Optical finish for acrylic -- vapor polishing and other techniques 

We do highly polished acrylic where I work. The secret is that you have to perform the mechanical smoothing to a very high degree before the flame or chemical polish and you cannot skip the in between grits (320 to get the saw marks, 400, 600, 800, 1500, 2400, 3200, 4000, 8000, 12000). 2000 grit is nowhere near good enough ending smoothness to achieve a good optically clear surface. If you took it to 8000 then polished with slurry compound charge on a cotton felt wheel that would be good or for even higher quality, you take it to 12000 grit, felt wheel polish then flame polish. Another issue is that trying to flame polish by hand is art you would have to learn by trial and error, consistent flame polish is best done by driving your acrylic at a consistent speed past the flame.

I use vapour polishing for plastic, and the best way I've found to do it is to use a large metal pot (though I have also used a giant pyrex beaker on occasion). I normally use a stainless steel pot. Put a few tablespoons of solvent in, and heat. Watch the vapour head as it rises up to the top, remove from heat to prevent the vapour head from over flowing. Simply suspend your plastic in the vapour head, I use wire to hang it from. Add more heat if the vapour head begins to drop too low in the pot. This can be done for long periods of time if necessary, and means far less solvent exposure for the operator, and consumes only small amounts of solvent. All the best!

it is SO rare to watch a youtube video so informative, clear, and concise. thank you so much for this! answered my acrylic polishing questions, which i've had rolling around in my head for awhile.

you know you're a good educator when you can do a video on polishing acrylic, something I don't see myself attempting any time soon, and hold my attention for over 8 minutes.

Looks like the scratch pattern from 400 grit or previous grit is still noticeable on 2000 grit scope. Try to avoid skipping grits, move in linear motions instead of circular, then perpendicular movement when switching grits. This is to help identify when the scratch pattern from the previous grit is removed. Only then can you step up to the next grit.

So glad to have people like Ben out there supplying some truly mind-forward entertainment and shop-talk. Great videos Ben!

*In my art 3D class taught by a famous artist in NYC when we worked with acrylic in the end we had a 2 step polishing process, both involving a rotatory fabric tool to which we had to apply special chalk, that chalk had 2 variants, one red and one was blue, this is what gave the final transparency, however it was also very easy to mess up the smooth surface, but this tool also evened out the imperfections, there are more standard tools for this*

I've been working with acrylics for a long time. This is good info. Only comments are sanding should be done in a figure 8 rather than single circle. The figure 8 is an old school lapping technique used to change approach angles of loose abrasives which will unevenly abrade the external edges of your surface. The other thing is that extruded acrylic sucks for glueing. The internal stresses blow the material apart once the material is chemically compromised by the glues/solvents. As for polishing... agreed, mechanical polishing is best.... flaming and chemical are nice for appearance but not optics. In my experience it is possible to solvent bond two pieces together with clamping but if multiple lenses are stacked a lense or optical oil makes a perfect transition. Avoid bonding if possible. Lastly, again, I liked this video and to add to the solvent bonding phase.... polishing closes the pores of the acrylic so the unpolished (but still flat machined) surfaces glue/bond much better. Great stuff here.

In my country, almost every schoolchild knows GOI polishing paste (read as GOI - which means that the State Optical Institute). This paste is very cheap (cheaper than sold on ebay), and it is sold in almost every hardware store. This paste is polished mirror and lens telescopes. It is very easy to use, available and cheap :).

Car clear-coat buffing compound is actually FAR superior to Novus, and will give a perfect finish. I use 3m finesse-it 3000 #1 #2 and #3 and the results are incredible even under a microscope. Way more economical also. I'm not paid by 3M by the way.

Just a side note, if you flame polish and try to use any solvent on the acrylic you will likely get crazing (cracking). You may have already observed this, but I thought it worth mentioning. Great videos! Thank you for sharing!

Very infomative. Thanks for taking the time to do this. Also thanks to all the other helpful comments.

Thanks a lot for this video. It is a really nice example of sharing your experience of polishing acrylic materials. Well done. I'm going to get 1200 and 2000 grid sandpaper as well.

Thanks! This is all useful information since I work a lot with acrylic. I haven't needed to polish any yet, but I might someday.

"Ok, bye." The sudden ending always gets me.

