Estimating Horsepower MAF vs Inj duty cycle
09-08-2022, 04:29 PM
#1
TECH Apprentice
Thread Starter
Estimating Horsepower MAF vs Inj duty cycle The Modern Automotive frequency output MAF sensor is a high precision instrument, Probably more accurate so than "ANY" Rubber tire driven Inertial Chassis dynomometer... Thus, it *should* be possible to indirectly "measure" flywheel horsepower with a known calibrated MAF sensor output, given reasonably decent AFR, and assuming clean burn, without misfire.
Presumably the MAF should be a more accurate measurement than fuel flow based on injector duty cycle times flow rate times rail pressure itmes.
.
Is there a simple formula relating AIRFLOW in pounds per hour or grams per second to horsepower ?
Any DYNO operators care to take a stab at this AIR vs Horsepower relationship, and document actual observations ?
.
Thank you.
Presumably the MAF should be a more accurate measurement than fuel flow based on injector duty cycle times flow rate times rail pressure itmes.
.
Is there a simple formula relating AIRFLOW in pounds per hour or grams per second to horsepower ?
Any DYNO operators care to take a stab at this AIR vs Horsepower relationship, and document actual observations ?
.
Thank you.
The following users liked this post:
G Atsma (09-12-2022)
09-14-2022, 12:38 AM
#3
TECH Resident
Actually, you can and fairly accurately I might add. 1 gm/sec of airflow is roughly 1.2 hp at the crank or 1 hp at the wheels assuming good air/fuel ratio and timing is on point. I have run many setups on the dyno with the datalog going. I have yet to see this more than ~10 HP off with a properly tuned MAF and proper injector data.
The following users liked this post:
LilJayV10 (06-28-2024)
09-14-2022, 11:26 AM
#5
TECH Apprentice
Thread Starter
Thank you FAST, THAT's exactly what I'm talkin about.
.
Anyone want to chime in this thread with observed DYNO numbers?
Let's see some HP vs Airflow ratios....
09-16-2022, 10:57 AM
#6
Actually, you can and fairly accurately I might add. 1 gm/sec of airflow is roughly 1.2 hp at the crank or 1 hp at the wheels assuming good air/fuel ratio and timing is on point. I have run many setups on the dyno with the datalog going. I have yet to see this more than ~10 HP off with a properly tuned MAF and proper injector data.
09-17-2022, 01:05 AM
#7
TECH Fanatic
Honestly can’t remember the formula off the top of my head, but yeah it works IF nothing is scaled. Also it’s just an assumption. As smoke said, lots of assumptions. The real question is, what is the purpose of the number? There are better ways to measure gains. Since airflow doesn’t necessarily change all that much with tuning, aka, mechanically, you are better served with acceleration rate or watching your IDC’s carefully for measurable successes. Airflow is going to be finicky literally everyday of the week. Your accel rate and IDC’s paint a better picture. I’m assuming spark is the reason here.
Trending Topics
09-22-2022, 01:01 AM
#8
The problem with using MAF or IDC as an estimate horsepower is that the conversion factor contains a bunch of assumptions about things like AFR, BSFC, spark timing, the accuracy of your MAF scaling, the accuracy of your injector specs, etc.
Spark timing: if you tune it right you could dyno it, compare the dyno to a data log, and get a horsepower-per-MAF ratio... Then you could remove several degrees of spark advance, and make noticeably less power with the same MAF.
MAF scaling / injector accuracy: if your injector specs are off by 10% and you use a wideband and data logging to dial in your MAF scaling... your MAF scaling will also be off by 10%. When you use AFR to tune your MAF scaling you're implicitly assuming that you know how much fuel is going into the engine.
That said... once you've done that initial comparison to get a baseline, you could then make some intake and exhaust hardware changes, re-tune your spark advance, and if you see 5% more MAF you reasonably guesstimate that you're making roughly 5% more power. I think you'd be in the ballpark. But you also have to keep in mind diminishing returns. 5% more air almost certainly means less than 5% more power.
If a few people post dyno charts and their max AFR from those dyno pulls, and if all of the horses-per-gram-per-second ratios come out similarly... then I guess you could say something like "my data logs show X grams per second so I might be making something roughly around (X *1.2) horsepower."
I'm not saying 1.2 is the right factor by the way, but it sounds plausible. I think my Subaru was closer to 1.1 when I compared a dyno pull to a data log, but that was over 10 years ago and I'm too lazy to digging through my old data.
And I think we'd probably find different ratios for NA vs low-boost vs high-boost setups. But it would be interesting to compare a bunch of data logs and dyno charts and see what numbers we come up with.
All of the above also holds for comparing torque to grams-per-cylinder, by the way. If you plot MAF and g/cyl over RPM, you get two curves that look a lot like power and torque on a dyno chart, and that's not a coincidence.
