<... as any variability in the water head pressure to the pump will skew the test
results ...>
Do you really think so?  I mean significantly?  I would intuit that your task is to
pump some certain minimum volume of water through the system.  Say you had a 5- or 10-
gal supply tank, and kept the water level in the tank at about the elevation of the
input to the pump, and piped the output from the rads (right where it goes into the
pump) up over the edge of the water tank.  You could then turn on the pump and WATCH
how much volume was pumped through the entire system.  You're not simulating pressure
drop across the engine and radiators.  You're observing the flow that you get in spite
of all that resistance.  In a closed system, it doesn't make any difference what the
system pressure is.  In this test, you're using virtually no pressure at the inlet
side, and just a few inches at output.  You could document it with a flow meter, but
you'd get a pretty good go/no-go idea just watching it.
Just a theory .... Jim S.

Marvin Kaye wrote:

> Posted for "Haywire" <haywire@telus.net>:
>
> Hi Bill;
>         This is an excellent suggestion. Unfortunately I've already completed the
> installation. If I have time before I run the engine I will remove it and
> subject it to these tests.
>         I was thinking of removing the hose from one rad and putting it in water,
> then measure flow out of the other rad, but this would have to be carefully
> supplied water as any variability in the water head pressure to the pump
> will skew the test results. Water would have to pump from a container at
> approx the same height as the engine & have an assistant keep container full
> to a constant height with a hose.
>         I think the only meaningful test results would be from a flow meter
> installed in a closed system. I'll see if I can get a suitable rotometer
> from work, although this type of flow meter creates a restriction that again
> will affect the results. It would be alot of work but I may be able to hook
> up a mag tube flow meter for a test. This would give accurate results.
>         I will eventually build a "trendlink" program which would record all flight
> & engine data, then data can be shown as a trend and it's relationship to
> other factors is very clear. This may be the best tool for analysing the
> performance of this pump in various flight (load) conditions.
>
> S. Todd Bartrim
> Turbo 13B rotary powered
> RV-9endurance (FWF)
> C-FSTB
> http://www3.telus.net/haywire/RV-9/C-FSTB.htm
>
> > If you can set up a flow system, where the pump removes water from a 55
> > gallon barrel, has a pressure gauge on the output side, a
> > throttable valve,
> > and a means of measuring the flow volume (either a flow meter or
> > a bucket and
> > stop watch), you can get some very valuable data.