|
|
Water Block Design / Construction Building your own block? Need info on designing one? Heres where to do it |
Thread Tools |
07-19-2002, 06:55 PM | #26 |
Cooling Neophyte
Join Date: Jul 2002
Location: missouri city, tx
Posts: 7
|
true, but you might have to reduce it alot, and i can tell you, drilling 2" deep holed in copper is hell, i broke 3 bits in it, one of them is still in there. you basically can go mmuch less than .1875" bits.
__________________
xp 1600+, still to be oc'ed, 768 mb ram, msi kt266pro2, baracuuda 4 40gig, linksis nik+cable modem, jaton gf2mx400 64 meg http://www.overclockers.com/tips835, cd+cdrw fixing to do water cooling |
07-19-2002, 07:36 PM | #27 |
Hottest Stank of them All
Join Date: Mar 2002
Location: Washington, DC area
Posts: 533
|
I use my giant "moon 'laser'" to drill my WB holes. (just getting my Austin Powers lines ready for when "Goldmember" comes out.
Seriously, what is the correct way to drill a 2" deep hole into copper? Drill press, of course, but what cutting oil/ technique?
__________________
Does a radioactive cat have 18 half lifes? --Kenny my pimpin' rig: ...previously poorly cooled... 486DX-2 66mhz @ 75mhz, 4Mb ram, shared, 256Mb hard drive Onboard VGA, watercooled, of course 16-color monitor, labtec speakers, 28kbs USRobotics modem Windows XP (think about it) Maze 3, DDen GF3 block, Eheim 1250, econoline van HC, 1/2" hose w/ 5/8" fittings Comair 172mm fan |
07-19-2002, 08:02 PM | #28 | |||
Hottest Stank of them All
Join Date: Mar 2002
Location: Washington, DC area
Posts: 533
|
Quote:
Quote:
Quote:
__________________
Does a radioactive cat have 18 half lifes? --Kenny my pimpin' rig: ...previously poorly cooled... 486DX-2 66mhz @ 75mhz, 4Mb ram, shared, 256Mb hard drive Onboard VGA, watercooled, of course 16-color monitor, labtec speakers, 28kbs USRobotics modem Windows XP (think about it) Maze 3, DDen GF3 block, Eheim 1250, econoline van HC, 1/2" hose w/ 5/8" fittings Comair 172mm fan |
|||
07-19-2002, 08:21 PM | #29 |
Hottest Stank of them All
Join Date: Mar 2002
Location: Washington, DC area
Posts: 533
|
BillA, WTF do you mean, wrong? And besides, this is WB THEORY, so we're just throwing ideas out.
Do you think that all of the heat from the core is dissipated into the water directly above the core, and that the outer spirals of a WB don't contribute to cooling? I didn't mean tiny holes, just not too big. I want to fit a couple or three rows of holes in the block. And why wouldn't more holes reduce resistance? ...additional channels in parallel usually have this effect, yes? (and don't merely answer, no!) Perhaps instead of giving one word, negative replies, you could elaborate, and enlighten us all w/ your vast knowledge of this (clearly) cut & dry science called watercooling.
__________________
Does a radioactive cat have 18 half lifes? --Kenny my pimpin' rig: ...previously poorly cooled... 486DX-2 66mhz @ 75mhz, 4Mb ram, shared, 256Mb hard drive Onboard VGA, watercooled, of course 16-color monitor, labtec speakers, 28kbs USRobotics modem Windows XP (think about it) Maze 3, DDen GF3 block, Eheim 1250, econoline van HC, 1/2" hose w/ 5/8" fittings Comair 172mm fan |
07-19-2002, 09:10 PM | #30 |
Hottest Stank of them All
Join Date: Mar 2002
Location: Washington, DC area
Posts: 533
|
I must admit, I don't think I've ever used google... I don't really know anything about it. Maybe I'll explore a bit.
[edit] ECUPirate don't make ME get grumpy -Administrator [/edit] Oh, and I can be cynical as well... maybe cynical isn't the word... hmmm.
