55 replies to this topic

[MatsNorway]

In response to bob riebes post in the crazy racing engines i just wanted to ask if the intake angles shown in the article is the norm?

I would say i find them sligthly dissapointing. the intake into the head is nearly 90 degrees


GSX R motor for comparison.
http://www.bikewalls...27_1024x768.jpg

The reason i am disapointed is that since there is no OHC and only two valves you would think they where able to put in some propper intake channels.

Does anyone have good cutaways of V8s with better angle preferably of the old school american type of engines.

Edited by MatsNorway, 07 September 2012 - 15:37.

[Tony Matthews]

Does anyone have good cutaways of V8s with better angle preferably of the old school american type of engines.

Check Dave Kimble, he's done a lot of production V8s.

[Lee Nicolle]

In response to bob riebes post in the crazy racing engines i just wanted to ask if the intake angles shown in the article is the norm?

I would say i find them sligthly dissapointing. the intake into the head is nearly 90 degrees


GSX R motor for comparison.
http://www.bikewalls...27_1024x768.jpg

The reason i am disapointed is that since there is no OHC and only two valves you would think they where able to put in some propper intake channels.

Does anyone have good cutaways of V8s with better angle preferably of the old school american type of engines.

The intake port does a near 90 deg bend to the valve,, as does the exhaust port doing a 90 away from the valve. But with the correct radius on the ports does not hurt flow at all.
Original 60s ports killed flow, often deliberatly to make the engine more driveable at low engine speeds. But those large cube drag engines are not too worried about that!!
Really in that sort of scenario there is no real advantage at all for OHC, in fact probably disadvantages.
A pushrod V8 is often slammed by the uneducated as being old fashioned and innefficient. Far from the truth. look at Nascar engines, 358ci, nearly 10000rpm and do it for 500 miles. And they seldom ever have a problem, And make huge power.
Even the true production based modified Chevs and Fords and Mopars that I play with are more efficient than most so called hi tech engines. And cost far less to play with too. The turbo engines make lots of power, often totally unuseable and are generally a handgrenade with the pin pulled.
Bike engines make plenty of power, but bugger all torque. Useless to drag around a 4 + passenger car at low rpm.

[Magoo]

In response to bob riebes post in the crazy racing engines i just wanted to ask if the intake angles shown in the article is the norm?

I would say i find them sligthly dissapointing. the intake into the head is nearly 90 degrees

Yes, that's the setup. On 2V pushrod racing V8s, generally speaking (NASCAR and NHRA Pro Stock) the more you stand up the valve toward vertical the better it works. In the current regime intake valve angles are in the area of the neighborhood of +/-10 degrees relative to cylinder CL. You can go even shallower and gain gross airflow but with carburetors, you lose fuel suspension and pump too much gasoline out the exhaust pipe.

This may all seem counterintuitive but a) airflow is not intuitive and b) there is a whole lot more to consider beyond short side radius.

[Bob Riebe]

The articles, especially the Goodwin one, if you read beyond the first picture page has some detail why they did what they did.

Any inline, whether four or six or even eight, does not have the packaging problems of a Vee engine but at the same time there have been V-8s with more straight shots to the valves but it would seem the current new engines are where they are for a reason.

The one article, and I forget which one, says one problem with carbs is there are none really big enough for the engines with 5.2 and bigger bore centers. One reason some are developing fuel injection systems.
Never thought I would ever hear that.

The Pro Mods and the ADRL have really given big block engines a new lease on life. There are new LS blocks designed to be taken over five hundred inches; Indy Cylinder Head produces new aluminum AMC blocks designed for over five hundred inches.
I am still amazed how they can get these long stroke engines to rev into and beyond territory that in the seventies was iffy for 305 Trans-Am engines.

Edited by Bob Riebe, 08 September 2012 - 15:20.

[Grumbles]

Current ProStock engines are producing from 2.87 to 2.9hp per cubic inch (nearly 180hp/litre) using 2 valves per cylinder, pushrods and rockers, carburetors and gasoline. And with cylinders the size of coffee tins. I find these engines - but even moreso the people that develop them - very remarkable.

[Kelpiecross]

[quote name='Lee Nicolle' date='Sep 8 2012, 10:35' post='5904486']


nearly 10000rpm and do it for 500 miles.

I have read that if the RPM wasn't limited by mandated diff ratios the max RPM would be more like nearly 11,000 - which is even more remarkable.

