Index Home About Blog
From: John De Armond
X-Source: The Hotrod Mailing list
Date: Feb 1992
Subject: Re: Blown '31 "ride"

> >>>Secondly, I had a 4 core radiator with dense fins made for my Z-car after
> >>>putting a turbo engine in it.  I had worse overheating problems than
> >>>before.  After extensive testing (dT across the radiator, delta air 
> >>>pressure across the fins, etc), I figured out that the dense fins (24 per
>I am not familiar with this type of testing.  dT and delta air??  I
>can guess what the basic idea of the tests are about, but can someone
>give me some specifics?  What type of shop can perform such tests?

I don't know of any commercial shop that can do this testing but it only
requires a little bit of equipment, namely a thermometer and a differential
pressure indicator.  I recommend buying a digital pyrometer from
someone like Omega Engineering for as little as $80.  It's better to spend
a little extra and get a dual input unit but not absolutely necessary.

To do the delta temperature (dT) test, simply insert a thermometer into 
the inlet and outlet of the radiator.  If using a digital pyrometer,
the thermocouple can be slid between the hose and the spigot and the 
pipe clamp retightened to seal it in.  Then just drive the car
under whatever conditions are giving you trouble and observe the 
readings.  My dual input pyrometer will display the difference between
the readings directly.

What you are interested in is the temperature loss across the radiator
and the actual temperature of the outlet.  The radiator should drop
at least 20 degrees and preferably more.  More importantly, the 
outlet temperature of the water should be below the setpoint of your
thermostat.  If you're running a 180 degree thermostat, an outlet
temperature of 165-170 should be observed even in the hottest weather.

If the dT is low and/or the outlet temperature is high, that indicates
insufficient heat transfer.  That could be either low airflow or blocked
tubes.  If the dT is OK, the outlet is lower than the thermostat setpoint
but the INLET temperature is significantly higher than the thermostat
setpoint, that indicates that there is insufficient coolant flow.

All the above, of course, assumes that the engine is in sound condition and
does not have a headgasket leak or something similiarly major.

The delta pressure (dP) measurement is made by placing two static pressure
taps - basically two tubes with their openings perpendicular to the 
flow (you pilots out there know all about static vs pitot tubes) - on
the radiator, one in front of it and another between the fan and radiator.
Care must be taken to make sure the tubes actually are static taps and
do not have any dynamic pressure on them.  This is best assured by 
making the tubes perfectly perpendicular to the air flow.

The pressure taps are connected to a differential pressure indicator
calibrated in inches of water.  I have a fancy digital manometer but
a simple incline manometer will do.  You can buy a small manometer with
a range of up to, say, 10 inches of water for under $50 at an HVAC
supply store.  Look in the phone book.  Or you can build one from 
a bit of tygon tubing and a piece of plywood.  If there is actual
interest, I'll post a construction article.  It'll be wordy and maybe
will have a postscript illustration so I'm going to save it for now.

The test is simply to drive the car under normal conditions and observe
the differential pressure across the radiator.  It should only be a 
couple of inches of water, 5 or 6 max.  The fan loses efficiency 
rapidly at that level of dP.  Common causes are debris blocking
the radiator, debris buildup between the fins or too many fins per inch.
Note that this pressure should be positive across the radiator.  That
is, the pressure outside the car should be higher than that inside the 
car.  It is not that unusual to see a car with a high nose actually
build enough pressure under the hood to negate the ram effect of 
cooling air entering the front grill.  

For this test to be meaningful, you must first insure that your
fan clutch is in good working order.  Particularly important if you
have a viscous clutch.  I jumped through hoops awhile back trying
to discover an overheating problem that indicated dP problems 
when the real problem was the clutch had leaked enough oil to not
be able to fully engage at speed.  The fan would turn at idle and 
you could hear it spin up when you raced the engine but it would
not spin fast enough to cool the car.  I finally replaced the clutch
on a hunch and the  problem went away.  I opened the clutch and 
found almost no oil inside.  90wt gear oil is good for refilling.


From: John De Armond
X-Source: The Hotrod Mailing list
Date: Feb 1992
Subject: Radiator testing

>How about too high of a flow-rate John?  Where the coolant does not
>remain in the radiator long enough to get rid of it's heat?  I think
>this is what I'm experiencing with my overheating problem.

