A question about long-stroke or undersquare engines

[karabiner98k]

As i have read in several articles, it is obvious that short-stroke or oversquare engines can rev higher and produce more power and less torque and usually used in sports cars but long-stroke engines are good for high low end torque and can't rev as high as short-stroke engines.

But i have seen some cars that have quite rev-happy engines (8000RPM) with so much power despite being undersquare.

Two examples are BMW M3 (E46) and Lamborghini Murcielago LP640

BMW M3 (E46) I-6 3.2L

Bore x Stroke = 87 x 91 (mm)

Lamborghini Murcielago LP640 V12 6.5L

Bore x Stroke = 88 x 89 (mm)

So, the above cars have long-stroke engines but with so much revs and power.

How can these cars manage to have the characteristics of an oversquare engine while being undersquare?

[brewnog]

They're only a bit undersquare, particularly the Lamborghini!

Aside from simple engineering factors (like reducing reciprocating mass and balancing components), the main factor here is that both I6 and V12 engines have perfect primary and secondary balance, unlike a straight four or V8.

[karabiner98k]

the main factor here is that both I6 and V12 engines have perfect primary and secondary balance, unlike a straight four or V8.
Thanks for your answer.
Could you explain this fact a little?
Why I6 and V12 engines have perfect primary and secondary balance?

[S_Happens]

Typically textbooks start with the generalization that rpm is primarily limited by mean piston speed. While this is not exactly true, (mean piston speed by itself can barely hint at accelerations seen by the rotating equipment nor does it give any idication of the geometry) it's a good place to start.

[brewnog]

I can, but the Wikipedia article is a good place to start. Let me know if you get stuck.

http://en.wikipedia.org/wiki/Engine_balance

[Ranger Mike]

good one brewnog..also con rod length ratio!

[mender]

As i have read in several articles, it is obvious that short-stroke or oversquare engines can rev higher and produce more power and less torque and usually used in sports cars but long-stroke engines are good for high low end torque and can't rev as high as short-stroke engines.

How can these cars manage to have the characteristics of an oversquare engine while being undersquare?

When discussing oversquare vs undersquare, you have to talk about the same cylinder displacement, otherwise there is no basis for comparison. The discussion then becomes one of stroke length and other things.

[jack action]

Size wise, if you want to compare engines, only the bore area is important.
Speed wise, if you want to compare engines, only the mean piston speed is important.

Check this post (http://www.physicsforums.com/showpost.php?p=2770016&postcount=6) I've wrote on another thread.

The advantage of an undersquare engine is the possibility or creating higher compression ratio (which is why practically all diesel engines are undersquare).

The advantage of an oversquare engine is a smaller engine (volume wise, hence weight wise as well).

[Lsos]

I'm surprised you didn't mention most extreme example that I know of: Honda/ Acura Integra type R (B18C5)

4 cylinder
Bore: 81mm
Stroke: 87.2mm
Redline: 8,500rpm

The simple answer to your question is engineering. If you use better materials and better engineering practices, you can end up with some extreme characteristics.

[mender]

Just think, if they'd only made the engine oversquare and bumped the piston size up to 95 mm or so the engine would have made much more power.

So why aren't newer engines oversquare?

[xxChrisxx]

Just think, if they'd only made the engine oversquare and bumped the piston size up to 95 mm or so the engine would have made much more power.

So why aren't newer engines oversquare?

No low down torque (which is what most people use for every day driving), as mentioned above.

If you meant keeping the stroke the same but increasing bore - then you are incraesing displacement...

[mender]

No low down torque (which is what most people use for every day driving), as mentioned above.

If the engine specs are the same other than the bore to stroke relationship, an oversquare engine will have the same torque as an undersquare engine.

So why aren't newer engines oversquare?

[Lsos]

I'm not enough of an expert to comment too much on your last post, but I believe undersquare do generally favor/ allow for more torque.

Whatever the case may be, sometimes there are other driving factors behind an engine design. The Acura/ Honda Type R that I mentioned, for example, is derived from the Integra GSR engine, which while still having the same dimensions, redlined at a lower 8000rpm and made about 25hp less. The engineers wanted a factory tuned "Type R" version of this car, and so they set themselves at squeezing more out of this engine. Despite its extreme characteristics, the engine performs extrmely well and reliably, and it was simply easier (and less expensive) to make more power this way than to completely redesign the entire engine/ car from scratch.

[mender]

... but I believe undersquare do generally favor/ allow for more torque. That is a myth that just won't die. Oversquare engines have the advantage in both torque and power production.

