New Tuning Tutorial: WeathermanShawn

[WeathermanShawn]

Good work there Joe.

Yea, I see the same unit problem using g*k/kPa. I can't tell if it is the LTFTBEN multiplication (%) factor or elsewhere. Like you said VE % works fine.

I'm using the same cal_pids.txt..

[TunasTwins]

What is the difference with how Tunercat display the table compared to Efilive?

Well I dont know what EFI Live displays unless someone has a stock file for an 05 4.8L The OS is 12592618....Quote from TC help file: The values is this table are expressed in grams per cylinder per second. These values can be converted to %VE which was used in earlier calibrations using the following formula:

%VE = 0.2871 x VE / CV

where VE is the value from the table and CV is the volume a one cylinder in liters.

It would be lots easier to conceptualize VE if it were in percentage.

[joecar]

Good work there Joe.

Yea, I see the same unit problem using g*k/kPa. I can't tell if it is the LTFTBEN multiplication (%) factor or elsewhere. Like you said VE % works fine.

I'm using the same cal_pids.txt..

It's not the LTFTBEN... it is the large constant being incorrect...


I analyzed the VE[g*K/kPa] equation, read on (see below for how I arrived at it):

when I use the following it equals with my B0101 exactly:
VE[g*K/kPa] = {SAE.MAF.gps} * {CALC.DAT.K} * 15 / {SAE.RPM} / {SAE.MAP.kPa} * {CALC.LTFTBEN}

where CALC.DAT.K is definend as {GM.DYNAIRTMP_DMA.C}+273.15

I'll edit the calc_pids.txt and post it later today.

In one cycle (4 strokes) (2 revs) all N cylinders are filled/emptied once.

Time for one cycle (2 revs):
t[s] = 2[rev] * 60[s/min] / RPM[rev/min]
     = 120[s*rev/min] / RPM[rev/min]
 
Mass of air thru engine in one cycle (for N cylinders):
VE[g] = MAF[g/s] * t[s]
      = MAF[g/s] * 120[s*rev/min] / RPM[rev/min]

Mass of air normalized for temperature and pressure in one cycle (for N cylinders):
VE[g*K/kPa] = VE[g] * DAT[K] / MAP[kPa]
            = MAF[g/s] * DAT[K] * 120[s*rev/min] / RPM[rev/min] / MAP[kPa]

Normalized airmass per cylinder (in an N cylinder engine):
VE[g*K/kPa] = MAF[g/s] * DAT[K] * 120[s*rev/min] / RPM[rev/min] / MAP[kPa] / N

When N=8, normalized airmass per cylinder becomes:
VE[g*K/kPa] = MAF[g/s] * DAT[K] * 15[s*rev/min] / RPM[rev/min] / MAP[kPa]


where: DAT[K] = DYNAIRTMP[K] = DYNAIRTMP[°C]+273.15[°C]

Units analysis shows the correct units on both sides of the equation.

So, in calc_pids.txt, after applying LTFTBEN and rearranging, VE[VE] becomes (for 1 of 8 cylinders):
VE[VE] = {SAE.MAF.gps} * {CALC.DAT.K} / {SAE.RPM} / {SAE.MAP.kPa} * 15 * {CALC.LTFTBEN}

The units [VE] is shorthand for [g*K/kPa]

Notice that VE[VE] is not dependent on air density or IGL.

[joecar]

To get VE[%]...

VE[%] is the actual airmass as a percentage of the theoretical airmass...

This can be calculated for the whole engine, or for an individual cylinder, the same result either way (the cylinder count cancels out)...

I did it for the whole engine as it is simpler:

Ideal Gas Law equation:
P[Pa] * V[m^3] = n[mol] * R[J/K/mol] * T[K] = (m[g] / M[g/mol]) * R[J/K/mol] * T[K] 

where:
    M = 28.96[g/mol] = molar mass of air
    R = 8.31447[J/K/mol] = Universal gas Constant
    T = DAT[K] = DYNAIRTMP[K] = DYNAIRTMP[°C]+273.15[°C]
    P = MAP[Pa]
    V = displacement[m^3] of engine (all N cylinders)

Theoretical air mass of engine displacement (rearrange IGL):
m[g] = V[m^3] * MAP[Pa] / DAT[K] * M[g/mol] / R[J/K/mol]

