Overclock.net › Forums › Cooling › Cooling Experiments › mineral oil-based cooler - senior project
New Posts  All Forums:Forum Nav:

mineral oil-based cooler - senior project - Page 4

post #31 of 53
Quote:
Originally Posted by Guruboy View Post
Well the point of this system is not to rival the performance of watercooling, it's the provide a decent price/performance ratio. Hopefully better than most budget-minded HS+F setups.

Well as long as that less dense oil is near the pipes and not the CPU, the CPU stays cooler, right? Very warm oil would make it to the next levels of pipes (or now probably heat pipes).

I already said heatpipes on the inside and fins on the outside.

I said "possibly aka copper wire" because heat pipes are going to be difficult to work with and will be much more expensive. They do the same job, but I should have said "or". Also, Where did you get the "100 times" number from?

That makes sense. My point is that I believe IHS > oil > Cu > air will be more effective than IHS > paste > Cu > fins > air.
I think a sealed mineral oil heatsink, that comes in contact with the IHS is going to be more expensive than a standard heatsink...i mean, I don't even know how you'd imply that's a selling point, since you'd either need professional installation, or the user would have to fill the reservoir, and seal the thing themselves.

As for "copper wire", it is less effective than a heatpipe. Heatpipes harness the heat of vaporization of a working fluid, that's probably 100 times more effective than conduction alone.

Again, I do not think a low density fluid is going to be better than thermal paste. In your thermal circuit, you've basically substituted mineral oil for paste. Mineral oil being less dense than thermal paste, in a thick layer, is going to ADD thermal resistance to the system, and decrease overall effectiveness of the cooler.
mitx
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel i5 3570K 4.5ghz ASRock Z77E-ITX Sapphire R9 290x Tri-X 1200/1500 Crucial Ballistix Tactical LP 2x8 GB 8-8-8-24-1T 
Hard DriveHard DriveHard DriveHard Drive
Samsung 840 Pro 256gb Western Digital Velociraptor Western Digital WD15EARX Western Digital WD15EADS 
CoolingOSMonitorKeyboard
Silverstone NT06-Pro Windows 10 x64 LG 29UM67 Logitech G710+ 
PowerCaseMouseAudio
Silverstone SG08 600W PSU Silverstone SG08B Logitech G700 Pioneer SE-A1000 
  hide details  
Reply
mitx
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel i5 3570K 4.5ghz ASRock Z77E-ITX Sapphire R9 290x Tri-X 1200/1500 Crucial Ballistix Tactical LP 2x8 GB 8-8-8-24-1T 
Hard DriveHard DriveHard DriveHard Drive
Samsung 840 Pro 256gb Western Digital Velociraptor Western Digital WD15EARX Western Digital WD15EADS 
CoolingOSMonitorKeyboard
Silverstone NT06-Pro Windows 10 x64 LG 29UM67 Logitech G710+ 
PowerCaseMouseAudio
Silverstone SG08 600W PSU Silverstone SG08B Logitech G700 Pioneer SE-A1000 
  hide details  
Reply
post #32 of 53
Thread Starter 
Quote:
Originally Posted by Guruboy View Post
Have you seen what copper costs these days? Mineral oil is practicly free in comparison.

