Background:

To achieve a less noisy working environment, I've recently assembled a
water-cooled PC (using a passive, external and very cool-looking
reservoir/radiator (no pun intended) called "Zalman Reserator 1",
together with a number of additional cooling-blocks for my GPU,VRAM
and harddrives).

Not only did I succeed in achieving a virtually silent (<18dB) PC, but
it also turns out that even with this low-performance (passive,
remember, only convection around the reservoir cooling the water)
solution, my CPU and GPU temps. are still a good 10 Deg.C. lower than
with any high-performance (and thus even higher-noise) air-cooled
solution I've tried. Furthermore, it now takes about 20 minutes from
the time i disable the cooling (switch off the waterpump) till the
BIOS over-temp. threshold is reached, compared to mere seconds for
air-coolers (which often isn't merely enough to save the CPU from
burning out, never mind allowing the OS to shutdown cleanly).

The last fact lead me to the conclusion that I didn't need waste my
money on the water-flowrate and -temperature monitoring equipment I
originally planned to pruchase. The only "monitoring" I have need of,
is some simple way to quickly, at-a-glance, ascertain that water is
indeed flowing through the system. All I need to know about the
flowrate, is that it is somewhere in the vicinity of what I usually
see (ie. no partial blockages or near-failing pumps etc.) Anything
more precise would be overkill.

This need to check for flow, but without the need for knowing the
precise rate, led me to the rather crazy idea I'm going to present
below, for what might be both a pleasant "visual effect" and a quite
useful monitoring-tool for water cooled PCs...


Concept:

... What if one could introduce some "element" to the cooling water,
that by its discernability from the water itself, would make apparant
the waterflow, as it went through a transparant (and possibly lit)
section of exposed hose? (Most water-cooled systems will have at least
a small section of externally visible and transparent pipe or hose. My
setup has rather a lot, due mainly to the reservoir being external.)

This would not only allow for a quick, failsafe and low-tech way to
inspect that water is indeed flowing through the system at an
acceptable (ie. "near usual") rate, but could also be made to provide
a pleasant and/or colorful display (in lieu of the current "modding"
trend), especially if the "element" in question could be made to
flouresce and/or change color with temperature, and similar "fancy"
(but possibly useful, you will note) features.


Motivation:

Well, mainly because I personally would find it of great use, and even
more fun, to have something like this running around with the water in
my already impressive looking water-cooler (See for your self:
http://www.zalman.co.kr/eng/product/view.asp?idx=63&code=021) But...

Seeing as water-cooling is the next big thing (if the fact that it is
so much better than air doesn't do it, the increasing heat-production
of modern CPUs and GPUs will soon make water the only reasonable
choice, indeed IMHO it already is if you saviour your hearing!) It is
tempting to consider the commercial aspects of a product like this.


Implementation:

My first inspiration was to use some kind of minute colored beads,
balls or granulate; but I've learned that this might not be feasible
for various reasons. At least, a single producer of "MicroBalls" for
the medical research industry, whom I asked about this, responded that
it was probable that [their products] would clot and/or gather in
certain areas of the system over time (and, I can only assume,
probably in the exact places you least want them to, such as the
hottest, often mazelike, parts of the cooling-blocks, at the surface
of the reservoir or in the pump).

My second thought therefore, was to use another fluid; something that
will never completely mix or disolve with water (or the necesary
anti-corrosion additive), but occur therein as "streaks" or "bubbles"
of color. Something along the lines of whatever they use in those
"desktop wave-machines" that produce nice slowmotion waves in two 
non-mixing fluids, or the lava-lamp stuff (only it shouldn't harden at
room-temperature, for obvious reasons).

