Just like the two-seat Honda Insight. Its 3-cylinder, 1.0 liter gas engine is the main power source. A 7 hp electric motor helps during acceleration and hill climbing. A continuously variable automatic transmission cuts down on energy lost during shifting and regenerative braking increases efficiency.

Visit Hybrid Vehicles for more information about EV.

Sounds good - lets see what it sounds like, and how much it costs.

I agree with Peter_xx. This could be incorporated into mid-size PHEVs as low fuel consumption gensets. Smaller, lighter, PHEVs could use micro-cars 600 cc ICE for gensets. There's no real needs for larger, heavier 4 or 6 cyls gensets.

Although efficient at full load, the Toyota Prius Atkinson cycle engine appears as a big dinosaur in comparison with this engine. Downsizing vs. anti-downsizing? Since this car reportedly has CO2 emissions below 120 g/km with conventional drivetrain, a mild hybrid version of the Ford would come very close to Prius and at a fraction of the incremental cost for a full hybrid. Does someone remember the recent article about the Ricardo concept car (Focus based)?

I have now put this car on the list of cars I would like to test drive. If I would like to buy, I do not know yet...

I think the idea that people resist new technology or small engines (with adequate power is a rationalization that just do not want small underpowered or overpriced cars.

Hybrids and EVs are not considered good values

It is not problematic to get customer acceptance for small displacemnt engines in an Audi.

I think the (or one) problem with switching a normal otto gasoline engine to Atkinson cycle is the compression ratio would be too high in otto mode or too low in Atkinson mode.

@MG,
The reason that mfg's don't make an engine switchable from Otto cycle to Atkinson cycle is due to the high geometric compression ratio inherent in the design, usually 13:1. In Atkinson mode, the CR is 10:1, while in the Otto mode, the CR will be 13:1, which is a bit too high. Direct injection will help, but at increased cost. Ethanol injection can be done in the Otto mode to prevent engine knock when high acceleration is needed, similar to the MIT ethanol boosted engine, in which ethanol injection is done when the engine is boosted, allowing higher CR than normal, hence higher expansion and more efficiency. Or, variable compression ratio (VCR) mechanism can be built into the engine, a bit complicated, but doable.
Good question, and I hope that auto mfg's will pursue at least one of the three options.

@Peter,
The lower peak output of Atkinson-cycle engine is corrected by the electric motor boost in an HEV, thus, there is no need to upsize the engine, which will incur more engine friction at part load. Furthermore, in the Toyota 1NZ-FXE which has lower peak rpm, engine friction is reduced by light-weight moving parts and low-tension piston ring, and offset crankshaft to reduce sideload on the piston skirt.

@ToppaTom,
Hybrids are definitely good values, according to the long-term data from many taxi companies that use the Prius in their service in nearly a decade. The much lower cost of operation will more than make up for the higher purchasing cost of the car. Maintenance cost is sharply reduced, and so is fuel cost!

@MG
Thanks for the tip! I actually have had access to many of those and have also digitized a couple of them in the past but not all of them. It is sometimes kind of difficult to compare maps like these, since they are so different (axes, scale, resolution etc.). However, there is a simple method one can use. Digitize a torque curve at approximately 2000 rpm (somewhat higher or lower rpm if the best “island”) and plot a curve. Then, compare curves for different engines simply by plotting fuel consumption vs. per cent torque. It gives a pretty good hint about which engine is better at low load or high load. A more sophisticated way is to run the engine in a model car in some drive cycle simulation software. In one analysis of this kind I did more than 10 years ago, I found that a turbocharged downsized engine in combination with mild (parallel) hybrid drive could become about equal to a full hybrid in fuel consumption. However, to get the full benefit, the ISG type of motor/generator would have to be bigger than what was possible at that time (e.g. Honda). Today, this is possible.

@Roger
Well it should be obvious to anyone that you can have a smaller ICE in a hybrid system than with conventional drivetrain. Of course I have realized that. But we have to compare apples and apples, not apples and pears. If you compare two hybrids, the Toyota “anti-downsizing concept” is hampered by the low power density of the Atkinson cycle. Not so much as it would be in conventional drive system but still it is a considerable drawback. I thought I was quite clear in my explanation about that. The fact that Toyota now is looking at turbocharging also proves my point.

@MG,
Thanks for the BSFC map info.
Let me distinguish between geometric CR and effective CR. In an Atkinsonized engine, the geometric CR is 13:, set by the ratio of entire swept volume of the piston over the dead volume on the top of the cylinder. However, the effective CR is only 10:1, due to the fact that the intake ports close late, well past BDC, so that some of the intake volume was pushed out of the intake port early on the compression stroke.

If you want to have an engine that can be switched between Otto mode and Atkinson mode, you must be prepared to deal with the higher effective CR of 13:1 in the Otto mode.

@Peter,
Let's consider the 1NZ-FXE 1.5-liter Atkinson cycle engine. It has a geometric CR of 13:1, but an effective CR of only 10:1, so in effect, its power output is only that of a 1.15-liter Otto engine, (10/13=77) or 77% of the power, because the effective intake swept volume is only 77% of the geometric swept volume. Now, let's compare the two engines of different sizes but similar peak output:

Due to the friction-reduction measures that Toyota has done on this engine, its friction is only ~80% that of an Otto-cycle 1.5-liter engine. At part load of let's say, 1/3 of full load, the viscous-friction drag is comparable in spite of the size difference. The pumping losses from moving the pistons against negative atmospheric pressure are also quite comparable in both engines even though their pistons are of different sizes and swept volumes, because the larger Atkinson engine requires higher manifold pressure than the smaller Otto engine for the same part-load output.
This can be verified on the BSFC maps for both engines. Atkinson engines actually have comparable part-load efficiency as Otto engines, and Toyota further enhanced this on their newer 2ZR-FXE - 1.8 L engine with even higher part-load efficiency.

