Block Desgin

Hixup · 06-21-2002, 08:46 PM

Hey, i was wondering if this would be a good desgin to cool a peltier. The in would be 1 tube split to maximize cooling by running more cold water over the pelt. it would be layed out like this:
_____
| Water Block
|
|-------
[ Copper Plate
[ Pelt
[ Chip
Heres the design.

Cyco-Dude · 06-21-2002, 08:59 PM

waterblock
pelt
cold plate
CPU

also, youll want to CLAMP the pelt between the waterblock and cold plate, as the pelt requires a lot of pressure to operate optimally.

Brad · 06-21-2002, 11:41 PM

the waterblock design is good, and very simple. Cyco dude has the right order for you

JimS · 06-22-2002, 06:38 PM

If there was some way to eliminate the 180 degree turns in the block, you could improve flow considerably.

crafty · 06-22-2002, 06:43 PM

Quote:

Originally posted by JimS
If there was some way to eliminate the 180 degree turns in the block, you could improve flow considerably.

Maybe trying to make it spiral - goes in on the outside spirals to the middle and goes out the middle.

#Rotor · 06-22-2002, 09:40 PM

with the utmost respect to those that I'm disagreeing with on this one....but don't try and design a block to give good flow, that inherently will mean that you need to get rig of all the turbulence in the flow pattern.....Rather design the block to give maximum turbulence and make the pump worry about flow rate... [PTCBTH]

Hixup · 06-22-2002, 09:42 PM

Ive done a redo, and thought about this more. The cold water will ender on the hottest part of the chip, the center and exit trough heach end. i tryied make make it flow a little bit better then before

bigben2k · 06-23-2002, 09:59 AM

Quote:

Originally posted by #Rotor
with the utmost respect to those that I'm disagreeing with on this one....but don't try and design a block to give good flow, that inherently will mean that you need to get rig of all the turbulence in the flow pattern.....Rather design the block to give maximum turbulence and make the pump worry about flow rate... [PTCBTH]

Agreed, 100% !!!

Block design shouldn't depend on the pump power.

I read somewhere a theory bit that says that turbulating can improve cooling efficiency by 100 to 500% (in theory of course!). Throughout everything we've all read, it's clear that a high flow rate will take heat out of the block better.

The problem, and the ultimate consideration, is how to get rid of that heat since the rad does absolutely nothing until there is a difference in temperature between the ambiant air, and the coolant.

I'm at a point where I'm considering testing with multiple heater cores running in parallel. The car rad is also starting to look good...

dcrainman · 06-23-2002, 12:44 PM

Increased turbulence = better cooling

Also I think I read an article somewhere about centered inputs not being the best way to design the block. I'm making one now thats just an old fashion W.

#Rotor · 06-23-2002, 12:53 PM

do not be fooled by thinking that it is high flow rate, that makes the block perform better.... high flow rate only induces more turbulence in the block, and that, the induced turbulence, is what really causes the block to perform. if you have a design that is absolutely smooth and that personifies laminar flow in a manner that would make them hydro engineers envy you, you will have astronomical flow-rates to brag with, but that block will utterly suck at removing heat. I say design a block with one thing in mind. That of moving heat into the water as efficiently as possible. and the only way to do that, is with lots and lots of turbulence. This is where science actually has 2 positive aspects complementing another, by designing the interior of a flowpath to induce turbulence, one invariably ends up with added surface area, For instance just adding a little dimple to a smooth surface, will also add a little but more surface area, but the turb. being caused in the wake of the dimple more than doubles the cooling effect of the surface behind it...

Now turbulence has the side effect of appearing as back-pressure to the pump, this is not always a bad thing, back pressure will slow down the overall flow rate of a rig, and if the blocks are responsible for this, then you are doing good, because it is in the blocks, that you need the flow velocity of the fluid to be transformed into turbulence, ending in superior heat transfer between block and fluid. This overall reduced flow will enable the radiators to also perform marginally better.

Hixup · 06-23-2002, 03:34 PM

i have heard that the maze 1 works beter then say, a base 2.

