Should GM build the next Cobalt around a 3-Cylinder?

[dwightlooi]

GM has been sucking for years. I know many of you don't to hear that or admit that, but it's true. GM sucked in the 80s. They sucked in the 90s. They have been dong all the bean counting and quality skimming possible in order to remain the "value leader" while operating out of a high cost country (USA). It's a moronic formula, but for over two decades they stuck to it. Today, GM vehicles (and most American cars for that matter) are viewed as technologically and qualitatively inferior products with a dubious reliability record, poor economy and resale values in the toilet. GM has finally awaken -- the GMT900s are OK, G6es are decent enough, the Aura is yet another step forward, the Solstice is a looker, the new CTS is world class, the Malibu has all the markings of a comeback car like the Taurus was to Ford in 1988. But to press the offensive, it's time GM think out of the box and aim not to simply match the competition or edge ahead slightly but to totally surpass them.

In order to beat the Civic and the Corolla in fuel economy numbers and perceived technological superiority it's time to pull out all the stops. And it's a tall order. The R18A engine in the Civic for instance is not only a VVT unit, it is a part time atkinson cycle engine. Meaning that at part load, the VTEC system engages a second set of intake camlobes which keep the intake valves open well into the intake stroke. This allows the engine to kick part of the intake air back out of the cylinder. The throttle body remains nearly fully open while the engine breathes easily through the unrestrictive intake system while providing cruise economy. Combined with a 5 speed automatic and impressive weight discipline in chassis engineering yields a car that is bigger than the G6 on the inside while weighing 2750 lbs and delivering 30 mpg in the city and 40 mpg on the freeway under the 2007 EPA standards.

Going through GM's list of new technologies and the latest parts bin we can give the new Cobalt the following:-

• Direct Injection
• Dual Cam VVT
• 4-valve DOHC valvetrain with roller followers
• 36 volts electrical system with alternator-starter-motor
• HCCI -- Homogeneous Charge Compression Ignition
• 6T40E 6-speed automatic transmission

But is it enough to beat the Civic?

Well, maybe it's time to think 3-cylinder!

A 3-cylinder engine has the following advantages over a 4-cylinder of the same displacement.

1. Lower frictional drag -- lower cylinder wall area, less valvetrain and reciprocating parts
2. Lighter and shorter block
3. Less parasitic loss from harmonic balancers (if equipped) -- one balance shaft at crank speed, vs two at 2 x crank speed.

In addition, the 2.4 liter Gen II Ecotec is a perfect candidate for building a 3-cylinder. The LE5 engine with one cylinder lopped off makes exactly 1.8 liters -- right where the Cobalt should be if it wants to play in the Civic/Corolla club. In addition, the HCCI engine with its tricky combustion management process is being engineered right now around the big Ecotec engine and keeping the same combustion chamber, piston and rods from the big 4-potters will reduce engineering time and costs.

Personally, I won't mind seeing a new Cobalt come on stage with a 3-cylinder Ecotec with mild hybridization (alternator-motor), Direct Injection HCCI and the 6T40 6-speed as standard. Such a 1.8 liter three-banger should deliver about 150hp, 140 lb-ft and help leap ahead of the Asians.

[CARBIZ]

Is this guy new? "GM has been sucking for years?"

My flame thrower is nearly out of fuel and I'd rather save it for a not so easy target....anyone else care to flame away?

[CaddyXLR-V]

Is this guy new? "GM has been sucking for years?"

My flame thrower is nearly out of fuel and I'd rather save it for a not so easy target....anyone else care to flame away?

he knows his stuff. I have no argument against it. IMO GM cars did suck in the 80s and most of the 90s. It is also true that GM can not compete on cost with most other automakers except maybe the germans.

[Dsuupr]

Though I would not totally agree with the sucking part, the Taurus was a 1986 model, I do understand the point of the 3-cyl. I wonder how marketing a 3-cyl would go, but on the other hand, maybe in todays high gas mileage minded group it might be more popular than a 4-cyl.