__________________
Does a radioactive cat have 18 half lifes? --Kenny my pimpin' rig: ...previously poorly cooled... 486DX-2 66mhz @ 75mhz, 4Mb ram, shared, 256Mb hard drive Onboard VGA, watercooled, of course 16-color monitor, labtec speakers, 28kbs USRobotics modem Windows XP (think about it) Maze 3, DDen GF3 block, Eheim 1250, econoline van HC, 1/2" hose w/ 5/8" fittings Comair 172mm fan |
07-19-2002, 11:11 PM | #31 |
Cooling Savant
Join Date: Feb 2002
Location: Dione, sector 4s1256
Posts: 852
|
__________________
There is no Spoon.... |
07-20-2002, 07:21 AM | #32 | |
Cooling Savant
Join Date: Jan 2002
Location: in a nice cool spot
Posts: 427
|
Quote:
I often have the work peice securley clamped to the drill plate and use one hand to feed the bit in and the other with the can of wd40 spraying into the hole as I drill. oh the drill needs to be correctly sharpened too.
__________________
feel free to icq/msn me, I'm always willing to toss around ideas. |
|
07-20-2002, 07:31 AM | #33 |
Cooling Savant
Join Date: Jun 2001
Location: Connecticut
Posts: 140
|
Chiller blocks require holes more than 2" deep. I have always used a drill press. Go very slow and used lots of lubricant, just like the post above states. When drilling very small holes the procedure above is also correct, withdraw the bit after every few seconds and clean the debris, use more lubricant and then start over. Good drill bits help as well. The $1.50 bits from Home Depot are not going to do to well for deep drilling.
I have learned this through trial and error mostly as I have snapped more than my share of bits. |
07-20-2002, 12:43 PM | #34 |
Hottest Stank of them All
Join Date: Mar 2002
Location: Washington, DC area
Posts: 533
|
thanx.
If you snap off a bit inside the hole, can you get it out? I'm guessing you couldn't drill it out...
__________________
Does a radioactive cat have 18 half lifes? --Kenny my pimpin' rig: ...previously poorly cooled... 486DX-2 66mhz @ 75mhz, 4Mb ram, shared, 256Mb hard drive Onboard VGA, watercooled, of course 16-color monitor, labtec speakers, 28kbs USRobotics modem Windows XP (think about it) Maze 3, DDen GF3 block, Eheim 1250, econoline van HC, 1/2" hose w/ 5/8" fittings Comair 172mm fan |
07-20-2002, 02:08 PM | #35 |
Cooling Savant
Join Date: Apr 2002
Posts: 110
|
If you snap a small bit (e.g 1.5mm) you will have just about trashed the thing your woking on.
Problem is the bit is so much harder then the Cu that if you try to drill it out, the carbide bit will just go sideways into the copepr , down the side of the broken HSS bit. As said - the higher quality the Bit the less chance of this happening. |
07-20-2002, 07:33 PM | #36 |
Cooling Savant
Join Date: Jun 2001
Location: Connecticut
Posts: 140
|
Yup, unless you get very lucky and the bit snaps off leaving enough of it sticking out to be able to grab it with a pliers or something similar, you are basically screwed.
Well this thread is depressing as hell. I now have to reconsider the way I built my chillerblocks as if myv65 is correct, my flow in the top few channels is just about nonexistant. They perform adequately but I am always looking for improvement. Back to the drawing board. |
07-20-2002, 07:40 PM | #37 |
Responsible for 2%
of all the posts here. Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
|
Here's a stupid question: why do you need a top row of holes?
I mean, you have your baseplate thickness, then you add to it to extend a row of cross-drilled holes, but why a second row? Why a third? Why not many? |
07-20-2002, 09:24 PM | #38 |
Big PlayerMaking Big Money
Join Date: Aug 2001
Location: irc.lostgeek.com #procooling.com
Posts: 4,782
|
Some interesting computer modeling here:
http://forum.oc-forums.com/vb/showth...0&pagenumber=4 Note the density of heat over the core. And throughout that thread much discussion of wb variables. ECU Pirate: I wouldn't consider prattling on without any technical leg to stand on to be theorizing; that is more like bullshitting. And I would suspect what is REALLY needed to improve wb design is expertise (the product of a theoretical background and experience). You will find it very hard to get help from those with expertise by behaving like a first grader. |
07-20-2002, 09:57 PM | #39 | |
Cooling Savant
Join Date: May 2002
Location: home
Posts: 365
|
Quote:
Don't despair. Theory will take me or anyone else only so far. It's more of a headstart in the right direction; it is by no means the destination. Ultimately nothing will beat real-life data (that is accurately obtained ). I think it may help to recognize that there are many approaches to designing blocks that will yield satisfactory results. The differences between the best blocks in terms of performance can be minor even though their approaches vary drastically. And as most folks know, some do better than others with low flow yet lose their advantage at higher flows. Of course, it's also possible to make a horrible block (and pretty darned easy I might add). Blessed are those with CNC machines at their disposal and the right test equipment to match. Us po' folks with drill presses (next on the tool shopping list) or less are really left at the mercy of block manufacturers or friends with better toys, er, tools. Anyway, don't sweat it. Make your blocks, test them, and try to make them better. I'd add a note to ECUPirate (and I missed that apparent exchange with BillA). More small holes doesn't mean less flow resistance. For a given total cross sectional area, you'll see the lowest flow restriction from a single round channel. You can approximate the effect of varying hole size and quantity with equations, but with the twists and turns in a typical setup it's generally nothing more than a good estimate. Qualitatively assessing a design is one thing, quantitatively assessing it is a whole 'nother ball game. |
|
07-20-2002, 10:49 PM | #40 |
Cooling Neophyte
Join Date: Nov 2001
Location: PA
Posts: 96
|
wow, tons of intresting ideas. i read only a few posts on the 1st page but nothing on the 2nd... i just wanted to see all the pics.