With a good, threaded fillport and a threaded cap, along with a filling tube/funnel, this would be a pretty simple install. You don't have to keep the motherboard stable in fear of messing up the thermal paste; you just screw on the reservoir and fill her up. Anyone who can install a HS+F could install this.
I am relying more on the mineral oil surrounding and being in contact with the CPU better than thermal paste does, rather than the thermal conductivity of the oil itself, which isn't even that much lower than water.
Creeping Death
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II 940 BE 3.7GHz Asus M3A78-T Visiontek HD4870 512MB 2x2GB OCZ Reaper DDR2-1066 
Hard DriveOptical DriveOSMonitor
WD VelociRaptor 300GB Sony DVD Burner Windows 7 Ultimate x64 2x Dell 2408WFP 
KeyboardPowerCaseMouse
Logitech Wave USB Antec Neopower 500W Modded Rocketfish Logitech G9 
Mouse Pad
(none) 
  hide details  
Reply
Creeping Death
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II 940 BE 3.7GHz Asus M3A78-T Visiontek HD4870 512MB 2x2GB OCZ Reaper DDR2-1066 
Hard DriveOptical DriveOSMonitor
WD VelociRaptor 300GB Sony DVD Burner Windows 7 Ultimate x64 2x Dell 2408WFP 
KeyboardPowerCaseMouse
Logitech Wave USB Antec Neopower 500W Modded Rocketfish Logitech G9 
Mouse Pad
(none) 
  hide details  
Reply
post #33 of 53
Thread Starter 
oops, double-posted
Creeping Death
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II 940 BE 3.7GHz Asus M3A78-T Visiontek HD4870 512MB 2x2GB OCZ Reaper DDR2-1066 
Hard DriveOptical DriveOSMonitor
WD VelociRaptor 300GB Sony DVD Burner Windows 7 Ultimate x64 2x Dell 2408WFP 
KeyboardPowerCaseMouse
Logitech Wave USB Antec Neopower 500W Modded Rocketfish Logitech G9 
Mouse Pad
(none) 
  hide details  
Reply
Creeping Death
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II 940 BE 3.7GHz Asus M3A78-T Visiontek HD4870 512MB 2x2GB OCZ Reaper DDR2-1066 
Hard DriveOptical DriveOSMonitor
WD VelociRaptor 300GB Sony DVD Burner Windows 7 Ultimate x64 2x Dell 2408WFP 
KeyboardPowerCaseMouse
Logitech Wave USB Antec Neopower 500W Modded Rocketfish Logitech G9 
Mouse Pad
(none) 
  hide details  
Reply
post #34 of 53
Quote:
Originally Posted by Guruboy View Post
Have you seen what copper costs these days? Mineral oil is practicly free in comparison.

With a good, threaded fillport and a threaded cap, along with a filling tube/funnel, this would be a pretty simple install. You don't have to keep the motherboard stable in fear of messing up the thermal paste; you just screw on the reservoir and fill her up. Anyone who can install a HS+F could install this.

I am relying more on the mineral oil covering the CPU MUCH better than thermal paste than the thermal conductivity of the oil itself.
part of the benefits of thermal paste in a modern heatsink system is that contact pressure provides a huge decrease in contact resistance, you're relying on oil resting on the surface of the cpu. I just don't see how the system wouldn't be hard to isntall. you have to make sure there are ZERO air bubbles in the system. If one exists, and you use this in a tower, the bubble moves to the surface of the cpu, and you have a hotspot. If you overfill just slightly, you have to clean up oil...which from experience with a fossil fuels lab, I know is annoying and messy.

See, the resistances caused by thermal paste, however small they are, can be overcome by applying more mounting pressure, further compressing the paste layer, and reducing contact resistance. Liquid cooling is a great system, it just needs flow. Fluid flow gives you a constantly greater delta T, as well as increased convection transfer coefficients...just look at what higher airflow does for fins.
mitx
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel i5 3570K 4.5ghz ASRock Z77E-ITX Sapphire R9 290x Tri-X 1200/1500 Crucial Ballistix Tactical LP 2x8 GB 8-8-8-24-1T 
Hard DriveHard DriveHard DriveHard Drive
Samsung 840 Pro 256gb Western Digital Velociraptor Western Digital WD15EARX Western Digital WD15EADS 
CoolingOSMonitorKeyboard
Silverstone NT06-Pro Windows 10 x64 LG 29UM67 Logitech G710+ 
PowerCaseMouseAudio
Silverstone SG08 600W PSU Silverstone SG08B Logitech G700 Pioneer SE-A1000 
  hide details  
Reply
mitx
(16 items)
 