[PurdueGuy]

I wouldn't take the NG Cobalt down to a 3cyl option only, but I could see developing a really nice 3cyl for use in smaller than Cobalt cars, and offering it as an option or base (depending on how it compared to the 4cyl) in the NG Cobalt.

[Robert Hall]

Maybe a 3-cyl version of the Aveo (or it's replacement? ). Hasn't GM done a 3cyl Corsa in Europe? I could see a subcompact model with 3- and 4- cyl variants (like a modern Geo Metro).

[siegen]

while operating out of a high cost country (USA).

That statement isn't exactly correct, and I'm sure someone will elaborate on it more. Compared to the foreign makes that operate inside the USA, GM spends more on factories and labor. Given a clean slate I'm sure GM could cut down operating costs considerably inside the US.

R18A engine in the Civic for instance is not only a VVT unit

Just to clarify, it does not have the same VVT (continuously varying) style system that GM would employ. The change in lift, duration, and timing on one intake valve per cylinder is the only implementation of vtec on the R-series engines. The K-series has the continuously variable valve timing/phasing as well as traditional vtec system. There is much talk of the Civic EX getting a higher level engine than the LX/DX (like Accord has two 2.4L engines), and it would probably be a K20 based engine with 160hp. That is entirely speculation though.

[SuperSport623]

mmmmm three banger!

[ehaase]

I would not be surprised to see Ford develop a Ecoboost 3 cylinder engine for the Fiesta/Verve or even the Focus.

[gm4life]

Just no. That is something we all are trying to forget 3 bangers, if they do a 4cyl. there is no need for a 3 cylinder.

[biff]

GM wan't all bad in the 80s and 90s. What other company matched them in fuel economy with V6 cars? Their 'low tech' pushrods do pretty well when mated to a 4 speed auto. I had 2 large cars with a 3.8 and 3.1 that got over 30mpg.

[76ChevyTrucker]

You know, first moment I read the title to this thread, I immediately thought Geo Metro. Truthfully I think that the 3-cylinder idea would be ideal in something smaller than the Cobalt, but I know what he's getting at here, the Ecotechs are bigger than what Honda is currently delivering, yet in no way performing quite like what H has been spitting out for a while now. My only concern is that while the 3-cylinder Metros got great mileage, they would hit a pebble the right way and get stopped, just don't build an engine like that and I think it might be a winner.

[dwightlooi]

You know, first moment I read the title to this thread, I immediately thought Geo Metro. Truthfully I think that the 3-cylinder idea would be ideal in something smaller than the Cobalt, but I know what he's getting at here, the Ecotechs are bigger than what Honda is currently delivering, yet in no way performing quite like what H has been spitting out for a while now. My only concern is that while the 3-cylinder Metros got great mileage, they would hit a pebble the right way and get stopped, just don't build an engine like that and I think it might be a winner.

Of the many things that govern how "civilized" an engine is, three things matter the most. Engine vibrations due to harmonic imbalance, Engine virbations due to block flex and the number of power impulses coming from the engine per unit time for at a given RPM.

The ONLY advantage a 4-cylinder has over a 3-cylinder is having 33% more power impulses per unit time. That is you will see the same spacing of power pulses from a 3-potter as a 4-potter running at 33% greater speeds. In otherwords, at 1500 rpms the 4-cyl makes as many power pulses per second as a 3-cylinder at 2000 rpm. Similarly, a V6 or I6 at 1000 rpm already produces power pulses with equal proximity as a 4-cyl at 1500 rpm. This we cannot fix, but really do you feel that an engine is much smoother at say 2600 rpm compared to say 2000 rpm? Probably not.

This leads us to the other factors -- harmonic balance and block stiffness.

