3D Printing Transparent Parts Using FDM/FFF Printer

Ever since I got my 3D printer, I have been wondering whether it was possible to produce transparent parts. While there are plenty of transparent filaments sold on the market, they typically do not produce transparent prints, but rather translucent ones.

The difference between transparency and translucency is in the material’s tendency to diffuse light.  Diffusion may result from several factors, the most dominant one in our case is the inhomogeneity of the print. Under normal settings, the printer leaves small cavities (bubbles) inside the print. Each cavity acts as a small lens (as it is curved) and scatters the light passing through, obscuring any image standing behind the print and making it… translucent!

Having in mind that the plastic itself is inherently transparent and that bubbles from the printing process are the only thing preventing us from achieving optically clear prints, I began searching for the right print setting to accomplish the task.

I started by searching the internet for prior attempts, but all I could find was this YouTube video. It was a great start, but I wanted to achieve clearer results.

Let’s Start

In short, the basic settings using CURA are:

CURA settings for transparent parts
  • Layer Height:    0.05 mm
  • Infill Density:    100%
  • Infill overlap percentage:    6%
  • Flow:    108%
  • Print Speed:    24 mm/s
  • Enable Print Cooling:    OFF
  • Printing Temperature:    255 °C  (on the highest end of the recommended range, I was using prusa’s ABS)
  • Build Plate Temperature:    100 °C
  • Nozzle Diameter: 0.4mm
Infill overlap (percentage)

The main purpose of the setting above is to make the laid plastic’s consistency as thin as possible by heating it to the highest possible temperature, and then forcing it into the grooves of the previous layer by applying over-extrusion (flow > 100%).

(I found it especially difficult to eliminate bubbles at the interface between the infill and the outer wall. To deal with this I changed the infill overlap).

Before printing, make sure that the hotend is clean, as leftover material in the part can damage its purity. After the first layer has been laid down, the hotend must be as close to the build plate as possible, but not too close as it might get clogged or even scratch the build plate. To find the sweet spot, I found it best to adjust the “live Z” (on prusa MK3/2) to the point where the air gap between the tip of the hotend and the build plate just disappears. This rule of thumb ensures that the plastic is forced into any gaps between the first layer lines and the build plate.

Post-Processing

The down-side of over-extrusion can be seen in the video below, in which a “plano-convex spherical lens” (flat on one side, and curved on the other) is being printed. Notice that the hotend is scraping the last layer, and is pushing on a chunk of excess material resulting from over-extrusion. The chunk is more pronounced on the first layers and typically ends up on the surface of the model. Worry not, as it can be easily removed with a crafting knife.

The next step is to turn the rough surface into a smooth one. In my first attempt I used acetone smoothing, and indeed, the surface got a lot smoother, but it also made the part look undesirably foggy.

 

A print fogged by the acetone treatment

 

Another technique I’ve tried and found successful was dry sanding using sandpapers raging from 600 grit to 4000, incrementally. It’s worth noting that cardboard can also be used instead of a high grit polishing paper.

Finished print of a lens
Before and after post-processing

The reason why I’ve made all of my lenses flat-bottomed is that printing curved shapes would require hard-to-remove supports. Additionally, while sanding a flat surface can be done easily on a flat platform, such as a table, there is no easy way to smooth a curved surface, and hand-sanding it can distort its geometry.

The end result is quite pleasing!

Although the quality of these lenses is not good enough for imaging (e.g. for the use of cameras), they’re certainly capable of focusing light!

So Let’s Have Some Fun!

Plano-Convex cylindrical lens

And nothing says fun more than LASERS! (pew pew pew) I’ve made a plano-convex cylindrical lens and a mount for holding the lasers parallel to each other. You can see how the laser beams are all focused into one spot. Success!

 

Focusing lens with parallel lasers

Another neat idea I had was to try and make something similar to those glass blocks that are laser-engraved with 3D figures. I downloaded a model of my favorite character – Mario. Using Blender I placed the model inside a cube and used the tool ‘Boolean’ to produce a cavity with the figure of Mario inside the cube. The result was not perfect, but I find it pleasing and unique.

 

One thing to note is that slicing softwares don’t like internal cavities. They see floating spaces as an error and try to correct them automatically. In order to “convince” the software to leave my shape alone I had to connect at least a small part of it to the outer shell. With the Mario figure, I simply moved it down so the shoes will touch the bottom part of the cube. I couldn’t find any option in Cura or Slic3r to disable error correction, and would love to find a way to allow internal cavities.

Looking for something more complex to make, I found this little thing and printed it using the process described above (this time I was a bit lazy with the polishing).

Solid print with shaped cavity

Below are some more examples: the prism on the right produces a small rainbow and the big lens can focus sunlight from one side of the room to the other. The bottom right print is just a simple round window that turned out almost perfect, because, as I mentioned before, flat surfaces are the easiest to sand.

