Water cooled induction motors, review and discussion of Tesla model S motor 1

[xj25]

Hi,

after seeing a video of Tesla Model S drive and some impressive specs I have done a little of investigation in
currently in production water cooled induction motors.

I took 250kW 4pole as reference for comparison, and found the following typical specs for some motor builder datasheets:

VEM, ABB, EMOD, 250kW 400V 50Hz, 4 poles water cooled
Torque about 1.6kNm
stator+axle length, from L=1100 to 1600mm
supports to axle length R=315mm
Weight, from 1030(ABB) to 1750kg

From WEG subsidiary AKH, I found a surprisingly small family with a "special" water cooling solution (see below datails), i.e.
WEG UTTWp 180.3-4 260kW water cooled with "special" design,
400V 4000rpm 621Nm x2.4breakdow torque,
r=180mm L=not shown?, 295kg

Tesla model S motor is told to have:
270kW 4 poles (396V battery supplied VFD + water cooled stator and rotor, some patents in rotor cooling released, see google)
up to 18000rpm? 440Nm
about 31.7kg?? very doubtful figure,
L=about 600mm r=about 180mm (from pictures, VFD is side by side with motor)

Summary for 250kW nominal power, 400V, 4 pole motors:
[ul]
[li]regular industrial water cooled (typical): Torque=1600Nm L=1300mm R=315mm 1300kg[/li]
[li]WEG special water cooling: Torque=621Nm L=? R=180mm 295kg[/li]
[li]Tesla model S: Torque=440Nm, about L=600mm R=180mm ?kg[/li]
[/ul]
Some questions that I would like to share:
- The nominal torque is limited fundamentally by the voltage if I remember well. Industrial motors of same rated power seems to have (x4) nominal torque comparing the tesla. Do that have any relation with motor size and weight for reducing them? saturation flux maybe?
- Typical breakdown torque of these industrial motors is 2.5 times nominal. As far as I know, with a suitable feedback control, a motor could be used near this torque, but thermal load will destroy it. However, what happens IF a suitable cooling system is provided?

Anyway even so, the Tesla drive has outstandings specs for its dimensions, and I would like to know if there is other fundamental aspects that are missing in my very basic discussion.

references:
Industrial motors 250kW 4poles:
Siemens 1LA4 312-4AN 4 poles 250kW forced ventilation
1.63kNm x2.3 bt 93.8% 1480rpm 2-6kV 50Hz
1650kg L=1610mm R=315mm,

VEM K21B 315 L4 4 polos 250kW, water cooled
400V 50Hz 250kw 96.1% x2,3bt 1485rppm
1750kg L=1466mm R=315mm,

ABB M3LP 315 MLA 3GLP 312 410- 250kw 4 poles, water cooled
400V 50Hz 1485rpm 96% 1.6kN x3.1 bt
1030kg L=1156mm R=315mm

EMOD 315S/4-460 250kW water cooled 250kW
400V 1480rpm 1.61kNm 95.3& water cooled
1710kg L=1215 R=315

WEG UTTWp 180.3-4 260kW water cooled with "special" design,
400V 4000rpm 621Nm x2.4bt 295kg
r=180mm L=?
With an enormous enhancement of the power density by a factor of 1.5 to 2.6* – compared to standard water-jacket cooled motors – three series provide the finest high-tech drive. The secret to success: The special, patented cooling system. These AC motors from AKH benefit from the cooling elements integrated in the stator lamination. The advantage: Absorption of the thermal losses directly at the location of occurrence. The effect: No additional dual steel pipes as with standard water-jacket cooled machines; a significant reduction in weight and installation size.
The induction motors in the UTTW series cover the power range from 24-240kW with a shaft height of 112 to 180mm, while the UTTW(p) series of premium induction motors – equipped with a water-cooled rotor – extends the range up to 300kW with the same shaft heights.

Tesla model S base, 270kW 4 poles (396V battery supplied VFD + water cooled stator and rotor, some patents in rotor cooling released, see google)
up to 18000rpm? 440Nm
about 31.7kg?? not likely, L=about 600mm r=about 180mm (from pictures)

Sources of Tesla drive info:

https://en.wikipedia.org/wiki/Tesla_Model_S#Powertrain

https://youtu.be/NaV7V07tEMQ?t=35s

https://chargedevs.com/wp-content/uploads/2014/10/Tesla-Model-S-Motor-Windell-Oskay-copy.jpg

https://chargedevs.com/newswire/elo...io-is-the-challenge-with-ac-induction-motors/

https://www.youtube.com/watch?v=zMQjcukphpA&t=1496s

 

[waross]

I would like to know if there is other fundamental aspects that are missing in my very basic discussion.

RPM
To compare motors it may be well to normalize the RPM.
An example:
Almost any good quality industrial induction motor.-
Take a motor rated at 60 Hz, 230 Volts and 1760 RPM, drive it at 120Hz and apply 460 Volts. You have just doubled the HP.
Same torque, twice the speed = twice the HP. Same current at full load.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[xj25]

In the given examples the "regular" industrial motors of ABB, EMOD, VEM, are all 4-pole motors of 1500rpms synchronous speed. The "special" WEG and Tesla motors declare 4000rpms the first, and can be calculated 5800rpms the second (to match 270kW power).
That gives for Tesla about 193Hz of "nominal" frequency for rated power (133Hz the WEG).

In this case, as maximun voltage is about 400V for all, V/F nominal power point will be about 400/193=2 (Tesla) or 400/133=3 (WEG), comparing 400/50=8 of one of the "regular" motors.
So Volts/Hertzs is 62% lower in the WEG and 75% lower in the Tesla than in a conventional motor.
How that affects to the motor design? Is that what explains the reduced weight and size?