The 3-cylinder engine produces 1st order vibrations with a frequency equivalent to the speed of the crank producing an end-to-end rocking moment. This is caused be a shifting Cg of the reciprocating assembly. We can balance this out with a single balancer shaft turning at crank speed. The 4-cylinder engine produces 2nd order vibrations with a frequency twice that of the speed of the crank producing vertical up-down moments. This is caused by the fact that the pistons accelerate at different speeds* when coming down from TDC and those going up from BDC! We can balance this out with two contra-rotating balancer shafts turning at twice the crank speed. When un-balanced, the 4-cylinder has the advantage in that its 2nd order vibrations are less severe the 1st order vibrations in a 3-potter. In general, 4-potters at 1.8 liters or less tend to achieve decent civility w/o balancers, at 2.0 liters they sometimes have it, at 2.2 or greater displacement they almost always have it (one exception, I believe, being the pretty truck like Nissan SR24DE in the older Altimas and 240SXes and that is one pretty rough motor). With proper balancers however, both the 3-cyl and the 4-cylinder can be smooth and civil engines.

The issue of block stiffness is really quite independent of the number of cylinders. Wobbly blocks flex in response to pulsatile pressure applied on the crank journals. This leads to relatively low frequency vibrations that yell DIESEL! A stiff block resists flexing and comes through with a higher pitched metallic whirl which we so endear.

*If you don't believe that pistons move at different speeds at the top and bottom of their travel, draw a circle with a long line going through its middle. Mark say a 20 degree point clockwise at the two points past the bisecting line. Now draw two lines of the same exact length signifying the connecting rods from these two points to the center line. Note the difference between the distance covered by the pistons.

[haypops]

I once drove one of those 3 cylinder Geo metro with over 100,000 miles on it. Seemed perfectly adequate. Likewise I had a 4 cylinder ca. 600 cc car and it too seemed smooth enough.

Although this thread is interesting, I wonder if the choice of number of cylinders isn't secondary to vehicle weight. Since glass weighs more than sheet metal, the recent shift to super thick windshields may need reversing. Likewise a change to polycarbonate side curtains may be in the cards. I suspect that NVH will be sacrificed before vehicle size is.

[dwightlooi]

I once drove one of those 3 cylinder Geo metro with over 100,000 miles on it. Seemed perfectly adequate. Likewise I had a 4 cylinder ca. 600 cc car and it too seemed smooth enough.

Although this thread is interesting, I wonder if the choice of number of cylinders isn't secondary to vehicle weight. Since glass weighs more than sheet metal, the recent shift to super thick windshields may need reversing. Likewise a change to polycarbonate side curtains may be in the cards. I suspect that NVH will be sacrificed before vehicle size is.

Weight, not engine size, engine technology or #of cylinders is the BIGGEST contributing factor to fuel consumption.

However, having said that, a 3-cylinder engine will be more efficient, more compact and lighter than a comparable 4-cylinder of the same displacement. The reasons to increase the # of cylinders has nothing to do with fuel economy, engine packaging or weight. It is just you get more firing pulses with more cylinders and with large cylinder sizes it is harder to control engine knock because the distance from the spark to the furthest reaches of the cylinder wall is greater. However, a 1.8 liter 3-pot will have the same knock resistance as a 2.4 liter 4-pot which we know to be perfectly manageable.

[daves87rs]

I wouldn't take the NG Cobalt down to a 3cyl option only, but I could see developing a really nice 3cyl for use in smaller than Cobalt cars, and offering it as an option or base (depending on how it compared to the 4cyl) in the NG Cobalt.

Agreed. People should still have the choice....

All tech aside-if the 3cyl runs just like a 4cyl ( and peppy-it has to have enough power to get out of it's own way, at least) That would be fine by me....

[dwightlooi]

Agreed. People should still have the choice....

All tech aside-if the 3cyl runs just like a 4cyl ( and peppy-it has to have enough power to get out of it's own way, at least) That would be fine by me....

