Go Back   Pro/Forums > ProCooling Technical Discussions > General Liquid/Water Cooling Discussion > Water Block Design / Construction
Password
Register FAQ Members List Calendar JavaChat Mark Forums Read

Water Block Design / Construction Building your own block? Need info on designing one? Heres where to do it

Reply
Thread Tools
Unread 12-08-2003, 08:41 AM   #1
grvydude18
Cooling Neophyte
 
Join Date: Dec 2003
Location: virginia
Posts: 17
Default Physics behind the cascade??

K, i have been reading many of the posts about the cascade, I still haven't really found much about the actual physics, as to why its better to jet it at the thin part? why does this help with heat absorbtion? if someone could explain it it would be helpful. Thanks a lot.
grvydude18 is offline   Reply With Quote
Unread 12-08-2003, 11:03 AM   #2
Aleck
Cooling Neophyte
 
Join Date: Nov 2003
Location: Malaysia
Posts: 29
Default

The thinner the material that the heat has to pass the faster it gets transferred

So thinner is better up to a point where the copper could not stand the clamping force of the wb onto the CPU die where it starts to buckle

Why are there some thick parts..?

This is so that the protruding nozzles could be submerged into the parts which are thinner (in this case dimples) to avoid the effect of crossflow or disruption of the jets due to the spent coolant trying to find its way out.
Aleck is offline   Reply With Quote
Unread 12-08-2003, 11:16 AM   #3
grvydude18
Cooling Neophyte
 
Join Date: Dec 2003
Location: virginia
Posts: 17
Default

but why jet it in? i understand the thinner stuff... but i don't understand jetting it on there? they made it sound like there was more then just lots of water flowing quickly, is there? or is it just another way to do it?
grvydude18 is offline   Reply With Quote
Unread 12-08-2003, 11:30 AM   #4
jaydee
Put up or Shut Up
 
jaydee's Avatar
 
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
Default

The jets create an impingment effect. If it is contained in a cylinder as the Cascade then the effect is improved as the water hits the bottom of the cylinder and then travels up the sides of the cylinder for a secondary cooling effect. The impingment can't be optimal unless it is in a cylinder with a flat bottom. It needs to be a FLAT bottom aswell for maximum impingment effect.

The jets need to go inside these cylinders to keep the water directed at it's target. If it is allowed to be raised above the hole then the exiting water will interfear with the jets. Find Cathars thread on the Cascade in the water block design section here.
jaydee is offline   Reply With Quote
Unread 12-08-2003, 12:13 PM   #5
bigben2k
Responsible for 2%
of all the posts here.
 
bigben2k's Avatar
 
Join Date: May 2002
Location: Texas, U.S.A.
Posts: 8,302
Default

When water flows inside of a pipe, there's a portion of it that flows right through, then there's a portion of it that just sticks to the pipe wall: that's called the boundary region.

Now you can try to pump a coolant faster, and minimize this boundary layer, but it's still going to be there, and you'll waste a lot of pumping power to minimize it.

So when you shoot the water, you're essentially "attacking" the boundary layer straight on, with minimal flow.


The concept behind Cascade is detailed here.
bigben2k is offline   Reply With Quote
Unread 12-08-2003, 12:51 PM   #6
Teus
Cooling Neophyte
 
Join Date: Jun 2003
Location: Belgium\Ieper
Posts: 65
Default

Quote:
Originally posted by bigben2k
The concept behind Cascade is detailed here.
ah, nice. so that means if you make a #rotor block, the jets need to be right above the drilled holes.

the holes in my CPU blocks are 3mm wide and I connect them with my dremel. I don't use cilinders like Cathar does, only a copperplate. what size should the jets above the drilled holes be? 3mm, 2mm, or four 1mm holes?

mod: after I've kept reading that thread, I'll definately use cilinders. prolly 1mm cilinders for 2mm holes

Last edited by Teus; 12-08-2003 at 01:08 PM.
Teus is offline   Reply With Quote
Unread 12-08-2003, 02:49 PM   #7
Khledar
Cooling Savant
 