And last but not least: a Fennec fox ring

47 Replies to “3D Printing Transparent Parts Using FDM/FFF Printer”

  1. This is amazing, and thanks for the explanation and walkthrough. One question though, any special filament used? Brand? PLA? PETG?

    1. hey man. he was using ABS sold on prusa. though I doubt that it is much different for different brands.

  2. I’m going to try this with some of eSuns transparent PETG. This is awesome. Thanks for showcasing this.

  3. On the clouded one that you used acetone vapor, you might try dipping it in liquid floor wax. Model airplane builders do that to get crystal clear canopies.

    Give it a try.

  4. Time to make a telescope. At least it would be better than Galileo’s 🙂 It might even be really good at least to look through. The primary could even be a mirror if you applied a reflective surface. Then you could go big. Imagine a 100% 3D printed telescope. 🦊

    p.s. I’m not sure if I’m talking to you, everyone, or myself. 🙂 All seem reasonable.

    1. That sound awesome!! We are really looking forward to see new applications with transparent printing!

  5. So you leave the print fan disabled the entire time?

    Are the same results possible with PLA filament?

    1. Yes, the fan was off the whole time.
      I haven’t tried to use PLA but I believe that it’s possible. I heard that the results are better with ABS of PETG.

  6. You can get floating voids in Cura by just inverting the mesh normals of your interior mesh, then joining both the interior and exterior mesh together before exporting your STL.

    Flipping your mesh normals in Blender is discussed on this thread (https://blender.stackexchange.com/questions/29132/how-do-i-flip-normals-on-this-model), and the function you’re looking for is spelled out a little ways down the page.

    Hitting Ctrl+Shift+N when the mesh is selected and you’re in edit mode should do the trick for you.

  7. Trying PLA. On a Prusa Mk3 with Slic3r PE.

    110% flow (extrusion multiplier in Slic3r) seems best, so far. But the rectilinear pattern is still easily discernible.

    Any tips?

  8. I’m thinking about printing a 45°-45°-90° prism.
    Are there anything for me to concern about in terms of size of the prism? What minimum size do you suggest?

    1. I would suggest to start with a small print, because the slow print speed it will take a long time.

  9. Great stuff. Maybe a better way to make a lens would be to print the negative of the lens and then use an epoxy/polymer such as acrylic to make the optical component. Your method is obviously incredibly good at making detailed parts so this detail could be used to make an appropriate mold.

  10. I have used Taulman PETT which I bake for 12 to 15 minutes at 220 degrees Fahrenheit on a clear glass plate. Lightly coat the plate with silicone if you don’t want to have the part stick to the glass. It will shrink slightly and sharp edges will round over but the resulting part will be very clear. This technique does not work with eSun PETG.

  11. Excellent! Very nicely done, I hope to see more of this. It would be great to see this as a profile/template in slicers and more people can test it and maybe even improve.

    I have never used it before but there is also a top surface ironing feature, could that maybe help with straightening out every layer? It would take more time but you could adjust the speed back from 24 mm/s to a much higher value.

  12. This is awesome! Well done.

    How sensitive is this process to moisture in the filament?

    I bet you could get a good polish on a curved surface by printing its match as well and placing an abrasive such as rouge or baking soda between them.

    I’m definitely going to try this.

  13. Have you tried printing a polisher for hard to polish surfaces? That’s the usual way of making lenses (in glass); a convex polisher for a concave surface and v.v.
    Loose abrasives are used.
    It would be fun do make a very fast aspheric lens. These require several polishers for different zones, because the curvature changes a lot from the center to the edge.
    Another thing to make would be a faceted gem. If you can get some slightly fluorescent filament, a UV LED would make it glow.

  14. Have you tried heat? I use a small butane torch to polish the edge of acrylic glass. It takes a light touch with the flame to get a very clear edge. With a bit of practice I got good results. I will make a few prints using your settings and try and see what I get.

  15. Quote: “One thing to note is that slicing softwares don’t like internal cavities. They see floating spaces as an error and try to correct them automatically. In order to “convince” the software to leave my shape alone I had to connect at least a small part of it to the outer shell. With the Mario figure, I simply moved it down so the shoes will touch the bottom part of the cube. I couldn’t find any option in Cura”

    Have you tried the “Mold” mode under special modes? It creates a mold of the .stl, letting you define distance of mold wall to object wall width, roof height and mold angle.

  16. Bonjour pourquoi utilise tu de l’abs ?normalement il faut utiliser un polycarbonate PC transparent pour un meilleur rendu de transparence .et une autre question est ce qu’en utilisant une buse de diamètre plus petite que 0.4mm le rendu serait meilleur ?

  17. Interesting…but that clear filament is sooooo yellow. Are you gonna try a more neutral transparent filament?

  18. What’s up, this weekend is nice designed for me, since this time i am reading
    this great educational piece of writing here at my house.

Leave a Reply

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