Searching \ for 'Design of a cooling system using expansion of comp' in subject line. ()
Make payments with PayPal - it's fast, free and secure! Help us get a faster server
FAQ page: www.piclist.com/techref/index.htm?key=design+cooling+system
Search entire site for: 'Design of a cooling system using expansion of comp'.

Truncated match.
PICList Thread
'Design of a cooling system using expansion of comp'
1998\11\11@162850 by Gerry Cox

flavicon
face
I need to cool an object to 4C below ambient. Ambient is in the range 20C to
30C. The object does not generate any heat of its own. Cooling does not have
to be rapid. I have sucessfully used peltier devices controlled by a
PIC16C73 running a closed loop with a temperature sensor and PWM controlled
mosfet drive to the peltier devices. However I have a problem dissipating
the heat from the hot side of the peltier devices in the space available
without increasing the ambient and defeating the object of it all :-(

So I am experimenting with the idea of using a compressed air supply, which
is available near the device, to achieve the cooling. I.e. from first
principles I understand that if I allow the air to expand, then its
temperature will fall. Does anyone know the maths involved and have any
advice on methods to expand the air and cool the object? To achieve
temperature control, I could turn a pneumatic valve on and of at varying
duty cycles monitoring the temperature with my PIC - low rate PWM. I suppose
my real question is how do I design the expander and heat exchanger.  An air
supply is available at up to 6 Bar with quite high flow rate. The
temperature of the compressed air before expansion is close to the ambient.
Ideas anyone?

Thanks in anticipation,
Gerry.

1998\11\11@172959 by Dennis Plunkett

flavicon
face
<Big snip>

A>my real question is how do I design the expander and heat exchanger.  An air
>supply is available at up to 6 Bar with quite high flow rate. The
>temperature of the compressed air before expansion is close to the ambient.
>Ideas anyone?
>
>Thanks in anticipation,
>Gerry.
>
>

6 bar, wow, sounds like a normal 12cfm air compressor! Just let the wind
blow free! As for 4 degrees, such a minimal change I do feel tempted to ask
why. Unless there is quite a large storage tank for the compressed air, I
don't see how it will sit at near abmient, unless the supply line is long
enough to provide a sort of heat exchanger.
>From the points made I can deduce that the item to be cooled is in some form
of enclosure, thus the need for the heat exchanger, rather than an air vent.
Yes what oyu want to do will work, anf if it is cold enough (Good wet bulb
readings) just add water and hey presto, snow! :-)
You will need to make you heat exchanger oil proof, or install a secondary
carbon filter.

Dennis

1998\11\11@180457 by Jason Tuendemann

flavicon
face
The density of air is dependant on temperature. The hotter the air the less
dense or thinner the air is (and the inverse). Therefore the way to make air
expand is to heat it.
So if you are cooling something the exhaust air will be less dense. The
cooling effect is due to more dense air across the device and the device
heating it and expanding it.
So you have to compress the air prior to hitting the device. I.e.. An
aircraft is susceptible to its carburetor icing up from 20 deg C and down
because the temperature of the air at the venturi is 0 deg C because of the
compression.
Possibly a book on carberation and venturi effect might be the go?

I would be interested to hear how you go.

Jason.

{Original Message removed}

1998\11\11@181419 by Mark Willis

flavicon
face
Gerry Cox wrote:
{Quote hidden}

 "It's Goldbergian inventiveness to the ??rescue??"!  <G>

 I'm pretty sure the applicable equation here is that Gas Constant
equation, PV=nRT.  So much for all MY chemistry knowledge!  <G>  Rusty
here, obviously.  Ya get lazy with a "chemistry fanatic" father I guess
<G>

 Pressure * Volume = Moles * a constant * Temperature (in Degrees K)

 (I don't remember what R is, a constant, my Father "liberated" my CRC
handbook again!  Moles should be a constant here though, unless you're
splitting atoms - or doing cold fusion <G>)

 So as you reduce pressure & increase volume, you want to greatly
reduce the pressure, somewhat akin to a 'fridge evaporator.  Volume will
increase, but I'm pretty sure you'll get 4 degree cooling fairly
easily.  Lots of noise potential though (There are honeycomb type
filters that can make a loud (hissing HF noise) air stream, pretty
quiet, though.  A good audio engineering book will also tell you how to
design an air plenum to quiet air conditioning duct fan noise problems &
etc., "The Master Handbook of Audio Engineering" IIRC was pretty
decent.  THAT's been a few years!)