So why are newer engines designed to be undersquare? The original question that started this thread hasn't been answered yet.

[xxChrisxx]

If the engine specs are the same other than the bore to stroke relationship, an oversquare engine will have the same torque as an undersquare engine.

So why aren't newer engines oversquare?

I think you'll find you are a bit off there. There are a coupe of reasons why stroked engines produce more torque, they are in Heywoods book on engine fundamentals.

One reason is that force decays away far quicker in an oversquare engine. So although you get a higher peak force value acting down the cylinder axis, most of the force is going into bending the crank rather than turning it. You get a pressure 'spike' close to TDC that decays away quickly, in a long stroke engine you get a force that decays away slower as the valume change per degree of crank angle is less.

There are a few other reasons but i'd have to read up on it in Heywood.

Think about it, if what you are saying WERE the case, all engines would have F1 style bore to stroke ratios. ie huge bores and tiny strokes. There would simply be no downside to doing it. It's for the very reason that they don't produce the same torque that engines are not oversquare.

You also can't really nail this down to one variable either, there are millions of interconnected reasons, as in an engine every variable affects almost every other variable.

[mender]

I think you'll find you are a bit off there. There are a coupe of reasons why stroked engines produce more torque, they are in Heywoods book on engine fundamentals.

One reason is that force decays away far quicker in an oversquare engine. So although you get a higher peak force value acting down the cylinder axis, most of the force is going into bending the crank rather than turning it. You get a pressure 'spike' close to TDC that decays away quickly, in a long stroke engine you get a force that decays away slower as the valume change per degree of crank angle is less.

There are a few other reasons but i'd have to read up on it in Heywood.


If the rod/stroke ratio is the same for both engines, the volume change/degree will be the same and pressure decay will also be the same.

.
It's for the very reason that they don't produce the same torque that engines are not oversquare.

Any other explanations as to why you think that? I see that statement a lot; I also build a lot of engines and dyno them. Did a new engine combo for a racing team just a few weeks ago that proves this wrong. Reduced the engine size by 5% to get a weight break yet lost only 2% torque and 1% power in the same rpm range despite increasing the bore to stroke ratio by 10%.

By the usual oversquare reasoning it should have lost at the minimum 5% torque output because of the displacement change and more because of the bore/stroke change.

.
Think about it, if what you are saying WERE the case, all engines would have F1 style bore to stroke ratios. ie huge bores and tiny strokes. There would simply be no downside to doing it. It's for the very reason that they don't produce the same torque that engines are not oversquare.

You also can't really nail this down to one variable either, there are millions of interconnected reasons, as in an engine every variable affects almost every other variable.

If power and torque production were the only considerations, every engine would be designed that way. There is another consideration that is more important; any guesses as to what that is?

[jack action]

I suspect that undersquare engine is limited by weakened parts. If the piston bore is increased, its height must be too to prevent rocking. But because the stroke is not increased, the piston pin has to move closer (in proportion) to the bottom edge of the piston to respect crankshaft clearance. The higher forces due to the larger piston are not supported by the (still the same) amount of material between the pin and the bottom edge.

Just an idea, nothing to support it.

I also build a lot of engines and dyno them. Did a new engine combo for a racing team just a few weeks ago that proves this wrong. Reduced the engine size by 5% to get a weight break yet lost only 2% torque and 1% power in the same rpm range despite increasing the bore to stroke ratio by 10%.

By the usual oversquare reasoning it should have lost at the minimum 5% torque output because of the displacement change and more because of the bore/stroke change.

Mender, can you tell me more about this engine. I'm in a debate on another forum about the fact that power is only dependent on bore and that stroke don't change anything (like I say in this post (http://www.physicsforums.com/showpost.php?p=2770016&postcount=6)). It looks like you have just prove that with your engine (it seems you only shorten the stroke without changing the bore, and by getting the same HP you are proving my point). Can you confirm that and give some numbers?

On the other forum, math is not seen as valuable tool to prove something, they want actual measurements.

[mender]

The engine dimensions changed from 4.040 x 3.5 to 4.100 x 3.25; displacement respectively was 359 and 343, power output was 668 hp and 664 hp @ 7500 rpm, torque was 505 and 495 @ 6100 and 6200 rpm. Rod length was 6" for both, so R/S ratio changed slightly. Piston pin was lower by.125" because of the shorter stroke, piston weight was very close to the same. CR was 13.5:1 for both, and same quench. Hp per cube went from 1.86 to 1.94, torque per cube at peak was 1.41 and 1.44; not bad considering the application and budget. Both power and torque per cube improved by going more oversquare.

