"Schertz, William W." wrote:
I set up my evaporator cores, hooked them to a garden hose, and did flow
measurements and pressure measurement tests on them, and have overlaid
the
pressure rise with flow rate across the core onto the pump pressure
curves
that I had previously published. These are shown in the attached figure
"Pump_curve_core_data_clr.jpg".
Some items to take note of on the chart.
1. At full tilt - 5594 rpm, the pump which can deliver 44 gallons per
minute
at zero head, can force only ~20 GPM through a single core, and ~31.5
GPM
through two cores in parallel.
2. I have plotted the actual data points taken on the chart. They are on
the
"One Core" line, and I was restricted in how much water I could force
through the core by the capacity of the water line in my hanger, and the
length of hose leading to the core. The curves are then, obviously an
extrapolation, using the fact that the pressure drop is expected to be
proportional to the velocity (flow-rate) squared. The data points are a
combination of the results for the two separate cores, combined for the
data
fitting.
3. At lower RPM's the flow is reduced dramatically, and these curves
make me
question the use of the "under-drive" pulley on the engine, to 'slow
down
the water pump'.
4. I did not have an ammeter to measure the current into the motor, but
the
1-hp motor did not have any problem driving the pump at ~5600 rpm.
5. I would not advocate having the radiator cap on the pump outlet side,
the
pressure that the cap would see would be the system pressure + pump
pressure, so it would be easy to create a condition where the radiator
cap
could be overcome and fluid vented. The more I think about it the
better I
like Ian's & Lynn's approach of using a pressurized bottle with a
head-space.
6. These curves show the advantage of having the cores in parallel, even
though I know of a good working example of cores in series (Ed
Anderson's).
I suspect that Ed's use of AN-16 fittings has contributed to a lower
entrance and exit losses than I am experiencing (basically transitioning
from 1-3/8 to 3/4").
Comments are welcome.
Bill Schertz
Keep up the good work Bill. We will soon have a ral handle on this
cooling problem.
Paul Lamar
Great work, Bill!
Yes, I am using AN-16 fittings on all my coolant system. It's really
good to see some actual data on the evaporator cores. It would appear that
if you are going to use them, then using them in parallel is a better
approach. Good point on underdriving the water pump. I believe my water
pump is under-driven by 20% as I am using the Racing beat main pulley (which
they claim underdrives by 20%). Given there is less need for coolant flow
at lower rpm (power) settings I am not alarmed by the fact the flow appears
to be decreased dramaticly, but do wonder whether 20% more rpm would help
during high power climb settings.
Tracy uses two in parallel using 5/8" dia lines (I believe) with success,
however, I know of several people who had problems with plumbing radiators
in parallel regarding getting the coolant flow to divide equally between the
two. It appears that the flow will "prefer" one or the other radiators and
most of the flow will go to one or the other. Further that apparently this
flow can change from one to the other radiators depending on flow rate
(rpm). One individual even tried a valve at the "Y" in order to fine tune
the flow to ensure it went equally into both radiators but found that while
they could set it for a narrow rpm range, that changing rpm further would
once again cause it to diverge.
I went with the series radiators mainly to simplify the plumbing. However,
your data makes me wonder how much resistance two in series are cutting my
flow rate. We know that the flow rate is the same in both serial radiator
cores. So, I suspect the second one may not further effect the flow rate
signficantly, but don't know for certain. So if like electronic resistors
that would imply that each core would drop 1/2 of the total core pressure
drop. We do know that whatever pressure is not dropped across
resistrictions in the engine and hose is dropped across the radiator cores,
whether two in series or two in parallel. However, in parallel then the the
total flow resistance would be probably close to 1/2 of two in series.
Anyhow, really appreciate your time and effort and sharing with us your
data.
Best Regards
Ed Anderson
I second that Bill.
The obvious is now repeat the test with a real
water radiator :-)
Actually I was rather amazed way back when I first
heard of Tracy using evap cores that they worked at all. The flow rate
taken by itself may be a relatively minor factor in the overall cooling
system. I suspect the measure of a good heat exchanger is the product
of the flow rate times the fin area. There is no doubt the fin
area of an evap core is high for its over all volume. I still doubt
that an evap core is an optimum solution and Tracy's on going troubles
with cooling suggest that. However I can see other problems with
Tracy's system such as air flow configuration so we are still in
the shade if no longer in the dark.
The next part of the equation is the water and air temp drop through
the cores. Both Ed and I are working hard on measuring that.
Paul Lamar
The AirCraft Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site. http://home.earthlink.net/~rotaryeng/
Copyright 1998-2002 All world wide rights reserved.