1. Attachments are working again! Check out this thread for more details and to report any other bugs.

RPM as related to 15-18kw output

Discussion in 'Gen 2 Prius Fuel Economy' started by rustystew, May 2, 2011.

  1. 2009Prius

    2009Prius A Wimpy DIYer

    Joined:
    Mar 25, 2009
    2,705
    509
    63
    Location:
    USA
    Vehicle:
    2009 Prius
    This is great! So maybe Ken's graph represents the total output from the engine, part of which will be used to propel the wheels and other part of which will be lost in the transmission and other places. Now planetaire just showed that the loss is RPM dependent; especially the loss is less at low RPM. Thus we can see that while the engine is most efficient at some high kW numbers, the best fuel economy for the entire car would be achieved at some (slightly) lower kW numbers.
     
  2. rustystew

    rustystew New Member

    Joined:
    Apr 22, 2011
    54
    4
    0
    Location:
    Galisteo NM
    Vehicle:
    2007 Prius
    Model:
    N/A
    Hey guys, thanks for helping me understand all this. There is just so much to it. I think a lot of this is still over my head a bit though.
     
  3. 2009Prius

    2009Prius A Wimpy DIYer

    Joined:
    Mar 25, 2009
    2,705
    509
    63
    Location:
    USA
    Vehicle:
    2009 Prius
    Thanks again to planetaire's kind help that enables me to collect some more data to share with everyone here. I analyzed about 10 hours of real life driving, mostly local roads with a small portion on the highway. First the relation between net kW and RPM seems quite simple:

    [​IMG]

    This is almost the same as planetaire's graph:
    The difference being the "kW" - whether it is "net" (passive CAN message) or "gross" (active CAN request) as we have discussed before. I have no idea how ScanGauge got its kW reading but from my experience the ScanGauge 15 kW corresponds to 12 net kW.

    Now the more interesting graph of ICE efficiency as a function of RPM:
    [​IMG]
    The ICE efficiency is calculated by dividing the net kW with the energy content in the fuel used (per unit time of course). There is some unknown offset in the fuel usage calculation so there are two curves of different color. Fortunately the offset does not seem to affect the trend much.

    It looks like the most efficient RPM is 2000 ~ 2400, corresponding to 15 ~ 20 net kW, or 18 ~23 gross kW, which agrees with the left region of the 230g/kWh area in Ken's graph:
    On the other hand the drop off of efficiency above 2400 RPM does not agree with Ken's graph, which suggests good efficiency up to almost 4000 RPM.

    The good news is that the efficiency doesn't vary more than a few percent over such a large RPM range that we don't need to overly concerned with hitting the "sweet spot". It's more like a "sweet range" I would say. :)
     
    3 people like this.
  4. ystasino

    ystasino Active Member

    Joined:
    Apr 17, 2007
    793
    49
    0
    Location:
    Tucson, Arizona
    Vehicle:
    2007 Prius
    Model:
    Two
    In my 45-55 mph 8 mile drive to the swimming pool, which I have monitored ~60 times, my experience suggests that 12.0 scangauge II kW is the sweet spot for mild acceleration and about 8.0 kW is the sweet spot for mild deceleration. Varying from 8.0 to 9.0 kW (both are IGN 14) at 50 mph results in a ~8 instant mpg decrease from ~73 to ~65.

    In my 55-65 mph 30 mile trip to DC and back, which I have monitored ~15 times, my experience suggests that 15.0 SG II kW is the sweet spot for mild acceleration.

    I'm now finishing my third 60+ mpg tank with hundreds of highway miles. I do not log single trips because I do not think that they are particularly reliable and we end up repeating the same error rather than averaging it out.

    I have tried to get the car to yield better efficiency in trips with ~26-28 SG II kW and it just hasn't worked at all. However, in a flat terrain without traffic or wind and similar temperature and humidity (ideal condition on a track) it is very possible that Ken's graph would yield practical benefits. I don't mean to be a downer, but I do not believe that there is a way to take real-life measurements and project them to the ideal graphs obtained in a lab.

