Why hybrids get better fuel economy on highway?

I have some half baked theories. Once I post them, sane members will debunk them one by one and we will both learn something.

I suspect no diesel car has as low a Cd as a Prius or Insight, .25 is a hard number to match, my last car had .43
[ame="http://en.wikipedia.org/wiki/Automobile_drag_coefficient"]Automobile drag coefficient - Wikipedia, the free encyclopedia[/ame]


Unless the road is level, the Prius is 'cheating' going both uphill and down. The pedal is a sign to the CPU of your intentions, but the CPU can run the ICE at full throttle and store the extra energy as battery charge, thus avoiding pumping losses, as the throttle is wide open, just like a diesel. Prius owners can find more battery charge at the top of a hill than before climbing it. On the way down, that battery charge can provide most or all of the energy needed to keep a car going downhill from losing speed. The Prius is used to restarting the ICE by computer whim, so it can rotate without gas being injected untill needed, just like a diesel. If the hill is steeper than needing any extra energy, the Prius will store more in the battery.

Unlike a normal automatic, which downshifts when you add gas, or a '5 speed' where gear ratios are chosen by the driver, the CVT can decide to run full throttle at lower RPMs, to avoid pumping losses. So even when you are on level ground, the pumping losses usually associated with gas engines can be minimized.

Most gasoline powered cars use the Otto cycle, the Prius uses the Atkinson Cycle idea. The 'downside' of the Atkinson Cycle is less power, but if you use more throttle to make up for that, you have less pumping loss, again.
[ame="http://en.wikipedia.org/wiki/Atkinson_cycle"]Atkinson cycle - Wikipedia, the free encyclopedia[/ame]

 

I've thought about that and I attribute it to a really efficient gas/electric gear-driven CVT transmission.

In general, highways are usually not completely flat. Give the Prius even a slight downward slope and it will lower the engine RPM to about 1280 RPM and stretch the fuel economy to about 2.5L/100km (as measured by a Scangauge II).

A 5-speed transmission is never in the perfect gear, even at highway speed in 5th gear. It has to be geared high enough to provide some power reserve or you'd have to downshift every time you needed to go up even a minor hill. So even on a truly flat road, it wastes power.

The Prius is always in a perfect gear regardless of highway conditions.

 

I can't address much of your post, but will address this one.

Normal Otto-cycle gasoline engines, when operating at partial throttle, waste a lot of power as air pumping loss. The intake manifold is under a strong vacuum, and pulling all that intake air through that vacuum requires considerable power. Power that is no longer available for propulsion.

In the North American market, this situation is made worse by our preference for large displacement engines, and having them geared for performance. Most of our cars spin at much higher RPM than necessary at highway speed, in order to have lots of torque available without downshifting. Americans have a name for cars that must downshift -- 'gutless'.

The smaller displacement of typical hybrids reduces pumping loss. Their Atkinson cycle reduces it further. And their CVTs are geared taller, allowing a lower highway RPM, even further reducing this pumping loss.

 

I believe the old non-hybrids that did deliver stellar MPG did have to downshift for relatively shallow hills. But such downshifting has become a mark of shame in America.

I could see the difference in my cars. Climbing Vail Pass in Colorado (above 10,000 feet, with nearly 1/3 less air than at sea level) in my fully loaded Subaru, it was able to maintain speed limit in high gear for all but the very steepest pitch (7% ?).

The one time my ancient Honda was there, it had to downshift 2 gears before the Subaru had to downshift 1 gear. No wonder the Honda had far superior real-life MPG, far more than the EPA stickers would suggest.

 

Atkinson cycle not miller cycle, miller cycle needs a supercharger.
Atkinson Cycle is a LOT more efficient than an Otto Cycle petrol engine.

That, in a nutshell is your reason.

 

The reason the Prius gets the numbers on the highway it does is the combination of all of the factors listed above:
Atkinson cycle is probably the #1 factor.
Low drag #2 factor (above a certain speed its the primary)
CVT allows for low highway RPM especially on the 2010
Downsized engine. Today under 100hp engine on a car the size of the Prius is unusual.

A 100hp diesel with a CVT and a .25 Cd might do better then the Prius but it certainly would be more polluting.

 

I suspect they have fewer computers between the pedal and the wheel, the Prius lets the computers do quite a lot of driving. The 1.8 liter Nissan Versa with the best gas mileage is the CVT, so all alone it seems worth 2 to 3 MPG.

