Josh Sirota wrote:
>
> Folks,
>
> People on the Z3 discussion board argue incessantly about whether or not
> dyno runs on these modern, OBD-II equipped cars are valid. Some say
> that the systems adapt so quickly that a dyno run tells you nothing, and
> others swear by the money they've spent on the dyno that prove that the
> $1000 they spent on an exhaust system was a win.
>
> Do any of you rocket scientists really understand how these modern fancy
> emission systems work and can you tell me with a fair amount of
> certainty whether or not it's a waste of money to dyno one of these
> cars?
Waste of money? That depends upon what you are wanting to learn.
If you are wanting to learn whether you get 3-5hp from an exhaust change;
yes, its probably a waste of money. First, the dyno itself has that much
imprecision. Thus, to get meaningful test results, you really need to run
numerous before and after tests to statistically reduce the dyno error.
But unless you own the dyno, that can get expensive.
Further, if you are wanting to test any modifications that effect air flow,
then you have to combat OBDII -- it _will_ indeed adapt to changes in air
flow with the goal of optimizing emissions. Does that invalidate the concept
of dyno testing? NO! All you really care about is rear wheel torque/HP, and
that's exactly what they dyno measures. However, it _does_ invalidate the
typical uses of the dyno:
dyno - make change - dyno - make change - dyno - ... - time's up
The problem with that is you are testing bogus conditions. The first test is
somewhat meaningful. The second test is only showing "how does my car perform
immediately after I've changed it from this way to that -- how does it perform
when I have this change but the car has not adapted for it yet." Unless your
plan is to make that change right before your first autox run, its fairly
meaningless.
So, what you need to do is:
dyno-dyno-dyno - make change - drive a couple hours - dyno-dyno-dyno - ...
Unfortunately, even that is somewhat flawed. Why? Because its fairly important
that during the "drive a couple hours" that you drive exactly the same as you
did
the couple hours before the previous dyno. And that the ambient conditions
during
the two dyno sessions and the couple hour drives are the same. That's tough.
Even more precisely, if you're worried about autox, you want to drive just like
you
do for the couple hours before an autox! Your car will perform differently at
the
autox depending upon whether you drove the hell out of it the previous several
hours
of driving or drove it like a grandma. That variance can easily swamp the dyno
results!
My recommendation: Buy a G-Cube, mount it in your car. Measure your
performance
right before and right after each autox you attend. Average them for you
performance
for that day. After a few weeks of this... make a change and then continue
your
testing. After a few weeks, draw your conclusions.
The advantage of this is that it tests the conditions you really care about:
given
how I drive during the week, how will my car perform on autox day? Of course,
this
is still flawed due to weather changes, so you have to be careful to record the
ambient temps not only when you tested, but also all week long.
The bigger advantage is that the G-Cube costs much less than a $1000 dyno run
and
you can also use it for analyzing your driving performance -- and most of us
will
get bigger gains from that than any exhaust change we could do.
Cheers,
Brian
P.S. The G-Cube is the soon-to-be-available g.Analyst successor built by the
maker of Geez! software (email team.net member Byron Short at
bshort@AFSinc.com). No association other than soon-to-be Geez!
customer.
P.P.S. No, I am not the OBDII authority that you asked for; if you want an
authoritative answer, ask Jim Conforti to respond; that consistently
shuts
down this same debate very quickly on the M3 list; I would guess it
would
have similar effect on the Z3 list.
|