gameover

if any MAF DTC is set the VE table is used directly, there is no complex filtering or scaling of the value.

i think you need to do a detailed analysyis of the units in the equation. You have a bunch of extra unknown factors thrown into the equation like if you convert the formula to g/cyl you lose the RPM and pickup a factor of 15. Also i still can;t work out how the extra factor of 8 you get cos you use the displacement of the engine rather than the cylinder volume... Combine that with the factor of 17.883 you have and maybe 17.833/15 is the missing 1.18 difference in the calcs??

I don't know for sure cos i still can't work out the units of that equation sorry... i have tried for a while now

NoGo

iTrader: (2)

I get the units to work out if I divide out R (which essentially makes it your equation), but other than that.........

Your formula definiantly comes out sexier on paper, but has some issues matching the real world data.

Mine has some issues on paper, but matches the real world data very well.

I guess the true test is going to be the "road" test. I will convert another open loop car to MAFless using my formula and this time and see how the actual a/f compares to the target a/f from the lamba table (open loop multiplier table). If it is dead on, then you'll have to get the hex-editor warmed up. If it is about 15% off, then obviously yours is correct.

Billiumss

iTrader: (11)

Um, my brain hurts!!!!!

Hey Kevin, any chance I could come up to your shop for a day and teach me about tuning and LS1 Edit? I'll buy you lunch! lol

PM me if possible...

Bill

deezel

If 2466 (or 12626) is the 100% VE value for a 5.7L engine, why do so many seasoned tuners think it should be close to 2900 or 3000? And why do calculations from logged data agree closely with the larger value? The value of 2466 is about 15% off from what most people have been using.

Using the above formula, 100%VE for a 5.3L would be (0.6625)*17833 = 11814 (or 2307 in Edit table numbers). I am looking at a stock 5.3L VE table that has 2162 as its maximum value. This translates to 93.7%VE! However, if you use 2903 as the 100% value, you get 74.6%VE. This is much more in line for a stock engine.
Similar comparison for 6.0L engine: formula above gives 100%VE = (0.75)*17833 = 13375 (or 2612 in Edit values). Max value on stock 6.0L VE table is 2679. This gives %VE = 102.5%! Not possible on a stock engine. But 2679/2903 = 92.2%VE (or 2679/3000 = 89.3%VE). Both a little high, but believable.

I think we are still missing something on the displacement and RPMs. You can get grams/cyl from MAF and RPM (using measurement from MAF) or from %VE, MAP, IAT, RPM and displacement (from Ideal Gas Law). But, this formula
{g/cyl x 8192 = VEvalue * MAP * 51.2 / (IAT * 32)}
seems to only use VE, MAP and IAT. I don't see how you can get to g/cyl from just VE, MAP and IAT. There has to be some form of RPM and displacement figured in there somewhere.

Bink

Bill -
Let me know if he accepts. If you'll let me attend the mini seminar I'll buy lunch and dinner!!

I'm grinning But not kidding.

joel

Bink

deezel-

Is the RPM and displacement incorporated in the VE Value - in the above equation?

joel

gameover

that's the point, the VE table is not VE at all. You don't need RPM to calculate g/cyl you need it to convert from g/cyl to g/sec.

Lets talk about the cranking VE table for a second and the PCM calculation for that.

The cranking airmass calc in the PCM is:
g/cyl = VEvalue * V * MAP / T

Where V is the Cylinder Volume of 0.708L.

Now this VE value is more like a true VE% because it doesn't include the Cylinder Volume. ie. it is the proportion of the cylinder volume. For the Main VE table the volume seems to be included in the value. Inconsistent and strange indeed...

deezel

I suppose it could be. Lets see...

The simplest form of this would be Table VE = %VE * RPM * Cylinder Volume.
This matches NoGo's equation.
Using NoGo's first data point again (5600 rpms) and the %VE calculated from the MAF reading (81.7%) we can test it.
Actual Table VE = 2349 and cylinder volume = 0.708L
Calc Table VE = 0.817 * 5600 * 0.708 = 3239 ... doesn't match.

