Why Prius can get 50 mpg even at constant highway speed?

The second law, equal and opposite force against the ICE rpm.

This is equivalent in the Prius to driving in "N" or in a manual transmission car pushing in the clutch so it is not engaged. No power flow, no torque.

Where does the counter torque that allows the ICE torque to flow though the PSD come from? The previous conditions means the ICE is decoupled from the drive path.

Ah! So we do have power flowing.

We don't have to as the ICE shaft is common to both. So 28% of the ICE torque is handled by MG1 and 72% is handled by the PSD as this is driven by the fixed gear ratios. Please understand it took me a long time to finally see how it works and it is just a frame of reference problem.

Using the ICE shaft as the common reference, the ICE power is a function of rpm * (MG1_torque/28%). We do not have to calculate anything else as the PSD gearing fixes the torque ratio and we can read out MG1 torque. MG1 rpm changes too quickly and we don't have fast enough sampling nor a good inertial model to calculate MG1-to/from-MG2 power flows by any other means that I am aware.

I have thought about how to trace the power flows more directly but realized the problem are the spider gears. We have a good handle on:

• ring gear - directly to MG2 and coupled to final reduction gears and differential
• spider gear carrier - directly to ICE
• sun gear - directly to MG1

What no one really likes to address are those spider gears on the spider gear carrier. If MG1 is dancing around like a maniac, the spider gears are going insane. If I were a professor with a wise-nice person student, I would assign them the task of developing and reporting on spider-gear motion in a Prius. . . . and contact university mental health services to check on the student the night before the assignment is due. It really is a hard problem.

Bob Wilson

 

I am ignoring losses. You can imagine a clutch locking the shaft mg1 down in this situation. This is equivalant to purely mechanical drive, not neutral. Power = torque times angular speed, so at 0 rpm, angular speed is 0 and there is no power. You can pick a constant to convert units. Now in the real world with mg1 we need a little electricity to hold it here since it is not mechanically locked. This power can be drawn from the battery or mg2. MG2 having mass is also going to add a little drag, so this is slightly less efficient than having simply the engine attached to the psd, mg1 locked mechanically and no mg2. So not ignoring losses and assuming no battery power, ICE supplies all its power through the ring gear, mg2 generates enough energy for electrical load including holding mg1 in place. This is very much like a single speed transmission with an alternator (mg2) acting as a load.

We also have the special case with no power flowing to the ring gear. Here the car with the brakes fully on and the accelerator pressed, with something physical locking down the ring gear (and maybe a little electricity). All of the ice power flows into mg1. Using those two equations we find that power = torqueIce x angularVelocityIce= .28 x torqueIce x angularVelocityMM1. Which reduces to rpmIce = .28 x rpmMG1 which you can find from the gear ratio If power was split at a constant ratio, we would get power at mg1 = .28 x torqueIce x angularVolocityIce = .28 x torqueIce x angularVelocityMM1, which is only true when rpmMG1 = rpmMG2.

 

You've just proposed an experiment I've not carried out, yet:

I'll put our NHW11 in "P" which uses a parking paw to hold the final gears. Then I'll do a 'forced charge' and record:

• ICE rpm
• MG1 torque
• MG1 rpm
• traction battery V
• traction battery A
• MG2 torque (should be zero but we want to confirm . . . I don't think we have MG2 amps)

I had never thought to run this experiment before but it will be fun to record and share the results.

Bob Wilson

 

Cool. Real experiments are often better than thought experiments.

As for the dance of mg1, its fairly straightforward, but looks funny in isolation. Your speed determines mg2RPM, the software decides based on your accelerator position and SOC the power requirements from the ice, and iceRPM to make that most efficient. mg1RPM needs to be whatever is left based on the psd gear ratio. The control software doesn't need to care the speed of mg1, as long as it isn't beyond its limits. If the ice power levels are set right, mg1 will generate or consume the correct amount of electricity. Pretty cool.

 

I ran the experiment on the way home at a couple of stop lights and discovered "P" does not support a forced charge. The traction battery current did not move when I floored the accelerator with our NHW11. I'll try the ZVW30 this evening

Bob Wilson

 

I knew the Gen 2 would only force charge in D, (I have not tried R) not P or N.

 

Upon further thought, I can run the traction battery down, say 50%. Then with recording instrumentation in place, start the car and monitor the "P" warm-up and recharge. Given the exothermic effects, I'll run the experiment Saturday morning when the temperatures are a cool and switch to my wife's car for the day:

• traction current :: initially with car not-ready to document 'overhead'
• traction voltage :: these provide the Whr saved
• MG1 torque
• ICE rpm
• MG1 rpm
• SOC
• starting and stopping traction battery temperature manually recorded

Bob Wilson