 Could you just blow the heat off the Peltier device with either
fan-blown air, or (less efficient but more effective perhaps) compressed
air through a metering valve (manually adjusted to always use "enough"
air), use a plenum box to quiet the output air stream if necessary, and
if needed change the Peltier devices' power input to get enough cooling?

 Alternate methods:  Heat Pipe to the outside with fan-cooled, wetted
heat sink on the end (Swamp coolers work pretty well.)  Or heat
conducting bar of some other high heat transferring material (A friend &
I joke about making IC's atop a Diamond layer, as it's a good heat
conductor, he says.  If so, it's an expensive heat conductor!)

 A water circulation scheme might be cheap? (Use water to cool the
Peltier devices, bond the Peltier device to a brazed or silver soldered
water plenum, 1/4" copper tube in & out of your device.)  Self-pumping
may be even achieveable with this (certainly works with high temperature
differentials!  Might be too slow here.  Have your output pipe higher at
the water vessel than the input pipe, you may've seen those campfire
water heaters that work with a coil of copper pipe to give you hot water
when camping?)  Jabsco makes good impeller pumps, among other
companies.  OTOH Water can be a mess *when* it leaks <G>

 Mark, spam_OUTmwillisTakeThisOuTspamnwlink.com

1998\11\11@182902 by Bob Blick

face
flavicon
face
Yes, there are commercial systems available. You can buy vortexes in any
number of sizes and ratings.

No need to reinvent the wheel. Do a search on "vortex cooling" and you'll
find a few sources of them.

Noisy little suckers, though.

-Bob

1998\11\11@183939 by Sean Breheny

face picon face
Hi Gerry,

At 09:26 AM 11/11/98 -0000, you wrote:
>I need to cool an object to 4C below ambient. Ambient is in the range 20C to
>30C. The object does not generate any heat of its own. Cooling does not have
>to be rapid. I have sucessfully used peltier devices controlled by a
>PIC16C73 running a closed loop with a temperature sensor and PWM controlled
>mosfet drive to the peltier devices. However I have a problem dissipating
>the heat from the hot side of the peltier devices in the space available
>without increasing the ambient and defeating the object of it all :-(

Did you try using a heatsink and fan to help remove the heat?

{Quote hidden}

The basic idea here I think would be to make the simplifying assumption
that the air is allowed to expand at constant temp(isothermically) (only
4deg out of 300 or so deg K is SMALL), and then calculate the amount of
air(volume) that needs to be allowed to expand thru a 5 bar change in
pressure per time to drain the ammount of power given off by the device.
Then, you could use a software PID (you may not need both I and D) with two
temp sensors to control a valve on the air outlet to keep a 4deg
differntial between device and ambient. The design of the nozzle thru which
the air is released might be tough. I'm not sure within how much distance
the majority of expansion will occur. You will still need a heatsink on the
device and blow the air thru it as it expands. The valve could be just
opened or shut, I don't think you need proportional control of it, as long
as you can open and shut it quickly enough.

Since there will be a feedback loop involved, you won't need exact math,
but you will probably still want a baseline figure. I believe that this is
the approximate math involved:

P1V1=P2V2

P1=initial pressure V1=initial volume
P2=final pressure V2=final volume

6V1=V2 (1 atm is abt 1 bar)

V1=(1/6)*V2

Now, as for the amt of heat drawn off per volume:

deltaW=integral(P*dV,V1,V2)

P(V)=P1*V1*(1/V)

So deltaW=P1*V1*ln(V2/V1)

V2/V1=6 so:

deltaW=P1*V1*ln(6)

Power=P1*(dV/dt)*ln(6)

dV/dt=Power/(P1*ln(6))

P1=about 6*101kPa=600kPa
ln(6)=1.8

dV/dt=(10^-6)*Power

Where dV/dt is the flow of gas in meters^3/sec out of your compressed air
source(volume measured BEFORE expansion)
and Power is the number of watts you need to dissipate.
Now, lets convert to liters/sec:

dV/dt=(10^-3)*Power

So, to maintain a temp. difference for a device dissipating 1 Watt, you
would need about 1 mL per second of compressed gas(or about 6 mL by the
time it expands to ambient press.)

I'm not sure if my numbers are correct, I did this rather quickly.

