Mustang and Ford Performance Forums banner

Increasing CR vs boost pressure

63K views 30 replies 19 participants last post by  eschaider 
#1 · (Edited)
Edit 10/19/07. To whom it may concern. The source of the material below is me.
Plagiarism = literary theft, or trying to pass someone else's words off as your own.
For myths on copyright, please see Brad's web page: => http://www.templetons.com/brad/copymyths.html
========================================================
This question seems to come up time and time again. Is it better to increase the static CR or boost pressure. There are a couple reasons why supercharged or turbocharged engines run lower static compression ratios. A static CR in the range of 8-9 is very common. Here are a couple considerations.

Consideration #1
Heat from compression by a supercharger or turbo can be removed (for the most part) through use of an intercooler. Heat from compression within the cylinder cannot. Also, the cylinder pressure at the end of the compression stroke (prior to ignition) goes up exponentially with an increase in static compression ratio, versus a linear increase with boost pressure. Therefore, increasing the static CR is going to unavoidably push you closer to the knock limit for a given fuel. In other words, the octane requirement goes up more by increasing the static CR than it does by increasing boost.

For example, increasing the static CR from 8.5 to 9.5 increases the temperature within the cylinder at the end of the compression stroke (but before ignition) by ~63°F, (assuming IAT2 = 130°F and ideal adiabatic compression with γ = Cp/Cv = 1.4. I won't bore anyone with equations. The situation doesn't change much even if IAT2 were only, say, 100°F. In that case, the increase in temp at the end of the compression stroke goes up by ~60°F for the same increase in static CR). Also, the pressure at the end of the compression stroke (before ignition) goes up by ~97 psi from 574 psi to 671 psi, assuming atmospheric and boost pressures of 14.7 and 14 psi, respectively. On the other hand, increasing the boost pressure from 14 to 15 psi increases the outlet temp of the compressor by only ~11°F, assuming AE=60% and IAT1 = 90°F. And by further assuming an intercooler efficiency of 80%, the increase in IAT2 is only ~2°F. Hence, the increase in temp at the end of the compression stroke will hardly change at all. Also, the increase in cylinder pressure at the end of the compression stroke only goes up by ~20 psi (from 574 to 594 psia) with this increase in boost pressure.

So summarizing the effects of increased temp and pressure at the end of the compression stroke for the two cases:
Increased CR from 8.5 to 9.5: ΔT = ~63°F and ΔP = ~97 psi
Increased boost from 14 to 15 psi: ΔT = ~2.4°F and ΔP = ~20 psi

A higher temp and pressure increase the likelihood of deadly preignition for a given octane fuel. And for those astute observers that know the physics I've applied, yes, although I've idealized things to keep it simple, (by not including effects such as heat loss thru the cylinder walls during the compression stroke or ignition and valve timing in the calculations), I'm sure they'll also recognize that this doesn't change the conclusion.

Consideration #2
Power is increased by two completely different mechanisms for the two approaches. Increasing the static compression ratio increases power via an increase in thermal-conversion efficiency. Increasing boost pressure increases power via an increase in mass-air flow rate. There's less gain in thermal-conversion efficiency (and hence power) via an increased static CR compared to the power gain by increasing the mass-air flow rate via an increase in boost pressure. For example, increasing the static CR from 8.5 to 9.5 results in an increase in thermal-conversion efficiency (for an ideal Otto cycle) of about 3.2%. On the other hand, increasing the boost pressure from just 14 psi to 15 psi, increases the mass-air flow rate by about 3.5%. If boost pressure is increased by 2 psi, (from 14 to 16 psi), the increase in mass-air flow rate will now be more than twice that compared to the increase in thermal-conversion efficiency, (~7% vs ~3.2%), and ΔT and ΔP still won't be as great as they are when increasing the static CR from 8.5 to 9.5. Therefore, not only can it be "safer" from the knock point of view, but a little more power is gained as well, (relatively speaking that is).

