Deutsche Version dieses Videos: • Optische Linsen herste... In this video I document my way of making optical lenses out of epoxy, using a mold of silicone rubber.
Nice job. I wonder if using a plaster of Paris mold made with another lens or sphere would produce an acceptable plano-convex lens. Plaster of Paris can replicate a diffraction grating well enough to produce colors so a lens seems plausible.
Great video, thanks. I'm also looking into casting a custom lense for a project. However think there are two suggestions I'd like to make. 1. Spend all the time sanding/polishing/coating the PLA printed part rather than the lense because silicon casting can produce a mirror smooth finish right out out of mold (with zero work needed) if the mold is case with a mirror smooth impression. Spending all the time getting the PLA lense perfect would have saved a lot more time with better results since PLA is heat smoothable and you can fix any issues before rather than later. 2. Kinda obvious for lenses, avoid cutting the silicon mold radially. A slice in half top to bottom would have worked just as well which will avoiding messing up the lense surface.
1. This will probably not work for two reasons: - there are still gases trapped in the silicone; so I got some tiny bubbles which caused mini-craters in the surface of the silicone once it was hardened. I didn't had a vacuum chamber to draw out the gases - perhaps that could indeed make things a lot easier already. Those tiny craters alone would mess up the surface too much already so you would need to polish it anyway. - also the separation wax adds just enough micro structure to the surface that the resulting epoxy-surface would be cloudy, thus requiring polishing as well (+ a layer of varnish, like I did). Without separation wax I found it to be nearly impossible to separate the two halves without destroying them (I wanted to reuse them). The 3D-Printed shape I did polish to near perfection, I think the shine should be visible in the video. 2. I think you mean to make the two halves "around the equator" of the lens, opposing to "along a meridian" like I did (essentially causing a line across the surface, yes.)? This is a good thought, and I thought of that as well, but decided against it, because I knew the surface would not be perfect anyway from previous trials and how I did it was easier to create the two halfes of the silicone. But yeah, it can be done either way. Good ideas though. And please try them out. Its definately a learning and trial/error process. I did quite a lot of trials and errors before these results.
@@aurigo_tech My bad, I skimmed through the part where you polished the PLA part. In that case I'll revise and say that it might be the thin coat of release agent you applied inside the silicon mold that might have caused the final casting to be not mirror smooth. I haven't casted a focusing lense but I have casted car light related parts and have gotten mirror smooth results right out of the mold without any release agent. But if you're finding you have to use a separation wax layer as a release agent, then maybe before putting the mold halves together, you could apply a super thin layer of the wax release agent and blow some hot air over it to melt and smooth out the wax layer. Also if you're gonna cut the silicon mold "around the equator" as you put it, you could stabilize the outer walls of the silicon mold by embedding a perforated piece of MDF or stiff metal sheet on both sides right into the silicon mold. That way when you cut the silicon mold height wise, you won't have to worry about a deformed mold when you clamp the mold together again. Another idea but haven't tried since I have a homemade simple vacuum rig. You could try vibrating (a 60hz massager, or a subwoofer playing below 100hz) the resin for short period of time and use a surfactant to remove some of the bubbles. Or maybe make a vacuum rig by modifying a simple car tire pump which is what I did.
thanks for uploading .. i think epoxy dip will give a better outcome since it's the same material also vibrating the fresh epoxy / heating it causes bubbles to rise to the top
you could build a frensnel lens like those of the TV series the great escapist episode 2 did, instead of using glass if you can use resin and if instead of building prisms you can make triangles directly?
Absolutly, but I think the smaller the lenses, the harder it will be to get the curvature and the surface perfect. Smaller than 3 to 4 cm diameter I would not go.
plastic wrap over a mason jar lid makes for a nice convex lens too. if you use acetone and abs filament, you can reduce your need for sanding. soak a sock, or piece of fabric in acetone, staple it to a metal coffee cans lid, then place your piece in the can and add the lid: it will be very soft, to cure more quickly, remove the can from the print. possibly have both ends of the can open. cooking times very, no heat.
