TMC is an independent, primarily volunteer organization that relies on ad revenue to cover its operating costs. Please consider whitelisting TMC on your ad blocker or making a Paypal contribution here: paypal.me/SupportTMC

motor efficiency

Discussion in 'Model S: Driving Dynamics' started by billarnett, Mar 17, 2015.

  1. billarnett

    billarnett Member

    Joined:
    Nov 15, 2008
    Messages:
    293
    Location:
    Emerald Hills CA
    I'm confused about torque sleep on the dual motor cars. My understanding is that it saves energy by using just one of the motors when cruising at a constant speed. Why is that good? I get it that the front motor on the P85D is more efficient than the big one in the rear. So that part makes sense. But not for the 85D where both motors are the same. But there's a more fundamental question here, too. In any given situation the amount of power required to maintain speed will be some fixed amount, P. And that implies a fixed current, I. Losses in the motor are proportional to I^2. But if we divide that power equally between the two motors each one has only I/2 and therefore losses of I^2/4 each or I^2/2 total for both. Using both motors gives the same power for half the losses. What am I missing here?

    And given all that, wouldn't 4 motors (one per wheel) be even better? And it would save a bit more energy by eliminating the differential and its losses. Plus torque vectoring and stability control would be easier.
     
  2. bp1000

    bp1000 Member

    Joined:
    Feb 14, 2015
    Messages:
    149
    Location:
    UK
    Motors are less efficient at low load. Turn off one on highway, use front at higher load, more efficient. Under acceleration, they can turn on an off for max efficiency depending on acceleration speed / load requirements etc.
     
  3. sandpiper

    sandpiper Active Member

    Joined:
    Sep 25, 2014
    Messages:
    1,464
    Location:
    Ontario, Canada
    The two motors have different ratio output gearboxes. If I understand correctly, the front one is geared lower so that at a given speed the motor runs higher torque, lower speed. This puts it in a different spot on the efficiency curve.
     
  4. Johan

    Johan Took a TSLA bear test. Came back negative.

    Joined:
    Feb 9, 2012
    Messages:
    6,890
    Location:
    Drammen, Norway
    #4 Johan, Mar 18, 2015
    Last edited: Mar 18, 2015
    motorcurve.gif

    (Image is for a DC motor but the principle is the same for AC induction.)
     
  5. jgs

    jgs Member

    Joined:
    Oct 28, 2014
    Messages:
    961
    Location:
    Ann Arbor, Michigan
    As far as I know, this is not yet supported by data or a knowledgeable source. See this thread. Real World Comparison of a S85 to an 85D Efficiency - Page 3
     
  6. Johan

    Johan Took a TSLA bear test. Came back negative.

    Joined:
    Feb 9, 2012
    Messages:
    6,890
    Location:
    Drammen, Norway
    I think you are talking past eachother. For the P85D the gearing is very likely (confirmed?) different while for the S85D it's very likely (confirmed?) identical between front and back.
     
  7. jgs

    jgs Member

    Joined:
    Oct 28, 2014
    Messages:
    961
    Location:
    Ann Arbor, Michigan
    If it's confirmed even only for the P, I'd like to see a source for that confirmation. I've looked and not found one.
     
  8. Johan

    Johan Took a TSLA bear test. Came back negative.

    Joined:
    Feb 9, 2012
    Messages:
    6,890
    Location:
    Drammen, Norway
    Even if the gear ratio is identical the front motor is without a doubt smaller than the rear motor so the point on the curve with max efficiency as a function of RPM will be different. But you are right I think that we just don't know the exact gear ratio for the front motor (the one that there are two of in the S85D). The exact gear ratio of the regular rear motor was only properly confirmed, I think, after Saleen announced their version of the P85+.
     
  9. jgs

    jgs Member

    Joined:
    Oct 28, 2014
    Messages:
    961
    Location:
    Ann Arbor, Michigan
    There are some interesting suggestions for non-intrusively determining the ratio further down in the thread I referenced.
     