now ppl were talking about the maximum surface block idea with all the cross channels and how the channels furthest from the inlet/outlet would probably get the most flow... now i know this idea is pratically impossible but the channels that would be getting the lowest flow rate could be made curved as to make the direction of flow change minimal... hell even a 45 degree angle drilled in from both sides would be better.... know that a 90 degree elbow is the worst thing you can do for flow rate.... hell just make them all 45 degree angles if you really want to... just my .02
__________________
i am myself not anybody else and thats who i am and not who you are so dont be like me because i am myself. |
07-21-2002, 10:59 AM | #41 | |
Hottest Stank of them All
Join Date: Mar 2002
Location: Washington, DC area
Posts: 533
|
Quote:
__________________
Does a radioactive cat have 18 half lifes? --Kenny my pimpin' rig: ...previously poorly cooled... 486DX-2 66mhz @ 75mhz, 4Mb ram, shared, 256Mb hard drive Onboard VGA, watercooled, of course 16-color monitor, labtec speakers, 28kbs USRobotics modem Windows XP (think about it) Maze 3, DDen GF3 block, Eheim 1250, econoline van HC, 1/2" hose w/ 5/8" fittings Comair 172mm fan |
|
07-21-2002, 12:28 PM | #42 |
Cooling Savant
Join Date: May 2002
Location: home
Posts: 365
|
OK, the direct answer to your question is "yes". Adding holes will always lower resistance and increase flow. This will not, however, assure you better performance. Remember that convection is largely influenced by velocity. Increasing holes and lowering resistances will raise total flow rate, but reduce peak velocity.
|
07-22-2002, 06:05 AM | #43 |
Thermophile
Join Date: Oct 2001
Location: Nuu Zeeelin
Posts: 3,175
|
Just a couple of thoughts at 11pm:
We all recognise that we want as much effective surface area as possible. We also realise that having a 26 foot long 1/8" channel winding throughout the block will completely kill flow rates, but of course will provide plenty of flow rate. Also, if we have a 2" long 1/2" radius tube going through a block of copper we will have huge flow rate, but shit all surface area. Which is why I feel we shouldn't be concentrating on maze style designs (not to offend DDen who make great blocks) but on blocks that have multiple paths. This is partially why most of my designs have 2 outlets and one inlet. The reason for the one inlet is because I feel a centre inlet will slightly decrease core temps. The reason for only 2 outlets is due to space, I think it is impracticle to have more than 3 or 4 barbs in total on a block, due to size restrictions of the block itself. But just because we have one inlet and one or two outlets doesn't mean we have to have just two channels. We can use just about as many channels as we want to get from the inlet to the outlet. And of course we have to consider the fact we can only fill up 80mm x 80mm of total space, and have to incorporate 4 holes in the design, at the dimensions shown here http://www.dansdata.com/images/coolercomp/holes500.gif. It is very hard to acheive all this in an effective design
__________________
2x P3 1100's at 1400, Abit VP6, 2x Corsair 256mb PC150 sticks, 20gb 'cuda ATA-III, 2x 40gb 'cuda ATA-IV in raid 0. 20" Trinitron. No fans 2x 2400+ at 2288mhz (16.0 x 143), Iwill MPX2, 2x Kingmax PC-3200 256mb sticks, 4x 20gb 60gxp in Raid 5 on a Promise SX6000. Asus Ti4200 320/630. Cooled by Water |
07-22-2002, 09:34 AM | #44 |
Responsible for 2%
of all the posts here. Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
|
Following this bit of info on fins (page 2, thx to myv65), I'm revising the theoretical design considerations, and adding it as a new (optional) item. Optional since some of us may want to forego fins.
