Toyota Claims Bigger Engine are Better for Fuel Economy
 

I was just looking through some of the official Toyota videos on youtube when I happened to stumble across this explanation of some of the tech in the new Prius. I already knew that Toyota had made some changes to the engine to favor a more powerful, quicker Prius. Despite that, the car received a [...]Related posts:

1. PLX Announces New Kiwi MPG Fuel Economy Gauge
2. Ford Upgrades Trucks for “Super Fuel Economy”
3. One Big Reason Not to Trust “Real World” Fuel Economy Testing

More...

I agree that to a point, a larger engine will give better FE. I have had several GM 2.8 and 3.1 MPFI engines in identical cars, and the 3.1 always gives better FE. The only difference between the powertrains is the piston stroke.

Same thing with some others, when you have an undersized engine for the vehicle, the FE takes a dump. Take a larger engine so you barely have to work it, and to a point you can get better FE.

I kinda doubt this with more modern engines. Older engines, that didn't run at stoichiometric all the time will obviously use more fuel at higher engine loads. This is much less common in engines today. They run 14.7:1 nearly all of the time unless they are seeing some higher rpms and high load. Smaller engines leads to increased load, and increased load means greater efficiency IF air/fuel ratios are kept the same. Its not absolutely universal, but pretty darn close.

Use the standard issue (drag-race to the stop sign) nut behind the wheel and compare smaller to larger engine, the larger engine might win out on FE by virtue of the ECU not having to spend 90% of the time in non-stoichiometric fuel enrichment.

Install an ecomodder/ecodriver nut behind the wheel of the car and compare smaller to larger engine and the smaller engine probably wins every time.

Hi,

I was at the Detroit show and saw the new Prius:
The problem is internal engine drag is proportional to speed. So what happens is as the rpm increases, the Atkinson cycle engine, which is very efficient at modest rpm begins to suffer higher internal losses. This reduces the brake specific fuel consumption:
http://hiwaay.net/~bzwilson/prius/pri_power_200.jpg
My NHW11 Prius has the 1.5L engine and you can see that the brake specific fuel consumption is falling off at higher rpm.

This chart shows another effect, which we suspect is fuel enrichment at high power settings above 3,7500 rpm:
http://hiwaay.net/~bzwilson/prius/pri_power_600.jpg
The high power region is where cooled exhaust gas is added to cool the exhaust and allow higher power output with less fuel burned. At high power settings, the engine has to use a rich mixture to keep from burning out the catalytic converter.

So the larger, 1.8L engine and appropriately improved hybrid transaxle lets the engine avoid higher, energy wasting rpms. With cooled exhaust gas, the engine continues into higher power settings but without the extra heat. What we don't know is if a similar system fitted to the existing Prius would provide a similar mileage improvement not counting the friction losses.

BTW, the Otto cycle engines suffer pumping losses through the throttle that the Atkinson cycle avoids. What this means is what works with the Prius engine won't necessarily work with an Otto cycle engine.

Bob Wilson

A rich mixture is used to maximize power output and cool the intake charge in order to minimize the risk of detonation, not cool the catalytic converter, which has typical operating temperatures around 1500-1600F, right around the maximum EGTs SI most engines will make. Having the larger 1.8L engine and taller gearing could very well help out with SFC, although the four-stroke Atkinson cycle is exactly like the SI Otto cycle minus the delayed closing of the intake valve after the piston has started to travel upwards again. The four-stroke versions allowed Toyota to cheaply and effectively de-stroke their 1.5L engines for use in the Prius, unlike the original Atkinson cycle that supposedly used a modified crank and had the intake, compression, power, and exhaust strokes in a single turn of the crankshaft.

Hi,
Lexus Vehicles : Lexus Advances Hybrid Drive with Comprehensive Improvements in New RX 450h / Toyota
When I had my 150 hp, Cherokee 140, we used a rich mixture at maximum power settings so it wouldn't burn out the valves. For maximum power at cruise, the instructions were to trim it 50F rich under peak. BTW, there is an excellent write up on "Exhaust Gas Recurculation" from AutoSpeed.

It means:

• 8-to-1 compression stroke - in an Otto engine, the throttle plate would cause a lot of pumping losses in low power regions whereas the Atkinson cycle makes a substantial reduction in throttle plate losses. This is especially useful in low power regions where low-drag vehicles cruse.
• 13-to-1 expansion stroke - provides a high expansion ratio so a large percentage of energy is extracted. Only diesels have a higher expansion ratio but they also have a problem with NOx formation. Longer durations at higher temperatures favors NOx formation.