In conclusion, I would contend that for a forced-induction application, that low compression is in general, the better way to go. :beerchug:
=============================================================
Boost vs compression ratio Part II

Since writing part I, there have been some comments made that I felt warranted a part II. Comments such as, "Good stuff, but peak combustion pressure and temperatures are far higher than they are pre-ignition." Or, "I have more gauges than a pimply faced teenager in his Honda Civic, so I can run closer to the ragged edge." Or, "The engineers didn't put 8.5:1 compression pistons in the Terminator, or 8.4:1 in the FGT & GT500 for any thermodynamic reasons." Or the classic, "It's all in the tune." Although most of these statements are not totally without substance, the basic conclusion is unchanged. Even though adjusting timing and AFR can reduce the tendency to knock, and although peak combustion pressure and temperatures post ignition are significantly higher than pre ignition, or in spite of how close one cares to run to the knock limit of a given octane fuel, (for a given fuel and AFR, etc.), the engine can make more power by reducing CR and increasing boost pressure, than the other way around for the same peak cylinder pressure. This is why it is common to see lower compression ratios on SI forced-induction motors. Obviously there are tradeoffs, however, as a direct result of the lower thermal-conversion efficiency. But when it comes to maximizing power and torque output at wide-open throttle for a given octane fuel, etc., lowering CR and raising boost pressure is the safer approach.

This conclusion is based primarily on two basic facts:
  1. Mean-effective pressure (MEP) goes in direct proportion to the mass of air ingested, (which is directly related to boost pressure), but goes up "sublinearly" with compression ratio, CR
  2. To good approximation, peak cylinder pressure goes in direct proportion to both.
And as mentioned in Part I, while heat from compression by a supercharger can be effectively removed through use of an intercooler, heat from compression within the cylinder cannot. As a result, peak combustion temperatures do not tend to rise significantly with increases in boost pressure, whereas they will go up with increased compression ratio. And as we all know, a lower peak combustion temp also reduces the likelihood of knock. One also needs to recall that power and torque are directly proportional to MEP, where the mean-effective pressure is defined as the "effective" pressure over the cycle, which is equal to the work generated over the cycle divided by the displaced volume. In other words, if the indicated MEP goes up X%, indicated power and torque will also go up X% at a given engine speed. Additionally, one needs to recognize that for a given octane fuel, AFR, etc., that as peak cylinder pressure and temperature are raised, eventually the engine going to knock, or detonate. This shouldn't be any surprise since this is exactly how a fuel's octane, (i.e. its resistance to knock), is measured. It is put in a special test engine whose compression is raised until the engine knocks. (The reference fuels used for comparison are iso-octane defined as having ON = 100, and normal heptane having ON = 0.) References: http://en.wikipedia.org/wiki/Octane_rating. Or section 6 here => http://blizzard.rwic.und.edu/~nordlie/cars/gasoline.html

For the ideal Otto cycle, it is very easy to derive expressions for the peak combustion pressure (P3) and temperature (T3), and mean-effective pressure. As in Part I, I'm not going to derive or show all the math, but simply get to the bottom line and show the results. For those that want the details, the interested reader is referred to any number of good text books on engine fundamentals, (Taylor's, Heywood's, etc.). I've also posted the equations, sans proof, in this thread: http://www.eng-tips.com/viewthread.cfm?qid=215499&page=1

Taking the ratio of mean-effective pressure to peak cylinder pressure, or vice versa, one will find that the dependency on boost pressure drops out and the ratio only depends on CR for a given fuel and AFR. (Note - timing does not factor in simply because MBTT at TDC for the ideal cycle, but this does not change the conclusion.)



where the thermal-conversion efficiency for the ideal cycle is given by, ηt = 1 - CR^(1-γ), cv is the constant-volume specific heat for the mixture, ηc is the combustion efficiency, Qhv is the fuel's heating value, and γ is the polytropic exponent which can be taken to have a value of around 1.25-1.3, over the full cycle, (Ref., H.M. Cheung and J.B. Heywood, SAE paper 932749).