Regards

 

[waross]

Take a 400 Volt motor @ 50 Hz, V/Hz = 8
Now split each winding in half and connect the halves in parallel.
The motor is now good for 200 Volts at 50 Hz, V/Hz = 4

Your normal 1500 RPM motors will turn about 5800 RPM with 193 Hz applied. The HP at 5800 RPM will be about 965 HP if the V/Hz ratio is maintained.
The windings will have to be split and reconnected in several parallel paths or the motor will have to be rewound with fewer turns of heavier wire.

Tesla 270 kW @ 193 Hz is comparable to 70 kW at 50 Hz.
What I meant when I said to normalize the speed is to calculate the HP for each motor at the same speed and or frequency before comparing motors.
You mentioned 18000 RPM. A 44 kW motor at 50 Hz is capable of 270 kW at 18000 RPM.
These figures assume that the rotor is physically capable of the speeds suggested.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[sreid]

http://www.motor-design.com/cmsAdmi...trtaction-systems-for-electrical-machines.pdf

 

[itsmoked]

I think one might also have to be wary of normal assumptions when discussing obscure motors like high performance vehicle traction motors. I could easily see obscene power claims that would seem far more generous than reality. A vehicle can actually only absorb a certain amount of energy (kinetic) before reaching any one of several limits like national speed limits or peak controllable velocity. These represent time limits on how much energy (power x time) can ever be demanded from the motor.

The result is a motor that you can declare is capable of X amount of power when in reality it's realistically constrained from ever needing to supply it for more than a very few minutes which is obviously not what a classic industrial motor might be faced with - possibly months at nameplate output.

There's also the sophisticated drives running the high performance traction motors that likely have all sorts of protections that will limit the motor's output verse time in a relatively smooth manner the average user might never recognize as power limiting.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

Actually the 6X is not real power. It is highly reactive. The maximum power may be limited by the breakdown torque which in a design A motor may be 280% of normal, and may be even higher in a specially designed traction motor.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[LionelHutz]

A Tesla is direct drive so the torque curve of the motor and inverter basically sets the acceleration rate of the car. Peak HP just doesn't mean that much. Tesla hasn't done anything too special. They make the 0-60mph times shorter by steeper gearing in car and "limiting" it's top speed. They increase the range by stuffing a whole bunch of battery into the car.

You are comparing general purpose industrial motors to a specific optimized single purpose motor and drive system. They don't compare, at all, which makes it pointless to try.

 

[xj25]

Thanks, got it. Two additional doubts.
1. If motor is designed for a higher freq. Its power/weight ratio will be higher.
Why is that not generally used in vfd Motors. I.e. with higher pole count to keep speed, or even with a reduction gear.

2. Comparing with given example, 44kW 50hz, increasing v/freq shall cause at least additional iron losses and Cooper losses, so cooling should be revised, right?

Tesla S is not direct drive, has a fixed reduction gear.

 

[LionelHutz]

Sure, it's power to weight ratio can be higher but you have to design a WHOLE bunch of things differently for a higher frequency motor. In general industry, if you used special motors at higher rpm's then you couldn't switch between motor control methods.

The Telsa doesn't have a transmission. Tesla cranked up the motor rpm to get the gear reduction ratio higher so the motor torque gets multiplied enough to make the car perform.

 

[xj25]

Seeing Tesla S Drive, and after the discussion seems, as you have well explained, that the Model S motor is a Tailor made
solution with a high freq. motor and special cooling. If you see the video, the inverter design is also quite remarkable and compact, sharing the water cooling with the motor.

They get a high power/weight ratio with a low profile motor than can be well integrated in the car. As Lionel explained, they use a 1 stage reduction (8 or 9:1 I read) to get the torque and speed needed at wheel level.

All that seems quite well thought. I am in railway bussines, and in my experience the motors used are 1500-2000rpms rated, VFD controlled and always with a fixed reduction gear. Typical power levels for LRVs i.e. are from 70 to 150kW depending if the motor drives an axle or a single wheel. See i.e. Link

This is not so different from Tesla application, but the norm is using auto ventilated aircooled motors. I would say that the same approach used by Tesla could be applied to this application, due to speed cycles, maximun speeds etc. are very similar to road cars in a city.

 

[itsmoked]

Ah, when you say "railway" to me I think of traction motors that tend to be 1.6MW.

Keith Cress
kcress -

http://www.flaminsystems.com

 

[waross]

Hi Keith.
1.6 kW sounds like a good ballpark for the diesel engine in a locomotive.
Divide that by 12 wheels and you get 133 kW per wheel motor.
I consider that to be still in the ballpark.

Bill
--------------------
"Why not the best?"
Jimmy Carter

 

[LionelHutz]

Yes, you could apply the same approach, but I would suspect that in LRT applications package size is secondary to simplicity and serviceability.

 

[davidbeach]

If I'm doing the math right, a modern diesel-electric locomotive will be in the 3.3MW (electrical to the traction motor) range. At one traction motor per axle for six axles, I'd sort of expect something 600kW ish per axle. Probably larger to allow for some pretty gnarly short-time ratings.

 

[EdStainless]

But the link was to 'light rail' systems, trams. Nothing like mainline freight.

A lot of special purpose motor look funny. I used to build well motors (and pumps) 5.4" OD x 20' long, 250 hp, 2200V 60A

= = = = = = = = = = = = = = = = = = = =
P.E. Metallurgy, Plymouth Tube