Drafting For Laser Cutting
Tim-M
Posts: 522
I know several of you are doing quite a bit of laser cutting... what software are you using to draft your cutting specific designs and what are your recommendations?
Thanks,
Tim
Thanks,
Tim
Comments
-Phil
As someone with little electronic drafting experience and at the beginning of the learning curve, I appreciate the input. I have started with CorelDraw but have wondered if other software is better suited for technical style parts that are more dimension critical and want to know where my learning time is best directed. I expect to use Corel mostly for engraving and marking, but my question is directed toward the cutting, and part outlines.
Any other advice for a drafting rookie?
Tim
I typically use Inkscape for rough drawings, then finish in an old version of TurboCAD (because my cutter is on a Windows 98 machine; the software I use is DOS-based and uses a real-time stepper controller). I have Corel Draw but only use that for vinyl cutting. I don't like its interface quite as much as Inkscape, but Inkscape (being open source freeware) has a number of bugs that Corel Draw and TurboCAD can overcome.
For real dimension-critical work, I'd look at a true CAD program. There are some free open source ones you can try.
So it really depends on your workflow. Start with your machine and its file format requirements, and work backward.
-Phil
It may vary for different laser cutters, but with my Epilog I NEVER do any exports or conversions of my drawings into another format to be able to send them to the laser. I print straight through using the Epilog driver just as if it was a plain old inkjet printer. Obviously the printer properties dialog has a few more options, lol.
I have studied the laser subject for several years, but until now have only worked with a Synrad Galvo system and remained within it's proprietary software.
Savings are in the works for a 'box' style machine, and so the reason for wanting your drawing feedback as it relates to cutting.
Rich - I looked very hard at the Kern flatbed machines mentioned in another thread, and in doing so, discovered a pretty serious design flaw and so turned another direction.
I fully support US made metal laser tubes also for three main reasons:
1 - The rise times of US made metal laser tubes is hugely superior to that of glass tubes, in the tens to hundreds of thousands of times better.
2 - Beam quality is far superior to that of glass tubes.
3 - The US made optics are much higher quality, although this seems to be improving some over time.
All of these add up to much better energy delivery, much more machine and material capability and far better cut quality and end result, at a much higher speed.
My glass tube machine cuts as fast and as well as any mainstream machine. When it comes to rastering, that is where a mainstream machine will excel, especially a Trotec.
1- Sometimes important for engraving, not cutting. Not all glass tubes are equal either, there are cheap ones and there are good ones.
2- Maybe, that is what people say anyway. When I test fired mine into a block of acrylic the crater left behind had a TEM-00 profile.
3- Optics in lasers are very simple - a single plano convex lens - maybe a beam expander. You won't notice any difference until you run hundreds of watts, maybe thousands. Assuming of course you don't get the absolute cheapest, worst quality lens to begin with. I could pay $400 for a lens or I could pay $40. Won't make any difference to the cut. Nothing to stop you from using US made optics if that is what you need.
Mainstream laser companies follow the "charge as much as you can get away with" pricing model. You'll notice that going up in bed size will cost many thousands of dollars, even though all the important bits remain the same. I would have had to pay more than eight times as much to get a mainstream machine and it still would have been smaller than mine - mine is more than twice as big as Epilog's largest model.
I highly suggest you join the Sawmill Creek forum. The engravers section is very active and there is a huge wealth of information there on all things laser. There are people there with cheap eBay machines and there are those who run large businesses with dozens of machines - and everything between.
What is this design flaw?
Wonder if it's a PostScript driver. Have you ever captured the output and looked at it? I would assume unless you're printing something rasterized you always want a vector output.
Me, it's all DXF in and G-Code out, on cranky and whiny stepper motors no less. This is the 20th century, after all...
http://openscad.org
Also, many current versions of popular vector drawing programs (like Adobe Illustrator) will let you export directly to .DXF for lasercutting and the like.
As far as Kern flat-bed machines are concerned, I was very interested in them myself up until, as part of my 'homework', I visited several laser shops who were running them. After detailed discussion with the owners, it was learned that they all exhibited the same trait of holding better tolerance down the center line of the table, and it got worse the farther away from the center line you worked. It was also unpredictable in which direction along the x axis it varied, and by how much. Several of the shops had had Kern service out for the problem many times over long periods of time, but with no improvement. This was enough of a problem that most of them had resigned to not doing closer tolerance work on the machines at all, or were limited to working down a narrow band of the center line of the table only.
As part of my 'homework' I ended up meeting the independent designer of the Kern machine who described the following, making it all clear: The original design did not have the above issue, but Jerry Kern felt it was too expensive to produce, and therefore changed it for cost reasons. The machine is built and sold, the gantry is driven with a single lead screw mounted below the table, down the center line. If you picture looking down on the machine from above, the bearing rails that the gantry rides on at each end, have enough play in them to allow the gantry to pivot about the lead screw, each time the gantry changes direction. This makes the motion of the gantry ends unpredictable and constantly varied, producing unhappy customers... so much so, that Kern later developed bed mapping in the control software. But because the problem is a design issue that produced unpredictable gantry end position, the bed mapping proved only to be an improvement and not a final fix. The original designer told me it was easy to prove, just walk up to a Kern machine and see how easily the you can wiggle the ends of the gantry back and forth...
That same designer built a machine for me which uses two lead screws, one at each end of the gantry, and it's rock solid. Each lead screw is driven by an independent motor, and the two are electronically sync'd in motion. As this design and machine are not production built, I couldn't see one on a sales floor before hand, but I did talk with several customers running them who proved fantastic accuracy and repeatability over the entire work area. I have been working to install my machine, but additional ruptured disks in my back have changed the picture quite a bit, so my personal results will be delayed for a while yet.
I have a similar issue with my machine. There is some variation on the X axis. If I cut several squares in a row each one may vary in width by a couple thousandths. Rarely is this an issue but one customer builds aquarium sumps with several baffles. If they are slightly different widths then the sides of the tank will not be straight. My solution is that I cut each baffle in the same place on the bed. I have lessened this issue slightly by switching to a chain drive for the X axis. Mainstream machines use a linear encoder so they sidestep this problem entirely.
-Phil
and previewing is easy (even in Emacs). Inkscape will import PostScript to generate DXF for a laser cutter,
and the design can be tweaked there if you want.
The nice thing is being able to tweak dimensions and hole sizes easily, you just /def them as constants.
Typically each laser cutter has its own beam width and acrylic stock varies in thickness, so its handy
to be able to tailor the details to your specific setup - if you want interlocking parts like joints in a box.
You can iterate easily, go gears/sprockets are doable.
To a programmer every problem is a programming one!
If the machine accepts something like a DXF then it'll have on-board conversion software that builds G-Code from the drawing, ie: CAD to CAM conversion. Many machines won't do that themselves but rather relies on external CAM software.
EDIT: Ah, I see Gordon has sort of mentioned this already in post #10.