Khledar's Avatar
 
Join Date: Mar 2002
Location: North Vancouver BC
Posts: 234
Default

#rotor blocks are quite different than cascades, they do not use impingment primarily; they are more of a pin style cooling methodology, with the twist of being easily manufactured (not really but once you get the hang of it it's not so labourious ). Cascade blocks have nothing to do with pin style cooling - the main portion of cooling occurs as the water hits the bottom of the holes. A pin style block absorbs the heat from the pins as the water travels through. Pin blocks benefit from a central inlet, and possibly from some sort of jets but the two designs are not the same in cooling theory.

Correct me if I'm wrong, but I think you're getting block styles a bit confused.
__________________
"mooooo" said the cow.
ERTW - UBC

P4 2.4B @3.01Ghz 167FSB :: Abit IC7 :: 2x256MB HyperX PC3700 :: ASUS 9600XT :: WD Raptor 2x36GB RAID0
Khledar is offline   Reply With Quote
Unread 12-08-2003, 02:55 PM   #8
Teus
Cooling Neophyte
 
Join Date: Jun 2003
Location: Belgium\Ieper
Posts: 65
Default

hmmm, I get it. but that doesn't mean both kinds of waterblocks can be combined?

I'm in my exams so I've got some time to think about that. why couldn't I use an impingment on the die of the CPU, and a #rotor block on the sides of it?
Teus is offline   Reply With Quote
Unread 12-08-2003, 02:56 PM   #9
grvydude18
Cooling Neophyte
 
Join Date: Dec 2003
Location: virginia
Posts: 17
Default

the impingment effect? what is that? i guess if someone can answer that im done with questions on this topic but yeah, i understand the jetting in the hole now, it makes a lot of sense, i the impingment effect is the jetting to attack the boundary layer? just say so and im done, thanks all a lot.
grvydude18 is offline   Reply With Quote
Unread 12-08-2003, 03:03 PM   #10
jaydee
Put up or Shut Up
 
jaydee's Avatar
 
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
Default

Quote:
Originally posted by grvydude18
the impingment effect? what is that? i guess if someone can answer that im done with questions on this topic but yeah, i understand the jetting in the hole now, it makes a lot of sense, i the impingment effect is the jetting to attack the boundary layer? just say so and im done, thanks all a lot.
The water hitting the bottom of the hole is the impingment effect. Like hitting a peice of wood with a hammer or running a car into a cement wall.
jaydee is offline   Reply With Quote
Unread 12-08-2003, 03:33 PM   #11
grvydude18
Cooling Neophyte
 
Join Date: Dec 2003
Location: virginia
Posts: 17
Default

but is the impingent effect the actual attacking of that boundary layer? or does it pull up other parts of physics i haven't yet learned in class? thanks to all thats helped so far, and hopefully im not the only one learning from this.
grvydude18 is offline   Reply With Quote
Unread 12-08-2003, 04:21 PM   #12
jaydee
Put up or Shut Up
 
jaydee's Avatar
 
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
Default

Quote:
Originally posted by grvydude18
but is the impingent effect the actual attacking of that boundary layer?
Yes.
jaydee is offline   Reply With Quote
Unread 12-08-2003, 05:32 PM   #13
joemac
Cooling Savant
 
joemac's Avatar
 
Join Date: Dec 2002
Location: Dallas
Posts: 339
Default

Quote:
Originally posted by Teus
hmmm, I get it. but that doesn't mean both kinds of waterblocks can be combined?

I'm in my exams so I've got some time to think about that. why couldn't I use an impingment on the die of the CPU, and a #rotor block on the sides of it?
We’ve played with the concept with mixed results. To drive the water toward the bottom of the block our design uses the pins as a sort of guide. Its not as good as a tube carrying the water and releasing it right above the core but it works to an extent. One of our early designed had the pins around the core that extended to the top of the inlet and the rest were half the size of the height of the inner chamber. This was good for flow (made the block less restrictive) but concerns over the boundary layer won out. The Pins all around the block help to minimize the boundary effect throughout the block by creating turbulence . This of course is not necessary for above core cooling were the only area of concern is above the core.

