Laser cutting while in motion

I am designing a special purpose machine for manufacturing a single product. One of the steps requires laser cutting a 4" circle from a slow moving ribbon of a cloth material. I have experimented on a standard C02 laser and the material cuts perfectly just laying on the bed. I intend to purchase a small new laser machine just for this purpose. I understand that the shape of the cut path that will need to be sent to the machine will be very odd as it must account for the motion of the material as it passes through. I feel I will be able to overcome this problem. What is giving me a headache is...In order for the cut to be consistent I will need to vary the cutting speed during the cut to continually change as the laser travels upstream->crosstream->downstream-crosstream. I am trying to figure out how to accomplish this.

I am contemplating buying one of the the cheap small format lasers from Amazon or ebay for the gantry and laser, then, if needed, upgrading the motion controller. I really don't want a large format machine with a power bed.

Can this be done without spending a fortune?

Is there a specific type/brand of controller or laser machine that would be needed?

Thanks

This is an interesting question. I would start at 12 o’clock move -X with the fabric moving at a constant speed Then go back to center then go +x and +Y at a faster rate then the fabric is traveling then back to center to finish up the circle.
Laser controllers cut at a constant head speed along a path. Since this is fabric it should cut at a range of speeds. I’d try a slower speed -X then faster speed +X. But your head speed will very.
Constant speed on the fabric shouldn’t be a problem and any controller would drive this cut. I think getting the cut to line up to cut a true circle on something as flexible as fabric will be the issue. All depends on how round the circle needs to be.


Sent from my iPhone using Tapatalk

If you look at it mathematically, I think the path described by the laser would be a rotated ellipse, the parameters of which would change with the speed of the cloth through the machine. An Excel spreadsheet might be able to generate the G code to run the laser.

EDIT: In thinking about this a bit, it would seem to me that given the small mass of the laser and material involved, moving the entire laser machine along with the cloth while cutting the circle, then making a rapid move back to the start position for the next cut, would make sense. This could be done pretty simply, possibly even purely mechanically, with a minimum of controls and hardware. The alternative might be a rather expensive motion controller and some custom software.

I think I have figured out how to do this. Essentially what needs to be done is to negate the relative motion of the ribbon in relation to the laser unit. Essentially the same as moving the machine but doing it with software. If I could independently set the steps per inch depending upon the direction of travel this would effectively cause the position AND speed to be compensated for without adjusting the G code.

For example:
Assume the ribbon is traveling in the X+ direction at 20 mm/sec.
I want the laser to cut at an effective speed of 30 mm/sec. Set the laser to take 30+20=50 per 1mm when moving in the X+ direction (making the laser move farther and faster) and 30-20=10 per 1mm when moving in the X- (moving less and slower).

This effectively subtracts for the motion of the ribbon and adds or subtracts it from the motion of the laser.

The trick to accomplishing this would be to modify the motion controller software to allow it to have different values for each direction. I could buy the cheap laser and then replace motion controller with an Arduino and modify the code.

Sounds like a good plan. It might require an encoder reading the cloth speed to coordinate the timing, but it should work.

The Marlin control software for Arduino supports a M92 code that allows you to set the steps per mm value on the fly from G code. If this was inserted into my code whenever the direction changes it should work.

I think I could get the cheap K40 laser and replace the controller and get what I need.

"Use M92 to set the steps-per-unit for one or more axes. This setting affects how many steps will be done for each unit of movement. Units will be in steps/mm unless inch mode is set with G20 (which requires INCH_MODE_SUPPORT)."

Quick question how cloose to each circle do you need. once the first one is cut the next one is going to be spaced further away as the head will have to go back to the start position or you are going to run out of bed. Could you use a stepped drive to move the material a certain distance then stop until the laser is done cutting

The purpose of cutting while moving is to avoid a start/stop which will create a registration issue. This ribbon of cloth material has something attached to it prior to being lasered and the positioning is somewhat critical. I have a much better chance of success if everything moves in a continuous motion. This is not as much about speed as it is quality. This is a single purpose machine, it will only make one thing in production batches. There are 4 different sizes to contend with.

I have considered the problem of getting back to the start position for the next cut. The feeding of the material will be controlled by steppers. I can move the ribbon through slow enough that the laser finishes cutting soon enough to return to the start position. i.e. if it takes 3 seconds to make the cut I will advance the ribbon so that it takes 4 seconds for it to advance the distance of a single unit giving the head 1 second to return home.

I assume the end goal is to have the laser cut centered on the attachment or at least relative to the attachment position. If registration is that critical, then locating the attachment with a camera might be good solution to locate the attachment edges. I used to do this with wood moving at 600 ft/min, +/- 0.5mm, something moving at the speeds you are talking about should be pretty easy as long as there is reasonable contrast between the attachment and the backing cloth. This would eliminate the need for continuous motion and any minor position errors in the attachment would be automatically corrected.

+-2.0 mm would probably be adequate for me. Right now we are designing the machine and allowing for plans A, B, C...

Plan A. Use stop/start, if not
Plan B. Try to laser on the fly or
Plan C. Use photo or some other kind of locating method to adjust the laser path.

The camera idea was originally my first alternative but doing this is beyond my personal skill set. I am not much of a programmer but I have access to some high school robotics kids that could probably do this in their sleep. I can good contrast by backlighting it.

This machine is for my own use so paying a programmer comes out of my pocket.

I would try Plan A, +/- 2mm is a pretty wide tolerance. Just moving the cloth with a stepper should be able to hold that easily.

You will need a capstan drive as the take up roller will change the placement