The next evolution to cathar's microchannel design - Page 2

jaydee · 01-12-2003, 02:39 PM

Quote:

Originally posted by hara
I assumed the core is all heating the same. I placed the thin base exactly on the centre because heat it is surrounded by heat. It's in the middle of the heat source. The reality is that half the area of the cpu core is L2 cache which doesn't produce much heat but anyways... I'll start researching where to machine it cause I'm sure it's a bitch.

I thought about a compination of thorns, mills and drill presses. What do you think?

Well a Mill can be used as a drill press. I m not sure what could be drilled though exept the barb holes and screw holes but if this is going to be done by a shop it would be better to have them do it with the mill. These small channels will be a pain, but they are strait so that is a bonus. I was impressed that morphin1 pulled it off so nicely with his mill. I would ask him what it took. I think you will have a good block here in any event.

hara · 01-12-2003, 02:52 PM

Is it more difficult to do than squirrel ?

jaydee · 01-12-2003, 07:27 PM

Quote:

Originally posted by hara
Is it more difficult to do than squirrel ?

No, this would be easier being the channels are strait. Curved channels add a extra level of difficulty as you are running more than one axis at the same time. Strait cuts all the load is in one direction and consistant. This will still be hard though and will require a mill with little slop in the axis's.

hara · 01-13-2003, 08:50 AM

A bit off topic, but how much does copper usually cost. for example how much will 2 78mm * 54mm * 10mm bars set me back?

bigben2k · 01-13-2003, 09:22 AM

Here's a rough representation of the heat gradient (red=hottest, light yellow=coolest).

Because your channel depth is deepest in the middle, the coolant will be facing a much hotter surface in the center.

Now that's usually good, but this goes back to a baseplate thickness issue: if it's too thin, the heat is too great for the coolant to pick it up (it can be compensated in many ways). If it's too thick, then the heat will spread some more within the rest of the block, resulting in a higher die temp.

Maybe you can follow up on this? Maybe I just don't understand your goal/applied principle here?!?

hara · 01-13-2003, 11:52 AM

The block will be slightly modified for accomodating the O rings, channel depts, screws to hold the block together AND MOST IMPORTANT, to force more water through the central channels. Keep in mind, these will be slight modifications In the meantime, BigBen, Don't forget this:

hara · 01-13-2003, 12:02 PM

One thing that bothers me is this:

1/2" id pipe cross sectional area: 113mm^2

My slit area: 17*2mm = 34mm^2

That's quite some restriction!

Cathar, what advice can you give me?

bigben2k · 01-13-2003, 01:18 PM

Same thing (see graph).

As for the fittings, it's ok. Cathar has 8 channels, with a final opening width around 2.0 mm, for 16 mm^2.

hara · 01-14-2003, 05:58 PM

Here's an update. By modifying my design a bit, more water is forced to pass throught the 3 central channels. It is difficult to see, if you wish I can attach the cad file.

hara · 01-14-2003, 05:59 PM

A new wide channel has been made in the empty space.

hara · 01-14-2003, 06:00 PM

From the other side. Copper stock should soon be with us...

UnloadeD · 01-14-2003, 08:45 PM

I don't know about the thermodynamics, but I'm wonering if something like this would help balance flow:

If you came down on both sides of the "divider" with a ball end mill you would have a bit of slope directing the flow towards outlet, likewise with the bottom opening in the top-plate. Anyway, just a thought, it came to mind while looking at morphling's block too. It just seems to me it would be better having the water directed rather than colliding below the outlet.
I'm not sure how the block will perform but the little I know about milling tells me it will be expensive to make. Also your drawings don't reflect the roundness of the tools on outside corners and such. Probably won't make much difference, but you should take that into account.

peace.
unloaded

hara · 01-15-2003, 04:22 AM

Thanks,
As for the roundness, I think it'll be better that way (rounded)

I'm sure It'll be expensive to machine but it is not a commercial block.

hara · 01-15-2003, 01:42 PM

Have you got hints on how to machine this? Guess the microchannels will have to be cut with a thorn. A mill would also have to be used.

hara · 01-16-2003, 08:43 AM

Something tells me nobody's interested in the outcome :shrug:

bigben2k · 01-16-2003, 09:59 AM

We're interested, but it's looking like you're about to hit a wall, and share the experience, so we all wait in anticipation.