Edit spelling:
__________________
www.aquajoe.com
joemac is offline   Reply With Quote
Unread 12-08-2003, 05:39 PM   #14
jaydee
Put up or Shut Up
 
jaydee's Avatar
 
Join Date: Dec 2001
Location: Spokane WA
Posts: 6,506
Default

Quote:
Originally posted by Teus
hmmm, I get it. but that doesn't mean both kinds of waterblocks can be combined?

I'm in my exams so I've got some time to think about that. why couldn't I use an impingment on the die of the CPU, and a #rotor block on the sides of it?
I have used impingment on my Direct Die cooling project and it doesn't work to well. It needs Copper to spread the heat first.

I don;t think combing a Rotor block with jets will do any good. Wouldn't be much different than just using a center inlet alone without any jets. :shrug:
jaydee is offline   Reply With Quote
Unread 12-08-2003, 06:14 PM   #15
prb123
Cooling Neophyte
 
Join Date: May 2002
Location: Michigan
Posts: 51
Default

Besides a Rotor block is very restrictive as is, if you add Jets to it your flow rate is going to be terrible. The small slits cause turbulence, and the design works well as is. #Rotor states that adding a center inlet does not help.
prb123 is offline   Reply With Quote
Unread 12-08-2003, 06:41 PM   #16
Cathar
Thermophile
 
Cathar's Avatar
 
Join Date: Sep 2002
Location: Melbourne, Australia
Posts: 2,538
Default

A number of things going on here that's different between #Rotor (pin-fin) and Cascade.

First and foremost there's the issue of the base-plate thickness. The Cascade uses a very thin base-plate, but is able to get away with doing so due to the honey-comb nature of the cup pattern. The ensures that the base-plate will not flex under normal circumstances. A pin-fin by nature needs a thicker base-plate as there's no inter-wall bracing between the pins.

A thicker base-plate means a greater cost in terms of a thermal gradient through the copper, but is also offset by better thermal spread through the copper.

Where it gets interesting is as the jetting is introduced, the impingement efficacy of the jets is such that the heat doesn't need to spread as far before it gets soaked up effectively by the water. So basically we get left with paying the cost of moving the heat through the thicker copper but reaping none of the rewards for improved heat spread.

This is exactly why thicker base-plates tend to be better at lower flow rates, but we see their cooling performance taper off rapidly at higher flow rates, whereas the thinner base-plate designs just keep on getting better. The challenge here of course is to make a thin base-plate design that performs at least as well as a thick base-plate design at low flow rates.

We also need to look at the flow. A pin-fin block is basically a lateral flow block, meaning the water flows sideways through the pins. This sideways flow interrupts the efficiency of jets and also pushes them sideways. Basically what you end up with is an effect that's not really different to just having a single inlet of water mashing down. The Cascade with its injection tubes and cups ensures that each jet region does its job effectively and independently of any other jet, and that the outwash doesn't turn the lot in a just single big mash water flowing down. The Cascade is an array of in-out injection points, with no injection point interfering with any other one.
Cathar is offline   Reply With Quote
Unread 12-08-2003, 07:27 PM   #17
joemac
Cooling Savant
 
joemac's Avatar
 
Join Date: Dec 2002
Location: Dallas
Posts: 339
Default

Quote:
Originally posted by Cathar
This is exactly why thicker base-plates tend to be better at lower flow rates, but we see their cooling performance taper off rapidly at higher flow rates, whereas the thinner base-plate designs just keep on getting better.
True, Performance increases as power input increases. That increase in performance does come with a price e.g. stronger pump, which in many cases requires more power. I am not saying that is a bad thing or that it is anything new. Its no different then adding a bigger pelt to a pelt system – sure it works but it comes with a price.
__________________
www.aquajoe.com
joemac is offline   Reply With Quote
Unread 12-08-2003, 08:50 PM   #18
Cathar
Thermophile
 