However, I'm no chemist, and wouldn't know how to procure, let alone
make, such (a) fluid(s)? So basically I'm looking for a fluid (posibly
two fluids, as the water could be replaced with any fluid with
suitable heat-carrying properties) with the following properties
(somewhat ordered after importance):

- The more plentiful (transparant) of the two fluids must have near
water-like cooling properties (perferably it should be plain water
with or without some additive). Generally it should not react with
various materials in a detrimental manner (specifically these often
used materials: Cobber, Aluminium, Steel, Silicone, PVC and various
other soft and hard plastics). Some kind of anti-corrosive additive
must be allowed or included in the mixture. The fluids must have
reasonable viscosities (and related properties?) so as to facilitate
"safe passage" through the pump as well as various kinds of
cooling-blocks. Much the same goes for the colored fluid, although
since it would not comprise much of the total coolant volume, its
heat-carrying and viscosity properties aren't as important. Both
fluids should be as non-toxic as possible (but since we are already
using a non-corrosive, we can accept some toxicity, such as by
ingestion and/or in concentration before being mixed etc.)

- The two fluids should stay seperate for at least several months
(IMHO the minimum acceptable interval for water changes), preferably
more than a year or indefinitely.

- The colored fluid must distribute in macroscopically discernible
pattern, such as the aforementioned "bubbles" or "streaks" (blotches,
threads, whatever), visualising the flow clearly for the naked eye.

- The above distribution/pattern must allow (perhaps even benefit
from) passing through a (typically medium-speed, impeller type) pump
every 10 seconds. (I suspect this might be the difficult one! A
possible solution might be something that would otherwise quickly
coalesce into large "blobs", but with nearly the same bouyancy as the
coolant )

- The colored fluid must be the heavier of the two, to ensure it
doesn't accumulate near the surface of the reservoir. Also, falling
quickly through the (in my case, 60 cm. tall) reservoir will allow
less of the colored fluid to be added (ie. it will be "reused" more
quickly than the clear fluid), as well as facilitating
"re-integration" at the bottom of the reservoir, before taking another
trip through the pump.

- The fluid(s) should either be commercially available, or producable
in a domestic kitchen setting.

The rest are If-Possible's, and would not ruin the concept by not being realised:

- If possible, the colored fluid should be available in a number of
opaque colors, while the clear fluid could either be completely
transparent or, better yet, available in some semi-transparent
contrasting colors. Flourescence by UV light would be a bonus,
especially if both fluids could be made to flouresce in different
colors. Even if flourescent, the colored fluid should still have a
distinct (ie. non-white non-transparant) color when normally lit.

- If possible, the colored fluid should be easily "detectable" by some
(non-expensive) electronic device (such as a photocell or CCD chip),
allowing it to signal the PC when the flowrate is low, possibly even
provide a flowrate reading. This basically entails the same as for the
naked eye, which is to say there should be a high contrast and clearly
defined border between the two fluids as they pass by the "detector".


Best Case:

The ideal (well, mine anyway) would be something along these lines:

- A colored fluid available in a number of normal/flourescent color
combos together with a concentrated additive for the water, likewise
available in a number of combos, but with semi-transparant color
variations. (Personally I'd like to have transparant water that glowed
slightly blue in UV with the colored fluid being bright red in normal
light, yellow-green in UV.

- When going through the pump, the colored fluid is "smashed" into
small 2-5 mm. "bubbles" that will be visible through the feed-line,
and then it coalesces rather quickly into gradually larger and larger
"blobs" (say 10 seconds to reach 10-20 mm.) Thus allowing different
"looks" in the feed- and return-lines, as well as ensuring that the
blob-size doesn't fall below the wanted (visible) minimum. The colored
fluid would be rather a lot heavier than the coolant (water), both for
the aforementioned reasons, and to possibly provide a "layered" look
to the return-water (assuming the colored fluid will have sunken to
the bottom half of the hose, and there joined to form a small "river"
rushing towards the reservoir).

- The colored fluid might have some heat-related state change. Such as
changing color (normal and/or UV) when reaching a certain temperature
(which should be somewhere near 35-40 deg.C. Preferably modifiable by
varying some component in the fluid mixture?) Or perhaps, like the
lava-lamp example, the viscosity and/or bouyancy might change,
providing a different "look and feel" when heated?