    I also think that this is the reason why some very very experienced hypermilers with deep knowledge of the Gen II's FE don't exactly agree on their results. Wayne has logged 1000s of hyper-miles with SHM and Hobbit's sweet spot refinement paper seems to be describing something that also works very well. My much more limited experience suggests that both are valid, but under limiting circumstances of flat terrain and ideal hill elevations respectively.

    All highway hypermilers should be able to get 57+ mpg on the highway using these techniques but there are some reasons why there's no consensus on how to best drive the Prius, beyond that FE. These reasons are related to the fact that our terrain varies wildly and our measurements are imprecise and the time we have in each trip is also varied.

    I'll be using Mytracks that measures elevation and other distances and start logging my trips better, but I still doubt I will refine this much more.
     
  5. 2009Prius

    2009Prius A Wimpy DIYer

    Joined:
    Mar 25, 2009
    2,705
    509
    63
    Location:
    USA
    Vehicle:
    2009 Prius
    Very cool. Maybe you could make a graph of MPG vs. kW from your experiments. That will be interesting to see.
     
  6. ken1784

    ken1784 SuperMID designer

    Joined:
    Dec 18, 2003
    2,940
    1,359
    67
    Location:
    Yokohama, JAPAN
    Vehicle:
    2010 Prius
    Model:
    N/A
    Regarding to fuel efficiency (MPGvsMPH), Wayne Brown's simulator tells us slower is better at highway speed (above 42mph threshold).
    Then, 60mpg at 60mph theory is generally accepted.

    Ken@Japan

    [​IMG]
     
  7. 2009Prius

    2009Prius A Wimpy DIYer

    Joined:
    Mar 25, 2009
    2,705
    509
    63
    Location:
    USA
    Vehicle:
    2009 Prius
  8. ken1784

    ken1784 SuperMID designer

    Joined:
    Dec 18, 2003
    2,940
    1,359
    67
    Location:
    Yokohama, JAPAN
    Vehicle:
    2010 Prius
    Model:
    N/A
    OK. So, has anyone achieved 70mpg at 66 ~ 68mph during the reasonable long distance?
    If yes, please show us the data.

    In my understanding, Prius driving resistance is approx 495N(newton) at 67mph.
    It is 14.8kW or 138Wh/km.
    The 70mpg result means Prius runs at 42.6% average fuel efficiency or 185g/kWh average BSFC range.
    I am saying it is impossible.
    Of course, short distance of 70mpg at 67mph is possible on downhill, tail wind, SOC change or other situations.

    Ken@Japan
     
    2 people like this.
  9. ystasino

    ystasino Active Member

    Joined:
    Apr 17, 2007
    793
    49
    0
    Location:
    Tucson, Arizona
    Vehicle:
    2007 Prius
    Model:
    Two
    The problem is that for many of us driving on flat terrain is not an option so our optimal mpg per speed varies.

    I'm currently leisurely looking into geomapping phone apps that could in theory be combined with OBD II data in one app and come up with an ideal efficiency range.
     
  10. planetaire

    planetaire Plug in 20 kWh 85 km/h or > 208km range

    Joined:
    Aug 11, 2009
    190
    139
    15
    Location:
    France
    Vehicle:
    2007 Prius
    Model:
    II
    Re: RPM as related to 15-18kw output. Sweet range.

    Splendid graph.
    Probably net efficiency is in blue color and engine in red ?

    I suppose the passive torque is after mecanical lost in planetary AND via MG1<->MG2 losts. But i am really not sure:

    Try this test: Drive with a constant rpm (you may have to move a little the pedal accelerator in order to maintain rpm) with a rpm that increase your speed (begining with a low one). Compare passive torque (net) and active torque (engine).
    Engine don't change his torque but net torque change. Losts in psd are a few % and depending more in torque value than in speed. But power transmited via MG may be a big % of engine power and, specialy with a low speed, the motor efficiency is falling (the motor efficiency can be found in argone tests).

    In this case the interesting efficiency depend upon engine efficiency AND MG1<->MG2 efficiency. In other words depend on rpm AND speed together.