Modern Atkinson cycle engines have different compression strokes from the power expansion stroke. This means less maximum power but substantially improved brake specific fuel consumption. Some of believe that the variable valve stroke and angle on intake and exhaust valves, already in some Toyota vehicles, will complete the picture and bring their Atkinson cycle engines into diesel BSFC ranges without diesel problems.

Bob Wilson

Well I'll be! I've never though about starving the cat of oxygen but that certainly shows me! ;)

Scaled for torque output, the engine in the Prius exhibits the same difference in BSFC compared to load a V6 engine designed over a decade earlier exhibits. Specifically, a ~10% increase in fuel consumption when comparing half load and full load, and a doubling of fuel consumption at around 15-20% of load. The Prius is a bit better, around 3%, at lower loads, but I'm not sure if this is the result of the greater expansion ratio or fewer cylinders/offset crank. The wider ovals at max torque are from limiting torque output and the offset crank.

Clearly the 1NZ-FXE would have lower fuel consumption than a normal 1NZ-FE since it's effectively destroked, which improves fuel consumption through fewer throttling losses, but also caps torque output across the powerband, and while the higher expansion ratio may help it doesn't appear to help a whole lot when we scale the output compared to other engines. Capping the torque output by reducing the effective compression ratio and the offset crank are where most of the gains appear to come from based on comparing it to decade+ older/larger engines from Toyota. The only way to determine how much the higher expansion ratio helps would be to compared the 1NZ-FXE to a version destroked to have the same effective compression ratio.

In terms of low drag vehicles cruising in lower power regions, that isn't much of a problem where a hybrid cycles on and off and stores power to be used in a battery instead of operating at a steady/lighter load. This, along with the taller gearing, is what Toyota can increase the engine size in the new Prius and increase or keep mileage the same. It's the same idea behind P&G. It doesn't matter how poor low load engine efficiency is if the engine is almost never operated there. Granted, their is a loss from dumping energy into the battery pack and out the motor, but as long as that loss is less than the increase in BSFC, then there's no reason not to do it.

Diesels have problems with NOx formation because of CI, not necessarily the higher CR. SI engines can have the same CR as modern diesels, but low engine out NOx because the A:F mixture can be distributed relatively homogeneously, which minimizes hots spots during ignition. CI engines otoh see ignition start more or less when the bulk of fuel is injected. Modern versions tend to run a bit richer to cool ignition and prevent as much NOx from hot spots during ignition. Anyway... Variable valve timing almost certainly won't get a SI engine near a CI engine in terms of overall fuel consumption/load, that's something we need HCCI, or something similar, for. On the plus side, GDI gets closer at low load, but that's the best in use implementation I've seen so far.

Using your charts, it looks:

• Prius - BFSC less than 230 g/kw-h between 2,200-3,400 rpm.
• 6-banger - BSFC less than 240 g/kw-h is between 1,400-2,900 rpm.
• 6-banger narrow - BSFC at 237 g/kw-h, 1,800-2,400 rpm.

Did I misread the charts?

When you get a chance, I'd recommend getting a copy of SAE 2004-01-0064 for this quote, "As a result, the minimum specific fuel consumption of 225g/kWh has been achieved. . . " (pp. 7.) This paper is the source of the first graph and does an excellent job of showing the specific systems in the Prius. More importantly, it shows how the Continuously Variable Transmission keeps the engine at the best BSFC over a very wide, rpm range (the operating range line on that first chart.) This operating line is the problem the old 6-banger could never solve with existing transmissions.

In the past, I used "destroked" to mean a mechanical change such as a shorter throw crankshaft to change the piston sweep. I use Atkinson cycle when the gas compression ratio is different from the power stroke gas expansion ratio.

The Otto version of the same 1NZ-FE is in our Toyota Echo and the early Scion series. The same engine is in the Yaris.

Upon further thought, I can somewhat agree. The gas ratios will have a greater impact than duration at high temperature and pressures. Just I remember my chemistry studies and some reactions have non-linear effects based upon pressure and temperature. Wikipedia notes:
But the mechanical aspects probably play another important part.

