Would double battery = double mileage?

[Frank Lee]

^wrong

[bwilson4web]

Originally Posted by codenamezero

. . . just get a hybrid car and just run it on battery, because a hybrid already has a gasoline engine built in, that is essentially your "generator trailer". . . .

Uh, this isn't quite accurate even though it is 'common wisdom.'

I have only studied the Prius in detail and the Honda IMA in passing so please understand the limits:

• Optimum BSFC - the traction battery and motors allow the Prius (more or less) and the Honda IMA systems to optimize their BSFC.
• EV only - somewhat concentrated in the Prius, the battery allows the ICE to 'shut the f*ck up' when modest power is needed. In the Prius world, we seek routes and paths that allow us to travel under the hybrid threshold speed, 42 mph, or 46 mph in the ZVW30.
• efficient CVT - although I can explain it, it takes a little more time. Most CVTs are not nearly as efficient as this Toyota invention but it has to do with the power split device.

Adding a motor and battery or even a generator would lead to poor performance. It is the integrated system of ICE and transaxle that gives the Prius unmatched efficiency. It is what is missing in the speculations and honorable competition.

Bob Wilson

[dcb]

Originally Posted by bwilson4web

It is the integrated system of ICE and transaxle that gives the Prius unmatched efficiency. It is what is missing in the speculations and honorable competition.

gonna chime in here, I think it is more a function of the right component parts for the job. A "right sized" atkinson engine pushing a bev with the bev drawing or adding as most efficient demands (not clear if recharging from gas ever make sense efficiency wise except to not kill the battery completely) would do fine with the right control scheme, there's always compromises.

The prius transmission conversions when it needs additional electric (which is fair bit of the time AFAICT to keep all the RPMs right) can get as bad as %70 per this article, though I are not a prius expert/marketer like bob, they may have added more gizmos to help out here
http://www.ae.pwr.wroc.pl/filez/2011...HEV_Toyota.pdf

"Also, the efficiency of the transmission is heavily dependent on the amount of power being transmitted over the electrical path. The latter has multiple conversions, with a low efficiency (70%) compared with the efficiency of the purely mechanical path (98%). Especially in higher speed regimes (>120 km/h) the efficiency (of the transmission alone) therefore drops below that of a generic automatic transmission with hydrodynamic coupler."

so I would probably like to at least try a stick shift version with a battery/generator disconnect for long boring hiway cruises

[dcb]

also note, that I think there is always some mechanical contribution (not sure) so that 70% might not be the lowest efficiency that the generator->controller->motor path sees.

[bwilson4web]

Thanks,

Originally Posted by dcb

. . .
The prius transmission conversions when it needs additional electric (which is fair bit of the time AFAICT to keep all the RPMs right) can get as bad as %70 per this article, though I are not a prius expert/marketer like bob, they may have added more gizmos to help out here
http://www.ae.pwr.wroc.pl/filez/2011...HEV_Toyota.pdf

"Also, the efficiency of the transmission is heavily dependent on the amount of power being transmitted over the electrical path. The latter has multiple conversions, with a low efficiency (70%) compared with the efficiency of the purely mechanical path (98%). Especially in higher speed regimes (>120 km/h [75 mph rjw]) the efficiency (of the transmission alone) therefore drops below that of a generic automatic transmission with hydrodynamic coupler."
. . .

Please include professor Raf Catthoor's authorship when sharing this paper even though it is in the PDF attribution.

Tracing back via the URL, it looks like this comes from professor Raf Catthoor's lectures:
Automotive Engeneering.
(NOTE: translations from Polish to English sometimes lead to interesting spellings.)

Yet this is what we find when we actually go out and measure NHW11 and NHW20 mileage:


Notice how the NHW11 mileage drop-off matches Catthoor's expectations but fails to predict the NHW20 mileage. There were significant changes between the NHW11 and NHW20 transaxles that the good professor missed:

1. Field-weakening replaced by higher voltages
2. Efficiency calculations need backup

Understanding how the Prius transaxle works is not easy and there are many subtle aspects that can trip up the unwary.