[bigleagueslider]

The pushrod OHV engines used in NHRA, NASCAR, USAC, etc. are legacies of the production-based engines these race series relied on years ago. With NASCAR, the valve cant angle, manifold flange location, cam shaft center location, etc. are all tightly regulated.

With drag race engine blocks/heads/manifolds there is a bit more freedom. But one design concern with these pushrod OHV V8's is the clearance between the intake port wall and the pushrod. This usually results in a tall, narrow intake port cross section. The combustion chamber geometry resulting from an off-center single large intake valve and single spark plug usually benefits from a bit of tangential intake charge swirl.

You should not be too harsh with regards to the performance of OHV V8's. They are still quite good in terms of weight, package size, cost, and even fuel efficiency.

[Wolf]

I'd like to make fun of those 'antiquated concepts', but then I remember I drive a car with air-cooled OHV I2 engine and love it to bits... And the engine doesn't even have a distributor!

[Bob Riebe]

If you check the history of the internal combustion engine "antiquity" has absolutely nothing to do with it, period.

Now less complexity, that separates the push-rod from the OHC.

Edited by Bob Riebe, 11 September 2012 - 18:46.

[Robin Fairservice]

According to Dave McLellan in his book "Corvette Inside", the Chevrolet Engine people were able to equal the Lotus four cam engine with a push rod engine that had less weight and a smaller width which made installation easier. He claimed that for every one pound of engine weight, the car became two pounds heavier, so you woukd need more power for similar performance.

[carlt]

of course -
that'll be the reason why all modern single seat race cars use pushrod engines

[desmo]

If the design of modern single seat race cars weren't arbitrarily constrained by engine displacement limits that favor specific outputs over almost all else, compact lightweight pushrod engines would make considerably more sense.

[Magoo]

According to Dave McLellan in his book "Corvette Inside", the Chevrolet Engine people were able to equal the Lotus four cam engine with a push rod engine that had less weight and a smaller width which made installation easier. He claimed that for every one pound of engine weight, the car became two pounds heavier, so you woukd need more power for similar performance.

Sure, simply compare the C4 Corvette with 32V LT5 V8 to the C5/C6 Corvette with LS series V8. The pushrod V8 wins big time.

Couple of things, though... that was not the sharpest DOHC V8 that ever came down the pike, especially by today's standards. And really, the final day of reckoning is coming for the pushrod V8.

Edited by Magoo, 11 September 2012 - 21:13.

[Bob Riebe]

And really, the final day of reckoning is coming for the pushrod V8.

If the vee eight dies possibly, if the vee eight does not die it is not gonna happen.

Unless highway speeds get so high vee eight engines have to spin in five figures, there is no reason in the U.S. to give up the simplicity plus weight and space savings of the push-rod engine.

Fords mod engine is huge and needs blowers to equal the push-rod engines in horsepower.
Not a good selling point.
Make the bore spacing larger and a huge engine become humungous.

[Robin Fairservice]

If the design of modern single seat race cars weren't arbitrarily constrained by engine displacement limits that favor specific outputs over almost all else, compact lightweight pushrod engines would make considerably more sense.

But the push rod V8 Corvettes have been very succesful in sports car racing in recent years.

The modern Corvette road cars are surprisingly fuel efficient considering the engine size and performance.

[Magoo]

Unless highway speeds get so high vee eight engines have to spin in five figures, there is no reason in the U.S. to give up the simplicity plus weight and space savings of the push-rod engine.

Pumping losses.

[bigleagueslider]

Pumping losses.

Magoo,

Pushrod V6/V8 engines were some of the first production engines to employ displacement-on-demand designs using collapsible tappets. This is now a very reliable and cost effective way to reduce throttling losses.

[Grumbles]

Another advantage that is often overlooked with 4-valve engines is high-rpm valvetrain durability. A 4-valve engine will survive quite nicely at highish rpms with very modest valvespring pressures. A traditional 2 valve pushrod engine requires much more lift and much more spring to maintain the same speeds, with the result that it simply isn't as reliably trouble free at sustained high rpms. That a Nascar valvetrain survives as well as it does is amazing, though the valvetrain alone is probably worth more than some cars I've owned... and the spring pressures used in Prostockers are staggering.

[Bob Riebe]

Pumping losses.

If an engines pumping loss ever becomes a serious reason it is or is not used, for legal reasons, society has become a bunch of pissant morons who are so busy dropping their pants and squating when politicians say crap, what they drive will matter little.