As the expansion ratio increases, the stress on the piston, rod, crank and cylinder head goes up right after ignition. It is equally likely that the 13-to-1 ratio seen with the 1NZ-FXE is a mechanical limitation. They didn't want to add the additional metal needed for a higher expansion ratio that might cause the engine to approach diesel weights.

Bob Wilson

Makes sense to me that if a guy were to keep putting in a smaller and smaller engine for a job that at some point the fe starts getting worse. Optimal piston speed might come into play here, although I admit I haven't done all my homework on that.

SVO, why don't you just say what you're going to say HERE?

The blog is more of a publicity tool and brings in a lot of people that wouldn't normally come to the forum and/or site. Its just a different way to help expand EM and get new users.

For me, it wasn't the blog as much as reading the forum threads and projects. I've found folks who grab their tools, perform the experiment and post the results. Finding a community grounded in reality makes it easy (perhaps too easy?) to share my data and reference material.

In fairness, I'm a moderator at GreenHybrid.com, a hybrid advocacy group, participate in PriusChat.com, the technical forums and Prius specific advocacy, and "Prius Technical Stuff," a group like ecomodder specific to the Prius. But I'm not here to 'bang the Prius drum,' an advocate, as much as another 'gear head' interested in the technology across a wider base than just the Prius.

Bob Wilson

No, you didn't misread the charts. However it appears you didn't understand what I was talking about. Specifically, you stated that the Atkinson cycle results in a substantial reduction in throttling losses in low power regions, however when comparing it to older engines the change in BSFC compared to load, adjusted for displacement/age of course, is more or less identical. In other words, the Atkinson cycle didn't do a lick for low load throttling losses as far as I can see, adjusted for the difference in overall efficiency/displacement. Granted, Toyota has increased overall engine efficiency, just like they did when comparing the two sixes made about a decade apart in the BSFC map, but all the Atkinson cycle does is allow for a cheap way to reduce the effective displacement, aka a cheaper way of destroking the engine.
Send me a copy! :D
Well, lesse, the old transmission over the same speed range would see BSFC increase about 9% compared to the minimum, specifically a minimum of 237g/kWh and a maximum of 260g/kWh, and the Prius also exhibits a ~9% difference with a minimum of 230g/kWh and a maximum of 245g/kWh. The biggest difference is the more "open" ovals as engine speed increases, which is due to the offset crank and lower friction losses, not the greater expansion ratio AFAIK. Course, like everything either one of us, or possibly both of us, could be wrong, so in short we need more info IMO.
A closer comparison would probably be the 2NZ-FE, although I haven't seen a BSFC map for it, and for that matter, since we don't know what the effective displacement of the 1NZ-FXE is, who knows what Otto cycle engine it would be equivalent to in terms of displacement. If we could slap something together with the same effective displacement, offset crank, all that jazz, less the greater expansion ratio of the 1NZ-FXE, then we could see what the difference is. Until then, all we can do is (fun IMO) internetz speculation.
I've never heard of a greater expansion ratio increasing stress outside of conventional engines. For those it's a given since the CR has to go up too, but if possible, it seems that having ER>CR would lower stress on the engine internals all things being equal.

All things being equal, an engine with a lower CR should see less stress on the internals, and by increasing the expansion ratio they're lowering EGTs/extracting more work over a greater time period, which should also decrease stress compared to extracting more work over the same time period. If anything I think the expansion ratio was dictated by what Toyota could do with the combustion chambers as opposed to the mechanical limits of the block.

Then we'll have to agree to disagree. I can't ignore my data:
http://hiwaay.net/~bzwilson/prius/pri_power_200.jpg
http://hiwaay.net/~bzwilson/prius/pri_power_600.jpg

I was hoping the Autospeed article on EGR might have also given a clue.
http://us1.webpublications.com.au/st...10577_10lo.jpg
Because half of the intake charge is pushed back into the manifold for the next cylinder taking in its charge, the throttle plate pumping losses have been effectively cut in half. Still, the results are quite telling on the road.

Just pointing out there are a bunch of 1NZ-FE engines around. Feel free to find a BSFC chart for them. However, the 2NZ-FE engines are an entirely different breed.

The 1NZ-FE engine was more likely developed for the Otto cycle engines. The hybrid version is a simple modification of an existing Otto engine and that is probably what set the maximum expansion ratio. After all, it takes a pretty high octane fuel to run without knock at 13 to 1.


Same answer, different path, no problem.

Bob Wilson