I need to do some more math...

deezel

Ok, I agree that RPM is not needed. I have convinced myself of that with pencil and paper. Still trying to put the Main VE table in terms of %VE, cylinder volume, and some constant...

Thanks for the cranking airmass equation gameover.

Bink

deezel- when I first read gameover's original post to the "sticky" thread I was on my way to bed - thinking about it. I Woke up next day and thought I could understand it. Wrote down some rough calculations and posted my thoughts to the sticky thread. Your posts today got me to thinking . I checked my early post to the sticky.... In that equation there is no RPM....but of course there is VE. How could I have logically eliminated RPM?? Because VE values are from the look up table and are RPM determined. Of course gameover's equation doesn't have RPM - it uses the referenced value at a particular RPM and kPA.
Gameover's PCM formula can't give us the VE values referenced by a predetermined table (at best it should only be accurate to 5%, 5kPa intervals -with the VE value known). But knowing how the VE values are used, by the PCM, we can properly utilise NoGo's VE equation.

Does this make sense?? I think we're trying to compare the base equation (NoGo's)to the functional equivalent (gameovers).

I'm not addressing this to NoGo also because I've already made too many stupid statements!!

Thoughts??? joel

deezel

Bink - yeah, I tried thinking about it that way, too. "Since the LUT uses RPM and MAP, maybe thats where the RPM comes from..." But, that doesn't explain where the RPM's go in the math. I agree with what gameover said - RPM is only needed to convert g/s to g/cyl.

If you take NoGo's equation: g/sec = VE*MAP*DISPL*RPM/IAT
and convert to g/cyl: g/cyl = (VE*MAP*DISPL*RPM/IAT) * (15/RPM)
= 15*VE*MAP*DISPL/IAT
you see the RPM's cancel out.

I'm still thinking about how the cylinder volume figures into the VE Table value...

Bink

Okay, but we still need RPM to determine g/cyl. How else would we do that??

VE = (g/cyl) / (15*MAP*DISPL/IAT) = (g/cyl * IAT) / (15*MAP*DISPL)

We still need RPM to determine g/cyl...right?? This equation won't work with out an RPM reference.
joel

deezel

Like you said before, the RPM reference is the look up table for VE. So, I don't think its needed... The 1/cyl units come from the cylinder volume (L/cyl).

Now we are getting close to an equation that resembles gameover's PCM calculation:
{g/cyl x 8192 = VEvalue * MAP * 51.2 / (IAT * 32)}
or, g/cyl = (TableVE * 51.2 / (8192 * 32)) * MAP / IAT
............ = (0.0001953 * TableVE) * MAP / IAT

I think units of MAP and IAT are kPa and K, respectively...

Gameover says cylinder volume is part of TableVE, so lets rearrange our equation (based on NoGo's calculation) above to this:
(g/cyl) = (nogoVE * 15 * MAP * DISPL) / (IAT)
..........= (15*(nogoVE * DISPL) * MAP) / IAT

So it looks like (0.0001953 * TableVE) from gameover's formula equals (15*nogoVE*DISPL) from our formula with possibly some units conversion factors needed....

HumpinSS

iTrader: (1)

TTT Any new updates

deezel

Ok, a little more thinking about this...
NoGo's massflow calc uses EditVE.
gameover says that TableVE (which is 5.12 * EditVE) contains displacement, so we could write it as
TableVE = %VE * DISPL * C1
From the PCM equation: g/cyl x 8192 = TableVE * MAP * 51.2 / (IAT * 32)
we can guess that the units on TableVE are (g/cyl)(K)/(kPa).
And since cylinder volume is L/cyl we can guess the units on the constant C1 above to be (g * K)/(L * kPa) = (g * K)/(J)
This is almost the units of 1/R, the gas constant. All we need is another constant, the molar mass of air M (28.96 g/mol) to have R's units of J/(K * mol).

