Cooling under pressure

UberBlue said:

Prresurizing water at ambient temperatures will not alter its chemical or physical properties directly pertaining to heat transfer; i.e thermal capacity, thermal conductivity, viscosity, ect.

Click to expand...

I'm almost 100% sure that an increase of pressure does change the chemical/physical properties of water. The effect of an increase in pressure or the same as an increase in the temperature although it takes a big increase in pressure to be equivalent to a relatively small increase in temperature.

Ok, I've never delved into it that far. Would you please cure my ignorance pertaining to the physical/chemical-temperature/pressure relationship? Are you thinking of Boyle's law?

well, i was most refering to tabulated value form one of my handbook, but the more i look at it, the less it makes sense to say that physical properties vary with increasing pressure. In fact they do change, but it is so small that it can be easily neglected. those value come form p,v,t curves (3D graph, v is specific volume not speed) and they are obtain from experiment mostly, although you could use Boyle's law in the case of ideal and non ideal (with the compressibility factor) gas.

So in short, you were right, increase in pressure, unless you go at extreme pressure, has neglegible effect on physical properties.

UberBlue said:

I don't remember the exact number but one gallon of water will produce hundreds of cubic feet of steam.

Click to expand...

1 gallon of water goes to 1245 gallons of steam, or a box about 11 feet on each side.

In any case, pressure will provide a lot more than you guys seem to give it credit for. Most of the pumps here are low pressure, even if they are high in the gallons per hour. The viscosity, or thickness, of the water will push back. So will every twist and turn. And turbulence will push back, too. So when you get a 300 gallon per hour pump with low pressure pumping through your system, the system pushes back. This back pressure may decrease the flow rate to well under 100GPH, depending on your pump.
A high pressure pump will be better able to compensate for this.

Think of a low pressure pump as a fan. Think of a high pressure pump as a tire pump.
Put a funnel in the valve of a flat tire and point the fan towards it. How fast does your tire fill up?
But the tire pump will continue to pump air in even when the tire is pushing back at 35psi.

Originally posted by Molybdym

bars, psi and nuclear explosions, this is the new thread to pay attention to!

Click to expand...

Speaking of nuclear explosions, that's the kind of pressure you would need to significantly change the properties of water. (other than boiling and freezing points.)

Caffinehog: I totaly agree with your low vs. high pressure pump statement. What we are talking about is pressurizing the whole loop to 10 - 15 bar static pressure. I.E. with the pump turned off, the loop has 15 bar at the pump suction, discharge, and every other infinite point inside the loop. Turn the pump on, and you create X bar differental pressure across the pump causing coolant to flow at a given rate dependent on loop restrictions. Now turn the pump off. Reduce loop pressure to 1 bar. The loop now has 1 bar at the pump suction, discharge, and every other infinite point inside the loop. Turn the pump on, and you create the same X bar differental pressure across the pump causing coolant to flow at the same given rate dependent on the same loop restrictions.

Speaking of insanely high pressures...I read a while back a research group pressurized water to something ungodly like 10 million atmospheres and discovered that H2O reorganizes itself into H6O3.

How would you 'pressurize' a water loop the way you are describing?

It would be easy to do. Just put a "T" inline with a bit of tubuing and an air fitting on the end. This would make a pressurized air pocket that would push on the water with X pressure.

Ok - so you would have air pressure within your system, not water pressure. Since water is virtually incompressible applying pressure to it will not have an effect on the water.

If anything you will get lots of air in your system as the high pressure will accelerate the rate at which it diffuses into the water.

Sure, if you put a pressure tap in the system with the pump off it will give a reading- but that is just the pressure of the air, not the water.

This pressure will not create additional head, flow rate, or anything. This is because there is no place with a lower pressure to create a differential. If you had a pipe filled with water with high pressure air at one end, and open to the atmosphere (or any other region of lower pressure) at the other end the water will be pushed to the region of lower pressure (think super soaker). This will not happen in the case of a closed loop - no pressure differential, no flow.

The key to this is the fact that water is incompressible, and in a closed system for a watercooling application - therefore no matter what reasonable pressure you apply to it there won't be any effect, at least no positive effect.

Water, like a lot of liquids, is *mostly* incompressible.

For one gallon of water, if you increase the temp by 20 deg C, you decreased the density in such a way that the water now occupies 1 gallon plus 5 drops (from machinist handbook, rev 26).

It's *almost* negligeable.

