We have in the past used several types of optical probe lenses for delivering and collecting laser light to an experiment for laser velocimetry. When the test surface was in focus, however, the collected light would fill mostly the laser fiber rather than the collection fiber(s). We have designed, developed and used for 8 years nested-lens probe assemblies that solve this problem. Our first version used a commercial AR-coated glass achromat, which we cored to remove the inner fourth of its area. The core was then reinserted with its optical center offset from that of annulus by an amount slightly less than the separation between the laser and collector fibers. The laser and collector fibers are placed in contact with each other behind the lens and have NA values of 0.11 and 0.22, respectively. Because most of the collected light now focused on the collection fiber, this system was far superior to the single lens systems, but was laborious. For the last five years we used injection-molded acrylic aspheric nested lenses, which are inexpensive in quantity and require little labor to install into a probe. Only an azimuthal rotation and positioning of the fiber plane are needed to incorporate the plastic lens into a probe. Special ray-trace codes were written and used to design the lens, and many iterations by the molder were required to develop the injection processing parameters to produce a good lens, since it was thick for its diameter. These probes have real light collection efficiencies of 75% of theoretical, work well over a wide range of distances, with collection depths of field matching theory. The lenses can take 100 watts of pulsed power many times without damage, since the lens is designed so that reflections from the lens surface do not focus within the lens. The collection fiber size is designed to work with our manybeam velocimeter facility reported in a previous Congress, where the collection NA times collection fiber size exceeds the acceptance of the velocimeter. The Doppler-shifted light enters the collection fiber with angles between 0.11 and 0.2, with little light in the 0 to 0.11 NA region. However, the manybeam velocimeter uses just the light in the 0 to 0.11 NA range, except when we link two analyzer tables together. A slight amount of mode scrambling of the Doppler shifted light converts the light into a uniformly filled NA equals 0.2 angular range before entering the velocimeter analyzer table. We have expended seven hundred plastic nested lenses in various experiments. The most recent version of the fiber cable assembly will be shown. Six situations will be discussed where multiple reflected frequencies were observed in experiments, illustrating an advantage of the Fabry-Perot vs. the VISAR method.
|