[Zach] over at his channel Breaking Taps has put up an extraordinary account on manufacturing some homemade acrylic lenses. In the end, not only does he produce some beautiful concave lenses, he also covers the complete manufacturing process, from milling the aluminium die used for injection moulding to tweaking the parameters associated with injecting the actual acrylic, he even goes over the limitations of optics produced in this fashion.
What caught our eye in particular, was how [Zach] used the finished product to practically demonstrate photoelasticity originating from the stress induced by the moulding process. You might be familiar with describing the optical properties of a material by a single number, i.e its permittivity. But what happens if in addition to altering speed, the material also alters the polarisation and direction of light depending on the stress distribution within the material? Whilst a quantitative answer gets a bit complicated you can check out [Zach’s] additional videos to visualise the answer in a pretty and colourful way, without resorting to fancy computer simulations! If however, you really want to persist with the simulation route, check out our article on stress analysis in a totally different setting using Finite Element Analysis.
You can try to anneal the lens in oven to remove the stress patters.
That’s a great idea, think I’ll give that a shot!
I have learned recently it looks like my favorite style of safety glasses (smith & wesson phantoms, only thing I use) might be being discontinued, and was thinking about doing something like this later in the year as a project for cnc mill Im saving for, but using an existing 3d scan I have of my face, and polycarbonate to make my own custom fitted safety glasses similar to the phantoms.
This desktop injection molder may be of use to me. Have some research to fo now, very cool project Zach
If you are making something so bespoke to your face injection moulding probably isn’t worth the effort. Will be of no use to anybody else, you don’t need enormous heaps of them which injection moulding is really all about – you throw out the first few each time as the system is purged of the last go, the small machines at least seem to only really start throwing good once everything warms up just right and then will probably heat soak the mould too much so only a heap in the middle of the run will be good (I’m no expert, but I did look into it and quickly realised it makes no sense at all if you don’t want lots and lots and lots and…… lots of the same part and to have to sort out all the failures – not for a hobbyist in general IMO but great if you are trying to go into a small business or just like playing with new toys)
You’d probably do better to stick with straight CNC work to shape and then some polishing to remove the tool marks.
Yeah, must everything be practical?
I cannot explain how much I detest normal safety glasses with the thick rims obfuscation of my view, and thick clunky plastic ear pieces that hurt.
Those glasses specifically give me no obstruction in view and they wrap around the sides of my face fully and perfectly while still staying close to my face closer than anything else on the market- which lets me attach Bausch & Lomb swing down magnifier for watch work with my eye an inch away from a 10000 rpm spinning part.
I don’t want to use anything else for the rest of my life and I would like to optimize what is already available- so it’s not about making these to make a profit.
I am aware of the losses that would be incurred but I don’t think you understand how much I really value good safety glasses! I already have them on for at least 8 to 10 hours every day- so to me it’s a viable project to make my own.
I used to be a plastics machinist for a time- and I can tell you from experience trying to mill polycarbonate to an optical finish is extremely difficult if not impossible even though it cuts very well. There will always be micro-cracking on facing tools no matter what geometry you use.
Thermoforming or injection molding is the only way to do this that is practical. It might be doable to machine a blank and then thermoform the blank afterwards to meet curvature and surface finish requirement for polishing.
‘must every be practical?’ – Of course not, just wished to point out the pitfalls of a solution I’ve looked into and decided against for those reasons. The fun you can have with such a tool can be justification in itself if you choose to spend your hobby? budget on it good for you!
I can well understand the want for good safety glasses. I was just expressing my opinion that CNC operations, perhaps followed by heat shaping and final polishing will probably be a better on your budget (probably give better results too after polishing) when you are only trying to make a tiny number of very bespoke and personalised to you items.
If you do go for the moulding process I’d love to see how you get on.
Though you might find just vacuum forming sheets around a blank will suit you best. This method does tend to thin the middle of a part quite abit, and not all materials form well but with a thick enough sheet of the right material to start with that won’t be an issue – Few iterations of the blank perhaps to remove any visual distortion. Its star features being its really really cheap, and the surfaces can be almost perfectly smooth right off the form so you shouldn’t need to do much post-processing.
Also acrylic can be polished by heat, so it might be win-win as long as there is no substantial amount of warping or outgasing…
Have you thought about all the optical products that can be improved by varying permitivity in a structured way in translucent and semi translucent material? If i were you i would patent it now, before I do hehe
The birefringence in polymers is caused by two sources:
– some of it comes from mechanical stress. If you bend it or pinch it, this pattern will change. Uneven cooling can cause this. With annealing this can go away
– some of it comes from orientation of the molecules
I think those stress patters if they could be controlled could be the basis for development of some interesting optical devices with practical applications. This has been an amazing read and i have no idea how it ended up in my google feed but i have to pass this on. Great work.
I am interested in the injection machine and type of plastic used.
Machine is a “Proto-Ject 150HP” from here: manninginnovations.com
Resin was clear acrylic, but unfortunately it was an eBay special so I don’t know the manufacturer, data-sheet, melt-flow index, etc.
I’m surprised he’s using a material with such high heat deformation as aluminum. Should be using a tool steel
I imagine the majority of HAD readers don’t have the equipment to machine tool steel. It is likely that many HAD readers have access to a CNC router that can machine aluminum.
Op here :) So aluminum was chosen for a few reasons, of varying importance:
– Cheap, had stock of right size in my shop
– As Alan mentioned, I have a router machine. It’s stout, and can do steel (I have videos on cutting mild steel with respectable finish), but still… Al is just easier and faster to work in, generally with better finish
– On a little desktop injection machine like this, the cycle time is so long that the mold hardly heats up from each shot. Heat deflection is not really an issue here given the tiny amount of thermal mass that is being injected into the mold. In fact, I really should have been _heating_ the mold to get better finish
– Further, a simple design like this (big spherical cavity) isn’t likely to deflect at all from heat, e.g. there are no thin segments that would see a big gradient across the cross-section.
And finally, aluminum molds are used in the industry a lot, although typically a different alloy than I was using. The main reason to avoid aluminum molds is because they wear out faster, and their higher heat coefficient means they require more stringent cooling in industrial machines. But it’s all moot for a little desktop machine like this :)
That is some really beautiful work. What CAM software do you use to generate the mold and the cut path?
Thanks! I used Fusion360 to model it and generate the toolpaths. 2D and 3D adaptives to clear out most of the material, a spiral finishing pass (with 2thou stepover) for the spherical lens surface and contour / “steep & shallow” to tidy up the gates/runners/vents/etc.
https://i.imgur.com/Vnwoakl.mp4
Direct link to the video in the article, since every single media player on the internet now has to be an overworked piece of trash that doesn’t work on any alternative browser or respect privacy, cookie limiting, adblocking, javascript limiting, or really anything the user does at all. Seriously, what is happening to all video players online? like 80% of them have become unusable unless you use the most current chrome stock with zero pro-privacy configuration.