UV laser for cutting plexigalss

[attah]

So I got this crazy idea.. cutting plexiglass with a laser.
I found out that normal plexiglass has almost 100% blocking/absorbtion below ~360nm (see: Link ) So I thought that maybe i should try my hand at building a cnc machine that cuts with a laser. From what I find it sems that I need a very powerful laser for this, and being a UV one that will be very expensive, and possibly almost impossible to buy anyway.

So I'm posting here to ask both wether this is more than just a bit insane, and where I might find a sufficiently powerful laser... What powers are we looking at, 1+ watts? (the budget isn't the main concern, but I'd really like for the entire build to stay well inside a months salary)

 

[flare09]

Uv diodes aren't cheap...below 405nm I believe are all custom made. And an array of 1w diodes enough to vaporize the plexi would be...expensive. plus the filtering to take all that vaporized acrylic out of the air....

 

[Cyparagon]

Although I've seen a guy make a successful engraver with frequency tripled YAG (355nm), you should use CO2 like the rest of the world

 

[Bionic-Badger]

Y'know, plexiglass is also opaque to far infrared (heat) too. That's why most people use CO2 lasers for cutting that material.

The only reason you should use a UV laser or something like that is if you need to expose something that requires a very short wavelength. Plexiglass cutting doesn't qualify.

 

[flare09]

Well considering the setup required for cbc laser cutter, upgrade the steppers and throw a spindle on it. And just cut it I bought a shapeoko which im sure it can do plexi just as easily as a laser cutter.

 

[Things]

Well considering the setup required for cbc laser cutter, upgrade the steppers and throw a spindle on it. And just cut it I bought a shapeoko which im sure it can do plexi just as easily as a laser cutter.

Not quite that simple, laser cutters tend to be designed a lot weaker structurally seeing as they only have to carry optics. They are also designed for speed, unlike CNC mills. Belt systems are no good for CNC mill use at all

 

[flare09]

Yesyes...im going to see how well it works out when I get it. If the belt drive slips I will move to threaded rods. Im making a flashlight that im going to post soon. Its made of about 18awg aluminum, stacked, and I hope the shapeoko can handle that so I can make a better one.

But I guess you're right. It does require a more sturdy base to cut rather than burn.

 

[attah]

So much response! You're all too kind

Exactly.. if I get the forces acting to counter the mechanism to near zero I can get good precision without beefy steppers and beams etc. Speed is not anything I have considered yet. So while it's nice, a somewhat slow machine will do fine.

Okay, probably I should use IR. Has anyone got any good exaples of a sufficiently powerful laser? I'm not finding all that much on the subject except for complete cutter systems.

But just to play with the thought.. about what power would do the trick for an IR laser? Is there anything at all like that availble? I read that some people here got nichia uv diodes, but has anyone seen their uv laser diodes in action?

 

[Bionic-Badger]

20W? 40W? Remember that these are CO2 lasers and put out far infrared (heat), not the 808nm IR laser diodes that are for pumping green lasers or other low-power applications.

 

[attah]

And heat has a greater refraction/diffraction/... related loss you mean?

A little correction.. I really meant to ask what power would be required in the UV spectrum.

 

[attah]

I just realized I kind of owe you guys an update. I restarted the project a while back and it's taking shape pretty good. I didn't go for UV, but still a low end of the spectrum diode laser. DTR's Nichia NDB7875 with driver and G2 lens.

Here is the damage it can do:

That's 3mm black cast acrylic. Extruded doesn't bulge as much.. but i like these edges better. The pi is 40mm high, and takes about 15 minutes to cut.

Here is the rig:

Note the hex key that used to be part of the construction.

And here i have cut acrylic parts to replace it:

Still some way to go before it is done, but I'd like to think I have sub-0.1mm accuracy now. Main focus at the moment will be fume extractor and making the software a bit more user friendly. (I also took the opportunity to learn how to the make software and stepper drivers, but i guess that's for another forum)

I wonder how much a NDB7A75 will improve the speed at which i can cut.. or if it simply would be impossible to focus to a small enough (again 0.1mm) and uniform spot. Anyone that have tried?

 

[trussmonkey25]

I'm working on a laser pantegraph. Lightning Stalker built the laser head and supply. It seems to cut cardboard and thin woods. The devil will be in what lens or lenses you use. Can you come up with adjustable focus controled by your software? Or have the head move up and down as it burns through thicker stuff? I dream of a desktop cnc using a diode laser for hobbiest.

 

[trussmonkey25]

Great work!

 

[attah]

Thanks! Yes, it's a b*tch to focus. Well, I have looked at it and it's not impossible.. but i have yet to find a good motor for that. And i'd all but run out of gpio on my raspberry pi. I think mine will remain manual, but perhaps with a worm gear. Moving it up and down seems like it would mess with the accuracy and the driver is not made for a third axis... although it could be redesigned. But i guess that does limit the thickness of what i can cut.

 

[Sigurthr]

The lens will also affect the cutting thickness capacity, as depth of focus comes into play. I.e. if the beam is very quickly convergent in terms of delta beam diameter per unit of distance, you'll have a shorter usable depth before the beam waist rediverges. Longer focal lengths tend to produce better depth of focii, but there's complex math involved for more accurate answer than this.

 

[upaa27]

The most convienient thing a uv laser would be used for is on aluminum as its photoelectric threshold is 266nm