Effects of Compression ratio on torque
 

I'm in the process of planning my 3.0 top end upgrade and have questions regarding how an increase in compression affects torque. After reading Bruce's book, I understand that if I were to install a set of 98mm P&C's at the same compression ratio 9.3:1, I could expect a a 6.7% increase in power. I assume this to be an increase in torque and horsepower. I find it difficult justifying the $2,500 in parts to get that little an increase.

My current setup is a stock '82 SC engine w/ 10k on the rebuild. PMO 40s, Electromotive crank fire ignition, SSIs, 901 trans.

So now I'm considering bumping the compression ratio to get more power. I realize that anything over 10:1 compression requires twinplug and I'm thinking this may offer more performance at a more affordable upgrade to the 3.2 SS. Since this is mostly street driven with occassional autoX, I need to preserve low end drivabeility and have decided a S cam or DC40 GE40 will be best for my application. I know the smaller intakes are the restrictive setup here so I either need to machine the ports or get 3.2 heads. Correct me if I'm wrong, but porting my current heads runs pretty close to a decent used set right?

Having to do the heads anyway, tapping an additional plug won't run much more, so the only extra cost I can think of at this point is the ignition system, pistons and machining the lower valve covers. Matched JE pistons at 10.5:1 and an additional Electromotive coil pack will run around a grand. What am I missing?

Anyway, back to the post of the title. What effect will a bump of over 1 point of compression add to the torque of the engine?

On my last dyno run, the engine pulled 214 lbs at 4400 and 201 hp at 5800, all at the flywheel.

I know I lose low end from the cam profile and the larger intakes but will the compression add back what I've lost?

As always thanx for taking the time to read my tangent;)

There are calculators on the net that will tell you what you can expect for an increase in compression. What you will find is that it's not much. My guess is that going up 1 point on compression will yield less than 10 HP. Increasing the displacement will give you much more than increasing the compression. Changing the cams and pistons will not do much unless you port the heads. The big power increases will only come at the higher revs with any of these changes and for that you need the big ports. Our spec engines have their peak torque at 5000-5500 RPM and don't get to peak HP until 6500 RPM and that is with stock SC cams.

-Andy

That depends on how you're going to increase the compression ratio by 1 pt. (9.5:1 to 10.5:1). For a 911, that's usually a function of higher compression pistons and usually in conjunction with a larger bore piston set. If you're going with larger pistons anyway, I'd go ahead and up the CR. Increasing CR using other methods are limited due to physical limitations and will present other issues to resolve.

However, increasing CR is one of the few modifications that should yield increased torque and HP in addition to increased fuel mileage due to the increase in VE (volumetric efficiency). All this assumes detonation is kept under control with twin plug ignition and min. octane fuel.

Sherwood

Really, there are two types of compression to consider: static compression and "dynamic" or actual compression.

Static compression could/should be called theoretical compression because it is really a function of volume of the cylinder with the piston at BDC vs when the piston of the same cylinder is at TDC. And here's the important part - with BOTH valves completely closed. In reality, with most performance cams, the intake and exhaust valves are at least partially open though some of the cycle. This is referred to as overlap.

Static compression is a bigger consideration with cams that are very mild and emissions-friendly and/or intended for use with fuel injection as they will have little or no overlap. This means that if you have a theoretical/volumetric compression of 10:1 and no overlap... your compression would be in fact 10:1.

But "bigger" cams intended for use with carbs have a lot overlap - degrees of rotation where both valves are at least partially open - and this situation bleeds off cylinder pressure. Which is why, for more power, you can mate a high-overlap & lift cam with a high static compression ration and not detonate on pump gas.

Hypothetically, if you have a big enough cam, you could run a static compression of 11:1 or even higher - because at low RPMs you're really bleeding off a lot of cylinder pressure due to the long overlap period. This is also why big cams "come on" at certain RPMs - they only operate efficiently within a certain range of stroke time vs combustion time. The idea is the compress the mixture and ignite with spark it before it can heat up enough to detonate... and you can probably see that this can only happen when the cycle times are really, really short - i.e. at high RPM.

I know this does not specifically answer you question... but it's important to realize that the cam choice actually determines the final/actual compression ratio. Static compression numbers are only the start of trying to evaluate the HP gains of any motor.

But the short take is that compression alone won't help much. Increase your displacement (preferably with a longer stroke) first for torque, then optomize for HP with compression (and cam choice).