If it were I, I'd try to use the peltier junction/fan idea ;-)

Good luck,

Sean








>
>Thanks in anticipation,
>Gerry.
>
+-------------------------------+
| Sean Breheny                  |
| Amateur Radio Callsign: KA3YXM|
| Electrical Engineering Student|
+-------------------------------+
Save lives, please look at http://www.all.org
Personal page: http://www.people.cornell.edu/pages/shb7
.....shb7KILLspamspam@spam@cornell.edu  Phone(USA): (607) 253-0315 ICQ #: 3329174

1998\11\11@184353 by Sean Breheny

face picon face
>  Pressure * Volume = Moles * a constant * Temperature (in Degrees K)
>
>  (I don't remember what R is, a constant, my Father "liberated" my CRC
>handbook again!  Moles should be a constant here though, unless you're
>splitting atoms - or doing cold fusion <G>)

Hehehehe,  I like the cold fusion reference!

R=Rydberg's (sp?) constant= 8.3 Joules/(mole*Kelvin deg)

Sean


+-------------------------------+
| Sean Breheny                  |
| Amateur Radio Callsign: KA3YXM|
| Electrical Engineering Student|
+-------------------------------+
Save lives, please look at http://www.all.org
Personal page: http://www.people.cornell.edu/pages/shb7
shb7spamKILLspamcornell.edu  Phone(USA): (607) 253-0315 ICQ #: 3329174

1998\11\12@055256 by Arnold Grubbs

flavicon
face
Gerry Cox wrote:
>
> I need to cool an object to 4C below ambient. Ambient is in the range 20C to
> 30C. The object does not generate any heat of its own. Cooling does not have
> to be rapid.
snip
{Quote hidden}

Well, you could try a vortex cooler... I don't know the name of the
maker, but we
use a small tube like device about 1 1/2 inches by 10 inches long for
the
cooling/heating system in an enviroment suit that the company I work for
sells.

You put in about 30-100 PSI to a fitting in about 1/3 the way down the
tube,
and out of one end comes cold air, and out of the other
end comes hot air.  Don't know how cold it is, but it produces frost on
the
cold end if left running very long. We use them in other parts of our
plant
for cooling adheasive coated foam used in makeing another product.
If you are interested, I will try to find out more about the thing and
where we get them from... maybe they would sell them in small amounts.


AG

1998\11\12@080247 by EFRAIN_TORRES

flavicon
face
  The company who sell Vortrex tubes are EXAIR Tel: 1-800-EXAIR.  It can
cool down air to -50 F with capacities up to 10,200 Btu/hr, no moving
parts.

1998\11\12@083806 by Mike Keitz

picon face
On Wed, 11 Nov 1998 09:26:24 -0000 Gerry Cox <.....gcoxKILLspamspam.....DEK.COM> writes:
>I need to cool an object to 4C below ambient. Ambient is in the range
>20C to
>30C. The object does not generate any heat of its own. Cooling does
>not have
>to be rapid. I have sucessfully used peltier devices controlled by a
>PIC16C73 running a closed loop with a temperature sensor and PWM
>controlled
>mosfet drive to the peltier devices. However I have a problem
>dissipating
>the heat from the hot side of the peltier devices in the space
>available
>without increasing the ambient and defeating the object of it all :-(

You won't have to drive the Peltier very hard to get 4C cooling, I think
that it still deserves consideration.  They work most efficiently driven
with DC to the cooling level required.  It does work to PWM all the way
to max and off, but you'll get more heat on the hot side compared to
driving with DC.  A simple switcher type of regulator chip with some
minimal output filter will work fine.

>
>So I am experimenting with the idea of using a compressed air supply,
>which
>is available near the device, to achieve the cooling. I.e. from first
>principles I understand that if I allow the air to expand, then its
>temperature will fall. Does anyone know the maths involved and have
>any
>advice on methods to expand the air and cool the object?

There is a device that does exactly that, I'm not sure what it is called
but I did play with one a long time ago.  It was intended to cool the air
going into a chemical-proof suit like is worn for toxic spill cleanups.
The person inside the suit could then breathe clean cool air (*).  A lot
of air at about 100 psi went into one side, most of it came out warm from
one end, but some was quite cold coming from the other end.  I was more
impressed that anything could expand air and make it warmer at the same
time.    Opening or closing a valve at the warm end that regulates the
output of cold air.

I also remember that Popular Mechanics or Popular Science did an article
expalining how it worked (something about "seperating the cold from warm
air molecules", but those publications aren't exactly hardcore science)
There is a spiraling flow of air inside that makes something happens in
addition to simple expansion.   Also the magazine had plans to build one
in a simple machine shop.   It takes a lot of air, probably 10 cfm.  I
don't know if a much scaled-down unit could be built readily.