  10. BerTX

    BerTX Member

    Joined:
    May 2, 2014
    Messages:
    636
    Location:
    Texas/Washington
    +1

    I don't have an efficiency curve for the AC motors Tesla uses (the curve above is for a DC motor), but typically the efficiency starts to drop below about 30% load. This is probably why the S85D should be a little more efficient under torque sleep than the P85D -- the 188 hp motor would be a little more loaded than the 221 hp front motor of the P85D at steady cruising speed. (yes, some difference from weight, but that is minor at cruising speed)
     
  11. lphe

    lphe Member

    Joined:
    Mar 18, 2015
    Messages:
    108
    Location:
    Eagan, MN
    From OBD II data I collect from my PHEV (not Tesla), I am able to observe electric motor efficiency. I can measure the mechanical power output of the electric motor as well as the electrical power input to the inverter. The efficiency (mechanical power output by motor divided by electric power input to inverter) varies from 0 when the motor outputs 0 kW of power to about 85% when the motor outputs 30 kW of power. At about 10 mph, the mechanical power output by the motor is about 1 kW and efficiency is 30%. At 30 mph, the mechanical power output is 4.3 kW and efficiency is 63%. At 50 mph, mechanical power output is 8.9 kW and efficiency is 74%. At 70 mph, mechanical power output is 13.7 kW and efficiency is 80%.

    So the motor operates more efficiently with increasing power up to about 30 kW. After 30 kW, efficiency starts dropping. Thus you would like to operate the motor at high output power. Pulse & Glide is one way to operate at high output power (but I don’t think it is going to be all that effective in increasing efficiency). When the motor is consuming power, the car is accelerating and the motor is outputting high power. During the glide (using neutral), little power is being consumed. So you are operating the motor in a more efficient operating range when consuming power. I have tried it and got mixed results. It may improve kWh/mile by 10% if you do it correctly.

    Maybe there are other tricks that the Tesla can use to operate the motor at higher efficiency more of the time.
     
  12. bp1000

    bp1000 Member

    Joined:
    Feb 14, 2015
    Messages:
    149
    Location:
    UK
    That is what i'm hoping thursdays announcement will be. A range mode that pulses the engine at an undetectable frequency to the user but its pulsing a higher load to maintain the speed instead of running at constant low load.

    Bit like a automated version of pulse and glide but they can do it at the millisecond level so its undetectable.
     
  13. billarnett

    billarnett Member

    Joined:
    Nov 15, 2008
    Messages:
    293
    Location:
    Emerald Hills CA
    But WHY is it more efficient at high output?
     
  14. AndreyATC

    AndreyATC Member

    Joined:
    Dec 19, 2013
    Messages:
    381
    Location:
    NY, USA
    If the load "thing" is true, does this mean MS is more efficient driving through hills?

    Also, can we really apply this to AC motors
     
  15. Johan

    Johan Took a TSLA bear test. Came back negative.

    Joined:
    Feb 9, 2012
    Messages:
    6,890
    Location:
    Drammen, Norway
    #15 Johan, Mar 18, 2015
    Last edited: Mar 18, 2015
    This not my answer, but a good one:

    "The other way to ask your question is why is efficiency low at low loads? Friction is the main cause of inefficiency at low loads. Losses due to friction are essentially constant with respect to load so at low loads, the majority of your input power may be used to overcome friction. As the load increases, friction plays a smaller and smaller roll in the overall efficiency. Granted, other inefficiencies begin to occur at larger loads ($I^2R$ losses, copper losses, stray load losses, etc.) but in a well-designed motor the efficiency will peak in the 80-100% load range."

    Source: classical mechanics - Why is an electric motor more efficient at higher loads? - Physics Stack Exchange

    This PDF has very good info too: http://www1.eere.energy.gov/manufacturing/tech_assistance/pdfs/motor.pdf

    Basically from reading up on this subject (it's interesting!) I've found that in industrial applications this is a big issue. In an industrial setting, for example a pump, some machine etc. you don't want an oversized motor - if it runs at low loads efficiency is a lot worse which can mean a lot of $$$ over time.

