Make Your Own Optical Lenses

hey everyone today we're making lenses out of epoxy i started this project three or four months ago and you can see a behind-the-scenes photo on my patreon around the end of september showing off a family of carbon fiber replicated mirrors using a similar technique and that's how i got started on this project was actually working on mirrors not necessarily lenses but because this process involves a lot of waiting around for molds and resins to cure usually 24 to 48 hours for each step of the project there's a lot of downtime so i started investigating if i could do lenses at the same time it just made sense if i'm molding up a mirror why don't i also try to mold up a lens at the same time let them cure overnight i originally wanted this to be a combined video showing both the lenses and the mirrors at the same time and the process for both of them because they're relatively similar but the mirrors have been a little trickier and i haven't quite dialed it in yet so i've decided to separate the two today's video will be just about making lenses and the next one hopefully soon will be about making the mirrors the process that we're using is known as replication molding and all this really is is taking some type of precision surface that has the surface roughness characteristics that you desire and the correct shape and molding your material of choice on top of that surface so that when you separate the two you get the desired lens or mirror whatever you're after out the other end and this replication technique is useful because you can fabricate optical quality components out of plastics rather than glass and so it's a lot cheaper and you can mass manufacture them a lot easier this is not a new technique this is basically how all injection molded lenses are made you can see an earlier video that i made or attempting to make injection molded lenses and there's a lot of complications to it because of that precision surface that's needed like whatever you're molding against in the injection mold or the mandrel has to be optical quality otherwise the thing you get out the other end is no good and so these mandrels or optical inserts as they're known in the injection molding industry are typically produced on a single point diamond lathe which is a lay that is just super super precise everything about it is controlled down to the nanometer so that a very fine tipped diamond insert can machine the surface of the metal to the point that it gets an optical surface right off of the lathe and then you can take this optical insert put it in your injection mold and shoot some hot plastic into it and you get yourself a precision plastic lens at the other end so there's nothing new about this it's just really hard to do in a home shop because i don't have a single point diamond lathe so i started looking around for ways to do a similar project but without using precision metal mandrels the mandrels don't have to be metal they're just convenient because you can turn them on lathes easily but you can really use any material that holds an optical surface and as it happens you can get glass mandrels precision optical quality for very cheap and of course these are just lenses but you can use a lens and the surface of the lens as a master mandrel to mold off of so that's what we're doing today we're taking lenses and we're creating molds out of them so that we can then replicate plastic lenses from the silicone molds the actual mechanics of molding are super simple there's nothing really to it it is a standard molding and casting technique you take your object of interest coat it with some kind of mold release pour silicone on top of it let it cure separate it pour the second half on top let that cure and once you have your two mold halves you can pour in whatever resin you want to use for the final part in our case we're using a clear epoxy and then you should get out the other end a finished lens in practice it's not that simple because there's a lot of subtle quirks and gotchas that will mess up an optical quality component so that's more of what the video is about today it's less of a tutorial and more of kind of a list of things to keep in mind if you want to try this yourself so you can avoid a lot of the mistakes that i ran into while working on this project but first let's look at a few of the lenses that i've created just to get a sense for what's possible and what some of the defects might look like far and away the best examples that i have so far are these small plano concave and convex lenses there are two different convex lenses that have slightly different focal lengths they're both about 12 millimeters in diameter and focal distance is 70 to millimeters thereabouts there are some defects a little bit of surface roughness and pitting little scratches here and there but optically they look pretty good they focus well on items in a distance you get reasonably crisp edges and definition while looking through it these are spherical lenses so there's a certain amount of aberration just inherent to them chromatic aberration where the colors kind of smear out but aside from the expected aberrations there's not a lot of unexpected problems you don't see like a fun house mirror effect the two convex lenses magnify and the concave lens demagnifies and it all looks pretty good if we pop them under the atomic force microscope you can see that while the epoxy lenses do have slightly higher surface roughness than the glass lens it's still remarkably low and pretty tolerable for what this is which is a cheap plastic lens if we zoom out to a scanning electron microscope view we can see that the surface while not perfectly smooth there's a lot of pits and valleys and little kind of pieces of crud stuck around it's really not a bad surface it looks