Milled Lithophanes

macboy · April 6, 2019, 2:38pm

I’m finally getting around to testing Corian on some lithophanes. I’m using 3dp.rocks to create the STLs and then ESTLCAM to generate the tool paths. I’ve figured out that you can just step the STL down the req’d amount and avoid a separate surfacing run. That was nice to learn (my corian is 1/2”).

What are you all using for generating your tool paths and how? I don’t like the margin that ESTLCAM adds in and am going to try adding a black border around the artwork and using no margin. That should fix that bit for me. I’ll know this evening I guess.

I read one thread on another machine/software setup and the guy was using a single finishing cut (no roughing) with a 45deg v bit that he “told” the software was a 0.8mm ball end. Interesting. I’ll give it a try on a second piece of Corian I have…it’s not white, sorta tan coloured and speckled.

cncncnc · April 6, 2019, 8:03pm

Can you please post some pictures of your work with Corian? I’d like to see how this turns out.

macboy · April 7, 2019, 12:03am

I sure can. The test today wasn’t detailed enough at all (1/8" flat roughing pass, 1/8" ball finish pass) but the fit and finish and feel of the piece is great. It’s without question a match to the one we bought almost ten years ago…who woulda thought I’d have the tools to make our own one day.

I have two more files ready for testing tomorrow - going to try the V bit, single pass trick too. I’ve got room for three more tests on the little chunk of Corian I grabbed (sink cutout from a countertop place). I’m working with 5" X 7" dimensions for ease of framing options. Then I’ll need to devise the backlight plan but I think I have a solid idea - just having troubles finding the cord set. I ordered a few to test form Banggood but have clearly been spoiled by Amazon Prime…it’s painful to wait for packages like this.

First world problems, I know.

steevyt · April 7, 2019, 9:48am

I second the request to see the Corian work. I’m pretty sure I can get a hold of some for free, or at least super cheap and would like to see the results.

macboy · April 7, 2019, 10:54am

I started with the V-Carve test just to get it out o the way. ESTLCAm figured it just over four hours but the LCD status bar looks like it’ll be under the 4 our mark. It’s showing promise surprisingly too. I was skeptical for the first 10-15 mm of Y cut but once I saw the details starting to come out I think the method might have merit. It’s not a time saver. I told ESTLCam that the bit was a .7mm ball end with a 3mm z step and 45% stopover (although I’m not sure how the tool-defined stopover and the stopover % defined in the free machining dialog relate to each other).

This is 12mm Corian. I made the STL 5mm thick, set the max z to 5.5mm and stepped the model down 5mm which should leave about 1-1.5mm at the thinnest. No presurfacing of the material. The first cut was heavy but after that very first line it’s just nibbling off little bits as it goes (ridiculously low stepover).

Here’s the image, contrast adjusted and black border put around to try to deal with the margin ESTLCam (set to zero for this test). The artifacting was from the original - I could have fixed it a bit but this is a freebie. And it’ll be interesting to see how the cutter handles it.

[attachment file=96015]

And here’s what it’s looking like after an hour and some:

[attachment file=96016]

I’m working final size in the off chance it works - then I’m done this one. It’s Sunday, nothing better to do than listen to the Dewalt buzz away.

This chunk was free and they have random scraps from time to time - they just leave it in a bin out back for hobbyists to come grab. If I get right into this though I’ll need a supply line which I can set up through them. They can order it in at 6mm already which is what I’d do but I think he said a 2’ X 12’ sheet was $350 or so. That cost would have to be factored into the final product price but it wouldn’t be huge. I think that’s something like 5 or 6¢ per square inch?

You guys in the US can order something called Candlestone which looks even better than Corian if what they say is true. Google it - you’ll find it right away. I contacted them about getting sheets shipped to Canada but never heard back from them.

macboy · April 7, 2019, 1:09pm

I think the ESTLCAM guesstimate was right. 4:17 in and just about finished to the shoulder level. Hahaha. It’s still looking promising though and the large black expanses feel very smooth to the touch.

[attachment file=96042]

macboy · April 7, 2019, 2:53pm

Well I’d call that a success. This was a 60 degree V bit and from what I can tell if I used a 30 degree instead, the vertical separations between black and white would be less stark but you don’t really need absolute vertical.

Total milling time was 5:46 so quite long. The next test which will be this evening and tomorrow morning will be using a 1/8 flat roughing pass and a 1.5mm flat finishing pass. That’s the finest I have currently. Different photo so not a direct test for comparison. Just final finish vs machine time in general.