Figure a 1.8 liter 3-cylinder DOHC 12V engine to be about 127hp w/o DI and about 150hp w/DI some advanced features like Dual VVT. That's in the same bracket as the leading 1.8 liters in the FWD compact segment. Thats assuming regular 87 octane and the same architecture (bore/stroke, combustion chamber shaping, etc) as the LE5 2.4 liter 4-cylinder engine currently in use in the Malibu, G6 and others.

[A Horse With No Name]

Personally even if it did work out from a fuel economy standard, it would be a pr disaster.

Better to upgrade/update the AVEO and use it there.

Chris

[76ChevyTrucker]

dwight I was arguin the civility of the engine. I know that the engines can easily be made smooth no matter how many cylinders you put into the block, just as your other data suggests, I'm just saying that he Cobalt seems a little too big/heavy for a 3 cylinder. The Aveo would be a better candidate for such a powerplant, but again, that's just my thought process and sometimes its flawed (can't help it, I'm human).

[cmattson]

I think a 3-cyl engine would have a stigma associated with it. It's unusual - and coming from GM, who has a less-than-steller quality record the last 2 1/2 decades, I think you'd find reluctant acceptance. I think GM would be better off following Ford (read: Ecoboost) on this one: smaller 4cyl, with direct injection and some sort of boosting.

Take a look at the Pontiac Solstice (manual transmission):

The base, 2.4l delivers 177hp, getting 20/28 mpg

In comparison, the turbo, DI-2.0 delivers 260hp, getting 22/31

Almost 100 more hp while delivering roughly 10% better fuel economy. What if GM ran a DI-turbo setup on a 1.3l-1.5l engine?

[CaddyXLR-V]

I think a 3-cyl engine would have a stigma associated with it. It's unusual - and coming from GM, who has a less-than-steller quality record the last 2 1/2 decades, I think you'd find reluctant acceptance. I think GM would be better off following Ford (read: Ecoboost) on this one: smaller 4cyl, with direct injection and some sort of boosting.

Take a look at the Pontiac Solstice (manual transmission):

The base, 2.4l delivers 177hp, getting 20/28 mpg

In comparison, the turbo, DI-2.0 delivers 260hp, getting 22/31

Almost 100 more hp while delivering roughly 10% better fuel economy. What if GM ran a DI-turbo setup on a 1.3l-1.5l engine?

It would be expensive. The Cobalt and Aveo are not $28,000 cars like the SKY RL and Solstice GXP. The Cobalt needs to stay in the price range of $10,000-$20,000, which it can't do with a DI turbo engine.

I think the question comes down to:

How do you get 130-150hp, with the best fuel economy as possible, as cheaply as possible?

Edited January 24, 2008 by CaddyXLR-V

[ToniCipriani]

I want to see a Diesel before a 3-cylinder. The torquey 1.9L CDTI in a Cobalt should be pretty fun.

[dwightlooi]

I want to see a Diesel before a 3-cylinder. The torquey 1.9L CDTI in a Cobalt should be pretty fun.

How about a 1.8 liter 3-cylinder Dual-VVT direct injection turbodiesel? Should make in the vicinity of 150hp and 240 lb-ft of torque on Unleaded #2. Probably need urine injection to meet emission standards though...

[turbo200]

nothing but good info here. I think the American public going into the next decade will have to be conditioned to accept all modes of differing displacements, sizes, types of engines, so I don't see why this wouldn't be a fantastic idea. I totally agree with your methodology of going beyond the call to really finally actually be ahead of the competition and prove you are a foward-thinking player instead of a follower.

The issue with people here is that humanity as a whole creates affinities to something and then can't be told when they are wrong or that affinity is somehow flawed. they are backpedaling striving to defend the thing they put so much faith in, when to accept the flaw and move beyond it is really the better thing to do, you never know what you can achieve when you see the flaws and choose to improve on it.

dwightlooi, great post.