One improvement over your demonstrated edge heat finish method is to scrape the side with a decent scraper before you apply the heat. This trick is used by special effects departments to make edges nearly perfectly transparent.. In the woodworking world, well done scraping on some type of wood can completely replace sanding. Great videos, by the way.

Great project, great video and a thorough process to follow. Thanks.

I did flame polishing in the sign business for 37 yrs. After sanding down to 600 grit, we found the best results for transparent gloss finish was straight acetylene, and methylene chloride for Polycarbonite Lexan.

Hi Ben: Another possible polishing method is to lap polish. Machine a concave surface of approximately the same shape, coat with RTV (or similar), cover with plastic wrap and press the convex lens into it and let harden (a few folds in the plastic probably won’t matter). Then lap your lens with Novus and the concave lap. Even with the problem of the aspherical surfaces gapping, with small rotating offsets you might get a good polish with minimal refiguring. Cheers, Mark

Thanks for the demonstration Ben! I've been wondering on techniques for awhile, and you just saved me from having to try it all myself!

Extremely useful video; almost exactly what i was looking for as I am looking to restore the window on a piece of vintage electronics, thanks!

You mentioned that extruded acrylic is best for laser-cutting. However, as an owner of a laser-engraving business, I am confident to say that extruded acrylic is far worse than cast acrylic for laser-cutting. I'm not sure why, but the fumes seem to be far less severe when using cast acrylic, the flame-polished edge is more consistently preserved, and engraving is dramatically better quality.

Nice comparisons and comments. Very useful. I'm doing something like that acrylic visors for my space Helmet. So polish will be important... Thanks for sharing!

Fantastic vid. You pretty much pointing out everything there is to know about doing this within the context of why you where doing it. I would suggest (if not already) that on a rounded surface for optical quality you would want to use liquid only means on a shaped die. For flatter surfaces (like head lights) you could go with sand paper (minimum 800+ grit finish). Also, it's best to not actually boil your chosen solvent unless you have some sort of "sprayer" end on the hose that will flatten out the direction of the fumes as the make contact with the surface. I personally use the double boiler method and try not to heat the water it's self over what would normally come out of the tap. This results in a dense enough fume but not so much that it effects the molecular structure of what i'm polishing beyond the surface.

I have joined acrylic with solvent for many years (as a craft teacher) and I have had only partial success. I can get an optically perfect joint about 4 times out of 5. You are right about the finish, if the surface is too flat the solvent can't wick all the way through. The other problem is that the solvent causes slight expansion of the plastic and a joint that was optically perfect at first, developed crazing inside the joint after a few years, caused I presume by locked in stresses. The solvent has to diffuse out through the full thickness of the plastic and this can take many months for thick plastic. Polishing joined pieces revealed that the joint line was still softer than the parent material after a month. Perhaps a period at elevated temperature might speed the process. I have had some success with super glues which are a different type of acrylic but still bond well. As for polishing I have once, just the once got a perfect finish straight off the lathe, the feed, tool and speed must have been just right. Mostly just the application of a thumb to the rotating plastic is all that is needed. Acrylic plastic absorbs a lot of water vapour and all machining and joining is best done on acrylic that has been oven dried at 100 C for a few hours. This also relieves internal stresses left after the casting process. HTH.

I know this is an older video, but for those coming across it years later, like me, I’m pretty sure cast acrylic is what’s recommended for laser cutting. I think extruded acrylic is often used for things like retail displays, etc.

Great video! I have several vapor polished manifolds for a flow cytometer I work on and they never look right, this gives me some great insight.

It's amazing, everything you do a video on becomes interesting.

In the laser optics world you use colloidal silica to polish lenses and for semiconductor wafers too. The tech is in the polishing pad prep and conditioning while polishing. As you polish, the pad will load with compound and start to change its polishing quality and by constant conditioning you can maintain a very high polishing process to achieve the best constant optical standards. Also, Evap carbon onto polished lenses to form sapphire coating for catch proof coating. Also all laser optics have what is called a stack, alternating dielectric and transparent metal layers to increase transmissivity through the lens. This well engineered stack is also applied using some other devices to improve deposition in the stack. And two, if you don't get all of the water vapor out of your vacuum chamber before deposition the stack will not get optimal adhesion and will blow of during use. good luck

