Impractical idea

FooTemps · 03-02-2005, 03:13 AM

I figured this idea would be highly impractical and costly but it's been bugging me lately...

Heatpipes + phase change = full system cooling using one off the shelf cpu cooling system (like prometia or vapochill)

Since Zalman already has a fully passive heatpipe cooled case someone could go the extra step and cool the whole system to low temps using a heatpipe system that junctions at the cold plate of the cooling unit.

Not sure about the effectiveness of the idea though since the heatpipes would have to be fairly long and that is a lot of heat going to one cooling unit. But since the compressor units can already achieve below freezing temperatures on just the cpu the combined heat of the system should be managable. In the end it may just be as effective as good watercooling but I'm just tossing around a crazy idea.

It also wouldn't be very cost effective for us DIYers since you'd have to basically buy and bend your own heatpipes for the whole unit and that may result in buying anywhere from 10 heatpipes for effective cooling on larger units (the number varies due to number of hard drives and how many pipes you want to use for the cpu, gpu, psu)

Long Haired Git · 03-02-2005, 03:53 AM

Given the tech available for download of specs, the Delta-T at the evaporator and condensors are too high to beat a good watercooling solution.
Heat pipes are awesome because they are maintenance free, no working parts, no power input heat pumps. They, like watercooling, can take heat from a tiny source and spread it over a large heat sink.
Air cooling does this, predominately, using the thermal conductivity of a material - typically aluminium. It suffers very little interfacing losses - basically one at the TIM and one to air, but suffers a high Delta-T due to the resistance of even good conductors like copper.
Water cooling does this, predominately, using a high-thermal-capacity fluid with mechanical transportation. It gains due to the low Delta-T at the water block and fluid transport, but loses due to mechanical nature and extra interfaces between water and waterblock and water and radiator.
Heat pipes have the same interfaces, but don't have any mechanical aides (jet impingement etc) at the waterblock side.

So, in short, for the application you describe, you're better off using a coolant.

Next, finding a working fluid for what you describe is hard. It has to boil at the temp you want to keep the "block" (evaporator) at, and condense at the cold end temp, and there needs to be a decent Delta-T between the two to generate the energy to get the liquid/gas to move in velocities and volumes to provide the transportation of the energy at the same rate as it arrives from the CPU (and environment, as you'll be sub delta).

Also, bear in mind that outside the working temp of the fluid, a heatpipe stops pumping heat. They fail not through mechanical breakdown, but in that the conditions exceed their parameters...

Da_Priest · 03-08-2005, 06:49 AM

you'd be better of making a chiller then.
You have to make blocks for your graka and cpu and stuff to connect your heatpipes to anyway, so you can makewaterblocks too.
instaid of hooking up you heatpipes to a cold plate, put your cold plate in a reservoir and run you 'water' trough it ...
You'll achieve much better temps than with heatpipes, the amount of time and money to put into it would be about the same prolly.

yet you'll still need a powerfull compressor. A modified vapochill or prometia won't handle the charge of a whole system... I think...

anyway, it's a crazy idea indeed

grtz

bigben2k · 03-08-2005, 08:40 PM

The heatpipe just adds an extra layer to the simple solution of "phase change" alone, which would perform slightly less efficiently.

It'd be a challenging exercise otherwise.

HAL-9000 · 03-08-2005, 09:09 PM

Sounds like a interesting idea. The heatpipe chemical would have to be different though wouldn't it? The idea I get from the literature is they work within very specific temperature ranges. Would the heatpipe fluid have a risk of freezing when it contacted the cold plate?

bigben2k · 03-08-2005, 10:16 PM

Exactly, hence the exercise.

