Page 209 - Chalcogenide Glasses for Infrared Optics

IR Imaging Bundles Made fr om Chalcogenide Glass Fibers 185

In considering the use of the polyamide plastic coating as the clad-
ding for the fiber, it was concluded that in the 3- to 12-µm range, the
percentage of energy lost in the plastic cladding due to the evanescent
wave coupling would be small relative to the total energy in the band.
The conclusion was based on measured FTIR transmission of a 500-µm
fiber coated with the plastic. However, the FTIR measured transmission
of a 1-m-long C1 bundle containing 2500 plastic-coated 100-µm core
fibers revealed that almost all the 8- to 12-µm energy had been absorbed
by the plastic due to evanescent wave reflection at the core cladding
interface. What had not been taken into account was the fact that light
traveling down 1 m of a small-diameter fiber reflects off the wall many
more times than a fiber with a large diameter. The number of reflections
N is proportional to length L divided by the diameter D. It takes twenty-
five 100-µm core fibers to provide the same area for the incident light as
for a 500-µm core fiber. The total plastic area on the twenty-five 100-µm
fibers is 5 times the plastic area on the larger fiber. The polyamide plastic
is no substitute for a clad glass when making an imaging bundle with
small-diameter fiber. Better the fiber is unclad where the index at the air
3
interface is equal 1. We had also expressed concern that fiber lengths in
individual ribbons should be exactly the same, or optical distortion
might result. That turned out to be an unfounded fear.
Another misconception that was holding us back was the tensile
strength of the fibers. We had previously reported tensile strength of
1
our fibers based on bend-to-break measurements. Gradually we
realized that the values were unrealistic and not indicative of true
strength. We began breaking fibers in tension and then calculating
tensile strength using the fiber cross-sectional area. We developed a
value for C1 and C2 fibers calculated from breaking force for all three
core sizes (30, 20, and 10 mil) averaged together. The results were

C1 = 15,000 psi ± 20%
C2 = 7000 psi ± 20%
Our conclusions from these measurements were that we would have a
difficult time drawing small-diameter fibers, < 100 µm, based on the
small calculated force to break number. However, we tested more C1
glass fiber unclad but plastic-coated and were surprised at the results. 2
3.5 mil (90 µm) 116,000 psi
10.2 mil (258 µm) 45,000 psi
19.8 mil (502 µm) 36,000 psi
Of course the plastic cladding increased the values. What impressed us
was the dramatic increase with decreased core diameter. Our hopes
that we could make small-diameter fiber ribbons were increased. We
had described C2 fiber as being one-half as strong as C1. New
3
measurements for glass clad C2 fibers, 90-µm core, yielded a value of