pretty smooth and if we look at one of the convex lenses that's been coated in silver for the sem and just look at it optically it makes a pretty good mirror's surface and focuses really well back at me and the camera and unsurprisingly if we look at the mold itself that these came out of the surface of the silicone is extremely reflective and smooth and looks optical quality itself so these were definitely a success the small lenses i did not have to try too hard to get dialed in correctly now as we scale up into larger lenses we start to see more of the problems that are intrinsic to this technique and unfortunately i don't have a lot of variety in my lens collection i have a lot of small lenses that are 10 millimeters and then i have lenses that are big at 100 millimeters and i don't have a whole lot in between so we get to see kind of the two extremes of this technique so you can see that some of the lenses have pretty considerable surface roughness that makes the surface hazy or just not very clear other lenses have pretty pronounced aberrations across the surface of the lens it gives it more of a fun house effect it's not a smooth perfect spherical surface okay so let's talk about the technique some and discuss the stumbling blocks you might run into while working on something similar the main goal to keep in mind here is that we need to produce a lens that requires zero post processing that means no polishing or lapping or doing anything to the lens we want to pull it out of the mold and have a perfect optical surface and the reason for that is simple you just can't polish your lap or grind plastic like you can glass with glass it's pretty easy although tedious to precisely grind down the shape that you want because it fractures and breaks and polishes in a very controlled repeatable manner and you just can't really do that with plastic it's too soft of a material it deforms and smears under your grinding or lapping process and generally just doesn't give you the same kind of control that you have with glass and so we want to avoid that completely we want something perfect out of the mold the two parameters that we care the most about are surface roughness and overall shape the shape is how much does it deviate from the original lens that we're molding from so if the original lens is a perfect sphere on the surface we want our molded part to also be a perfect sphere on the other hand we have surface roughness and this is how rough or smooth is the surface of the part at a micro level and this is important because it directly impacts contrast of the optical image a very rough surface will give you a hazy low contrast image because all the light is scattering in different directions whereas a nice clear surface will let a lot of light through and you'll get a good crisp contrasty image these are technically unrelated you can have really good figure but really poor surface roughness or vice versa and i actually have some examples of that this particular lens has a really quite nice surface figure but for a variety of reasons the surface roughness came out very rough and so it's a hazy almost unusable lens because of that for this project i'm using a silicone resin to make the molds this particular silicone is known as mold max xls from smooth-on and i chose it because it has very low shrinkage and it holds up to epoxies and other aggressive resins better than other silicones meaning it won't shrink as much each time you pull apart out of the mold all the normal casting things apply so you should degas your resin very thoroughly because you don't want any bubbles on the surface of the part interfering with it make sure you have proper molding techniques everything's vented correctly so you don't get bubbles in the middle of it i had a lens where i didn't mold it particularly well and there's a large bubble right in the center which kind of ruins this particular trial these are tin cure silicones which means that one of the reaction products while it's curing is alcohol so you have to make sure that the mold itself is fully and thoroughly cured before you use it otherwise alcohol will actually degas into your epoxy and that kind of messes everything up and i think actually that's one of the reasons one of my lenses came out with a lot of internal stresses was due to is kind of bubbling through it while i was gassing so that's all just normal molding procedure let's talk about some of the interesting things about this particular project so we're using glass lenses as our mandrel and this presents a unique problem for us you see silicones are great because they can replicate details to a remarkable degree like nanoscale features you can replicate in silicones which is why i chose it silicones are also great because they don't stick to anything they're notoriously kind of unsticky you can pour it on just about any surface and then peel it right back off and it's fine there are however two things that silicone will stick to if you pour silicone on top of another piece of silicone they'll bond together chemically and if you pour it on top of glass like our lenses it will also bond to the glass which is unfortunate for us the reason for this is that silicone tends to react to anything else that also has elemental silicon in it and since glass is silicon dioxide it likes to bond to it and this can be a sufficient bonding that when you try to separate it it will rip off chunks of the silicone and basically ruin your mold so we need to use some type of mold release so that the silicone will come off of our lens but this is a challenging prospect for us because a typical mold release like this doesn't give us a nanometer smooth surface if you try to spray or paint or dip a lens in mold release when it dries you'll get a pretty substantial amount of layer lines