is a continuous rating or a peak (short-time) rating?

Can you supply a data sheet for the motor, with e.g. continuous-and-peak torque-and-power versus rpm?

Regards,
-- Dave Keenan

Here's the reply I received:

Dave,

Unless you are a magician, you won't be able to carry enough batteries to worry about a steady state 150 KW motor load.

150KW can be a continuous rating if it has a large enough supplementary cooling fan.

The motor produces approximately 200 FT/LB or torque @ 5,000 RPM.

Further data will follow as we are using a new supplier for these motors and still do not have all the information.

PS. It takes only approximate 15 HP to push a full size automobile at a steady 60 MPH.

Best Regards,

Edward J. Dempsey

Translation of archaic units: 200 lb ft is about 270 Nm, 15 hp is about 11 kW and 60 mph is about 96 km/h.

[acmotor]

ACpropulsions AC150 is 50kg for 150kW peak.
http://www.acpropulsion.com/company/press-releases.php

It will be interesting to see if the 400Hz offering is the same price !
i.e. US$25k with controller if you can get them to sell you one.

I would say that 150kW would only just satisfy some amp heads !

weber,
Please keep us posted !!!!!

[antiscab]

yeh, i mean 150kw would be fine for a motorbike, but a lil underpowered for a car

Matt

[acmotor]

[weber]

Regarding the 400Hz motor: 270 Nm * 5000 rpm * 2 * pi /60 = 140 kW (peak)
So it sounds like field-weakening starts at around 5000 rpm (with 400 V), which might be about 180 Hz (4 pole) allowing for slip. If so, and assuming Tmax/Tn = 3, its performance sounds similar to a 400 V 15 kW (cont.) 60 Hz motor that's been rewound for a voltage that's a factor of 3 lower, but still run at 400 V. Have I got that right?

And it would be worthwhile for direct drive even though you might never go above 180 Hz (5000 rpm). So the 39 kg weight does seem very low. Bet the price is the killer.

Funny how it has no cooling fins and yet it is air cooled.

[Richo]

Sounds about right.
Also have to consider if you could mount that motor without it breaking those little screws.

Hey Matt will you be putting a racing harness on your electric bike

[Johny]

Funny how it has no cooling fins and yet it is air cooled.

They did say "150KW can be a continuous rating if it has a large enough supplementary cooling fan."

Maybe inferred it needs another size cooling fan if you want more than a fleawatt.

[woody]

Found a fact sheet on MEPS:

There are a number of motor types that are exempt from
MEPS including:
   submersible motors;
   integral motor-gear systems (non separable);
   variable or multi-speed speed motors;
   motors rated only for short duty cycles (IEC60034-2 duty rating S2)
   rewound motors or motors sold as second hand.

Edit: Quote format around the fact sheet quote

[Richo]

Dave,

Unless you are a magician, you won't be able to carry enough batteries to worry about a steady state 150 KW motor load.

150KW can be a continuous rating if it has a large enough supplementary cooling fan.

The motor produces approximately 200 FT/LB or torque @ 5,000 RPM.

Further data will follow as we are using a new supplier for these motors and still do not have all the information.

PS. It takes only approximate 15 HP to push a full size automobile at a steady 60 MPH.

Best Regards,

Edward J. Dempsey

Actuallty that seems a bit rude
1. It's not his problem how we are supplying 150kW. In all reality 150kW is quite realisitc LOW CAPACITY - HIGH PERFORMANCE CELLS.
2. His website says that this is for "automotive purposes".
3. It maybe 15HP to push it at 60MPH but you need more than 15HP to get to 60MPH in a realistic time - hence the motor he is selling and advertising as for use in "electric automobiles".

The 150kW would be the peak and approx 45kW nominal.
All the other motors he sells are advertising the peak.
They are used for saws/drilling and so aren't used 24/7.
270Nm I think sound like a bit of torque boosting

[Johny]

The 'Electric Beemer' topic appears to have stolen this threads content to some degree but I thought I should post this here.
Having lost a bit of faith in ABB's new 132 frame motors, I emailed Rob at Trojan (has 11kW, 132 frame, 4 pole, 67kg, 73NM Tmax/Tn=2.6).

