Overclock.net banner

21 - 40 of 57 Posts

·
Premium Member
Joined
·
22,477 Posts
Also, Ira, check this out:<br />
<a href="http://www.frozencpu.com/products/1447/cpu-swi-16/Swiftech_MCX775-V_-_Pentium_4_LGA_775.html?tl=g40c14s52" target="_blank">http://www.frozencpu.com/products/14...l?tl=g40c14s52</a><br />
<br />
<br />
Thanks Franken, Got it book marked for the Intel guys.<img src="/images/smilies/thumb.gif" border="0" alt="" title="Thumb" class="inlineimg" />..Very nice!! If Spode had cut his down some more it would of looked like that, he gave up on it to soon...<img src="/images/smilies/biggrin.gif" border="0" alt="" title="Big Grin" class="inlineimg" />
 

·
Premium Member
Joined
·
5,282 Posts
That design would be pretty bad. Shape-wise you want to maximize surface area and allow for good airflow - which is why manufacturers make HS's with a lot of parallel plates. The plates give a good Surface-area to volume ratio and the flat parallel plates minimally restrict air flow. The fractal tree design would choke the HS. The heat would be trapped inside of the mass of "branches" and ruin the efficiency.
 

·
Registered
Joined
·
63 Posts
Discussion Starter #23
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>FrankenPC</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391558#post1391558" rel="nofollow"><img class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" border="0" alt="View Post" /></a>
</div>
<div style="font-style:italic">You could place a fan at the base of the "tree" blowing gently upwards towards the copper "capillaries". I think that would be one hell of a aircooled heatsink. But, think how soft those fine branches would be. You could accidently crush it.</div>

</td>
</tr>
</table>
</div>True true... If the turbine was used... At least it could serve the purpose of physically guarding (somewhat) against crushing... Although (it would be much cooler (as in nice), for the turbine to be made see-through.. maybe lexan.. lol
 

·
Premium Member
Joined
·
5,573 Posts
I've often wondered if you could use a low temp vaporizing liquid in a heat pipe assembeled in such a way the the copper block has two stages. On the bottom is the "heater" which vaporizes the liquid...that gas travel's up several heat pipes into a condenser which then lets the condensed liquid flow back down a separate set of tubes into a insulated block (plastic) on the top of the copper block with a one way valve to allow the condensed liquid back onto the CPU block. So, you would have a simple circulation system power by expanding gasses and not a pump.
 

·
Registered
Joined
·
63 Posts
Discussion Starter #25
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>rabidgnome229</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391572#post1391572" rel="nofollow"><img class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" border="0" alt="View Post" /></a>
</div>
<div style="font-style:italic">That design would be pretty bad. Shape-wise you want to maximize surface area and allow for good airflow - which is why manufacturers make HS's with a lot of parallel plates. The plates give a good Surface-area to volume ratio and the flat parallel plates minimally restrict air flow. The fractal tree design would choke the HS. The heat would be trapped inside of the mass of "branches" and ruin the efficiency.</div>

</td>
</tr>
</table>
</div>Absolutely... My only counter to this would be the branch multiplier and length ratio... In the picture it is 3x ... maybe 2x would be better?<br />
<br />
Either way, you are right... <br />
<br />
...revision thoughts pending....
 

·
Registered
Joined
·
63 Posts
Discussion Starter #26
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>FrankenPC</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391580#post1391580" rel="nofollow"><img class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" border="0" alt="View Post" /></a>
</div>
<div style="font-style:italic">I've often wondered if you could use a low temp vaporizing liquid in a heat pipe assembeled in such a way the the copper block has two stages. On the bottom is the "heater" which vaporizes the liquid...that gas travel's up several heat pipes into a condenser which then lets the condensed liquid flow back down a separate set of tubes into a insulated block (plastic) on the top of the copper block with a one way valve to allow the condensed liquid back onto the CPU block. So, you would have a simple circulation system power by expanding gasses and not a pump.</div>

</td>
</tr>
</table>
</div>Brilliant idea.... <img src="/images/smilies/thumb.gif" border="0" alt="" title="Thumb" class="inlineimg" /> ...
 

·
Premium Member
Joined
·
5,282 Posts
After some calculus I found out that cylinders maximize surface area better than plates - so maybe a design with many parallel cylinders would be better than many parallel plates - however the fractal design still chokes air flow - changing a variable or two will not fix that.<br />
I don't see why the fact that it is fractal would improve the design at all. There are better designs than the fins, but fractal is not one of them. It is by nature a very complex (hard to manufacture) and dense (restricts air flow) shape - not good for heatsink design
 

·
Premium Member
Joined
·
8,095 Posts
it's all about surface area....hence the fins...<br />
<br />
it's a logical way to do it...I guess you could do "prongs"..but then you can't suspend them, and use heatpipes and stuff..
 