Field-Weakening

This was first described in "Development of Electric Motors for the TOYOTA Hybrid Vehicle "PRIUS"" Kazuaki Shingo, Kaoru Kubo, Toshiaki Katsu, Yuji Hata, Toyota Motor Corporation where they left this gem:

Originally Posted by above

. . . In the flux-weakening control range, motor efficiency is lower than at the maximum torque outputting current phase angle. . . .

What happens is at higher rotation speeds, these electric motors generate a back-EMF, a voltage that opposes the power current flow. So Toyota uses a technique they call "field-weakening" which applies power outside of the optimum angle before the back-EMF is high enough to block the power current.

To solve this problem in the NHW20, Toyota uses voltage doubling to the motors. The motor back-EMF is still there at high speeds but the higher voltage from the inverter means it does not have to use the inefficient, field weakening until much higher speeds.

In the meanwhile, the NHW11 still uses the straight traction battery voltage and at high speeds, has to use field-weakening. This is why above 65 mph (104 km/h) the mileage begins to drop off:

• 52 MPG @65 mph (104 km/h)
• 39 MPG @75 mph (120 km/h)

Observations agree that the older NHW11 has an efficiency fall-off but Prof. Catthoor missed the primary mechanism, field-weakening. There are supporting aspects that are seen in the data such as:

• ICE in less efficient power region - at higher speeds, the ICE suffers an efficiency hit and in some cases, uses fuel enrichment to keep the exhaust temperatures down.
• MG1 speed limit - nominally 6,000 rpm in the NHW11, it was increased to 10,000 rpm in the NHW20. This means as the vehicle speed increases over 65 mph, the engine has to spin faster than idle, ~980 rpm, to prevent overspeed. A faster turning engine is less efficient. In contrast, the NHW20 goes out to 85 mph before it hits this limit.

Only the NHW11 suffers from this high-speed, efficiency hit. The NHW20 and NHW30 don't or at least don't until speeds approach 85 mph (~136 km/h.) It is inaccurate and misleading to make this claim against all Prius.

Efficiency calculations: "98%" mechanical, "70%" electrical

It helps to measure efficiency such as "Report On Toyota/Prius Motor Torque Capability, Torque Property, No-Load Back EMF, and Mechanical Losses" J.S. Hsu, C.W. Ayers, C.L. Coomer, R.H. Wiles, Oak Ridge National Laboratory, S.L. Campbell, K.T. Lowe, R.T. Michelhaugh, Oat Ridge Institute for Science and Education, ORNL/TM-2004/185.

Originally Posted by pp_13

The gear system is a source of mechanical losses. At 6000 rpm, with a 30-kW load, the total mechanical loss (excluding the electromagnetic losses) of the motor and its gears is 7.4%. . . .

So the direct, mechanical efficiency is 92.6%, not the "98%" claimed by professor Raf Catthoor. But I think I know where the error came from.

As a general rule of thumb, mechanical gears are treated as a 2% loss per set of teeth, a rule of thumb I've seen in the past. One formal paper (I don't have handy but can find if someone needs a reference,) points out the gear stage loss occurs in the interaction of the teeth. Primary losses are hydraulic by squeezing out the oil followed by hysteresis from stressing the teeth. The good professor failed to count the gear stages which are additive.

Now the good professor states:

The latter [electrical path rjw] has multiple conversions, with a low efficiency (70%) . . .

