Introduction to Prius Power Flow

A frustrated automotive engineer once claimed that you had to have a doctorate to understand the Prius transaxle. Learning how it works can lead to a headache but it is understandable. Hopefully, the following charts and data will clear things up.

Power Split Device

The key to understanding is how the planetary gear, power-split device works. The ICE powers the planetary gear carrier and the planetary gears are in a computer controlled dance between MG1 driving the sun gear, and MG2 driving the ring gear.

(I edited in the ICE power shaft from images of these parts found at: http://privatenrg.com/01-03MGs%20&%20PSD.jpg)

Normal Power Mode

In this mode, MG1 provides the counter torque to the engine, always 28%. This counter torque generates power that electrically passes to MG2. The remaining 72% of torque passes to the ring gear and via the chain drive to the reduction gears and differential.

In order to make the power schematic easier to understand, I have shown MG2 connected to the chain drive pickup. In fact, it is connected to the ring gear in an very compact arrangement. However, this does not change the basic power flow.


Energy Re-Circulation ("heretical" or "overdrive")

Unfortunately, normal mode leads to higher engine power settings and speeds. To put the engine in a lower-speed, highly efficient region, the transaxle can use MG2 as a generator and pass power back to MG1. MG1 then combines this with the ICE power and it 're-circulates' via the chain back towards the wheels.

When some of the earliest Prius owners noticed this it was called "heretical" mode. Toyota engineers published a paper and called it "energy re-circulation" mode. Regardless, it provides a very efficient operating mode.



Transition Between Normal and Re-Circulate Mode

The following data was captured from a Graham miniscanner and shows the transition from 'normal' mode, used for higher power operations, to "energy re-circulate." Acceleration and hill climbing are high-power, normal modes of operation but once the target speed is achieved, the car transitions to "energy re-circulate mode."


'Kevin17' pointed out that in "energy re-circulate" mode, MG1 is working to keep the ICE from spinning faster. This additional load on the ICE keeps the ICE rpm low so it can generate maximum power for the fuel burned AT A LOWER POWER SETTING. It is similar to 'lugging the engine' only done so the ICE is never overstressed. Thanks 'Kevin17' for the insight. Also, Hobbit posted, ". . . the engine was running barely above idle -- 1100 RPM or so -- the throttle was significantly open, shaft torque was high, and RPM was being kept down by the "heretical mode" electrically-routed overdrive. . ."

Optimum System Efficiency

As the Toyota engineers pointed out, optimum system efficiency requires balance and tradeoffs. Sometimes one part has to operate in what appears to be a less efficient mode so another part can perform at peak efficiency. It is the total system performance that is important.



Example of Regeneration

One of the better ways to see regeneration in action is to descend a hill. In this example upon entry to the hill at 65 mph, the transaxle was put in "B". The battery was at a nominal 60% charge and quickly reached the 80% limit. Thereafter, no further changing could occur and the rest of the braking was from spinning the engine.




Climbing The Hill

To drain the battery from a nominal 60% down to 40%, drive up a hill at high speed. In this case, there was a small 'leveling' when the vehicle attempted to transit to "re-circulate" mode. Then the hill climb continued.



Hopefully, this makes things a little clearer.

About Efficiency

Dept. of Energy measurements indicate MG1 and MG2 typically have 95% efficiency as either a generator or motor. Switching power supplies like the motor drive circuits can achieve 90-95% efficiency. So though there are electrical losses, they are fairly modest, especially considering only a fraction of the power passes through this path.

As for mechanical losses, this remains an area of investigation. Some claim that each gear stage is only 95% efficient. This is not something I have direct knowledge of so I'll leave it unchallenged.

Bob Wilson

 

<div class='quotetop'>QUOTE(usbseawolf2000 @ Mar 5 2007, 06:28 PM) [snapback]400609[/snapback]</div>

The data came from an ordinary drive and wasn't 'scripted.' But the MG2_rpm is recorded which lets us plot the mph.

NOTE: I updated the chart to show the mph over the same interval. Just refresh the browser to get the new graph.

. . . Given the sharp cut-off, I probably put the car in "N" to turn off all energy flows and perform an energy neutral glide.

<div class='quotetop'>QUOTE(usbseawolf2000 @ Mar 5 2007, 06:28 PM) [snapback]400609[/snapback]</div>

Google for "Development of New-Generation of Hybrid System" works for me. As for how to get into 're-circulate' mode, I have a four-line, Graham scanner configured for:

• MG2_Nm - positive sign for "normal" and negative for "re-circulate"
• MG1_Nm - positive sign to start ICE and negative for either "normal" or "re-circulate"
• Batt_I - indication of battery energy contribution or charging
• MG1_rpm - indirect indication of ICE rpm
• MG2_rpm - speed
• Batt_V - battery voltage for energy calculations

I can easily monitor the vehicle energy flows, in rough engineering units, while driving. So far, it looks like there are three acceleration modes to investigate: (1) normal, (2) MG2_Nm close to 0, and (3) MG2_Nm negative. Each is slower than the previous but I'll need to record injector timing to see how efficiently I can achieve a given kinetic energy . . . something for next month's posting.

Bob Wilson

 

Hi Gary,
<div class='quotetop'>QUOTE(MPG > HP @ Mar 5 2007, 11:02 PM) [snapback]400728[/snapback]</div>

With my older NHW11 and the Graham scanner, I have access to over 50, decoded OBD signals:

http://ecrostech.com/Products/MiniScanner/Intro.htm

It has been a terrific road-map to how the Prius systems work and their modular construction. One thing it has done, along with the Toyota Maintenance Manuals and Toyota SAE papers is show the modular construction and integration of the Prius. So think of the cruise control as just a device that tries to maintain a given speed and sends its inputs to the hybrid vehicle controller, the master mind. The hybrid vehicle controller manages everything going on.