03-02-2005, 03:13 AM	#1
FooTemps Cooling Neophyte Join Date: Jan 2005 Location: WA, USA Posts: 49	Impractical idea I figured this idea would be highly impractical and costly but it's been bugging me lately... Heatpipes + phase change = full system cooling using one off the shelf cpu cooling system (like prometia or vapochill) Since Zalman already has a fully passive heatpipe cooled case someone could go the extra step and cool the whole system to low temps using a heatpipe system that junctions at the cold plate of the cooling unit. Not sure about the effectiveness of the idea though since the heatpipes would have to be fairly long and that is a lot of heat going to one cooling unit. But since the compressor units can already achieve below freezing temperatures on just the cpu the combined heat of the system should be managable. In the end it may just be as effective as good watercooling but I'm just tossing around a crazy idea. It also wouldn't be very cost effective for us DIYers since you'd have to basically buy and bend your own heatpipes for the whole unit and that may result in buying anywhere from 10 heatpipes for effective cooling on larger units (the number varies due to number of hard drives and how many pipes you want to use for the cpu, gpu, psu)

03-02-2005, 03:53 AM	#2
Long Haired Git Cooling Savant Join Date: Jan 2003 Location: Sydney, Oz Posts: 336	Given the tech available for download of specs, the Delta-T at the evaporator and condensors are too high to beat a good watercooling solution. Heat pipes are awesome because they are maintenance free, no working parts, no power input heat pumps. They, like watercooling, can take heat from a tiny source and spread it over a large heat sink. Air cooling does this, predominately, using the thermal conductivity of a material - typically aluminium. It suffers very little interfacing losses - basically one at the TIM and one to air, but suffers a high Delta-T due to the resistance of even good conductors like copper. Water cooling does this, predominately, using a high-thermal-capacity fluid with mechanical transportation. It gains due to the low Delta-T at the water block and fluid transport, but loses due to mechanical nature and extra interfaces between water and waterblock and water and radiator. Heat pipes have the same interfaces, but don't have any mechanical aides (jet impingement etc) at the waterblock side. So, in short, for the application you describe, you're better off using a coolant. Next, finding a working fluid for what you describe is hard. It has to boil at the temp you want to keep the "block" (evaporator) at, and condense at the cold end temp, and there needs to be a decent Delta-T between the two to generate the energy to get the liquid/gas to move in velocities and volumes to provide the transportation of the energy at the same rate as it arrives from the CPU (and environment, as you'll be sub delta). Also, bear in mind that outside the working temp of the fluid, a heatpipe stops pumping heat. They fail not through mechanical breakdown, but in that the conditions exceed their parameters... __________________ Long Haired Git "Securing an environment of Windows platforms from abuse - external or internal - is akin to trying to install sprinklers in a fireworks factory where smoking on the job is permitted." (Prof. Gene Spafford) My Rig, in all its glory, can be seen best here AMD XP1600 @ 1530 Mhz \| Soyo Dragon + \| 256 Mb PC2700 DDRAM \| 2 x 40 Gb 7200rpm in Raid-0 \| Maze 2, eheim 1250, dual heater cores! \| Full specifications (PCDB)

03-08-2005, 08:40 PM	#4
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	The heatpipe just adds an extra layer to the simple solution of "phase change" alone, which would perform slightly less efficiently. It'd be a challenging exercise otherwise. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

03-08-2005, 10:16 PM	#6
bigben2k Responsible for 2% of all the posts here. Join Date: May 2002 Location: Texas, U.S.A. Posts: 8,302	Exactly, hence the exercise. __________________ WBTA (Water Block Testing Alliance) Nordic Hardware WC 101

03-08-2005, 06:49 AM	#3
Da_Priest Cooling Neophyte Join Date: Jul 2004 Location: Izegem, Belgium Posts: 8	you'd be better of making a chiller then. You have to make blocks for your graka and cpu and stuff to connect your heatpipes to anyway, so you can makewaterblocks too. instaid of hooking up you heatpipes to a cold plate, put your cold plate in a reservoir and run you 'water' trough it ... You'll achieve much better temps than with heatpipes, the amount of time and money to put into it would be about the same prolly. yet you'll still need a powerfull compressor. A modified vapochill or prometia won't handle the charge of a whole system... I think... anyway, it's a crazy idea indeed grtz

03-08-2005, 09:09 PM	#5
HAL-9000 Cooling Savant Join Date: May 2004 Location: San Diego, CA Posts: 202	Sounds like a interesting idea. The heatpipe chemical would have to be different though wouldn't it? The idea I get from the literature is they work within very specific temperature ranges. Would the heatpipe fluid have a risk of freezing when it contacted the cold plate?

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