    I agree.
    That's the conclusion i haved reading this graph: (with a maximum efficiency near 2000 rpm)

    [​IMG]

    -Efficiency is the red line.
    -engine Nm/2 blue (engine torque, not net torque)
    -power in green

    :)
     
    1 person likes this.
  11. ystasino

    ystasino Active Member

    Joined:
    Apr 17, 2007
    793
    49
    0
    Location:
    Tucson, Arizona
    Vehicle:
    2007 Prius
    Model:
    Two
    I think Hobbit's article refers to more practical terrain and traffic conditions, while SHM works on flat terrains.

    I think on a flat terrain it might be possible to do 60-65 mpg at 65 mph for a long time. I realize that 60-65 is a wide range, but I have no flat terrain to practice near where I live.

    I don't think that IGN is an accurate enough metric to go by not even on flat terrain. I'm certain it can be refined. IGN 14 on flat at 50 mph refers to a range of scangauge II kW (~4.7-9.0) and a range of iMPG (~104-70). I prefer to use 8.0 kW for coasting at that speed and 12.0 kW for accelerating.

    I agree that 70 mpg at 66-68 mph is not feasible with a Gen II.
     
  12. ystasino

    ystasino Active Member

    Joined:
    Apr 17, 2007
    793
    49
    0
    Location:
    Tucson, Arizona
    Vehicle:
    2007 Prius
    Model:
    Two
    Re: RPM as related to 15-18kw output. Sweet range.

    What is "efficiency"?

    ICE efficiency is not the only factor here. Rate of charge and discharge of the traction battery is also a factor. I don't think you can isolate ICE efficiency without taking into account the efficiency of the energy conversion and thermal losses at higher RPM.

    More importantly, many people get higher mpg when coasting at 1500 RPM at 50 mph rather than accelerating and decelerating between 2000 RPM and something else.
     
  13. planetaire

    planetaire Plug in 20 kWh 85 km/h or > 208km range

    Joined:
    Aug 11, 2009
    190
    139
    15
    Location:
    France
    Vehicle:
    2007 Prius
    Model:
    II
    Re: RPM as related to 15-18kw output. Sweet range.

    The graph i publish are with constant Soc and a 63% value, so there is no charge and no battery discharge.
    Of course ice efficiency is not the only factor.
    And this grah show ICE efficiency alone vs rpm. It is a 2D graph.

    For net efficiency you nead a 3D graph at least.

    :)
     
  14. ken1784

    ken1784 SuperMID designer

    Joined:
    Dec 18, 2003
    2,940
    1,359
    67
    Location:
    Yokohama, JAPAN
    Vehicle:
    2010 Prius
    Model:
    N/A
    I am not saying anything about terrain, traffic conditions or technique.
    I am saying just a physics.
    From point-A to point-B, Prius requires 13800Wh per 100km drive at 67mph without altitude change.
    If the average mpg result was 70mpg, Prius has to run at 42.6% average fuel efficiency or 185g/kWh average BSFC range.
    I am saying it is impossible.

    If Prius could run 230g/kWh range or 34% fuel efficiency, the peak mpg number at 67mph will be 55mpg.

    Ken@Japan
     
  15. 2009Prius

    2009Prius A Wimpy DIYer

    Joined:
    Mar 25, 2009
    2,705
    509
    63
    Location:
    USA
    Vehicle:
    2009 Prius
    Thank you Ken and ystasino for the sanity check! I just went back and re-read Hobbit's long writing:
    Sweet spot refinement
    which I quote some key paragraphs here that caused me to (mistakenly?) think he got "70 at 70":
    Hobbit did not give any exact data for "pure" highway driving at high speed. Only the 70 MPG number quoted above. I actually asked for some more details in this thread that I started:
    http://priuschat.com/forums/gen-ii-...super-highway-mode-lower-speeds-50-55mph.html
    but so far there is no answer.

    ystasino, since you have been accumulating driving experience with Hobbit's technique, could you help collect some data? It would be very interesting to compare, for example, 15 kW pulse vs. 11 kW pulse, using the same technique and keeping approximately the same overall average speed, to see if 11 kW pulse would really give lower fuel economy than 15 kW pulse does. Its OK if the overall average speed has some variation since we can always compare your result with Wayne Brown's curve. Thanks!


