Optimal RPM for Gen3 Engine?

Optimal for what, engine longevity, mpg, efficiency of acceleration, all of those?

 

The 1.8 liter (Gen 3) is the one on the right:



When you've got a transmission with fixed gears, you kind of move around in the 2-D space of that graph; you might be running the engine at higher or lower RPM than is optimal for the power you need at that moment, depending on what gear you're in.

When you have a CVT being controlled by the car's power management control ECU, as in a Prius, that doesn't happen as much. The ECU knows that heavy-line path that goes through all the RPMs that are optimal for different levels of torque demand, effectively making a 2-D region 1-D; the ECU just kind of moves the operating point along that path, using the CVT to match the wheel speed to the RPM best suited to the current power need from the engine. So the car is already doing a lot of that thinking for you.

 

I'm remembering Hobbit's work as being primarily on the Gen2, with the 1.5L engine. The BSFC (brake specific fuel consumption) graph in Chap's post here shows that the Gen3's 1.8L engine has a greatly expanded 'optimum' RPM zone, with anything from near idle to 4000+ RPM being close enough.

 

It has a greatly expanded 230 g/kWh area, but also has a little zone within that where the efficiency's even better.

 

Even that new expanded zone is fairly wide, from ~1200 to ~3000 RPM, which easily covers the bulk of my driving.

 

That's right.

 

I invariably have the horizontal bar graph of fuel economy showing, and try to keep it in the 1/2 to 3/4 zone.

 

Figured I'd comment on my own experience with the Gen 3 Prius as far as engine RPM. For a while I had a habit of running the engine at roughly 1300-1400 rpm for maybe 3 minutes at a time once or twice a day, during my commute. I was getting good overall fuel economy even though I think the engine was outside its most efficient RPM range I would strongly recommend you don't do that every day. (Basically, I'd speed the car up to ~60mph and then slowly bleed off speed over a couple of miles. Fuel economy was good partly because I'd arrive at my exit off the expressway without a whole lot of extra speed to bleed off.)

I'm absolutely convinced using the engine to produce power (i.e. not just idle) for substantial time periods at under 1500 rpm caused EGR problems. I had also used the car as a delivery vehicle where, due to very low average speeds, it'd wear down the traction battery and then have to charge it off the engine at not much over 1000RPM.
I've now picked up the habit of keeping the engine above 2000rpm most of the time if the engine is needed at all - even at relatively slow speeds I try to push the engine up to 2000rpm and then glide (i.e. pulse and glide more or less) which I feel improved my fuel economy a bit over driving at steadier speeds with lower typical engine RPM.

 

Top of the bar before the orange part usually hovers around 2000rpm depending on speed.

 

I'm not sure there's that much of a difference if you just look at trying to maintain the optimum RPM. The car is programmed to balance the RPM, eCVT effective gear ratio, and other things based on your demand. Looking at just the RPM may not give you what you're looking for. I'm sure there's an optimum RPM, but it's an RPM for that particular speed limit, for that particular road incline, and a whole host of other things that change block by block, mile by mile.

Try to keep the bar nearer the middle/left of the range on Mendel's picture and out of the PWR range.

 

The lower your RPM the better your MPG.

I can keep my RPM lower than 2000 at 80 MPH.

The lowest you can keep it before the ICE shuts off is 970-1000 RPM

 

As you see in its manufactured efficiency curve in post #3, the best efficiency of 220 grams of dinosaur bones per kilowatt hour is achieved over a wide RPM range that covers power demands from just over 10 kW to just over 30 kW.

Within that range, you get the best efficiency at whatever RPM is where the amount of power you're asking for intersects the bold path on the graph. If you're asking for, say, 20 kW, that's about 2000 RPM. There are other ways to get 20 kW out of the engine (at higher or lower RPMs along the 20 kW curve), but not as efficiently as where that curve intersects that path.

The car already knows that.

 

See now I'm curious if I can get a ScanGauge to show live g/kWh... I mean why not reduce it out like that?

 

It does, but at some point you've got to account for real-world situations where the power needs vary. I appreciate what you''re trying to do - maximize the efficiency for the process of driving down the road. That's the type of thing that drew me to buying a Prius in the first place.

Engineering is often a trade-offs between competing interests. Running nearer the optimum RPM would make the ICE more efficient, but would require more capacity in the electrical drive system to take up the slack between idling at a red light, coasting down hills, and accelerating at a green light and driving up a hill on the interstate.

That also gets into the balancing act as to how much capacity the batteries, motors, and wiring has with how much the vehicle costs. Which gets into the balancing act between vehicle sales, volume pricing, and the technology that's available at the time.

 

I'm not sure where you could get data to plug in to those calculations, in some easier-to-calculate-with form than the Toyota graphs shown upthread.

I mean, I guess you could blow up the graph and trace the outlines of the 220 g/kWh, 230, 240, 250, 260 contours and save them as parametric curves. Then when you have a given RPM and engine torque demand (both of those are PIDs that exist)*, you can find that operating point on the graph, determine the two efficiency contours either side of you, and interpolate to your operating point along a path constructed to meet both as close to perpendicularly as possible.

From what I remember of the arithmetic you can do on a ScanGauge, it lets you multiply a PID by some factor, and maybe add something. All you usually need to convert from one measurement unit to another. I don't remember it including vector operations or calculus.


* weirdly, the "requested engine torque" PID is documented to have units of kW, which would make it a "requested engine power" PID instead. I concluded they really did mean "torque", and the unit should be Nm. So multiply it by the RPM (and scale appropriately) to find the right place on the "power" axis.