[bigleagueslider]

Another advantage that is often overlooked with 4-valve engines is high-rpm valvetrain durability. A 4-valve engine will survive quite nicely at highish rpms with very modest valvespring pressures. A traditional 2 valve pushrod engine requires much more lift and much more spring to maintain the same speeds, with the result that it simply isn't as reliably trouble free at sustained high rpms. That a Nascar valvetrain survives as well as it does is amazing, though the valvetrain alone is probably worth more than some cars I've owned... and the spring pressures used in Prostockers are staggering.

Grumbles,

The DOHC 4-valve is definitely best for a race engine. But for a production auto engine that spends 80% of its life below 2000rpm, it's not quite so important. Regardless, the DOHC 4-valve does give slightly better performance than a pushrod OHV 2-valve engine in terms of combustion/thermal efficiency. The lower component inertias and spring forces in the DOHC 4-valve also result in lower friction losses in the valvetrain.

[Bob Riebe]

Having read that there are now also automobile/truck/tractor etc. comp. engines, that can be put in a passenger vehicle if one so chooses, with five inches and four hundred thousandths bore spacing, at what size does this hit its lime. Both mechanically and simple bulk-wise?

Edited by Bob Riebe, 12 September 2012 - 21:27.

[Magoo]

Magoo,

Pushrod V6/V8 engines were some of the first production engines to employ displacement-on-demand designs using collapsible tappets. This is now a very reliable and cost effective way to reduce throttling losses.

Not good enough. With four camshafts and multiple intake and exhaust valves per cylinder you can do so much more.

[Lee Nicolle]

If the vee eight dies possibly, if the vee eight does not die it is not gonna happen.

Unless highway speeds get so high vee eight engines have to spin in five figures, there is no reason in the U.S. to give up the simplicity plus weight and space savings of the push-rod engine.

Fords mod engine is huge and needs blowers to equal the push-rod engines in horsepower.
Not a good selling point.
Make the bore spacing larger and a huge engine become humungous.

The Ford OHC engine is just a big lump. Goes ok but too heavy and bulky. And superchargers add to the complexity. The GM pushrod V8 is far simpler and has more power. And the economy is the same for both. Which is better in hoighway cycle than a lot of 4 cyl engines.
Ford went from the tiny little Windsor to the modular lump. In the Aussie Falcon it does nothing for the handling.
Though the Windsor was lacking a bit in the power dept in the end.

[Bob Riebe]

Not good enough. With four camshafts and multiple intake and exhaust valves per cylinder you can do so much more.

While it costs more to build, repair and takes more space in already pathetically crowded engine bays.
No real reason for it before and none now as far as passenger vehicles go.

[Wuzak]

The Ford OHC engine is just a big lump. Goes ok but too heavy and bulky. And superchargers add to the complexity. The GM pushrod V8 is far simpler and has more power. And the economy is the same for both. Which is better in hoighway cycle than a lot of 4 cyl engines.
Ford went from the tiny little Windsor to the modular lump. In the Aussie Falcon it does nothing for the handling.
Though the Windsor was lacking a bit in the power dept in the end.

Well, the one used by Ford Australia was the truck engine...

The current engine, the Coyote, is different and quite a deal lighter, I believe. In fact, it weighs much the same as the LS-whatever it is up to now in the Commonbore.

[Lee Nicolle]

Well, the one used by Ford Australia was the truck engine...

The current engine, the Coyote, is different and quite a deal lighter, I believe. In fact, it weighs much the same as the LS-whatever it is up to now in the Commonbore.

Yes, it is the 5.4 not the asthmatic 4.6. And the 4.6 is a lot bigger than the Windsor too, and heavier still. From an alloy engine.

[Bob Riebe]

The Coyote is nothing more than a heavily worked over alloy mod.

Still really the same engine; while the alloy ones we have had for a long time, were never that heavy.

[Wuzak]

Yes, it is the 5.4 not the asthmatic 4.6. And the 4.6 is a lot bigger than the Windsor too, and heavier still. From an alloy engine.

Wasn't the 5.4l used in teh Falcon the iron block version?

[Magoo]

While it costs more to build, repair and takes more space in already pathetically crowded engine bays.
No real reason for it before and none now as far as passenger vehicles go.

Four-valve DOHC V8s will obsolete two-valve pushrod V8s just as four-valve DOHC inline fours have obsoleted two-valve pushrod inline fours.

And just as four-valve DOHC V6s obsoleted two-valve pushrod V6s.

