I've been kind of jumping all over the last week. I sanded down the exterior of the secondary hull sides and top and got the walls behind the windows thinned down on one side. I then turned my attention to the warp nacelle lighting. I did a quick test of an acrylic diffuser and the DotStar LED strip, and it worked amazingly well, even considering the fabrication of the test article was extremely rough. I started on the "production" diffusers and got one of them knocked out. I need to pick up some more aluminum tape and decide how I'm going to permanently attach the LED strip.
I also experimented with filling in the grid lines on my test mule secondary hull bottom. I used Perfect Plastic Putty and it was not robust enough to hold up to the scribing. I may give it one more shot with some epoxy putty, or I may just leave the grid lines alone (the softness bugs me, but it may be more work than it's worth and the potential for disaster is high).
Finally, I played around with some of the secondary lighting systems, some things were a success and some were failures. On the success side, I bought some 1.2mm fiber optic strand which fits perfectly in the existing holes for the nav lights. When the ends are rounded and sanded, they make excellent "bulbs". On the failure side, I tried several different methods of lighting the shuttle bay and arboretum, none of which were entirely satisfying. I'm ditching 12V COB strip because I don't want the control and added power difficulties. The RGB DotStars have a light quality issues with white, the RGB mixing always looks a touch off and in places like the shuttle bay and arboretum, you can see the separate emitters, even with a hefty amount of diffusion. I also tried some small Adafruit backlight modules, but they're not bright enough for the shuttle bay (they work OK in the arboretum. Because of these problems, I'm ditching the DotStars for the window and interior lighting. I ordered some RGBW Neopixels, which should solve the white light problem while still giving me the programmability that I want. I will have to run one separate data line for the Neopixels, but they can use the same +5V and ground as the DotStars (and I have plenty of GPIO pins on the Arduino to control them). I'll run some more experiments when I receive them.
Starboard side (thinned) vs. port side (unthinned) secondary hull halves. This is a good example of the difference thinning the walls down makes.
Back side of the starboard hull showing the ground out areas.
Diffuser test. Even with the unpainted grilles and no light blocking it looks really good and is more even to the eye than the picture shows.
Test light module. Again, it looks even better in person.
Test module with the light off. The back and edges are covered in metal foil.
"Production" version of the diffuser. I made them longer and drilled holes to use the glue posts to solidly mount them. I also sanded and polished the edges down to 7000 grit to help eliminate any hot spots from the rough edge tooling on my test. The diffusion side is sanded to 800 grit, and I may spray it with some glass frosting paint (Rustoleum from a rattle can).
Snug in place using the glue posts.
Close up of the polished edge. "Smoove"
Close up of the mounting. I trimmed back the glue post gussets by the width of the diffuser, they make a nice solid stop to positively locate it. The gussets on the other pin sandwich the diffuser in.
Test of the LED backlight module to light the arboretum. The brightness is acceptable here, but it's not enough in the shuttle bay, so I don't think it's worth the trouble to add in another new light type (plus it's not controllable other than PWM for brightness).
Bare backlight module. They come in a couple of larger sizes as well. Even though they don't work for me for this application, they'll stay in my stash and my brain for potential use in other projects.