Using these results, one can plot the ratio of indicated mean-effective pressure to peak cylinder pressure, iMEP/P3, vs compression ratio. From this plot, one will see that as CR goes up, the ratio iMEP/P3 will go down. (See plot below).



What does this mean? It means for any given maximum tolerable peak pressure (for a given octane and AFR), that iMEP will be higher at a lower CR than at a higher CR, independent of boost pressure. Said another way, this means peak cylinder pressure climbs faster than indicated mean-effective pressure does as CR is increased, whereas both P3 and iMEP will climb at the same rate with boost pressure. Therefore, for any given fuel and AFR, etc., one can make more power & torque at any given engine speed by reducing CR and increasing boost pressure, than the other way around for the same peak cylinder pressure. The tradeoff is a lower thermal-conversion efficiency, which translates to a higher specific fuel consumption (pounds per hour per horsepower) and a "doggier" response at part throttle.

Although the above conclusion was based on analysis of the ideal cycle, a more complete thermodynamic model including finite burn duration, heat loss, spark timing, etc, will show the same trends and lead one to the same conclusion. As an example, cylinder pressure and temperatures vs crank angle for two engines with the same iMEP, but different CR and boost pressures are shown below. As can be seen, the engine with the higher CR has higher combustion pressures and temperatures, making it more likely to knock for a given octane fuel.




=============================================================
Boost vs Compression Part III - Measured data

From posts in another related thread:

Measured data, example 1, turbocharged Cobra:
Let's use the publicly available, actual data as posted on this web site. Using a turbo Cobra as an example, consider the data in the Terminator Summary of turbo data thread. (Link => Summary of turbo data)

From this data, the "typical" turbo Cobra running 15 psi makes around 670 rwHP, (with stock displacement and CR.)

Now say one increases CR to 10:1 from 8.5:1, and pulls a few deg of timing to avoid knock.
How? Click link => Spark timing impact on knock

Say this increase in CR nets a 6% gain as published in a recent magazine article. This then translates to 1.06*670 = 710 rwHP. So yes... a nice gain of 40 rwHP.

But now lets keep timing at this same reduced amount, but instead of increasing CR, increase boost until one gets to the same peak cylinder pressure & temp. So how much boost is added? Use the well-known, effective-compression ratio as given by,

CReff = CR(1 + Pboost/Patm)

which comes about from how peak cyl pressure scales with both boost and CR. (Note - it also can take into account valve timing, by using the dynamic-compression ratio for the value of CR.) As Ed (eschaider) has pointed out, this effective-compression ratio is what VP (and others) uses to "rate" their various fuels.

If one goes through the math, they'll find that this would mean running around 20 psi at this same reduced timing, but with stock CR of 8.5:1, instead of 10.0:1. (Sorry...yes...a little math. The "prime" ( ' ) symbols indicate the "new" values of a quantity.)

Pb' = [(CR/CR')(1 + Pboost/Patm) -1]Patm = [(10/8.5)(1 + 15/14.7) -1]14.7 = 20.2 psi

Again from the regression analysis of actual data shown in the turbo data summary thread, (link provided above), the change in power with boost is roughly 22 rwHP/psi. So this means with the same reduced timing, the CR=8.5 motor with 20 psi would now be making 670 + 22*(20 - 15) = 780 rwHP. So a gain of 110 rwHP.

Example from measured Turbo data - summary:
Baseline case; CR = 8.5:1, boost = 15 psi, ~670 rwHP
Higher CR; CR = 10:1, boost = 15 psi, ~710 rwHP
Higher boost; CR = 8.5, boost = 20 psi, ~780 rwHP

Measured data, example 2, twin-screw KB Cobra:
Here's a similar example of the same analysis for a blower car from chassis-dyno measurements made on the same car. (I didn't have data at 15 and 20 psi, but I do at 17 and 23. Using the effective-compression formula given above, the higher boost would come out to 22.6 psi for this case, but I think you'll give me the 0.4 psi.)