It's only a matter of time.

There is no sense in arguing with me about it. I didn't do it.

[Bob Riebe]

Four-valve DOHC V8s will obsolete two-valve pushrod V8s just as four-valve DOHC inline fours have obsoleted two-valve pushrod inline fours.

And just as four-valve DOHC V6s obsoleted two-valve pushrod V6s.

It's only a matter of time.

There is no sense in arguing with me about it. I didn't do it.

No one is arguing, the basis of your rhetoric is flawed so you can think what you wish but the reality simply says you are wrong.

[desmo]

If pushrods ever become a marketing liability in the target demographic, they'll be gone like yesterday. I'm not sure engineering has much to do with such decisions.

[Rasputin]

If pushrods ever become a marketing liability in the target demographic, they'll be gone like yesterday. I'm not sure engineering has much to do with such decisions.

I believe that's a very good point, I can recall when the four-valve hysteria on production cars began in the 80s, they all had these "16V DOHC" marking in the rear,
not that middle-class martin had a the vaguest what it meant, but it impressed the suburban neighbors apparently.

As for the push-rod V8s, the single cam solution with rockers and all is a neat and simplistic solution, which has bothered many engineers over the years because the
reciprocating movement of the rods is seen as a bad thing per definition. This is why the Wankel was born btw, from the naive concept that rotating is always better.

First of all, with modern FEA and materials, the reciprocating mass can be kept at a minimum, which often makes the valve itself the speed-limiting factor anyways,
why I'm certain that a purpose-built racing push-rod V8 of today's technology with pneumatic valves could rev almost as good as a OHC, without any limiting rules.

Secondly, the idea of rotating motion being superior to reciprocating is rather childish, just look at nature, ever seen an animal with wheels or a fish with propeller?

[MatsNorway]

The reason i made this thread if anyone cares is that i wanted to talk or hear about intake shape and angle. It is one of the reasons you can get more hp pr liter in a 1L engine that you can buy basically for a month or two worth of sallary. compared to a race engine that has been developed in... 60years ish.

[Grumbles]

The reason i made this thread if anyone cares is that i wanted to talk or hear about intake shape and angle. It is one of the reasons you can get more hp pr liter in a 1L engine that you can buy basically for a month or two worth of sallary. compared to a race engine that has been developed in... 60years ish.

Like anything to do with engine design - especially production based engines - it's all a bunch of compromises. It's interesting to see the changes made in the way these compromises have been traded against each other over the years. For example, 30 odd years ago it was thought that the rod-to-stroke ratio was much more important than it's thought to be today, and I guess you could argue that with the ports of the time a longer rod did have some advantages. Builders used tall-deck blocks to enable rod ratios of around 1.8 - 2.0:1 and the resultant wider head spacing meant the intake manifold runners had a pronounced curvature. At the time though, the long rod was deemed more important than a straight runner.
The situation today is basically reversed - the rod is considered merely something to connect the crank to the piston, and it and the block deck height are being made as short as is physically possible. With the heads closer together the intake runners basically fall straight down to the head (and I'm talking typical PS style tunnel ram here, though it applies to single planes too). The other major change over the last decade or two - especially with inline-valve heads like the classic SBC - is the change in the valve inclination angle, and again it's interesting because of the way the conflicting compromises have been handled. The original factory SBC had the valves inclined 23deg from vertical to form the classic wedge chamber. You might expect that the port could be improved (ie. made straighter) by increasing the inclination but in fact they have gone the other way. Current aftermarket heads have the valve inclined anywhere up to 10 deg or so less than the old 23 deg heads, and while the ports have generally been raised as well they still have quite an angle. Again it's a tradeoff - the modern, more vertical valve angle gives better high lift flow simply because it doesn't get so close to the cylinder wall at those higher lifts. As well, it allows a shallower, more compact combustion chamber. These modern heads perform dramatically better, even though the port isn't necessarily straighter. If you look at the evolution of four-valve heads it's much the same; valve angles have become flatter even though it means more port curvature for a given package size.

I think it's important to remember when comparing the likeable old lumps (like the SBC and SBF) to the more modern V8s that the pushrod engines have had the benefit of decades of development. In the case of the little Chev it's over half a century. Can you imagine what would be possible from say a Coyote or a Northstar if the same attention had been lavished on them for the same length of time? I tend to agree with Magoo - four valve engines aren't the future, they're right now. The only reasons I can think of that the old pushrod V8s (god love'em) are still competitive is that they are relatively compact and that they haven't had a lot of serious competition. Can you imagine a 2 valve pushrod engine being competitive in say four cylinder superbikes or sportscars? It just ain't gonna happen.

