PDA

View Full Version : Acrylic Drop Problems



Shannon Galway
05-10-2018, 12:40 AM
I am cutting some 3/8" acrylic with beautiful results until the very last second. I have the acrylic standing off the bed by about an inch to prevent flash back on the backside from my honeycomb bed. When the cut finishes, the acrylic drops. The problem is the laser is still on for a split second after it stops moving (charging down I assume), when the part drops it falls out of focus and a large blister is left on the surface of the acrylic that dropped where the out of focus beam has hit the surface. I need to stop this from happening. Surely I'm not the only one that has seen this. Ideas?

Kev Williams
05-10-2018, 3:43 AM
suggestions:

1- mask the acrylic with blue tape or transfer tape. If not the entire piece, just put a 2x2" or so piece or two over the start/end point, so the laser hits the tape instead of the surface of the cut part...

2- pick the best place to start/end the cut, if there's not a node there add one, then break it apart, and move one of the points away from the other, about .3mm/.012" should do-- for good measure, after moving the start/end points apart, add a tangential exit to the end point:
385641
accounting for beam width you'll have a tiny bit of acrylic uncut; the part should still fall out, but slow enough that the laser will be off before it does.

Ian Stewart-Koster
05-10-2018, 5:43 AM
^^^ Yes, what Kev says...
E-cut, the great plug-in for Corel Draw has a node extension module to do exactly that, so th ecuts start and end in the waste, instead of overlapping the perimeter of the piece

alan corder
05-10-2018, 7:54 AM
How about supporting the acrylic under the bit that drops as well, then it can't drop.

Jerome Stanek
05-10-2018, 10:45 AM
Build a pin table I made one using pop rivets and it works pretty good

Shannon Galway
05-10-2018, 11:47 AM
Thanks for the replies. I've seen several pin table ideas but I'm not sure that will solve my problem. My part is shown here:

http://mylocalmaker.com/photo-drop

The OD of the circle is 3" and the wall thickness is 0.75". Printing more than one of these at a time makes it very difficult to line up the pins so the main part doesn't drop or tip.

I found this idea and am going to give it a try this evening, I'll come back with results:

http://www.instructables.com/id/Laser-Cutting-Table-the-Better-Version/

I've tried leadins/leadouts and that doesn't solve it. When the part drops 1" the beam spot is so large that it marks the part on the way out the leadout. I'll try leaving a small tab, but I'd like to try not to have to do any surface finish work on all these parts.

An unrelated question, how on earth do you include pictures in a post on this forum?

Keith Downing
05-10-2018, 3:03 PM
You can always pause the job 2/3 of the way through and add a little bit of masking tape to the part that as already been cut. If you're gentle, it will not disturb the piece but will usually add just enough holding power to keep the piece from moving until the laser is done cutting. Maybe not ideal, but shouldn't be a huge inconvenience either if this is a paid job.

As others have said, the pin table all but solved this problem for us. It does take a little strategy, but all you really need is 2 or 3 pins to hit the back of your piece and they will hold it up.

Jerome Stanek
05-10-2018, 4:04 PM
my pin table has pins 1/2 inch apart each way small stuff doesn't fall down

Kev Williams
05-10-2018, 6:37 PM
I'm curious as to why a 1" drop? You only need to be 1/16" from the table to prevent flashback-

Anyway, I have an idea, you'll need some sacrificial material, such as 1/4" thick plex--

see if this makes sense, this is a support plate/jig you can laser cut to support the substrate and the cut pieces...
385680
--you need the parts not to fall- this jig will hold everything up, nothing will fall...first you cut the 'cavity' pattern out of the plex, then throw out the 'gray' area, which leaves the cavity for the beam path...

this is all the sections together--
385674
--Once you cut the pattern and throw out the cavity pieces you'll be left with undersized replicas of your parts, you don't have to, but if you tape the replica pieces in place you don't have to worry about them moving, just cut the parts, remove them and the scrap substrate, run the next batch...

--of course the size of your support jig is up to you and the substrate sizes you're working with...

I make cardboard patterns like this for certain jobs--

Doug Fisher
05-10-2018, 11:33 PM
>>An unrelated question, how on earth do you include pictures in a post on this forum?<<

When you are creating a post, there are three lines of icons. One is Insert Image.

>>I'm curious as to why a 1" drop?<<

+1

Tony Lenkic
05-11-2018, 10:44 AM
Whet posting click on "Go Advanced" button. You will get icon for attachments.

Keith Outten
05-11-2018, 11:10 AM
The picture below has two Corian vector tables I designed and built, the one on the left is my second model that I modified the supports under the table.. Corian absorbs the energy and prevents flashback, its sacrificial material but it lasts a very long time. One of your friends who has a CNC Router can make one for you.

The original table on the right I have used to cut hundreds of thousands of 5/8" tall blue plastic letters, numbers and icons for ADA signs, hence the blue stain.

