Hello! / Laser Marquetry

[Dave Fifield]

Hi Folks,

I found this forum during a web search on laser marquetry and simply had to join. Even though I just got here, the place feels like home already. Hope to make all your aquainances soon!

I purchased a laser engraver to enhance my woodwork. One of the things I like to do is marquetry. I have the basic skills down using the fret saw/scroll saw double bevel method but want to try my hand at some more intricate pictures using the laser to cut the veneers and hopefully speed the process up some. I have a 35W Epilog Legend 24TT and all the usual software (Corel 12, Photoshop 6, Photograv 2.11 etc.).

I've spend quite a while reading through tons of really informative and highly useful threads on the forum - many thanks to all you experienced laser/Corel experts! Most, but not all, of my newbie questions have been answered in the threads. Here goes with the two big questions (with lots of sub-questions!) that are still bugging me:

a) Laser "Kerf" and how to deal with it?

I made a very simple picture (a square, triangle and circle on top of each other) as a test piece. I vector cut the veneer pieces easily enough (one sheet of veneer at a time), but when I put them together to make the picture, the laser "kerf" (gap between the pieces) seems much larger than I was led to believe it would be.

I ran another test job with a bunch of vector cut lines, each one mapped (via color) to a different speed setting on the laser. I tested cuts from 20% speed in steps of 10% up to 100% (power was 100% for all cuts) and found that there is quite a bit of difference in the kerf width. As expected, the 100% speed / 100% power cut gives the narrowest kerf and easily cuts the thin veneer material. I measured it at about 5 mil on the top of the veneer and about 3 mil on the bottom. Why is it thicker on the top?

I had ran the initial test piece using 30% speed / 100% power, so the kerf was about 15 mil per the second test - no wonder it was "sloppy" marquetry! I re-ran the square/circle/triange job using 100% speed / 100% power, but even with the narrowest possible kerf, the gap is really too big for good quality marquetry.

I imagine that some of you have already come across this issue and solved it. The solution I plan to try is to increase the size of the inlaid piece by a fraction (how much? how to do in Corel Draw 12?) so that the pieces fit snugly together. Is there a simpler way? What am I missing?

b) I have the rotary attachement for my laser, but it's only any good for round objects up to about 6" in diameter. That will work fine for vase-like turned objects, but most of my turnings are much larger bowls/vessels.

What I think I need is a turntable rotary attachment, and another mirror (prism?) to turn the laser beam through 90 degrees. Then I could use the z axis motor (table height) and rotary turntable to engrave stuff all the way around much larger diameter bowls.

Has anyone done anything like this before? I realize it would probably be very difficult, requiring firmware changes in the laser as well as driver changes on the PC. Is this something I/we should lobby the laser engraving machine manufacturer's for?

Cheers for now,
Dave

[Joe Pelonio]

Welcome Dave. I have a 45 Watt 24tt, and no rotary so I'll address your first issue. Think of the laser's beam as an hourglass. The center, or thinnest point is what you focus to the material. That wider area of beam toward the bottom of the material is what makes the "kerf." The thicker the material the worse it is. For the kind of work you are doing there are little tricks like the one you already found, using more power and less speed than needed, sometimes cutting with the bad side up will help.
I'm currently doing 10 jigsaw puzzles 12x18 with digital prints applied to 1/8" acrylic and with more power and wet transfer tape on top I can get a good fit without damaging the images. On the other hand, I cut a lot of 1/4" acrylic and while I haven't measured it it does seem better when I manually focus to 1/8", spreads out the "kerf" between the top and bottom. I did some 1/2" acrylic in two passes, manually focused to 1/4"
on the second, and there was a dual step kerf.

[Bob Tate]

Hi, and welcome. I have done some work like this and need to get back to it. Different woods burn differently, that is, larger kerfs or smaller, I suppose related to the amount of pitch in the wood. I also assume it has something to do with the amount of moisture in the wood. I cut my own veneers and let them dry sufficiently, even though our average humidity is only about 28%.
The angle (slope) in the vector cut is because of how the laser beam focuses. Look at a flashlight beam. It will be wider until it gets to the focus point, narrowing to that point, then getting wider beyond the focus poing. I have tried manual focus and had some success but that is tedious. Take advantage of the "feature". Burn (raster) the pocket for the piece. Vector cut the piece upside down and it fits much better. Another issue, as I said earlier, various woods burn differently. Control this with PPI. My Epilog manual gives me settings like 500 PPI to Vector Cut wood. That is hot, charring edges. The charred edges give you a bigger kerf. At higher settings the edges are black or close to it. I cut quarter inch Hard Maple at 75 PPI and the kerf was so small I had to force the pieces out of the sheet. They were cut cleanly, the kerf was just tight. The pieces kind of squeaked on their way out. If you cut Alder at the same settings you do get a larger kerf but it is still acceptable if you are cutting the pieces upside down. I think if you cut some thinner wood like veneer with the lower PPI settings and do it upside down you will be surprised with some great results.
If you need any further information I will be glad to help you. I would love to post some pictures of my work but I am terrible at taking pictures.
Bob

