Question about VE air, MAF air, and Dynamic air

[barum]

There are several ways to calculate (approximated = assuming some fixed ratio between IAT and ECT) GMVE.
These are the formulas I've been using with 2.24 version, but I haven't done the conversion to 3.x series variable names yet.

I took me a while to gather all the pieces from here and there (thanks Marcin for your work) to be able to update the GMVE table from almost any log with correct filtering.


Calculated GMVE (from Dynamic Airflow)

[PID.2320]/([SENS.70]/15*0.875*[SENS.30]/0.28705/(273.15+[SENS.11]))*875/0.28705

GM Volumetric Efficiency = Dynamic Airflow[g/s] / ( RPM[1/min]/15 * Displacement[l] * MAP[kPa] / 0.28705 / (273.15 + IAT[C]) ) * Displacement[cm3] / 0.28705


[PID.2320]*(273.15+[SENS.11])*1000*15/([SENS.70]*[SENS.30])

GM Volumetric Efficiency = Dynamic Airflow[g/s] * (273.15 + IAT[C]) * 1000 * 15 / ( RPM[1/min] * MAP[kPa] )



Calculated GMVE (from MAF)

[SENS.40]/([SENS.70]/15*0.875*[SENS.30]/0.28705/(273.15+[SENS.11]))*875/0.28705

GM Volumetric Efficiency = Mass Air Flow[g/s] / ( RPM[1/min]/15 * Displacement[l] * MAP[kPa] / 0.28705 / (273.15 + IAT[C]) ) * Displacement[cm3] / 0.28705


[SENS.40]*(273.15+[SENS.11])*1000*15/([SENS.70]*[SENS.30])

GM Volumetric Efficiency = Mass Air Flow[g/s] * (273.15 + IAT[C]) * 1000 * 15 / ( RPM[1/min] * MAP[kPa] )



Calculated GMVE (from cylAir which is derived from Dynamic Airflow)

[PID.2321]*(273.15+[SENS.11])*1000/[SENS.30]

GM Volumetric Efficiency = Cylinder Airmass[g] * (273.15 + IAT[C]) * 1000 / MAP[kPa]



Calculated GMVE (from WBo2)

([PID.6200]+[PID.6201])/2000*[PID.6210]*[AUX.20121]*(273.15+[SENS.11])*1000/[SENS.30]

GM Volumetric Efficiency = (InjectorPW_B1[ms] + InjectorPW_B2[ms]) / 2000 * InjectorFlowRate[g/s] * WideBandO2AFR * (273.15 + IAT[C]) * 1000 / MAP[kPa]


([PID.6200]+[PID.6201])/2000*[PID.6210]*[AUX.20122]*[PID.6001]*[PID.68]*(273.15+[SENS.11])*1000/[SENS.30]

GM Volumetric Efficiency = (InjectorPW_B1[ms] + InjectorPW_B2[ms]) / 2000 * InjectorFlowRate[g/s] * WideBandO2Lambda * CommandedAFRHiRes * CommandedEQRatio * (273.15 + IAT[C]) * 1000 / MAP[kPa]


([PID.6200]+[PID.6201])/2000*[PID.6210]*[AUX.20122]*[SENS.121]*[PID.68]*(273.15+[SENS.11])*1000/[SENS.30]

GM Volumetric Efficiency = (InjectorPW_B1[ms] + InjectorPW_B2[ms]) / 2000 * InjectorFlowRate[g/s] * WideBandO2Lambda * CommandedAFR * CommandedEQRatio * (273.15 + IAT[C]) * 1000 / MAP[kPa]


Depending on the mode (OL or CL, SD, Hybrid or MAF only) it is necessary to choose the best suited calculation method. Generally cylAir gives the best quality but I have used also an average of all four.

[schpenxel]

That was the most win I've seen posted here in years. I had figured the MAF one out but that was about it.

Cheers to you, sir.

[barum]

You're welcome.

