Warming up engine and batteries downhill

I don't think the batteries would need to be warmed up, they will do that themselves when they charge or discharge. The ICE, as you mention, does need run to warm up the catalytic converter for the first 50 seconds or so when cold. It runs during that time without providing any power to the wheels or the battery. So durning that 50 second warmup Ken@japan has suggested that it might be better to just set in the driveway and let it warm up. That would save the conversion loss encountered when running on the battery then recharging it.

The exception, of course, is if you live at the top of a hill. So in your case, when the car is cold, I would just leave it in ECO, or whichever mode you use,and proceed down the hill. You would be warming up the engine and putting some Amps in the battery at the same time. That's something most of us can't do.

 

You may want to try an differential test:

• EV to same hill starting point
• Hit "D" to let engine start
• Roll down hill in "D" holding speed constant with brake
• Roll down hill in "N" holding speed constant with brake

With my 2003 Prius, rolling down the hill in "N" with the engine in idle, it continues the warm-up and burns significantly less fuel. The 'loss' of regeneration energy pales compared to the fuel savings.

Bob Wilson

 

I accept your results, but now I wonder why? Was the engine cold when you tried your tests? could the ZVW30 be different in that respect?

When the car is in the 50 second warm up from cold mode (stage 1a) it runs with retarded timing and provides no power whatever to anything that I have discovered. When I drive in stage 1a using D (in any driving mode) all the power for the wheels comes from the battery, unless you really put your foot down.

I tried this morning with the ICE cold, stage 1a, and shifted from P to D to N and there is no difference, RPM 1280, timing ATDC by some amount determined apparently by coolant temperature. I don't live on a hill so I can't do the cold start downhill test.

 

This is something I've tried to replicate with our ZVW30 but I haven't tried my small, hill, yet:

My hill is not as long as I would like. I've tried some coasting tests the other direction, on the level parts of my street but I've not seen a significant result. So I'll be testing my hill too . . . or at least try.

I can tell that "N" during the NHW11 warm-up reduces the fuel burn. I just have yet to see the same thing in our ZVW30.

Bob Wilson

 

I don't think the ICE provides zero power during 1a. It provides a low amount of power to the wheels and/or battery. If you sit still, the ICE is spinning mg1 and putting its little power into battery. If moving, its small output is split between the wheels (MG2) and battery (MG1) based on the fixed connections within the 3 devices in the PSD.

All Ken is saying is that it cannot put out enough power to be efficient. There is no denying that the ICE is spinning, so the gear in the psd is being turned. Therefore, work is being done.

So I would stay out of the throttle, leave it in D, and enjoy getting to move during 1a without throwing tons of fuel into it.

 

I just went and looked and ICE power during 1a (in park) increases slowly from 1.5 to 2.5 kW. This is way down on the efficiency curve, but it is still producing power and torque (7-10 ft-lbs) which is presumably going to the battery through MG1 when it isn't moving. I doubt if it really matters whether you sit in the driveway and use your ~2 kW to charge the battery, or drive along sending your 2 kW to the battery and the wheels through the PSD. Either way it is horribly inefficient, but part of the warm-up cycle you really can't do anything about.

 

Shift into "N" and the ICE will have no MG1/MG2 counter torque. However, this also defeats the ability to stop the ICE when it is warm enough to stop. Regardless, a similar pattern is in the NHW11 data:

You can see the effect of the energy taken from the ICE by MG1 torque.

Bob Wilson

 

The ICE starts to warm up the cat. It has nothing to do with the engine or battery. Unless it's winter. In that case, the ICE will be run if heat is "demanded" by the climate control system.

Why warm up the cat? If the cat is cold and you try to accelerate, you will release a lot of pollutants. If you just run the ICE without trying to generate a lot of power (not accelerating) you only release a small amount of pollutants. It takes about 10-15 sec. to warm up the cat. The ICE will then shut down if it isn't needed.

Many of us think Prius was designed to get the best mileage. That is wrong. Prius is designed to generate as few pollutants as possible, THEN get good mileage.

 

I've started testing hill-rolling warm-up. The trick is to use EV to get on the hill downgrade; start the ICE and; test roll-down in "D" and "N". Curious, I had a screw-up this afternoon trying to run a test.

I used EV to creep out and on the street but by accident, went over 9 mph and the ICE started. I stopped quickly and hit EV which shutdown the ICE with less than 5 seconds of running time. Then I used EV again and reached the downgrade without triggering the ICE. So I slipped it into "D" and started rolling down the hill . . . the ICE did not start. I did see a slower roll-down speed due to a slight regeneration current.

My Auto Enginuity is able to record some useful data. I'll see about capturing some metrics.

Bob Wilson

 

That is really interesting. You inspired me to go back and measure the ICE power during the warmup cycle and log the data. I also tried to replicate your Park vs Neutral data. However, interestingly, with the car in Neutral, the ICE never starts until I put it Park (or drive), then I can put it in Neutral. Even more interestingly, there is a PID (#692) in the Engine ECU (234) that gives a boolean value for "Request Warmup". I've put this on some of the plots below. For the "Park" condition, the "Request Warmup" exactly indicates the stage 1a time from ICE start to ICE stop. For the "Neutral" condition, the request warmup indicator is TRUE at the same time as the Park condition, i.e. about 8 seconds after the power button is pushed, but the ICE does not start the stage 1a warmup until the car is actually shifted into park or drive. Weird.

Anyway, here are some graphs of the data I took. This is for a GenIII (ZVW30) manufactured Oct. '09.