<stamp> <stamp> <stamp> <stamp> 

There!  That rumor is finally dead!  I know the hotrod magazines like to 
parrot it but it just ain't true.  Think of it in terms of fluid dynamics
and thermodynamics.  What does higher flow get you?  turbulence.  What
does turbulence do?  It strips away the boundary layer of stagnant water
on the surface of the radiator tube.  As to not having time to get rid
of the heat, do a simple thought experiment.  if the water is moving 
twice as fast as before, it only has half the time as before to deposit
its heat.  BUT.  It will also come back around twice as often.  It 
all balances out.  The net win is the increased turbulence.

One reason this rumor persists is that the 2 most popular purported cures 
actually work sometimes.  The first "cure" is to slow the water pump with
a larger pulley.  The problem that is actually solved is the fact that
many stock pumps will cavitate at even moderate RPM.  A cavitating pump
is very inefficient.  The second "cure", restricting the outlet of the
engine, pressureizes the block a bit more so that at higher engine
speeds, the boiling point of the water is raised.  That works to preventing
local boiling on stagnant areas.  Steam bubbles can slow or stop
flow through small passages.  The increased pressure also cuts down on

"Stock Car Racing" did an exhaustive article on the subject.  The guy 
who owns Stewart Racing pumps actually built a water pump dyno and
demonstrated all of these effects.  There is also a significant
body of literature within the SAE addressing this problem.

>>Care must be taken to make sure the tubes actually are static taps and
>>do not have any dynamic pressure on them.  This is best assured by 
>>making the tubes perfectly perpendicular to the air flow.
>Could be tough with a angled radiator... :^(

No.  You don't want the taps facing into the flow, you want them to extend
in from the side or top, in other words, perpendicular to the flow.
Any angle off perpendicular to the flow will introduce a component of 
dynamic pressure.

>Had the same problem on the clutch-fan on my Corvette..  The unit
>looks like it is sealed to me though John, how did you get yours open?

The one on my Datsun is held together with 4 6mm bolts.  Just unscrew
'em and pry the clutch open.

>Great article John -- I love this kinda' talk!

Thanks.  Been waiting for this forum for a LOOOONG time.


From: John De Armond
X-Source: The Hotrod Mailing list
Date: Feb 1992
Subject: Re: Radiator testing

>Just one point here on Johns posting.  I believe the reason I have a little
>trouble accepting the premise that too high of a flow through the radiator
>would not cause overheating is this:
>If I shut off the flow of hot water *to* the radiator, that water trapped
>in the radiator should cool (if allowed, down to ambient.)  Granted,
>the water trapped in the engine is now getting hotter.  If we relase a
>'spurt' of cooler water into the engine, displacing an equal volume of
>hot water, then close the thermostat, this cooler water now begins to
>absorb heat energy while the 'spurt' that went to the radiator is cooling.
>Now, which one is more efficient?  Remember, the coolant in the block
>recirculates until it has absorbed enough heat to open the thermostat --
>the radiator does not have this luxury of being able to hold coolant until
>it's at a low enough temperature to send back to the engine...

>I very much value Johns opinion, and  not wanting to start a hotly disputed ;^)
>radiator war where everyone gets steamed ;^) I'm just going to cool down ;^)
>(I can't believe I punned three times in one sentence!) -- I'll just report
>back on what I find and we can take up the discussion from there.

hey, we're a long ways away from the name calling stage. :-)

Let me walk you through a thought experiment.   Consider the instance
where the flowrate is just such that the outlet temperature of the 
radiator is near ambient.  Hot water going in, cold water going out.
maximum efficiency, right?  Wrong.  Assuming that the water loses its
heat linearly, the top part of the radiatior where the hot water is
dumps most all of the heat while the lower part where the water is at
ambient dumps almost none.  The part in between works proportionally.

Similarly, in the engine, where the cold water hits the block, maximum
heat is absorbed.  Where the now hot water exits, little heat is absorbed.
Cooling is non-uniform and worse, areas in the head that have the highest
heat load (such as around the exhaust ports) may suffer localized boiling.
Once a film of steam forms, almost all cooling is lost.

Now let's change things around a bit.  The flow is now so high that 
the radiator inlet and outlet temperatures are practically the same.
Consider what is happening now.  All areas of the radiator are at the
same temperature.  That means all areas of the radiator are contributing
equally to dumping  heat.  Maximum efficiency. 

Similarly, the water temperature in the block, by definition, is uniform.
The high water velocity scrubs the heat generation surfaces of steam and
air bubbles  and stagnant boundary layers.  Heat transfer is again optimized.