Great ideas but I imagine it will be very hard to make very specificly shaped lenses (as I did). For a more or less arbitrary convex lens, yes you are right, there is certaily some steps that can simplify the process quite a lot.
it shouldn't be too far off, measure from the meniscus to the vertex of the curve, measure the diameter, and oula. a perfect isosceles. ;) Something else I've been meaning to try is to use a 3d printed curve for a mold, then use a microwave as a kiln.
Hello thanks for the cool video. How many times can I use this silicon mold without ending up withe a cloudy resin look? would the silicon mold`s surface roughen up over time? Thanks
The surface of the lens will not be perfectly smooth in any case when you take it out of the mold. Besides that this method is definitely not for mass production. I guess with careful handling of the mold a few dozen at most. For mass production one would probably need a solid aluminium mold or something like it.
Thanks! I am really not sure about UV. Epoxy generally gets a yellow tint faster under UV influence, thus it might be that it is actually opaque to UV and absorbs it. But I don't know for certain.
Very cool video! Thanks for making it. What is that handheld power sanding tool called, a mini-belt sander? I think I need that, for many different projects!
Thanks! I appreciate it. Jep, its a mini belt sander, very useful indeed, but only for smaller things like these lenses or for corners etc. on larger pieces. For removing more material or working on larger surfaces it however is still too slow/small.
sanding the lens increases the eccentricity as I assume the pressure of your fingers is not perfect during sanding. There has to be a better way of achieving the right surface... By the way, what is the purpose of these lens?
Hardened Epoxy is hard enough and the fine sandpapers are fine enough so the amount of material you remove by hand is so miniscule that it doesn't affect the overall shape in a noticeable way. Well, if you want to make a industrial grade microscope or telescope or something similar this manual approach will not bring a perfect result in any case, so I guess then hand sanding would be an issue, yes. The imperfections through the molding process have to be removed however somehow and the surface needs to get a perfect finish. This is actually more important than small imperfections in the shape of the lens I found out (at least for hobby grade stuff like I did here). Without a perfect finish on the surface the image you see will simply suck. I made these as a test and for another project which I have put aside for now, unfinished. I plan to resume working on it though and maybe also make an updated version of this video.
@@aurigo_tech I'm intrigued about the process and metrology of making lens. I always wanted to make a telescope from scratch, but the problem is the "how to be guaranteed of the surface".
I don't know. The material in the volume of the lens would need to be perfectly even for it to have an acceptable optical quality. I don't know if a layer-per-layer approach could deliver that. I have seen translucent resin prints but not transparent. But I don't know if that was caused by the surface (which could be solves as I did in the video) or the material itself. So it remains to be tested. I don't have a resin printer though.
I think so, the material is very sturdy and can be cut without having to wonder if it shatters. I would cut it first though and then apply the varnish finish (or epoxy dip, as someone commented here). However, when following the approach I did in the video you could also just make the 3d model and 3d print a rectangle from the start and there would be no need to cut anything.
Yes. There may exist UV blocking varnishes etc. that could prevent it or slow down, but thats not something I investigated further yet. After some years now the lenses from the video have yellowed a bit but are still clear to look through.
Most useful video about making lenses. Believe me. I've watched and watched. Another thing...I'm not good at math, so I'm not sure what does one hundred past two two mean. Not that skilled in English either.
Hm, I have never heard of a venn lens. I assume you mean its a lens following a venn diagram - that would be what I calles spherical lens then. "Curved lenses" is also not a term I have heard of, as all lenses are curved in one way or another. Parabolic lenses are indeed a thing - they would be aspherical.
