Cylinder head CFM to HP?

[PRH]

I can’t think of any of these little quickie formulas that I use to make decisions for a build.

They keep getting dragged back up...... so I throw in my .02

[BradH]

FWIW, the 2 HP per CFM seems to be a decent "potential" HP estimator for my street/strip combinations over the years. We (PRH and I) dynoed two of my builds for which it seemed a reasonable sanity check:

452 ci, 10.8 CR, 266 at .050 x .600 sft, small-volume 305 CFM heads ==> 620+ HP and 570 torque after multiple carbs & intakes tested

452 ci 11.5 CR, 265 at .050 x .650 solid roller, medium-volume 350 CFM heads ==> 680+ HP and 600+ torque... and my carbs tested weren't very close on their tunes, either; a dialed-in carb and the right spacer "should" be pretty close to 700 HP, but that will need to be proven on the track

[Diodedog]

FWIW, the 2 HP per CFM seems to be a decent "potential" HP estimator for my street/strip combinations over the years. We (PRH and I) dynoed two of my builds for which it seemed a reasonable sanity check:

452 ci, 10.8 CR, 266 at .050 x .600 sft, small-volume 305 CFM heads ==> 620+ HP and 570 torque after multiple carbs & intakes tested

452 ci 11.5 CR, 265 at .050 x .650 solid roller, medium-volume 350 CFM heads ==> 680+ HP and 600+ torque... and my carbs tested weren't very close on their tunes, either; a dialed-in carb and the right spacer "should" be pretty close to 700 HP, but that will need to be proven on the track

Did you flow through the manifold.

[BradH]

Although I don't have the data easily available, I did a fair bit of testing of the first heads with a couple of different intakes. The 305 & 350 #s are the heads only.

[MadBill]

I have trouble with the fact that displacement isn't part of the equation. Say my SBC heads with intake flow 300 CFM, so my power potential is ~ 600. Doesn't it matter whether it's a 283 or 427?

[cgarb]

Potential is the key. The same 300cfm heads on a 400 and on a 283 have the potential to make the same 600hp. You would just have to spin the 283 faster to get it.

[mt-engines]

I have seen 1hp per CFM and 2.7hp per CFM on a 8 cylinder engine. I have seen 2.19HP/CFM go to 2.38 just by slowing the heads down, CFM remained the same. I don't use CFM to assume HP potential, experience and RPM will approximate what kind of HP an engine will make. I have seen CSA be more consistant with HP than CFM. But that's my findings.

I could throw out a bunch of examples. Bottom line is you cant stick a 300cfm head on a 350sbc with a 196@.050 cam and expect 600hp.

[mt-engines]

Potential is the key. The same 300cfm heads on a 400 and on a 283 have the potential to make the same 600hp. You would just have to spin the 283 faster to get it.

Is it the CFM or CSA holding the 400 from turning the RPM required to make the HP? or not enough camshaft? You can have the Biggest lungs in the world, but if you are breathing through a straw or can only take short breaths you will faint.

[cgarb]

Potential is the key. The same 300cfm heads on a 400 and on a 283 have the potential to make the same 600hp. You would just have to spin the 283 faster to get it.

Is it the CFM or CSA holding the 400 from turning the RPM required to make the HP? or not enough camshaft? You can have the Biggest lungs in the world, but if you are breathing through a straw or can only take short breaths you will faint.

Or the rocker geometry could be not correct or one spark plug could be missing or even the cam could be installed a tooth of. All of theses this 2 second quick general rule of thumb does not account for.