Just to give an extreme example, consider this.  I'm a nuke by training
and have spent much of my career making nuclear plants work a bit better.
So I'll use one as an example.  In a pressurized water reactor system,
heat is transfered from the fuel to steam via an intermediate, sealed
water loop called the primary coolant system.  There is a heat exchanger that 
is called the steam generator that takes hot reactor water in on one
side and boils lower pressure water on the other.  This scheme isolates
the radioactive primary coolant from the clean secondary steam system.

This system transfers 24,000,000,000 watts of heat continuously with 
less than a 20 degree F drop in coolant temperature!  Of course, the 
flow is in the millions of gallons per minute.  The fuel pins produce
heat at a rate of about 20 kilowatts per linear inch.  The extremely high 
flow is designed to mitigate all the same problems we've been discussing.
A very serious condition in a nuclear plant (and in a car engine) is called
DNB or Departure from Nucleatic Boiling.  Nucleatic boiling is similar
to what you see in a pot where steam bubbles seem to form at particular
points or nuculi.  This is very efficient in terms of heat transfer because
most of the surface remains wetted.  DNB means that boiling has expanded
to general film boiling (easy to see if you plunge a red hot piece of 
metal into water.  The shiny surface under the water for a few seconds is
film boiling.)  Film boiling means there is a layer of relatively poor
thermally conductive steam between the metal and the water.  In a 
nuclear plant, a DNB incident can result in melted fuel.  In a car
engine, it car result in local and severe oveheating.

Phil, you might be running into a problem I chased on my BMW for 
quite some time.  I'd get quick spikes of high temperature under certain
conditions.  The BMW temperature gauge is very fast.  What was happening
was because of a weak radiator cap that would not hold pressure, steam
bubbles were forming and collecting at the highest point in the engine
which happened to be where the thermometer sender was mounted.   I'd see
the temperature spike almost instantly to full scale and sit there until
I revved the engine a bit.   That would sweep away the steam and I'd
get a regular reading again.  If your temperature gauge is slow to respond,
it might be averaging such spikes to look like a uniformly high 
temperature.  The clue that I ignored for awhile was the fact that when
I opened the radiator to add water, there was practically no pressure.

>Having written the above and reading it over, I thought I'd give the people
>at "Custom Auto Radiator" (a company that designs and builds custom made
>radiators for antiques, hotrods and restored automobiles) a call and ask to
>speak to their engineer.  I told him about the problem I'm seeing, and
>expecting him to try to sell me a new radiator ;^) his first response was
>that the flow through the radiator is too high!  He made several suggestions
>as to the fixes that could be done.

Hey, what can I say?  All kinds of superstitions aboud.


From: (Eric W Youngblood)
X-Source: The Hotrod Mailing list
Date: Jul 1993
Subject: RE: street comp. ratios
X-Sequence: 5691

In article <>, (The Hotrod List) writes:
|>  (Eric Y's engine runs >200F w/ 180F t'stat)
|>  Sounds to me like more cooling is needed. If the cooling system won't hold
|> its setpoint and you've got a good t'stat (maybe a high flow one is worth
|> trying) then there's a problem with dumping the heat. More rows or more
|> airflow, probably the latter since the A/C seems involved. Maybe a pusher
|> fan tied to the A/C compressor clutch line would help you out.
|> ----------
|> Posted by: emory!!jws (James W. Swonger)

This puppy has always run warm in the summer. Im sure the siamesed bores
are partly responsible. As far as heat dumping goes its set up with a
standard waterpump, standard size pulley, viscous clutch fan with shroud.
All grill openings are unobstructed and the OEM wind dam is in place.
(2nd generation Trans Am)

I put a new radiator (4 core) in two years ago when the engine was rebuilt
17k miles ago. I also installed a trans cooler and a electric pusher fan to
help keep good air flow in traffic.

[You didn't let 'em talk you into buying an high fin count radiator, did you?
I let a radiator shop sucker me into that one with my Z.  Went from (foggy
memory alert) 16 fins per inch to 24 fins per inch at the same time
I added another row of tubes.  Car wouldn't cool worth a damn.  I did some
measurements with a differential pressure gauge and discovered that the
radiator had a hell of an air pressure drop across it, a LOT more than
stock.  I had another one made with the right number of fins but with the
extra row of tubes and it worked fine.  DP was pretty close to stock.  JGD]

I can sit idling at stoplights all day in 100 F w/ AC on full blast with no
problem.  But, if I get cruising at about 3000 rpm for an extended period it
will peek over the 200F mark.  Simply dropping the rpms to about 2500 will
allow the temp to drop below 200F again. Of course, w/o the AC in the picture
it'll run 180-190F range.

