Can heatpipes function outside vertical alignment?
Can heatpipes function outside vertical alignment?
I've seen "CPU Cooler Charts 2008" and believe Tom has provided thorough evaluations of various coolers in practical settings. However, I feel the topic of heatpipe alignment was not fully addressed and might have been misunderstood.
Heatpipes rely on convection and gravity to function properly. The fluid needs to remain in contact with the heat source (the CPU) to absorb heat, evaporate, rise as vapor, be cooled by the heatsink/fan, condense back into liquid, and then flow back down to the heat source to restart the cycle.
From my perspective, most system boards are installed vertically in tower cases, causing heatpipes to run horizontally across the heatsinks. No matter the orientation, the fluid tends to gather at the lowest point. With every cooler Tom's reviewed, that lowest point is only present on a horizontal system board. If the board is mounted vertically, none of these coolers collect the fluid at the CPU, making them less efficient than they could be.
It would be valuable to see additional tests with heatpipes used in horizontally-mounted systems. While it's clear the coolers would perform significantly better then, it might surprise how much impact this orientation has.
Tom's may discover that some coolers previously rated poorly actually function quite well when designed for vertical installation.
Additionally, why isn't there a specific design for heatpipe CPU coolers tailored to vertically-mounted systems?
To address these issues, most heat pipes—especially those used for PC cooling—include a wick inside the tube that helps pull the cooled fluid back toward the hot end through capillary action. Wicked tubes perform slightly better than unwicked ones when placed correctly, but they still show a minor decline in efficiency, particularly when inverted. This variation is probably not significant enough to affect results, especially if you're relying mainly on motherboard temperature sensors. It would be useful to review the tests referenced, though.
I have an E6850, 3 GHz, 65 nM second generation Core2Duo (assuming the 6600 and 6700 are first generation and the E8500 is third generation). In theory it should produce some heat, but with a Noctua NH-U12 it might not even require a fan. The case is vertical and the heat pipes are horizontal.
Theoretically there is a difference. You won't notice this in the usual PC temp range. But if you go to 300C+ then you start to see the difference (usually 10-15C+).
The wicking action cancels out any orientation concerns. About two years ago, using an Artic Freezer 64 Pro cooler, I confirmed no variation in temperatures during idle or load, whether the M/B was vertical or horizontal. If you're worried, perform your own tests—you'll see the results.
Completely in agreement, keep in mind the graphics cards. They seem to rely on a horizontal setup (which is likely more typical now due to the rise of tower cases). This might account for the need for water cooling. There are now 'semi-flexible' gas pipes in construction—maybe we should think about applying similar ideas to PC heat-pipes.
I completely understand and recall the graphics cards. They seem to rely on a horizontal setup, which is likely more typical now due to the rise of tower cases. This might be why water cooling is necessary. There are now 'semi-flexible' gas pipes in construction, and it could be worth considering similar options for PC heat-pipes.