·
Registered
Joined
·
63 Posts
Discussion Starter #29
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>rabidgnome229</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391612#post1391612" rel="nofollow"><img class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" border="0" alt="View Post" /></a>
</div>
<div style="font-style:italic">After some calculus I found out that cylinders maximize surface area better than plates - so maybe a design with many parallel cylinders would be better than many parallel plates - however the fractal design still chokes air flow - changing a variable or two will not fix that.<br />
I don't see why the fact that it is fractal would improve the design at all. There are better designs than the fins, but fractal is not one of them. It is by nature a very complex (hard to manufacture) and dense (restricts air flow) shape - not good for heatsink design</div>

</td>
</tr>
</table>
</div>Hmm.. <br />
<br />
So... the next logical conclusion would be that a high performance passive/active heat sink is a balance between maximizing surface area and the actual aerodynamics of the formation itself.... hmm.. more pondering to do...
 

·
Premium Member
Joined
·
5,282 Posts
My recommendation would be a tuniq-tower style HS with rows of parallel tubes creating a rectangular area to replace the plates.<br />
<br />
A though occurs, tho. It is possible to fit more plates into a given volume than cylinders. My optimization was based on a constant volume of material - so given say 10 cm^3 of copper tubes would be the best design. What may be more useful would be to find the shape that maximizes surface area within a volume of space. Back to calc...<br />
<br />
P.S. - Just in case anybody was planning on actually creating a design my previous optimizations found that<br />
1 - given a volume V of material taking the form of a square plate length x and depth y the optimal (most surface area) dimension is x = V^(1/3)<br />
<br />
2 - given a volume V of material taking the form of a cylinder with radius r and length x the optimal dimension for r is given by r = V^(1/3)
 

·
Premium Member
Joined
·
22,477 Posts
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>rabidgnome229</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391658#post1391658" rel="nofollow"><img class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" border="0" alt="View Post" /></a>
</div>
<div style="font-style:italic">My recommendation would be a tuniq-tower style HS with rows of parallel tubes creating a rectangular area to replace the plates.<br />
<br />
A though occurs, tho. It is possible to fit more plates into a given volume than cylinders. My optimization was based on a constant volume of material - so given say 10 cm^3 of copper tubes would be the best design. What may be more useful would be to find the shape that maximizes surface area within a volume of space. Back to calc...<br />
<br />
P.S. - Just in case anybody was planning on actually creating a design my previous optimizations found that<br />
1 - given a volume V of material taking the form of a square plate length x and depth y the optimal (most surface area) dimension is x = V^(1/3)<br />
<br />
2 - given a volume V of material taking the form of a cylinder with radius r and length x the optimal dimension for r is given by r = V^(1/3)</div>

</td>
</tr>
</table>
</div> <br />
Good stuff there!!! A flattened cylinder might be a Happy compromize.....
 

·
Premium Member
Joined
·
5,282 Posts
NNNNOOOO!!! Why did you have to suggest that. Now I'm gonna make myself do calc with eliptical prisms. I hate conic sections <img src="/images/smilies/mad.gif" border="0" alt="" title="Mad" class="inlineimg" />
 

·
Registered
Joined
·
63 Posts
Discussion Starter #33
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>rabidgnome229</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391683#post1391683" rel="nofollow"><img class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" border="0" alt="View Post" /></a>
</div>
<div style="font-style:italic">NNNNOOOO!!! Why did you have to suggest that. Now I'm gonna make myself do calc with eliptical prisms. I hate conic sections <img src="/images/smilies/mad.gif" border="0" alt="" title="Mad" class="inlineimg" /></div>

</td>
</tr>
</table>
</div>Wouldn't we be moving into wing territory? Cool!
 

·
Premium Member
Joined
·
22,477 Posts
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>rabidgnome229</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391683#post1391683" rel="nofollow"><img class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" border="0" alt="View Post" /></a>
</div>
<div style="font-style:italic">NNNNOOOO!!! Why did you have to suggest that. Now I'm gonna make myself do calc with eliptical prisms. I hate conic sections <img src="/images/smilies/mad.gif" border="0" alt="" title="Mad" class="inlineimg" /></div>

</td>
</tr>
</table>
</div> <br />
....<img src="/images/smilies/tongue.gif" border="0" alt="" title="Stick Out Tongue" class="inlineimg" />......<img src="/images/smilies/biggrin.gif" border="0" alt="" title="Big Grin" class="inlineimg" />.......
 