    Thank you again planetaire without your help I won't be able to collect these interesting data to share. The blue and red curves in my graph was actually both "net" (passive). The difference was one was with the fuel usage offset and the other without.

    I just programmed in the "gross" (active) power and will report back later on the analysis you suggested. I have to think about how to present the power - rpm - speed 3D graph! :)

    One thing I did notice was that the ScanGauge reports higher power than both the "net" and "gross" power read from the CAN messages. It's about 3 ~ 3.5 kW more than the "net" power. It's much harder to analyze since the ScanGauge updates very slowly and there is no way that I know of to record the ScanGauge readings.
     
  16. SageBrush

    SageBrush Senior Member

    Joined:
    Jun 4, 2008
    11,627
    2,530
    8
    Location:
    Southwest Colorado
    Vehicle:
    2012 Prius v wagon
    Model:
    Two
    I also find this sentence confusing. I don't think Hobbit was saying he averages 68 mph while averaging 70 mpg. His point (seems to be?) that his terrain will sometimes push the car up to 68 mph given his requirement that the engine always operate at the most efficient possible (or not at all,) and that over distance this is the best trade-off. Using 34.5 kwh/gallon, 38% efficiency, and 138 wh/km energy consumption at 67 mph, the most optimistic result works out to 95 km/gallon, or 95*.62 = 59 mpg.

    Put another way: for best fuel economy, if you have to choose between sub-optimal ICE operation or more air drag -- pick more air drag.

    Oh, and if you just *have* to get over 60 mpg at 67 mph, you can. Drive at my altitude, or run on hyper-inflated bald tyres ;)
     
  17. ystasino

    ystasino Active Member

    Joined:
    Apr 17, 2007
    793
    49
    0
    Location:
    Tucson, Arizona
    Vehicle:
    2007 Prius
    Model:
    Two
    Does this mean that the Prius won't accelerate above 67 mph on a flat terrain, without tail wind or drafting? At what tire pressure and what driver and fuel weight?

    The way kW are displayed in SG II using Hobbit's xgauge, my non scientific experience is that it will accelerate above 67 mph. But now that I have a terrain readout, I might be able to test.
     
  18. SageBrush

    SageBrush Senior Member

    Joined:
    Jun 4, 2008
    11,627
    2,530
    8
    Location:
    Southwest Colorado
    Vehicle:
    2012 Prius v wagon
    Model:
    Two
    Yep, that is what it means. Assuming constant 14.8 kw power. Just remember though that the power was measured at the driveshaft.
     
  19. ken1784

    ken1784 SuperMID designer

    Joined:
    Dec 18, 2003
    2,940
    1,359
    67
    Location:
    Yokohama, JAPAN
    Vehicle:
    2010 Prius
    Model:
    N/A
    Prius will accelerate at 67 mph on a flat terrain when it supplies more than 14.8kW power.
    I don't have the test condition numbers, such as tire pressure, fuel weight or others..

    Ken@Japan
     
  20. ystasino

    ystasino Active Member

    Joined:
    Apr 17, 2007
    793
    49
    0
    Location:
    Tucson, Arizona
    Vehicle:
    2007 Prius
    Model:
    Two
    It's not easy because there's no flat terrain within 2 hours of highway driving from where I live. It's further complicated because I need someone to record the MPH, SOC, iMPG and kW as I am driving across some really busy interstates.

    My best tank (63.0 mpg for 512 miles with 70% interstate driving) involves pulsing at 12.0 kW (or as close to that as possible to maintain an reasonable speed) and holding steady or decelerating slightly at 8.0 kW. I am also using WS in steep or prolonged downhills.

    It would be interesting to find a flat terrain pulse at 12.0 kW for 2 minutes and compare the mpg to a sequence of 1 minute 15.0 and 9.0 kW pulses.