my family operates a plastics fabrication company in detroit, we specialize in custom fabrication and work with acrylic alot when polishing any surface it is best to prepare it by sanding regardless of how you will be doing the end polishing for flame polishing we generally will sand a routed or saw cut edge with 220 grit sand paper untill all the marks from the tool are gone then we repeat this process with 320 grit or 00 steelwool this provides a extremely fine surface for the flame to polish... flaming a piece will put stresses in the plastic so you do not want to flame any surface that mar come into contact with weldon 3 or 4 as the solvents will cause the acrylic to craze in many cases. i have a bunch of info on how to work with diffrent types of plastics if you have any questions let me know and ill try to help some of the things we do at our shop include custom fabrication, plastics welding, forming, heat bending, machining, and more. if you have any questions about how to work with acrylics, polyethylene, polypropylene, Hdpe, pvc, phenolic/micarta, polycarbonate, or another type of plastic me or my family probably have experience working with it and can offer some information to you

Instantly fell in love with your videos and subscribed. Please keep videos coming. Youre a great person to learn from.

I dont know how he makes polishing plastic interesting to watch, but he did it. Ben, I dont know what these lenses are for, but It made me think that you making a telescope would be really neat!

One issue for the "lenses" you made is the correct profile, on a standard CNC the centres of the machined part surface won't be bang on. On the inner surface, though a lathe cuts on one side making the other a mirror of it, if the tool does not reach the centre then there will be a slight flat at the centre and if the tool goes beyond the centre the two halves will have a pointed profile. If you measure either half on a surface co-ordinate machine it will give you the points so you can correct the profile by altering the start and finish points in the program. Someone mentioned using a formed polishing "dolly for the grit" this will help solve that problem but you would have to set the profile of the dolly in the same way i mentioned for the part profile earlier. Probably a bit of overkill for the occasional part you are making but i do production parts that are similar and use a high precision cnc lathe to get the profile correct to within a micron. My final polish is done with 0.5 micron diamond paste (it needs acropol in the mix so specific to acrylic) Did you get an effective finish for your project?

This is amazingly thorough! Do you know if cast polyester resin behaves in a similar way?

Old video, i know. But you could try micro-mesh. It's what they use to polish stuff like an aircrafts canopy. Though keep in mind, micro-mesh follows a different grit standards than regular sandpaper. 3200 grit micro-mesh is equivalent to 2000 grit sandpaper. But 4000 grit micro-mesh is finer than 5000 grit sandpaper. And micro-mesh goes up to 12000 grit. Speaking in terms of microns, 12000 grit micro-mesh is 2 microns. 2000 grit sandpaper is 12 microns. Micro-mesh is very gentle too. It doesn't abrade away much material. Just enough to polish. So you're not gonna have a big risk of changing lens geometry.

Awesome! Just as was thinking of ideas on how to make lenses with a CNC, you upload this! But, I haven't even started metalshop yet... and I don't have any acrylic to machine... haha. I'm definitely archiving this on my hard drive for later, because I plan on experimenting with optics soon after I learn how to properly use the major tools in a machine shop. This is exactly the video I was looking for! Thanks!

Brilliant Ben! I have never made lenses but I have always polished acrylic using sand paper and polishing compound, Seems to work great and its just like working on your car, always move to a finer grit (ideally 1200-2000) before you polish off and you get the best results!

This is such handy information! Thank you for posting this. :)

SCIENCE rules with that MAN.Many THANKS!

I love your adjustable lamp base. :)

Thanks for making this video! I’m cutting back some acrylic shelves that are slightly too narrow and wanted to know the best way to make it crystal clear again after cutting it. I’m going to use the novus system with ultra-fine sandpaper. Thanks again!

Couple of things Doing a vapor bath generally your supposed to apply the vapor in a container so the vapor can rest on the part since the vapor is heaver then air. I use this method for 3d printing with abs which uses acetone but I've experimented with other chemicals and it more or less the same. I use a presto steamer to heat the chemical and enclose the part. It comes with a tempered lid. You'll only need to do this for 30 seconds at a time about 3 times at 40 minute intervals and it's best to freeze the lid so the liquid condenses and reduces the potency. For sanding they have sand paper all the way up to 12,000 grit which would give you a higher polish. One thing you mentioned is that you jumped from 600 grit to 2000 girt. That's generally not a best practice as your spending more time for less results. You should start with 600,800,1000,1200 elec it will save you quite a bit time and remove the previous grit faster.

This is a very informational video. Thank you. I've got a question about getting an optical finish with an acrylic/polycarbonate tube - it is just a case of trying to sand down the inside of the tube, or is there another method of getting rid of machining marks?