Shannon Galway
05-12-2018, 1:57 PM
I'm curious as to why a 1" drop? You only need to be 1/16" from the table to prevent flashback-


Kev,
I am cutting these parts from 3/8" acrylic. The power needed to get through with a nice edge finish is 80w at 3mm/s. Unless I have a solid 1/2" gap between the honeycomb and the material I am getting flashback everywhere it crosses the honeycomb, where the honeycomb support rods are I need 3/4" minimum. Is this out of the ordinary?

Full disclosure, I am quite new at laser cutting, but have extensive experience with CNC routing/milling. That table is sitting right next to this one. I will look into the corian pin table.

For now I have just split the job into two and kept the last line before the drop happens. I put some tape across and finish the cut. Works well.

The wire support idea didn't work. Flash marks anytime the laser crossed a wire. Very hard to get the material level to the laser as well. Took 30mins and $20 at the hardware store, lesson learned.

This just might be the friendliest, most helpful forum I have ever been on! I'm happy the cranky "experts" don't waste their time here. My father always told me, and expert is defined as -- an Ex is a has been, and a spurt is a drip under pressure.

Kev Williams
05-12-2018, 11:42 PM
Ok, I get the 1" drop thing, old goo all over the honeycomb from previous cutting-- just the other day I looked at the wood welding rod holder that's in my garage with new eyes- it dawned on me to try a handful of 3/32 aluminum welding rod to hold up the 1/8" Rowmark I cut off the honeycomb, and as I'm cutting into that old goo with some air above it now, it's like little bombs going off in each hole as the laser passes by! The welding rod works great by the way, but would still flash back onto your clear--

But I do have a simple fix for the 1" drop: I'm not sure how large an area you're cutting at once, but all you need to do is cover the honeycomb with cheap $4 sheets of trophy aluminum- bare side up- and let the laser hit that. If you make a 1/4" thick cutting jig like I described above- or 3/8" thick since that's what you have on hand- you'll get zero flashback, and since the parts won't fall your cuts and parts will be pristine.

You'll need to clean the goo off the aluminum occasionally with some DNA, just wipe & done. How 'occasionally', you'll figure that out as you go.

Covering the honeycomb means smoke extraction may raise some issues, but as long as you don't cover the whole table you should be fine.

Kev Williams
05-13-2018, 12:46 AM
Just made a quickie cutting jig since my goofy drawings might not be so clear ;)
so I drew this up-- the red lines will cut the part,
the blue lines will cut the cavity--
385794
I used a piece of 1/8" Rowmark for the jig, and some 1/8" clear extruded
for the parts. Jig is cutting--
385795

-jig is cut. the thin pieces are going away--
385796

--the thick pieces are going to hold everything up-
I centered them in place by eye, good enough for now-
note the clear plex is in place to be cut...
385797

-and is being cut, note the cut path--
385798

-cut is finished and parts are supported-
385799

-finished parts-
385800

--- in use, taping the center sections to the 'floor' will obviously keep them in place. Also, my cavity width is only 1/8", a wider cavity is preferable since the beam will likely flare wider the farther down it travels past the kerf. but these parts were kinda small so I couldn't get it much wider! :)
And maybe wood instead of acrylic wouldn't be a bad idea. Or get crazy and use aluminum-- There could be some issues with the power you're using to get thru 3/8" material, heat buildup in the cavity may prevent using a solid floor. Taller jig and use the honeycomb...?

Anyway, your tape-before-the-fall is working, so maybe not fixing what ain't broken is the path of least resistance here :D (I'm always for that!)

Shannon Galway
05-13-2018, 1:34 AM
The picture below has two Corian vector tables I designed and built, the one on the left is my second model that I modified the supports under the table.. Corian absorbs the energy and prevents flashback, its sacrificial material but it lasts a very long time. One of your friends who has a CNC Router can make one for you.

The original table on the right I have used to cut hundreds of thousands of 5/8" tall blue plastic letters, numbers and icons for ADA signs, hence the blue stain.

Keith,
How tall are the spikes? Are there holes in the troughs of the spikes for air flow?

Shannon Galway
05-13-2018, 1:36 AM
Anyway, your tape-before-the-fall is working, so maybe not fixing what ain't broken is the path of least resistance here :D (I'm always for that!)

This is where I'm at right now! I have a functioning work flow, until it doesn't work anymore it stays.

Keith Outten
05-14-2018, 8:49 AM
Shannon,

My Corian vector table is machined with a "V" router bit so what you see are thousands of pyramids. There is a path of air movement under the material being cut but its very small, even so the table has worked great for cutting small parts, letters and numbers. You can adjust the depth of the pyramids by using different angle router bits but the number of pyramids will change with the depth. I've never seen any flashing or had melting problems to date with my tables. I don't know of any other design that provides more contact points.