Bill Freeman wrote:
Paul,
Were you disagreeing with my comments on boiling and a
higher boiling temp coolant? Not sure what to make of your
comment. I *did* look at the piece you sent to me privately
before you posted it. I generally agree with your source
(Heinz), I just now saw that the heat transfer increases by
TWO orders of magnitude as nucleat boiling comes on
line rather than "just" one order of magnitude. I think we are
in general agreement on this, except I contest your comment
on steam creating an "air pocket". :-) Russell agrees with
me, but as he says - a nit, not worth much picking. The only
part of Heinz that I question is the continued existence of the
steam bubbles for large distances away from the hot spot.
I would expect the cooler (sub boiling temp) turbulant gross coolant
would quickly condense the bubbles unless the liquid coolant
was extremely close to the boiling temp itself. Typical auto coolant
temps are 200-220F and under pressure the EG/water
has something like a 260F boiling temp. Steam bubbles should
quickly condense when washed into the turbulent stream of
liquid coolant. Heinz says they stay intact until they reach the
surface in a coolant tank. The conditions that form the bubble
are only controlled by temp and press, not by any free surface
of trapped air that might exist somewhere in the engine. Bubbles
would rise, of course, but would not necessarily seek out some
air pocket, whether in the block or in a coolant expansion tank-
just a high point. And remember I stipulated adequate coolant
flow rates and radiator size plus airflow being adequate.
re: "part of Heinz that I question"
The difference may lie in the temperatures cited. You say that "typical
auto coolant temps are 200-220F". Is that average system temperature,
or
is that a "typical" temperature at a particular point in the system? Is
it
possible that the local coolant temperature near the nucleation site
("near" in this case means within the water jacket region, i.e. a dozen
millimeters or so) has reached the coolant's boiling temperature? Under
those circumstances, the vapor bubbles from the nucleation site would
remain bubbles until the coolant had been cooled slightly, which might
be
the radiator tank or possibly the hose to the radiator.
My problem with Heinz Heisler's treatise is that it doesn't discriminate
between the initial nucleate boiling, where the bubbles depart the
surface
and quickly re-condense ("quickly" in this case being on the order of a
few
millimeters distance), and later nucleate boiling where the bubbles do
not
re-condense until the bubbles (and the surrounding coolant) enter the
radiator.
In the general scheme of things, sure, the later nucleate boiling is
even
better since the heat of condensing isn't dumped into the coolant until
the
coolant is in the radiator. But for *our* purposes, this later nucleate
boiling has a huge potential problem: if the vapor bubbles get hung on
some
structure, they can coalesce and form a "steam pocket". The steam
pocket
then promotes a "pseudo-film" boiling local condition, which can rapidly
result in localized overheating.
And just to make things even more difficult, there are not good names
for
the initial and later phases of nucleate boiling, making conversation
about
it a hassle.
Russell Kent
The heat transfer folks have pretty well studied this subject to death. It
is a big concern in the design of steam boilers and the cooling of nuclear
reactors. But I was a little suprised to see the depth of interest in this
group. At the risk of running off at the fingertips, here's what little I
know of the subject:
"Sub-cooled nucleate boiling" is used to describe the condition where the
bulk coolant temperature (anything away from the laminar boundary layer that
is on the order of a millimeter or two thick) is well below the boiling
(saturation) temperature. This is by far the most effective heat removal
mechanism in a liquid cooled system because the bubbles rapidly carry away
the heat of vaporization and deposit it in the fast(er) flowing bulk fluid.
The bubbles also generate turbulence and break up the laminar boundary layer
to promote faster, better cooling even through the convective layer. By
depositing the heat in the bulk fluid, it is rapidly swept away to whatever
heat sink is available (i.e. the rad). Russell, you read the graphs
correctly: This mode of heat transfer is about 2 orders of magnitude more
effective than just keeping everything single-phase and in the liquid state.
A real mind boggler!
"Saturated nucleate boiling" is used to describe a condition where the bulk
liquid is at or very near the boiling point. Here the danger is exactly
what you pointed out: vapor bubbles do not rapidly collapse, but instead
coalesce and begin to form larger bubbles or voids. This can result in
"chugging" or "slugging" flow, and can eventually lead to drying out of the
heated surface (bad... hot spots). For our purposes it also is bad because
it causes rapid expansion of the overall liquid volume (read "boil over").
Up until the point that the bubbles start to coalesce and/or blanket the
surface of the hot stuff, this is a more efficient way to remove heat than
subcooled nucleate boiling, but only barely so. And of course it isn't a
very stable mode. So this is to be avoided.
There is a continuum of conditions that can cause loss of cooling. The two
ends of the spectrum are referred to as "DNB" (departure from nucleate
boiling) and "dry-out".
DNB is characterized by having extremely high heat transfer rates into a
cool liquid. I can't imagine such a situation in an IC engine though. Here
the heat source is jamming so much heat through so fast that the surface
flashes over into a very thin but stable film of steam and virtually stops
any further heat removal.
Dry-out is more typical of the "hot spot" concern in engines. It occurs
when the liquid is already at the boiling point and even a relatively small
amount of heat input starts making very large steam bubbles (at atmospheric
conditions water will expand to 1,000 times it's original volume when
converted to steam at the same temperature). The bubbles will invariably
stabilize over whatever hot spot already exists and further insulate it,
which makes it hotter, etc.
So: keep the liquid *cool*. It may very well be that small bubbles form
and quickly re-condense. Great! This makes everything work better. But
keep the liquid well below the boiling point (20 degrees F or more?).
Unless there is something really bad in the mixing/combustion process, I
doubt that it is possible to cause enough heat to get into a DNB situation,
so don't sweat that.
Hopefully this answers some of the questions on "what do we call it
when...?"!
Tom Kendall
Now this is why we need those thermocouples down near the combustion chamber guys.
Water temp alone is not enough.
Paul Lamar
The AirCraft Rotary Engine NewsLetter. Powered by Linux.
ACRE NL web site.
http://home.earthlink.net/~rotaryeng/ http://www.linux.org