[Bob Riebe]

SIZE, weight and COST are in the past, present and future.

OHC offer nothing as an advantage as far as V-8s go.

There are already multi-valve push-rod heads commercially for sale and they are not setting the world on fire.

[Grumbles]

...OHC offer nothing as an advantage as far as V-8s go...

So if it wasn't for all those damned rules and regulations the designers of the current F1 engines would be using pushrods? And there are pushrod V8s producing specific outputs on a par with 4-valve engines?

[MatsNorway]

Grumbles.

Yes the evolution is towards less valve angle. Im guessing its because they want to close the valves as late as possible.

But it does not mean you can`t get straighter intake channels. Its a pushrod thing it seems those 90 deg bends.



Not sure what engine it is but it got way better angle. ofc there are benefits and downsides to this too. by having a steep angle like this one you get less space on the center side of the valve so im guessing the flow is poorer at that side compared to bend in the head.


And why is the valvespring on the valve. by moving it to say the other side you can go more crazy on the intake. you can`t do that as easily on a OHC



"Yes the evolution is towards less valve angle"

when you think about it it is not that great. You have to get smaller valves. you get the cyl wall distrupting flow on that side. and you get less good flow on the inside (bowl) due to space issues inside the valve port.. I guess it shows the power of aggressive valve timings.


I at least think that it gives less gains to turbo engines as they would want a hemi shaped chamber anyway.



I guess its often due to rules that many of the pushrods got a single carb intake.

Edited by MatsNorway, 23 September 2012 - 12:36.

[Fondles]

Not sure what engine it is but it got way better angle. ofc there are benefits and downsides to this too. by having a steep angle like this one you get less space on the center side of the valve so im guessing the flow is poorer at that side compared to bend in the head.


And why is the valvespring on the valve. by moving it to say the other side you can go more crazy on the intake. you can`t do that as easily on a OHC

The main difference is that push-rod engines typically use swirl type heads and multi-valve engines typically use tumble & dumb type heads. The ports for those are very different with the T & D being angled far more steeply onto the valve thus making the heads taller. If they are well designed they aren't a lot taller than a push-rod head though.

The best of the 2V engines, Nascar and V8 Supercars and so-on, match the BMEP of the best of the 4V engines. They are an extremely competitive package. The main advantage of 4V engines lies in ones that are less restricted by revs and that is typically engines with shorter strokes, or in other words smaller capacity engines. If you can't let them rev then they will only make about the same power output-per-capacity as a 2V engine.

[Bob Riebe]

So if it wasn't for all those damned rules and regulations the designers of the current F1 engines would be using pushrods? And there are pushrod V8s producing specific outputs on a par with 4-valve engines?

I am speaking of production cars not spec. racers.

Pro-mods are purpose built racing engines and use push-rods, as do AA Fuelers even in areas where OHC are allowed so the OHC is not some magic potion or the SOHC fuel engine designed in Australia would be ruling the roost but again as far as mass production cars go, no real reason to use them.

[desmo]

DOHC 4vs make perfect sense for high revving short stroke engines working under a restrictive arbitrary displacement limit.

Don't BMEPs strongly favor lower revving engines?

[Canuck]

Lower revving or lower piston speeds?

[gruntguru]

BMEP (think of it as torque per unit displacement) is a useful metric for judging whether an engine design has acchieved its potential in terms of volumetric efficiency, combustion efficiency and mechanical efficiency. It is not really useful for comparing two engines designed for different operating speeds. For the higher speed engine valve-timing, intake & exhaust runners etc would be optimised for a higher operating speed where friction losses are greater ie it would probably produce higher BMEP if it was optimised/re-tuned to a lower speed.

[desmo]

If BMEP essentially = specific torque, and that makes sense, then I've read or heard from multiple sources that that specific torque metric has changed remarkably little for engines built since the early/mid 20th century right through to today. If you think of an IC engine as a heat engine that runs at a nearly fixed AFR and thermal efficiency, this makes thermodynamic sense, and I think provides a useful and sobering perspective for assessing real progress.

[Kelpiecross]

If BMEP essentially = specific torque, and that makes sense, then I've read or heard from multiple sources that that specific torque metric has changed remarkably little for engines built since the early/mid 20th century right through to today. If you think of an IC engine as a heat engine that runs at a nearly fixed AFR and thermal efficiency, this makes thermodynamic sense, and I think provides a useful and sobering perspective for assessing real progress.

