Date: Mon, 20 Feb 2006 17:23:41
To: Prius_Technical_Stuff@yahoogroups.com
From: hobbit@avian.org
Subject: warm-air testing results, finally

Global warming certainly makes more difficult to do cold-weather testing
in these parts, but I finally got a chilly enough day to go out and really
determine whether or not this warm-air intake hack has any real merit or
not.  I needed a day to meet various conditions: 1> cold enough, with an
afternoon high still at least subfreezing; 2> clear weather with low winds;
3> having *time* to actually go out and do some test runs.

Test location: that same stretch of I-95 through Topsfield/Georgetown in
northeastern MA, which is nice and straight and not too traffic-choked so
I could hold a constant speed for most or all of it.

Test purpose: two identical round-trip runs, one with the warm-air intake
in place and one with the stock [cold] air intake that comes from behind
the RH headlight, all other conditions being held as constant as possible,
to measure any differences in mileage.

Test framework: near the end of a tank of gas, fully warmed up, battery
SOC nominal 60%, windows closed, no cabin-heat demand other than whatever
happens to leak through the vents in fresh-air mode.  Ambient temp indicated
on the MFD bobbled between 28F and 30F during the entire test duration,
meaning it was probably more like 26F since that tends to read a little
high.  Laptop monitoring Intake Air Temperature [IAT] continuously so I
could determine the delta between the two intake modes.  This time, since
I was going to tank up at the end of this test anyways, I reset the MFD
consumption screen and let the car measure average mileage for each run.

Each run was intended to be a full round trip back to the same point, to
go the same distance over the same terrain and eliminate prevailing-wind
effects on mileage.  Speed was held at a constant 65 mph with the cruise
control once out on the highway, with occasional dips to 64 to drift away
from *other* people who had locked their own at the local speed limit of 65.
About the same amount of 64 vs. 65 was done on each run.  Braking for the
main deceleration for all highway exits was done in Neutral, to avoid
pushing a lot of energy into the battery and keep it at its nominal 60%.
[Slower local driving to turn around still involved a little regen braking,
but much less than getting off a highway exit ramp.]

Run 1 [warm-air hack]: 33 miles, 47.7 mpg.  IAT bounced around quite a bit,
but averaged 74F over ten or so times I glanced at the laptop and scribbled
down the current reading.  There must be a bit of interesting turbulence
back there around the exhaust header.

Run 2 [stock intake]: 25 miles, 47.9 mpg.  IAT started around 40F and quickly
drifted down to hold very steady at 32F.  This trip was cut short because
the low-fuel warning came on during the "up" leg of the run and I pushed it
as far as I dared on the "back" leg before pulling off at an earlier exit
with a gas station a few miles short of my return point.  But I had already
observed the accumulated mileage average doing almost exactly the same thing
as on the first run, and I believe 25 miles is still enough to eliminate most
transient effects from terrain.

Along with tanking up, I pulled the plastic trim rings off the wheels, as
a completely unrelated change but inspired by a recent Priuschat thread
about how they just let dirt accumulate behind them and seem unlikely to
contribute all that much to airflow mitigation aroud the wheels.  I also
pulled off the stock intake tube again and left the airbox inlet just sitting
open, with neither tube hooked up.  [Clarification: yes, air filter and
air-box cover still firmly in place -- I certainly wouldn't run around
with the throttle body completely nekkid to the elements!]

Without any intake plumbing at all, IAT once again fell and held steady at
32F even after the coolant temp was back up to 85C as shown on my meter.
This tells me that airflow into the area around the air-box apparently does
*not* come through the radiator, or I would have expected IAT to rise again
once I had been under way for a while.  Air through the radiator must take
a different path, and go out underneath the engine or something.

I continued back toward the original startpoint, but breezed by its exit and
headed toward home.  Average was 42.4 mpg at that passing point but as I
noticed that it always seems to start low and climb after a reset, this makes
sense.  I continued on past there with the cruise set at 65 as long as I
could, until having to slow down for more traffic.  That was maybe 10 miles
later, by which time the average was at 47.0.

Conclusion: the warm-air intake kit does absolutely NOTHING for the steady-
state efficiency of the Prius.  I would posit that with the lower intake
temperature and denser air, more fuel would be injected per charge to
maintain a stoichiometric burn, but since that produces more power then
the throttle would simply be closed a little more to keep engine output
in check to hold the same highway speed.  The tradeoff either way apparently
balances to zero out on the highway.  I believe I had a significant enough
IAT delta between runs that if there was to be any effect on mileage, I would
have easily see it.  Unfortunately I didn't try to log and then average
out a TPS reading, which might have been revealing, but it wouldn't have
really mattered in this mileage-comparison test anyways.

So, do we have any hard numbers from the radiator-blockoff side of the
house?  We don't have too much winter left, if you wanted to get out there
and collect some data...

Further tests to do might be trying to time full warmups from dead-cold
to 85C, with and without the warm-air intake.  That may be where it could
still help.  Some means of eliminating the thermos bottle's contribution
may be useful -- such as letting it pump in, and then immediately shutting
down for a while to let the block lose that stored heat and begin from a
true "cold soak".  It seems that on really cold startups the system isn't
bothering to pump in at all anyways, which seems odd since I'd expect the
bottle to be able to hold at least *some* heat while sitting for, say, only
6 hours.  Perhaps the cold penetrates far enough in to the storage-tank
outlet sensor that the system makes the [possibly erroneous] decision that
there's no stored heat worth pumping??  You'd think they would have dealt
with that...

On the way home from all this, a largish green SUV passed me, with
ENVIRONMENTAL POLICE emblazoned across the back.  Hee hee...

_H*