the 3 main points of custom block design

I have here a dilemma, how to design a water block to have the best possible heat transfer. Material is obvious, copper. My next three questions are; how big should the channels be, how fast should the water more thru these channels, and what kind of interference and turbulence should be created above the core.

For size of the channel, the dimensions are height and width, though my illustration has only one, this is trivial.

For water velocity, this is simply how fast the water is moving, not mow much is moved per unit of time, that is determined by velocity and size. I ask this because I have heard different things about how fast it should be moving,.

Lastly, turbulence, I can put the micro channels above the core in parallel to the flow, for less resistance, or perpendicular to it to create a lot of turbulence.

Since I don't have either a degree in thermodynamics or hydrodynamics I would like it if someone could answer there. What I am certain of is that these are the three most fundamental questions for a water block IMO.

So basically you're asking for people to tell you how to make a kick-*** efficient waterblock? Isn't that what many of us are trying to figure out for ourselves?

Some keywords to plug through google:

1) Microchannel
2) Jet Impingement

Read up on both, understand it, and you're on your way to figure out what you need to do and more importantly, why you need to do it.

microchannel i understand, and know fairly well from other blocks like that in the hydrocool 200 kit. they seem to work well. however a real explenation of jet impingment is harder to find.

tho it would seem that smaller is better from what i was able to understand, and faster too.

The first link I ever saw about it, was actually about jet inpingement designed as emergency cooling of a Korean nuclear power plant.

The principle is that in order to achieve maximum turbulence without using an outrageously powerful pump, all you do is project water right at the surface to be cooled: as the water jet hits the wall, it goes into turbulence, from the sudden change of direction.

The reason that turbulence is important, is because it reduces the thermal resistance, allowing the heat to travel more easier, from one layer to the next.

What I can tell you, beyond what's already available in these forums, and in the miscellaneous articles under the watercooling section right here, is that there is no hope of achieving turbulent flow strictly from a pump: it's just not cost efficient.

So you're down to turbulators, and jet inpingement.

Good luck!

so if i'm reading you right. the best way to do it might be to have the water enter the block right above the core where the heat comes from, which is sometihng known on these boards. and to have the water impact a flat surface right above te core that is very thin, say 1/16 thick with no fins or chanels in the way there. than have fins and channels for added surface area close by the core so after the water slams into the flat palte above the care it goes thru the fins and or micro channels and than out of the block.

i have heared that the heat from the core does not spread thru the block fery far beyond the area of contact with the core, so how big an opening should i use for the "jet"? I am thinknig that this will determine how bit the flat area above the core is, so than can have the fins/channels start just beyond this area.

That's going to depend on the pump you'll be using. You can google for a formulae to calculate the pressure drop of the nozzle.

As for restrictions, if you go that route, you'd miss the point of White Water. The fins extend the surface area from which the heat is drawn.

In fact the latest design that Cathar recently presented, as well as I've posted (at ProCooling), directs an array of small jets into cups drilled into a copper baseplate. The idea is to force jet inpingement not only against the baseplate, but also up against the walls of the holes.

Augmentation

I'm trying to put this to good use on a block of mine right now.

Impingement (Jet) Cooling

This one too.


Fin Calc

This should help also.

Does one of those things say that it is possible to bond two different pieces together for the heatsink? I am wondering because I want to make a design but it would require adding fins to it and i remember hearing that if it was milled out of a solid block it would be more heat efficient.

mmcshmi11 said:

Does one of those things say that it is possible to bond two different pieces together for the heatsink? I am wondering because I want to make a design but it would require adding fins to it and i remember hearing that if it was milled out of a solid block it would be more heat efficient.

Click to expand...

Yes, the bonding agent or solder would just be another thermal resistance.