So, it looks like TableVE = (%VE * DISPL) * (C2 * M / R). If we know a few good data points for TableVE vs. %VE, we can determine the constant C2...
Using NoGo's data again - (5600 RPM data point)
EditVE = 2349, MAP = 100kPa, IAT = 300K, MAFcylair = .6696 g/cyl (~81%VE), TableVE = (5.12 * EditVE) = 12027
solving for C2
12027 = (81%VE)*(.708 L) * (C2) * (28.96 g/mol) / (8.314 J/K mol)
C2 = {(12027 g*K/(cyl*kPa)) * (8.314 J/K mol)}/{(28.96 g/mol) * (.708 L/cyl) * (.81)} = 6020.75 (unitless)
So,
TableVE = 6020.75 * (%VE * DISPL * M) / R and,
EditVE = 1175.9 * (%VE * DISPL * M) / R
Both TableVE and EditVE have units of (L*g*K)/(cyl*J), and you can see the displacement in there (L/cyl).

I haven't rigorously tested these values... just extracting the units from the math we have been talking about.
cheers,
deezel

deezel

More math..

EditVE = 1175.9 * DISPL * M/R * %VE = 4096 * DISPL * %VE

Now that looks like a hex-scaled PCM calculation.

The above equation for EditVE was derived from gameover's calculation from the PCM code. Previously, I determined from NoGo's massflow equation that EditVE = %VE * 2902. If you take the newly derived equation and figure in cylinder volume, you get the same result.
EditVE = 4096 * (0.708) * (%VE) = 2900 * %VE

So, I think I have shown (in a long roundabout way) that NoGo's massflow calculation (based on the Ideal gas law) is consistent with gameover's PCM calculation.

This means that:
EditVE = 4096 * (DISPL) * (%VE)
TableVE = 5.12 * EditVE = 20971.5 * (DISPL) * (%VE)

And that the 100% EditVE value for a given engine is 4096 * DISPL (in L/cyl).

GrannySShifting

iTrader: (10)

Eitehr that or max VE in a 5.7 liter table divided by deezels definition of 100% VE (2918 for a 5.7 liter)

2345/2918 is roughly .83 or 17% right? Is one of you solving for 100% of max VE of this particular engine, and the other solving for an outright percenage of full volumetric effeciency?

marcink

Any consensus on this subject yet?
I am especially interested if the cylinder volume is somehow included in the VE table values (LS1 Edit) as my stroker motor is being build right now...

iateyourcat

I just want to know where I can buy the spreadsheet.

deezel

I'm still interested to see if NoGo, gameover, Bink, or anyone else may have any comments.
I'm still confused about gamover's definition of 100% VE with the 17833 multiplier. But I think my definition of EditVE holds up for a 100% VE calculation. I will show this below...
Previously I posted:
Based on my further analysis the 100% VE values would be these:
EditVE = 4096 * (DISPL) * (%VE)
100% EditVE = 4096 * DISPL (displacement values from Edit files used)

5.3L - 4096 * (0.66592) = 2727.6 = ~2728
With 2162 max in the VE table, that gives a 79.25% max VE on a 5.3L. Better than the 74.6 I had before.

6.0L - 4096 * (0.74588) = 3055
With 2679 max in the VE table - 87.7% max VE on a 6.0L. More beleiveable than the 89-92% I had before.

I assume 0.708 is the Edit value for cylinder volume on a 5.7L (from gameover's post). (I can only read truck files)
So that would make 4096 * (0.708) = 2900 the 100% VE value for a 5.7L. And 2349 (from NoGo's data) gives a VE of 81% at 5600 RPM's for that car, which matches the VE from NoGo's massflow calculation as well.

Comments, corrections, and criticisms welcome.