For this analysis, the data is taken from the same vehicle, on back-to-back pulls, on the same dynamometer, and covers the full rpm range - not just peak power. Since it is from actual measured data, it includes all additional effects such as any potential belt slip, increased SC drive power, etc. Data for the CR=10:1, 17 psi case came by scaling the CR=8.5, 17 psi case up by 6%. (A worked example of how power scales with CR, including actual measured data, is given later in this thread in post #23, Link => Post #23.)



As one can clearly see, the higher-boost/lower-CR makes significantly more torque than that of the higher-CR/lower-boost case, all across the rpm range.

=============================================================
Boost vs Compression Part IV - Impact of valve timing

From thread in 2011 Mustang Forum. Click link => Boosted Coyote Engines using VVT

Yes, valve timing also has an impact and can reduce the risk of knock. One can delay the inlet-valve closing (IVC) event, which reduces the so-called dynamic compression ratio. Basically, when you delay the IVC, you aren't running as high a (dynamic) compression ratio anymore which as explained above, helps reduce peak cylinder pressure & temp, thereby reducing the risk of knock. Below is an example for the 5L Coyote engine in the new 2011 Mustang GT.



Therefore, the engine calibrators can make use of the variable cam timing on the 5L Coyote engine to reduce its relatively high-compression ratio, to a lower compression ratio when using forced induction.
 
See less See more
6
#2 ·
Very nice, as always, Black2003.

In your comparison, I assume the 2 different mods produce a similar increase in power at WOT? If not, please provide us with a comparison that can be expected to produce similar gains.
 
#3 · (Edited)
Hey Philly - Of course it would ultimately depend on the type of compressor used (Eaton, KB, Whipple, or turbo). But for the most part, yes…the comparison of upping the static CR from 8.5 to 9.5 should give a similar increase in indicated power as going from 14 to 15 psi. (The relative increases were 3.2% vs 3.5%, respectively). In terms of brake power, a supercharger would definitely net you something less than the 3.5%, of course, but it would depend on what kind. Although it could be done, I didn't bother calculating the increase in drive power. A turbo would probably see most of that 3.5%.
========================================================
7-5-08: Please ignore. Graphics below for in-line posting above.
 

Attachments

#5 ·
Re: Increasing static CR vs boost pressure

Someone i know tried to put a 20psi blower on a 11.5:1 NA motor, they had good gas (C16) and very little timing but it still went all to pieces. Thats a good example of how too high of a CR is bad, no matter what the fuel.
 
#6 ·
Re: Increasing static CR vs boost pressure

From a practical standpoint (cars that will be driven on the street) I can vouch for not dropping compression too much. If you run an 8.0:1 motor on the street with an appropriate amount of boost, it will feel doggy off the line (before boost comes on). At least when it is compared to a 9.0:1 motor (even running a bit less timing).
The thermal efficiency of the forced induction components also has a huge amount of importance when dealing with these cars. An Eaton (ported or not) is going to whip up IATs, forcing the intercooler/ heat exchanger system to do more work. Pulleying a thermally-challenged blower only compounds the problem. That is why 16psi from a Heaton does not equal 16psi from a KB or Whipple or even a properly-sized turbo. Luckily enough, that means you car run more boost AND timing from the bigger blower, and keep the oily bits inside the motor where they belong...
 
#7 ·
FYI - Added Part II to keep this altogether in one thread.
 
#9 · (Edited)
Resurrection!

Good info!

Just my .02 though.....
The scope of this information (atleast what I'm deriving) is "max power" with also references to pump fuel.

Those who are using pump fuel are most interested in "street drivability/performance";

I think when considering one extreme end of the spectrum, the other must also be considered.

Can you make more power with 3000 CFM, and 6 to 1 compression. Maybe. Does that help a guy that is looking to put together the best compromize for pump fuel (and street performance). NOPE!

With the scope of pump fuel and street driveability there has to be a happy medium, driving satisfaction vs. peak WOT performance.

