[urq] Re: Turbocharging & Elevations

[QSHIPQ at aol.com]

Dave:
If you look up the exact requirements for applying SAE 1349 correction 
factor, there is no way any chassis dyno is giving you the proper one.  SAE J1349, 
J1389 and J1995 have very specific requirements to make correction factors 
accurate.  These correction factors on commonly available chassis dynos are almost 
impossible to be accurate.  Inputs are required for engine specific fuel 
consumption, where and what sensors are placed, your exact method of cooling, and 
specific gravity of fuel in relation to the reference fuel.  In your world of 
91 octane that alone probably creates a large discrepency.

Dynos have some simple factors built into them, but they aren't engineering 
standards as applied to 1349, 1389 or 1995, they are rough corrections that 
helps one possibly compare one's car to someone elses.  That's a different 
purpose than you are trying to use it for.  Dyno companies don't routinely share the 
formulas they use to 'correct' it, cuz there would be a lot of data input 
that just isn't there, period.  The problem as I see it, is the raw data is more 
useful to you than trying to correct it.  A quick search of those using 
chassis dynos confirm, the best way to use a dyno is to use the same one, under as 
close to the same conditions as possible.  Don't 'correct' a darn thing.  

Most papers presenting a 'corrected' equation try to stay as close to 77 
degrees and 990mb for testing, then give a claim that a 5% variance to the results 
using SAE1389 correction factor.  I think that's just fine, since it requires 
someone else to do the hard work.  Me, I bet you have less than 5% variance 
testing your car on the same dyno at roughly the same temp, at whatever 
altitude your dyno resides..  

I'm not arguing that J1349 isn't a useful tool in engineering and SAE 
standards.  I just think that 1349 is useless and not yet applicable to a chassis 
dyno with a S car without many more inputs.  

Dynos good tools, I personally believe a simplistic and innacurate correction 
factor dulls the sharpness of it.    More fun with a dyno?  Do a before and 
after with a lightweight flywheel.  Massive increase in horsepower, no other 
changes.  Move 1 block down the street and put it on a different dyno under the 
same temp and pressure, different number.  Add in a turbo Scar on a chassis 
dyno designed  for 3500lbs of normally aspirated power, these numbers lose 
validity.  Pick one up there to tune with, and stick to it Dave.  Uncorrected is 
the only 'real' and repeatable number you have.

Peace

SJ


In a message dated 11/14/2004 12:37:29 PM Central Standard Time, Djdawson2 
writes:
I'm keeping it very simple.  I'm applying the same principals across the 
board... every time.  Listing variances leaves issues wide open to speculative 
interpretation.  Applying a mathematical principal to correct and minimize those 
variances closes the door.  You realize (I hope) that you're fighting the best 
tools developed by the most organized engineering groups.


Several of us 'old timers' can look at these dyno results and conclude that 
power curves are less than ideal.  You and mike 'may' have more work to do.  
The question is where?  You automatically think that needs to be the dyno, I 
don't necessarily agree.  Finish the paper first, you're already jumping to 
conclusions.


It doesn't take an old timer to evaluate a dyno sheet.  If I were to purchase 
an engine package based on 1/4 mile times, I'd be in a bad place in the end.  
"Hey, that EVO does 10 second 1/4s."  Great... look closer.  It launches a 7k 
and shifts at 9200.  That has absolutely no value to me.  Similarly, Javad is 
tuning an engine.  It's purpose is unknown to me... but "grocery getter" I'm 
betting it's not.  Look at his dyno sheet... you want that in your daily 
driver?  Not me.  It may meet his goals, but it wouldn't meet mine.  I couldn't 
know that based solely on his 1/4 mile times.

I'm gonna sign off this thread with a simple agreement to disagree.  
Correction factors are a simple scientific principal.  They are required to interpret 
data.  They are found in EVERY aspect of scientific pursuit, and that suits me 
fine.
Take care,
Dave in CO