Spark timing: if you tune it right you could dyno it, compare the dyno to a data log, and get a horsepower-per-MAF ratio... Then you could remove several degrees of spark advance, and make noticeably less power with the same MAF.
MAF scaling / injector accuracy: if your injector specs are off by 10% and you use a wideband and data logging to dial in your MAF scaling... your MAF scaling will also be off by 10%. When you use AFR to tune your MAF scaling you're implicitly assuming that you know how much fuel is going into the engine.
That said... once you've done that initial comparison to get a baseline, you could then make some intake and exhaust hardware changes, re-tune your spark advance, and if you see 5% more MAF you reasonably guesstimate that you're making roughly 5% more power. I think you'd be in the ballpark. But you also have to keep in mind diminishing returns. 5% more air almost certainly means less than 5% more power.
If a few people post dyno charts and their max AFR from those dyno pulls, and if all of the horses-per-gram-per-second ratios come out similarly... then I guess you could say something like "my data logs show X grams per second so I might be making something roughly around (X *1.2) horsepower."
I'm not saying 1.2 is the right factor by the way, but it sounds plausible. I think my Subaru was closer to 1.1 when I compared a dyno pull to a data log, but that was over 10 years ago and I'm too lazy to digging through my old data.
And I think we'd probably find different ratios for NA vs low-boost vs high-boost setups. But it would be interesting to compare a bunch of data logs and dyno charts and see what numbers we come up with.
All of the above also holds for comparing torque to grams-per-cylinder, by the way. If you plot MAF and g/cyl over RPM, you get two curves that look a lot like power and torque on a dyno chart, and that's not a coincidence.
09-24-2022, 08:22 AM
#9
TECH Resident
The problem with using MAF or IDC as an estimate horsepower is that the conversion factor contains a bunch of assumptions about things like AFR, BSFC, spark timing, the accuracy of your MAF scaling, the accuracy of your injector specs, etc.
Spark timing: if you tune it right you could dyno it, compare the dyno to a data log, and get a horsepower-per-MAF ratio... Then you could remove several degrees of spark advance, and make noticeably less power with the same MAF.
MAF scaling / injector accuracy: if your injector specs are off by 10% and you use a wideband and data logging to dial in your MAF scaling... your MAF scaling will also be off by 10%. When you use AFR to tune your MAF scaling you're implicitly assuming that you know how much fuel is going into the engine.
That said... once you've done that initial comparison to get a baseline, you could then make some intake and exhaust hardware changes, re-tune your spark advance, and if you see 5% more MAF you reasonably guesstimate that you're making roughly 5% more power. I think you'd be in the ballpark. But you also have to keep in mind diminishing returns. 5% more air almost certainly means less than 5% more power.
If a few people post dyno charts and their max AFR from those dyno pulls, and if all of the horses-per-gram-per-second ratios come out similarly... then I guess you could say something like "my data logs show X grams per second so I might be making something roughly around (X *1.2) horsepower."
I'm not saying 1.2 is the right factor by the way, but it sounds plausible. I think my Subaru was closer to 1.1 when I compared a dyno pull to a data log, but that was over 10 years ago and I'm too lazy to digging through my old data.
And I think we'd probably find different ratios for NA vs low-boost vs high-boost setups. But it would be interesting to compare a bunch of data logs and dyno charts and see what numbers we come up with.
All of the above also holds for comparing torque to grams-per-cylinder, by the way. If you plot MAF and g/cyl over RPM, you get two curves that look a lot like power and torque on a dyno chart, and that's not a coincidence.
Spark timing: if you tune it right you could dyno it, compare the dyno to a data log, and get a horsepower-per-MAF ratio... Then you could remove several degrees of spark advance, and make noticeably less power with the same MAF.
MAF scaling / injector accuracy: if your injector specs are off by 10% and you use a wideband and data logging to dial in your MAF scaling... your MAF scaling will also be off by 10%. When you use AFR to tune your MAF scaling you're implicitly assuming that you know how much fuel is going into the engine.
That said... once you've done that initial comparison to get a baseline, you could then make some intake and exhaust hardware changes, re-tune your spark advance, and if you see 5% more MAF you reasonably guesstimate that you're making roughly 5% more power. I think you'd be in the ballpark. But you also have to keep in mind diminishing returns. 5% more air almost certainly means less than 5% more power.
If a few people post dyno charts and their max AFR from those dyno pulls, and if all of the horses-per-gram-per-second ratios come out similarly... then I guess you could say something like "my data logs show X grams per second so I might be making something roughly around (X *1.2) horsepower."
I'm not saying 1.2 is the right factor by the way, but it sounds plausible. I think my Subaru was closer to 1.1 when I compared a dyno pull to a data log, but that was over 10 years ago and I'm too lazy to digging through my old data.
And I think we'd probably find different ratios for NA vs low-boost vs high-boost setups. But it would be interesting to compare a bunch of data logs and dyno charts and see what numbers we come up with.