Manual Transmission 26/31
Automatic Transmission 24/32
Continuously Variable Transmission 28/34

I am not aware of another car with all three transmissions, but perhaps some one is.

 

Thanks, but regardless of the terminology, the Atkinson cycle engine is responsible for the lions share of the increased efficiency of the Prius. The Atkinson cycle results in a lower power output for a given capacity and lower torque, but the electric assistance from the hybrid transmission means this lower power output does not hinder the acceleration of the hybrid car, in fact the high torque of electric drive compliments the atkinson cycle engine perfectly.

By the way, Miller Cycle is very much like an Atkinson Cycle with a super charger. It is all about valve timing and compression ratios, or rather expansion ratios. I think both use offset crank positioning.

 

First your observations are correct and these details will help:


Key elements:

• Both 1.5L Prius, NHW11 and NHW20, have a distinct fall off in efficiency at higher power settings. The engine has to move into rpm ranges that are less efficient. The NHW11 drops faster in part because it needs more power to handle the draggier body shape. It also has a less efficient transmission.
• The 2010 Honda Insight really suffers with a rapid fall off in performance.
• The Yaris with the same 1.5L engine block as the Prius is running an Otto cycle and is not even in the running.
• Notice that at 65 mph, the diesel is falling below the 1.5L Prius performance. This is probably because the 1.5L Prius engine is running at peak efficiency that soon falls off at higher speeds.
• The 1.8L Prius, ZVW30, is obviously pacing and exceeding the TDI diesel across all three speed data points. This is what you are probably curious about.

There are multiple features that contribute to the 1.8L Prius performance. It is the sum of the parts:

1. Atkinson cycle - lower compression loss with high expansion ratio
2. E-CVT - this is not a mechanical CVT transmission
3. Cooled exhaust gas recirculation
4. 27 kW traction battery, 36 extra HP on demand
5. By-pass valve in muffler

I. The Atkinson cycle delays closing the intake valve for part of the compression stroke. Part of the fuel-air mixture is pushed back into the manifold and pulled into an adjacent cylinder. This reduces the engine pumping losses but it also gives a lower compression ratio, typically 8 to 1, with a very high expansion ratio, 13 to 1.

The lower compression ratio reduces the risk of knock and allows using a higher energy content, lower octane fuel. The high expansion ratio, 13 to 1, maximizes extraction of energy but not so high as to cause excessive NO{x} generation. Thus a fuel efficient, 3-way catalytic converter can be used and avoid the NO{x} scrubbing systems used with clean diesel.

II. The Prius transmission is an E-CVT, electronic CVT, not a mechanical CVT like the Honda. One difference is no clutch and an extremely wide range. Instead, it uses two computer operated motor generators MG1 and MG2 and at any point in time, 28% of the engine power flows between MG1 and MG2 while 72% flows through a planetary gear called the Power Split Device.

The engine turns the planetary gear carrier with the spider gears. At the center is the sun gear driven by MG1. The outside, ring gear, eventually drives the differential and MG2. The operational scenario is:

• When MG1 is electrically disconnected, the engine spins the spider gears but no torque is transmitted because the sun gear offers no counter torque.
• As soon as MG1 begins to work as a generator, it causes a counter-torque that the engine driven spider gears send to the ring gear ... moving the car forward.
• The power from MG1 is sent to MG2 and combines to help move the car forward.

The E-CVT has a mode called "energy recirculate" that functions like an overdrive. MG2 begins generating power that it sends to MG1 to force the engine to turn slower. However this mechanical power just passes back through the Power Split Device so it remains very efficient. It is the wide range of the E-CVT that allows it to keep the 1.8L engine in its most efficient performance zone that helps on the highway.

The ZVW30 replaced a drive chain, improved lubrication and reduced gear stages. It is ~30% more efficient than the NHW20 transaxle.

III. The ZVW30 engine has a cooled exhaust loop used at high power settings to reduce exhaust gas temperature. The primary reason is to avoid having to use a rich mixture to cool the exhaust and preserve the 3-way catalytic converter. It also helps to reduce NO{x} generation. I suspect the dip we see with the NHW11 and NHW20 above 65 mph comes from enriching the mixture to reduce exhaust temperature. But on the highway, the cooled, recirculated exhaust keeps the fuel efficiency up. I understand some diesels use exhaust gas to help reduce NO{x} but I've not made a study of it.

IV. Next comes the extra 36 hp in the traction battery. What this means is an electronic boost used for passing or high-speed, very steep, hill climbs. This means the engine doesn't have to be larger than 1.8L to achieve excellent performance.