[schpenxel]

Oh, I had figured out the wideband one also, almost.. had wrong units on IAT so that's why it never worked right

[barum]

That's why it's a good practice to check any formula using the correct (SI) units.

[schpenxel]

True, true.. I should have written it all out by hand. I tend to screw up units less then.

[JBZ]

barum...awesome thank you. now let me study that for the next 5 weeks

and it looks like my math may have been a little off and that is most definitely why it was a little screwy...this is what I had for my GMVE calc based on WB. and to be honest I don't use AFR normally but make it jive I had to in this formula.

([50119.234] * ([50151] + [50152])/2) * [6210.71] * [50011.240]/[50030.91]
EQ ratio (AFR) * (inj PW B1 + inj PW B2/2) * injector flow rate * intake air temp/ MAP

[schpenxel]

Well, I just checked two of those formulas and they worked perfect. Win.

I'm going to try it on something less dialed in later and see what I get.

[sssnake]

GM Volumetric Efficiency = (InjectorPW_B1[ms] + InjectorPW_B2[ms]) / 2000 * InjectorFlowRate[g/s] * WideBandO2Lambda * CommandedAFRHiRes * CommandedEQRatio * (273.15 + IAT[C]) * 1000 / MAP[kPa]

Barum - I am struggling with the equation above. Will the CommandedAFRHiRes not deal with commanding other than stoich or is it just to address alcohol percentage? (BTW - I don't have access to COmmandedAFRHiRes in my OS or I would just check).

EQ ratio (AFR) * (inj PW B1 + inj PW B2/2) * injector flow rate * intake air temp/ MAP

JBZ Try this: EQ ratio (AFR) * (inj PW B1 + inj PW B2)/2 * injector flow rate * intake air temp/ MAP. Your parens were keeping you from getting an average injPW.

[GHuggins]

Are you supposed to be using IAT or MAT or does it matter - I think I was using MAT (manifold air temperature)?

[schpenxel]

[2320.71]*(273.15+[50011.241])*1000*15/([50070.56]*[50030.91])

([50151.254]+[50152.254])/2000*[6210.71]*[50127.238]*[50121]*[50118.239]*(273.15+[50011.241])*1000/[50030.91]

Are you supposed to be using IAT or MAT or does it matter - I think I was using MAT (manifold air temperature)?

I was wondering the same.. but since my IAT is in the manifold, IAT=MAT, so I didn't ask

[JBZ]

Are you supposed to be using IAT or MAT or does it matter - I think I was using MAT (manifold air temperature)?

[2320.71]*(273.15+[50011.241])*1000*15/([50070.56]*[50030.91])

([50151.254]+[50152.254])/2000*[6210.71]*[50127.238]*[50121]*[50118.239]*(273.15+[50011.241])*1000/[50030.91]




I was wondering the same.. but since my IAT is in the manifold, IAT=MAT, so I didn't ask

Me 3 was going to ask but yall beat me to it

[barum]

Answer: The temperature of the air entering the combustion chamber => if you have MAT, then that's the one to use. Note, that GM uses a factor (depending on the airflow if I remember correctly) to weigh between IAT and ECT, so at higher airflow the estimated temperature of the air entering the combustion chamber is relatively close to IAT and vice versa. If you have MAT, then I suppose that factor table should have been balanced differently (or possibly even using weighing factor of 1/close to 1 for MAT => zero weight for ECT).

[GHuggins]

Next Question.... Since most of these formulas off of MAF - are well - using MAF without any sort of correction factor - wouldn't you possibly want to multiply your final number by an air fuel error? Or possibly a Fuel Correction Error? It's just that when I tried this before - that was the only way to get it within 2% the first pass or two... I wish I still had that formula - took days and work with Carson to come up with it using Marcin's original formulas - actually talked to Marcin a little at the time about it come to think of it - didn't write it down and lost it with the transition to 2.25 when it was causing you to lose all of your custom math parameters...