06-21-2002, 08:46 PM	#1
Hixup Cooling Neophyte Join Date: Jun 2002 Location: By the computer Posts: 51	Block Desgin Hey, i was wondering if this would be a good desgin to cool a peltier. The in would be 1 tube split to maximize cooling by running more cold water over the pelt. it would be layed out like this: _____ \| Water Block \| \|------- [ Copper Plate [ Pelt [ Chip Heres the design. __________________ AMD Athlon XP 1900+ Water Cooled Soyo SY-K7V Dragon plus! 512 DDR radoen 9500 Soft Mod Give Me Money

06-21-2002, 11:41 PM	#3
Brad Thermophile Join Date: Oct 2001 Location: Nuu Zeeelin Posts: 3,175	the waterblock design is good, and very simple. Cyco dude has the right order for you __________________ 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

06-22-2002, 09:42 PM	#7
Hixup Cooling Neophyte Join Date: Jun 2002 Location: By the computer Posts: 51	Ive done a redo, and thought about this more. The cold water will ender on the hottest part of the chip, the center and exit trough heach end. i tryied make make it flow a little bit better then before __________________ AMD Athlon XP 1900+ Water Cooled Soyo SY-K7V Dragon plus! 512 DDR radoen 9500 Soft Mod Give Me Money

06-23-2002, 03:34 PM	#11
Hixup Cooling Neophyte Join Date: Jun 2002 Location: By the computer Posts: 51	i have heard that the maze 1 works beter then say, a base 2. __________________ AMD Athlon XP 1900+ Water Cooled Soyo SY-K7V Dragon plus! 512 DDR radoen 9500 Soft Mod Give Me Money

06-21-2002, 08:59 PM	#2
Cyco-Dude Cooling Savant Join Date: Feb 2002 Location: Dayton, Ohio Posts: 836	waterblock pelt cold plate CPU also, youll want to CLAMP the pelt between the waterblock and cold plate, as the pelt requires a lot of pressure to operate optimally.

06-22-2002, 06:38 PM	#4
JimS Cooling Savant Join Date: Jun 2001 Location: Connecticut Posts: 140	If there was some way to eliminate the 180 degree turns in the block, you could improve flow considerably.

06-22-2002, 09:40 PM	#6
#Rotor Cooling Savant Join Date: Feb 2002 Location: Dione, sector 4s1256 Posts: 852	with the utmost respect to those that I'm disagreeing with on this one....but don't try and design a block to give good flow, that inherently will mean that you need to get rig of all the turbulence in the flow pattern.....Rather design the block to give maximum turbulence and make the pump worry about flow rate... [PTCBTH]

06-23-2002, 12:44 PM	#9
dcrainman Cooling Neophyte Join Date: Jun 2002 Location: ma, usa Posts: 50	Increased turbulence = better cooling Also I think I read an article somewhere about centered inputs not being the best way to design the block. I'm making one now thats just an old fashion W.

06-23-2002, 12:53 PM	#10
#Rotor Cooling Savant Join Date: Feb 2002 Location: Dione, sector 4s1256 Posts: 852	do not be fooled by thinking that it is high flow rate, that makes the block perform better.... high flow rate only induces more turbulence in the block, and that, the induced turbulence, is what really causes the block to perform. if you have a design that is absolutely smooth and that personifies laminar flow in a manner that would make them hydro engineers envy you, you will have astronomical flow-rates to brag with, but that block will utterly suck at removing heat. I say design a block with one thing in mind. That of moving heat into the water as efficiently as possible. and the only way to do that, is with lots and lots of turbulence. This is where science actually has 2 positive aspects complementing another, by designing the interior of a flowpath to induce turbulence, one invariably ends up with added surface area, For instance just adding a little dimple to a smooth surface, will also add a little but more surface area, but the turb. being caused in the wake of the dimple more than doubles the cooling effect of the surface behind it... Now turbulence has the side effect of appearing as back-pressure to the pump, this is not always a bad thing, back pressure will slow down the overall flow rate of a rig, and if the blocks are responsible for this, then you are doing good, because it is in the blocks, that you need the flow velocity of the fluid to be transformed into turbulence, ending in superior heat transfer between block and fluid. This overall reduced flow will enable the radiators to also perform marginally better.

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