high mileage version

[SccrMan13]

Would the engine from the newest impreza fit? I know there are some people getting 39+ mpg out of that engine with the cvt. Could be the perfect high mileage combo for fit if it bolts up. Obviously the availability would be slim for the time being but should be some wrecked ones our there by 2014. I wonder how that combo would fair. 148hp and cvt.

[dbjr63]

I talking to Dave at the open house they are going to do a high mileage with the Ford Eco Boost. personally, this is the version I am going to buy. 818 + 2.0L turbo Ford Eco Boost (@ 250HP).

[PhyrraM]

IIRC Dave was more interested in the 1.6 liter and 1.0 liter versions of the Ford. Can't see that great of milage out of the 2.0 (comparative, of course). The Explorer and Edge using the same 2.0 get just shy of 30. Because highway milage is more a function of aerodynamics than weight, I can see maybe 35 from the Focus ST.

As far as the new Impreza motor goes...it should work without too much issue if the transmission can be made (a) to fit and (b) 2WD. The electronics might be a chore to figure out, but doable with enough determination.

[BipDBo]

If you were to theoretically build a high mpg 818 right now, you would need to gets several parts from a subaru donor, but get the engine, and probably the transmission separately either new or from another donor. The Subaru boxer is a great performance engine, but efficient it is not. This may change soon.

What makes the Ecoboost engine so efficient is the use of a turbo along with direct injection. The turbo gives the engine higher power density, so that they can use a lower displacement engine for a given application. The direct injection not only more precisely controls the amount of fuel dispensed, but it allows the engine to retain a high compression ratio. Without, DI and engine with forced induction needs a lower compression ratio, which robs efficiency, making it no more efficient than a larger displacement, naturally aspirated engine of the same power output.

When Subaru partnered with Toyota, The BRZ got Toyota's direct injection technology. Subaru is apparently developing direct injection of their own, and undoubtedly applying it to a turbo engine. There is apparently an STi engine in the works that will have the same current 300 hp output but do it wit only 1.6 liters rather than 2.5. Subaru will need to make more advancements in efficiency to keep up not just with the market but also the EPA. Their crappy belt CVTs aren't going to suffice. I think that we will be seeing small, direct injected, turbo'd engines with Ecoboost-like efficiency coming from Subaru very soon.

[PhyrraM]

From everything I've read the CVT is working fine. It drastically improved mileage over the 4-5EATs, didn't slow the cars down, and 'normal' folks review it as very peppy for a 4cyl (vs. the same motor in the previous Outback/Legacy which was reviewed as slow).

It's us enthusiasts (AKA forum junkies) who seem to have the problems with it. (as expected, I guess)

[HelluvaEngineer]

If you were to theoretically build a high mpg 818 right now, you would need to gets several parts from a subaru donor, but get the engine, and probably the transmission separately either new or from another donor. The Subaru boxer is a great performance engine, but efficient it is not. This may change soon.

What makes the Ecoboost engine so efficient is the use of a turbo along with direct injection. The turbo gives the engine higher power density, so that they can use a lower displacement engine for a given application. The direct injection not only more precisely controls the amount of fuel dispensed, but it allows the engine to retain a high compression ratio. Without, DI and engine with forced induction needs a lower compression ratio, which robs efficiency, making it no more efficient than a larger displacement, naturally aspirated engine of the same power output.

When Subaru partnered with Toyota, The BRZ got Toyota's direct injection technology. Subaru is apparently developing direct injection of their own, and undoubtedly applying it to a turbo engine. There is apparently an STi engine in the works that will have the same current 300 hp output but do it wit only 1.6 liters rather than 2.5. Subaru will need to make more advancements in efficiency to keep up not just with the market but also the EPA. Their crappy belt CVTs aren't going to suffice. I think that we will be seeing small, direct injected, turbo'd engines with Ecoboost-like efficiency coming from Subaru very soon.

In Japan, the Legacy GT uses a CVT. I believe that one uses a chain instead of a belt. It is either a 250HP or 300HP engine. Subaru's CVT is nothing like the others in the market.

