In a message dated 10/31/02 2:52:07 PM, kenwilfy@COMCAST.NET writes:

<< Then I took the
Cibie lamp and siliconed it in place with this industrial silicon that we
had. >>

Oh yes, let's all run out and glue our headlamps in place.  People, this
design problem was worked out about 70 years ago.  The solution is called a
headlamp bucket.  If you can't fabricate one yourself, search the wrecking
yards for one that can be modified to serve your needs.

Many moons ago, I was charged with implementing a pilot program to establish
a baseline for vibrational analysis of the primary coolant loop at a nuclear
power plant.  The analysis was done during a stage of pre-fuel testing when
the primary coolant pumps were run to create enough steam, merely from flow
friction, to roll the turbine.

Because it is beaucoup verboten to weld to a nuclear piping system without
big time hassles, to attach the vibration transducers we were provided with
some of the most expensive, multi-pole, rare earth magnets ever devised by
man.  The magnets did their damnedest to hold the transducers in place, but
it soon became apparent that they needed some help.  The experts from GE
Industrial Adhesive were summoned and the attachment of the magnets was
bolstered by the best RTV they could provide.  This scheme failed, also,
which began a series of various surface conditioning and RTV glopping
experiments.

The boys from the university finally gave up, used up their grant money is
more like it, and went home saying that they had gleaned enough data from the
transducers which had hung on.  Sure.

My point?  Screw, don't glue.

George