    In a car such as the Model S, this must be an issue: you want the peak performance output of 320 kW+, the very quick 0-60 etc. but at the same time 99% of the time your loading the motor with just a fraction of this output...

    My amateurish guess is that in a brushless AC motor (like Tesla's) the efficiency curve is flatter due to less friction losses than one with brushes.
     
  16. lphe

    lphe Member

    Joined:
    Mar 18, 2015
    Messages:
    108
    Location:
    Eagan, MN
    Yes—the car is more efficient in converting electrical energy to mechanical energy when going uphill. But your overall kWh/mile will go down if you drive on too hilly a road.
    While efficiency is greater when climbing the hill, you have to convert more electrical energy to provide the additional (potential) energy to climb the hill. Since efficiency is still less than perfect, some of the electricity used to climb the hill is lost (mostly in the form of heat). If the energy loss required to climb the hill exceeds the gain from increased efficiency, then you will use more electrical energy overall for the hilly road vs. a flat road and kWh/mile will go down.
     
  17. apacheguy

    apacheguy Sig 255, VIN 320

    Joined:
    Oct 21, 2012
    Messages:
    3,729
    Location:
    So Cal
    Were there not rumors that the efficiency of single motor P85s would also see a bump as a result of the lessons learned from torque sleep? I'm wondering why there was no mention of an effect on the S85 as well.
     
  18. AmpedRealtor

    AmpedRealtor Active Member

    Joined:
    Jun 30, 2013
    Messages:
    4,711
    Location:
    Buckeye, AZ
    There are some speculating in the TM Forums that Elon will announce a kind of "micro sleep" for the non-D fleet. Per those remarks, the motor will be kept in its higher efficiency band through pulsing many times per second. Simple example... Suppose it takes 20 kWh to maintain forward momentum at speed, but 20 kW is not where the motor is most efficient. If you pulse/sleep that motor 10 times per second, the motor would have to draw 200 kW per 1/10th second pulse to maintain the same momentum as a non-pulsed motor at 20 kW. Again, quick and dirty and my math may not even be correct. But that's the idea, to keep the motor drawing an amount of current that would keep it in an optimal efficiency range.
     
  19. lphe

    lphe Member

    Joined:
    Mar 18, 2015
    Messages:
    108
    Location:
    Eagan, MN
    #19 lphe, Mar 18, 2015
    Last edited: Mar 18, 2015
    Suppose the motor is turning at 7000 rpm. Due to friction, unless you provide electricity to power the motor it will slow down. Assume that it takes 1 kW of electricity just to maintain 7000 rpm. Then the mechanical power output of the motor is 0 kW and it is consuming 1 kW of electricity. Efficiency is 0%. If you want to go 10 mph, you need the motor to provide 1 kW of power. That means you need 1 kW of electricity just to counteract friction and another 1 kW of electricity to provide the mechanical power needed to propel the car. In this case you get 1 kW of mechanical power output from the motor by feeding it 2 kW of electricity, i.e. efficiency is 50%. If instead you drive on the freeway, you might need 14 kW of power to maintain speed. Friction will increase with higher RPMs. In this case, you might need 2 kW of power to counteract friction and 14 kW to provide the output power for the motor to propel the car, i.e. efficiency is now 14/(2+14) = 88%.
     
  20. Mattias

    Mattias Member

    Joined:
    Mar 15, 2015
    Messages:
    21
    Location:
    (Sweden)
    100$ says range has not increased by Thursday.

    The Variable-Frequency-Drive already does PWM to control the engine. And also, electric motors has been around for over 100 years, and this would be used widely in industrial applications. To this date, I have not heard of any "micro-sleeping" electric motors, so no I doubt it´s possible.

    BUT, it would be cool for future electric cars to have a "cruise-drive", a smaller motor for high speed cruising. Or a basic 2-speed gearbox, with the other gear activated by "range-mode".
     

Share This Page