[riptide motorsport]

Seems exspensive

[skullandbones]

Please pardon my ignorance regarding the "high milage 818". I have been hearing about it since I got here in July of last year. I understand the basic concept, I think. You have a vehicle with a very efficient engine but still some performance capabilities (250 hp pretty good). However, what I don't get is the use of the vehicle. Are there long distance rallies or events that support this type of project? When I heard the talk about it from Dave and others in the FFR circle I was a little confused as to what generated that direction for one of the 818 versions. Maybe it's a European thing. Like I said, I probably missed the explanations somewhere in the conversation so help me out. Thanks, WEK.

[PhyrraM]

Once you step away from the thought that the 818 is only a track star, and can be built as a normal road car, the high milage versions start to make sense.

Even at 150 HP, an 1800 pound car will out perform most other road cars.

Most of us here on the forums talk about track performance, but I really believe 90% of cars sold will never see one.

[BipDBo]

In Japan, the Legacy GT uses a CVT. I believe that one uses a chain instead of a belt. It is either a 250HP or 300HP engine. Subaru's CVT is nothing like the others in the market.

CVTs cannot, by nature, use chains. Chains interlock with cogs. There is no way for a chain to slip without something breaking. There is also no way to make chains with cogs work as continuously variable because the ratio is dependant on the number of teeth. A CVT works by running a belt between two pullies, which each change size. One pulley grows while another shrinks. The pullies are actually a pair of cones which move relative to each other. When they are compressed together, the belt rides at a higher radis, thereby increasing the effective size of the pully. What you are seeing is a belt that looks like a chain. It transmits force to the pully with friction contact through pads on its sides. A belt made ut of a chain with rubber pads on the sides is no different than a steel reinfoced V-belt. CVTs always work by friction rather than meshing like chain cogs or gears. This is their inherant flaw. They will work flawless for a time, but the friction points, which work under tremendous pressure, eventually wear out. Then they are very expensive to repair. Short term reliability can be good with CVTs, but longevity will be limited. The rate of wear is not dependant on the number of "shifts" but rather proportunal to the number of miles driven. Nissan currently uses CVTs in most of their cars. You will be able to very closely predict when they will fail based on the odometer.

The other problem with CVTs is the diminishing return of benefit. The idea is that you can get the optimal rpm. With 6-speed transmissions you can get pretty close. Now they are coming out with as many as 11 speed transmissions. The highway mpgs, which is what gets advertized, as far as the transmission is concerned, is really dependant on the final gear ratio. The Chevy Cruze Eco uses a simple 6 speed manual. 6th gear is such that you can cruise at highway speeds with very low rpm, and therefore get very high mpg. Gears have lower friction loss than belts, so a properly geared transmission with clutches (no fluid torque converter) should easily beat a belt CVT in real world efficiency, all while being cheaper and longer lasting.

A lightweight, low power economy car may do alright with a CVT. A CVT may also work alright in a lightweight 818 so long as the power output is low and you don't drive it hard. I still would never buy a car with a CVT, especially one that is heavy, powerful or powers all four wheels. When the wheels don't slip, the belts will.

By the way, a eCVT or power-split transmission as seen on a true hybrid like the Prius and Volt is a completely different animal. It is the most mechanically simple transmission of them all. It really shouldn't even be called a transmission IMO. It is a fixed planetary gearset. It never changes gears. It just takes power from both a gas engine and an electric motor and allows the two power sources to trade rpm. Electronics can fail (although they rarely do), but there is very little that go wrong mechanically with a planetary gearset.

[Twinspool]

CVTs cannot, by nature, use chains. Chains interlock with cogs. There is no way for a chain to slip without something breaking.

Uh oh. I better go install a belt in my differential since gears don't slip without breaking teeth! ;)

[BipDBo]

Uh oh. I better go install a belt in my differential since gears don't slip without breaking teeth! ;)

Differential slip occurs when one wheel spins faster than the other. Unless you start breaking teeth, the average rpm of the two wheels is always at the same proportion to the average rpm of the input driveshaft. It works like a hybrid planetary gearset, but transfering power in the reverse direction. Instead of transfering two power sources to one output, it transfers one power souce to two outputs (two wheels), allowing the wheels to trade rpm between eachother. The only slipping allowed by a differential occurs between the tire and the road.