<div class='quotetop'>QUOTE(MPG > HP @ Mar 5 2007, 11:02 PM) [snapback]400728[/snapback]</div>

There is not enough information to answer this question. It is a complex function dependent upon speed, grade and vehicle maintenance and one of the reasons why having a scanner is so important for getting quantitative data. At least with the NHW11, Toyota SAE papers and insights gained from this and other Prius technical forum, we're gaining a clue. Graham Davies, Ken, Hobbit and other early adopters have done seminal work and I've been lucky enough to build upon their efforts.

<div class='quotetop'>QUOTE(MPG > HP @ Mar 5 2007, 11:02 PM) [snapback]400728[/snapback]</div>

The closest I know of is Wayne Brown's fairly extensive Prius model:

http://privatenrg.com/

Some of my work has been building upon Wayne's earlier work with my own data. But if you have an interest in this subject, I'd also recommend reading the Miller and Miller paper, "Comparative Assessment of Hybrid Vehicle Power Split Transmissions" (Google for sources.) Ultimately, it comes down to using your Prius as an experimental device and doing the hard work of measuring what happens in well defined test environments. which has been the greatest teacher of us all.

BTW, the ICE was already running at the stop because it had yet to reach a temperature where the normal hybrid vehicle control laws could auto-stop the ICE. But the data had cleanly captured the transition from "normal" to "energy re-circulate" (aka., heretical) mode. Being visually oriented, I thought folks might enjoy seeing what happens.

Bob Wilson

ps. BTW, I had created the PSD graphic just a day or so before Florian Steiper's excellent transaxle disassembly photos. Also, Ken had asked about one of the earlier power flow schematics and that sent me back to the Toyota SAE papers. So together, we've advanced our understanding.

 

<div class='quotetop'>QUOTE(bgdrewsif @ Mar 6 2007, 04:11 AM) [snapback]400779[/snapback]</div>

Don't feel bad, it is not easy. It took seeing 'torque' and other data before I could get this far. So how far do you get before it makes no sense?

Understanding the Power Split Device, the planetary gears is the foundation. Do you see how the parts assemble and with the exception of the planetary gear carrier, freely rotate around the ICE power shaft? The planetary gears on the carrier freely rotate but on the inside is the sun gear operated by MG1 and the outside has MG2, the ring gear, that transfers power to the reduction gears, differential and wheels. So to get ICE power to the wheels via MG2, MG1 has to provide 'counter torque' to the planetary gears. This is how the 'split' part of the power split device works.

The largest part of the ICE torque passes through the ring gear and MG2 towards the wheels. The rest of the ICE torque passes to MG1, which works as a generator. The power of MG1 can then be routed to MG2 to recombine and pass on to the wheels. But in the meanwhile, the relative speed of MG1 and MG2 keeps the ICE operating in 'the sweet spot.'

Bob Wilson

 

Hi folks,

Graham Davies did some excellent work describing PSD operation:

http://prius.ecrostech.com/original/PriusFrames.htm

For example, Graham explained how to derive the rpm relationships:

MG1 = 3.6 * ICE - 2.6 * MG2

MG1 - MG1 rpm
ICE - ICE rpm
MG2 - MG2 rpm

MG1_torque = 28% * ICE_torque (again from Graham Davies)

Any two rpm values determine the third rpm. Also, the MG1_torque tells the ICE_torque. Using the Graham's mini-scanner data, it was easy to calculate the energy flows.

Bob Wilson

 

Thanks H*,
<div class='quotetop'>QUOTE(hobbit @ Mar 29 2007, 08:02 PM) [snapback]414585[/snapback]</div>

I recorded 35 miles of data with the throttle position and graphed the ICE power vs. throttle%:



There appears to be an unusual, high power, trend at high throttle settings to try and open the throttle plate more but without a significant power increase. But the interesting area is in the 25-35% throttle settings.

In the range of 25-35% throttle,obviously not "wide open," the higher power settings are occurring in energy recirculate (aka., "heretical") mode. This suggests MG1 torque enables higher ICE power output at any given throttle plate position:



This is counter to what I would have expected. I'll edit the first posting to reflect this new information.

Bob Wilson

 

<div class='quotetop'>QUOTE(okiebutnotfrommuskogee @ Mar 31 2007, 09:10 AM) [snapback]415433[/snapback]</div>

The planetary gears on the planetary gear carrier are spinning like mad because anytime the car is moving the ring gear is rotating. But with no counter-torque from the sun gear, the planetary gears have nothing to push against. The planetary gears spin but can not apply torque via their bushings to the planetary gear carrier and ICE shaft. So normal friction and compression drag lets the ICE sit there without spinning or transferring any force to slow the vehicle down. The only motions are the ring gear rotation spinning the planetary gears around their bushings and they in turn spin MG1 freely in the opposite direction. Only the planetary gear carrier, connected to the ICE shaft, remains immobile.

Bob Wilson

 

<div class='quotetop'>QUOTE(MPG > HP @ Apr 1 2007, 06:58 PM) [snapback]416177[/snapback]</div>

We use topographical maps from our test runs. To the best of my knowledge, there are no built-in altitude measurement devices or systems.

<div class='quotetop'>QUOTE(MPG > HP @ Apr 1 2007, 06:58 PM) [snapback]416177[/snapback]</div>

Toyota has released a number of engineering papers and several US agencies have done extensive analysis. The rest has been done by individuals trying to expand the envelope of their vehicles. You can't make a change without understanding what you have and how it works.

Bob Wilson