RomRaider

your engine seems to be breathing quite well.

do you have a mod list handy?

also, in general, if people have recalibrated their maf sensor translation table, we'd need to know that to make valid comparisons.

ken

did some crunching with some of the info from jeff's site this morning.

i think i've got some good results.

================================

first,

i used the info from from http://www.stealth316.com/2-air-fuel-flow.htm to give me a good equation for air density:





disregarding pressure, this gives us an equation of:

density--simple form (no pressure correction)


adding pressure correction:

density--with pressure correction


converting to mrp instead of map (this equation assumes you are at sea level!)

density with mrp instead of map


all of the above equations result in units of g/L.

===============================

next,

take logged mafg (g/s) divide by calculated density (g/L) to get true volumetric flow rate, l/sec

===============================

now, calculate what the engine SHOULD be inhaling--ie, theoretical volumetric flow rate

rpm/2 * 1.994l = taf (l/min)
(rpm/2 * 1.994l) / 60 = taf (l/sec)

===============================

finally, divide true volumetric flow rate by theoretical volumetric flow rate to net VE.

===============================

after plugging and chugging and scatter-plotting on a 15 minute log file, i got this:



i didn't like the big scatter involved, and although in theory the throttle plate doesn't make a difference anymore (since we are taking air density at the manifold into consideration), i eliminated all tps <50%. this looks much cleaner:



i think with a lot of lines of log at 100% tps one could build up a quite accurate VE chart pretty quickly. the log file i used was a short commute into work and didn't have many lines >50%tps, and NO lines at WOT.

i think the big advantage of the above method is that the "artificial" inflation of VE is already corrected for by calculating density using boost pressure as a factor.

key assumption: iatc represents matc--how good your IC setup is and how prone to heatsoak may have big effects here. iatc may be in fact higher or lower than matc which cannot be accounted for.

hth
ken

My maf table is stock. EJ257 shortblock though on my WRX heads, so my displacement is close to 2.5L. I think that's why my numbers are higher than yours and so much like the Forester XT numbers - there is about 20% more load than your 2.0L, which makes sense due to the 20-25% more displacement.

Mods I have include TGV delete, knife-edge throttle body inlet, ported WRX heads (light porting), ported stock exhaust manifold, TD05H-18G 8cm^2 turbo, and eBay TMIC.

I would expect to pick up some VE from all that porting, but who knows? Maybe some of it has to due with my altitude? (~1350 ft)

'06 WRX
stock td04

~ 0.102 over the life of my car ---- pretty close to what Airboy got. It's interesting to see the slope changes over some of the different setups and tunes the car has seen.

Here's mine '06 WRX; bellmouth DP w/hiflow cat; IAT ~68* F.



So ride5000 in the above equations you are attempting to calculate the VE of the engine if it was NA?

I have been using pretty much the same equations only using atm pressure in place of MAP to calculate the volume of air the turbo is sucking in, so I can overlay the data on the compressor map. Seeing you using MAP has me a little worried that I have been going about this all wrong.... Anyone know if the m^3/s on the compressor map is supposed to be air taken in or air pushed out?

that's about a 0.19 slope in the 16bit scale... nice!



exactly. by using the manifold pressure to calculate the density you've eliminated the "boost" factor.



on compressor maps that use volume instead of mass for an X-axis, the reference is "air taken in."

Yeah not quite as steep as a couple of the other guys with 06s, but I was pleased by my Y intercept... looks like my car is breathing well when in vacuum and low boost for whatever reason.


Phhhewww would have sucked if all my calculations and graphs were wrong. Thanks for the info and starting this thread it is interesting to see how the cars stack up... Hopefully some more people contribute.

I end up somewhere around 0.8 to 0.9.

I've been adding a VE column to some logs, then importing them back in my log analyzer, and displaying VE data on a MAP vs. RPM table. Do this on two logs with different AVCS maps, then compare what the VE table looks like to see which one is flowing better. It seems pretty effective.

I think I just use:
(Displacement(L) / 2 * 1.292g/L) / MAF g/s
Though I don't have access to the file right now, that seems right. I just used STP for air density. I have also sometimes MAP correct the number (just take the above * MAP(mmHG)/760mHG. Again, not clear if you should use 760mmHG or the atmospheric pressure you have at your altitude on a turbo car, so I went with STP as my reference state. The ECU uses 760mmHG commonly as well.

another scatter plot... a few more points this time from a log i took yesterday. tps 60% and up.

I forgot to post my results too.

Here is a 2005 USDM LGT. Mods are AEM CAI, Borla Headers, and Crucial shorty downpipe. Stock 3rd cat and stock catback.

just so folks know, i noticed this morning my f*** webhost is having dns issues AGAIN.

i'm shopping for a new one right now.

the image links and whatnot will be back soon.

-ken

I hope no one minds if I bump this thread up.

Bump.