Step 2: 3D Printing

If you want to 3D print a lens you'll need a nice printer. Most resin and stereolithography-based printers will have the resolution necessary. If you want to try with a filament(Makerbot style) printer, you are going to have to cast transparent resin, which is totally doable.

While there are many ways to make a custom lens, 3D printing is perhaps the coolest. You can export a 3D model of a lens and watch it print without too much planning or physical labor. Additionally, additive manufacturing allows for more complicated lens designs with undercuts and double-sided geometry.

However, this flexibility comes at the cost of expensive material and lower optical clarity.

My Experience

Here at Autodesk's Pier 9 Workshop, we're lucky to have access to an Objet Connex 500 3D printer. I used this photosensitive resin printer along with VeroClear transparent material to create many of the lenses for Smaller and Upside Down.

Printing was simple. I just pressed print and let the lens accumulate overnight.

Even on our fancy printer, however, the lenses ended up dull and striated with artifacts from the layer-based printing process. 5 hours of sanding and polishing (covered in step four) was required to make the lenses transparent. Even then, the printed lenses don't match the optical clarity of milled ones.

Other people have been 3D printing optics too. A few months after I started printing lenses on our Objet, the team at Formlabs made an excellent step by step guide for 3D printing a monocle. A cool startup called Luxexcel prints lenses professionally using specialized machines.