[BradH]

People are better spent focusing effort on a simulation programs rather than a rule of thumb. We are not interested in potential, we are interested in what is achieved. You will learn so much from simulations such that flow never comes into your vocabulary In many ways it is better than your average dyno data as you cant see whats happening at the average dyno as they usually dont record pressures etc

FWIW, a lot of work w/ my own copy of EA Pro and collaborating with someone who had a reasonably current version of Dynomation using the same data (mine for the then build-in-progress) reinforced some beliefs that I've had from a good bit of running EA Pro using real-world data and results I'd seen from a predecessor of Dynomation:

EA Pro
- Required a lot of gaming the inputs from the actual data in order to generate baseline results that matched my engine dyno data "reasonably well"
- Even with all the mods to the inputs as noted above, the sim always over-estimated peak torque output and underestimated peak HP RPM by 400-600 RPM
- The general trends were realistic, but the outputs were far from being able to take at face value

Dynomation
- Almost always said the engine would perform better with a wider LSA cam, despite real-world testing to the contrary
- Was better at predicting peak HP RPM, but way over-estimated the peak HP #

Neither was that great, IMO. Or maybe something about a standard-port BB Mopar doesn't "fit" the traditional SBC model I suspect these programs are correlated against. Doesn't matter, I'll trust what I see from a reputable engine dyno w/ an experienced operator. I've relegated these "affordable professional" sim programs to one step above toys. I don't think my budget can swing any program that would actually give me legit outputs for actual inputs. My PipeMax v3.98 (IIRC) was a much better bang-for-the-buck in terms of providing me w/ valuable data to factor into my build.

Re the 2 HP per CFM "rule of thumb"... you know what it's still good for? If someone says they're going to build a 700 HP NA engine with 280 CFM heads, I know they probably "can't get there from here", unless we're talking about a high-end engine development program. And when I see someone builds a combination with 340 CFM heads that barely breaks 600 HP, it's a pretty good indicator that combination isn't well thought out in one or more areas (too much port volume for heads, etc.).

Naturally, JMO and JM $.02.

[cgarb]

Re the 2 HP per CFM "rule of thumb"... you know what it's still good for? If someone says they're going to build a 700 HP NA engine with 280 CFM heads, I know they probably "can't get there from here", unless we're talking about a high-end engine development program. And when I see someone builds a combination with 340 CFM heads that barely breaks 600 HP, it's a pretty good indicator that combination isn't well thought out in one or more areas (too much port volume for heads, etc.).

Naturally, JMO and JM $.02.

Exactly....

[digger]

People are better spent focusing effort on a simulation programs rather than a rule of thumb. We are not interested in potential, we are interested in what is achieved. You will learn so much from simulations such that flow never comes into your vocabulary In many ways it is better than your average dyno data as you cant see whats happening at the average dyno as they usually dont record pressures etc

FWIW, a lot of work w/ my own copy of EA Pro and collaborating with someone who had a reasonably current version of Dynomation using the same data (mine for the then build-in-progress) reinforced some beliefs that I've had from a good bit of running EA Pro using real-world data and results I'd seen from a predecessor of Dynomation:

EA Pro
- Required a lot of gaming the inputs from the actual data in order to generate baseline results that matched my engine dyno data "reasonably well"
- Even with all the mods to the inputs as noted above, the sim always over-estimated peak torque output and underestimated peak HP RPM by 400-600 RPM
- The general trends were realistic, but the outputs were far from being able to take at face value

Dynomation
- Almost always said the engine would perform better with a wider LSA cam, despite real-world testing to the contrary
- Was better at predicting peak HP RPM, but way over-estimated the peak HP #

Neither was that great, IMO. Or maybe something about a standard-port BB Mopar doesn't "fit" the traditional SBC model I suspect these programs are correlated against. Doesn't matter, I'll trust what I see from a reputable engine dyno w/ an experienced operator. I've relegated these "affordable professional" sim programs to one step above toys. I don't think my budget can swing any program that would actually give me legit outputs for actual inputs. My PipeMax v3.98 (IIRC) was a much better bang-for-the-buck in terms of providing me w/ valuable data to factor into my build.

Re the 2 HP per CFM "rule of thumb"... you know what it's still good for? If someone says they're going to build a 700 HP NA engine with 280 CFM heads, I know they probably "can't get there from here", unless we're talking about a high-end engine development program. And when I see someone builds a combination with 340 CFM heads that barely breaks 600 HP, it's a pretty good indicator that combination isn't well thought out in one or more areas (too much port volume for heads, etc.).

Naturally, JMO and JM $.02.

