tigers
[Top] [All Lists]

Re: Hot Tigers

To: Steve Laifman <laifman@flash.net>
Subject: Re: Hot Tigers
From: rpalmer@ames.ucsd.edu (Bob Palmer)
Date: Thu, 4 Sep 97 17:29:36 PDT
Steve, et al.

Are you just trying to stir up trouble, or is this really just an innocent 
question???  :>)  :>).  Throwing in some happy faces so at least some of you 
will know I'm just kidding.  The argument you pose is the one most often 
offered to explain the reason the restriction works.  However, I believe 
this is actually a specious argument.  Now before you get bent out of shape 
over my use of the word "specious", let me quote from my Webster's Noew 
World Dictionary:  "seeming to be, sound, correct, logical, etc. without 
really being so".  I do not intended by the use of this term to defame or 
otherwise impugn you or members of your family.  With that disclaimer, let 
me explain why I think the "dwell time" argument is faulty and, in fact, 
more flow is better.  One of the things we know for sure is that the rate of 
heat transfer is proportional to temperature difference.  This applies both 
on the engine side and at the radiator.  Heat is carried away from the 
radiator by forced convection.  Keeping all other conditions the same, the 
rate at which heat is transfered from the radiator to the air is 
proportional to the temperature difference between radiator and air.  This 
seems pretty obvious.  I notice, for example, that my water temperature runs 
about 90 degrees F over the local air temperature.  This will vary somewhat 
with humidity and elevation (air density), and whether you have a tail wind 
or head wind, etc.; but I'm sure you get the idea.  What I'm leading up to 
here is that the radiator works best when its hottest, and hot all the way 
across from inlet to outlet.  To the extent that the temperature decreases 
as it flows through the radiator, you lose efficiency.  Now I know what you 
are probably thinking; if the water doesn't cool off while it goes through 
the radiator, how can it cool the engine?  This is where the "specious" 
thinking comes in.  Let's break the problem up into four components; engine, 
coolant, radiator, and air.  The engine produces heat.  The coolant 
transfers the heat to the radiator.  The radiator transfers the heat to the 
air.  Seems simple enough.  Now consider what happens as we  vary the 
coolant flow rate from zero to infinity.  At very slow flow, the coolant 
temperature in the engine is the engine temperature and the coolant 
temperature in the radiator is at air temperature.  In this example, 
essentially no heat is removed from the engine.  Obviously, this case 
doesn't work.  As we increase coolant flow rate, its temperature difference 
decreases continuously, appoaching a uniform temperature at a sufficiently 
high flow rate.  Under this circumstance the coolant is at an intermediate 
temperature between the engine and the radiator and has ceased to be a 
limitation on the transfer of heat between the engine and the radiator.  
Whether the coolant temperature is exactly half-way between the engine 
temperature and radiator temperature would depend on the relative heat 
transfer efficiencies at the engine and at the radiator.  At least the 
radiator, I think we can assume almost no temperature difference between the 
coolant and the radiator.  There may be a more significant temperture 
difference inside the engine; especially in dead-end sections, etc.  But, 
I'm probably getting into a little too much detail here.  In general, with 
sufficient flow, the coolant will come to the optimum temperature for 
transfering heat between the engine and radiator and other factors will then 
become limitations on the removal of heat from the engine; e.g., the flow of 
air past the radiator.

Based on this simple analysis alone, the use of restrictors in the cooling 
loop would seem counterproductive.  However, too many varifiable examples 
exist that show a positive effect of restrictors.  Does this mean that the 
above reasoning is also specious?  I don't think so, otherwise I would not 
have offered it.  What I do believe is the above reasoning is basically 
valid and that a reduction in flow rate per se is never benficial.  However, 
there must be one or more factors other than flow that benefit from the 
restriction.  I have offered some ideas as to what these may be, but these 
are  just speculations on my part.  It's at least fun to try and divine 
what's going on.

