Water Flow and Cooling

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

From Roger Gluth

In the late 1970’s, I was involved with the first solar heating systems to be installed on
military family housing. Working as a supervisor for a heating and A/C
contractor at Ft. Stewart, Ga., I was privileged to spend a few days
with Carrier Air Conditioning field engineers as they set up all the
instrumentation required by the Army Corps of Engineers.

The systems we installed were flat collection panels with 100% drain
down. The test units were fitted with an “ISTA meter” on both the supply
and return side of the storage tank. The meter was of European
manufacture and calculated heat transfer in BTU’s based on temperature
change and flow rates.

Also installed in the test systems were adjustable flow meters along with a truck load of thermisters (the
floors of the test units were literally carpeted with cables). (Another
interesting instrument, of which I can not recall the name, measured the intensity of the sun.
This was all necessary to satisfy the Corps of Engineers that the systems
would “harvest” heat as called for in the specifications.

I was surprised to discover that maximum heat harvest was not realized in
direct proportion to water flow. In fact, we needed to slow the flow rate
in some cases to achieve the best results. Sorry I can’t provide any
empirical data for you, but it must be out there somewhere, and from my
experience flow rates do make a difference.

From Glenn Loftus

The subject of the effect of water flow rate on heat transfer in a radiator
can be complex. Good texts are available from McGraw-Hill, but I don’t
think you want to get into that much detail.

Note that a radiator’s
effectiveness is usually MUCH more limited by heat transfer resistance on
the air side than on the water side or in the metal tubes. It’s important
to have good air flow through the radiator and keep it relatively clean.
See here.

In reality, you don’t need to calculate the detailed transfer coefficients
to answer the question of whether more flow is better.
Increasing flow will increase cooling, but there is a law of diminishing

Keep the flow rate high enough that the flow is turbulent rather than
laminar. Laminar flow is layered and does a terrible job for heat transfer.
Set your flow rate based on the amount of heat that you need to transfer and
the temperature rise that you want in the cooling liquid. For cooling
towers, we try to design for 10°F rise, but sometimes you have to accept a
bit more to keep a heat exchanger small.

I haven’t worked with many
air-cooled applications, but you’ll probably have to accept at least double
that amount because air cooled radiators aren’t as efficient as a cooling

Look at your heat source. If you’re putting 100 watts of heat into 1 gallon
per hour of water, the water will increase in temperature by 41°F as it goes
through your cooling block. If you double the flow rate, the temperature
rise is half as much. If you double the heat (say for some really BIG
peltiers on a massively overclocked CPU) then the temperature rise doubles.

Note that we’re talking about the temperature RISE. This temperature rise
is ADDED to entering temperature of the cooling liquid going into the block
on your heat source.

If your radiator is not able to get rid of heat as
fast as it is added by your source, the circulating liquid will gradually
get hotter and hotter.

Eventually, it will reach an equilibrium, because
the hotter the water is, the slower that heat can be transferred from the
hot source to the circulating water in the water block.

Also, the hotter the water is going into your radiator, the faster that heat can be
transferred from the hot water through the radiator to the environment.

Incidentally, if you’re shopping for radiators, they may be rated in BTUs
per hour. One Watt = 3.41 BTU/hr.

Update- 7/17/00: Some tests and charts to illustrate the above here.


Leave a Reply