Strostkovy
Titanium
- Joined
- Oct 29, 2017
For conduction, the surface finish doesn't matter except for in extremely high airflows. The reason for this is the boundary layer (the air near the surface that doesn't move) is thicker than the surface features that provide the extra surface area. I spoke with a few engineers at a heatsink manufacturer and they told me that even the coarse wiggles some people put in heatsinks meant for passive cooling do absolutely nothing in thermal testing.Some further thoughts.
Radiation is only one method of heat transfer. There is also direct transfer from the surface to the surrounding air and then convection currents will carry it away. A rough surface will have more surface area and therefore would, in theory, transfer more heat that way. So a chrome surface, bright or black is not ideal for that. A rough surface, like sandpaper, would have more area and would transfer more heat. Again, the "flat" surface finish is probably better.
As for a layer that is under the final coat, it comes in play in the conduction of heat from the interior of the part to the exterior surface. This is kind of like a resistor in electronics, it opposes or facilitates the flow of the heat. But this is a different property than it's radiation properties. Those two properties may or may not track. Here the thickness of the layer also comes into play: the thicker it is the more it will block the conduction of heat. Also you should consider that the final, exterior coating must conduct the heat from the interior to it's own surface so the actual material of this layer is in play.
So long as the coatings are well bonded to the materials, they matter very little for conduction unless you have a highly optimized forced air setup. The coatings, being very thin and very wide, can conduct a lot of heat across them with minimal temperature rise, even if they are a fairly insulating material. It all ends up being negligible against the boundary layer.
Consider a block of PET plastic, at just 0.19 W/m K. This is really terrible thermal conductivity. (copper is 400 W/m K). Now apply it as a coating, 0.003" thick. That coating can dissipate 2500 watts per square meter of coating per degree Celsius of temperature rise. Running the numbers for any metal plating puts that number in the megawatts.