Thanks for the explanation, that does make sense. I'm going to have to look into one of these Graham miniscanners if they aren't obscenely expensive. Or maybe CAN view would work. But what you have me thinking is that we might be able to estimate the round-trip efficiency for a regen--->battery--->propulsion cycle (kinetic to kinetic) through a controlled test. The crude methodology might be to start at a known speed (and kinetic energy) below 41mph and a known battery state of charge, allow the vehicle to coast down to a lower speed (and KE) threshold while noting the new battery SoC, then accelerate under EV and note the speed attained when the battery reaches its starting SoC. It would be desireable (more like crucial) to estimate energy losses to things like rolling resistance and aero drag in order to isolate internal system efficiency. Even better to record battery voltage and current throughout the test series to help give an idea of the magnitude of losses in either direction.

I'm just rattling this off without doing a proper experimental design so no doubt there are holes to be filled.

Actually I'm impressed and pleased. One of the harder things in life is to find collaborators who understand technology and how to approach these kind of problems. When I see someone with "the lights are on and someone is home," I want to shout out, "Welcome fellow engineer!"

You might want to wander over to the Yahoogroup, "Prius_Technical_Stuff" and take a look at the files and search through some of the message archives. This will 'open your eyes' and help you see some of the really interesting experimental stuff that is going on. In particualar, take a peek at what "Hobbit" has been doing.

We'll still share results and summaries here. What we learn is important and can help less technically inclined folks 'get a clue.' But the folks in that group 'play with fire' and sometimes 'the smoke escapes.' We're a friendly group of geeks.

Bob Wilson

Operation Iraqi Oil Freedom:

Automatic, stock, project car.

My other 1500 cc car:

Automatic, stock, backup car.
Free speech, dialog and knowledge thrives without the poison of SPAM.

Sounds like a good place to hang out and talk shop, so to speak. I'll stop over to see what you guys are up to!

Never good when you let the smoke out of electronics.

Done it a few times, and they never seem to work the same again.

The genie who makes you wish you hadn't done that. <grins>

Bob Wilson

Operation Iraqi Oil Freedom:

Automatic, stock, project car.

My other 1500 cc car:

Automatic, stock, backup car.
Free speech, dialog and knowledge thrives without the poison of SPAM.

They aren't obscenely expensive (IMHO), but they are not available for the third generation (2004+) Prius.

Jim

Lifetime fuel mileage:

After learning how to hypermile:

To simplify that first graph Bob posted, he's driving 30mph consistently - no change in speed or accelleration - yet the gas engine is turning on, and off, over and over again. While it's on it's pushing the car and any extra energy goes to the battery. Then the Prius computer shuts off the gas and drives on electric, then back to gas again. All this happens without you doing anything special - you just drive the car.

No - yellow arrow bad. Green arrow good. If you see a red arrow, you don't want a yellow one coming from the battery. Below about 15mph though, it is better to have just a yellow arrow from the battery and no red arrow (accellerating on electric-only because it's more efficient at very low speeds).

More or less. If I Pulse and Glide below 42mph I get my best mileage. If I'm going to be above 42 I do better at 60-85mph mileage-wise. But the more important factor in all this is maintaining a consistent speed whenever you can - accellerate to the speed you want and maintain it. So for example, accellerating constantly until you see a red light, then decelerating to the light, means at no point were you going a constant speed. That's going to get you bad mileage. That's also how nearly everyone in the US drives.

I originally thought this was an issue too - but it isn't. It turns out starting the gas engine on the Prius is really, really efficient. In fact, if you were travelling at 45 the Prius might be charging the battery, and at 39 it might turn the gas engine off because the battery has some extra charge - you might do very well mileage-wise driving that way.

One last note on temperature. When it's 70F or above here I find pulsing and gliding on local streets gets me fantastic mileage. When it's 40F here (lately) I actually do better on the highway because the gas engine won't stop running to keep the car warm. At least on the highway there are no stop lights and I can drive a mostly consistent speed. So to simplify, if it's warm out, driving under 42mph will probably get you your best results. If it's cold out, 65mph is your ticket.

Prius 2006, bought September 14:

OK, now I'm confused. I thought my better option (not necessarily best option) was in those cases where the ICE is running (and gliding is not an option), having a yellow arrow to the battery would produce better FE?

The idea being the the excess energy produced by the ICE is partially going back to re-charge the battery (as opposed to simply being wasted as excess consumption). In practical terms, when my ICE is running (sending an orange arrow to the wheels) I will either have a yellow arrow running from the battery to the EM or a yellow arrow to the battery. My original take from the thread is that the yellow to the battery is the preferred option in this scenario.

That's actually good news as about half to 2/3 of my driving is highway based and its been under 30F here for several weeks now. I've have noticed I get much improved FE when I am running at highway speeds and P&G from 80-65 (which I have the option to do quite frequently as I am the only one in the right hand lane). It also helps due to the number of hills I have to traverse each day during my commute. I can anticipate my glides up the inclines and then use the following decline to get my speed up again without over working the ICE.

UTpiper
Likely never to be a hypermiler but desperately trying to keep my average FE above 50 mpg during the winter.

Sort of - I think you just have the arrow colors confused but get it just fine. If you accellerate gently a red arrow comes from the engine to the wheels, and a green arrow comes from the EM to the battery. If you see a yellow arrow between the EM and the battery, it's going to be pointing the other way - at the EM - and that means you're harming your efficiency.

Pulsing and Gliding is very effective under 42mph because you can turn the gas engine off. I'm not sure that Pulsing and Gliding is effective at high speeds, because the gas engine is running constantly. I have the most luck above 60mph by accellerating gradually to my target speed (let's say 75mph) and maintaining it. Others with more knowledge of the Prius' workings can probably better comment on the efficiency of P&G at high speeds though.

Prius 2006, bought September 14:

I don't have that much more knowledge about high-speed modes and it is something still under discussion. There are some who claim to regularly achieve a high-speed glide and that it can be used to improve high-speed MPG. I just don't do enough high-speed driving to tell.

My data indicates there are two high-speed glides:

A) partial ICE power - effectively no wheel torque, ICE running
B) battery-assisted, ICE - no wheel torque but some MG1 current to keep the ICE turning over

Although "B" has no fuel burn, it is followed by ICE powered recharging of the battery. The "A", partial ICE power mode, does not have to put a charge on the battery. What I don't have is a total energy model that indicates whether mode "A" or "B" has a lower fuel consumption over a long distance trip.

The only key I've found is avoiding running the ICE above 2,500 rpm and especially avoiding 3,500 and above. My data shows a fall-off in specific fuel consumption as the ICE rpm approaches and exceeds 3,500 rpm. This is not a function of 'glide' or 'pulse' but what energy is being requested for the current load. IMHO, the best strategy would be a cruise control that would also back-off the speed setting to keep the ICE below some given rpm:

2,500 rpm - very green but slow
3,200 rpm - pretty green but lethargic up hills
3,500 rpm - dull green

The data indicates this would provide a directly measurable cruise improvement. Notice that speed has nothing to do with it, ICE rpm only.

Bob Wilson

Operation Iraqi Oil Freedom:

Automatic, stock, project car.

My other 1500 cc car:

Automatic, stock, backup car.
Free speech, dialog and knowledge thrives without the poison of SPAM.