3D printing (almost) perfect transparent objects

I can’t know how many readers know about 3D printing. If the beginning of this article is common knowledge for you skip paragraphs until you find something interesting.

Since 3D printing became available to hobbyist not so long time ago, people are trying more than simply printing solid colorful plastic objects. From flexible materials to self folding objects. One difficult task has been making transparent prints. Even though there are transparent materials, the nature of printing process itself cause rough surfaces, which breaks light and leads to suboptimal transparency. FDM (Fused filament fabrication) prints can be coated, smoothed with chemicals. This usually works with vase mode printing, where a thin 1-2 layer thickness objects. Solid objects are impossible to obtain glass like transparency. Often the printed lines have air pockets around them. These pockets have different breaking index than the plastic. Light beams break at interfaces with different breaking indices. This gives the material a turbid look. Filling up those spaces would have resulted in a true transparent material.

Turbidity of FDM prints

These air pockets are very difficult to avoid. Only thin walls can be penetrated with filling coatings such as polyurethane coating. However, thicker materials are much more difficult.

Things are different for SLA (stereolithography) printers. Laser SLA printers can print in continuous lines and MSLA (masked stereolithography) printers in smaller dots. It looks like as if the printed object is consisted of rounded cubes.

Structure of resin under light microscope with varying print settings

Longer exposure times and slimmer layer height can decrease pockets between these physical voxels. However, this is not really required, as resin remains in these pockets until curing and solidifies during the curing process. Still these objects aren’t clear yet. The reason for that is the surface. If you can coat the surface with a breaking index similar to the solid material smoothly, you can achieve transparency.

Polyurethane coating is used for vase mode FDM printing but that didn’t work for my thicker SLA prints very well. So I tried to coat them with, well…, resin, which somewhat worked. Instead of print-solidifying, I brushed some resin over the prints. Then cured, washed and cured again. Using this method I made this corona virus model.

Design adapted from gregbejtlich on Thingiverse: https://www.thingiverse.com/thing:4166787

This was printed in 2 parts (envelope) and another 2 parts (single strand genome in red).

Virus model cut

One problem with brushing is, that it creates bump over the surface and is highly susceptible to air bubbles. Photocurable resin is quite viscous. A drop of it over a surface doesn’t really spread evenly. Plus the brush introduces tiny air bubbles. Dipping into resin removes many tiny features. So this works with such structures but not with others, where a better solution is required.

If the object is flat, then it is much easier. All you need is a piece of glass. By pressing the object with some uncured resin against the glass and illuminating with UV from the bottom, you can create transparent surfaces.

Flat surface smoothing

One issue with this approach is to ensure perfect corners and edges. Some manual handling and smoothing may be required and can be labor intensive. I personally don’t like to damage resin structure as some hazardous trapped chemicals might be exposed. In such cases, I highly advice to clean your prints with a separate (from after print washing) IPA batch at the end.

If viscosity is the only problem and resin coating works, can we reduce the viscosity? My idea was to use isopropyl alcohol (IPA). If IPA is compatible for cleaning and after cleaning curing, then diluting resin with IPA shouldn’t be a problem. IPA is much less viscous and is quite volatile. The latter property is used to get rid of it before curing. Here is the workflow:

  1. After printing, cleaning and curing, dilute clear resin with IPA (1:2-1:4)
  2. Coat the model with the dilution (either dip in or pour over the object
  3. Let it drip for 5 minutes
  4. Carefully transfer the object into the curing device
  5. Wash with IPA and cure again
  6. Repeat 2-5 until satisfied

This has the advantage that resin spreads with the help of IPA much more evenly. IPA within the thin resin coat evaporates quickly. Much less details are lost thanks to the thin coating. One major drawback for me is that it takes awful amount of time and manual labor. I guess, one can automate this easily with 2 tanks of liquid (IPA and IPA-Resin mixture) within a wash and curing station.

So why did I go all this trouble?

First, I wanted to see if it is possible to obtain transparent materials for my future projects. Second, I wanted to make my own optics. My own small lenses for camera applications. I will post my tests on this regard.