Procedures for performing more conceptually broad activities that don't fall in a single category, like "model to print workflow" or "achieving dimensional accuracy".

Going from Model to Print

flickr:17065122323
Illustration by Henrik Kniberg
  1. Create or obtain a 3D model that you wish to print
  2. Convert the model to STL format
  3. Evaluate slice surface area for printability
  4. Rotate model to reduce surface area (if necessary)
  5. Hollow model to reduce surface area and add drain holes (if necessary)
  6. Evaluate overhangs for printability
  7. Create supports to handle overhangs (if necessary)
  8. Lay out model on build plate
  9. Determine and set XY and Z resolution
  10. Slice into PNG files
  11. Obtain desired resin for printing
  12. Move projector height to necessary distance for XY resolution (if necessary)
  13. Validate and correct (if necessary) focus
  14. Calibrate the build plate level and zero height
  15. Determine necessary exposure and lift settings for resin and part
  16. Set printing software to match Z resolution from slicing
  17. Set exposure and lift settings for resin and part
  18. Verify projection grid size and output
  19. Add resin to vat
  20. Move build plate to Z=0
  21. Run print sequence
  22. Rinse part in IPA
  23. UV cure part
  24. Remove part from build plate

Tuning for Dimensional Accuracy

Calibration grid, test part, measure positive space, measure negative space, …

Determining Settings for a Print

If you are starting from scratch (new part, unknown resin), here are some things to think about when trying to figure out print settings

  • Exposure time - primarily determined by three things:
    • Amount of photoinitiator in the resin mix by the manufacturer
      • For a resin that supports DLP printing, if they supply an exposure time for a given XY resolution for a printer with a similar projector (like the B9Creator) that would be a good starting value
      • Otherwise, there will have to be guesswork. I purchased Venus Creator from the manufacturer in Taiwan with no instructions. When I e-mailed them asking for DLP settings for the Titan 1 (or any DLP printer with an Acer 6510), they only said "we think less than 5 seconds". So I tested using a small 1.8mL test part, starting at 5 seconds, then 3 seconds, then 1 second. At 1 second, the part still formed, but was a bit soft, so I settled on around 1.5 seconds.
    • Balance between hardness(and warping) versus vat release force
      • The shorter you cure, the less each layer will stick to the film on the vat floor. This will reduce the amount of lift height you need, and reduce the per-layer separation stress on the part.
      • The longer you cure, the harder the part will be and the less it will change in shape after UV exposure. This in general means that longer layer cure time will result in less overall part warping.
      • Thus, you are balancing two different types of defects for printed models. You will get better surface finish if the individual layers are exposed to less separation force (e.g. less cure), but the overall part's shape will be more consistent (less warp due to resin shrinkage) if you cure each layer longer (each layer will be more fully cured in the desired shape and less likely to form internal stresses over time).
  • Lift height - primarily determined by 2 things
    • Location of print in the vat
      • This is the smaller of the 2 concerns, but in general a part that is closer to the fixed end of the vat will require less lift (but be exposed to sharper separation forces). A part that is closer to the flexing end of the vat will require more lift (but be exposed to more gradual separation forces).
      • There is also a balance here. The fixed end of the vat pulls the part more straight up, and thus suddenly pops the part off the vat floor. That means the part is exposed to a sudden "upward pull" shock, but experiences less bending. The floating end of the vat peels the part more gradually, slowly separating the part off the vat flow. This means the part is exposed to a more gentle separation, but experiences more bending.
      • Problems relating to the "upward pull" shock will lead to print failure (e.g. part pulled off the build plate). Problems relating to bending are more likely to be slight imperfections in flatness (mainly on softer cure settings. if you cure long and hard, this likely won't happen)
    • Per-layer separation force
      • This is the more complicated and important issue because this will change at various points in the part.
      • Harder (longer) cure times will increase the per-layer separation force, requiring more lift
      • Larger per-layer surface area will increase the per layer separation force, requiring more lift
      • A really hard cure across the entire build plate will need 8-10mm of lift. A soft cure on a low-surface area part like the chiral sphere might require less than 1mm of lift.
      • I've never encountered the need to lift more than 10mm
  • Lift speed
    • Slower lifts reduce the risk of separation-related flaws or failures, but faster lifts complete faster.
    • During the earlier part of the print process, especially when the build plate is under the resin, lifting too fast can cause print failure.
    • Generally 25mm/m or less is recommended for the first 4-10mm, then 50mm/m or less for the remainder. For large surface area parts, I often use as little as 10mm/m throughout the entire print.

Delay time is mostly unimportant for the Titan 1, because it does not have a mechanical sweeper system that moves the vat in the X or Y dimensions. The lowering of the build plate will push resin back into position, so we don't need to wait for gravity to settle the resin. It is possible that if you have a really viscous resin (like Form 1 clear which is supposedly as thick as honey), you may need more delay. However, my general experience has been that increasing this setting doesn't have any noticeable effect. The main idea here is that whenever you push or move the resin, it takes time for the resin to react and finish moving. The more viscous the resin, the longer this will take, and the more delay you need. In practice, the design of the Titan 1 eliminates most settling issues.

Lowering speed also large doesn't matter, I've used both 150mm/m and 200mm/m for all layers, and not seen any difference.

Different resins also have different ranges of "base" separation forces - meaning some resins are just naturally stickier than others. That means that there is a factor for all of the properties, where separation force matters, that varies by resin.