|
|
Mark, I'll embed comments in your e-mail below - and add some words about
"best coolant leak indicator", triggered by Finn's e-mail earlier this
morning and comments by Ed recently.
David
----- Original Message -----
From: "Marvin Kaye" <marv@lancaironline.net> [ for Mark S. ]
To: "Rotary motors in aircraft" <flyrotary@lancaironline.net>
Sent: Thursday, March 04, 2004 8:23 AM
Subject: [FlyRotary] Re: Ideal Cooling System Plumbing (was Re: [FlyRotary]
Re: overflow connections
> Posted for Mark Steitle <msteitle@mail.utexas.edu>:
>
> David,
> I guess it comes down to how soon the system will go into melt down. I
see
> it like a water cooler... why do they put the spigot on the bottom instead
of
> the top. You'll get more water out of the cooler if you draw off the
bottom.
> So,
> in your cooling system, by drawing off the top of the radiator,
The system I propose draws COOLANT off of the BOTTOM. The AIR goes out the
top, at least when you are filling the system on the ground with no flow.
The main purpose is to be able to FILL the system and PURGE ALL AIR - right
out the top of the engine - the highest spot you can find.
>if a leak were to develop you would run out of water sooner than if you
were drawing off the bottom.
If we are talking about a slower leak (not a hose blowing or coming off,
i.e., NOT and INSTANT LOSS of all coolant), then if the coolant level is
simply steadily decreasing in the engine block, then, with both my a/c cores
below the engine, the cores will be the last to empty and the pump is
pulling out of the BOTTOM of the cores.
- The engine seals will already be toasted by the time the pump sucks
the last coolant out of the radiators.
> From my brief experience with the Mazda rotary water pump, it doesn't
appear to
> take much air in the system to shut the pump down. I would like my system
to be
> able to continue circulating water for as long as possible, to enable me
to get back
> on the ground.
>
> I remember when auto manufacturers switched to cross-flow radiators. It
was
> normal for the coolant level in those systems to be about 6-8" below the
top
> of the tank. If they tried to make the top hose the outlet, they would
suck
> air and overheat/ruin the engine. I don't know if your system would have
this
> same characteristic or not.
- My next comments below will explain why there shouldn't be any
significant air in the rads. HOWEVER, if plain old EXPERIENCE were to ever
indicate that air was being trapped in the top of rad, then an AIR VENT LINE
could be installed at high point in rad or a hose there that would feed back
into the bottom (below liquid level) in pressurized expansion tank - would
be a "2ND AIR VENT LINE", with same function as the air vent line from
highest point in engine block back to the expansion tank, below liq level.
Both lines would be flowing a small amt of coolant that would be bypassing
the rad, not being cooled, so, again, the caution to use SMALL AIR BLEED
LINES or RESTRICTOR in the lines.
> But, if you were to get a significant amount of air into the radiator, it
would have the same result.
Air in the system: We have hoses at the water pump inlet and outlet, and
the pump is high up on the engine in a "plugs normal installation".
- Where does the hose from the pump outlet feed back into the engine
block - high or low on engine? (Basic question that I can't visualize
because I don't have an engine yet)?
- Likewise, where is the "hot coolant outlet" from the engine block
where the hose connects to the radiator? Should be high up near the top
like on all cars?
Anyway, the assumed short hose from pump outlet to somewhere on engine could
have a leak; likewise, the hot coolant hose that runs DOWN from near the top
of block to the rads could have a leak anywhere below that outlet from block
all the way down to connections to the radiators.
- As long as there is a head of coolant in the engine, there is going to
be a head of coolant down at the rads, and so the pump is going to have
coolant at the bottom of rads to suck UP to the pump. I'm assuming there
won't be any air at the BOTTOM of the rads because the air should never
reach the rads until the top of the coolant head leaks out down to the level
of the rads. Air "bubbles", on the other hand, might be circulated to some
unknown degree - BUBBLES may or may not "counterflow" UP to the block - if
the cores are sloped 10 to 15 degrees as suggested then at least there won't
be any "pockets" for collection of air in the rads and hoses near them.
- I don't envision any significant pump-stopping air at the bottom where
pump is sucking from as long as there is any coolant in the block. The
pump is going to suck from bottom of rads and pump out of its outlet into
the engine block and/or out the "leak" until there is ZERO coolant left in
the rads, at which time the pump stops ciruclating - but not because of
"system configuration", rather for lack of liquid.
- Again, I simply assume any AIR THAT IS BEING SUCKED IN AT THE LEAK
will be up ON TOP of the coolant, not down at the bottom of the liquid
column where the pump sucks from.
> Or maybe I don't fully understand your design?
>
> Mark S.
>
Now, to address the discussion of "What is the best indication of a leak?"
1) TEMP - sensor closest to the "combustion chamber" that goes berserk when
coolant level drops - the "coolant temp gage that changes to a CHT gage when
the coolant level drops to uncover it". This sensor, and its proper
"implementation" into your gage and/or warning system, is quite good.
2) PRESSURE - drops below "cap relief pressure" - Good, IF YOU ASSUME that
a SLOW LEAK won't let the system keep PRESSURE up while COOLANT LEVEL is
going down.
- I recently replaced the radiator on my '93 Ford Escort that was
leaking right in the middle of the core - took months to find. At first
could smell the coolant and had slow decrease in expansion tank until float
swithc turned on the "check coolant" light. Shop couldn't find any wet
spots. Obtained Magnaflux water soluable dye (Zyglo), and finally found
it - SYSTEM WAS PRESSURIZING, but COOLANT WAS GOING DOWN IN THE
(unpressurized, older style EXPANSION TANK. Really only found leak when it
"let go" and punched out a flap of aluminum fin and REALLY leaked on
driveway and let out hot coolant that showed steam coming out front of
radiator. Now I have a "lizard green carport"! (0nly temporary - it washed
off.)
3) LIQUID LEVEL - switch to turn on light.
- My old style ('93, '95 Escort) unpressurized expansion tank has fluid
level sensor (don't know if physical float or thermister) - in event of a
real leak, where coolant is being lost, the expansion tank will have coolant
sucked out after every engine shutdown/cool down, until the level goes down
enough to activiate the level sensor.
- I'm assuming that a more modern pressurized expansion tank (late 90's
Ford Taurus, for example) will also have a float or thermister to activate a
coolant level warning.
- THIS IS THE BEST SYSTEM: It tells you coolant level in EXPANSION
TANK has decreased too much, BEFORE the level in the BLOCK has even begun to
suck air. This is the ideal warning - warned BEFORE the engine is damaged,
gives "more" time (depending on speed of leak) to take action to land and
save the engine, if that is possible, given where you are and where the
closest landing spot is.
........-- It remains to be seen if the stock Ford pressurized tank will
fit under the cowl - or if I'll have to build a custom tank and cannibalize
the "coolant level sensor" from the Ford tank into my custom tank.
........-- The Ford tank is translucent plastic - maybe can "fire sleeve" it
or otherwise physically "wrap" it (in an aluminum box?) so as to protect
from a flash fire (momentary) in engine compartment - same level of
protection that fire sleeve will give to my rubber coolant hoses.
David
|
|