Craig Markwardt replied to Jeff Root:

Note that the 20% figure is for the area of the RTG which is
visible to the back of the dish, etc., and is not the most
important thing to measure here, but was a response to Bruce
Woodcock's assertion that:

every RTG is not a spherical black body, but rather has fins
that are "edge on" to the antenna, which means only 2.5% of
the surface area of the RTG is actually facing the antenna.

My 20% figure may be high, but Bruce's 2.5% figure has *got*
to be *way* too low. Or else he mis-stated what the figure

I wouldn't be too quick to dismiss the 2.5% amount without a more
rigorous factual basis.

I'm not dismissing it-- I'm saying it looks wrong, so it should
be checked.

The majority portions of the RTGs which face the dish are the
endcaps, which are designed to be *non*-radiative.

That has very little effect on my analysis. I don't recall
taking into account that the ends were designed to not radiate,
but it was obvious that the area they expose to the dish and
body is much less than that of the sides and fins.

This leaves the edges of the radiator fins, which are pretty
thin indeed, compared to the total surface area.

Are you assuming that infrared from an RTG is radiated only
perpendicular to the surfaces? When I look at an RTG from near
the body of the spacecraft, I see a good deal of surface area
of the fins and of the cylinder itself. About 20% of the total
surface area, as I said.

Don't you agree that 5% looks more like it?

I have no idea. It's a realm where eyeballing it won't help.

I don't have any problem estimating the solid angle subtended
by the body of the spacecraft as seen from an RTG, based only
on what I can see.

-- Jeff, in Minneapolis

..