"Evaluationg of 2004 Toyota Prius Hybrid Electric Drive System", R.H. Staunton, C.W. Ayers, L.D. Marlino, J.N Chiasson, T.A. Burress, ORNL/TM-2006/423, Appendix B, pp. 69-85 of Oak Ridge Labs report lists transaxle efficiencies in multiple speeds. There is no single efficiency value because it is a function of multiple variables and the efficiency tables range from:

• 55.5% - lowest found
• 92.7% - highest

In reality, the electrical path efficiency is a function of operational conditions and to claim a single number "70%" is at best misleading and at worse, leads the ignorant to false claims and conclusions. For example, it would be accurate but misleading to 'cherry pick' lab testing data and claim measured Prius transaxle efficiency is:

• 92.6% - mechanical
• 92.7% - electrical
• 92.65% - average
• 92% - four-stage, manual transmission and differential losses

Here, 'cherry picking' data makes the Prius more efficient than a manual transmission but I'm not making this claim.

CONCLUSION

Professor Raf Catthoor made a reasonable attempt to understand Prius operation but field, lab testing and the Toyota papers provide details and insights he missed. Sad to say, his data points, "98%" mechanical, and "70%" electrical do not reflect "in real life" and leads the unwary astray.

Bob Wilson

[dcb]

Originally Posted by bwilson4web

Thanks,
Please include professor Raf Catthoor's authorship when sharing this paper even though it is in the PDF attribution.

Please cram the suggestion that I treat this link special with walnuts. That has to be the weirdest request I have heard yet, where is your head at? Standard forum etiquette is to provide a link plus some verbiage for context, which is precisely what I had done.

Originally Posted by bwilson4web

and leads the unwary astray.

I think you just did. My point was that there are losses when you use an engine to drive a generator to drive a motor to drive the wheels, and the prius has to do this very thing at higher speeds. Boost converters lose efficiency when they have to increase the output ratio also. You made the claim "unmatched efficiency" when, as was apparently missed in your attempt to sell more priuses, but is still supported by " * 55.5% - lowest found
* 92.7% - highest" and if 92.7% is zero electrical then 70% sounds like a reasonable swag to me.

There are plenty of operating conditions where direct drive, plus right sized atkinson will beat a priuses pants off for efficiency, especially in the high speed arena. I appreciate your love for data, but that does not mean you are completely unbiased when it comes to the prius or when choosing your words to describe it.

I'm afraid you have not supported your claim that "Adding a motor and battery or even a generator would lead to poor performance." or "unmatched efficiency". In an honorable competition you would not compare a prius to a $500 generator-BEV come hilbilly series hybrid. To determine if it is unmatched, I would pit you against a driver with the tiniest mote of education (especially in comparison to all the useless by comparison information rabbit holes various prius versions and authors and sycophants will lay in front of you) in a similiar CDA BEV using an atkinson parallel range extender with basically a stick shift. Feel free to add an efficient generator too if it makes sense, but especially for constant speed vs mpg readings, you still want to have the right sized engine transferring power to the wheels with minimal conversions and losses. This basic fundamental fact has not changed, conversions cost efficiency. If you want to drive 80 then choose your engine wisely, or choose "multiple" smaller engines or see what you can do with electronic valves, and take up the slack with the electric motor. I don't know what conservationists are saying driving 80 is a swell idea though.

[bwilson4web]

Originally Posted by dcb

Please cram the suggestion that I treat this link special with walnuts. That has to be the weirdest request I have heard yet, where is your head at? Standard forum etiquette is to provide a link plus some verbiage for context, which is precisely what I had done.

What happened is the paper does not include a title nor author in the body. So it took a little digging to find out this is Polish professor Raf Catthoor's lecture materials. So then the common EU 'opinions' made sense as well as the simplistic analysis.

Originally Posted by dcb

. . .
I think you just did. My point was that there are losses when you use an engine to drive a generator to drive a motor to drive the wheels, and the prius has to do this very thing at higher speeds. Boost converters lose efficiency when they have to increase the output ratio also. . . .

I held off on discussing how the Prius transaxle works since my critique was about the good professor's work. He didn't have measurements to backup his claims yet. The papers and metrics I provided filled in what he is missing and sad to say, reality is at odds from his claims. A cautious person might seek copies of my cited papers to gain understanding.