In my sig you can see one of my gone-but-not-forgotten cars is a 347ci Ford V8 Mustang. That's a 302 (5.0 liter) with an extra 1/2 inch of stroke. It was a torque beast... pretty close to 400 ft/lbs. I could break the tires loose rolling along in 3rd gear when above 3k RPMs. It's the only car I've ever driven that truly scared the snot out of me. :)

Your barrels can be opened up to 98mm, if they are Nikasil, and new JEs to match with the higher compression ratio will make for a better motor. If you are going to change out the pistons for the compression bump, you may as well go the 98mm route.

Do the twin plug and head upgrade.

Rather do it correctly once, and have a motor that is fun to drive.

I had a 3.2ss in my race car with 11:1, supercup cams, twin spark & throttles. This gave about 280+ HP with a flat torque curve all the way up to 7 000 rpm.

joe wrote,

Anyway, back to the post of the title. What effect will a bump of over 1 point of compression add to the torque of the engine?


Hello Joe,

The most accurate rule of thumb I've seen is a 1 point increase in compression results in a 4% increase in torque. Torque is directly related to displacement, it is very difficult to actually increase it. Most mods just move the torque peak to a higher rpm, resulting in a higher calculated hp.

Paul

Wow, I never knew the 95 mm Nikasils could be bored out to 98mm.

Does this weaken the cylinder walls at all?


Andy, I realize the biggest gain will come from flowing more charge through the intakes. And after reading some of the posts above, my concern is to maintain the torque I have after putting a cam with more overlap and an intake that is less responsive at lower rpms. My ultimate goal is to make more power in the higher rpms up to 6500-7k w/o losing grunt.

I read about 3.2 SSs making 240 ft/lbs torque and I think that is above the safe limit for first gear in my 901. And I like the 901 enuf to keep it.

Alusil cylinders can be bored as well, and plated with nikasil. Only the 3.2 95's can be bored, the 3.0 ones with the head gasket groove can not....

Cheers

So I'm back to square one:confused:

A bump of 1 point won't make a huge difference. For comparison, see how much torque/liter (more accurately the BMEP) curve shifted when Porsche moved from the high CR 2.2S to the lower CR 2.4S, or from the 2.0E or 2.2E to the 2.4E. Here's some charts that I made from the factories published data.
http://forums.pelicanparts.com/uploa...1199484148.jpg

My conclusions...
1) Bumping the CR will help the "off-cam" performance (aka to most people as "torque" or driveability) a lot more then the peak HP performance.
2) Note that the no matter what the CR, the torque curves seem to converge above the peak torque engine speed and track the same from there up to the peak engine speed. I think because in this zone of an engine's performance envelope, performance has more to do with combustion quality then in the compression pressures, assuming that you don't go over the top and start causing detonation. So in the case of the 2.2's and 2.4's, the heads are exactly the same, and apparently the differences in the piston dome are not enough to make a significant difference in the combustion chamber shape, and the resulting quality. Curiously, of the 5, the 2.0E has what most would consider to be the most compromised combustion chamber shape, not to mention the smallest valves. Never the less, it doesn't seem to have hurt it. Now in the case of your 3.0, the combustion chamber is flatter (and in the case of the 911's Hemi head, this means better from a combustion perspective), and so I would expect the change to be smaller still.

Joe - the question is at what RPM was that 3.2ss that you mentioned making 240ft/lbs of torque? That might give you some way to triangulate the approximate config of such a motor, and give you some clues about what you need to do in order to achieve that kind of goal.

I'd surely bet that your 240 quote it certainly was not down low... at say 2000 RPM. Maybe at or just before the HP peak... which for an SS would probably be above 6K.

-Troy

Troy,

Thats a great point. Peak torque on the SS was at 6100 rpm. I will have to check the torque curve of a 3.2. it's probably not much different than a 2.2 E which is making 85% of peak from 3000 rpm and above.

John,
Interesting charts and thanks for posting. However, the data would be easier to extrapolate and compare the difference if the baseline engines cited were more similar. Each engine has a different displacement (2.0, 2.2 2.4) via larger bore, some with longer stroke. What percentage of torque increase is based on increased stroke versus increased CR + other factors?

Sherwood

Sherwood, note that I've plotted the BMEP, not the torque, This correlates to torque/liter, the units would be different but the graphs the exact same if you were to plot torque/liter. Since peak torque is largely a function of displacement, I've divided out displacement. You can see how closely peak torque correlates with displacement since they all wind up almost equal once you divide out the displacement. This means that the differences in the engines' torque curves on the charts have to be the result of the other factors, because displacement has been removed from the data.

If you want to calculate the exact percentages, you'll need to do some sort of correlation and regression analysis. Having tried this, it's not as simple you might think.