>An air
>supply is available at up to 6 Bar with quite high flow rate. The
>temperature of the compressed air before expansion is close to the
>ambient.

(*) Obligatory safety notes:  Unless specially processed, industrial
compressed air has quite a bit of oil in it and should not be used for
breathing.   In this project, it may be necessary to pipe the spent air
out of the box rather than just let it go where it may and deposit water
and oil on everything.  Be sure that nothing can burst if the air outlets
are blocked.


___________________________________________________________________
You don't need to buy Internet access to use free Internet e-mail.
Get completely free e-mail from Juno at http://www.juno.com/getjuno.html
or call Juno at (800) 654-JUNO [654-5866]

1998\11\12@100751 by Brian Striggow

flavicon
face
On Wed, 11 Nov 1998, Gerry Cox wrote:

> I need to cool an object to 4C below ambient. Ambient is in the range 20C to
> 30C. The object does not generate any heat of its own. Cooling does not have
>
> So I am experimenting with the idea of using a compressed air supply, which
> is available near the device, to achieve the cooling. I.e. from first
>

>From my empirical experience with being whupped on by broken air lines, I
would say that it would be difficult to reduce the temp to 4degC w/ just
a diffuser.

But try doing a web search for vortex coolers.  These use very moderate
amounts of compressed air to generate split streams of very cool and
very hot air.  As far as I'm concerned, they work by magic.  One common
application is to cool personnel that have to work in protective clothing
in devilish environments.

bcs

1998\11\12@104858 by Richard A. Smith

flavicon
face
On Wed, 11 Nov 1998 15:12:45 -0800, Mark Willis wrote:

>conducting bar of some other high heat transferring material (A friend &
>I joke about making IC's atop a Diamond layer, as it's a good heat
>conductor, he says.  If so, it's an expensive heat conductor!)

Dude this is no joke...  It's being done here at the University of Arkansas in t
he High Density Electronics Center (HiDEC).  They
make Multi-Chip-Modules and aparently heat is the real problem.  One of my frate
rnity brothers wrote a simulation program for the
laser they use to drill and cut the diamond.  It was his MS thesis.


--
Richard A. Smith                         Bitworks, Inc.
EraseMErsmithspam_OUTspamTakeThisOuTbitworks.com               501.521.3908
Sr. Design Engineer        http://www.bitworks.com

1998\11\12@132517 by Peter L. Peres

picon face
On Wed, 11 Nov 1998, Bob Blick wrote:

> Yes, there are commercial systems available. You can buy vortexes in any
> number of sizes and ratings.
>
> No need to reinvent the wheel. Do a search on "vortex cooling" and you'll
> find a few sources of them.
>
> Noisy little suckers, though.

Uh, a vortex cooler works on different principles than normal expansion,
no ? It's not a thermodynamic thingy, it exploits the statistical
distribution of 'local' density in a gas, which also depends on
'local' temperature and separates the different 'parts' of the gas by
centrifugal force, thus obtaining two fluxes, one of hot, and one of cold
expanded gas. This is very nice but it requires a lot of air and
mechanical tuning (tuning the air input won't do normally).

am I wrong ? Has anyone used such a cooler in a power electronics project
on this list ?

Peter

1998\11\12@144143 by Bob Blick

face
flavicon
face
> > Yes, there are commercial systems available. You can buy vortexes in any
> > number of sizes and ratings.
>
> Uh, a vortex cooler works on different principles than normal expansion,
> no ? It's not a thermodynamic thingy, it exploits the statistical
> distribution of 'local' density in a gas, which also depends on
> 'local' temperature and separates the different 'parts' of the gas by
> centrifugal force, thus obtaining two fluxes, one of hot, and one of cold
> expanded gas. This is very nice but it requires a lot of air and
> mechanical tuning (tuning the air input won't do normally).

Correct, except that they are fairly easy to adjust, choking off the hot
side output is what you do.

They work very well. A real benefit is the cold air rushes out, it's very
gratifying to feel ice-cold air!

They use a lot of air, but the Carnot cycle is not very friendly either
when put into common practice.