  
CPUMotherboardGraphicsRAM
Intel i5 3570K 4.5ghz ASRock Z77E-ITX Sapphire R9 290x Tri-X 1200/1500 Crucial Ballistix Tactical LP 2x8 GB 8-8-8-24-1T 
Hard DriveHard DriveHard DriveHard Drive
Samsung 840 Pro 256gb Western Digital Velociraptor Western Digital WD15EARX Western Digital WD15EADS 
CoolingOSMonitorKeyboard
Silverstone NT06-Pro Windows 10 x64 LG 29UM67 Logitech G710+ 
PowerCaseMouseAudio
Silverstone SG08 600W PSU Silverstone SG08B Logitech G700 Pioneer SE-A1000 
  hide details  
Reply
post #35 of 53
Thread Starter 
Quote:
Originally Posted by ryboto View Post
part of the benefits of thermal paste in a modern heatsink system is that contact pressure provides a huge decrease in contact resistance, you're relying on oil resting on the surface of the cpu. I just don't see how the system wouldn't be hard to isntall. you have to make sure there are ZERO air bubbles in the system. If one exists, and you use this in a tower, the bubble moves to the surface of the cpu, and you have a hotspot. If you overfill just slightly, you have to clean up oil...which from experience with a fossil fuels lab, I know is annoying and messy.

See, the resistances caused by thermal paste, however small they are, can be overcome by applying more mounting pressure, further compressing the paste layer, and reducing contact resistance. Liquid cooling is a great system, it just needs flow. Fluid flow gives you a constantly greater delta T, as well as increased convection transfer coefficients...just look at what higher airflow does for fins.
Well I'm relying on half a kilogram of mineral oil, which is more mass than a lot of heatsinks out there except for the heaviest ones like the TT120 and TRUE, but the contact pressure point (no pun intended) is well seen.

I understand the air bubble concern. I have no idea how easily air bubbles form in mineral oil as opposed to water, or if my system will make them less likely. I don't think that getting an air bubble out of this system would be too difficult though, if they actually form.

Surely running the mineral oil through a tubing system would be more effective, but that means you have to have the skills to set up a watercooling system, and I'm trying to keep setup simple and parts cost low.
Creeping Death
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II 940 BE 3.7GHz Asus M3A78-T Visiontek HD4870 512MB 2x2GB OCZ Reaper DDR2-1066 
Hard DriveOptical DriveOSMonitor
WD VelociRaptor 300GB Sony DVD Burner Windows 7 Ultimate x64 2x Dell 2408WFP 
KeyboardPowerCaseMouse
Logitech Wave USB Antec Neopower 500W Modded Rocketfish Logitech G9 
Mouse Pad
(none) 
  hide details  
Reply
Creeping Death
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II 940 BE 3.7GHz Asus M3A78-T Visiontek HD4870 512MB 2x2GB OCZ Reaper DDR2-1066 
Hard DriveOptical DriveOSMonitor
WD VelociRaptor 300GB Sony DVD Burner Windows 7 Ultimate x64 2x Dell 2408WFP 
KeyboardPowerCaseMouse
Logitech Wave USB Antec Neopower 500W Modded Rocketfish Logitech G9 
Mouse Pad
(none) 
  hide details  
Reply
post #36 of 53
Quote:
Originally Posted by Guruboy View Post
Well as long as that less dense oil is near the pipes and not the CPU, the CPU stays cooler, right? The more dense oil, which we both know conducts heat betterm will replace the warm oil near the CPU. The CPU is vertical so the effect won't be as efficient, but it will be present. Very warm oil would make it to the next levels of pipes (or now probably heat pipes).
No... the oil won't circulate enough. The temperature delta won't be enough to induce flow. If you look at watercooling, the temperature delta is less than 3C, if that. That little difference in temperature in such a small space won't be enough for circulation. The temperature differences within the fluid are too small to induce much movement.


Quote:
Originally Posted by Guruboy View Post
I said "possibly aka copper wire" because heat pipes are going to be difficult to work with and will be much more expensive. They do the same job, but I should have said "or", not "aka". I know what heat pipes are. Also, Where did you get the "100 times" number from?
http://en.wikipedia.org/wiki/Heat_pipe#limitations I have read other places that the difference between a heatpipe and equivalent copper bar is well over a magnitude.


