My "I have a dream" block
 

Sorry if this has been addressed before. Heavy-Hitters are invited to wade in.

Ever since I held a sintered copper water filter disk in my hand, blew through what appeared to be a solid device and pondered the surface area inside it, I have wondered how it would perform in a water block. Examine the diagram and ponder the process. Chide in on the parameter than need to be determined:

The ideal size of the copper beads that would be sintered together to form the matrix? IE .5mm beads, 1mm beads, etc.

The ideal thickness of the resultant disk

The ideal thickness of the silver or copper plate that the disk would be bonded to.

Essentially, water is driven down into the copper (or more costly silver) sintered disk. The top would prevent the water from doubling back out through the top of the disk. Some of the water would eventually reach the bottom of the disk deflecting off of the copper (or more costly silver) plate, exit the disk in random directions on all sides, being routed by the ovoid sleeve towards the exit barb.

http://hill195.home.mchsi.com/ideas/sinter.jpg

Food for thought...

Hoot

Not sure I follow the drawing to well. From the side view it looks like you got 7 layers magically floating in the block. Top view looks like a bunch of wavy channels.

So this design looks to create lots & lots of turbulence right?

did either of you read the post?


big pile of copper beats connected together

The perfect Super-heavy restrict flow, for several parallel blocks?

Huummmnn.....

but how would you get teh water to run towards the bottom of the block?
isn't that where the heat is?

Looks sort of like copper foam to me. I reckon MadHacker may have a point. You might find it works better if you pass water across (rather than down then across) the block.

Any ideas on how to bond the foam to the base? For that matter how does the sintering process work?
It would be incredibly restrictive alright but it's a very cool idea.

After I first saw the sintered copper disk, comprised of what must be thousands of tiny copper beads, I did a little reading on the process. The individual beads are shaped in the form you want them to wind up as, then heated so that they barely reach their melting point, but not far enough so that they become entirely liquified. Wherever they touch one another they melt first and if you stop the heat once that has occurred, they are all bonded, quite strongly to one another. If you examine a model of spheres piled together, they only touch each other in small places, leaving plenty of space for air or liquid to pass between them, while providing lots of surface area in contact with whatever is pasing through them. The idea with the block was that water entering the matrix would drive downward and quickly divert into many random directions, eventually reaching the bottom of the matrix. Thats why I said the optimum thickness of the disk would have to be determined through experimentation, such that it wipes the liquid a lot while not deterring it from striking the baseplate too much. Probably the biggest challenge would arise in finding a satisfactory method for bonding the disk to the baseplate. Loading the beads onto an already made silver plate probably would not work since the sintering temperature of copper is higher than the melting point of pure silver. You could take the sintered disk and place it into the mold used for casting a molten silver plate and pour the silver onto it. Hopefully the viscosity of molten silver would be great enough that it would not flow too far into the copper matrix. Another approach would be to saturate the disk with wax and then mill off the wax on one face of the disk, exposing the copper beads. Then, if you have access to a plating tank, you could plate that face until the pores totally clog with silver and finally the deposition would build up until the surface was plated thick enough to mill back to flat. Once you have that accomplished, you could warm the disk and melt out the wax. I don't have all the answers, just a starting point for ponderance. The diagram did not convert very well from a .bmp to .jpg format to allow the matrix to show the detail of its composition, but again, it was just a starting point. If you go to some of the sintering web pages and look at finished products, you can get a better mental image of where I was headed.

Hoot

Yes, but didn't understand, same as drawing. Sintering is something I have no knowledge of so wasn't sure what he was talking about.

So what I am not sure about tnow is how these sintered beads connect to the base of the block?? Are they sintered to the block aswell? What happens to the thermal properties of the copper after sintering and how well will the heat transfer from the copper base to the sintered part?

Biggest trick would be...
 

Getting more significant bonding to the base w/o melting beads into one big mess, err I mean mass :shrug: :D
Must take into account total surface contact - Guessing some serious testing. Consistent melt rate/time seems like a real bugger.

Also would simple stream be better vs Cathar's implementations (many streams impinging)

use a copper base and heat from the bottom of the block up to top till the slintered part begins to melt as is done in the slintering process

More importantly: what PUMP? :P

May I suggest naming it the "pachinko block"?

Doh! (I remember seeing that b4 too)
 

:evilaugh: :rolleyes:
Thank Lolito

Yes, thanks Lolito for the link, though it left me disappointed that my idea was not an original one. Oh well, back to pondering....

I do love the "Pachinko Block" idea! :p

Hoot

I predict dismal results; this is not going to induce enough turbulence to lower the thermal resistance between the copper and water. Note Lolito's similarly linked product with a 22 deg C dT between die and air...

That was another question, not my interpretation. I suppose reading it again I could have chose better phrasing. Everyone has a post like that every once in a while so I doubt anyone here will hang me for that, & if so I'll just making my own noose ;).

Looks like a good design to me. I've always been one to believe that there are lots of different designs out there waiting to be found & this is keeping that theory alive :). I hope the design works out well.

Here's a link to a site giving some explanation to the sintering process.

Hoot

hey, faut lire les articles de Rosco sur les wb Ã* microcanaux ;)

(turbulence isn't the only factor...)

I see two 'problems' with this design.

The highest heat transfer coefficient is likely 'wasted' where the water first strikes the copper beads, at the top layer. By the time the water hits the bottom plate, it will likely have slowed a great deal, decreasing turbulence and the good heat transfer characteristics that come with it.

My second point, which explains why I said 'wasted' in the first paragraph, is that by only having (in theory) point contact between the beads this will make for horrible heat transfer from the base block up to the top of the beads.

Of course my first point may be totally wrong and my second point depends on the manufacturing process, the number and size of beads, etc...

Interesting idea still.

That's air, though, and it's 0.15C/W, and that's on a 49mm^2 die, with 6PSI mounting pressure. Seems to be harsher (indicating worse results) than we'd be experiencing, no?

The sintered part does not have to be as thick as a slice of bread. The optimum thickness would have to be determined through experimentation, but I'd imagine it would only be somewhere around a few milimeters. Perhaps omitting the baseplate and opting just to plate up the bottom of the sintered part until its sealed would be more effective. Sintering manufacturers make the parts in any number of form factors and I'd bet having one side of the part a solid cast is an option. That kind of eliminates the prospect of using an existing form, such as the disks employed as filters though.

Hoot