- Feel free to surprise me with your creativity here. As you can
understand the concept is rather loose. Anything that will give the
coolant a colored or otherwise visible "texture" or "pattern", making
the otherwise invisible flow plain to see, while at the same time
providing something nice to look at (while those tardy applications do
their "please wait.." stuff). I'm of course also open to suggestions
of the granulate/particle/bead kind, as long as the mentioned issues
have been properly addressed.

Bottomline: It is the resultant effect and properties, as (hopefully
adequately) described above, that is important, not the method choosen
to produce them.


I will be looking forward to trying your "recipes" in my cooler...


Thanks in advance,


inventus-ga.

---

Clarification of Question by inventus-ga on 14 Dec 2004 13:07 PST

In response to Comment #1 (zn833-ga):

I like the idea with the Galileo thermometer (I assume this is the one
often seen realised in glass, where a number of balls, filled with
colored fluid and a temperature "badge" attached, will rise to the top
as temperature rises).

Unfortunatly, my reservoir is made of aluminium (being a "reserator" =
reservour + radiator, it needs to be, as the heat is carried away by
convection through the ribs on the outside of the cylinder, see the
Zalman site for pictures!)

But aside from the fact that I could easily add a second reservoir to
the loop, this one made from a transparant material, it might be
possible to make a reserator in a transparent material as well? I will
certainly think about this.

I guess using thin-walled glass (pyrex for instance) in a shape
suitable for convection (ie. rippled, with a lot of surface area)
might provide a decent radiator effect (aside from looking awsome AND
allowing the Galileo thermometer to be seen!) Now I just have to find
someone selling large cylinders of pyrex? (Perhaps a labatory supplier
might have something?)

Do you happen to know what fluids are used in a Galileo thermometer?
If the commercially available ones use normal water for the clear
fluid (the one in which the glass balls float) I guess I could simply
buy one of those, break the big glass container, and put the balls
into my reservoir. That would really make for an easy "mod" (Although
I'll have to research such a transparant reservoir, but that shouldn't
be too hard to find or make.)

=

With regard to your "cooking oil" suggestion: Although I did imagine
the flow to be directly visible (ie. looking directly at the fluids
passing through the tube) as opposed to being vissible in the
reflection/shadow cast by a light shining through the tube, I guess
that might be one way of doing it.

While I wouldn't want to add cooking oil to my water-cooler unless I
was sure it worked (might be a bit messy), it should be possible to
carry out som preliminary tests in the kitchen. If I get the time to
do this, I'll post a clarification with my findings.

Thanks for your suggestions.

inventus-ga.

If you can see into the resevior, and it is of sufficient size and
volume, you could use the Galileo thermometor effect as a means of
measuring tempurate.  I know that isn't quite what you had in mind,
but by monitoring the temp and knowing the approximate temp your
machine should run at, you should be able to tell whether the cooler
is doing a sufficient job.  Plus they look cool.

As far as something to tell if water is flowing.  I think a major
problem in coming up with some liquid is that you are dealing with
water as a coolent.  There are few fluids that are heavier than water,
but will remain fluid at room temp.  Furthermore, you need some sort
of insoluble or organic substance so that it doesn't mix with the
water.  These are relatively harder to come by in an everyday
environment because they tend to have harmful properties.  I too came
up with the lava lamp fluid, but like you said it solidifies at room
temp.  Another possibility is cooking oil.  I know it is less dense
than water and will float, but as the water heats, more and more
should go into solution and pass through the system.  It may be some
what difficult to visualize, but here's what I have come up with, and
it also could have a cool effect, though difficult to quantify.  There
are two options.  First, color the water and use a white light. 
Second, don't color the water and use a colored light.  If you shine a
light through the water/oil mixture it should give a shimmering effect
when projected on a wall (tindle effect).  If, however, the water
stops flowing, the oil will float to the top, and you will no longer
see a shimmering.  Your light must, therefor, shine through a tube
somewhere near the bottom of a water column.  I can't guarantee this
will work, but if you are willing to try it, and it does work, it
should give a cool effect.

Just put vegtable oil in it. That does not mix with water. exparament
with different types of oils and you will find a good solution.