NavList:

The Kollsman MA-1 sextant has been brought to a functional condition after a refill of the mirror unit. Here are some of the actions taken...

After optical modeling supported a hypothesis that the mirror unit was underfilled or empty, the two fill plugs were removed and very little fluid dribbled out. The plug seals were replaced with -001 BUNA-N orings. To determine that no major leaks were present, a tube fitting to transition from the #6-40 UNF fill port to 5/32 OD x 3/32 ID urethane tubing was fabricated. Not knowing how much pressure the unit was designed to withstand, an initial leakdown test was done with an inflated balloon as a pressure source (thus avoiding direct connection to my air compressor/regulator). A Magnahelic differential pressure gage showed an initial balloon pressure of 20" H20. After the balloon did not immediately deflate, the tubing from balloon was clamped off. The pressure in the mirror unit and gage then decayed from 20" to 10" over the next 24 hours indicating a minor leak somewhere.

After consulting pressure vs altitude charts, and realizing this equipment could be flown in unpressurized aircraft to 30,000 plus feet, I realized I could be more aggressive with the test pressure! A 28 hour pressure decay test showed a drop from 75" H20 to 35". When the test was done with the lens facing down, some fluid was seen wetting out the area adjacent to the lens. The unit was cleaned and tested repeatedly for various durations to eventually find that the fluid was emanating from both lens sealant injection points and not from around the exterior lens surface. These injection holes are about .050" diameter x .050" deep. Given that the silicone oil is a great mold release, I didn't expect to get any adhesive to stick. I also wanted any modification to be reversible. What I needed was mm sized rubber stoppers. After checking that thermoplastic polyurethane (TPU) was compatible with the fluid, I obtained a short length of 1.75 mm dia NinjaFlex(TM) TPU 3D printing filament. I heated a section of this with a hot air gun while applying tension to produce a necked down section. I then cut this with a razor blade to get a diameter that just fit the injection hole and didn't bottom out. I then cut off the tapered "stopper" about 1.5mm long. With tweezers this was placed over the port and pressed in with the flat side of a screwdriver. I lost several to the carpet!

Another 28 hour test showed an almost identical pressure decay, but the lens end of the mirror unit was now dry (there never was any sign of leak at the bellows end). Suspecting the remaining slow leak could be in the test apparatus, the test was repeated with the tubing connection at the lowest point. Fluid backing up the tubing was detected confirming a small leak in the test equipment (which had several junctions). Repressurizing the test system showed the fluid returning to the mirror cavity just as the pressure reached the initial value.

The overhaul manual instructs to retract the bellows to a specified distance while filling. This is to maintain positve pressure on the fluid at altitude after the chamber is sealed and the spring loaded bellows is released. The dilema I have is that the bellows is already in the specified position while in its free state. I chose not to preload the bellows at this time. I filled the chamber with a syringe coupled to my tube fitting at the lower fill port and added fluid until all air was expelled from the upper port. It took about 16cc to fill. Test of the focal length of the mirror unit showed the anticipated increase compared to the partially filled state.

After cleaning up the spillage the mirror unit was reinstalled in the sextant. Power was applied to the lamp and I was rewarded with a sharply focused horizon reticle! The mirror response to sextant motion is now "over damped" (in the mechanical dynamics sense that the mirror doesn't overshoot the final rest position).