Prius, better hiway milage via brake band MG1?

[P-hack]

Hi EM!

I recently picked up a 2004 prius, interesting car, I am used to small stick shift cars (and P&G and etc).

I was trying to get my head around it and reading all the online opinions, including varying opinions about driveline efficiency.

referring to:
eahart.com/prius/psd/

It seems apparent that it is basically a series hybrid at 0mph, and that even at higher speeds it relies on generator torque to drive the wheels directly and thus is a balance of series/parallel.

So, the question is, can hiway efficiency be improved by not relying on the electrical transmission of torque? I.e. add a brake band to MG1?

Example:
Lets say ICE bsfc peak is at 3000 rpm
MG1 is braked at 0 rpm
MG2 is at 73mph/4300rpm

I am not sure how many hp you need at 73 mph, 35? maybe the engine would be 50% loaded? so you go from 230g/kwh to 265? (%15 difference), so maybe go a little faster?

But the idea is that with mg1 physically locked, it becomes essentially direct drive (with mg2 "freewheeling") and no xfer of power.


And on the other end, if you could put a brake band on the ICE, would it reduce the electrical demands when trying to drive 41mph on electrical only.


Note, I am not familiar with all the software/control nuances, so this is just a high level question about the drivetrain hardware (before I contemplate hacking the crap out of my prius) Perhaps there are already brakes in place, or the holding/opposing torque current demands are not large.

Thanks.

[Daox]

AFAIK there are no brakes. I'm sure that could be done and you would see a small efficiency increase. Doing so would be quite complicated and controlling the whole mess would be a nightmare though.

[P-hack]

re: 41mph, ok it doesn't look like it matters if mg1 is not helping to drive the vehicle. So an ICE brake could add a couple hp by allowing both electric motors to push the car forward in EV mode (unless I completely don't get it), but otherwise not help efficiency.

[P-hack]

Well, control *should* be as "simple" as disabling the motors (ensure mg2 can freewheel ) and activating the mg1 brake, at 70+mph.

Probably a more definitive answer about existing driveline losses at those speeds/loads would be prudent first, somehow.

[P-hack]

curious, looks like they did something like this on the volt, lock up the ring gear, for EV mode though, not direct drive long distance hiway on gas mode though.

http://image.motortrend.com/f/348033...lustration.jpg

[sheepdog 44]

I've been told that the electric torque to lock up one of the MG's is so ridiculousy efficient it has just about no effect on mpgs. Someone locked up a freestanding transmission electrically with no electrical components attached. Just set it on a bench and attach a hand crank as a demonstration of regen braking. Once fully locked up, the strongest man there could not turn the damn thing! I had no chance at all beyond driving the regen braking part. There were no physical connections to lock it up, just the force between magnet and coil, no batteries attached.

A physical brake would net near zero mpg increase, and i may be mistaken, but hinder the electrical side of the prius.

A better approach to get more mpg's would be to try and make pulse and gliding easier. From what i can gather, it's a very precise skill to learn and limited to certain speeds.

[mort]

Hi P-hack,
It's a good idea, but you need a transmission in this car, the Atkinson engine operates almost full throttle all the time it is producing appreciable power.

Originally Posted by P-hack

I am not sure how many hp you need at 73 mph, 35? maybe the engine would be 50% loaded? so you go from 230g/kwh to 265? (%15 difference), so maybe go a little faster?

The engine is never 50% loaded. Power is regulated by changing rpm. If you are zipping along at 60 mph, using 15 hp, which you can get at about 1500 rpm, then hit a hill and need 50 hp the engine must rev up to 3700 rpm.

But, this report: A fundamental case study on the Prius and IMA drivetrain concepts from Technische Universiteit Eindhoven (about 2004) shows that MG1 has pretty poor efficiency at low rpm (or locked) and high torques. It looks like 50% efficiency at locked rotor and 80 Nm (60 ft-lb)

So I dunno what to do about that.

-mort

[P-hack]

Yah I am assuming you could engage the mg1 brake after you get up to cruising speed, and that it would be useable under some range of conditions (mostly long flat-ish stretches, though at %50 it has reserve capacity for some hills, plus possibly mg2 assist), and that the control headaches could be sorted (i.e. allow ice to run at %50 and whatever electrical magic).

That link looks useful though for some armchair engineering.

[Cobb]

Are you talking about stopping mg2 to prevent "background charging? Many insight overs find that reduced fuel economy and happens with a bad ima as well as a starter battery. The drag alone is enough to prevent lean burn.

My thinking is since its a ECVT, you would get your best mpg at the top speed before it starts to "shift", or charge ratios.

[P-hack]

Hmm, I am thinking that by locking mg1 (the sun gear) to the case at speed, that the engine basically drives the wheels directly, and let mg2 freewheel (if that makes sense on a PM motor). Eliminating the losses for that portion of the torque that is getting converted to electrical and back to mechanical, as well as holding current for mg1. It is a controller nightmare in reality, but conceptually it is simple enough. If the inverter doesn't care at speed when the car is off, then you could lock mg1 and switch over to a megasquirt+ for the ICE.

Yah, a pedal controller sounds like a good low hanging hack too (via some sensor taps and cruise control actuator?), will have to look into that in a different thread.