You've got a flow balancing problem:
1) you'll find that there's more flow through the fins closest to the outlet (how much, dunno, probably not significant)

2) your nozzle plate doesn't help getting more flow in the central channel, the deepest, even though each channels exit is the same size.

You haven't addressed the heat distribution issue. This design might work just fine, assuming that the central channel's extra heat can be picked up by the surrounding channels; it's hard to tell.

hara · 01-16-2003, 11:12 AM

Big ben,
1) The top compensates for distributing the flow evenly by acting in the opposite way of the bottom.

2) true

I'll see what I can do

bigben2k · 01-16-2003, 11:20 AM

Quote:

Originally posted by hara
1) The top compensates for distributing the flow evenly by acting in the opposite way of the bottom.

How?

hara · 01-16-2003, 01:49 PM

water will tend to pass through the fins fartest from the outlet

hara · 01-30-2003, 09:53 AM

This is an update. My friend "got" (stole) enough copper to make four of these blocks

1cm thick.

Next problem: Cutting the b1tch

hara · 02-04-2003, 01:30 PM

What endmills should i use to cut it? Material and size

bigben2k · 02-04-2003, 01:34 PM

2 0r 3 flutes, TIN coated, if possible.

LiquidRulez · 02-04-2003, 08:08 PM

Go with a high helix >30 degrees.....prefferably 45.
Im a solid carbide kind of guy myself, but the more flutes you use..the stronger the end mill is. But also, the more flutes, the less chip clearance and they have a tendency to "gum up" quicker.

HSS Tin coated bits are pretty hard also, but they dont hold their cutting edge as long as carbide does in my experience..and carbide can be had for as cheap as HSS plain bits if you shop at the right place.

Be very very patient with copper. You WILL break alot of bits if you try to cut too much at once when in a hurry.(youre going to be breaking quite a few anyway)

When you do buy your endmills....buy about 3 times more than you think the job will require......youre going to need em'!

jaydee · 02-04-2003, 08:16 PM

Quote:

Originally posted by LiquidRulez
Go with a high helix >30 degrees.....prefferably 45.
Im a solid carbide kind of guy myself, but the more flutes you use..the stronger the end mill is. But also, the more flutes, the less chip clearance and they have a tendency to "gum up" quicker.

HSS Tin coated bits are pretty hard also, but they dont hold their cutting edge as long as carbide does in my experience..and carbide can be had for as cheap as HSS plain bits if you shop at the right place.

Be very very patient with copper. You WILL break alot of bits if you try to cut too much at once when in a hurry.(youre going to be breaking quite a few anyway)

When you do buy your endmills....buy about 3 times more than you think the job will require......youre going to need em'!

Pretty much sumed it up. I am not sure they make a 45 degree helix endmill that small though.

I can get 60 degree in 1/4", 45 degree in 1/8", but havn't seen anything under 1/8" above 30 degree. I maybe looking in the wrong places though.

01-16-2003, 09:59 AM	#41
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	We're interested, but it's looking like you're about to hit a wall, and share the experience, so we all wait in anticipation. You've got a flow balancing problem: 1) you'll find that there's more flow through the fins closest to the outlet (how much, dunno, probably not significant) 2) your nozzle plate doesn't help getting more flow in the central channel, the deepest, even though each channels exit is the same size. You haven't addressed the heat distribution issue. This design might work just fine, assuming that the central channel's extra heat can be picked up by the surrounding channels; it's hard to tell. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

01-30-2003, 09:53 AM	#45
hara Cooling Savant Join Date: Oct 2002 Location: Malta, Mediterranean Posts: 662	Update This is an update. My friend "got" (stole) enough copper to make four of these blocks 1cm thick. Next problem: Cutting the b1tch

02-04-2003, 01:34 PM	#47
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	2 0r 3 flutes, TIN coated, if possible. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

01-12-2003, 02:52 PM	#27
hara Cooling Savant Join Date: Oct 2002 Location: Malta, Mediterranean Posts: 662	Is it more difficult to do than squirrel ?