The pre-ignition starts about >190 when the ignition is tuned for max
performance.  So as the weather gets hotter I detune the ignition.

[Does the temperature go down when you retard the timing?  If so, that
probably indicates too much vacuum advance.  The engine is probably suffering
inaudible detonation and that is enough to overheat it.

If it's not timing, it may be pump cavitation.  You might try installing
a restrictor in the upper radiator hose.  This increases the head pressure
on the pump and will tend to supress cavitation.  JGD]

It only loses a couple tenths in the 1/4 mile, and low end throttle is a
bit off.

I was just looking for a way to avoid seasonal tuning.


From: John De Armond
Subject: Re: Is my 1 ton Dodge van/RV overheating?
Date: Tue, 21 Dec 1999 18:31:19 EST

Dan Simmons wrote:

> I went to my mechanic today and he thinks the radiator has build ups of
> deposits so that the flow is restricted. He suggested changing the
> radiator and to go with the heavy duty radiator. He though the price of
> a new radiator would be something like $230 dollars. I checked the price
> in a store though and it was $318 for the part.

Dan, I think someone is tugging at your wallet. Three bills for a
small truck radiator is pretty steep.  Shouldn't cost over half that
for a re-coreing job and perhaps $250 for a new one if you're having
to pay list.

Unless $300 comes to you a lot easier than it does me and you want
to attempt to solve the problem by tossing parts at it, I suggest
you do some testing to narrow down the problem.  What you need to do
is measure the radiator inlet and outlet temperatures.  This is easy
to do with a pair of thermocouples and a suitable readout.  You can
get a thermocouple adapter for a digital voltmeter for about $30. 
Even Home Depot sells 'em.  

To measure the temperature, loosen the hose clamp to the radiator
and pull the hose loose.  Take the thermocouple wire and wrap a turn
of double-sided mounting tape around it about 2" up from the end. 
Stick this in the hose and slip the hose back on the radiator nipple
so that the tape will lay between the hose and nipple under the
clamp.  Tighten the clamp down.  The tape will flow with the heat
and pressure and make a permanent, watertite seal.

Run the vehicle until the overheating condition you're interested in
is happening.  Record both temperatures.  Actually, have an
assistant do so because you want the temperatures while the vehicle
is in motion.  Thermocouples respond almost instantly to changes so
you can't just slow and then read.  Interpret the readings as

Assume a 180 deg thermostat.

Inlet a little greater than 180, outlet approx 120 or lower - Normal
operating conditions.  The thermostat is throttling the flow to
maintain block temperature and the radiator is disposing of the
heat.  Your temperature indicator is probably not indicating

Inlet much > 180, outlet near ambient.  Suffering from low flow. 
Either the radiator is stopped up or the water pump is
malfunctioning.  The radiator is disposing of the heat that is
arriving at its inlet and so the outlet water is near ambient.  If
you feel the radiator, you'll probably be able to feel a sharp
transition from hot to cold near the inlet tank.  A confirming
indication is if the inlet hose balloons slightly and the outlet
hose sucks in slightly when racing the engine.

Inlet much > 180, outlet somewhat below or greater than 180 - 
Radiator is overloaded or the airflow is insufficient.  The radiator
is receiving more heat than it can dispose of and so the outlet
water is remaining hot.  Make sure the mechanical fan clutch is
fully engaged, that the electric fan (if applicable) is working and
that you've done nothing to disrupt the smooth airflow to the
radiator.  In an effort to improve economy, modern vehicles rely
much more on ram air than did older vehicles and therefore, major
disturbances in the flow path, say, a bicycle carrier, can cause

Inlet temperature much > 180 with momentary spikes even higher -
localized overheating, usually around the exhaust valves and
sparkplug holes, is causing the formation of steam.  This is very
serious because it is regenerative.  Once steam starts forming, it
insulates the surface which results on more steam, etc.  The usual
cause is a pressure cap not holding the specified pressure.  have
the cap checked and replaced if it doesn't meet specs.  If it is
holding the specified pressure, then it is OK to try a cap a couple
of PSI higher.  Some caution has to be used if your radiator has
plastic tanks that are crimp-sealed to the core.  Extra pressure can
blow these seals.

