Heat pipe orientation.. Does it matter?

K

Korben Dallas

Gawd
Joined
Oct 25, 2010
Messages
561
Ok

I'll be building a brand new rig come christmas time and have most of my stuff planned out already. One of the stuff is the Cooler Master Hyper 212 plus. I've been reading reviews and searching [H] on this particular cooler for a while now since and decided it was the best bang for buck.

I was researching the best way to apply thermal paste for direct touch heat pipe shenanigans and came across one website that said the orientation of the heat pipes is better when they are horizontal.
thisdv.jpg



Now I know from looking at pics of people's rigs here that the favored orientation is vertical.
thatr.jpg



I wish I had bookmarked that site but alas, Mary J had my full attention. Anyway, this site mentioned that gravity had something to do with the better performance when the heat pipes were oriented horizontally.

What do you guys think?
 
I am doubtful of this, as I have seen no such effects on my rigs, also, the temp will more likely be more so effected by the orientations based on fan layout than heatpipe. Keep in mind one of the main uses of heatpipes were for satellites (in zero gravity). If you question this, when you receive the cooler, test in both orientations and see for your self what gives you better temps.
 
heatpipes do not move liquid based on gravity, they move it using a wick action so the orientation will not matter.
 
Google "heatpipe orientation" - third hit for pointed to this article showing a 9c difference favoring horizontal orientation.

Ideally, you would want the lowest point of the heatpipe to be contacting the heat source, with the rest of the heatpipe above. It's true that heatpipes work well enough against gravity, but remember - the liquid inside the pipe wants to rise once it vaporizes, and the liquid, being relatively cooler and denser, wants to fall. The wick structure of the inner walls are reasonably capable of overcoming the force of gravity, but it's less effective than it would be if it didn't have to work against gravity.

In your second diagram, the top half of the heatsink should perform better than the horizontal orientation, whereas the bottom half should perform worse, moving less heat to the heatsink fins. The horizontal orientation presents a sort of gravitationally neutral configuration.
 
i'd like to see more tests on the subject. it may be that the design of that heatsink is poor whereas heatsinks with pipes that come out of two sides fare better. but as is, there certainly would not be any issues with a horizontal application, assuming the fan was compatible.
 
from what I've gathered most heat pipe coolers work best in the horizontal mount, but there are an odd few that show no difference in either orientation. either way, load temperature differences between either orientation only differs by less than several degrees at best.
 
to test this even further you could simply lay any system with a standard heat pipe tower on it's side so all the heat pipes point up and see the max efficiency of the cooler.

imo there is no effect. it should work upside down as well as right side up. if there was truly a 9C difference that was only related to orientation then i would say those heat pipes are bad.


just test it out by laying your PC on it's side so heat pipes point up and see the difference in max temps. hell 95% of GPU coolers have heat pipes upside down.

in Necere theory this should provide noticeable improvements. if it doesn't it simply means the heat pipes are working as intended and work in any orientation.
 
gesicht said:
to test this even further you could simply lay any system with a standard heat pipe tower on it's side so all the heat pipes point up and see the max efficiency of the cooler.

imo there is no effect. it should work upside down as well as right side up. if there was truly a 9C difference that was only related to orientation then i would say those heat pipes are bad.


just test it out by laying your PC on it's side so heat pipes point up and see the difference in max temps. hell 95% of GPU coolers have heat pipes upside down.

in Necere theory this should provide noticeable improvements. if it doesn't it simply means the heat pipes are working as intended and work in any orientation.
Click to expand...

That's a good test - someone get on that and let us know what the results are.
 
Okay, hard data

From Enertron Inc., a thermal solutions and heat pipe manufacturer, from this PDF, located in the library section of their site:

OwZ8N.png


xje68.png


So it depends on the length and type of wick the heat pipe uses. These charts only compare gravity-assisted vertical and horizontal orientations, and don't address operation against gravity, however.

This Frostytech article claims most commercial heatpipe heatsinks use the metal mesh variety of wick, which, according to the above charts, perform 10-20% better in gravity-assisted vertical vs. horizontal orientation. It's probably safe to say the difference is even greater when comparing operation against gravity. That link has some interesting pics of the interior wick structure as well; worth a look.


Here are some choice quotes from other sources:
It is challenging to make heat-pipe heat exchangers operate against gravity, where the heat is transferred from the top portion of the heat exchanger to the lower portion. For this orientation, sintered-wick heat pipes are required, which substantially increases the cost.
Click to expand...
Source

One advantage of a heat pipe with a sintered powder wick is that it can work in any orientation, including against gravity (i.e., the heat source above the cooling source). The power transport capacity of the heat pipe will typically decrease as the angle of operation against gravity increases. Since groove and screen mesh wicks have very limited capillary force capability, they typically cannot overcome significant gravitational forces, and dry out generally occurs.
Click to expand...
(emphasis mine)
Source


TL;DR: Heat pipe orientation does matter, but depends on length & wick type, which you can't really know without opening (destroying) the heat pipe. Metal mesh wicks are (probably) the most common and are affected significantly (10-20%+) by orientation. Sintered metal powder wicks are less affected by orientation, and actually seem to perform better horizontally, but are more expensive and less common.
 
Good find.

That testing is the difference between the motherboard laying flat on a table, with the tower standing upright, and the motherboard on the edge with the tower sticking out to the side (the typical case orientation). Doesn't really address the rotation of a tower in a typical configuration - the difference between the heatpipes being horizontal or vertical as they cross the CPU. Probably lower impact of gravity in that case, since the heatpipe is predominantly horizontal to gravity in both rotational orientations.
 
You must log in or register to reply here.
Back
Top