Des - I am not exactly sure what you are saying ("specific torque metric?") but I think BMEPs and thermal efficiencies have improved enormously since the "early/mid 20th century". TE/BMEP is linked directly to the compression ratio and typically early/mid 20th century CRs were in the 5:1 to 8:1 range - now they are around 10:1.
I had a 1937 Austin 7 Ruby which I think was 5:1 CR. It would refuse to start if the petrol was a few weeks old.

[gruntguru]

If BMEP essentially = specific torque, and that makes sense, then I've read or heard from multiple sources that that specific torque metric has changed remarkably little for engines built since the early/mid 20th century right through to today. If you think of an IC engine as a heat engine that runs at a nearly fixed AFR and thermal efficiency, this makes thermodynamic sense, and I think provides a useful and sobering perspective for assessing real progress.

"Real progress" has largely been improved breathing. To utilise improved breathing you need to increase engine speed - even a poor flowing design can achieve near 100% VE if you run it slow enough and choose valve timing to suit the low RPM. (The speed where maximum VE occurs will be almost exactly the same speed as peak BMEP.)

So increasing peak BMEP requires an increase in VE (not easy when even an archaic design can approach 100%), an increase in combustion efficiency or a decrease in friction.

[Greg Locock]

Des - I am not exactly sure what you are saying ("specific torque metric?") but I think BMEPs and thermal efficiencies have improved enormously since the "early/mid 20th century". TE/BMEP is linked directly to the compression ratio and typically early/mid 20th century CRs were in the 5:1 to 8:1 range - now they are around 10:1.
I had a 1937 Austin 7 Ruby which I think was 5:1 CR. It would refuse to start if the petrol was a few weeks old.

BMEP is not directly related to thermal efficiency, that is, for a given engine running at the best thermal efficiency for that rpm, bunging in a bit more fuel will almost always give more torque and hence bmep, at an obvious cost in efficiency. A good example is the Prius engine, pretty good (best SI automotive engine) for efficiency, rather unimpressive in terms of bmep.

[Kelpiecross]

BMEP is not directly related to thermal efficiency, that is, for a given engine running at the best thermal efficiency for that rpm, bunging in a bit more fuel will almost always give more torque and hence bmep, at an obvious cost in efficiency. A good example is the Prius engine, pretty good (best SI automotive engine) for efficiency, rather unimpressive in terms of bmep.

I think I see your point - however - in the case of an experimental engine (for example) with a variable CR, keeping the various parameters constant and just varying the CR - increasing the CR increases the TE and in turn increases the BMEP surely? So it is fair to say that BMEP and TE are directly related via the CR?

[Greg Locock]

I think I see your point - however - in the case of an experimental engine (for example) with a variable CR, keeping the various parameters constant and just varying the CR - increasing the CR increases the TE and in turn increases the BMEP surely? So it is fair to say that BMEP and TE are directly related via the CR?

I think you need to hold at least one thing more constant, so are you tuning for max efficiency each time, or stoke, or max torque? Assuming all at one speed. It sounds like a good time to delve into the LARC archive http://naca.larc.nasa.gov/search.jsp where that is the sort of crazy stuff they used to do.

Roughly speaking I'd expect both curves to resemble inverted parabolae, but the peaks may be at different CR, and the gradients may be different.

Coo that was a good guess, Heywood figure 15-12, optimum CR on one engine is about 16 for efficiency, 18 for bmep.

[Kelpiecross]

I think you need to hold at least one thing more constant, so are you tuning for max efficiency each time, or stoke, or max torque? Assuming all at one speed. It sounds like a good time to delve into the LARC archive http://naca.larc.nasa.gov/search.jsp where that is the sort of crazy stuff they used to do.

Roughly speaking I'd expect both curves to resemble inverted parabolae, but the peaks may be at different CR, and the gradients may be different.

Coo that was a good guess, Heywood figure 15-12, optimum CR on one engine is about 16 for efficiency, 18 for bmep.

Thank you for the link - I hadn't realised that these reports were available free. What other similar reports are free?

I was involved in, and also witnessed, a lot of testing of an SI engine with a very high CR (around 18:1) and the comparison testing of the same type of engine with a CR of 8:1 or so. The various thermodynamic effects etc. of very high CRs become very clear.