as it dries and so there is a noticeable roughness that's left behind by the mold release which then directly transfers into the mold the lens that i mentioned earlier that had a pretty hazy finish to it was done using mold release if you look at the mold you can see faint kind of hazy stripes or circular oblong lines across the surface and that is the nano scale relief created by the mold release that was on the surface of the glass and indeed if we look at the molded part that comes out afterwards it has the exact same lines transferred to it there are ways around this i've read that there are mold releases that have very low solid content meaning it doesn't have a lot of dissolved solids in the mold release itself which are better for this because they deposit to a thinner smoother surface i couldn't find any of those easily available and i didn't want to go chasing down different manufacturers so i quickly decided that mold release wasn't going to work for me instead i chose to sputter a thin layer of metal across the surface of the glass and this is in effect a mold release for us the silicone doesn't like to stick to anything except itself and glass and so it's not going to stick to this layer of metal and the sputter machine will deposit a very thin and uniform layer of metal so we don't get any like layer lines in between it it's a pretty perfect copy of the underlying surface just in metal i chose to use silver mainly because it's cheap and sputters quickly and you can chemically etch it away later if you don't want it which was a nice way to recover lenses without permanently ruining it it also doesn't stick to glass particularly well so it'll separate away from the glass easier than other metals like chromium which will bond pretty firmly to glass but in principle you can use just about any metal that you can sputter on top of the glass layer so that takes care of the top surface for the edges and the bottom i still use just the regular liquid mold release because that's not the optical surface so it's fine to use that there with the mode release taken care of we can turn our attention to the next major problem which is shrinkage of both the mold as well as the molded part shrinkage is a property of pretty much any curing polymer system as the polymer chemical reaction happens it shrinks and that's just a fact of nature there are ways that you can control it to minimize shrinkage but it's gonna happen so the key here is to make sure that it happens as little as possible by choosing a resin or a silicone that has a very low level shrinkage and then ensuring that when it does shrink it shrinks in an even controlled manner kind of uniformly across the part you can see that on these two different molds one of them has a pretty nice reflective optically looking surface whereas the other one is noticeably warped and wavy this mold i wasn't thinking and i used a accelerator to speed up the curing of the silicone and of course the accelerator isn't uniformly distributed everywhere no matter how well you mix and just the nature of the chemistry means that you will get uneven curing in places and so you can see this surface is distinctly not spherical so that's tip number one make sure you don't use any type of accelerator related tip number two is to get a slow cure resin so you want all of your resins that's using this to cure as slowly as possible because that means the chemical reaction isn't running too quickly which means that the shrinkage will be controlled and even and relatively minor silicones aren't overly affected by temperature they'll speed up a little bit if you move them to a warmer location but epoxies are dramatically affected by the curing temperature if it's cold enough they won't cure at all and if you throw them in an oven they'll be cured in 20 minutes so temperature control is a really important part here for the molding of the final part you want to make sure you're at a nice relatively cool temperature to let your slow curing resin just do its thing over a long period of time there's a fair amount of research on this particular on the mirror side where for example nasa has looked at special formulations of epoxy that cure over the time span of weeks to minimize the amount of shrinkage just to give you kind of an idea how important that is and to make life harder the actual curing process is exothermic meaning that it gives off heat as it cures and this is problematic for things like lenses because most lenses have a thin section on the outside and the center of the lens is quite a bit thicker meaning that the center of the lens is going to cure at a different rate than the outside of the lens it'll cure a little bit faster now this isn't a huge difference in thickness it's maybe a couple millimeters but that's enough to introduce stresses into the epoxy as it cures and because this is an optical component like we saw back on the injection molding video any stresses inside of a clear polymer will introduce aberrations we can see this very clearly if we use polarized light and a polarizer in front of the camera if polymers have different amounts of internal stress that stress is manifested by a slightly different index of refraction so when you look at it under polarized light you actually see different colors because the light that's hitting the camera has refracted at a slightly different angle so you start to see colors appear and this is a a visual indication of all the internal stresses in this particular lens due to an uneven or too quick curing schedule other lenses that are much clearer do not have nearly as much internal stress so you can see some other examples where there's really no change under the polarizer maybe a little bit on the edges but it's pretty much the same homogeneous consistency of polymer all the way