Moon Lamp – Failures

For a Christmas present, I wanted to print a moon enclosure for a lamp. There are a number of variations on textured moons, and different scales and configurations depending on the hardware different makers used to create a glowing light effect in the lamp. These are all fantastic. This isn’t about finishing one of those. This is about completely failing to modifiy one of the more popular, and large scale, moon models into a form suitable for a lamp.

A full scale print of the Dexter’s Moon model won’t fit within the print volume of the Prusa i3 m2, and reasonable scale versions at high quality could take days to print. I wanted to make sure the model and materials were going to work well, so I took an iterative approach; first printing a small scale version with the various settings I hope will work for a high quality and durability print.

Starting Small

The 15% version with consistent height layers turned out fine – too small for detail to show through, but it confirms the need for better quality layers at the top of the print, where the curvature of the sphere makes even high quality layers stretch out and look bad.

Experimenting with Variable Height

The 50% version with variable height layers on the top and bottom failed twice – layer adhesion seemed to be an issue, and the model portions kept pulling away from the support layer, catching on the PINDA probe and tearing.

Eventually, printing a 50% scale version with variable layer heights for the top only turned out fairly well; the north pole of the moon was a bit thin, and needs a few extra layers, and “thin walls” detection needs to be turned on, otherwise quite a nice print. The final version will need to scale up just a bit more.

Better Layers at the Poles

Ok, a 25% scale test with Cura, to see if we get better top layers.

Ugly tearing, but better quality at the poles. One approach to improve the north pole layers would be to reslice with an inverted moon, to get the surface quality from the base, and not worry about hte quality of the top part of the print (the bottom of the moon), which will get removed with the final build. The bottom layers have plenty of support and typically print much cleaner than the top layers.

Final Scale

Scaling models is funny; sometimes problems only show up at very specific scales, where multiples of layer heights don’t quite line up, leading to, among other things, entire sections of the model not adhering. In the case of the Dexter’s Moon model, at 75% scale and 0.20mm layers, a gap appeared in the sliced model, about 15% of the way up the moon. The gap was big enough that the layers above the gap didn’t adhere to the layers below the gap. After three failed prints, with the same characteristics, at this scale, I opted to use a different model that was pre-scaled and remodeled for the specific purpose of making a glowing moon lamp.

Chalk this up as a 20+ hour failure.

  • Maker: Dexter_New_Materials
  • Source: The moon
  • License: Creative Commons – Attribution – Non-commercial
  • Model: Dexters_Moon_v2
  • GCode: –
  • Materials:
    • Prusa White PLA 1.75mm
    • Inland Gray ABS 1.75mm
  • Print Date: 2017/11/20 – 2017/11/25
  • Print Time:
    • 15% scale test: 1 hour 8 minutes
    • 25% scale test (ABS): 1 hour 2 minutes
    • 50% scale test: 6 hours 21 minutes
    • 75% scale: multiple aborted attempts
  • Estimated Time:
    • 15% scale test: 1 hour 1 minute
    • 25% scale test (ABS): 1 hour 6 minutes
    • 50% scale test: 6 hours 1 minute
    • 75% scale: 10 hours 23 minutes
  • Estimated Filament:
    • 15% scale test: 1.6 meters
    • 25% scale test (ABS): 3.38 meters
    • 50% scale test: 22.1 meters
    • 75% scale: 35.79 meters
  • Slicer: Slic3r/PrusaControl
  • Attempted Slicer Settings:
    • various scales
      • 0.20mm layers
      • 0% infill
      • include supports (zig zag or pillars)
      • variable quality layering for top with larger prints
      • detect thin walls
      • extra perimeters if needed
      • top horizontal layers: 5
    • Upside Down
      • 4 perimeters
      • 75% scale
      • detect thin walls
      • extra perimeters if needed
      • 4 bottom horizontal layers
      • no infill
      • 10mm brim
      • 0.20mm layers
      • supports from print bed
      • pillar supports
  • Printer: Prusa i3 mk2

One Comment Add yours

Leave a Reply

Your email address will not be published. Required fields are marked *