Bottom line: In spite of its larger sizes, Atkinson-cycle engines behave efficiency-wise as if they are being downsized to ~77% of their geometric displacement.
The reason for going to Miller cycle by adding a turbocharger to Atkinsonized engine is pure lust for higher specific power. The price to pay for this is higher cost of turbocharger and higher engine operating temperatures that may have a negative impact on engine longevity or reliability. I love the Atkinson engines because they run cool, low-stressed, and will last longer.

@E-P,
Agree. Miller cycle is for those who wants extra power, while wanting a higher level of efficiency than those offered by turbocharged vehicles, and are willing to afford the additional cost of turbocharging.

For those who just want to stick with the highest in efficiency at the most affordable price, an Atkinson cycle with electric boost would be best, since the electric motor and the battery do more than just provide power boost: They allow for braking energy recuperation, and the battery provides ample of power for the AC compressor when the engine is stopped at a traffic stop.

A micro hybrid with too small a battery requires that the engine be idling to provide power for the belt-driven AC compressor, and that wastes a lot of fuel, since an idling ICE gets the worse efficiency.

@Peter,
Love may be blind, but figures don't lie.
Friction can be reduced on engines with lower loads and lower rpm for any given displacement. With lower loads, the bearing surfaces can be made smaller, hence less viscous friction drag. With lower rpm, inertial forces will be reduced by the square of the rpm, hence lower friction on con rod, crank pin and crankshaft bearings. With lower rpm, lower-tension piston rings can be used...

The friction loss of air going back out of the intake valves is quite negligible. At 2000 rpm, the friction loss of air moving thru the intake valves is only 1/5 the loss when the engine is running at 4500 rpm in proportion to the power produced, quite negligible. However, at 4500 rpm, which is the engine's peak power output, the intake air is moving in so fast that its inertia does not allow it to reverse direction until well after 30-35 degrees after BDC, meaning that proportionally a smaller volume of air would be forced out of the intake ports than at lower rpm's, still keeping this loss at a low percentage.

Anyway, in comparing some Otto-cycle engines of similar displacement to the 1NZ-XFE, I can see your point, in that the 1NZ-XFE is not quite as efficient at low loads in comparison with some other Otto-cycle engines.

The implication of this is that Atkinson cycle engine in HEV's must be downsized even more if efficiency is to be improved.
The hypermiling of the Optima Hybrid recently shown in Guiness World Record demonstrated that the pulse & glide method can increase the hwy mileage of the Optima by 61%, meaning that the 2.4-liter 160-hp engine must be reduced in half, to a 2-cylinder 1.2 liter 80-hp engine, in order to realize maximum cruise efficiency. This can be done by doubling the power of the motor and the battery to 80 hp in order to maintain the same acceleration performance, to provide a combine power level of 160 hp, still very adequate!

However, Toyota managed to increase the size of the Prius engine from 1.5 liter to 1.8 liter and still gets higher efficiency out of it, even at lower loads, instead of less, as we would have predicted. Go figure!
Apparently, there are more in the world of engines than are dreamed of in our philosophy!

Peter XX: Start a blog. Put that diagram, and any others you do, on the blog. Annotate the diagram in the text so that search engines can find it.

If you're doing the work anyway, add it to the searchable global knowledge base.

I have been quite busy with other things lately but, finally, I was able to make the comparison between the Prius engine, a naturally-aspirated engine and a highly turbocharged engine. Note that engine maps with BSFC are not published very often, so the choice of engines is by no means “optimal” but it proves general trends. I chose a GM 2.2-liter engine (MY 2010) with MPI injection and a Hyundai 2-liter turbocharged engine (MY 2011). All engines are approximately the same size, so the impact of engine size will be left for later discussion. The GM engine is quite old but I digitized the diagram long ago and have used in in many comparisons in the past. It is a reasonably “good” engine for its size and age. The Hyundai engine was just recently published in the MTZ journal, so I could not resist digitizing it for this comparison. I might look for a more proper comparison when I have more time but this is what I had available for the moment. If someone could hint me about a neutral site to upload a diagram with the results, I would be glad to do so.

The results: As we would expect, the Prius engine is better than the other two at high loads. If you would compare FC vs. BMEP, the Prius engine would also be better than the GM engine at light loads but when the comparison is made vs. load (%), the crossover point is at about 25% load. The GM engine is better at lower loads, albeit that the difference is not great. This is in spite of the measures to reduce friction in the Prius engine, while the GM engine - considering its age - does not have such measures and that the “Atkinson effect” with higher expansion ratio also contributes positively at light load. The low power density and pumping losses for the Prius engine outweigh the positive contributions mentioned. The turbocharged engine has the highest fuel consumption at high load, as expected, but it is better than the other two at low load. This is why we need the Miller system to improve full load fuel consumption! The crossover point vs. the GM engine is at ~65% load and at ~50% vs. the Prius engine. The difference at low load is quite dramatic. For example, at ~20% load the BSFC advantage for the turbocharged engine s in the order of 60-80 g/kWh. This is why we need turbocharging and downsizing!

As I said in a previous e-mail, we would have to “run” the engines in some drive cycle simulation software to be able to properly quantify the differences in conventional and various hybrid drive systems. Needless to say, different driving patterns are also important in such comparisons.

@EP
One of my principles in participating at GCC has been not to refer to my own work. I do not always succeed with that objective but I prefer to cite studies made by others. I this case, I did not have any study to refer to. I have no intention to start a blog.