Quote:
Originally Posted by Guruboy View Post
That makes sense. My point is that I believe IHS > oil > Cu > air will be more effective than IHS > paste > Cu > fins > air.
It's not. Try to calculate the the thermal resistence of the path. It's not easy to do though. Here is rough example....
CPU Heat -> .001" 10 W/(m·K) TIM -> copper heat spreader -> air
CPU Heat -> .5" .1 W/(m·K) Oil -> copper inteface-> air

Heatsinks work by spreading heat from a source. In your design, there's a 1:1 hot surface area to cool surface area for the copper.... not good. Basically, the copper is just a heat passthrough. A good design would try to spread the heat using the copper.
Once again...
(13 items)
 
  
CPUMotherboardGraphicsRAM
i7 920 [4.28GHz, HT] Asus P6T + Broadcom NetXtreme II VisionTek HD5850 [900/1200] + Galaxy GT240 2x4GB G.Skill Ripjaw X [1632 MHz] 
Hard DriveOSMonitorKeyboard
Intel X25-M 160GB + 3xRAID0 500GB 7200.12 Window 7 Pro 64 Acer H243H + Samsung 226BW XARMOR-U9BL  
PowerCaseMouseMouse Pad
Antec Truepower New 750W Li Lian PC-V2100 [10x120mm fans] Logitech G9 X-Trac Pro 
  hide details  
Reply
Once again...
(13 items)
 
  
CPUMotherboardGraphicsRAM
i7 920 [4.28GHz, HT] Asus P6T + Broadcom NetXtreme II VisionTek HD5850 [900/1200] + Galaxy GT240 2x4GB G.Skill Ripjaw X [1632 MHz] 
Hard DriveOSMonitorKeyboard
Intel X25-M 160GB + 3xRAID0 500GB 7200.12 Window 7 Pro 64 Acer H243H + Samsung 226BW XARMOR-U9BL  
PowerCaseMouseMouse Pad
Antec Truepower New 750W Li Lian PC-V2100 [10x120mm fans] Logitech G9 X-Trac Pro 
  hide details  
Reply
post #37 of 53
Thread Starter 
I never said the oil would circulate like nuts, but are you saying that absolutely no flow of oil will occur between oil that is closer to the heatsink and oil that is farther away? All I'm saying is that there will be more flow than say, a solid copper heatsink. You say the temperature difference will be too small to induce flow, citing that 3C number-from-somewhere in WC systems. Are you telling me that the heat from the CPU will be distributed throughout the mineral oil? Is that where the small temperature difference comes from? Or are you telling me that is will not transfer quickly? Because if not, there will be a large temperature differences and this density-induced flow WILL occur.

It's worth noting that mineral oil has about five times the specific heat capacity of copper, so it takes a lot more heat to raise the temperature of the mineral oil. Attempting to dissipate the heat from a CPU throughout a volume of mineral oil is more effective than spreading the same amount of heat throughout the same volume of copper. Since my system has a larger volume of mineral oil than is found in most commercial copper heatsinks, this gives a large advantage to my system over a copper heatsink.

I understand that those are rough calculations, but you seem to have slashed about 25% of the thermal conductivity of oil while adding 25% to that of AS5 and the like. I will ask my physics teacher about the thermal resistance of this system.
Creeping Death
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II 940 BE 3.7GHz Asus M3A78-T Visiontek HD4870 512MB 2x2GB OCZ Reaper DDR2-1066 
Hard DriveOptical DriveOSMonitor
WD VelociRaptor 300GB Sony DVD Burner Windows 7 Ultimate x64 2x Dell 2408WFP 
KeyboardPowerCaseMouse
Logitech Wave USB Antec Neopower 500W Modded Rocketfish Logitech G9 
Mouse Pad
(none) 
  hide details  
Reply
Creeping Death
(13 items)
 
  
CPUMotherboardGraphicsRAM
Phenom II 940 BE 3.7GHz Asus M3A78-T Visiontek HD4870 512MB 2x2GB OCZ Reaper DDR2-1066 
Hard DriveOptical DriveOSMonitor
WD VelociRaptor 300GB Sony DVD Burner Windows 7 Ultimate x64 2x Dell 2408WFP 
KeyboardPowerCaseMouse
Logitech Wave USB Antec Neopower 500W Modded Rocketfish Logitech G9 
Mouse Pad
(none) 
  hide details  
Reply
post #38 of 53
Quote:
Originally Posted by Guruboy View Post
I never said the oil would circulate like nuts, but are you saying that absolutely no flow of oil will occur between oil that is closer to the heatsink and oil that is farther away? All I'm saying is that there will be more flow than say, a solid copper heatsink. You say the temperature difference will be too small to induce flow, citing that 3C number-from-somewhere in WC systems. Are you telling me that the heat from the CPU will be distributed throughout the mineral oil? Is that where the small temperature difference comes from? Or are you telling me that is will not transfer quickly? Because if not, there will be a large temperature differences and this density-induced flow WILL occur.