So, uh.. why not 3d print the mold & modle? A stand could easily be incorporated into bottom and could double as a bracket to support the mold during curing. I'd also recommend building a vacuum chamber as it will draw out all the bubbles. Its not hard or very expensive: a lower end rotary vane vacuum pump is like $200 (CAD)/ $160 (USD), and the actual chamber only needs to withstand a max of 14pis, and a small volume at that. Roughtly speaking a 6"×6"×4" chamber would give ya a lot of versatility and if your mold was say 4"×4"×2" then the actual volume of vacuum would be (6×6×4) - (4×4×2) = 144 - 32 = 112 inches³. You could probably get away with 3d printing that if you really wanted too. It may also be worth it for you to throw together a high precision lens polisher. 4 servo motors (to position the head use one for x axis and one for y, the other two for much more refined controls. If you had both of those set up so that the first two affected the second two with a gear ratio of 100:1 and 10'000:1, isolated it from vibrations (suspend the whole thing by steel wires), and worked really slow you could produce some seriously stellar lenses. Oh ya, and you should also spray um with uv resistant clear coat so they don't turn yellow in a few years
It's all a matter of effort and desired quality of outcome. What I wanted to archieve was a finish in geometric and surface quality, so the optical effect would be clearly visible, though not perfect. For glasses or camera lenses it would be too rough and unprecice, but for what I wanted to do it worked out just fine. I was aware of using underpressure to reduce gas bubbles in the epoxy while doing this video but decided it was not worth (yet) the effort, considering what I wrote. For my purposes a few tiny bubbles in the lens were still totally acceptable. A dedicated lens polisher might be overkill for any purpose short of camera grade optical lenses or for just some individual pieces. As shown in the video a simple varnish dip will return a decent result on a pre-sanded surface. Regarding the UV resistance you are absolutly right, they have yellowed a bit now after 3 years, though the lenses are still absolutly clear and usable. But don't get me wrong - all good suggestions! Again it just depends on the outcome required and the effort/means one wants to invest.
@@aurigo_tech I should have mentioned that your video inspired me to attempt to make my own telescopic lenses 😅. Your proof of concept got me thinking about it and that's what popped into my head to get them to that sort of quality... though I forgot they also oxidize over time. If I do wind up attempting it ill document it and share it with you.
@@ltsgobrando Very cool! And yes, please share your results. :) Regarding oxidation I am not sure - certainly nothing like it visible on my ones (after 3 years). Apart from UV epoxy seems to be quite resistant to corrosion. But then again I have them only lying around in my house and don't subject them to much different conditions.
In principle yes. Try it out. While doing so you will quickly learn the tricky parts of making lenses. To get optical grade lenses you have to work really accurate.
Use Teflon coating machine after varnish .hope this will help. Teflon Coting gives glaze to paint's it may also work on varnish. Teflon pads are cheap you can also do with hand or at any Vachel workshop.
Depends on the intended purpose I think. Such a big lens would probably need to be a fresnel lens (to save material). In any case with at-home techniques (like limited size for 3D-prints) I would approach this by separating the lens into 4 or more pieces and later glue them together. When the lens is not too thick and/or the exact shape not so important then its just a matter of finding a way to make the mold and then pour in the epoxy.
Same as the lens itself basicly, starting with rougher sandpapers, going to finer ones and finishing with polishing paste + bits on a dremel. However, because the lens will require extensive sanding/polishing anyways one could skip the polishing for the 3d-printed shape. Just sand it down to the finest grid paper and it will be enough.
Zünder Anyway problem solved . Do you think you can find a glass cube and cast a resin one and compare the refraction index of the glass vs resin? 🙏🤗 I would love to know they have almost the same index so we can just copy the engineering of existing lenses
Nah, just sanding+polishing + in the end I tried to make it smoother with spray varnish. It surely looked better but did not improve smoothness. Refraction Index: no need to test it as the properties of the epoxy waterclear are known. When I remember correctly its 1,5ish - so pretty much like glass. Hardest thing in lens making is to get the shape perfect so the image is not distorted (most comparable yt-videos fail at this, and mine are also not perfect). When you have to combine 2 or 3 or more lenses the errors add up and the image would be unrecognizable. So, to get to industrial quality (like in camera lenses) is actually super hard.
You could spray it with some kind of semitransparent color, I am no expert on what would work, but many sunglasses seem o have such a coating. Or you could add colors to the epoxy itself (I want to try that myself at one point). But this would not be the best sunglasses as epoxy is not UV resistant + you might damage your eyes (I don't know if it blocks UV totally).