Cathar's Avatar
 
Join Date: Sep 2002
Location: Melbourne, Australia
Posts: 2,538
Default

Quote:
Originally posted by joemac
True, Performance increases as power input increases. That increase in performance does come with a price e.g. stronger pump, which in many cases requires more power. I am not saying that is a bad thing or that it is anything new. Its no different then adding a bigger pelt to a pelt system – sure it works but it comes with a price.
Hmmm, perhaps you missed the second part to the sentence, which is why it was there:

Quote:
The challenge here of course is to make a thin base-plate design that performs at least as well as a thick base-plate design at low flow rates.
Achieve that, then people can achieve good performance at low flow rates/weaker pumps, but those with stronger pumps can reap the benefits of having the stronger pumps in the first place.
Cathar is offline   Reply With Quote
Unread 12-08-2003, 09:43 PM   #19
joemac
Cooling Savant
 
joemac's Avatar
 
Join Date: Dec 2002
Location: Dallas
Posts: 339
Default

Quote:
Originally posted by Cathar
Hmmm, perhaps you missed the second part to the sentence, which is why it was there:



Achieve that, then people can achieve good performance at low flow rates/weaker pumps, but those with stronger pumps can reap the benefits of having the stronger pumps in the first place.
No – I did not miss that, I just pointed out that with current water block designs you need more to get more
__________________
www.aquajoe.com
joemac is offline   Reply With Quote
Unread 12-08-2003, 11:00 PM   #20
Cathar
Thermophile
 
Cathar's Avatar
 
Join Date: Sep 2002
Location: Melbourne, Australia
Posts: 2,538
Default

Quote:
Originally posted by joemac
No – I did not miss that, I just pointed out that with current water block designs you need more to get more
Sorry - you lost me on that one. What are you trying to say? Are you saying that it's different somehow for other blocks?

Last edited by Cathar; 12-08-2003 at 11:23 PM.
Cathar is offline   Reply With Quote
Unread 12-08-2003, 11:45 PM   #21
joemac
Cooling Savant
 
joemac's Avatar
 
Join Date: Dec 2002
Location: Dallas
Posts: 339
Default

Quote:
Originally posted by Cathar
Sorry - you lost me on that one. What are you trying to say? Are you saying that it's different somehow for other blocks?
No - :shrug:
__________________
www.aquajoe.com
joemac is offline   Reply With Quote
Unread 12-09-2003, 08:49 PM   #22
rocketmanx
Cooling Savant
 
rocketmanx's Avatar
 
Join Date: Nov 2003
Location: Indiana
Posts: 127
Default

Here's a pretty good explenation of jet impingment cooling- Vision Engineering

The boundary layer is a ctually at the area of impact. The boundary layer as I've read about it and understand, is a layer at the point of contact between the medium to be cooled (copper in this case) and the cooling medim- water. There's a layer at the surface where the water and copper aren't actually "touching" each other due to a very fine layer of the "boundary" between the 2. It's kinda like "approaches zero but never equals zero". The impact thins this boundary layer through impact or "impinging" on the boundary layer, in other words it forces it's way through the layer a little and helps cooling.
I think of it as, if 2 pieces of perfectly flat copper were pressed together very hard the 2 pieces would still have a very tiny "boundary" layer wetween them because they aren't atomicaly bonded like a solid piece of copper. The same with the water running over a piece of copper, but worse. The force of the impingement jet basically gets the water closer to the copper. We all know that the more layers heat/energy has to travel through the more get lost in the transfer from one layer to the other.

Maybe a bit to long an expleantion there but I was very interested in finding out why Cathar's block worked so well. So I did a little research
If I'm off in my understanding please correct me as I could only find bits here and there.

Last edited by rocketmanx; 12-09-2003 at 08:55 PM.
rocketmanx is offline   Reply With Quote
Unread 12-11-2003, 02:29 AM   #23
Aleck
Cooling Neophyte
 
Join Date: Nov 2003
Location: Malaysia
Posts: 29
Default

I think you got it wrong there rocketmanx...