It's worth noting that mineral oil has about five times the specific heat capacity of copper, so it takes a lot more heat to raise the temperature of the mineral oil. Attempting to dissipate the heat from a CPU throughout a volume of mineral oil is more effective than spreading the same amount of heat throughout the same volume of copper. Since my system has a larger volume of mineral oil than is found in most commercial copper heatsinks, this gives a large advantage to my system over a copper heatsink.
Since oil is less conductive than copper, you will need to rely on convection. There won't be enough temperature variation zones in that relatively small block. Over time the heat from the CPU will be distributed pretty evenly... unfortunately, that temperature will be close to the temperature of the source. With any liquid cooling, you want the liquid to move to the cooler area and bring the excess heat with it.

Think about this.... you have a watercooling system with a CPU, tubing, and a radiator. You are relying mainly on conduction here. The resulting temperature will be around 60C+ and the water will get quite warm. However, if you add a pump, you know are utilizing convection since the water flow moves the heat to cooler areas. The resulting CPU temperature will be around 40C and the water temperature will be only +3C over ambient. (This is from a personal experience with a X2 3800).


Your design provides a better means of storing thermal energy than a traditional heatsink. However, a heatsink has an one or two order magnitude better at distributing the thermal energy into the atmosphere. This is more important in cooling.

The big thing is your are building a superior heat absorber. However, it can't dissipate the energy rapidly enough. All basic cooling rely on transporting thermal energy from a source to an area of lesser thermal energy as fast and efficiently as possible. All basic cooling final area of lesser thermal energy is the surrounding air. Therefore, you need to maximizing getting the heat from the CPU to the air. Your design gets heat from the CPU to a relatively small area better. However, that volume of oil won't be able to hold as much energy as a room of air. IMO, the transfer of heat from the oil to the air is lacking.
Once again...
(13 items)
 
  
CPUMotherboardGraphicsRAM
i7 920 [4.28GHz, HT] Asus P6T + Broadcom NetXtreme II VisionTek HD5850 [900/1200] + Galaxy GT240 2x4GB G.Skill Ripjaw X [1632 MHz] 
Hard DriveOSMonitorKeyboard
Intel X25-M 160GB + 3xRAID0 500GB 7200.12 Window 7 Pro 64 Acer H243H + Samsung 226BW XARMOR-U9BL  
PowerCaseMouseMouse Pad
Antec Truepower New 750W Li Lian PC-V2100 [10x120mm fans] Logitech G9 X-Trac Pro 
  hide details  
Reply
Once again...
(13 items)
 
  
CPUMotherboardGraphicsRAM
i7 920 [4.28GHz, HT] Asus P6T + Broadcom NetXtreme II VisionTek HD5850 [900/1200] + Galaxy GT240 2x4GB G.Skill Ripjaw X [1632 MHz] 
Hard DriveOSMonitorKeyboard
Intel X25-M 160GB + 3xRAID0 500GB 7200.12 Window 7 Pro 64 Acer H243H + Samsung 226BW XARMOR-U9BL  
PowerCaseMouseMouse Pad
Antec Truepower New 750W Li Lian PC-V2100 [10x120mm fans] Logitech G9 X-Trac Pro 
  hide details  
Reply
post #39 of 53
hmm, interesting idea. i can see alot of flaws though which could have been solved by some basic research, most of which Duckie has already pointed out. First off is about, Convection, Density and Viscosity, the only way for this design to be effective is for the motherboard to be laid flat, then you get an even heating of the oil, with the motherboard on a vertical plane only 2/3 maybe 3/4 of the oil will be used as the rest will be sat cold at the bottom, the oil will circulate- very slowly, the viscosity is the main factor here as to convect oil takes alot of energy.

secondly is that the surface area is far too small to be efficient, this is related to The Rainhill Trials in England where the first Steam Locomotives competed to win a race, 'The Rocket' won because the boiler used more heat pipes and therefore more pressure was put into the cylinders, this is basically what your idea is based on, an inverse boiler.