Rpm is limited to 8200 for valve train reliability (road course engine), so I couldn't take full advantage of the shorter stroke. I backed off on the cam specs ever so slightly as well, 2 degrees less intake duration and .010" less lift which likely accounts for the slight drop in hp. Everything else stayed the same; heads, intake, carb, exhaust system, ignition, air cleaner, oil pan and dry sump system.

I looked at your other post; to put it simply, the engine is an air pump and the more air that goes through, the power it makes. And since power is related to time, the more strokes per minute, the more power the engine makes as well. It's usually cylinder head flow that determines the hp limit.

Pistons can be quite short without strength issues, it's more about supporting the ring package properly. Given the advances in piston and rod design, the main limiting factor now is the valve train (as mentioned in your other post) and that is coming along nicely as well. It wasn't that long ago that 10,000 rpm was an incredible number; now we have production car engines that are approaching that.

[jack action]

Thank you mender, that is very helpful.

[mender]

I think you'll find you are a bit off there. There are a coupe of reasons why stroked engines produce more torque, they are in Heywoods book on engine fundamentals.

One reason is that force decays away far quicker in an oversquare engine. So although you get a higher peak force value acting down the cylinder axis, most of the force is going into bending the crank rather than turning it. You get a pressure 'spike' close to TDC that decays away quickly, in a long stroke engine you get a force that decays away slower as the valume change per degree of crank angle is less.

There are a few other reasons but i'd have to read up on it in Heywood.


Okay, I have a copy of Heywood's now; do you have page numbers? I've been leafing through the 900+ pages but haven't seen anything specific yet.

[brewnog]

First couple of chapters will have you sorted. It's actually quite a good bedtime book.

[jack action]

One reason is that force decays away far quicker in an oversquare engine. So although you get a higher peak force value acting down the cylinder axis, most of the force is going into bending the crank rather than turning it. You get a pressure 'spike' close to TDC that decays away quickly, in a long stroke engine you get a force that decays away slower as the valume change per degree of crank angle is less.

There are a few other reasons but i'd have to read up on it in Heywood.

Think about it, if what you are saying WERE the case, all engines would have F1 style bore to stroke ratios. ie huge bores and tiny strokes. There would simply be no downside to doing it. It's for the very reason that they don't produce the same torque that engines are not oversquare.

You also can't really nail this down to one variable either, there are millions of interconnected reasons, as in an engine every variable affects almost every other variable.

If the rod/stroke ratio is the same for both engines, the volume change/degree will be the same and pressure decay will also be the same.

If power and torque production were the only considerations, every engine would be designed that way. There is another consideration that is more important; any guesses as to what that is?

First, Chris, as mender pointed out, you're wrong about stating that bore/stroke ratio have something to do with the decompressing rate with respect to crankshaft angle. If you do the math, they are only 2 variables that affect that:
rod length/stroke ratio (smaller --> higher decompressing rate) and compression ratio (higher --> faster decompressing rate).

After thinking about it and doing a simple analysis, the bore/stroke ratio is limited by the surface-to-volume ratio (SVR) of the combustion chamber. For a given bore (which determines the power potential of your engine), the SVR increases with the bore/stroke ratio (higher SVR means more combustion chamber cooling - hence, energy loss - and creation of "dead spots" where the fuel mixture doesn't burn). By the same token, it should also reduce the squish zone. The problem gets worst as the CR increases (that's why diesel are normally undersquare).

[brewnog]

Sorry Jack, you're wrong here. Power is dependent on both bore and stroke. The quoted book would be your best source of enlightenment.

[xxChrisxx]

Okay, I have a copy of Heywood's now; do you have page numbers? I've been leafing through the 900+ pages but haven't seen anything specific yet.

My copy is back in Manchester, I've recently moved house so ive not shifted all my engineering books yet. It should be pretty near the front though.

I haven't had chance to read back up on it yet, it's why i've not posted anything on this.

[mender]

Okay, I've read the first two chapters and browsed most of the rest. He describes the relationships but doesn't make any claims that I saw. Let me know when you've located your book, I'm curious.

The overriding consideration that I was alluding to is emissions and fuel efficiency, both of which are better in undersquare engines and is a major reason for that trend in production engines even in the more exotic applications. Another significant advantage is reduction in octane sensitivity, allowing a higher specific output on pump gas.

Power and torque production are better in oversquare engines, which is why the vast majority of racing engines follow that trend.

[mender]

Sorry Jack, you're wrong here. Power is dependent on both bore and stroke. The quoted book would be your best source of enlightenment.