Boundary layer is actualy a thin layer of fluid that slows down on the surface of the solid that the fluid is flowing on.

The slowing down of this fluid is due to the adhering of the fluid molecules nearest to the surface of the solid due to the intermolucular forces.

This boundary layer can be reduced by introducing turbulance..
Aleck is offline   Reply With Quote
Unread 12-11-2003, 05:50 AM   #24
satanicoo
Cooling Savant
 
satanicoo's Avatar
 
Join Date: Nov 2002
Location: portugal
Posts: 635
Default

Quote:
Originally posted by Aleck
I think you got it wrong there rocketmanx...

Boundary layer is actualy a thin layer of fluid that slows down on the surface of the solid that the fluid is flowing on.

The slowing down of this fluid is due to the adhering of the fluid molecules nearest to the surface of the solid due to the intermolucular forces.

This boundary layer can be reduced by introducing turbulance..
And the impingement is the best kind of turbulance to break the boundary layer. It simply breaks it and "new water" (fresher) comes in contact with copper.
satanicoo is offline   Reply With Quote
Unread 12-11-2003, 06:45 AM   #25
Cathar
Thermophile
 
Cathar's Avatar
 
Join Date: Sep 2002
Location: Melbourne, Australia
Posts: 2,538
Default

Quote:
Originally posted by satanicoo
And the impingement is the best kind of turbulance to break the boundary layer. It simply breaks it and "new water" (fresher) comes in contact with copper.
Agreed. Although I dont really view impingement as turbulence. Impingement doesn't necessarily require turbulent flow for it to be effective.

Impingement is costly though in terms of pressure drop, but there is no free lunch here.

A further part of the Cascade's impingement effect is the creation of turbulence in the incoming jet, at the cost of some of the jet's power. When the jet rushes down into the cup, it shears against the water coming back up out of the cup. In this fashion the Cascade uses the cups to both create a dual impingement effect at the middle base of the cup, and the cup walls, as well as churn the water up to more effectively scavenge any heat that makes it up the cup walls.

Again, I am struggling to find the exact link to the paper that I read, but basically it was showing that the effect of extra turbulence in the jet stream by standing it off from the base and having it shear against the water around it was actually more effective at stripping the boundary layers despite the force of the jet's effect being slightly reduced. I believe that I've been able to mimic and observe the paper's statements through experimentation myself. The paper was talking about simple submerged jet behavior and not really talking about jet-in-a-cup behavior which is something else that I had to explore.

So basically if the jet is too close it doesn't gather any turbulence as it descends. The further away the jet is, the more turbulent the jet becomes but the more power the jet loses. There is a cross-over point between the two effects where as the jet is moved yet further away, the added turbulence is no longer enough to offset the loss in the jet power. The actual ideal distance is related to the jet diameter and the jet velocity, with lower velocities achieving better performance with closer jets, and higher velocities with more distant jets. Picking a design point amongst all that mess of variables is the challenge.

Anyway, the end summary is that it's all about impingement and turbulence and attempting to find the right balance in the design to make the most of the "typical" pumping pressure that gets applied. It's about stripping away the boundary layer most effectively. No matter what you do, the boundary layer is always there, and the best we can hope for is to reduce it as much as possible across as much of the surface area that's closest to the CPU heat as possible.

There is danger in going too extreme though in that in much the same way that water shapes mountains and valleys out of granite for the Earth, metal erosion can become a factor.
Cathar is offline   Reply With Quote
Reply


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

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -5. The time now is 09:57 AM.


Powered by vBulletin® Version 3.7.4
Copyright ©2000 - 2024, Jelsoft Enterprises Ltd.
(C) 2005 ProCooling.com
If we in some way offend you, insult you or your people, screw your mom, beat up your dad, or poop on your porch... we're sorry... we were probably really drunk...
Oh and dont steal our content bitches! Don't give us a reason to pee in your open car window this summer...