Overclock.net banner

1 - 20 of 57 Posts

·
Registered
Joined
·
63 Posts
Discussion Starter #1
I'm no expert, though I wish I was, or at least I could speak to one......<br><br>
Anyways...<br><br>
I've been looking at different heat sinks, anywhere I see one (i.e. on things other than computers)....<br><br>
I'm of the opinion that heat sinks are generally finned because of a size constraint (i.e. commercial application)..<br><br>
However, given the following constraints/considerations, what is the most efficient heat sink design/formation (that man is aware of)?<br><br>
Copper construction<br><br>
Limited amount of metal (i.e. 5 lb of copper)<br><br>
Stationary design (as in, not moving or any huge risk of being blown over)<br><br>
Passive (no fan involved)<br><br>
Must support itself under the force of gravity (doesn't need any structural support, other than the primary metal itself)<br><br>
No budget constraint (i.e. doesn't have to be built in a commercially efficient manner and/or on a assembly line)<br><br>
No size constraint (i.e. MAKE AS MUCH OF THE 5 lbs OF COPPER COME IN CONTACT WITH THE ROOM'S AIR)<br><br>
Must be heat sinking a single central heat source (i.e. my horizontally mounted CPU)<br><br><br>
My only conclusion is one that is nature-inspired... and that is:<br><br>
A copper tree with a square footprint (proportional to the size of the CPU's chassis)<br><br>
Think fractal tree...<br><br>
Can anyone help me disprove this?<br><br>
Thanks....<br><br><img alt="" src="http://www.littlestick.com/Pictures/TreesBig/DenseTree.jpg" style="border:0px solid;">
 

·
Registered
Joined
·
2,338 Posts
i would think you would use a lot of your copper making your base....... it makes sense to a degree, but at this point, you have to have a lot of heat travel up the "trunk" and through the branches to become dissipated, maybe if you had a bunch of tiny trees, but at that point, you may as well go back to a lot of fins, maybe fanned out so more air can get through them.<br><br>
just imo <img alt="" class="inlineimg" src="/images/smilies/thumb.gif" style="border:0px solid;" title="Thumb">
 

·
Registered
Joined
·
63 Posts
Discussion Starter #3
<div style="margin:20px;margin-top:5px;">
<div class="smallfont" style="margin-bottom:2px;">Quote:</div>
<table border="0" cellpadding="6" cellspacing="0" width="99%"><tr><td class="alt2" style="border:1px inset;">
<div>Originally Posted by <strong>ae804</strong> <a href="showthread.php?s=5f012238fa36186cbb85825ea8d4fba8&p=1391220#post1391220"><img alt="View Post" class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" style="border:0px solid;"></a></div>
<div style="font-style:italic;">i would think you would use a lot of your copper making your base....... it makes sense to a degree, but at this point, you have to have a lot of heat travel up the "trunk" and through the branches to become dissipated, maybe if you had a bunch of tiny trees, but at that point, you may as well go back to a lot of fins, maybe fanned out so more air can get through them.<br><br>
just imo <img alt="" class="inlineimg" src="/images/smilies/thumb.gif" style="border:0px solid;" title="Thumb"></div>
</td>
</tr></table></div>
<br>
Sounds valid to me... hmmm....
 

·
Premium Member
Joined
·
4,746 Posts
That is an interesting idea...<br><br>
I'm sure if someone made a tree-type HS with a flat head and a fan mounted on top of it, it would perform very well. However, I think it's too far from the norm to really be considered for most companies. I personally thought Swiftech was gonna have trouble when they came out with the MCX series HS's (<a href="http://www.newegg.com/Product/Product.asp?Item=N82E16835108062)" target="_blank">http://www.newegg.com/Product/Produc...2E16835108062)</a>, even though they seemed well-deigned and looked to offer decent performance.<br><br>
You provide something to think about though <img alt="" class="inlineimg" src="/images/smilies/cool.gif" style="border:0px solid;" title="Cool">
 

·
Registered
Joined
·
63 Posts
Discussion Starter #5
<div style="margin:20px;margin-top:5px;">
<div class="smallfont" style="margin-bottom:2px;">Quote:</div>
<table border="0" cellpadding="6" cellspacing="0" width="99%"><tr><td class="alt2" style="border:1px inset;">
<div>Originally Posted by <strong>kidwolf909</strong> <a href="showthread.php?s=5f012238fa36186cbb85825ea8d4fba8&p=1391231#post1391231"><img alt="View Post" class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" style="border:0px solid;"></a></div>
<div style="font-style:italic;">That is an interesting idea...<br><br>
I'm sure if someone made a tree-type HS with a flat head and a fan mounted on top of it, it would perform very well. However, I think it's too far from the norm to really be considered for most companies. I personally thought Swiftech was gonna have trouble when they came out with the MCX series HS's (<a href="http://www.newegg.com/Product/Product.asp?Item=N82E16835108062)" target="_blank">http://www.newegg.com/Product/Produc...2E16835108062)</a>, even though they seemed well-deigned and looked to offer decent performance.<br><br>
You provide something to think about though <img alt="" class="inlineimg" src="/images/smilies/cool.gif" style="border:0px solid;" title="Cool"></div>
</td>
</tr></table></div>
WOW! Those are cool! And thanks for the comment...<br><br>
I was thinking of making almost an art project out of it... I mean how many unused scraps of copper wire could one find if they really tried? And the tree could just continue to grow over the years... It'd be called CPU Banzai or something .lol
 

·
Registered
Joined
·
63 Posts
Discussion Starter #6
Cool!<br><a href="http://www.josechu.com/moving_fractal/index.htm" target="_blank">http://www.josechu.com/moving_fractal/index.htm</a><br><img alt="" src="http://www.josechu.com/moving_fractal/moving_fractal_tree_phi.gif" style="border:0px solid;">
 

·
Premium Member
Joined
·
22,477 Posts
Hey take a look at this, it's basically the tree design...<a href="http://www.spodesabode.com/content/article/twist" target="_blank">Spode's Abode - The Twist Waterblock</a>
 

·
Premium Member
Joined
·
5,809 Posts
Thats a good design, but you would have soo much mass in the base of the tree design that it would weigh a ton! And i feel that for the size of it, a pinned block would do better. A simple block such as one off of a vga card seems to be optimal. It has maximum airflow over all of the fins, unlike the tree design where there will be "dead spots" of air flow du to some fins blocking the inner ones.<br><br>
And another thing... just how in the hell would to manufacture that tree desighn without it costing a fortune!?
 

·
Registered
Joined
·
63 Posts
Discussion Starter #9
<div style="margin:20px;margin-top:5px;">
<div class="smallfont" style="margin-bottom:2px;">Quote:</div>
<table border="0" cellpadding="6" cellspacing="0" width="99%"><tr><td class="alt2" style="border:1px inset;">
<div>Originally Posted by <strong>ira-k</strong> <a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391423#post1391423"><img alt="View Post" class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" style="border:0px solid;"></a></div>
<div style="font-style:italic;">Hey take a look at this, it's basically the tree design...<a href="http://www.spodesabode.com/content/article/twist" target="_blank">Spode's Abode - The Twist Waterblock</a></div>
</td>
</tr></table></div>
Sweet <img alt="" class="inlineimg" src="/images/smilies/smile.gif" style="border:0px solid;" title="Smile">
 

·
Registered
Joined
·
63 Posts
Discussion Starter #10
<div style="margin:20px;margin-top:5px;">
<div class="smallfont" style="margin-bottom:2px;">Quote:</div>
<table border="0" cellpadding="6" cellspacing="0" width="99%"><tr><td class="alt2" style="border:1px inset;">
<div>Originally Posted by <strong>pbasil1</strong> <a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391424#post1391424"><img alt="View Post" class="inlineimg" src="http://static.overclock.net//img/forum/go_quote.gif" style="border:0px solid;"></a></div>
<div style="font-style:italic;">Thats a good design, but you would have soo much mass in the base of the tree design that it would weigh a ton! And i feel that for the size of it, a pinned block would do better. A simple block such as one off of a vga card seems to be optimal. It has maximum airflow over all of the fins, unlike the tree design where there will be "dead spots" of air flow du to some fins blocking the inner ones.<br><br>
And another thing... just how in the hell would to manufacture that tree desighn without it costing a fortune!?</div>
</td>
</tr></table></div>
I'm not sure that I understand what design you are referring to by "pinned block"... but You are right, there would be blockage and it would be expensive to create....<br><br>
Of course, you referred to maximum air flow... and this heat sink's function would be passive and omni-radiance... Not directional wind cooling... Hmmm...<br><br>
However, they are good points, indeed<br><br>
Thanks for your help
 

·
Premium Member
Joined
·
2,687 Posts
<a href="http://www.physorg.com/news81096760.html" target="_blank">IBM thinks so.</a> Read it, and while its not exactly what you are talking about its close. Any way the 3 biggest things that affect heatsink performance are how well the heat gets from the cpu to the heatsink (thermal paste or other), total surface area, and how well the material absorbs, conducts and gives off heat. copper is really good at absorbing and conducting heat, but not nearly good as aluminum at giving it off. Thats why we are seeing more of these hybrid Al/Cu towers now. There are other materials that are better than copper, silver for instance. Also an active heatsink is better than a passive, as the overall goal is to get the heat into the air. A fan will provide more cool air for the heat to go into, and it will get the hot air moving away form the heatsink.
 

·
Premium Member
Joined
·
5,809 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>Peritus</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391439#post1391439" 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'm not sure that I understand what design you are referring to by "pinned block"... but You are right, there would be blockage and it would be expensive to create....<br />
<br />
Of course, you referred to maximum air flow... and this heat sink's function would be passive and omni-radiance... Not directional wind cooling... Hmmm... <br />
<br />
However, they are good points, indeed<br />
<br />
Thanks for your help</div>

</td>
</tr>
</table>
</div>Oops i meant to say finned, not pinned. For a passive block, maybe it would be smart to have sort of a chimney over the cooler, whatever desighn it be, to create a natural convection effect from the heat rising, pulling cool air in from the bottom. I geuss for a strictly passive design, the tree idea should work pretty well, as long as the base/trunk of the tree isnt too tall. If its too big, it will obsorb alot of the heat, but wount have enough surface area to dissipate the heat fast enough, so the block will become saturated.
 

·
Premium Member
Joined
·
5,573 Posts
The bifurcated fin design is already proven in radial heat sinks. i.e.:<br />
<a href="http://www.frozencpu.com/products/1472/cpu-tta-29/Thermaltake_Blue_Orb_II_CPU_Heatsink_w_Blue_LED_Pentium_4_LGA775_AMD_754939940AM2_CL-P0257.html?tl=g40c14s52#blank" target="_blank">http://www.frozencpu.com/products/14...40c14s52#blank</a><br />
<br />
TT call's it "crotch finning"...but it's the same concept.<br />
<br />
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 />
But, what you theorize is a much more elgant solution. The problem being HOW to get that formation synthetically. I wonder if you could grow it in a charged copper solution using a seed of some type?
 

·
Registered
Joined
·
817 Posts
being a mechanical engineer I know a thing or 2 about heatsink design, most of the ideas above would need a fan to work. The biggest limitation in building larger heatsinks is getting the heat throu the heatsink. Modern heatsinks use heatpipes to transport the heat and then have small fins attached to the pipes. <br />
A 5lb block of copper could take 20min to even be warm to the touch on a CPU. Perfict for suicid runs but not for continued use. For fanless operation the best design would be like the Zalman Reserator1 with a solid copper core and radial fins traveling up the core. The heat disipated would rise (hot air rises) creating passive airflow. Careful not to crush your core if you build this thing.
 

·
Registered
Joined
·
63 Posts
Discussion Starter #15
<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>The Bartender Paradox</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391445#post1391445" 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"><a href="http://www.physorg.com/news81096760.html" target="_blank">IBM thinks so.</a> Read it, and while its not exactly what you are talking about its close. Any way the 3 biggest things that affect heatsink performance are how well the heat gets from the cpu to the heatsink (thermal paste or other), total surface area, and how well the material absorbs, conducts and gives off heat. copper is really good at absorbing and conducting heat, but not nearly good as aluminum at giving it off. Thats why we are seeing more of these hybrid Al/Cu towers now. There are other materials that are better than copper, silver for instance. Also an active heatsink is better than a passive, as the overall goal is to get the heat into the air. A fan will provide more cool air for the heat to go into, and it will get the hot air moving away form the heatsink.</div>

</td>
</tr>
</table>
</div>Great post.. Thanks!
 