Barum, all I can say is Thank You Sir Not many people would have openly published this on the forum... Wonder if HP could incorporate this into some of the regular provided maths?

So MAT is definitely the best one to use when applicable?

Thanks Again Barum

[barum]

Barum - I am struggling with the equation above. Will the CommandedAFRHiRes not deal with commanding other than stoich or is it just to address alcohol percentage?

That's the idea, if I remember correctly (it's been a year or so when I came up with these) plus all the other AFR/lambda related variables are used to "normalize" the situation in PE mode as we want to get the GMVE values for all areas (that are actually accessed) in the table. I have a flex fuel sensor installed, so the commanded AFR is not the leftmost value in the AFR table (14.6xx or what ever the stoich value for 100% gas is).



Here's what the result looks like. Yes, some areas are not accessed and have to be more or less guessed based on the rest of the table. Yes, the every day traffic area has been smoothed at times, but this is what it looks like after calibration. As you can see, it's quite different to OEM VVE result.

[GHuggins]

That's nice to see... You get the same kind of "decel tail whip" that I get tuning off of just the wideband using your calculated methods...

[barum]

Greg, since it's all about VE with some GM built-in multipliers, you can iterate towards the (more) correct values w/o afr error (assuming of course that you done your homework on MAF scaling). I wouldn't use afr error to correct MAF derived measurement since CL introduces extra "noise" to the estimation meaning you would need much more hits per each cell to average it out (plus if you have all the DFCO's & CFCO's active, it will unnecessarily skew the results). If you want to take the lambda/afr approach, then use those formulas. And as I've mentioned, you can play with excel by calculating averages and weighted averages from all of these, or just compare each and see if there's one that doesn't belong to the group. If I remember correctly, WBo2 will give slightly different (richer = higher GMVE values) than all the others. My WBo2 (TechEdge 2C0B) is in the right merge collector and bank 2 seems to be needing slightly more fuel which may explain the higher GMVE result over the averaged engine, too. At the end you'll have to find what works best for each configuration.

As Marcin has explained, the actual temperature of the air entering the combustion chamber depends on the air flow (=load), so you might consider having a different formula (taking also the ECT into account) for calculating the GMVE in those areas where airflow is relatively low. Of course if you export the log to csv file and use Matlab, then you can use actual weighing factor table to have more precise results. CL is very good at correcting most of the errors in VE and injector data so there's a limit for how far you need to go with this. Another thing to keep in mind are the various error sources (like CL control quality, everything affecting the airflow seen by MAF) making the signal noise ratio worse => you can hone these endlessly and depending on the relative position of the sun and moon, you can have couple of % different values the next day.

The beauty of these air flow based formulas, for me at least, is that you can iterate the VE towards the right ballpark (with the right filtering of course) with data from everyday logs.

[GHuggins]

Barum - my wee brain may need some one on one training or some way to visually see what your explaining if that's OK... I know before I was using the injector flow rate for the formula and as long as the engine had stock injectors - I could get the VE table withen 2% on the first pass with the MAF disabled / the second you threw in a set of A/M injectors - no matter the brand - you could more or less throw the formula in the trash (this was the part I talked to Marcin about I believe) and found that as long as you corrected the data via fuel trims or air fuel error or lambda error - everything was hunky dory again - this is why I was asking about needing a correction while tuning in the MAF at the same time while using the MAF for the base formula...

SO, if the formula is calculated straight off of the MAF flow rate, then how accurate does it need to be dialed in to begin with? This is why I was asking about the correction factor???

Thanks Again

[Eaglegoat]

If you're using the MAF or the dynamic airflow you will need to multiply by lambda reading. If you follow the formula he posted, the lambda reading is incorporated and will auto correct.

[Eaglegoat]

Also, if you have MAT available you need to use MAT. It's what the ECM uses to calculate the fueling, therefore it's what you need to use to calculate fueling.

I've been using these same formulas for several years. It took a very, very, long time for me to put it all together and I just figured everyone else had it figured out already.