Hardware Canucks

Hardware Canucks (http://www.hardwarecanucks.com/forum/)
-   Air Cooling (http://www.hardwarecanucks.com/forum/air-cooling/)
-   -   Heatsink Mod? (http://www.hardwarecanucks.com/forum/air-cooling/59118-heatsink-mod.html)

tangrisser January 13, 2013 10:56 AM

Heatsink Mod?
 
This is just a thought more than anything else.

Would it help with cooling if the heatsink fins on a tower style heatsinks are on an angle and/or if the sides were blocked to focus air through the fins?

I'm just thinking rather than 'guiding' air through the heatsink, what if the air was actually hitting the fins. Would the surface area increase sufficiently to the point where modding a heatsink is worth while?

How do you feel about if the sides were blocked? I feel like it might interfere with the actual cooling but at the same time, it might do better because air is focused into the middle rather than dispersing all over the place. Do most fans do a good enough job at creating air pressure that this is negligible?

It would be pretty interesting to see someone try with awesome mod skills because I got none. :ph34r:

Perineum January 13, 2013 11:55 AM

I think a slight angle would be fine, but anything too dramatic is just going to increase noise and possibly cavitation. Blocking the sides from the fan to the back of the heatsink would probably help somewhat but remember, the heatpipe tower heatsink has pretty much been perfected now.

BrainEater January 13, 2013 01:22 PM

I think the idea you are after would be a laminar flow heatsink. (think 'smooth air')...

Without a lot of testing in a controlled environment ( $$ ) it would be blind luck to build an efficient one.Chaotic gasflow is good for some things , cooling is one.

:thumb:

gingerbee January 13, 2013 01:47 PM

ya there just there too increase surface area so i second the blind like factor

tangrisser January 13, 2013 11:05 PM

Quote:

Originally Posted by BrainEater (Post 684080)
I think the idea you are after would be a laminar flow heatsink. (think 'smooth air')...

Without a lot of testing in a controlled environment ( $$ ) it would be blind luck to build an efficient one.Chaotic gasflow is good for some things , cooling is one.

:thumb:

I was thinking about how liquid cools and thought about that. I suppose fans should create turbulence in the shape of vortex rather than in laminar flow.

It took me some time to understand laminar flow from Google and Wikipedia because they describe is as liquid effect but I understand what you mean after reading Wikipedia few more times. :punk:

This guy did what I was saying about the sides but he went as far as covering the space between the heatsink and the fan:
Improving the 212+'s efficiency...? - AnandTech Forums

Masteroderus January 14, 2013 06:13 AM

The guy is saying 5 degrees cooler at load when he taped up the sides of the heatsink.

I'm sorry to be so critical, but he didn't post his ambient temperatures, so the data is useless. The real number to look for is change in temperature (load temp - ambient temp of room).

So let's put our physics thinking caps on and try and figure out what is going on inside the heatsink.

The CPU generates heat and transfers it to its integrated heat spreader, which transfers heat to the thermal interface material, which transfers it the heat pipes and the metal base plate in which the pipes are embedded.

The heat pipes transfer heat to the body structure, which is finned to maximize surface area. Heat does rise, and the heatsink will work, somewhat, without the fans plugged in, because as the hot air rises it makes a pressure differential near the surface of the heatsink that sucks "fresh" air in ("in" being to the surface of the heatsink).

Covering the side fins does focus the flow of the air more, but it also removes the ability for the air to travel orthagonally (at 90 degrees to the fan) through the heatsink.

Think about the mass of air going through the heatsink. It would equal the flow rate (m^3/sec) times the density of air (kg/m^3) = kg/sec. This same mass of air would go through the heatsink regardless of the side shrouds (approximately). But one path has more surface area of heatsink, the path that includes the sides open.

This is why coolermaster, a very reputable and old cooling company, did not include 2 cents worth of duct tape, or a 2 cent molded plastic shroud with the 212. 5 degrees is A LOT for a 2 cent mod, and I am fairly sure that if the side shroud mod worked it would be included with the 212. If the 212 was 5 degrees cooler at load it would mean BIG business for coolermaster because it would basically kill off a lot of competition.

Now, if you were paying close attention you can see from my analysis what you can do to increase the efficiency of your heat sink. If the mass flow of air through your heatsink is the product of density and volumetric flow rate, you have a free parameter to play with without modding the 212; density.

PV=nRT => V/n = PRT => V/n ~ P

You can increase pressure inside your case which will, both theoretically and from first hand experience, increase the efficiency of your air cooler. Just add more intake fans while keeping the cpu exhaust fan an exhaust fan.

Anyways, I hope this helps, and I'm not making this stuff up, it's basic thermodynamics which I have formally studied.

Dragonstongue January 14, 2013 06:16 AM

vortex is a good thing to have, if designed from the get-go to get said vortex smoothly, almost like a prop in water would be, however, just like said prop in water if it cuts to hard or spins to fast for the cut that it is making, cavitation could be a result.

Also, quite a few modern heatsinks close up the sides and just leave fins to wick the heat, sometimes dimpling the fins themselves for more surface area or (not the mouths or push/pull area)to aid in smoothing/collecting the airflow(less resistance just like a golfball) there is only so much that can be done, the heatpipes themselves are doing the vast majority of work.

Now if you could make the heatpipes into the sides of said heatsink and allow them to wick thier liquid into the fins(like a rad per fin type thing) then you could probably boost the cooling power exponentially, but as it stands, there is very little to boost them, they are at thier pinnacle more or less, the size of the hatsink and its fins i.e wide,tall,thick and a couple other factors.

KaptCrunch January 14, 2013 09:05 AM

other issue is recycled air on the HS when the prescott's were furnaces simple paper towel roll fitted purfect on intel fan housing dropped cpu temps in case


laminar flow corrected

NyteOwl January 14, 2013 06:04 PM

Once of the main concepts in laminar flow is to create as little friction as possible between the fluid and the surrounding surfaces. Without sufficient contact between a heatsink and the cooling fluid (be it air or water or soemthign else) there will be poor heat transfer. This is one case where laminar flow may be a disadvantage.


All times are GMT -7. The time now is 05:33 AM.