<p>So, this process could be used to make your own prescription lenses, if you have a current copy of your prescription?</p>
<p>I am not certain, but my intuition is that the quality is too poor for that kind of application.</p>
<p>To help with the polishing process, you could make &quot;tools&quot;. These are shaped to &quot;fit&quot; over the lens surface. For aspheric surfaces, you also need some sort of compliant layer. What I would do is cast Hydrocal plaster over the printed lens (after the first rough sanding). Put a piece of broomstick into the plaster to serve as a handle. After this is set, glue on some sort of foam for compliance, and then your sandpaper, cut in small squares. You work this &quot;tool&quot; over the lens, stroking toward and away from you, turning it a little each stroke (if the lens is rotationally symmetrical), as you slowly walk around the lens. (Mount it on a barrel so you can walk around it). The more aspheric the lens is, the more compliance you need (thicker foam). You can use loose abrasives and a cloth over the tool to finish up the polish.</p><p>The tool diameter should be about 5/6th the size of the lens, unless the curve is extremely aspheric. More aspheric surfaces finish better with smaller tools.</p><p>This process I describe is the one used to produce &quot;real&quot; lenses in industry. You can test the lens before it is fully polished by wetting it with soapy water.</p><p>The faceted convex lens can be polished with a tool 5/6ths the size of a facet, one facet at a time. This tool would want to be quite &quot;hard&quot;.</p>
<p>This is awesome! Try to enter in the 3D Printed Contest. I'll vote for you and I think you would earn a prize.</p>
<p>Its sweet</p>
<p>I've never tried it, but I would think you could do the same thing with clear casting resin. You could sculpt your lens shape out of modeling clay then cast a mold of that and then use the mold to cast your lens.</p>
<p>Could this be adapted for a peephole on a door somehow?</p>
<p>I haven't tried any multi-lens systems yet, but a giant peep hole would be pretty cool! I'm sure it is possible.</p>
<p>I have been casting aspheric lenses in epoxy for some time now. I had dismissed the idea of CNC machining the blanks, but I see now that I'll have to give this a second look.</p><p>I am most impressed by your use of ray tracing software for optical simulation. I tried using POV-Ray for this, but I never found anyone else online that had described how they did it, and I never sorted out all the details myself. I should revisit this someday, too.</p>
<p>Hey Noah, I added a video to the design step that might clarify the process i used a bit.</p><p>I never considered using epoxy, but I might give that a shot. I am ordering some optical silicone soon to play with that. The optics industry is starting to make squishy lenses for durability, and I want to see how they work. </p>
Yes! please send me some links and brand names that offer these products. This is something's I've been trying to find. I have been unable to find any optically transparent silicone or urethane rubbers except for the thick one-part crystal clear caulking used in fish tanks -- unsuitable for my needs. I need something I can cast with... I have come across a few of these rubber lenses in products like webcams and, once, in the 1D linear lens over the sensor array in a flatbed scanner.<br><br>My other unsolved challenge has been to find an easy and effective method for silverring plastic surfaces to make reflective lenses. I have cast some decent 12&quot; parabolic reflector lenses, but I have not tried mirroring them yet. I'm going to experiment first with vapor deposition. After that I might try galinstan, although the examples I've seen hint that it works better as second surface mirror. The first surface always looks streaky and wer -- and it is wet since it stays liquid at room temperature.... Then I'll give the classic silver nitrate method a try, though I'm not sure if it works on plastic.
<p>I haven't tried it yet, but I am requesting samples of <a href="http://www.quantumsilicones.com/qsil-213/" target="_blank">this</a> silicone.</p><p>For mirroring, I intend to try <a href="https://www.youtube.com/watch?v=M9Gteg_Hpls" target="_blank">this</a>. They say it works on acrylic, but I haven't checked yet.</p><p>Add some images if you try some silicone! I'd love to see.</p>
Yes, for high volume work, if you need a lot of identical lenses, you might consider printing a lens in reverse to form a mold. Then use your mold to cast lenses from epoxy or other resin. You'll need a vacuum chamber to degas to eliminate bubbles. A pressure chamber is also helpful. You can use either one alone or combine for super bubble free castings. If you are start with just one I'd go for vacuum -- It's easier to build a vacuum chamber than a pressure chamber.
<p>Awesome! Just recently I have been searching for lenses to try and 3d print on thingiverse. I didn't find any. Do you have an STL file for the lens in the photo on step 7? My purposes will be mostly decorative, but I would like to try and print some in clear filament.</p><p>Another thing, have you attempted any fresnel lenses? I feel like those would lend themselves well to being CNC milled or 3d printed as they use grooves instead of curvature. I could be wrong on that but I'm still curious as to whether or not you've attempted it.</p><p>People say we have reached a point in time where its almost impossible to invent anything new, but we continue to reinvent old things in new and amazing ways!</p><p>Thanks!</p>
Cutting the fresnel lens should be simple enough... polishing it for any kind of optics may be a bit harder.<br>They still use curvature, but broken up over the many circular lens rings, instead of one large curve.<br>The clear filament, you will still have to check the optical characteristics of. Different plastics react differently.
<p>I don't think any filament based printer can make an optically clear lens, but hey, new things are happening every day. </p><p>The only way I know to get lenses out of a filament printer is polishing your print(in whatever color) and casting it in some sort of clear resin using a rubber mold.</p><p>As for fresnels, I agree with Ironsmiter. Polishing would be nearly impossible. Continuous smoothness makes sanding possible. </p>
I thought of a way to print a polish fresnels. Print each ring as physically separable from the others. If printing a 10 inch lens integrated into 1/2&quot; bands gives 9 nesting rings and a solid central rod. Also print a polishing jig. This would be a wedding cake shaped stand that would be the inverse of the fresnel. The polishing jig stand would not need separable rings.The height of each ring would be calculated such that when you place the fresnel rings on each ring would be sit with a refracting surface curve restored to a continuous parabolic surface. Then you polish this surface optically smooth and when you're done you pull the rings off the jig and renest them to form a fresnel. The gaps between rings could hang the polishing wheel but it should be easy if you're careful.
<p>It depends on how finely you subdivide the curve. I have some large glass fresnels with half inch and one inch bands. Polishing would still be quite a bit more difficult, but not impossible, and perhaps even practical if your application is for light concentration or projection -- if you don't need to focus virtual images then the optical requirements are a lot more lax.</p>
<p>Hi Robb. Great post! I got an alert in for <br>&quot;3D printed lenses&quot; as I am active for Luxexcel. We can 3D print <br>lenses that require no post processing. I don't want to plug the company I work <br>for here but maybe we can talk sometime?</p><p>I have made a comparison of different <br>transparent 3D printing techniques you might find interesting. </p><p><a href="https://www.luxexcel.com/3d-printing/transparent-3d-print/" rel="nofollow">https://www.luxexcel.com/3d-printing/transparent-3d-print/</a></p>
​I'd love to talk! My email address is hello at robb dot cc!<br/><br/>I have seen some physical samples of your prints, they are phenomenal. <br/><br/>My collaborator and I wanted to see y'all at Photonics West, but we got bogged down at other booths.
<p>Perfect I'll drop you a mail next week. To bad you didn't manage to come around at PW.</p><p>Would be nice if we could make something together/ do some experiments :-)</p><p>Did you see the comparison for the 3D printed materials we made?</p>
<p>I saw the comparison a few weeks ago while doing some research. Very thorough! Makes the objet resin look almost opaque by comparison. Look forward to talking more!</p>
<p>these are making distorted lenses not magnifires</p>
<p>That's all magnifiers do. They distort the light coming through them, and spread it out. The only difference is how the light is distorted. It wouldn't be a problem, making a magnifying lens with this method. </p>
<p>Thank you for sharing a new way to lens the world.</p><p>I really enjoyed your ible.</p>
<p>wow, robb, this is so cool that i'm going to ask my son {robbie} to make one of these &amp; insert it into our <strong>front door</strong>! that way, when obnoxious people come knocking &amp; ringing long after dark, the first thing they see is a freakishly distorted face looking menacingly @ them! <strong>x^D</strong></p>
<p>wow that's science fiction! </p>
<p>This is awesome!!!</p>
this is so awesome! it makes me wish I had more access to these types of tools!!!

About This Instructable




Bio: Robb is an Artist-in-Residence st Autodesk's Pier 9. He went to Carnegie Mellon to study Art. He mostly does tangible artifacts that are often ... More »
More by Robb:Flight-case Toolbox and Workstation How to Put a Bumper Sticker on your Bicycle Upcycled Vintage Suitcase Boombox 
Add instructable to: