Custom (very) Quark esc heatsink...

[zeropointbug]

Originally Posted by GriffinRU

even more complicated, because now you add gas to equation and gravity and direction and sensitivity to temperature range. What works or doesn't work for CPU is not true for every application. With heatpipes selecting correct fluid is very important, more important that pipe is long enough that fluid easily goes through all stages and there is some in liquid form not in gas othewise it stops working...

They use wicking in heatpipes now, the texture inside the pipe. Also, these things are very efficient in transferring heat to some other place (fins). gravity has little to no affect on heatpipes now.

Why do you think this is SO complicated? Heatsink design is trucky yes, but the basic understanding of heat transfer is simple.

[GriffinRU]

Originally Posted by zeropointbug

They use wicking in heatpipes now, the texture inside the pipe. Also, these things are very efficient in transferring heat to some other place (fins). gravity has little to no affect on heatpipes now.

Why do you think this is SO complicated? Heatsink design is trucky yes, but the basic understanding of heat transfer is simple.

Because, I've done some designs in the past and most of them didn't pass mil specs due to the little gravity effects and other heat related problems. And if it is simple then why there is no simple heatsink for CPU, but we see all kind of crazy designs flying around while copper and aluminum are still have the same properties with air included.
I didn't said that it is less efficient, I said it is complicated and has its limits due to its better efficiency in limited range of applications or temperatures.

[zeropointbug]

Well, to be honest, MOST heatsinks for CPU nowadays are just designs, that's all they are. Lot's just looking cool, decent cooling capabilities, nothing great. Most heatsink design is coming from PC industry now.

Like i said, heatsink design is tricky, as there are several ways you can make it perform: such as LOT'S of airflow, NOISY, and guaranteed to work, with moderate fin count. OR another extreme, like these large tower heatsinks, using heatpies densely around the CPU, then efficiently transferring the heat as vapor to a large set of fins. This method works amazingly well, and better uses a given surface area of fins, than a simple heatspreader, fins design.

What kind of experiment did you do?

[GriffinRU]

Originally Posted by zeropointbug

Well, to be honest, MOST heatsinks for CPU nowadays are just designs, that's all they are. Lot's just looking cool, decent cooling capabilities, nothing great. Most heatsink design is coming from PC industry now.

Like i said, heatsink design is tricky, as there are several ways you can make it perform: such as LOT'S of airflow, NOISY, and guaranteed to work, with moderate fin count. OR another extreme, like these large tower heatsinks, using heatpies densely around the CPU, then efficiently transferring the heat as vapor to a large set of fins. This method works amazingly well, and better uses a given surface area of fins, than a simple heatspreader, fins design.

What kind of experiment did you do?

Problems can be found here under "Limitations" and at the bottom check "See also" and "External Links" heat-pipe selection guide

As far experiment - you know the answer, right :)