The first is of the ICE power (in kW) with the car in Park. The ICE runs but it is generating very little torque. The ICE torque is on the right axis and it peaks around 10 ft-lbs. The spike of negative torque (and thus negative ICE power) at the beginning of the warmup cycle is MG1 spinning up the ICE to start. It peaks at about -100 ft-lbs.



The 2nd graph is the MAF (Manifold air flow) for both the Park and Neutral conditions. This is similar to Bob's injection-time as it is related to the amount of fuel consumed. As in Bob's results, the Neutral condition uses less Fuel/time, but stage 1a lasts longer, and the ICE doesn't shut down after stage 1a. The MAF data doesn't account for changes in enrichment so it isn't an absolute measure of fuel consumption, although after the engine is warmup up the canonical air/fuel ratio of 14.7 almost exactly reproduces the instantaneous fuel gauge on the Gen III MFD, at least as far as I can tell by eye. Note that I've also plotted the parameter "Request Warmup" for the Park condition so you can see how well it corresponds to the ICE running. The "Request Warmup" for the "Neutral" condition isn't plotted but it corresponds exactly to the fall in the MAF at around 90 seconds.



The 3rd graph is the battery state-of-charge for both the Park and the Neutral conditions just to show that the battery charges in Park but not in Neutral. This was, of course, expected. The Warmup Request for Park is plotted again.



The 4th graph is the RPM data.



The 5th is interesting. This is a parameter entitled "injector volume (ml) Cylinder 1" (ECU 234, PID 115). I have no idea if this is ml/injection or ml/time. I can probably figure it out by knowing the cylinder volume and the canonical air/fuel ratio. I'll work on it. However, it looks a lot like Bob's injection time plot, as it should…



The 6th plot I had hoped would be the same as Bob's plot. I used a parameter called "Injection Time (D4)" (ECU 234, PID 116). I have no idea what the "(D4)" means and the units are quoted as "81s" which I also don't understand. However, comparing the values to Bob's chart it would make sense if the units were in 0.1 ms. At the moment this is a mystery



Okay that's enough fun for now. Note that all the data were taken with a DashDaq mounted on my dashboard and written to SD card. The data were plotted in IGOR on a Mac.

 

It looks like you managed to cancel the warmup request with the EV button. Was this with your GenII or GenIII? I'm guessing it is on your GenII. Is it repeatable? I wonder if it works on a GenIII...?

 

I think it's ml/sec. If it were ml/injection, it would be 0.15x700=105ml/min per injector at 1400 rpm.

 

Yes, I came to the same conclusion as follows:

From the MAF data:

MAF = 0.85 lbs/min of air
MAF/cylinder = 0.21 lbs/min = 95 g/min of air
air/fuel ratio = 14.7
gasoline density = 0.72 g/cc
==> fuel consumption from MAF = 9.0 cc/min = 0.15 cc/sec

Which is precisely what the injector volume PID is giving. That is just great since now we have a direct measure of fuel consumption and I can also calculate the enrichment from the MAF data and the injector volume! This is great!

Now we can integrate the data in graph 5 and get the total fuel consumption for stage 1a in Park and Neutral:

Park = 34 cc (for all 4 cylinders)
Neutral = 24 cc

This is what it takes to heat up the catalytic converter, and in the first case to charge the traction battery by ~0.8%. Cool!

 

Our ZVW30, we don't have the NHW20.

The only way to kill the NHW11 warm-up is to spoof 70C once the engine coolant reaches 40C.

Bob Wilson

 

Sorry if this is a dumb question: why are these warm-up ideas not akin to starting the car in the am and letting it sit for a bit before driving off ?

 

You may not like the answer:

• Yes - warm-up in "P" is a common practice encouraged for those who must climb a hill in the morning
• No - a trick those who can descend a grade can use in the morning

But it has a practical application for PHEV or those with some EV capability. They can use the EV or hill potential energy . . . possibly . . . to let the ICE warm-up with a minimum, mechanical energy load. The elapsed time to reach 70C (or 40C depending upon warm-up cycle) won't change much. However, the fuel saved can make a measurable improvement in overall mileage.

Bob Wilson

 

Bob,

Warming up the ICE with electric or potential energy makes fine sense to me. I thought running the ICE under low load was the discussion topic.

I didn't follow the yes/no part of your comment. Yes to what, No to what ?

 

Sorry, I did not mean to be so obtuse:

• If you live at the bottom of a hill or a situation that requires early application of significant power, warm up the Prius in the driveway, ~60 seconds, to auto-stop. This avoids inefficient ICE power modes because the catalytic converter O{2} sensor feedback is not there or the coolant is below 40C which causes the ICE to run inefficiently to generate heat. If you know how long it takes to normally warm-up, you could shift into "N" and reduce the fuel burn yet still warm the engine within seconds of how long it takes in "P" or "D".
• In contrast, if you live a the top of down grade, especially a gentle down grade, try to pull out in EV mode (to keep ICE off) and once you get the gravity assist, hit "D" or go faster than 9 mph to trigger the ICE to start up. If conditions permit (it is illegal in some states,) shift into "N" and let the ICE warm-up in minimum fuel consumption mode.

Anytime the engine is running and in "P", "D", or "B", it is going to burn more fuel per second than if it is in "N". However, the heating effect of being in gear is nearly unmeasurable while the fuel consumption is up ~30-50%. So if you have to warm up the ICE, try to use either "N" or some 'no load' driving mode.

Bob Wilson