·
Registered
Joined
·
63 Posts
Discussion Starter #35
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>Peritus</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391696#post1391696" rel="nofollow"><img class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" border="0" alt="View Post" /></a>
</div>
<div style="font-style:italic">Wouldn't we be moving into wing territory? Cool!</div>

</td>
</tr>
</table>
</div>How about heatpipe-capillary-based feathers...made of copper of course...<br />
<br />
<img src="http://bougie.servibec.net/CL_harfang/44.jpg" border="0" alt="" onload="NcodeImageResizer.createOn(this);" />
 

·
Registered
Joined
·
63 Posts
Discussion Starter #36
Trees and aerodynamics....<br />
<br />
<a href="http://www.nhm.ac.uk/nature-online/life/plants-fungi/Fathomseminar-treesmagnificentstructures/session4/no-plant-fathomseminar-trees-session4.html" target="_blank">http://www.nhm.ac.uk/nature-online/l...-session4.html</a>
 

·
Registered
Joined
·
63 Posts
Discussion Starter #37
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>rabidgnome229</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391658#post1391658" rel="nofollow"><img class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" border="0" alt="View Post" /></a>
</div>
<div style="font-style:italic">What may be more useful would be to find the shape that maximizes surface area within a volume of space. Back to calc...</div>

</td>
</tr>
</table>
</div>This seems illogical to me... I purposely ignored this constraint for the same reasons we spoke about earlier, "choking"... Wouldn't having them spaced out allow for more "room temp" or cooled air to cross the heat sink?<br />
<br />
The only exception, other than actual space constraints, that I can imagine would only occur, would be in an active heat sink .. with a fan..<br />
<br />
The reason being, that the fan has a limited area upon which to blow... In a passive heat sink, room air is more omni-present (if that combo of words is even possible)..
 

·
Registered
Joined
·
63 Posts
Discussion Starter #38
Here's a random picture that I thought interesting...<br />
<br />
<img src="http://www.implosionresearch.com/ccp51/media/images/product_detail/EMFLS-L.gif" border="0" alt="" onload="NcodeImageResizer.createOn(this);" /><br />
<br />
THE FOLLOWING IS FROM THIS LINK: <a href="http://www.coolingzone.com/Guest/News/NL_MAR_2002/Ake/Mar_Ake_2002.html" target="_blank">http://www.coolingzone.com/Guest/New..._Ake_2002.html</a><br />
"Conclusions<br />
<br />
A bypass is always associated with a performance loss.<br />
<br />
A porous body approach can be used to develop a reasonably successful bypass flow theory.<br />
<br />
A heat sink with a bypass is always optimised for one particular air velocity. The potential loss when used at another velocity is usually moderate.<br />
<br />
The fin thickness is a critical parameter and significant gains can often be made if it is optimised.<br />
<br />
Overviews of heat sink performance are very appreciated in the front-end design process on the condition that they are created fast enough to keep the discussions alive."
 

·
Registered
Joined
·
63 Posts
Discussion Starter #39
Somebody's studying trees.. <br />
<br />
<img src="http://www.tamabi.ac.jp/idd/shiro/cg/cac/tree.jpg" border="0" alt="" onload="NcodeImageResizer.createOn(this);" />
 

·
Registered
Joined
·
2,338 Posts
<div style="margin:20px; margin-top:5px; ">
<div class="smallfont" style="margin-bottom:2px">Quote:</div>
<table cellpadding="6" cellspacing="0" border="0" width="99%">
<tr>
<td class="alt2" style="border:1px inset">

<div>
Originally Posted by <strong>FrankenPC</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391580#post1391580" rel="nofollow"><img class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" border="0" alt="View Post" /></a>
</div>
<div style="font-style:italic">I've often wondered if you could use a low temp vaporizing liquid in a heat pipe assembeled in such a way the the copper block has two stages. On the bottom is the "heater" which vaporizes the liquid...that gas travel's up several heat pipes into a condenser which then lets the condensed liquid flow back down a separate set of tubes into a insulated block (plastic) on the top of the copper block with a one way valve to allow the condensed liquid back onto the CPU block. So, you would have a simple circulation system power by expanding gasses and not a pump.</div>

</td>
</tr>
</table>
</div>...that's what a refridgerator does. It has a vaporizor and a condensor and the fluid just runs around in a loop. It could be done, and probably fairly effeciently as long ask you kept the area small (only cooling 2 in square instead of 2 cubic yards...), but that's what phase is (i think)
 
21 - 40 of 57 Posts
Top