·
Registered
Joined
·
63 Posts
Discussion Starter #16
<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>pbasil1</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391477#post1391477" 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">Oops i meant to say finned, not pinned. For a passive block, maybe it would be smart to have sort of a chimney over the cooler, whatever desighn it be, to create a natural convection effect from the heat rising, pulling cool air in from the bottom. I geuss for a strictly passive design, the tree idea should work pretty well, as long as the base/trunk of the tree isnt too tall. If its too big, it will obsorb alot of the heat, but wount have enough surface area to dissipate the heat fast enough, so the block will become saturated.</div>

</td>
</tr>
</table>
</div>Agreed... So.. The perfect fan shape, for the tree, might resemble this?<br />
<br />
<img src="http://mywebpages.comcast.net/All4Magic/iturbine.JPG" border="0" alt="" onload="NcodeImageResizer.createOn(this);" />
 

·
Registered
Joined
·
63 Posts
Discussion Starter #17
<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=1391481#post1391481" 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">The bifurcated fin design is already proven in radial heat sinks. i.e.:<br />
<a href="http://www.frozencpu.com/products/1472/cpu-tta-29/Thermaltake_Blue_Orb_II_CPU_Heatsink_w_Blue_LED_Pentium_4_LGA775_AMD_754939940AM2_CL-P0257.html?tl=g40c14s52#blank" target="_blank">http://www.frozencpu.com/products/14...40c14s52#blank</a><br />
<br />
TT call's it "crotch finning"...but it's the same concept.<br />
<br />
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 />
But, what you theorize is a much more elgant solution. The problem being HOW to get that formation synthetically. I wonder if you could grow it in a charged copper solution using a seed of some type?</div>

</td>
</tr>
</table>
</div>Sweet designs<br />
<br />
Thanks... Yeah, it would be hard to manufacture... Your charged copper solution sounds very cool... I'll have to look into that...
 

·
Premium Member
Joined
·
5,809 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>Peritus</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391530#post1391530" 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">Agreed... So.. The perfect fan shape, for the tree, might resemble this?<br />
<br />
<img src="http://mywebpages.comcast.net/All4Magic/iturbine.JPG" border="0" alt="" onload="NcodeImageResizer.createOn(this);" /></div>

</td>
</tr>
</table>
</div>Meh not really, just a simple tube over it with the bottom and top open would create a convection flow, just like BSman described, the tube or the cover would just improve this some. If your really looking to build this, ill have my dad sit down and brainstorm on a few designs, or hell he may know a few already. He is an engineer with many degrees, one being in aeronautics,and many others i cant seem to remember, but he was in charge of maintenece and design of some cooling applications at his new job, and some simlar things while he was a pilot in the navy.<br />
<br />
Right now his company is traning over trying to figure out how to cool a small computer control unit, that is the size of a graphing calulator... but the thing is, it going to be placed in the engine compartment of an Abrams tank. All is well while the motor is running because there is mass amounts of airflow from the cooling fans, but once the engines is shut off, no more airflow, and everything bakes....
 

·
Premium Member
Joined
·
5,573 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>Peritus</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391535#post1391535" 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">Sweet designs<br />
<br />
Thanks... Yeah, it would be hard to manufacture... Your charged copper solution sounds very cool... I'll have to look into that...</div>

</td>
</tr>
</table>
</div>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.
 

·
Registered
Joined
·
63 Posts
Discussion Starter #20
<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>BSman</strong>
<a href="showthread.php?s=f0ff622456a527c34b14aeab359ee59d&p=1391497#post1391497" 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">being a mechanical engineer I know a thing or 2 about heatsink design, most of the ideas above would need a fan to work. The biggest limitation in building larger heatsinks is getting the heat throu the heatsink. Modern heatsinks use heatpipes to transport the heat and then have small fins attached to the pipes. <br />
A 5lb block of copper could take 20min to even be warm to the touch on a CPU. Perfict for suicid runs but not for continued use. For fanless operation the best design would be like the Zalman Reserator1 with a solid copper core and radial fins traveling up the core. The heat disipated would rise (hot air rises) creating passive airflow. Careful not to crush your core if you build this thing.</div>

</td>
</tr>
</table>
</div>Thanks for the contribution...<br />
<br />
You bring up something I was already pondering.... Heat pipes...<br />
<br />
Okay.. so, within the limits of modern manufacturing... AKA what's the smallest diameter copper tube man can create?..<br />
<br />
So instead of the tree being a collection of wires, it's a fractal-faceted heat tube....<br />
<br />
Once you got to the branch hierarchy where man was unable to manufacture a smaller heat tube, you would go to wires (which hopefully would be able to be inserted into the end of the last level of tubing....<br />
<br />
In essence, you've just helped me MULTIPLIED the number of tree "levels".. AKA the 5 lbs of copper was now is used MUCH MORE EFFICIENTLY and is now is contact with MUCH MORE "open air"<br />
<br />
Hmm.. <br />
<br />
Oh yeah... the turbine, if used, would be motorized....
 
1 - 20 of 57 Posts
Top