Radiator cap

[Stephen M. Brown]

From: "Stephen M. Brown" <hurricanesteve at geocities.com>
To: <bigbroncos at unix.off-road.com>
Subject: Re: Radiator cap
Date: Thu, 16 Oct 1997 10:25:26

> From: Tom Cloud <cloud at peaches.ph.utexas.edu>
> Subject: Re: Radiator Cap
> >Too fast and you don't take ANY heat from the metal and you over heat. 
You
> >are correct, delta-T is required for cooling, but there is a finite time
> >period the liquid must be in contact with the metal for optimum
> >thermodynamic transfer of heat to take place.
> sorry, this is over-simplification -- and just not true !!  If you flow
> coolant across a surface -- and the coolant stays in contact with the
> surface -- the amount of heat removed is a function of delta-T  PERIOD !!

Hmmm....amount of heat removed = heat load = Q = U*A*(Log Mean Temperature
Difference)
Therefore amount of heat removed is a function of the temperature
difference  (log mean, of course, calculated using both inlet and outlet
temperatures) & AREA available for heat transfer & overall heat transfer
coefficient (which is a function of velocity, among other things, which
should have some optimum point -- too slow and you're laminar = poor heat
xfer, too fast = big pressure drop, requires more energy in pumps).

> Therefore, the faster the coolant flow (remember, it has to stay in
contact)
> the higher the delta-T .... ergo, the more heat removed.  Where does this
> stuff about it having to stick around for a certain period of time come
> from?  The Wizard of Oz???  Certainly, the faster the coolant flows past,
> the less heat each unit of coolant absorbs -- i.e. the coolant won't get
as
> hot -- but that's the point, isn't it??  ... to keep the "coolant" cool!

Ok, simple example.  Let's fix Q (heat load) and A (area).  Do this by
fixing 23725 scfm of air at 80F flowing across 50 ft^2 of exchanger area
(non-specific shape, mass & energy balance only).  Q=278250 Btu/hr. Exit
air temperature fixed at 90F

Case 1:
16 gpm of 50%EthyleneGlycol and 50%water is cooled from 160F to 118F.  We
calculate an LMTD of ~58F.  We calculate a U (overall heat xfer coeff) of
96 Btu/(hr*ft^2*F). 

Case 2:
32 gpm of same mixture is cooled from 160F to 139F. (exit temperature is
cooler).  LMTD is ~64 (higher).  U is 86 Btu/(hr*ft^2*F) (lower).  

What am I saying?  Well, I am saying that no matter what mental model you
build, Q must equal U*A*LMTD.  If Q and A are fixed (as in the example), if
LMTD goes up, U must go down and vice versa.

> Now, in a closed system, where the coolant T0 (temp at entry to the
> heat source) is not constant but rather a function of a secondary heat
> exchanger (the radiator), another factor enters -- the variation in T0
> (the beginning coolant temp as it enters the block).

This system is difficult because (1) you have two exchange areas, the block
and the radiator (not to mention ambient losses), and (2) because you don't
reach equilibrium in the radiator if the thermostat is switching on and off
while you drive.  

> 
> I think I buy the cavitation thing  ....  it allows me to keep my
> hard-headed reasoning  ;-)  .... the disclaimer I've been making all
> along is that the coolant ** HAS TO STAY IN CONTACT WITH THE HEAT
> SOURCE **.  So, with the demon, "cavitation", I can smugly allow
> those less informed than me am think that faster is worser -- cause
> it just could be.  Now, faster is not the same as "more" -- and we
> can all agree that "more" is better, can't we?  And by this, I mean
> more volume.  This may not be possible without extensive redesign
> of the cooling system -- if you want to keep the velocity the same.

With respect to cavitation, not only is it bad for the pump (it can erode
pump surfaces quickly), IF the fluid stays a vapor, you will get HORRIBLE
heat xfer.   Why do you think you can stick your hand into a 500F oven for
a few seconds, but cannot stick your hand into 212F water?  There will be a
limit to how fast you can pump stuff into the block, not only because of
cavitation, but also because of  pressure constraints in the system itself.

> Tom Cloud
> From: "John Carroll" <jac at wavecom.net> (by way of Tom Cloud
<cloud at hagar.ph.utexas.edu>)
> Subject: Re: Radiator Cap
> 
> Disregarding the undesireable effects of pump cavitation in high 
> temperature coolant and uneven coolant flow through the block:
> 
> Does the the theory of constraining the velocity of water 
> through the the radiator and block in order to assure that there is 
> time to acquire heat, also call for the velocity of 
> air through the radiator to be controlled to assure that the 
> air has time to pick up heat from the radiator? :)  

i was under the impression that a large volume of coolant did stay in the
radiator under operating conditions.  Air to water exchangers have terrible
heat xfer.   I was not under the impression that the coolant spent much
time in the block.  

Steve

=====
http://www.off-road.com/4x4web/ford/bigbroncos
http://www.bronco.com