[HelluvaEngineer]

CVTs cannot, by nature, use chains. Chains interlock with cogs. There is no way for a chain to slip without something breaking. There is also no way to make chains with cogs work as continuously variable because the ratio is dependant on the number of teeth. A CVT works by running a belt between two pullies, which each change size. One pulley grows while another shrinks. The pullies are actually a pair of cones which move relative to each other. When they are compressed together, the belt rides at a higher radis, thereby increasing the effective size of the pully. What you are seeing is a belt that looks like a chain. It transmits force to the pully with friction contact through pads on its sides. A belt made ut of a chain with rubber pads on the sides is no different than a steel reinfoced V-belt. CVTs always work by friction rather than meshing like chain cogs or gears. This is their inherant flaw. They will work flawless for a time, but the friction points, which work under tremendous pressure, eventually wear out. Then they are very expensive to repair. Short term reliability can be good with CVTs, but longevity will be limited. The rate of wear is not dependant on the number of "shifts" but rather proportunal to the number of miles driven. Nissan currently uses CVTs in most of their cars. You will be able to very closely predict when they will fail based on the odometer.

The other problem with CVTs is the diminishing return of benefit. The idea is that you can get the optimal rpm. With 6-speed transmissions you can get pretty close. Now they are coming out with as many as 11 speed transmissions. The highway mpgs, which is what gets advertized, as far as the transmission is concerned, is really dependant on the final gear ratio. The Chevy Cruze Eco uses a simple 6 speed manual. 6th gear is such that you can cruise at highway speeds with very low rpm, and therefore get very high mpg. Gears have lower friction loss than belts, so a properly geared transmission with clutches (no fluid torque converter) should easily beat a belt CVT in real world efficiency, all while being cheaper and longer lasting.

A lightweight, low power economy car may do alright with a CVT. A CVT may also work alright in a lightweight 818 so long as the power output is low and you don't drive it hard. I still would never buy a car with a CVT, especially one that is heavy, powerful or powers all four wheels. When the wheels don't slip, the belts will.

By the way, a eCVT or power-split transmission as seen on a true hybrid like the Prius and Volt is a completely different animal. It is the most mechanically simple transmission of them all. It really shouldn't even be called a transmission IMO. It is a fixed planetary gearset. It never changes gears. It just takes power from both a gas engine and an electric motor and allows the two power sources to trade rpm. Electronics can fail (although they rarely do), but there is very little that go wrong mechanically with a planetary gearset.

Here is a link with pictures. They use a chain http://drive2.subaru.com/Sum09/Sum09_whatmakes.htm

[BipDBo]

Here is a link with pictures. They use a chain http://drive2.subaru.com/Sum09/Sum09_whatmakes.htm

It looks like a chain, but functionally, it is a belt. Like I said, a chain meshes to cogs. This uses friction pads touching a smooth pulley surface. On each side of that chain/belt are angled rubber-like contact pads that are equivalent to the sides of a rubber "V" belt. They use a construction similar to a chain to reduce stretch. It's the same thing as when they put steel or fiber reinforcement within a rubber V-belt. It's also the same basic design as Nissan CVTs Nissan's belts look a bit different, but they work the exact same way.

[jsquared]

Tomato, tomato. :D

Audi's CVT also uses a chain, the "pulleys" grab the pins at the end instead of the typical cog tooth grabbing the pin at the center.

[Twinspool]

It looks like a chain, but functionally, it is a belt. Like I said, a chain meshes to cogs. This uses friction pads touching a smooth pulley surface. On each side of that chain/belt are angled rubber-like contact pads that are equivalent to the sides of a rubber "V" belt. They use a construction similar to a chain to reduce stretch. It's the same thing as when they put steel or fiber reinforcement within a rubber V-belt. It's also the same basic design as Nissan CVTs Nissan's belts look a bit different, but they work the exact same way.

But if the CVT is on a conveyor belt, will it take off?

[PhyrraM]

But if the CVT is on a conveyor belt, will it take off?

Mythbusters says yes.