I think Jack is referring to power being dependent on bore size and mean piston speed.

[brewnog]

I think he's referring to power being dependent on bore size and mean piston speed.

Fair enough, but I'd picked up on this further up the page...

I'm in a debate on another forum about the fact that power is only dependent on bore and that stroke don't change anything (like I say in this post).

[mender]

I interpreted that to mean that stroke in itself doesn't dictate a change in the torque curve other than by the obvious displacement difference. I assume that you read his post.

Maybe Jack can expand on his thinking.

[jack action]

I interpreted that to mean that stroke in itself doesn't dictate a change in the torque curve other than by the obvious displacement difference. I assume that you read his post.

Maybe Jack can expand on his thinking.

I'm not sure I understand clearly on what I should expand, because I always thought it was pretty obvious, so I will just repeat myself. In SI unit the power is define as such:

Power = torque * rpm

Torque can be related to the engine displacement & BMEP; and rpm can be related to mean piston speed (vp) and stroke (S):

Power = (BMEP * Engine displacement / (4 pi) ) * (vp / S * pi)

Replacing the engine displacement by its definition based on # of cylinders (N), bore (D) and stroke:

Power = (BMEP * (Ap * S) / (4 pi) ) * (vp / S * pi)

where: Ap = pi/4 * D² * N

The stroke from the engine displacement cancel the one from the rpm, leaving:

Power = BMEP * Ap * vp / 4

So, physically, only the total bore area (bore & # of cyl) is of importance. If you increase the stroke but do not reduce the rpm, then vp will increase and that will put the engine in an all new category, not comparable to the previous one. Furthermore, once you will reach the 30 m/s mark, you won't be able to increase vp further as your piston will travel faster than the flame front.

If you increase the stroke but keep the same vp, you will get the same power but the torque will increase because the rpm will be reduced. So, if you change only the stroke from Sold to Snew and keep the top end components the same (camshaft & cylinder head), you will basically only see a shift of the power curve by a factor Snew / Sold (the intake & exhaust lengths are tuned with rpm, so it may make some little difference if the change is too radical and the engine is well tuned). But it doesn't change anything as it is only a matter of choosing a proper gear ratio to go back to the original values of torque & rpm.

IMHO, stroke is basically a gear ratio that is more efficient. Of course, long stroke means bigger engine and more weight so it has to be taken into consideration as well.

And there is no magic in this as:

volume flow rate = Ap * vp / 4

So:

Power = BMEP * volume flow rate

Which is true for any engine (4-stroke, 2-stroke, CI, SI and even gas turbine).

Yes, a bigger engine draws more air per rev, but it also rev slower, so in the end it has the same flow rate.

[mender]

That's what I understood you to say before, Jack, but thanks for restating it.

As has been said, there are quite a few variables that add up to a particular torque curve but the bore/stroke ratio doesn't have a direct effect on the shape of the curve.

[mender]

My copy is back in Manchester, I've recently moved house so ive not shifted all my engineering books yet. It should be pretty near the front though.

I haven't had chance to read back up on it yet, it's why i've not posted anything on this.
It took a while, I've been busy with many things but here it is.

Internal Combustion Engine Fundamentals by Heywood, page 824:

"Thus for well-designed engines, where the maximum values of mean effective pressure and piston speed are either flow limited (in naturally aspirated engines) or stress limited (in turbocharged engines), power is proportional to piston area and torque to displaced volume."

Reinforcing what I said earlier:
If the engine specs are the same other than the bore to stroke relationship, an oversquare engine will have the same torque as an undersquare engine.

The oversquare engine will have slightly less friction hp loss so it will produce slightly more net torque than an undersquare engine with the same displacement, again from Heywood, confirming that:
Oversquare engines have the advantage in both torque and power production.

Just in case anyone was still wondering.

[jack action]

It took a while, I've been busy with many things but here it is.

Internal Combustion Engine Fundamentals by Heywood, page 824:

"Thus for well-designed engines, where the maximum values of mean effective pressure and piston speed are either flow limited (in naturally aspirated engines) or stress limited (in turbocharged engines), power is proportional to piston area and torque to displaced volume."

Reinforcing what I said earlier:


The oversquare engine will have slightly less friction hp loss so it will produce slightly more net torque than an undersquare engine with the same displacement, again from Heywood, confirming that:


Just in case anyone was still wondering.

I agree with the citation from Heywood. But when you say «If the engine specs are the same», you have to specify what engine specs. Because if mean piston speed is considered (and it should), oversquare or undersquare, the friction losses are the same for both.