Incompressibility refers to density variability with respect to pressure, not temperature. We all know water goes through density changes with temperature, very small while in a phase, and large when the phase changes.

water will go from 998.2kg/m^3 to 999.4kg/m^3 at atmospheric pressure and 25psi respectively, that is why i said it is negligable, so Bigben is right, water is mostly imcompressible

drop water down to 4C, and you have a density of 1000kg/m^3. Note that temperature plays a much bigger role than pressure.

Last night I read a post citing a performance improvement in whitewater blocks when operated around 20 PSI.

You want to remember where you read that and link to it here eaglescouter? And are you sure that what you read was referring to a system pressurized to 20 psi, excluding the pump, or did it mean the pump could produce 20 psi head?

Cathar's post on the thread titled Whitewater W/B Experiences reads (in part):

"If one were to use a 20PSI pump, the block could be improved upon by greatly reducing the channel width. A 20PSI pump would allow a reduction of channel width down to around 0.2mm wide with fin height to match. The cooling plate itself (including the fins) would only need to be around 1/16" thick. The jet nozzle could be reduced in size and/or arrangement. Performance gain would be around 3C over the White Water for a 100W heat source."

So..... I'm no expert, but...... Cathar appears to suggest a modified whitewater operated at 20 PSI could acheive a 3C performance gain.

Thus I add 2+2 (hoping I don't get 9) and observe that this performance gain could probably be achieved through operating a pressurized system at 20PSI, allowing me to use a smaller pump in this pressure system concept. I assume that the entire system is pressurized (including tank and radiator).

Be nice! But do tell me if this is a faulty assumption.

Unfortunately that is a faulty assumption. He suggested using a 20 psi pump to move water in the coolant loop, which will cause higher flow rates and correspondingly higher velocities. Furthermore, he also suggests narrowing the nozzle and channels, which would increase the water velocity even more, at the cost of a higher resistance. Of course, with a 20 psi pump these types of flow restrictions would be no problem.

Your assumption that you could mimic the same behaviour with a pressurized system and a weaker pump is not correct however. Centrifugal pumps operate by causing a pressure difference with makes the water flow. However, it doesn't matter what the overall system pressure is as far as the pump is concerned. Think of it this way: if you pressurize the system to 20 psi, there's the same pressure throughout the loop with the pump off, so the fluid is not going to flow anywhere... turn the pump on and it's going to cause the same pressure differential as it would if the loop were at atmospheric pressure... or 50 psi, or 10000 psi.

Interesting:

I assumed in a pressurized loop, that the pump would cause flow and could only increase pressure if there were a constriction in the system, thus the constriction would cause the pressure increase not the pump.

I realize that some waterblocks create constriction thus create back pressure between the pump and the block.

Is it necessary that the water block restrict flow? Is this an essential design element in a water block?

(Sorry if these are dumb questions or assumptions)

Actually, all waterblocks create a restriction that's proportional to the fluid velocity squared. As such, it is not an essential design element, but a constraint that must be considered in your design. I think that you would do well to read some of the articles explaining how a watercooling system works... alternatively you could pick up a book that describes the fundamentals of fluid dynamics. Without understanding how the flow and pressure within a system relate to each other I feel you'll have a hard time designing any original and meaningful blocks.

"I think that you would do well to read some of the articles explaining how a watercooling system works... alternatively you could pick up a book that describes the fundamentals of fluid dynamics. Without understanding how the flow and pressure within a system relate to each other I feel you'll have a hard time designing any original and meaningful blocks."

Ummm, this thread is posted under general cooling, not under technical discussions, thus beginners should be welcome.

I don't intend for my water block to end world hunger, cure aids, or earn me the nobel prize. I just want to produce a useable waterblock for my home PC and happen to have access to a milling machine.

Other than the friction of a fluid flowing through a channel (or tube), I did not know it was necessary for a water block to create resistance. I had assumed that I could size my water lines and block channels to avoid creating any additional resistance (except for the obvious resistance caused by the directional changes within the block).

If beginners, non-engineers, and folks who have not been in a physics class in a very long time are not welcome on this thread, then please move it to the technical discussions forum.

Most of us are participating in these forums and in watercooling for fun and personal use, not for a career.

I appologise if you took that wrong. It was meant to help, not to criticise. Designing something by trial and error is perfectly valid... I was only trying to suggest that one can improve their initial assumptions with a little understanding of the topic at hand. I do not feel that my comments were overly technical... I believe that this is the point we are at with watercooling, and I certainly do not want to see it take any steps backwards. In any case, that's enough for me... I will leave you to it. Good luck with your designs.