All of the above also holds for comparing torque to grams-per-cylinder, by the way. If you plot MAF and g/cyl over RPM, you get two curves that look a lot like power and torque on a dyno chart, and that's not a coincidence.
EDIT, Actually I should rephrase part of the above. The reason gms/cyl, torque and INJ PW all trend the same curve is because they all follow the Volumetric Efficiency of the engine. Where the volumetric efficiency is the highest is where torque peak will occur and around peak hp is where peak MAF flow will occur.
Last edited by Fast355; 09-24-2022 at 08:38 AM.
The following 2 users liked this post by Fast355:
Full Power (09-24-2022), G Atsma (09-24-2022)
06-28-2024, 02:29 AM
#10
TECH Apprentice
Thread Starter
Any DYNO Operators/ Pro-Tuners want to re-visit this Airflow x (Constant)= Flywheel Horsepower equation ?
Lots of data being collected out there, certainly someone's got some correlation observed.
Lots of data being collected out there, certainly someone's got some correlation observed.
06-28-2024, 07:23 AM
#11
TECH Addict
its a lot of data to go through, would be a full time job. ive got data from just a few cars and im unlikely to ever really dig deep for presentation. theres a significant step in effort between studying data and assembling something to present to others via forum, youtube, powerpoint, etc
unless you are always using the same ecu with similar combos its going to be hard to generalize power based on other variables. plus every dyno is different. i have a 468 ls7 on mine now ive been working on and it is interesting comparing what other shops have made with strong ls7 combos. i had a base c6 ls3 with blower recently and it made in the lower end of the range of what i found similar combos doing on the net, which is about what i expected. lots of dyno heros out there and mine has shown to be on the conservative end of the spectrum.
i guess its something i might look at a little closer today. in the past ive always used fuel flow to estimate power, but with the various ethonal mixes running around that gets complicated too.
unless you are always using the same ecu with similar combos its going to be hard to generalize power based on other variables. plus every dyno is different. i have a 468 ls7 on mine now ive been working on and it is interesting comparing what other shops have made with strong ls7 combos. i had a base c6 ls3 with blower recently and it made in the lower end of the range of what i found similar combos doing on the net, which is about what i expected. lots of dyno heros out there and mine has shown to be on the conservative end of the spectrum.
i guess its something i might look at a little closer today. in the past ive always used fuel flow to estimate power, but with the various ethonal mixes running around that gets complicated too.
06-28-2024, 11:41 PM
#12
No, it really isn't. While it's true that you get similarly shaped curves if you plot them over RPM, they measure different things.
This thread is all about estimating torque and power based on easily logged parameters. Grams/cylinder and MAF are great because they tell you how much air the engine is ingesting. You can multiply g/cyl by a conversion factor to get a rough estimate of torque. You can multiply MAF by a conversion factor to get a rough estimate of horsepower.
Since VE tells you how much air the engine is ingesting as a fraction of the engine's displacement, you'd need to multiply that by the aforementioned conversion factor and then ALSO multiply it by the engine's displacement. That's just introducing an extra parameter that complicates the process for no benefit, and makes it awkward to compare those conversion factors between engines with different displacements.
There are times when it's helpful to look at VE itself. This thread is not one of those times.
This thread is all about estimating torque and power based on easily logged parameters. Grams/cylinder and MAF are great because they tell you how much air the engine is ingesting. You can multiply g/cyl by a conversion factor to get a rough estimate of torque. You can multiply MAF by a conversion factor to get a rough estimate of horsepower.
Since VE tells you how much air the engine is ingesting as a fraction of the engine's displacement, you'd need to multiply that by the aforementioned conversion factor and then ALSO multiply it by the engine's displacement. That's just introducing an extra parameter that complicates the process for no benefit, and makes it awkward to compare those conversion factors between engines with different displacements.
There are times when it's helpful to look at VE itself. This thread is not one of those times.
06-28-2024, 11:51 PM
#13
I don't think that's really necessary though. Just pick a random dyno run that has MAF, horsepower, and RPM, and plot them, together with a "horsepower / MAF" value. That last curve is the thing we're all curious about - the conversion factor that turns MAF values into horsepower. Of course it'll just be a horsepower estimate for every engine other than the one example, but it'll probably get us in the ballpark.
Better yet, pick two dyno runs - one from an LS with its factory cam, and one from an LS with a big aftermarket cam / heads / intake / etc. The difference between the two curves will be interesting, and might give us factors that we can use to make upper and lower horsepower estimates based on our own MAF data.
We're just looking for estimates here, not scientifically proven alternatives to using dynos.
Better yet, pick two dyno runs - one from an LS with its factory cam, and one from an LS with a big aftermarket cam / heads / intake / etc. The difference between the two curves will be interesting, and might give us factors that we can use to make upper and lower horsepower estimates based on our own MAF data.
We're just looking for estimates here, not scientifically proven alternatives to using dynos.