A friend of mine passed on anothe anecdotal story about a particulary 
variety of Ferrari V12 heads that had a cooling problem.  The heads were 
sectioned and flaws in the casting were revealed that left lots of pockets 
where water could stagnate.  The heads were redesigned and the problem went 
away.  I don't really see how this story relates to our particular problem 
though.  The problem with the Ferrari heads would lead to hot spots, not 
overall cooling problems as with Tigers and I guess Panteras, etc.  One of 
the possible ways that the use of a restrictor helps is by building more 
pressure inside the engine, especially the heads, to reduce local boiling.  
This isn't an original idea with me, but seems plausible.  Another thought, 
which I think I originated, is that by adding a restriction before the 
pressure relief cap, you increase the overall pressure everywhere in the 
system.  This could have a really beneficial effect on the inlet side of the 
water pump.  Cavitation, and consequently loss of flow, would worsen as the 
coolant temperature rises and also as the pressure at the inlet of the pump 
decreases.  The article I mentioned in my previous e-mail talks about 
pressure differentials of up to 40 psi produced by the water pump.  If we 
take this as being true, then let's consider what happens if we take our 
simplified cooling system and add a restriction.  Assume, for the moment 
that neither the engine or radiator are significant restrictions.  Suppose 
we start throttling down at the point where the thermostat usually is.  
Initially, the pump is flow limited and is producing very little pressure 
differential.  Now as we increase the restriction, the pressure difference 
increases.  Over some range of restriction, there may be very little change 
in flow because the pressure difference at the pump increases to compensate. 
 Let's assume that this is basically true up to, say 40 psi.  Assume also, 
that we have a 15 psig pressure cap at the usual location; i.e., after the 
restriction.  In this example, the pressure in the cooling system following 
the restriction, through the radiator, and incuding the inlet side of the 
pump is all at 15 psig.  At the pump we have a 40 psi pressure difference, 
so inside the engine the pressure is 15 + 40 = 55 psig.  Contrast this case 
with the case where all the pressure drop is across the radiator.  In this 
case we have 15 psig at the inlet to the radiator (and everywhere else in 
between the pump and the radiator, including the engine) and a 40 psi drop 
across the radiator so the pressure between the outlet of the radiator and 
the inlet to the pump is 15 -40 = -25 psig.  I can definitely see the 
possibility of problems with this situation.  Any given "real world" 
situation probably lies somewhere between these extremes.

Does the above model really give the right explanation for the beneficial 
effect of a restriction?  Maybe so, maybe not.  It does seem plausible, but 
I would be happy just to be able to have elevated the discussion a little 
bit above what has preceeded which has seemed to me to be logically 
inconsistent.  (I'm trying to be as gentle as possible here.)  If we were to 
fully instrument a car with pressure and temperature sensors, we could 
probably really nail this topic down.  I think as a practical matter, must 
of us find it easier to do the things that make sense and/or copy what 
others have done that seems to work.  However, using anecdotal information 
without any theoretical foundation can be dangerous.  Seldom does one do a 
controlled experiment with changing just one variable at a time and it's 
easy to make erroneous assumptions when this is the case.  If you have at 
least the basic ideas right you can avoid at least some of the pitfalls.  
One example in my case comes to mind; I wanted a new water pump and was 
thinking about getting the Motorsport impeller and having my old water pump 
rebuilt with this impeller.  So I called up the local rebuild shop and asked 
about a price.  The guy in the shop immediately wondered why is was using a 
high performance impeller.  I explained my application and he started off on 
the usual crap about flowing the water too fast through the radiator and 
what he usually does for high rpm motors is to cut every other vane out of 
the cheap stamped metal impeller.  I listened as politely as I could, and 
then called Sunbeam Specialties and ordered their HiPo pump.  Seems to work 
fine.  Can't say if it is the absolute best, but I have no indication of a 
problem from idle up to the 7-8000 rpm range.  Somehow it strikes me as a 
little ridiculous to take the advice of someone at the local water pump 
rebuild shop versus the Ford engineers who are getting the big bucks to 
design parts like impellers that work.

Well, I've gone on at much more length than I intended.  Now it's someone 
elses turn to expound for awhile.  Hope this has been at least a little 
amusing if not enlightening.  By the way, I will get you the information 
about my electric fan.  I also like the idea of two fans side-by-side which 
has been submitted previously.

Bob

>Jim,
>
>RE: Laminar vs turbulent flow, I am not sure that is accurate (see Bob's
>memo). However, I did use a reducing washer on the custom rear
>engine-front radiator car I built, and it did improve cooling. Maybe the
>decrease flow rate increases the "dwell" time of the hot water in the
>heat transfer section of the system (radiator) and allows more heat to
>be removed per pass. This would decrease exit temperature. Think about
>it, there are a lot of washers in use in modified cars, and are
>available in speed shops. They all don't do it if it doesn't work. I do
>believe a valid thermal case can be made for dwell time in a heat
>exchanger. What do you think, Bob?, Jim?
>
>Steve
>-- 
>Steve Laifman         < One first kiss,       >
>B9472289              < one first love, and   >
>                      < one first win, is all >
>                      < you get in this life. >
>                    
>
>_/_/_/_/_/_/_/
>     _/                 _/_/_/       _/_/_/       _/
>    _/        _/      _/     _/     _/    _/     _/_/_/_/
>   _/        _/       _/    _/      _/  _/      _/
>  _/_/_/_/_/__/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/
>                         _/
>                    _/_/_/
>
>


<Prev in Thread] Current Thread [Next in Thread>