V. At high power settings, the 1.8L Prius has a valve in the muffler that opens to reduce back pressure. At highway speeds, the tire and air noise mask the slightly increased engine noise.

It is the combination of parts that makes the Prius efficient. The advanced 1.8L engine, E-CVT, and traction battery that makes the whole work more than the individual parts. But there is a better one coming ... variable intake duration engine.

Today, the intake valve open and close durations are fixed but they can change relative to the crankshaft angle. In the future, the open and close durations will change so no throttle plate will be needed. Right now these new engines are being tested in non-hybrid vehicles.

Bob Wilson

 

Could you provide the source of that information?

Perhaps you misundertood? Atkinson cycle efficiency peaks at 37% while Otto cycle peaks at 32%. That gain is actually about 15% not 5%.

It is also important to note that eCVT almost disconnect the ICE from the wheels and can run at a very wide RPM range. 3k rpm uphill, <1k downhill and 1.5k on a flat surface. You can google for a video I made on the highway at 65MPH on cruise control.

Due to the hybrid components; MGs and HV battery, torque is no longer controlled by throttle + gears (transmission). Instead, torque on demand from the MGs is utilized; allowing full throttle at all (majority of the) times. This is the advantage of series hybrid such as the Volt. The only difference could be is the Volt ICE may run at a more constant RPM.

 

Interesting topic. It has been discussed many times before, but that probably only reflects the incomplete understanding of many people, including myself.

Bob Wilson's post is probably all you need. Other differences that come to mind that play into the end result:

• Diesel cars are heavier than similar sized petrol cars
• Current pollution devices for diesels take more energy than petrols, and this is probably only going to get worse.
• Diesel cars that come to the US are geared for 'peppiness' and 'performance' rather than fuel economy. Hybrids can suffer the same fate, e.g. the Insight II. A slew of very small diesel cars are available in Europe that are rated for 60 mpg+ on the highway. 0 - 60 mph in 12 - 14 seconds is typical.
• All engines have efficiency curves dependent on power, torque, and rpm. While the diesel has a higher theoretical maximum than a petrol engine, what matters on the road is the integral of the actual instantaneous efficiencies. Prius averages 30+ percent.
• The Prius transmission is a model of efficiency

Other comments:
As BW's graph suggests, each car has it's own best highway speed, and they differ from car to car. A 50 kw (75 hp) diesel may not look so great at 70 mph, if high rpm is required. OTOH, it might rule the roost at 60 mph.

2. I think your aero calcs are off. If the Prius Cd is .26, and a VW Cd is .30, the difference is 15%.

Lastly, if you are really interested in this topic, track down a book written by Toyota on Prius development. They are very clear in making the point that they looked for efficiencies *everywhere*, not just the hybrid powertrain. Electronic steering, LRR tyres, and electric variable AC are just three examples among many. Every single electric and mechanical component in a car, in addition to the shape and weight, determine eventual fuel economy. Restricting yourself to 'fuel energy' or 'compression ratio' misses the forest for the trees.

 

Absolutely! For example, the 1.8L engine has no belt for driving the water pump or air conditioner compressor or power steering pump or an alternator (MG1 is the alternator.) I believe the book is called "The Prius That Shook The World"

Bob Wilson

 

Hi Marcinpisz,

Not sure where your getting that 3000-4000 RPM. My Prius on flat terrain, in no-wind conditions at 70 degree F on a sunny day does 1280 rpm at 53 mph. That would scale to 1700 rpm, but in actuallity the RPM will be greater than that, because wind-drag is not constant with speed but increases. I have never driven my Prius at 70 mph in these conditions, while I had a tachometer readout installed.

Now, put any hill in there, or head wind and it goes up from there. Make it colder, or set the sun, and the rpm goes up too. Make it hotter, or increase humidity (while keeping the AC off) and the RPM goes down. Which is why it confuses me no end, why the Volt does not have something like a 1.0 liter turbo diesel ?

The Prius transmission allows the car to put the engine at the optimum efficiency torque and rpm. These two are linked to get best efficiency. In a fixed geared car, the transmission is typically geared below the optimum engine efficiency torque, as that is close to the engine stall torque of an Otto cycle engine. The consequence is your running down the highway with the throttle valve mostly closed, and allot of pumping losses.

I think the stuff your missing is that engine efficiency is only a small part of car-system efficiency (mileage). Good mileage is as much about how the car can put the engine in the optimum torque and rpm when it needs engine power, as how efficient the engine is optimully.