This was my email:
----------------------------------------------------------------
Hello Robert
I'm the guy still doing an AC EV.
I wonder if I could ask you another question.
What current would you expect the 132 frame 4 pole 11kW (MD1C(MD1A)132M2-4)
to draw at or near breakdown torque? (Nominal current at 415 V is 20.3 Amps).

A primitive calculation just multiplying rated current by Tmax/Tn (which is 2.6 for this motor) gives me 52.8 Amps but I am told it will actually be higher.

The ambition is to have one rewound for 300 V Star / 173 V Delta with the idea of a rated speed of around 3400 RPM at 118 Hz. This would place the car at 90k/h at rated speed. The relationship between current as maximum torque effects this target rewind voltage due to controller limitations (currently 89Amps).

I am hoping that you still have some MD1C(MD1A)132M2-4 on the shelf.
Thanks
John

--------------------------------------------------------------------
Here is his response:
--------------------------------------------------------------------
Hi John,

Good to see you’re still persevering with EV.

The 11kw 4pole 132frame is 20.3FLA @ 415VAC 50Hz.

Break down torque is calculated by multiplying maximum torque/full load torque. Which is correct as you have stated 52.78A in your email: This is a safe calculation, I would not exceed it, and these tolerances are usually within 5%.

Yes I still have some 2 and 4 pole 11kW 132frame motors.

Your concern will be more complicated as higher RPM (Hz) will increase your slip once you run over 65-70Hz as it’s outside the design criteria of the motor, you may also over flux the motor so you may loose speed and torque.

Once the motor is rewound to 300V star and 173V Delta your full load currents at 50Hz would be 28.1A and 48.7A respectively. Your full load toque output will still be the same at 73Nm Regardless of connection.

Although you could induce higher voltages and work outside the normal design parameters of the motor; you will lose efficiency in some manner, possibly over heating the motor and losing torque and speed, I’ve seen motors doing less rpm at 100Hz than what they were doing at 80Hz.

Unfortunately I don’t have any data that’s really going to help your scenario. I can tell you that increasing the voltage and frequency at the same time will maintain the current and increase the rated power of the motor.

For example this motor is rated at 11kw on 415V/50Hz, but if we export the motor overseas and re-rate at 480 / 60Hz, you pull the same current but the motor is now about 13.2kW (the torque is still the same but you now achieve it at 20% higher speed) I do know from a recent application that with reduced alloy frames of motor you can expect a higher percentage slip over 100Hz compared to a cast iron 160 frame motor. Because of this there is no way I would guarantee under load your 3400rpm @118Hz

Once you go over 70Hz regardless of what voltage you pump in, you will really start to lose your torque.

If we offered a 2pole 11kW ran it to 60Hz your final output would be 3480rpm with full load torque of 36Nm. You could rewind this motor to be 208V / 40.6A

If we offered a 4 pole 11kW ran it to 117.5Hz your final output should be 3395rpm, but the motor will start to fluxing out after 70Hz especially under load, If you keep the current the same and increased the voltage which I believe is your intention the motor will run much hotter but you may be able to regain some of your lost torque-but because the motor would be fluxing out there no way of telling how much torque you would still have with out trying it ( this could be a very costly failure or a breakthrough for the EV market. You would need a full dynamometer capable of torque induction and motor voltage, frequency and current regulation-which would be your drive)

If you still have your heart set the 132 frame but you’re still lacking the torque would you consider putting 2 motors back to back? With the windings connected parallel. Another option would be putting a motor behind each wheel?

I have done recent calculations on a larger car (ford falcon wagon or similar) but using a 2 pole motor 30kW, as the unloaded current is nearly the same as the 18.5kW that’s actually required for normal use. The upshot is the regenerative breaking has a higher charge capacity-and there is an extra 60% torque for overtaking and going up hills if required.

Hope this helps and not confuses.