The happy medium is my holy grail at the moment as I decide on piston dish sizes :)

I'll likely go with an 11.5 cc piston dish, and a thicker head gasket to net around 8.7 to 1 CR, (which I can change my mind and revert to stock gasket thickness and higher CR if desired).

Although I do admit, if a car is doggy, this can be masked somewhat with gearing, auto-trans with loose converter, etc. It will still affect the drivability and satisfaction on the street.

The bigger problem is enthusiasts who can't decide if they are driving, or racing there car, hahaha.

Good info regardless!
 
#10 ·
I’m glad you found the thread interesting, 4vMn12!

The analysis actually applies to any fuel though. With any fuel, as peak cylinder pressures and temperatures are increased, the engine is eventually going to knock. Again, this is basically how a fuel’s octane rating is measured and is what determines the octane requirement of the motor.

As was mentioned, and as you are apparently aware, there are tradeoffs under part throttle (no boost) when reducing the static compression ratio of the motor. This is due to the drop in thermal-conversion efficiency, ηt, which is a very non-linear function of compression ratio. To help illustrate this, one can see from the graph below that ηt really starts to drop off quickly at lower CRs, and approaches zero when CR = 1. (This is simply because the gas is no longer doing any work on the system when CR ≡ Vd/Vc = 1. In other words, at CR = 1, work = 0, and hence, ηt ≡ work/Qin = 0.) Clearly such low compression ratios are to be considered extreme cases, however, as they are not exactly practical nor appropriate with modern fuels and engine designs.

The value you have chosen for you motor is quite typical. Thanks for your input!
 
#11 ·
Great info! I bumped my compression some becuase i couldnt raise boost any more since i was sticking with a already max pullied ported eaton. While rebuilding the short block i had to buy pistons anyways so it was sort of like free power. If i was going TS or turbo i never would have changed the comp ratio.
 
#16 ·
03yllwcobra - With a centri blower, I can understand why you wanted to bump up the CR. Since boost at low rpm is virtually non-existent, they tend to be doggy down low. I had one, so I understand. With a positive-displacement blower, that isn't so much of an issue. (The bypass valve is not very big, so it will come out of vacuum fairly easily.) Even at part throttle, my '03 is significantly more responsive than my '01 was, which had a CR of 9.85 IIRC. If you're detecting signs of detonation, there are a few things you can do. 1) reduce spark advance, 2) enrichen the air/fuel mixture, 3) reduce boost pressure, or 4) up the fuel octane. Options 1-3 will result in a loss of power, of course. And there are some that fancy alcohol injection w/ centri blowers.
 
#17 ·
The Highest timing im going to run is going to be 14* but the lowest boost psi i can run is 23 psi unless procharger makes a bigger cog pulley, i usually run 110 in the car all the time but i think that what caused my stock motor to blow up because the o2s werent switching fast enough to keep a/f
 
#18 ·
I'm not a tuner, but I would suspect for those conditions on 110, you would be OK. I assume you're running an intercooler.
 
#20 ·
I worked on a TT mach1, dual 50 trims I believe. With the stock 10+:1 motor it made about 440rwhp at 11:1, 7psi, and 14*.

With an 03 cobra shortblock it made 380RWHP with the same tune, even with 5-6* more timing cranked in, it still made only 420rwhp, though I could have run 2-4* more timing at such a low boost level, I did not feel like there were great gains from doing so, gains were starting to taper off.

That goes to show how much power lies in compression.

Now in his case he just turned up the boost to 12psi and made over 600rwhp, this goes to show how inefficient the compressors were at low boost.
 
#21 · (Edited)
Thanks for the data Justin.

Going from 8.5:1 to 11:1 increases the thermal conversion efficiency by roughly 8.3%, (using gamma = 1.3). So assuming all else is equal, scaling gives ... 380*1.083 = 411 at the same rpm, which isn't too terrible considering it's just a basic, back-of-the-envelope estimate. (~6.6% error.) If I get ambitious, maybe I'll run it through the more detailed thermodynamic model I have.