Some other things to be aware of.

Sometime around the time when your vehicle was made, mfrs started
using highly nonlinear sensors in order to make the gauges read
"correctly" in order to reduce the number of nuisance service calls.
On modern Ford light trucks, for example, the oil pressure "sender"
is really a switch.  The gauge either reads "right" or it reads
zero.  At the same time, the definition of "overheating" has been
changed from "gauge moved from where it normally runs" to "gauge is
in the red."  Non-linear senders move the needle only a little until
a knee in the response curve is hit and then the indicator rapidly
runs to full scale.  Point being, the gauge may or may not mean
anything.  A known good temperature indicator such as the
aforementioned thermocouple is the only way to know for sure.

The new inexpensive infrared pyrometer thermometers that have
recently come on the market are invaluable for a number of things on
the vehicle.  My local machine shop is selling the Raytek
fixed-emissivity unit for $79.  It functions by measuring the
infrared radiation emitted by the hot object without having to make
contact.  In this situation, it can be used to quickly map the
temperature profile of the radiator to determine whether some of the
tubes are stopped up.

Awhile back I pulled the radiator on my 68 Plymouth Fury because the
tubes were stopped up.  When I pulled the inlet cap I found that the
tubes were clogged by shards of RTV, gasket particles and so on. 
Obviously introduced when a previous owner had the heads or intake
off and was not careful about removing debris from the cooling
system.  This is increasingly common as the tube size gets smaller. 
This could have been fixed by backflushing had I taken a look inside
first.  I have a borescope that lets me do that but a dental mirror
and a small lamp soldered on the end of some stiff wire will do as
well.  The little mini-maglite flashlight bulbs are perfect for

If the radiator is plugged with hard water and/or rust deposits,
they can easily be removed by boiling out.  The radiator shop can do
that or you can do it yourself.  Prestone used to sell a 2 part dry
flush system.  One part consisted of oxalic acid.  This dissolved
the lime.  The other half contained sodium carbonate  to neutralize
the acid.  With the light alloys now being used on the engine, it is
not a good idea to put this stuff in the system like the old
Prestone stuff was.  But the radiator can be removed, the lower hole
plugged and the thing filled with the solution.  If you can't find
oxalic acid, phosphoric acid works at least as well.  It is
available from farm stores such as Tractor Supply or your local
co-op as "milkstone remover".  The stuff that Tractor Supply sells
is about 45% phosphoric acid and costs about $3 a gallon.  Be aware
that salvaging a radiator with this technique is a 50-50 
proposition.  It is common to find pinhole corrosion holes under the
lime deposits.  Costs almost nothing, though, and the radiator has
to come out anyway so very little to lose in trying.


From: John De Armond
Subject: Re: Is my 1 ton Dodge van/RV overheating?
Date: Wed, 22 Dec 1999 02:02:38 EST

Dan Simmons wrote:
>   Neon John <> wrote:
> <snip>
> > Unless $300 comes to you a lot easier than it does me and you want
> > to attempt to solve the problem by tossing parts at it, I suggest
> > you do some testing to narrow down the problem.  What you need to do
> > is measure the radiator inlet and outlet temperatures.  This is easy
> > to do with a pair of thermocouples and a suitable readout.  You can
> > get a thermocouple adapter for a digital voltmeter for about $30.
> > Even Home Depot sells 'em.
> <snip>
> Sounds like a good idea. I'll see if I can find one of those
> thermocouple adaptors. My idea along this line was to buy a new
> temperature sensor and calibrate it in hot water using an ohm meter and
> a thermometer. (That only works up to 212 degrees though.)

I did something similar in the bad old days before I could afford a
thermocouple meter.  I used Mopar sender units because they are in
1/4" NPT fittings.  The procedure involved punching a hole in the
radiator hose and sealing the sensor through the wall with washers,
RTV and thin (1/2 thickness) nuts.  It sacrifices the hoses and the
response is slow but it is a cheap way out.  Actually one of my old
hotrod engine customers is using this setup permanently and has for
years but I'm not that brave.  I can just see the assembly being
spit out in 100 degree weather.

BTW, saw something really funny at Wally world tonight.  A small
roll of duct tape packaged and labeled as "hose bandage".  Duh! 
Funny part was the price was more than for a full roll of duct tape.

> Thanks for the extensive reply John.

You're welcome.  Gotta try and toss a few crumbs of substance in
here occasionally to balance things out :-)


Index Home About Blog