Note:
    1[Pa] = 1[N/m^2] = 1[J/m^3] since 1[J] = 1[Nm]
    1[kPa] = 1000[Pa]
    1[m^3] = 1000[L]

So, converting to [L] and [kPa]:
m[g] = V[L] * MAP[kPa] / DAT[K] * M[g/mol] / R[J/K/mol] * 1000[Pa/kPa] / 1000[L/m^3] 
     = V[L] * MAP[kPa] / DAT[K] * M[g/mol] / R[J/K/mol] * 1[Pa/kPa*m^3/L]
     = V[L] * MAP[kPa] / DAT[K] * 28.96[g/mol] / 8.31447[J/K/mol] * 1[Pa/kPa*m^3/L]
     = V[L] * MAP[kPa] / DAT[K] * 3.4831[g*K/kPa/L]

So, mass volumetric efficiency of engine (for N cylinders):
VE[%] = VE[g] / m[g] * 100[%]
      = MAF[g/s] * 120[s*rev/min] / RPM[rev/min] / V[L] / MAP[kPa] * DAT[K] / 3.4831[g*K/kPa/L] * 100[%]
      = MAF[g/s] * DAT[K] * 3445.2[s/g*kPa/K*rev/min*L*%] / RPM[rev/min] / MAP[kPa] / V[L]

Units analysis shows the correct units on both sides of the equation.

If you did this per cylinder, then the expression would be:
VE[%] = (VE[g] / N[cyl]) / (m[g] / N[cyl]) * 100[%] = VE[g] / m[g] * 100[%]

and it would come out the exactly same since N[cyl] divides out.

So, in calc_pids.txt, after applying LTFTBEN, VE[%] becomes (for 5.7L displacement):
VE[%] = {SAE.MAF.gps} * {CALC.DAT.K} / {SAE.RPM} / {SAE.MAP.kPa} * 3445.2 / 5.669 * {CALC.LTFTBEN}

Notice that VE[%] is dependent on air density (expressed by the IGL).

[WeathermanShawn]

Wow..

I think I need to break out some of my math books!!!.

I'll probably have to wait to see how you get it worked out. Its been a while since I crunched that much algebra..

[5.7ute]

Well I dont know what EFI Live displays unless someone has a stock file for an 05 4.8L The OS is 12592618....Quote from TC help file: The values is this table are expressed in grams per cylinder per second. These values can be converted to %VE which was used in earlier calibrations using the following formula:

%VE = 0.2871 x VE / CV

where VE is the value from the table and CV is the volume a one cylinder in liters.

It would be lots easier to conceptualize VE if it were in percentage.

Using edit/properties Efilive can display in
g*K/kpa
% of theoretical maximum
g/cyl
g/sec
I thought that by doing the log & converting the units in the Efilive tune tool you would be able to do a straight cut/paste into the tunercat file. The tunercat units make no sense to me. You would think it would be either g/cyl or g/sec.

[TunasTwins]

I thought that by doing the log & converting the units in the Efilive tune tool you would be able to do a straight cut/paste into the tunercat file. The tunercat units make no sense to me. You would think it would be either g/cyl or g/sec.

I just discovered the properties table and thought it was my aha moment for the day. It doesnt work however. I have made the VE table, smoothed it three passes and need to convert it back to grams/cylinder/sec and then go for a log. My MAF was nearly dead on but I'll make sure to take care of that too. Hopefully I can get some good/verifiable data for WeathermanShawn's theory.

[TunasTwins]

Here are the first two logs. both drives were about 25 minutes mostly interstate driving last night. I've looked at the fuel trims and have adjusted the VE table to get me closer in the lower map higher rpm areas. I am going to flash and log again shortly (hopefully). I have a remote start Viper alarm I've been meaning to get installed for two months so maybe I'll drive to the guy today. Its a good 30 mile drive so I should get a much better log this round.

[WeathermanShawn]

To me (other than KR) this looks like an good tune.

MAF, and Trims are nailed down (LTFTBENS), and when you look at his DYNAIR and CYLAIR they are as perfect as you can get.

The CALC.VE Table looks good.
Thanks TunasTwins...

[Wolfie]

What am I missing here...
the pids:
GM.CYLAIR.DMA
GM.DYNCYLAIR.DMA
GM.DYNAIRTMP_DMA
are not valid for me.