01-13-2003, 08:50 AM	#29
hara Cooling Savant Join Date: Oct 2002 Location: Malta, Mediterranean Posts: 662	A bit off topic, but how much does copper usually cost. for example how much will 2 78mm * 54mm * 10mm bars set me back?

01-13-2003, 12:02 PM	#32
hara Cooling Savant Join Date: Oct 2002 Location: Malta, Mediterranean Posts: 662	One thing that bothers me is this: 1/2" id pipe cross sectional area: 113mm^2 My slit area: 17*2mm = 34mm^2 That's quite some restriction! Cathar, what advice can you give me?

01-14-2003, 08:45 PM	#37
UnloadeD Cooling Savant Join Date: Oct 2002 Location: MidWest USA Posts: 176	I don't know about the thermodynamics, but I'm wonering if something like this would help balance flow: If you came down on both sides of the "divider" with a ball end mill you would have a bit of slope directing the flow towards outlet, likewise with the bottom opening in the top-plate. Anyway, just a thought, it came to mind while looking at morphling's block too. It just seems to me it would be better having the water directed rather than colliding below the outlet. I'm not sure how the block will perform but the little I know about milling tells me it will be expensive to make. Also your drawings don't reflect the roundness of the tools on outside corners and such. Probably won't make much difference, but you should take that into account. peace. unloaded

01-15-2003, 04:22 AM	#38
hara Cooling Savant Join Date: Oct 2002 Location: Malta, Mediterranean Posts: 662	Thanks, As for the roundness, I think it'll be better that way (rounded) I'm sure It'll be expensive to machine but it is not a commercial block.

01-15-2003, 01:42 PM	#39
hara Cooling Savant Join Date: Oct 2002 Location: Malta, Mediterranean Posts: 662	Have you got hints on how to machine this? Guess the microchannels will have to be cut with a thorn. A mill would also have to be used.

01-16-2003, 08:43 AM	#40
hara Cooling Savant Join Date: Oct 2002 Location: Malta, Mediterranean Posts: 662	Something tells me nobody's interested in the outcome :shrug:

01-16-2003, 11:12 AM	#42
hara Cooling Savant Join Date: Oct 2002 Location: Malta, Mediterranean Posts: 662	Big ben, 1) The top compensates for distributing the flow evenly by acting in the opposite way of the bottom. 2) true I'll see what I can do

01-16-2003, 01:49 PM	#44
hara Cooling Savant Join Date: Oct 2002 Location: Malta, Mediterranean Posts: 662	water will tend to pass through the fins fartest from the outlet

02-04-2003, 01:30 PM	#46
hara Cooling Savant Join Date: Oct 2002 Location: Malta, Mediterranean Posts: 662	What endmills should i use to cut it? Material and size

02-04-2003, 08:08 PM	#48
LiquidRulez Cooling Savant Join Date: Sep 2002 Location: In Hell Posts: 322	Go with a high helix >30 degrees.....prefferably 45. Im a solid carbide kind of guy myself, but the more flutes you use..the stronger the end mill is. But also, the more flutes, the less chip clearance and they have a tendency to "gum up" quicker. HSS Tin coated bits are pretty hard also, but they dont hold their cutting edge as long as carbide does in my experience..and carbide can be had for as cheap as HSS plain bits if you shop at the right place. Be very very patient with copper. You WILL break alot of bits if you try to cut too much at once when in a hurry.(youre going to be breaking quite a few anyway) When you do buy your endmills....buy about 3 times more than you think the job will require......youre going to need em'!

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)