[Rodne Gold]

Hiya , welcome
In the first place , a program like Engravelab or profile lab is way better for marquetry/inlay than corel as it has built in routines for inlay work and you wont have to compensate for male/female pieces or for kerfs.
A 1.5" lens will also give you a much smaller spot size and a cleaner cut with veneers under 1-1.5mm thick. Ideally for cutting you want 100% power and you want to vary speed , however as BOB says , there is a setting called PPI that is vital for wood , it sets the amount of pulses a laser produces in an inch of travel (a laser does not cut by firing continously , it drills many small holes), too few pulses and you get a stamp like perforation , too many overlapping and you char the piece. The reason the kerf slopes is not likely to be beam diversion on thin sections with a std 2" lens , its most likely a misaligned beam which is not striking the final lens dead centre. If it is beam divergence , there is a very simple "cure" and that is to focus on the material upper surface and raise the table by 1/2 the material thickness.
My rotary works like a lathe and we have dumped the horrible rubber cones and fitted proper jaws etc for more utility. I doubt the laser guys will listen to your request for the turntable , its the first time I have ever heard someone wanting to do this and it would really be a mission to implement.
Joe pex is generally a different case re slanted cutting as it acts to some extent like a wave guide and thus beam divergence isnt as big an issue. We also did some big jigsaw puzzles with digiprints apoplied and found the best way to do these was print , laminate and then cut on a Roland print and cut machine and then laser cut the pex and hand apply , you have to cut the digi print with an offset inwards with respect to the lasered pex pieces as if its exactly the same size its almost impossible to apply and even if you do get it right , the slightest projecting edge over the pex will make the piece snag. We had discolouration when trying to cut the puzzle in one go. We used a lam over the puzzle to protect the print when the puzzle was used and found the double layer of cast overlam and the pvc didnt laser well at all. We did however have the pex printed with a uv flatbed digi printer and got great results in a single pass.
We laminate veneers with 3m VHB double sided tape and kiss cut , no gluing etc when you apply and the lam always smoothes out ripples or warps , we use a cold pressure roller to do that , but a hand roller works well. We also use CnC overhead routers/engravers for pocketing for thicker veneers as they leave a very smooth surface for the inlay unlike lasers , it also has very precise depth control and grain is not a problem ( a laser will tend to "follow" grain in terms of depth.)
Engravelab/profile lab drives both machines and compensates for the bit used in the cnc work and even has strategies for corner sharpening. I highly recommend it for your applications. We also*shocke and horror* often use formica in place of veneers as it can be laser cut , but also laser engraved and colour filled with paints and waxes as well as having a constant colour and grain across a large production job.

[Lee DeRaud]

Here goes with the two big questions (with lots of sub-questions!) that are still bugging me:

a) Laser "Kerf" and how to deal with it?

I made a very simple picture (a square, triangle and circle on top of each other) as a test piece. I vector cut the veneer pieces easily enough (one sheet of veneer at a time), but when I put them together to make the picture, the laser "kerf" (gap between the pieces) seems much larger than I was led to believe it would be.

I ran another test job with a bunch of vector cut lines, each one mapped (via color) to a different speed setting on the laser. I tested cuts from 20% speed in steps of 10% up to 100% (power was 100% for all cuts) and found that there is quite a bit of difference in the kerf width. As expected, the 100% speed / 100% power cut gives the narrowest kerf and easily cuts the thin veneer material. I measured it at about 5 mil on the top of the veneer and about 3 mil on the bottom. Why is it thicker on the top?

Ok, there are two effects in play here. As others have noted, if the focus of the beam is at the top surface, the beam diverges between there and the bottom surface. But, as you say, that's backwards of what you normally see, and in any case is a very minor effect on stock that thin. The beam width is actually much less (at the focus point) than the kerf width: the kerf is mostly material that has been vaporized by the heat. The beam loses power as it works through the wood, both through absorbsion and divergence, with the result that the kerf is narrower on the bottom.

I had ran the initial test piece using 30% speed / 100% power, so the kerf was about 15 mil per the second test - no wonder it was "sloppy" marquetry! I re-ran the square/circle/triange job using 100% speed / 100% power, but even with the narrowest possible kerf, the gap is really too big for good quality marquetry.

You'll want to reduce the power for cutting veneer. You've done your experiment with power fixed, now start at 100% speed and step the power down until the veneer is just barely cut...should be a lot better, and the "mirror cut" technique will let you use the thinner kerf on the bottom. (Worst kerf width I've ever seen is about 9 mil...and that's on like 1/4" stock. Might want to check your focus settings as well.)

I imagine that some of you have already come across this issue and solved it. The solution I plan to try is to increase the size of the inlaid piece by a fraction (how much? how to do in Corel Draw 12?) so that the pieces fit snugly together. Is there a simpler way? What am I missing?