Looking closely at this one it looks like I could reduce the thickness in the STL - the woman’s eye could use more detail. Of course part of this is the shoulder of the bit used here and the depth of the STL (4mm) as well as the high contrast image.

[attachment file=96054]
[attachment file=96055]
[attachment file=96056]

macboy · April 8, 2019, 9:21pm

Now we’re getting somewhere! I think I know exactly what I need to do to get where I need to be on these. This is so, so close now.

The specs:
7" wide X 5" tall
12mm Corian, not presurfaced to 6mm

STL prepared 5mm thick with 0.8mm min thickness. ESTLCam settings above.

The roughing pass was one minute shy of 30. The finishing pass:

[attachment file=96286]

Here’s the first one I did last night. Used the 1/8" flat end mill for roughing and the 1/16" flat end mill for the finishing. This means the bit is still about 1.5mm dia and it shows in the image. It looks like one of the poster daub filters in Photoshop. Close, but not close enough.

[attachment file=96287]

So I re-ran the roughing pass on a new spot on my test piece and then recreated the finishing pass using this bit (I THINK I got this from Ryan?? If I did, it’s a 1/8", 45deg v bit):

[attachment file=96288]

I set the bit up in ESTLcam as a 0.4mm 45degree bit with a 2.5% stepover. All other parameters “standard”. The result:

[attachment file=96289]

I could visually see the change in the amount of detail coming out while it was running. The ladies’ hair, the details on the girl’s shirt on the left and the facial features. Such a difference!

Now have a look at dude top right. See how his hair shows the angle of the bit I used? The instructions I started down this rabbit hole said he used a 30 degree bit and I know that would minimize this ledge angle I’m getting here.

[attachment file=96290]

Still, looking dead on, it looks great. Oddly if you look from side to side the way the faces are cut into the dark hair, they sorta follow you around like those Jesus pictures where the eyes follow you around. Kinda cool, kinda creepy.

[attachment file=96291]

One other thing that I will try and am certain it will make these perfect is to drop the thickness of the lithophane to 4mm instead of five. That, combined with the 30 degree V should minimize the angled ledges enough to be passable. I did read somewhere last summer about a guy that blurs off the edges where steep ledges would form - that’s a possibility too. You just need to minimize how much of the edge of the cutter gets put into service.

Fun fact about the Corian…it makes snow! The low stopover makes for a very fine cut and the stuff just powders off. I’m actually going to remember this and will certainly use some of it for use with my stop motion stuff (one day…that’s way down on the round tuit list). Ironically it’s not the first time my MPCNC has been snowed in. But last time it was REAL snow. HA!

[attachment file=96292]

The roughing passes create very coarse strands of Corian which are statically charged so they stick to EVERYTHING. Including the rails. The machine needs a good cleaning now - both axis are rough rolling from impacted material on the bearings and tubes. It does clean off well but definitely gets impacted.

Another GREAT day in V1 land. Thanks yet again Ryan!

jamiek · April 8, 2019, 10:37pm

Beautiful work!

I am curious, what does it look like if you illuminate from the opposite side and view flipped? I am seeing halos along the edges that my brain cannot explain with physics. I am wondering about lateral diffusion of light in addition to the usual darkening due to thickness.

macboy · April 8, 2019, 10:57pm

The halos are just the angle of the ledge there. Where it steps from black to white it should be a vertical edge but with a V bit you cannot achieve that so you get an angle instead. This causes the halos you’re seeing. I feel that if I reduce the depth of the STL file to 4mm and then use a tighter angled bit (30 deg) I should be able to minimize those artifacts.

Lighting lithophanes from the other side makes the visual very diffused and blurred as far as “they” say. I’ll try it though and post for all.

forcerouge · April 9, 2019, 12:04am

My neck hurts now.

Impressive work though.

barry99705 · April 9, 2019, 4:50am

That’s weird, I thought the dark parts were supposed to be the tall parts and the light stuff was the engraved.

macboy · April 9, 2019, 9:00am

Weirder still - it IS! I think the angled ledges create some odd perspective shift that messes with you which is why the girl on the left looks like she’s turning her head if you rotate the image slowly. I should try to video that. It’s crazy.

Sorry for the sideways pics…I’m too lazy to open them and rotate them all. Maybe I can do that on my phone. I’ll try.