TIP: Use a wide "figure '8' " pattern when polishing acrylics to avoid surface deformation. The circular motion used here, causes a buildup of material in the center of the workpiece.

Another good option for a very flat sanding surface is samples of polished granite worktops. Perfect for small parts as the samples are usually around 100x100mm and cost a quid. Ebay has lots.

is that Novus stuff the same type of compound that comes with headlight polishing kits? I have seen those get really nice finishes on opaque acrylic

Nice work. Very thorough and informative. Do you know if it's possible to solvent 2 thin pieces of acrylic together to make a thicker piece without compromising too much clarity? I would think if you clamped the pieces together tightly it would eliminate any air bubbles.

Do you think that the 2000 grit paper will help buff out the mistakes from weld on adhesive?

I'm a recent sub and I just wanted try say that I'm really enjoying your channel!

Confirmed! Polishing is best compared to acetone vapour. So here is what happened. I've always been using the green compound and buff wheel on welding mask, headlights, the case that holds my swim goggles, the actual swim goggles. All of these gave me satisfactory results until I had problem with Arena swim goggle internals. After a perfect polish it was like looking through 600 grit cloud. It will not work. So I heat gun a can of red bull with acetone and it 85% fixed it, but I tried on the already polished swim case for example, it didn't make it any more clearer, in fact holding it for too long makes it melt, scratches appear. Now I don't know which is polycarbonate/acrylic or if there is a way to tell. My adice is if you can't sand/polish only then bring out the acetone method. It's also more annoying to use, it takes a lot to heat up the acetone and the fumes don't heat up the surface perfectly. Make sure you move and don't stay on one spot. I'm actually puzzled why some headlights still look hazy/matt finish after a polish and some looks like the day I bought them. I assume some stuff is not pure 100% polycarbonate or acrylic. Not to mention Acetone ain't cheap so the fumes method is not cost effective on large pieces. Hope this helps

how did you mount the microscope objective to your camera? is it a special kit, or did you make it yourself? if the latter, can you make a video on that?

What are these lenses going to be used for? And would it be possible to build a lens that corrects for the chromatic aberration using two different blends of acrylics? Given that you would calculate the shapes of the lenses with for example OSLO...

You should deposit an anti reflective coating on the surface of your lenses with you vacuum chamber. An optical stack of TiO2 and SiO2 works well.

I made the adapter myself. Sure, a video on extreme macro photography would be a good idea.

Would the polished acrylic make better lens if covered with release agent and used as a mold for some sort of clear resin? (obviously it would have to be a negative or used as a buck for a mold.) Wonder if it would be possible to make concave lenses by spinning the mold as the resin cures?

What about using a jewellers polish wheel as a finish? I found that also worked very nice with plastic resin on the buffer....? It is much higher in fine finish than 2000 wet grit......Any thoughts? We used to use that method years ago on good plexi and wow,,,,that looked near perfect!

Thank you, it is the best basically method for me.

Have you tried cerium oxide paste too as a polish ?

Ben, may I suggest a topic to build microlens for plenoptic cameras.

believe it or not a headlight cover restoration kit is a good source of the polishes and pads.

Hi ben, do you think it's possible to make a telescope mirror this way? That would be quite revolutionary!

Great info, thanks! Trying to produce transparent bores in acrylic/polycarb after taking it off the lathe, though the bore is 3mm. Will see if I can sand in there though seems like slower RPM is helping a bit.

There is a thread over on homeshopmachinist net where one of the guys does vapor polishing. He does it in a tin can. The thread is titled "I like chemical polishing" Might be worth a look. At work we make a yearly trophy for Nike and the top crown and bottom stand are both made from cnc machined acrylic. One of the guys flame polishes it with a hydrogen-oxygen torch with a pretty standard brazing tip. He is very quick with the flame using multiple passes never dwelling, like spraying paint.

Solvent welding is exactly what I need. Did you ever make one

Will 2000 grit work well with glass as well?

using a concave shaped sanding block and a random orbit sander may yield better results. I once saw a mirror being polished for a telescope that incorporated these methods.

I just found your channel and there are quite a few of your videos that interest me. Did you ever post a video on gluing two sheets of cast acrylic together without losing optical quality? Or can this be done by the DIYer? I want to glue two sheets of acrylic together while placing designs made of very small gauge wire between them. (30 -34 gauge) without causing the acrylic to lose it's clarity. I looked through the videos but didn't find it but I may have missed it. Thanks in advance and thanks for helping keep science in our lives.