The "contour" function is your friend: learn it and use it. The question of "how much" will be answered by the kerf figures you get from further experiments.

Something else to consider: lasering a large-area pocket in your substrate takes bloody forever. I've had success with using the contour function to cut 6-8 multiple kerfs around the inside of the pocket at 0.02 intervals or so and then using a small router to take out the rest. I use a Dremel tool with the router attachment and a 1/8" bit. For smaller areas an Exacto knife does the trick. There's much less burning this way also.

[Joe Pelonio]

Rodne,

On the puzzles, this is the 3rd job for this customer. We use large format inkjet prints on adhesive, then UV/Waterproof overlam that's very thin. I apply that to the 1/8" plex sheet, transfer tape it, mist the transfer tape with water, and it cuts beautifully thru everything with no discoloration at all. Hand applying each piece (would have to be perfect) is a lot of work.

[Mike Mackenzie]

Dave,

Lee said it all mirror your image and cut from the back side that way the angle goes down into the pocket. He is also correct with using the contour tool. Your beam width is usally about .005 so if you take 1/2 that and contour to the outside you will end up with a really tight fit. You also need to use as little power as necessary to cut through the material the more power you use the wider the cut.

Hope this helps I am attaching a pic that one of our customers did using a ULS laser system.

[Alan Young]

<!--StartFragment --> Has any one used the offset or scaling feature of the driver to compensate for the kerf. I have a laserpro so it might not be available to others. I have tried it and had some good results. I use it for inlay but have not used it for side to side fit like marquetry.

Alan

[Dave Fifield]

Wow, thanks for all the wonderfully informative replies! You guys are awesome!!

I think I have the process for cutting veneer for marquetry sorted out now. I did a whole bunch of tests at 100% speed and varying the power (not sure why I didn't think of doing that before - thanks for the pointers Joe, Bob, Rodne, Mike). I left the frequency at 2500Hz for these initial tests. It seems that 100% speed with 30 to 40% power (depending on the veneer thickness) gives me the best cut - it just cuts through cleanly. I then ran a bunch of experiments with the laser pulse frequency with the speed/power set to 100%/30%, varying the frequency from 50Hz on up, and found the best frequency to use at 100% speed is between 700Hz and 1KHz. This gave me a really thin kerf and not too much edge burning or roughness.

With these settings, I re-did my initial marquetry test piece upside down (cutting from the back). It came out beautifully, without having to resort to any contour line jiggerypokery! I did some contour experiments too, so I know how to use that nice feature now also - I can see that's going to be very handy in the future.

Here's a photo I took of the test piece (don't mind the damage to the square - that was due to my over-zealous use of a scraper to get the tape off the front!).




I rushed this project a bit, so it's far from perfect, but, it proved to me that the kerf issue is conquerable/manageable.....thanks to all your help!!

It's real late and I'd better get some shut-eye before tomorrow! I'll add some more Q's later (about beam centering and focus plus axis linearity - yup, I'm going to get all teknikal on ya!). Thanks again everyone!

Cheers,
Dave

[Rodne Gold]

On the puzzles, this is the 3rd job for this customer. We use large format inkjet prints on adhesive, then UV/Waterproof overlam that's very thin. I apply that to the 1/8" plex sheet, transfer tape it, mist the transfer tape with water, and it cuts beautifully thru everything with no discoloration at all. Hand applying each piece (would have to be perfect) is a lot of work.[/quote]

Joe , I must try that , thanks for the tip. Do you use paper based xfer tape?

[Joe Pelonio]

Rodne,

Yes, since I'm a sign shop I use what I use for vinyl, it's rtape #4775.
I tape the prints just before applying to the acrylic, as the adhesive will strengthen with time. When peeling the heat from the laser cuts makes it peel off easily, have never had it pull off image or laminate.

[Lee DeRaud]

<!--StartFragment --> Has any one used the offset or scaling feature of the driver to compensate for the kerf. I have a laserpro so it might not be available to others. I have tried it and had some good results. I use it for inlay but have not used it for side to side fit like marquetry.

Unless you're cutting the pieces one at a time, scaling doesn't help: the scale factor you need depends on the size of the piece. Example: if you have a 1" circle, you can scale it by 101% and get the same effect as a 0.005" contour. But a 1/2" circle will require 102% scaling to fill the same kerf offset.

[Joe Pelonio]

Rodne, here's one of those puzzles:

[Abdul Baseer Hai]

Hi Alan

I have a laserpro explorer 50 watt. you mention the scaling feature.
i dont know how to use it. could you kindly describe the feature in detail.
thanks

abdul

[Alan Young]

Abdul

After Lee's post I am not sure I under stand how it works. I was cutting out some letters for an inlay sign I was making and manged to make it work. It sounds like the contour feature in Corel maybe the best way to go. I have not found any documentation on the scale feature of the driver. I just used trial and error. I would think it would have some application purpose.

Alan