I don't understand why people are still using displacement as a viable yardstick to compare engines. Displacement is composed with 2 independent variables: the bore area and the stroke. Multiplying the bore area by 2 will increase power and torque by 2 as well. But multiplying the stroke by 2 will increase the torque by 2 but do nothing for the power. Why? Because the stroke is essentially a «gear ratio» where the linear force of the piston is converted to rotational torque of the crankshaft. Yes, you can double the torque by doubling the stroke, but you will have to halve the rpm as well; just like a gear set does.

Fundamentally, it goes as follow:

The power of the piston, with its linear motion, is Force X Velocity where the Force = Pressure X Bore area. That's it. Once you are there, the only thing you can do is transform that power between force and speed, in linear form or rotational form. To find the torque, all you do is an energy balance for one cycle where:

Erotational = Elinear

crankshaft torque X angular displacement = piston force X linear displacement

The angular displacement per cycle is 1 rev for a 2-stroke and 2 revs for a 4-stroke. The linear displacement is the stroke length (because the piston force is defined as the mean effective pressure acting during the power stroke only).

In the end:

power = BMEP X bore area X mean piston speed / # of stroke per cycle

torque = BMEP X bore area X stroke / # of rev per cycle

Where it «just happens» that bore area X stroke is the engine displacement.

That being said, I know that bore/stroke ratio has an influence on the combustion chamber shape and probably also on some stress factor on the mechanical parts. That is in those fine details that a discussion on bore/stroke ratio becomes interesting (where it is BMEP and maximum mean piston speed that will be affected).

[mender]

I agree with the citation from Heywood. But when you say «If the engine specs are the same», you have to specify what engine specs. Because if mean piston speed is considered (and it should), oversquare or undersquare, the friction losses are the same for both.

That being said, I know that bore/stroke ratio has an influence on the combustion chamber shape and probably also on some stress factor on the mechanical parts. That is in those fine details that a discussion on bore/stroke ratio becomes interesting (where it is BMEP and maximum mean piston speed that will be affected).

You're preaching to the choir if your post is only directed at me.:smile: And you basically reiterating what Heywood stated.

And I did specifically specify the specifications here:
"The oversquare engine will have slightly less friction hp loss so it will produce slightly more net torque than an undersquare engine with the same displacement, again from Heywood, confirming that: "

[jack action]

You're preaching to the choir if your post is only directed at me.:smile: And you basically reiterating what Heywood stated.

And I did specifically specify the specifications here:
"The oversquare engine will have slightly less friction hp loss so it will produce slightly more net torque than an undersquare engine with the same displacement, again from Heywood, confirming that: "

I'm just thinking out loud, not trying to convert anyone in particular. I just think it's misleading to give so much importance to displacement when trying to determine engine performance.

Where I differ from Heywood is that it's not the bore/stroke ratio that gives the advantage, it's the increase in bore area. 2 engines with the same bore area, one with a bore/stroke ratio of 1:1 and the other of 2:1, will produce the same power output and are equivalent on my point of view. The fact that one has a displacement or a bore/stroke ratio twice as big as the other is irrelevant. Even if the smaller engine has less torque and a higher rpm, it's nothing a gear set with a gear ratio 2:1 won't cure.

Where does the «less friction hp» comes from for an oversquare engine? The only way I can think of is by keeping the same connecting rod length, which implies that the rod/stroke ratio would increase, reducing lateral forces. Is that what you mean? Because from what I know, everything being equal, only the mean piston speed should affect the friction losses.

[mender]

Where does the «less friction hp» comes from for an oversquare engine? The only way I can think of is by keeping the same connecting rod length, which implies that the rod/stroke ratio would increase, reducing lateral forces. Is that what you mean? Because from what I know, everything being equal, only the mean piston speed should affect the friction losses.

Since torque is proportional to displacement, it's only reasonable to compare the torque output of the engines using the same CFM passed through the engine and therefore the same rpm.

For a given displacement, the mean piston speed will be lower at any rpm for an oversquare engine but the gross torque output will be the same as the undersquare engine. Friction losses will be lower at the same rpm because the piston speed will be lower, meaning that the net torque output will be higher for the oversquare engine.

R/S ratio is assumed to be the same.

[jack action]

I see your point.

That's what I meant by «If the engine specs are the same»: I assumed the mean piston speed was constant, you assumed volumetric flow was constant.

On your point of view, you get a slight increase in torque and power because of the smaller friction losses;

On my point of view, I get exactly the same torque but a huge increase in power because of the increase in volumetric flow.

I enjoy looking at things through different eyes! :surprised