Thirdly, the heatpipes need to be optimized for airflow, remember hot air is less dense than cold air so the heatpipes exhausting on the top and around the unit are a bad idea, here is what i think would be better:

Using this design the fan would be located on the bottom and vent air up through the pipes and then exhausted through the top of the case.

My final thought is about expansion, have you thought about the heat expanding and contracting the oil, some kind of pressure valve will be needed or a gasket of some sort.

This is a nice innovative idea though and i like to see this kind of thing, with some work this could potentially be successful in terms of your goal. +rep
    
CPUMotherboardGraphicsRAM
Q6600 G0 @ 2.9Ghz (1.25V) 32'C/55'C EVGA nforce680i SLI (A1) 2x BFG 8800GT OC 512mb Edition (625/900) SLI 4x 1Gb Crucial Ballistix Tracers (PC2 6400) 800mhz 
Hard DriveOptical DriveOSMonitor
1.5Tb Total (WD + Maxtor) 2x LiteOn 20x DVD Writer (SATA) Vista Home Premium 64bit Samsung T240 HD/ Sony Bravia 32" HDTV/ Samsung 932 
KeyboardPowerCaseMouse
Logitech Media Elite USB Silverstone Zeus 650W Modified Thermaltake Armor Razer Diamondback 
Mouse Pad
Razer Mantis Speed 
  hide details  
Reply
    
CPUMotherboardGraphicsRAM
Q6600 G0 @ 2.9Ghz (1.25V) 32'C/55'C EVGA nforce680i SLI (A1) 2x BFG 8800GT OC 512mb Edition (625/900) SLI 4x 1Gb Crucial Ballistix Tracers (PC2 6400) 800mhz 
Hard DriveOptical DriveOSMonitor
1.5Tb Total (WD + Maxtor) 2x LiteOn 20x DVD Writer (SATA) Vista Home Premium 64bit Samsung T240 HD/ Sony Bravia 32" HDTV/ Samsung 932 
KeyboardPowerCaseMouse
Logitech Media Elite USB Silverstone Zeus 650W Modified Thermaltake Armor Razer Diamondback 
Mouse Pad
Razer Mantis Speed 
  hide details  
Reply
post #40 of 53
I'd be interested to see a run in a finite element modeling program to see the temperature distribution when a fan is running and with a CPU's thermal load applied.
BladeRunner v3.0
(11 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7-5930K @ 4.6GHz Core, 4.4GHz Cache ASUS X99 Sabertooth Sapphire R9 380 Dual-X OC G.Skill TridentZ 32GB DDR4 @ 13-15-13-33-1T 320... 
Hard DriveCoolingOSKeyboard
Samsung 850 Pro 512GB Noctua NH-D15S Windows 10 Home 64-bit Logitech G910 Orion Spark 
PowerCaseMouse
EVGA SuperNova 1000W T2 NZXT Phantom 820 Black Logitech G5 
  hide details  
Reply
BladeRunner v3.0
(11 items)
 
  
CPUMotherboardGraphicsRAM
Intel Core i7-5930K @ 4.6GHz Core, 4.4GHz Cache ASUS X99 Sabertooth Sapphire R9 380 Dual-X OC G.Skill TridentZ 32GB DDR4 @ 13-15-13-33-1T 320... 
Hard DriveCoolingOSKeyboard
Samsung 850 Pro 512GB Noctua NH-D15S Windows 10 Home 64-bit Logitech G910 Orion Spark 
PowerCaseMouse
EVGA SuperNova 1000W T2 NZXT Phantom 820 Black Logitech G5 
  hide details  
Reply
New Posts  All Forums:Forum Nav:
  Return Home
  Back to Forum: Cooling Experiments
Overclock.net › Forums › Cooling › Cooling Experiments › mineral oil-based cooler - senior project