A major component of my project for this summer
is the production of various optics (primarily collimating and focusing lenses, but also optical flats to be used later in the production of beamsplitters) which will work at wavelengths in the far infrared region. For reasons of cost and convenience, we decided to try and mold the necessary optics from TPX
resin. TPX is convenient because it is transparent to visible wavelengths as well as to far infrared - this means that we can do alignment work with visible
lasers (probably Helium-Neon lasers
), which makes the whole alignment process much easier. It's cost effective because it can be injection molded readily, which means that we can make the optics ourselves instead of paying someone lots of money to do it for us. To hear the manufacturer (Mitsui Chemicals
) tell it, we just need to get the mold, melt the plastic, and shoot it in there and we'll have nice transparent optics. Of course, it's not that simple.
It never is. My goal in this node is to describe a few of the snags we've hit, and our potential solutions.
Clearly, for any optical material to work, it has to be transparent. For plastics, this means specifically that (among other things), there must be no molded-in stresses. Techniques to avoid molding in stresses include adding more gates to the mold, venting the mold, and basically anything that can keep the flow lines relatively straight and parallel. Turbulence is the enemy here. Our first mold design was not at all ventilated, and had a single gate at the top of the mold. This design trapped air in the mold, which burned the plastic when compressed enough, rendering the resin opaque and useless. Additionally, the flow lines were swirly - there were eddies all over the place. This is not good mold design. We quickly added vents near the bottom of the mold (which is circular in shape), but kept the gate as it was. This improved the quality of the plastic drastically, but there were still swirling flow lines near the gate and far from the vent. Our next modification will be to add more ventilation and make the gate wider, which will hopefully cut out the appearance of flow lines entirely.
Another problem is that injection molding requires enormous pressure - we are currently working at 2250 PSI. This, of course, opens up a whole slew of mechanical problems. Our first mold had optics, backed by aluminum discs, which were in turn backed by 3 quarter-inch screws each. After a few shots, the screws buckled badly. We fixed that by using beefier screws, and shortening the distance over which they had to provide support. Another potential problem is finding optical surfaces which can withstand the pressure of injection molding. So far, we've only done flats, and they've stood up pretty well... but curved surfaces could be another story entirely. We shall see.
Yet another thing that can wreck a molding is thermal problems. Currently the TPX resin leaves the nozzle at 520 degrees Farenheit. The mold has to be at a uniform temperature, ideally around 150 degrees Farenheit, or the molding will shrink unevenly and warp. When we were using all aluminum, this was no problem - but our optical surfaces are glass, and they are pretty well thermally isolated from the rest of the mold. Right now the only way to get the whole mold up to temperature is to shoot a dummy molding, which heats up all the surfaces equally, and then hope the mold cools fairly evenly.
Those are the major issues we've encountered so far - I'll node more as they come up (and I'm sure they will...) Once we get this process down, though, I smell a patent... and I can corner the market on far-infrared molded optics! MUWAHAHAHAHA!