Some thoughts about a laser scanner for SLS

I just wanted to "get rid" of some thoughts about an idea to build a laser scanning setup for SLS. It's really just an idea and I haven't even found a solution to implement this yet. This idea probably won't work out but maybe it's still of some value to someone.

The main problem with two rotating mirrors on a galvo is that your focus distance is varying with X/Y so you would get different intensities / sizes. Mounting the laser higher with a larger focus distance will reduce that effect and might actually be the best way to go about this - adding a bit of height is cheap. AFAIK the industrial solution is to use F-Theta or macro lenses but these are expensive. A scanner is ideal because the moving mass of the mechanical elements is small (similar to a deltas mass is smaller = faster acceleration / speed).

The "simplest" approach is to use a X/Y gantry e.g. a coreXY gantry. Of course you reduce laser speed vastly with this! You'll at least have to move the Y axis with the print head. Basically you need lots of material that you don't really need. The advantage is of course we know exactly how to build this because of FDM printers and can reuse lots of knowhow, cheap parts and electronics and firmware. An additional advantage would be that potentially in the future we could add an CMYK inkjet print head to do full color SLS printing. Some folks are working on this but it's somewhat far off.

So my idea is to move the mirror towards the laser while rotating it in either X direction so the focus distances stays the same. How such a mechanism would work I don't know but the formula would be:

mx(x) = -0.5 * x2 / (z - x)

x would be the x part of the x/y signal and Z is the height of the laser above the print bed (15 in the graph shown - it gets flatter the higher up you put your gantry).



So if the height of the laser would be 25cm above the print bed and you want to have a print bed of 20cm x 20cm you'd have to move the mirror about 3.3cm closer to the laser source at x = 10cm.

The wxMaxima code to visualize this (FD = focus distance, LX = ???)

FD = z - LX;
A : (z - mx)^2 = z^2 + (x - mx)^2;
B : solve (A, mx)[1];
mx(x, z) := ''(rhs(B ));
plot2d(mx(x, 25), [x, -15, 15], [y, -6, +6]) $

So the next step would be to get to the formula to the rotation. And then visualize this in an animation to show this movement. Maybe you can visualize by showing how a lever would move or the distances would look like or where circles attacked to the edges of the mirror would meet. I really just wanted to share this idea which may be crazy or stupid. Maybe someone smarter can work this out further. Maybe I'll go back to it later.

Edited 1 time(s). Last edit at 06/24/2015 09:47AM by Dejay.

... while this can work with high quality gas- and NdYAG-lasers and pretty expensive fiber-lasers, 'adding height' or focus distance is really bad with laser-diodes - the beam quality is so poor, that with a focus distance of >70mm the shape of the spot will resemble not a round spot, but a line/rectangle with roughly 0.4mm x 0.1mm and even worser with growing distance

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Quote
VDX
... while this can work with high quality gas- and NdYAG-lasers and pretty expensive fiber-lasers, 'adding height' or focus distance is really bad with laser-diodes - the beam quality is so poor, that with a focus distance of >70mm the shape of the spot will resemble not a round spot, but a line/rectangle with roughly 0.4mm x 0.1mm and even worser with growing distance

Thanks VDX. I had not considered this. Somehow I keep forgetting the essentials of optics haha - focus and image size are two different things so of course if you increase the focus distance you also increase the spot size. A moving laser head allows for a much shorter focus distance.

I have a question: You mentioned beam shaping before and I saw a prism like setup to scale the beam size on one direction only (but still rather expensive). But am I right in the assumption that if you would shape the beam to be really thin before hitting the mirrors you would get a problem with the mirrors breaking fast? Right?

So I guess it's back to a X/Y gantry as the go to solution for a cheap SLS printer until cheap beam shaping / lenses become available. That should allow you to use cheap diodes with a short focus distance. Hmm, maybe have a scara / wally arm instead of a gantry. Without a heavy hotend this could at least be faster than a X/Y gantry.

Edited 1 time(s). Last edit at 06/24/2015 12:29PM by Dejay.

... yes, high powers with 'thin' beams gives enough energy density to heat/wreck the mirrors - I've killed some of the 8mm wide mirrors of the cheap chinese show-laser galvoscanners with a 3mm wide beam with averaged 8W@1070nm from my fiberlaser or with 15W@975nm from a fibercoupled laserdiode ... and some 20mm-mirrors too with a 8mm wide beam and doubled power

I

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]

Thanks. So that means you'd have to have the lens after the mirrors - which means the lens has to be rather large?

I guess a moving laser head is really much easier to implement.

... in a common galvo-head the (f-theta!) lens is shortly after the mirrors and so designed, that the resulting focus moves exact in a plane with fixed distance.

With my first setups I've used a beam-expander to grow the beam diameter from 3 or 5mm to double to triple size, then a lens with 250mm focus distance 'started' to converge the beam, which was fed across the moving mirrors with some 80% of the previous diameter, and the resulting focus was something like 150mm below the mirror head.

The angular distortion was maybe 3mm from outside to center of the working are of roughly 120x120mm, so enough for testing.

I have several f-theta optics for the fast moving galvo-heads, but I've too assembled a much slower moving head with servos and big 45mm-mirrors, where I can use a simple lens before the mirrors on a moving stage as auto-focus to eliminate the angular distortion per software ...

---

Viktor
--------
Aufruf zum Projekt "Müll-freie Meere" - [reprap.org] -- Deutsche Facebook-Gruppe - [www.facebook.com]

Call for the project "garbage-free seas" - [reprap.org]