Page 183 - Chalcogenide Glasses for Infrared Optics
P. 183
Glass Pr ocesses for Other Applications 159
samples 5 to 10 cm in length cut from homogeneous glass made using
our plate process, and we measured transmission with the infrared
spectrophotometer and the laser. The measurements at 10.6 µm
agreed with one another. AMI found that glass compounded in a
cylindrical tube in a rocking furnace and quenched in billet form was
not homogeneous. When slumped and extruded into rods, the rods
would not be homogeneous as well. The same problem would exist
in fibers drawn from the rod. Striations in the fiber would scatter the
laser light, lowering the measured transmission. Glass delivered as
specified to Codman, Galileo Electro-Optics, and Infrared Fiber Systems
was mostly in billet form. The cast glass rods from AMI were probably
not homogeneous as well. The decision was made that fiber had to be
drawn from homogeneous glass made in our plate process in order to
have a good chance of producing fiber matching the starting glass
transmission. In 1987, we notified Rutgers, Galileo Electro-Optics,
and Infrared Fiber Systems that we would only deliver glass preforms
produced in this manner. We were going to have to learn to draw our
own fiber in order to succeed.
One factor holding us back was the cost of a standard silicate fiber
drawing tower. Most were 30 ft tall and cost $100,000 to $150,000,
much too expensive for a small company with no fiber business dedi-
cated to paying the expense. Presumably, the height was such because
the fibers were drawn at high temperatures and at high speeds as
already mentioned. The fibers had to be cool before reaching the
drum where they were wound. Thinking all this over, we decided in
1988 to build our own. The height was only about 10 ft because the
fiber to be drawn was at a relatively low temperature and pulled at a
slow rate, a few meters per minute. Extra height was not needed. The
tower was constructed from channel iron welded together in our
shop. We decided that since the rod drawing process had not pro-
duced good results for others, it would not work for us either. We
decided to base our process on a cylinder of homogeneous glass from
our plate process placed in a metal cylinder with a hole in the bottom,
as much as 1 to 5 mm in diameter. The glass would be under inert gas
pressure, heated until the glass softened and sealed the bottom hole.
At the right glass viscosity, tweezers were used to pull a fiber slowly
through the hole out of the bottom. The diameter of the fiber would
be much smaller than the hole, so the surface of the fiber would remain
pristine, not touched by any surface. The AMI fiber drawing tower is
shown in Fig.7.3. The chamber is at the top with the fiber emerging
from the hole in the bottom. The fiber is carefully guided through the
two open split dies down to the slowly rotating drum and attached to
its surface using a small piece of tape. Rotation speed and position on
the drum are precisely controlled by the unit made by Sancliff, Inc. As
the fiber leaves the chamber, the diameter is continuously monitored
by a HeNe “laser mike.” The operator adjusts parameters of pressure,
glass temperature, and draw speed to produce the core diameter