Predicted effects of changes in Kinematic Viscosity on h(conv coeff) at constant Density, Conductivity and Heat Capacity.

http://www.jr001b4751.pwp.blueyonder.co.uk/Cascade2.jpg

What you've plotted is the basic variation in convectional efficiency with viscosity. What I'm having troubles with solving is what's occuring within the G* blocks before the convectional action. Increased viscosity will definitely be interfering with the longevity of the "turbulation" effects that assist with the block's increased convectional activity.

Have completed re-drawn plans for the G4/G5 which should improve things slightly, and should not be at the expense of "plain water" performance.

Still, it is rather an interesting relationship, and definitely something for people to consider before dumping heaps of glycol in their cooling loops.

Suggest "longevity" is included in the Impingement(Flomerics) calculations.
The Velocities(4.17m/s and 6.67m/s) are those in the nozzle.

Would suggest that Flomerics is attempting to model "classical" jet impingement only.

Yes, it is, also, my understanding is that it is not intended for Submerged Jet Impingement(SJI).
Probably a "fumble to far" to suggest any major "longevity" contribution.
Will wait for your model, when wb's details are released, before further sums.

I think it suffices well enough for SJI, at least in a relative sense. It seems to roughly agree with my experiences, with performance differences for varying jet standoff heights being the notable exceptions.

So yeah, acceptable as a good guidleline for initial starting points.

If C/W(TIM) ~ 0.1: yes.
However if C/W(TIM)~ 0.05 , I think I prefer Sieder-Tate( eg Re(3x1mm Entry) from here )
Graphs are still being developed.

Edit: Typo: "C/W(TIM)~0.05" did read "C/W~0.05"

I am unsure if this contributes anything to your divagations....
I refer to Incoherent's graph showing relationship between Re and kinematic viscosity. Isn't there an equilibrium between kin. visc. and Re number? If kin visc. decreases then velocity increases and hence Re goes up as well.
Just some idle thinking here, if worthless just ignore and do not reply :)

Yes it would. I'm not quite sure what you mean by equalibrium in this case though. More a strengthening I'd say. Or?

Also something for the testers/reviewers out there to consider as well. A block could possibly show massivelly improved (CPU-H2O delta T) performance simply by increasing the water temperature.
I'd love to see if there really is a difference. All very well to play with numbers but it's nice to see real life.
(I discovered something else today. Hot water can freeze faster than cold water. An old wives tale that is actually true but not understood.)

well I was waiting to see if the obvious would be discussed
sure the viscosity changes, and with good instrumentation AND good technique the changed performance is readily apparent
- this suggests an 'optimum' temp (IF the wb has been so designed)

tweek away, are the blind fumbling for the light switch which is not even connected ?

sorry guys, these are numbers for the sake of numbers

I'm just waiting for my blocks to get back to me...

Well, numbers are interesting, and are quite illuminating themselves. The temperature/viscosity relationship is moot however, since our end goal is (normally) lowest load temperature, not highest convectional efficiency. I think 'optimum' coolant temperature within the bounds of reality is as cold as possible.
The viscosity in itself however, is highly relevant.

Not "numbers for the sake of numbers".
Numbers were used in the design:
Playing with a performance number(h(con coeff)) and kinematic viscosity

I think I understand what is being played with,
but when the desired/reasonable result is arrived at via fiddling with the factor of interest; . . . . .

ignore my skepticism, it is not new news on this topic - (h

What country are they coming from? Customs issue or just delayed shipping?

Hmmm, maybe reading too much into my statement.

Much of my fumbling math was done assuming 25C water, and the kinematic viscosity that goes along with that. That was what was used to drive the basis of the design, which then got refined through emperical results (and a rough attempt at adapting the maths to fit). I hadn't really anticipated 25% glycol viscosities and the effects that could have.

Sunny Saskatoon apparantly.

Have decided not to wait any longer and have submitted the refined design to the machinists already. I figure by the time that the blocks get back I can do a shoot-out between the old and the new, with plain and glycol-gooped water, and decide then on which way to swing.

The new block could perform worse (and I have a slight feeling that it will) for plain water, but in any event I'll find out either way.

To be shipped by dogsled.......waiting for snow aparently :dome:

I'll be back at home around midnight; will check on a tracking number then Cathar.

I thought hot water froze faster than cold water because it was evaporating away (and decreasing the volume of liquid water that is involved)? Pretty sure I recall something like that from a P Chem discussion...

The idea of hot water freezing faster, while true, is retarded. Put less water in. Then it will freeze even faster and you'll get the same amount of ice!

It doesn't account for all of it, the Mpemba effect occurs in closed containers as well.

interesting. The more you know, right? It's been a long time but I seem to recall doing the math on an open system and evaporation could explain it. It's biased though because we were given the numbers to work with rather than actually doing said experiments. Guess the P Chem prof was just seeing if anyone would actually go out and look for more research on the topic...

Hmmm, can't remember where I read it now. The explanation that I read for hot water freezing faster had to do with the crystalline nature in which water molecules align themselves.

Apparantly water is a very highly structured liquid if left to sit. When it's cool, (and sitting still) it apparantly conducts heat much more readily through its structure. When the water starts off hot, the exited molecules break down the crystalline structure of the water, which takes a long time to re-establish.

So basically what is meant to happen is that the initially cool water freezes throughout, meaning that the entire mass of the water has to be brought to the freezing point. In the initially hot-water scenario, the outer edges freeze more quickly.

Seemed a little wishy-washy to me though.

I remember playing with my TEC water-chiller, attempting to freeze up a few litres of water in a bucket. The water would keep on flowing down to around -2.5C or so, until the water froze up solid inside the TEC chiller, forcing the flow to stop.

I would then see some little ice crystals exit from the loop outlet into the reservoir just as the flow was stopping, and then the water turned into what can be best described as a set of icy sheet layers, of a highly crystalline nature. The whole reservoir of water froze up into this sloughy set of ice sheets, and somewhat more curiously, when this process happened, the water temperature jumped up to -0.1C almost instantly.

It was like the kinetic motion of the liquid was enough to stave off the phase-change into solid form for a while, until when the motion was removed by the flow being blocked, then that catalysed the near total phase-change of all the water in the bucket into icy sheets. Now when I say "sheets", these are not layered horizontally. They were more like at a 45 degree isometric angle kinda deal.

Yeah, I know, kids playing with water, but it was fascinating to watch the phase-change into a solid take place, and then to measure the temperature jump, and the form of the frozen water.

I'm not sure how useful this is but here's another hot water freezes faster than cold discussion.

The results from this less advanced test suggest water has to be very warm ~140-212 deg F (60-100 deg C) to realize any appreciable decrease in freezing time compared to cold tap water.

man,what a tease this is.i live in saskatoon,and know somewhere in this great little city,lurks not only phaestus,but soon the new cascade :cry:
fyi,the igloos are melting pretty quick today :D

KnightElite lives here too. We should get together for some beers + LAN action sometime. I have a finished basement and plenty of room (and mediocre gaming skills)

that sounds great.not sure if i am up to transporting my rig tho,
26"chenming case with an external watercooling box.
nice to know this type of action going on here

Ok , numbers are a decoration.
Have deleted mine.