I’m trying to clear up the dust cover on my turntable. This looks like a viable solution!

Great video... And indeed please do the solvent welding video to ! Thanks !

is it possible to use acid like in glass production that uses hydrofluoric acid?

I would love to see if these methods work to make acrylic fresnel lens

Hi, have you tried making some sort of sanding attachment for the CNC lathe? I would imagine that the sanding would be more consistent if you follow a cnc profile with that attachment. Come to think of it, all 3 methods would benefit from being attached to a CNC profile.

Do you think that a record player dust cover could be flame polished? I'm in the process of sanding a large area that was solvent damaged, and the sanding is taking a long time. Could flame polishing be considered a shortcut?

I just LOVE your video's!!

This method of vapor polishing is primarily used for polycarbonate (lexan)... edges or surfaces.

Do these techniques hold true for polycarbonate? I would imagine that flame polishing would still work.

Does anyone know if it's possible to fix a tiny chip in a piece of acrylic? I dropped a drill bit on a piece of thick acrylic sheet and of course it's exactly where I didn't want it to be visible. It's perhaps a mm or so in size. Thanks!

Boss video! Thanks for the info.

you can buy #3000, #5000 and #7000 sandpaper on ebay relative cheap too, i recently bought so sheets for acrylic

A peak-reading IR thermometer, following on the flame would give a nice consistent result. I found when I was just casually doing this to repair scratches in safety glasses that best results were obtained with a very strong flame passed over the surface much more rapidly. (using a butane/propane torch) With the acrylic in the highest heat area of the flame.

can acetone be used to weld acrylic?

astonishing: i knew tht Emanuel Swedenborg managed to build lenses. Today it is the first time I humbly know something more spcific about Best REGARDS

can you use certain chemicals to etch acrylic?

Use powder abrasives, cast a tin die the same radius as your lenses, and use felt with powder abrasives + water or oil (oil is slower and more messy to clean up, but I think it gives better results). Its much easier than it seems, tin can be cast so easily and already with an embedded centered axe to use with a drill or something. The last abrasive would be rouge or cerium oxide. So with one tin cast and several felts and several grades of powder abrasive, you will make very decent lenses.

I have a boat with a wrap around acrylic window. The style is beautiful, but the window is really jacked up. It’s pitted really bad and foggy. Replacing it would be expensive, and I’d like to see if there’s a way to polish it. Any Thoughts?

I remember seeing a analog device used to evaluate lenses. Light was beamed trough the lens and a uniformity pattern, or something similar, was projected on the other side of the device. I think it was on a "How it's made" episode. I guess that with your experience you could easily build such a device... if you could identify it from my vague description. :)

According to my brain, I for some reason absolutely need to know how to do this.

Hi, i have a plastic helmet that I got for a cosplay that i stupidly ruined. I ruined it by putting a clear coat acrylic on it didn't like the way that looked and used Acetone to remove it which clouded and melted the plastic then to try and get rid of that put more clear coat on it to try and make it clear. all of these were suggestions from people who obviously don't know how to deal with plastic, I dont either. Anyway, any suggestions on a good method of getting it actually clear again?

Great vid! Please do a video on solvent welding acrylic - I frequently laser cut acrylic and join it with solvent glue, but for some reason the solvent fills the join nicely at first, but as it starts to try, bubbles start forming/seeping in and kinda ruin the join. Not sure if maybe the laser cut edges are maybe a little too "jagged" and the water thin solvent is just drying up instead of bonding.

I am guessing this is good info for polishing other plastics such as BC 408 which has a polyvinyltoluene base.

Informative video, thanks

What are these lenses going to eventually do?

hey ben could you make a video on refractor telescope? would be possible with your lenses...would be awesome!!

Could you possibly do a video on etching acrylic/plexiglass?

After drilling a hole long ways through acrylic. How can you polish the drilled hole? So it becomes clear.

So are you trying to make F-theta lenses for photolithography here?

Really nice video, thanks very much for sharing.

Ben, when I remove scratches from motorcycle helmet shields, I wet sand up to 3000 grit before polishing. Try that.

Good stuff!

I'm making lenses out of polycarbonate with a belt sander, Thinking about trying the upturned vapour bucket technique.