I'm not clear which motor made 600 going from 7 to 12 psi, but the turbo 03 & 04's seem to gain about 23 rwHP per psi, (=> Ref link). So a 5 psi increase would translate to 115 rwHP. So in either case, the engine definitely picked up more than what is typical!
 
#22 ·
fpv_gtp - Thank you.
 
#23 · (Edited)
Copied relevant post below from CAT600's thread:
Link => Compression ratio and Boost

I agree with you, but the statement I was replaying to sounded like he was talking NA motors where a higher CR motor would do better, cuz it makes more pressure and more power. This is tru which is why I was saying if you know take a low CR motor and cram air in with a blower there is twice the air to make power with. When you get into it even more the lower air temps allow even more power.
Yes, that's correct. With the same inlet pressure, power and torque go up with increasing CR due to the commensurate increase in thermal-conversion efficiency. (Power is proportional to mass-air flow rate times therm-conv efficiency.)

There is a very nice write up on a test that MM&FF did at their site. (Link => MM&FF Ref ) On a 5.4L NA motor, they raised CR from 8.3 to 11.8, and observed power and torque to go from 478 hp and 395 lb-ft of torque, to 543 hp and 439 lb-ft of torque. Again, that's due to the increase in thermal-conversion efficiency. Assuming an effective ratio of cp/cv = 1.3 over the entire cycle, when one does the math, the thermal-conversion efficiency will be seen to go up by ~11.3%, from CR = 8.3 to 11.8. So theoretically, (first order), the baseline power of 478 scales to 478*1.113 = 532 hp, and baseline tq of 395 scales to 395*1.113 = 439.6 lb-ft, (where the error is a mere 2%, and 0.1%, respectively.)

Summarizing, by going from a CR of 8.3 to 11.8, the measured and calc'd results are:
Code:
Parm	Meas	Calc	Error
Power	543	532	2%
Torq	439	439.6	0.1%
So it's not magic or BS theory, and the experimental numbers agree. The article also rightly points out the dangers of increased risk of knock, and what this means as far as the octane requirements of the engine. Make no mistake about it...power/torque absolutely do go up with CR, (all else being equal). And so does cylinder pressure. So one needs to be cognizant of limits such as an octane (knock) limit, strength-of-parts limit, etc.

And of course the equal-inlet-pressure case they tested isn't being discussed here. They didn't address the low CR/high boost vs high CR/low boost comparison at the same peak cyl pressure. As far as that case goes, SlowSVT is certainly on the right track...more air "wins" over the lower therm-conv efficiency of the low CR/hi boost case, at the same peak combustion pressure as the hi CR/low boost case...
 
#26 ·
I just had a flashback to my undergrad Mechanical Engineering courses. If only all professors were gearheads, I would've paid more attention!

Always good to dig back into my old books though after reading a post like this and refresh my memory - now if I could only find a job where I could apply this type of knowledge!
 
#28 ·
I am curious what your opinion is on how much boost I can run using 9.3:1 compression on pump gas (93 octane)? I would guess 14 psi? There are variables such as heads, cams, timing, etc., but in general I need to know to dial in my next street tune. Want to keep her safe, but not strip all the gusto out of her. I keep reading more boost and less timing is better. My race tune has 22psi and 9.3:1 compression, but octane is being supplemented by Torco and 100% VP M1 methanol.
 
#29 ·
What is the limit - or the best power for a given setup? I am not yet familiar with the math required to do this myself. But by manipulating the following equation that you provided: Pb' = [(CR/CR')(1 + Pboost/Patm) -1]Patm = [(10/8.5)(1 + 15/14.7) -1]14.7 = 20.2 psi, I could build a table of CRs and Pbs. I don't know how to relate this back to hp as you did in the example. As you point out nt goes to zero at CR=1, so somewhere there is a best power for this setup...a graph of HP vs CR/Pboost would be enlightening. I don't care about "drivability"...just steady state hp at 1 engine speed and 1 manifold setting. FYI: we are at 6:1 and lots of boost with great success, but we don't know where to go from here or even how far we can go with this setup.
 
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top