QUESTION:
How was it possible to design the aero braking used to slow down 
the Probe as it entered the Jovian atmosphere when the atmosphere had
never been tested before?  How did you know the atmosphere's
uniformity and density?

ANSWER from Charlie Sobeck on January 25, 1996:
Clearly, as the question suggests, we could not fully test the Galileo
Probe's aerobraking/thermal protection performance in advance.  The
performance of the heat shield had to be inferred from a number of
partial tests.  The basic heat shield material had been used in less
severe conditions several times before (Apollo, Pioneer Venus,
military applications), and had been tested in special ground
facilities at NASA/Ames Research Center.  Additionally, several
reduced-scale tests were performed at Ames to identify the proper
shape of the heat shield.  These tests included firing a model of
the probe down a test range instrumented with cameras, where the
behavior of the atmosphere around the probe could be observed under
conditions similar (in some ways) to entry at Jupiter.  The results
of these various tests, were used to verify the predictions made by
detailed computer simulations.  When the computer models were verified,
they were then used to predict the performance of the Galileo heat shield
at Jupiter.  Finally, because of the uncertainty associated with
extrapolating from all these partial tests, a generous safety margin
was added to the heat shield.

At this point, the data we have gotten back from the Probe indicates
that our predictions were generally pretty good.  About half of the
heat shield seems to have burned away, as we expected.