Cheers

Rob Rathbone

---------------------------------------------------------------------
It would be REALLY helpful to have some actual data from one of the existing Industrial AC EVs on the road.
Not lots of accel and de-accel but one straight and level run under maximum throttle that shows us time, speed, voltage and current.
We could then compare this to our speadsheet models and see if the predicted performance is accurate.

I have actually revised my target rewind voltage to 360/208 (Star/Delta) to optimise acceleration to 80k/h. It's a real balancing act if there are any controller or battery limitations at all.

Comments on Rob's email?

[antiscab]

interesting that he says the motor will "flux out" at frequencies above 70Hz, regardless of the V/Hz ratio we can maintain.
anyone have access to a dyno?
i doubt it would have to have the motor controller built in, we could just use a standard VFD.

Matt

[Richo]

From what I see the 22kW 2-pole 132-frame has been removed from ABB catalouge.
Also the 11, 15 and 18.5kW have been recategorised as High output.

Im interested in the local rewinders testing facilities to see if they can test the motor before it goes in the car.

May have to go for a visit.

[Johny]

The Lenze motors that are sold as 50hz/87Hz motors show identical torque when used at 87Hz. The 87Hz is really 240V windings run off a 400 volt VFD which is exactly what we are talking about. So are the Lenze motors special in this regard or is this just uncharted territory for Trojan?

[weber]

I don't believe anything special happens around 70 Hz if you're still giving the motor its rated V/f. ABB show graphs of continuous torque up to 100 Hz, although those are for _constant_ voltage above 50 Hz.

All that I'd expect to happen as you increase both V and f in proportion is that certain kinds of power losses will gradually increase, and the probability of a catastrophic insulation breakdown will rapidly increase (unless there are special precautions on the part of the rewinder, and possibly high frequency series inductors on the VFD outputs). The additional loss would be mainly hysteresis loss which is proportional to frequency, and eventually eddy current and skin effect losses will become significant too.

Let's say you are considering a choice between overvolting a 2-pole motor by a factor k > 1, versus overvolting a 4-pole motor by a factor of 2k (which would give approx the same rpm at the field-weakening point). Then if you needed the light weight, you might risk the large overvoltage on the 4-pole, but the 2-pole should be more efficient [Edit: I originally had "2-pole" and "4-pole" reversed in this sentence]. And the 4 pole will not be 1/2 the weight, as you might think, unless you happen to drop to the most powerful motor in the next lower frame size or some such.

Of course these are gross generalisations and your situation may be different

In either case (2 or 4 pole) if you're keeping your gearbox you want field-weakening to occur near (and preferably below), the maximum rpm of the motor (its "centrifugal limit"). So if that was 4500 rpm max then you wouldn't want to overvolt a 2 pole motor by much more than a factor of 1.5 or a 4 pole by much more than a factor of 3.

A common overvoltage factor is 1.73 ~= sqrt(3) because that's what you get if you run a star-rated motor in delta at the same voltage. This is not too unreasonable for a 2 pole motor as the slip at max torque will result in the actual motor speed being under 4500 rpm even though synchronous speed is 5200 rpm.

If you're ditching the gearbox then you'll need a more powerful motor and you will want field weakening to occur at about half the motors maximum speed. So you wouldn't use a 2 pole at all. A 4-pole with a 1.73 overvoltage would make more sense.

Have I missed anything?

[bga]

This is what I'd expect: voltage is propoptional to speed.
i.e. 87Hz / 50Hz =~ 415V / 240V

I'm going to assume a 4 pole motor here, but the same holds true for 2 poles.

As far is the rotor is concerned, it doesn't care what absolute speed it's doing, up to things breaking or frying.
Whats important to it is the difference between its speed and the speed of rotation of the field (the slip speed).

