alventures wrote:

Giving a little more thought to the cooling system philosophy, I see the
following options; which I'll post here for further review.

*Coolant system with pressure cap, overflow to a vented container.  It
comes to full cap pressure, but a slight reduction in coolant temp and
the pressure drops quickly.  It will go very slightly below ambient as
coolant is drawn back in.  Yes, as you power on again you get full cap
pressure again, but it results in pressure cycles between 0 and cap
pressure.  This is the common approach used on most vehicles, and the
pressure variations may not be an issue.

*Pressurized reservoir (expansion tank) as part of the system.  The air
in the tank is compressed as the coolant expands.  It only reaches cap
pressure when the system is as hot as it has previously been (assuming
no refill in between).  But it always has some pressure, and the
pressure fluctuation with load changes is significantly reduced.  This
is the approach used by Mercedes and Volvo; and some others.

*Use the same basic approach as the first one above, but use a
pressurized overflow container.  Use 20+ psig cap on the coolant system;
15 psig on the overflow (expansion) bottle.  This is the system used the
Jag XJS.  This approach maintains pressure on the system and reduces
pressure fluctuations.  Seems like the best approach to me; just a
little more complicated.  I'm told a number of late model cars use this
idea - something about coolant longevity.

 I had planned the second option above, but am reconsidering.  Seems
good for a car where you don't make large altitude changes; but fly your
plane hard at 15,000' and then drop to 5000' and run at high power and
you won't get full pressure.  So the third approach may be best.

I've tried to find a plastic expansion bottle used by M-B and Volvo, but
the configurations don't fit well for my application; so I may use my
fire extinguisher bottle for the pressurized overflow.  Would be nice to
have translucent tank, or a site level gauge.

In whatever approach, I will have the system pressure cap connected to
the pump inlet line.

Al

Another issue Al is where the expansion tanks goes in the system.
My thought is to place them on the inlet side of the pump after
the rad(s). That way maximum advantage is taken of pressure in the
block to increase boiling temperature with no HP increase for the pump
or a reduction in overall flow rate as the use of an artificial restrictor
in
the system would suggest. The rad and the block pressure drops then will
be the only restrictions in the system to maximize water flow and
minimize pump HP consumption.

The expansion tank can take the form of a long 2 or 2-1/2 inch tube
mounted on the firewall.

Paul Lamar

The AirCraft Rotary Engine NewsLetter.  Powered by Linux.
ACRE NL web site. http://home.earthlink.net/~rotaryeng/
Copyright 1998-2003 All world wide rights reserved.

FWIW...
Check out the late model Chevrolet Silverado (5.3L) truck cooling system
design.  The Silverado comes with the 5.3 engine which has the aluminum heads
and I imagine that GM felt it was very important to keep the air out of the
system.  (The 5.3 cooling system may be different from the others) It only has
one cap which is on the see-thru plastic purge tank.  It has a crossflow
radiator like most cars have these days, but GM has added a small line,
approx. 1/2", which comes off the top of the radiator "return" tank, and
connects to purge tank, about 1/3 down from the top.  There is another line,
approx. 3/4", coming off the bottom of the purge tank which connects to the
water pump inlet.  The two large hoses remain (wp outlet -> rad inlet, rad
outlet -> wp inlet).  If you have room you can use the "see-through" GM tank,
or get a smaller aluminum tank from a racing equipment supplier, or make your
own.

The benefit of this system is that most of the flow still goes through the
larger hoses while the purge tank goes about extracting the air from the
system.

My 2000 Chevy C-1500, 5.3L with 93,000 miles has never lost a drop of
coolant.  From that I would say that it seems to work pretty well.

Mark S.

 
The AirCraft Rotary Engine NewsLetter.  Powered by Linux.
ACRE NL web site. http://home.earthlink.net/~rotaryeng/            
Copyright 1998-2003 All world wide rights reserved.