So we have: 1- flow restriction vs pressure and turbulence 2- flow path length vs backpressure 3- base plate thickness vs heat spreading / resistance / capacity 4- Channel pathway for optimal cooling (crossflow ?) vs flow path length and restriction 5- Channel shape vs liquid properties of water (or wiper fluid) 6- Cross sectional material print 7- Flow restriction versus fin design (where applicable) Thanks to RedTalon, we covered that in #4 and 5. Thanks to ECUPirate, and for myv65's correct reply. Brad: I agree that more area is better, but I think that this essentially reaches a point where more holes don't add any cooling whatsoever, it only reduces the flow restriction, like you said. Since we know that we need to achieve a high flow speed to maximize cooling, then I think that we need to forget about a second or third row of cross-holes, because it defeats that purpose. Further, since the flow would be at a relatively high speed, then there must be a limit of how deep the water needs to be, in relation to the surface that is being cooled. Example, if we had a single channel, and the water was flowing at X speed, where the channel is 0.3" deep, versus a channel that is say 3 inches deep, then obviously the latter is a complete waste of pump power, and so on. Brad, I think that you're right, in that a 2 channel design is probably good for what you are doing, but as I just stated, unless you make your block thinner, and your channels shallower, then all that you are doing is reducing the flow speed of the coolant, which is bad. pHaestus, that's an amazing link. It clearly illustrates the dynamics of heat transfer, as it comes in the block from the CPU's core. I think that it is critical that it be considered, and that it would be a waste (as above) to allow coolant flow to end up cooling an already cold part of a block. To all: So how can we determine the depth of the coolant required as a function of its speed? |
07-22-2002, 09:49 AM | #45 | |
Responsible for 2%
of all the posts here. Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
|
Quote:
I'm not so much worried about the mounting holes right now, but otherwise, space is a limitation: the block must fit within a specific area. Given that the heat comes from the core, I don't see any reason why the block should exceed the CPU print. How do you figure 80mm by 80mm? |
|
07-22-2002, 03:06 PM | #46 |
Responsible for 2%
of all the posts here. Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
|
(Just so that it doesn't get lost, from 2 posts above)
How can we determine the depth of the coolant required as a function of its speed? |
07-22-2002, 04:42 PM | #47 |
Responsible for 2%
of all the posts here. Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
|
I may have an answer...
[edit] :removed link because it turned out to be about something else entirely [/edit] This is where I'm going to point out that three rows of cross-drilled channels may be a waste, simply because the heat dispersion of the first row makes the other ones redundant, or at least, infinitesimally insignificant. Last edited by bigben2k; 07-22-2002 at 06:49 PM. |
07-22-2002, 08:05 PM | #48 |
Responsible for 2%
of all the posts here. Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
|
I got a new link:
http://webbook.nist.gov/chemistry/fluid/ It doesn't cover what I'm looking for, but hey, it's a start! |
07-23-2002, 04:38 PM | #49 |
Responsible for 2%
of all the posts here. Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
|
Hum... I have a partial answer here
In short, turbulent flow is achieve by either (1) a high flow speed, or (2) induced turbulence. This will dictate the optimal height of a channel. |
07-24-2002, 02:54 AM | #50 |
Thermophile
Join Date: Oct 2001
Location: Nuu Zeeelin
Posts: 3,175
|
well, the Alpha 8942 takes up 90mm x 80mm, so maybe thats the max size for P4's, but most athlon boards don't allow anything above 80mm x 80mm, although my iwill mpx2 allows for 2x 70mm x 100mm waterblocks.
As for the shallower channels, thats just increasing back pressure. (just one side of the story, you present the other side of the story)
__________________
2x P3 1100's at 1400, Abit VP6, 2x Corsair 256mb PC150 sticks, 20gb 'cuda ATA-III, 2x 40gb 'cuda ATA-IV in raid 0. 20" Trinitron. No fans 2x 2400+ at 2288mhz (16.0 x 143), Iwill MPX2, 2x Kingmax PC-3200 256mb sticks, 4x 20gb 60gxp in Raid 5 on a Promise SX6000. Asus Ti4200 320/630. Cooled by Water |
Currently Active Users Viewing This Thread: 1 (0 members and 1 guests) | |
|
|