Yeah I do not recommend EApro it is rubbish .

I would use engmod4t and I would focus on learning how and why things work and the interrelationships that will help you see the engine as a combination and not a sum of individual parts

[BradH]

Looked over engmod4t and like what I see. However, the next $400 I spend are going to be on a new driveshaft to match my new differential.

[smeg]

People are better spent focusing effort on a simulation programs rather than a rule of thumb. We are not interested in potential, we are interested in what is achieved. You will learn so much from simulations such that flow never comes into your vocabulary In many ways it is better than your average dyno data as you cant see whats happening at the average dyno as they usually dont record pressures etc

FWIW, a lot of work w/ my own copy of EA Pro and collaborating with someone who had a reasonably current version of Dynomation using the same data (mine for the then build-in-progress) reinforced some beliefs that I've had from a good bit of running EA Pro using real-world data and results I'd seen from a predecessor of Dynomation:

EA Pro
- Required a lot of gaming the inputs from the actual data in order to generate baseline results that matched my engine dyno data "reasonably well"
- Even with all the mods to the inputs as noted above, the sim always over-estimated peak torque output and underestimated peak HP RPM by 400-600 RPM
- The general trends were realistic, but the outputs were far from being able to take at face value

Dynomation
- Almost always said the engine would perform better with a wider LSA cam, despite real-world testing to the contrary
- Was better at predicting peak HP RPM, but way over-estimated the peak HP #

Neither was that great, IMO. Or maybe something about a standard-port BB Mopar doesn't "fit" the traditional SBC model I suspect these programs are correlated against. Doesn't matter, I'll trust what I see from a reputable engine dyno w/ an experienced operator. I've relegated these "affordable professional" sim programs to one step above toys. I don't think my budget can swing any program that would actually give me legit outputs for actual inputs. My PipeMax v3.98 (IIRC) was a much better bang-for-the-buck in terms of providing me w/ valuable data to factor into my build.

Re the 2 HP per CFM "rule of thumb"... you know what it's still good for? If someone says they're going to build a 700 HP NA engine with 280 CFM heads, I know they probably "can't get there from here", unless we're talking about a high-end engine development program. And when I see someone builds a combination with 340 CFM heads that barely breaks 600 HP, it's a pretty good indicator that combination isn't well thought out in one or more areas (too much port volume for heads, etc.).

Naturally, JMO and JM $.02.

Yeah I do not recommend EApro it is rubbish .

I would use engmod4t and I would focus on learning how and why things work and the interrelationships that will help you see the engine as a combination and not a sum of individual parts

Sorry I cannot agree there i have an engine shop where we are fully equipped with dyno etc. I have built countless engines using pipemax and EApro and have been within 5 hp on every build. You just have to work out some of the parameters that are wrong and adjust them accordingly.
So no it is not rubbish.

[BradH]

Yeah I do not recommend EApro it is rubbish .

I would use engmod4t and I would focus on learning how and why things work and the interrelationships that will help you see the engine as a combination and not a sum of individual parts

Sorry I cannot agree there i have an engine shop where we are fully equipped with dyno etc. I have built countless engines using pipemax and EApro and have been within 5 hp on every build. You just have to work out some of the parameters that are wrong and adjust them accordingly.
So no it is not rubbish.

When I first plugged in all the actual data into EA Pro, it was shockingly far off from the actual dyno results I had for that combination. So, using the EA Pro user manual where it gave different options for how to "tweak" the inputs to make the sim output more reflective of the real results, I started making changes... lots of changes. Eventually I got something that was gamed enough to use as a baseline for my next round of mods. And the "real" changes to those inputs couldn't be used as is, either. It got to the point where I accepted that if you have to apply that many Band-Aids to the actual inputs to make the sim produce more realistic output, you're not really simulating "your" engine anymore.

If the products works well for you, that's great... and I'm envious. I don't even bother with it anymore. The older PipeMax I have provided solid guidelines for cylinder head CSAs, etc., that I found beneficial. I haven't looked into what else the latest 4.x version can do, though.