-Bob

1998\11\12@152313 by goflo

flavicon
face
Peter L. Peres wrote:

> Uh, a vortex cooler works on different principles than normal expansion,
> no ? It's not a thermodynamic thingy, it exploits the statistical
> distribution of 'local' density in a gas, which also depends on
> 'local' temperature and separates the different 'parts' of the gas by
> centrifugal force, thus obtaining two fluxes, one of hot, and one of cold
> expanded gas. This is very nice but it requires a lot of air and
> mechanical tuning (tuning the air input won't do normally).
> am I wrong ? Has anyone used such a cooler in a power electronics project
> on this list ?

Built one about 35 years ago - Power semiconductors were hard to find at
the
time :) but if memory serves your explanation is correct - It's called a
"Hilsch vortex tube", BTW. For the application mentioned it'd be
interesting
to pencil out the trade-offs: Compressed air is a fairly expensive
commodity,
but Peltier cooling is not particularly efficient, although extremely
reliable
and long-lived (100,000 hours +) in a proper design.
Really depends on the app - The quantity of heat to be pumped is'nt
mentioned
so anybodys guess...

Jack

1998\11\13@023324 by

flavicon
face
       <snip>
>   Alternate methods:  Heat Pipe to the outside with fan-cooled, wetted
> heat sink on the end (Swamp coolers work pretty well.)  Or heat
> conducting bar of some other high heat transferring material (A friend &
> I joke about making IC's atop a Diamond layer, as it's a good heat
> conductor, he says.  If so, it's an expensive heat conductor!)
>
>
I don't know about IC's, but here at work they mount laser diodes onto
diamonds for that exact reason.  Obviously we aren't talking huge flashy
jobs that wouldn't be out of place among the crown jewels!

Mike Rigby-Jones
mrjonesspamspam_OUTnortel.co.uk

1998\11\13@030455 by William Chops Westfield

face picon face
IIRC, "they" are doing all sorts of interesting and promising things with
thin film diamonds...  Shouldn't be that surprising - industrial class
diamonds have been cheap enough to cut tile with for a long time now, right?
Don't make an DeBeers (sp!) investments based on the industrial value of
natural diamond!

BillW

1998\11\13@035619 by Gerry Cox

flavicon
face
I would like to thank everyone who replied to my request for help on the
cooling problem.
A lot of you gave me very thorough answers which must have taken some time
to prepare. I am most grateful.
There is an amazing amount of knowledge amongst you PIC'ers.
I had never heard of a Vortex cooler and am now researching this
possibility.
I will let you know the results in due course.
In the meantime, thanks again to you all.

Best Regards,
Gerry.

1998\11\13@053140 by Peter L. Peres

picon face
If the amount of heat to be removed is really small, then a supply of
volatile liquid and a wick cooler will do much better than air. This
requires maintenance but I suspect that a 10 liter canister of propane or
such needs to be filled once a month or so.

I don't know anything about where this is to be used, so it might not be a
good idea after all.

Peter

1998\11\13@124952 by Morgan Olsson

picon face
At 11:38 1998-11-13 +0000, you wrote:
>If the amount of heat to be removed is really small, then a supply of
>volatile liquid and a wick cooler will do much better than air. This
>requires maintenance but I suspect that a 10 liter canister of propane or
>such needs to be filled once a month or so.

It also helps detecting a short cirquit..  KA-BOOOM !

>I don't know anything about where this is to be used, so it might not be a
>good idea after all.

There are probably better choices...

If only a feww degrees, maybe possible to just let water evaporate?
Use destilled water, and a "clunk" construction to keep level in a small
"basin " from which a piece of cotton fabric by capillary effect gest wet,
and can be directly on the object to cool.

I used that method some yrs ago to cool my salt water aquarium.
A small fan boosted it in summer time.

...but of course in electronic we don«t like water...

Maybe if mounted on the outside surface?

A good thing is that it don«t consume power.

/Morgan

>Peter

       Morgan Olsson                   ph  +46(0)414 70741
       MORGANS REGLERTEKNIK            fax +46(0)414 70331
       H€LLEKS           (in A-Z letters: "HALLEKAS")
       SE-277 35 KIVIK, SWEDEN               @spam@mrtKILLspamspaminame.com
___________________________________________________________

1998\11\19@130336 by Peter L. Peres

picon face
If the amount of heat to be removed is really small, then a supply of
volatile liquid and a wick cooler will do much better than air. This
requires maintenance but I suspect that a 10 liter canister of propane or
such needs to be filled once a month or so.

I don't know anything about where this is to be used, so it might not be a
good idea after all.

Peter

More... (looser matching)
- Last day of these posts
- In 1998 , 1999 only
- Today
- New search...