Imagine that you are sitting on the rotor and rotating with it. Think of that the magnetic field is doing around and through the rotor.
You'll notice that the field is only rotating very slowly, somewhere around 30 to 50 RPM, being the difference between the synchronous speed (1500 RPM) and the rated motor speed, say 1470 RPM.
This also means that if you were to put a DC field on the stator, as in DC injection (plug) braking, the motor would develop full torque at about 30 RPM. (A low voltage battery and a spanner should demonstrate this)

What's going to happen then we speed the field up to, say 100 Hz? The synchronous speed becomes 3000RPM, the same as the 2 pole version at 50 Hz. Because the motor is a big inductor, additional terminal voltage is needed to drive current into the motor and achieve full field strength, in this case it's about double the voltage needed for 50 Hz.

Assuming that we have a high enough voltage power supply:
a) The rated speed becomes 2970 RPM, and the slip is about half that of the 2 pole version.
b) There will be a bit more eddy current and hysteresis losses in the stator because the magnetic field is changing more quickly. But because it started out pretty efficient, this will make the motor run a only bit warmer. Very high speed field, say 400Hz would make the bigger difference and may cause problems.
c) The power out of the motor approximately doubles.
d) getting carrined sway ccould result is such high terminal voltages that the stator insulation breaks down.

An easy wasy to figure out if this is true is to put your accounting cap on:
It's not a perpetual motion machine so...
    Watts out = Watts in - losses.
Where,
      Watts in = Volts * Amps
      Watts out = Torque * Speed (radians per second)
      Losses = Small (motors run a bit fast don't catch fire)
      and Speed =~ Volts (from the very beginning)
Converting:
      Volts * Amps =~ Torque * Speed
      and Speed =~ Volts.

This is all we need to roughly determine the entire motor behavior.

Note that a part of the motor losses are due to slip:
1470 / 1500 = 2% of the total 8% losses this is loss in the rotor. My guess is that the other 6% are divided roughly equaly between resistive and magnetic. THe balance of these varies according to speed and current.

(see Crompton parkinson for a 92% efficient motor that's on the shelf and reasonably priced in Australia)

4 Pole has more torque for a given frame size
eg:
160LR Long frame from C-P
2Pole 3000RPM 18.5kW 92.5% eff 60.2Nm * 3.1 = 186Nm
4Pole 1500RPM 15.0KW 92.0% eff 97.5Nm * 2.9 = 282Nm
Looks like diesel vs petrol!

Also, motors have to operate in that linear band near 'synchonous' or the efficiency will be really lousy.
You can calculate the slip-torque slope by knowing the rated torque and speed (slip). This will relatively linear up to the maximum torque point, so if Mx/Mn=3, the slip at maximum will be about 3 times that at rated output.

Edited.

[weber]

From what I see the 22kW 2-pole 132-frame has been removed from ABB catalouge.
Also the 11, 15 and 18.5kW have been recategorised as High output.

Thanks for letting us know about the new 1/2009 Industrial Performance Motors (IPM) catalog. But there is another ABB motor catalog of interest to us: the General Perfomance Motors (GPM) catalog. These are supposedly off-the-shelf designs with only a few options available (but they seem to be the options we need).

The only 132 frame motors that have been dropped from the new IPM catalog are those that were, and still are, in the GPM catalog. The GPM catalog is still dated 2008. We should keep an eye on this page for an update.
http://www.abbaustralia.com.au/product/ ... f527f.aspx

By the way, Coulomb pointed out to me a mistake in the old IPM catalog, which is still present in this latest one. The 3GAA 131 317-**E an 18.5 kW 2-pole 132-frame is listed as having a torque of 72 or 72.6 Nm. This is not possible for an 18.5 kW motor at the given speed. It should be more like 61 Nm.

[acmotor]

All solid reasoning guys.

If you run a 2 pole motor as above at 3000RPM it can supply 186Nm but the 4 pole can only supply 286/4 Nm = 71.5Nm at 3000RPM.

Ok you say, that is all obvious. It does suggest though that the 2 pole would work well with a gearbox in 2nd and 4th.

4 pole suits direct drive with lower voltage windings, I think we all agree, so you can get more revs.

Potentially the 2 pole could be half the weight of the 4 pole for the same kW (depending on manufacturer). Has someone done a 2 pole industrial 3PIM on a gearbox that works ? (gemini was 1.1kW 2 pole with 20:1 gearbox so doesn't quite count)