Anyhow, as you mentioned, what's left for differences in the configurations that will affect the shape of the curve? Based on analysis that I've done in the past (do a search under my name for "selecting cams" or " intake velocity", I think the list of factors that you started would look like the following for air-cooled 911 motors.

1) Cams: In my opinion, cams are what drive the engines performance, and have the biggest impact on where in the rev range the torque will peak. Where the torque peaks and how long it stays up (see #2 below) then drives your HP numbers.
2) Intake port velocity: This needs to be "just right" for proper mixing while not choking off the engine. If the velocity is too high (the ports are too small) you'll never draw enough air to reach the HP numbers at peak RPM that the engine may be capable of. If the ports are too large, your engine will have a big hole in the torque curve when off-cam at low rev's. I believe that it's not so important to get the velocity numbers exactly right, as it is to avoid making too high or too low. If you avoid the wrong answer, then it will be OK. In the "S" chart above, you'll note that the 2.2S was lagging the 2.4 below peak torque RPM, and slightly better above peak torque RPM. Given that they have exactly the same heads and ports, the intake velocity on the larger 2.4S is slightly higher on the 2.2S. As a result I suspect that the increased turbulence from this increased velocity below the peak torque RPM point results in the 2.4S having better combustion quality. On the other hand, above the peak torque RPM point, the torque falls off because the intake charge is pretty much traveling as fast as it can, so the revs increase, the amount of charge that actually makes it into the combustion chamber during that every shortening time while the intake valve is open begins to decrease. With this reduction in intake charge is a corresponding reduction in torque. Here are a couple of charts that I made a few years back that compare the torque curves to the average intake velocity:

http://forums.pelicanparts.com/uploa...1199543539.jpg

Note that in the case of the "S" engines, the BMEP is above 160 PSI when the average intake velocity is between about 50 to 100 meters/second. In the case of Tony G's 3.0 running through 40 mm carbs (and I assume using venturis which are about 36 mm's), the engine is much fatter at lower rev's, and runs out of steam at higher revs. But he's also running a completely different cam from the "S" cam of the other engines. But the two extreme cases of the carb's '67 2.0s and 2.7 RS are interesting. Because of it's comparatively large ports the '67 2.0 S never even gets to 100 m/s, but at the low end it doesn't reach 160 PSI until the ports are flowing about 50 m/s. On other hand, the somewhat port restricted 2.7RS just hits 160 PSI when the parts are flowing 50 m/s, and passes 160 PSI again on the downside when the ports are flowing 100 m/s. Here's the same data, plotted differently that might be a little clearer.

http://forums.pelicanparts.com/uploa...1199544503.jpg

For comparison, here are the same charts for the "E" spec engines.

http://forums.pelicanparts.com/uploa...1199544525.jpg

http://forums.pelicanparts.com/uploa...1199544959.jpg

3) Stoke: This isn't as simple as "long stroke = more torque". If you divide out the change in capacity resulting from increasing the stroke, you'll most likely see the torque and HP numbers are the same if you increase the bore or increase the stroke. The operative thing with stroke is travel of the piston near TDC, and how this interacts with the intake flow and valve events. "Long stroke" engines tend to linger near TDC a little longer then short stroke engines, which allows the cam more time (in degrees) to open the valve before the piston reaches peak acceleration on the intake stroke. This can affect the shape of the torque curve if you can get everything to operate in synchronicity. Increased piston velocity on the other hand, in addition to increasing wear on the engine at higher rev's, can also end up drawing faster at certain parts of the crank rotation then the ports can flow at high, especially at high RPM which may negatively impact your peak HP and high RPM performance. Finally, longer stroke engines can get by with a smaller piston dome then short stroke engines, which results in more compact combustion chamber, which improves combustion and is less prone to detonation.
4) Valve size: Bigger valves allow more flow at lower lifts, which can make an engine act like it's got a "bigger" cam, potentially without sacrificing low RPM performance. The downside is that you can only fit a certain sized valves in a given head and the increased mass can result in an increased propensity to float the valves at high RPM. The increased flow may also result in less turbulence at low RPM's which will hurt your part throttle running at low RPMS. Once again it's not necessarily one is generally better then the other, it depends on the total combination of cam profile, intake velocity (aka: porting), stroke and combustion chamber design.

That's my $0.02

On the subject of 3.0 liter engines, the factory provides a useful example. The 3.0RSR race engine made 230 ft/lbs of torque. The only way you can exceed that on a road car engine would be to use a carpenter's pencil when drawing the torque curve.