At 03:08 PM 11/23/99 -0800, Doug wrote:
>Um ...  It appears without any sort of rigorous analysis
>that there's a min/max at work here.  I.e., there'll
>come a point where too many holes won't help.  Loss of
>too much material and heat conduction paths being
>interfered with too many holes.

I did not attempt to maximize, by analysis, the surface area. The article I
mentioned showed the maximum cooling effect with the maximum hole size. I
don't think that "too many holes" would be a problem, until one had no
material strength left. For a given hole size, each additional hole *adds*
conduction path (that is, it replaces board material with a copper tube);
the idea that it interferes with conduction seems spurious to me. The holes
also make the second side effective, for which see below.

>Just a suggestion, but it might be worthwhile to specify
>the exact orientation of said board - horizontal or vertical.
>I'm *assuming* you mean horizontal.

I'd expect the effect of the holes to be increased with a horizontal board,
because, well hot air rises! But a vertical board should also see
substantial improvement, again, because of the effect of thermally
connecting the second side of the board.

> Now without proper
>convective paths, the bottom plate will do little as
>a second dissipative surface.

I don't think this is true. Suppose that the bottom of the board is closed
and air cannot flow. This air will heat, and where will the heat go? Unless
the bottom of the chassis is insulated, it will transfer to the chassis and
from there to the ambient air. The bottom might cool better than the top!
If air can flow across the bottom, it will because of the convection
induced by the heating. If air rises from the top of the board, it will
also pull air through the holes from the bottom.

It's hard to quantify these effects, but it would be easy to test them!

>Start adding holes, use
>an approx virtual hole of only 60% air flow through
>the hole (I think that's about right), and that's
>when air flow begins to make effect.  Start adding
>too many holes or holes that are too large and you
>may end up as I stated earlier.

I'd expect the problem to be with holes that were too small rather than too
large. There are two effects from the holes. One is to transfer heat from
the top of the board to the bottom of the board. This alone is a good
reason to use holes, if one can put plane on both sides of the board. The
other effect is to increase the exposed surface; but if the air does not
move across the exposed surface, this may not increase heat flow. The
smaller the hole the more likely that air will not move effectively, I
would think. This could why the research found better effect with larger
holes.

Considering the case mentioned also by Doug of two boards oriented
vertically, next to each other. Suppose we have the hole pattern I
described in the heat sink; suppose that the sides of the board facing each
other get hotter than the other side of the board. This would cause air
flow up across those surfaces, pulling air through the holes to some degree.

[log in to unmask]
Abdulrahman Lomax
P.O. Box 690
El Verano, CA 95433