Originally Posted by dcb

. . .
You made the claim "unmatched efficiency" when, as was apparently missed in your attempt to sell more priuses, but is still supported by " * 55.5% - lowest found
* 92.7% - highest" and if 92.7% is zero electrical then 70% sounds like a reasonable swag to me.

There are plenty of operating conditions where direct drive, plus right sized atkinson will beat a priuses pants off for efficiency, especially in the high speed arena. I appreciate your love for data, but that does not mean you are completely unbiased when it comes to the prius or when choosing your words to describe it.

I'm biased by my direct observations and the papers previously referenced that have full attribution.

Originally Posted by dcb

. . .
I'm afraid you have not supported . . .

That is OK. There is no requirement for understanding what the Prius is doing. Experimentation is good and I hope folks are giving these alternates a good try. But it would be helpful to review some basic principles of operation that professor Raf Catthoor didn't cover very well.

Professor Raf Catthoor correctly identified that there are two power paths from the power split device:

He was also correct that the relative torque is fixed, 28% going the electrical path and 72% the mechanical. But both torque paths are taken at the same angular rotation speed from the ICE and the power split is proportional. This has a profound impact on efficiency calculations:

• 28% electrical * electrical_efficiency (55-97%) = electrical power path
• 72% mechanical * 92.7% (measured) = mechanical power path

In normal mode, electrical power from MG1 is routed directly to MG2 where it joins back with the mechanical power:

Counter torque is needed from the generator to get any mechanical torque. This is done by generating power that simply recombines in MG2.

There is another mode often called "heretical" where the roles are reversed. The car is traveling fast enough that MG2 can generate the power needed to drive MG1 to force mechanical path torque:

The electrical path direction is reversed. But conservation of energy leads to an interesting observation:

MG2_power = MG1_power + traction_battery_power

We see this in the data:

It has the effect of overdrive by letting the ICE spin slower.

What I'm trying to suggest is what came out of the Toyota paper:

The whole system needs to be tweaked and tuned. There is no single fix that will provide Prius efficiency.

Bob Wilson

[dcb]

I do not hope to untangle that for anyone else, but I maintain that a driver with little more than stick shift skills and similiar equipment can still outperform the prius, precisely because the power split device requires power and conversions to hold various RPMs. The prius efficiency is compromised for convenience and thus is NOT "unmatched efficiency".

Also speaking of leading astray, a picture that shows all those conversions plus 100% ice going to the wheels is doing just that.

Not happy with the constant and misleading in their own way prius commercials

[dcb]

let me also point out an apparent discrepancy here (not the only discrepancy I have seen), even though this is probably the wrong thread

from seconds 335 to 355 you are indicating an increase in speed from maybe 43 mph to 49 mph. At the same time you are indicating the ICE wattage going from a peak of ~13.5kw down to ~10kw. The battery isn't doing anything terribly interesting either at that time, so I have a strong suspicion that you are doing some "double booking" when calculating ICE power. at 355 you are taking something over 10hp from the driveline to generate about 10hp with mg2 to pump ~7hp (conditioning efficiency of 70% prior to mechanical conversion?) into mg1. Meanwhile speed increased and ICE decreased. It doesn't seem to add up.

[ShadeTreeMech]

Y'all are barking up the wrong tree and ignoring something important.

The reason to have an electric vehicle is to make use of electricity from the power grid, which is produced very efficiently, but as I understand it, is also subsidized to allow poor people the ability to afford it. Producing electricity will always incur energy losses as per the law of conservation of energy. This law states energy cannot be created nor destroyed so no matter how electricity is produced, some energy must be lost.

According to the wiki on the nissan leaf, using a 120 volt charger connection will consume 1.4 kWh to add 5 miles to the range. Using the 220 volt charger would add 11.7 miles to the range and consume 3.3 kWh. These amounts of electricity are easily produced by a portable generator, but add very little to the range. And this is dealing with a highly efficient electric car. There is a generator that can easily produce that power for the cost of .3 gallon per hour, which would give you a mpg rating of 33 mpg at a speed of 11.7 mph. Here is a picture of it.