through notably my best examples that have the least amount of stress are also the thinnest they're either the small plano convex and concave lenses that i showed up at the very beginning or they're these large but very thin lenses that i fabricated a kind of logistical issue that you have to deal with is that silicone is not particularly rigid it's quite soft and flexible you can warp it all around which is problematic for fabricating lenses because when you sandwich two halves of the mold together whichever side is on the top will sink a little bit just because this isn't mechanically very strong and so the top surface tends to sag just due to gravity and that this isn't very robust and that sagging will change the shape of the lens there's different ways you could address this you could use a stiffer molding material so there are silicones in a wide range of stiffness you can use some type of backer material to help mechanically keep this nice and firm some like metal or rigid plastic to hold everything together i think it should be noted that you don't have to use silicone so i have to use silicone because it can replicate extremely fine levels of detail and it doesn't stick to much so it's a convenient molding material which is why you know prop masters and stuff use it but there are other materials you could mold from you could for example use an aluminum filled epoxy which will give you a really stiff mold that doesn't suffer from any of the kind of sagging or distortion issues of a silicone but the problem is that the mold release issue becomes even worse so you can still do the metal sputtering trick on top of your lens when you first mold it and that will separate fine but the problem becomes when you make your final molded part pouring an epoxy into an epoxy they're just gonna bond together chemically you'll never get those parts separated without some type of mold release and so you could sputter the mold itself with a thin layer of metal but you would need a pretty large chamber to get the mold inside and my chamber is just not big enough to do that so that wasn't really an option the last thing you could do is just to cut out one of the molding steps this whole time we've been taking a positive lens and making a negative mold which allows us to create a positive plastic lens but every time you have an extra molding step you're introducing more chances for error both in shrinkage and surface roughness and particles and just kind of all the associated problems so if you can get rid of this mold and go straight from your mandrel to your part you'll have a much better finish both on service roughness and surface figure that's essentially what i'm doing for the mirror project i'm making carbon fiber replicated mirrors directly off of a precision mandrel essentially as a way to remove one of the molding steps and limit the amount of error that creeps into the mirrors because mirrors are even more susceptible to surface roughness and figure than lenses are but that's all for the next video which will hopefully be out soon if you are interested in replicated carbon fiber mirrors and you're not subscribed go ahead and subscribe so you don't miss it should hopefully be out sometime [Laughter] sometime yeah good so that's where we are today the lenses aren't perfect but the smaller ones and honestly even some of the larger ones are looking pretty decent and i was tempted to sit on this until i'd really dialed it in but i know i personally have a tendency to not share things until i think it's perfect which means a lot of stuff that i think's interesting never actually gets shared because i never get it perfect so this was my attempt to kind of share some of the knowledge i've gained while working on this and i don't get some information out there for folks that might be interested in similar i think if you wanted to continue to refine this there's really two main avenues that you could work on the first one is getting a detailed breakdown of different resins and their properties and finding the resin that works really well with your setup so i just used an optical-ish clear resin from smooth-on that had reasonable properties but i didn't test any other resins so that would be step one is finding a whole bunch of different epoxies and vinyl esters and polyesters and testing them all to see which one really works well and then i think the biggest thing really is just really controlling the process itself so active temperature control so that the epoxy cures uniformly at the correct temperature for the entire duration of the curing maybe embedding sensors in the molds that you can track temperature maybe cooling loops so that everything stays at exactly homogeneous temperature like you want as uniform of environment as possible making sure all your resins are mixed entirely correctly completely degassed you know et cetera et cetera so really refining the process of this i think would eliminate a lot of the errors that i ran into with like stress or surface roughness big thanks to folks on patreon i really appreciate all the support you've given me it helps me work on these projects that take a really long amount of time as i mentioned the top i posted the first start of this project back in september three or four months ago and these things just take a long time sometimes and so having patreon support and a place that i can share these behind the scenes looks just keeps me motivated and is really helpful so thank you i appreciate it okay cool well it's getting cold in my garage so i'm gonna wrap this up stay tuned for part two where we look at those carbon fiber mirrors and otherwise thanks for watching see you next time

Channel: Breaking Taps

Views: 440,112

Keywords:

Id: mfAGivG9Koc

Length: 24min 28sec (1468 seconds)

Published: Mon Dec 06 2021