Only very few FDM materials are truly transparent. When printing a line next to line, inclusions of air can and will create outer boundaries that would be visible, so it is paramount to print in such a fashion that prints are as tight as possible. Another problem is, that under normal conditions the extrusion-extrusion boundaries are quite visible and that have different refraction indices inside the part. To remove those boundaries and gaps, it needs special printing parameters that result not only in filling every gap but also partially melt all adjacent lines to create a truly solid part. As such, FDM prints with multiple layers for walls are extremely hard to make fully transparent.
With extreme tuning, the internal voids can be reduced to almost none or truly none, and inter-wall bonding increased to be nigh invisible, but this leaves outer walls.
Extreme settings for clarity
Stefan from CNC-Kitchen had a whole talk about these specialty print parameters with which he achieved very low-opacity prints in some, but not always all directions. As a baseline, he used settings proposed by Rygar1432. He started using a very transparent PCTG, with already very good results, and then experimented with PETG, which had a slightly less transparent look to it after optimizing the settings.
The exact composition of the print material was found to have tremendous effects on the transparency, thus different manufacturers and even rolls from different batches might behave differently and require different settings. Further, some filaments absorb moisture over time, which will degrade the print result by creating voids of entrapped water vapor. As printers also heat slightly differently due to the variance of the heater cartridges' resistance, your settings might not be perfectly applicable to another printer of the same type and might require adjustment. The most important basic settings discussed were:
- extrusion width & layer height
- Compared to the standard 0.4 mm or 0.45 mm, a wider extrusion of 0.5 mm was used, but the layer height was significantly reduced to 0.12 mm at the same time, resulting in the plastic getting pressed tighter against the previous print and allowing fusion to other walls
- Using only a single outer perimeter made the print mainly composed of infill and reduced the different geometries that could create pockets
- printing speed & part cooling.
- Compared to classic prints, the printing speeds had to go way down to 15 mm/s, which is about 1/4 of typical print speeds. The slower print speed assists with keeping the extruded plastic hotter in the area of the nozzle, assisting with inter-extrusion bonding. This was further assisted by disabling part cooling. However, this resulted in very bad results of bridging material.
- Infill Geometry
- An aligned rectilinear infill pattern was necessary to remove almost all voids, which was coupled with no top and bottom surfaces to again, reduce different print geometries that could create voids in the print.
The most important factor was the extrusion multiplier, which for Stefan's machine was best at about 101 %, as above that he started to experience some deformation of the part's outer shape, mostly in the shape of rough surfaces.
A slight bit of polishing and cleanup was at times required, as Stefan retold. Also, again, the outer perimeter clearly shows in different alignments of the print, best shown with this shot. While you can see almost undisturbed through the part down like, sideways the outer perimeters are very opaque.
All in all, you could get a very well-transparent print in some directions, but the outer surfaces are always somewhat rough.
Some filaments can be treated after printing to create a single, well-fused outer surface that is nigh fully transparent. This precludes prints with many walls unless the interior is already clear, but the results are known to be stunning for prints with a single or few walls. Do note that the order of filaments mentioned here is by the toxicity of their solvents, so the higher up in the list, the less dangerous it is.
The Hourglass by Mageb uses a chemical quirk of polyvinyl butyral (PVB): it is soluble in isopropyl alcohol to a good degree, resulting in softening and smoothening of the surface if the right exposure time is used. It's hard to predict if and where the IPA can creep into the walls to remove cavities without destroying the print's shape, so this effect is generally best used on Vase-mode prints. This of course limits the wall thickness and strength of the print.
It's well known that Acrylonitrile butadiene styrene (ABS) dissolves in Acetone. In a similar fashion to PVB, Acetone liquid or vapors thus can be used to smooth out the outer deformities of a print and create a smooth surface, though not only is Acetone much more dangerous than Isopropanol due to toxicity, ABS itself is a more challenging printing material to PVB and can emit some dangerous fumes in printing. Since ABS has lost ist appeal to most hobbyists, I could not find any report on possible clouding issues from the treatment.
Yes, Polyethylene terephthalate glycol-modified [copolymer] PETG is chemically smoothable. However, the chemicals that can dissolve PETG are even more toxic than for ABS: Dichloromethane, methyl ethyl ketone (MEK), toluene, and cyclohexanone are extremely toxic and nobody would guarantee you that the print will stay clear and that the prints won't cloud from the chemicals penetrating the print.