The Atkinson Cycle (as implimented with a Miller Valve timing) in the Prius gives a broader power range for optimum efficinciency, than an Otto Engine. Argonne National Lab reported that the Prius engine still had 25 % efficiency (down from 30 %) at 12.5 hp. In an Otto engine of the same size and peak power, you can pretty much reverse those numbers. Drop the Otto engine from 76 hp to 25 hp, and at 25 hp the efficiency will be close to 12.5 %, even though it started out at 25 %.

This is why they talk about Synergy so much in the Prius. Every thing contributes advantagously. The Aktinson Engine, the hybridisation, and the eCVT.

Another comment is pumping. A Diesel is pumping allot of air it does not need for combustion. And even though the Diesel is not limiting that flow in anyway, it still passing all that gas through. A low RPM Prius at optimum load has its throttle valve nearly open. And the RPM is limited through computer action by the vehicle load. So, there is effectly little throttle loss difference between the Prius and the Diesel, yet there is allot more RPM's and allot more gas volume being pushed through the Diesel by the fuel.

This last point is probably why the Hybrid Diesels in the Patnership for the Next Genertion (PNGV) program got 80 mpg on the Old EPA test, versus the Prius 61/51 . They could run the small diesel slow and loaded, combusting all the air, and not wasting any pumping.

 

Hi Again,

Another way to say some of that stuf above, is that a traditional drive train locks in the engine speed to a ratio of the car speed. You can add more gears, but if your locked into one cruising gear there is still low-hanging-fruit.

Specifically, the wheel torque required varies dramatically with the roadway conditions. Anybody who has driven a Prius on the highway understands this. As the Prius engine is up and down in RPM, and as the Prius engine has a pretty flat torque and efficiency curve, one can infer the Prius is matching the car's power requirement at optimum efficiency on a moment by moment basis.

A traditional transmission car cannot do this. As even though the speed does not change, and the engine RPM does not change, the engine and wheel torques do have to change. In an Otto cycle car that means the throttle is opening and closeing and creating pumping losses. In a Diesel car, that means the fuel is pumping a displacements worth of air, times RPM , unvaringly, and only the amount of it to be combusted is varying. No matter what the torque load, the engine is still pushing the same amount of air through it.

And this is one of the low-hanging fruits that the extreme hypermilers pick. Specifcally, while on the highway, the shift into neutral and turn off the engine. Now no air is pumped through the engine. Do not do this with an automatic transmission car, unless you know you have one of the few towable automatic transmission cars!

 

At 60MPH on a level road the Prius only runs at about 1500-1600 RPM. At 70 it runs at 1800-2000RPM. This is incredibly low for an engine with an equivalent displacement of only ~1.1L (1.5 together with Atkinson pushing out 1/3 of the inlet charge). My BMW 328 was geared so it ran faster than that!

The Prius engine speed varies greatly for a given road speed - even a slight slope may cause it to go up to 3000RPM or drop to 960RPM.

I think this is where the majority of the highway improvement comes from over a conventional car. The battery helps here also in that there needs to be less excess capability available because an additional 20HP or so is instantly available if needed.

kevin

 

Bob,

Minor correction - it is 28% of the torque flows into MG1, 72% to the final drive. At various combinations of road speed and load MG1 will not be rotating so no power will be going through that path at all - all the engine power will go through the efficient mechanical path

kevin

 

MG1 can not generate torque without power. Take the power away and MG1 will free wheel. Since 28% of the engine torque goes one way via MG1 and 72% the other, the ring gear, the ICE rpm is the reference needed to calculate the power flows and there is only one ICE rpm.

When you use the PSD model, you'll notice the minimum vehicle speed is ~25 mph to get MG1 near 0 rpm. But 25 mph is more than fast enough to generate the power from MG2 to 'freeze' MG1:

Toyota Prius - Power Split Device

The MG1 control law only has to maintain the necessary counter torque for any given ICE rpm. Just monitor the current and voila. The MG1 rpm is unimportant as long as the ICE counter-torque is maintained. The traction voltage has to be high enough to force the current through the back EMF from MG1 rpm.

The "zero" MG1 rpm is the transition between normal and "heretical" mode and though these transitions happens often enough, the power flows are driven by the torques from the spider gears to the ring and sun gears. It is really dominated by the MG1 torque.

In my studies, the only use for MG1 and MG2 rpm are to calculate the ICE rpm. The interesting stuff is MG1 torque whose fixed relationship to the ICE makes monitoring the ICE shaft power possible.

Bob Wilson