What's new
What's new

CNC Tube Laser Having Issues Keeping Holes On Centerline

controlled_burn

Plastic
Joined
Mar 25, 2019
I'll start off by mentioning that I have a Mazak SpaceGear U44 2.5kw laser. It can do flat, 3D, chuck/tube jobs, and has a 4'x4' bed. Lately, I've been scratching my head as to why it can't keep circular holes on the centerline of a part I'm programming. I've attached some photos of what my setup looks like, but basically, I have a chuck with a 1" square tube loaded in it and up against a live center on the other end (not shown in the simulation screenshot). The part that I'm programming has a bunch of circular holes, on various faces of the tube, but all on the tube centerline. When I'm done cutting (and I cut 8 parts to get a good sample size), I get holes off centerline anywhere from .010" to .080".

Here's what I know so far:

1. The tube is straight. I swept an indicator along the entire length of the tube using the laser nozzle and y-axis, and the tube only runs out by about .005" along the 4' piece that I'm working with, which is pretty impressive. I also rotated the chuck and did the same thing with the same results.

2. The software that I'm using to program is outputting correct g code (verified with a different program's backplot viewer). The machine is being told to put the holes directly on centerline.

3. When the machine is moving/cutting, I'm watching the DRO to see what actual values are showing up. The y work offset is the centerline of the part and y reads 0 whenever the nozzle moves down the tube and is not cutting a hole (y values would obviously not be zero while making a hole in the x/y plane).

4. I don't believe there is anything wrong with the machine with respect to servos, encoders, ball screws, etc. We keep out machine well maintained and have it serviced every 4,000 hours.

Anyway, if anybody could provide some insight or point me in the right direction, I'd appreciate that. Right now I'm just spinning my wheels.

Thanks!

4348_1.jpg

4348_2.jpg

4348_3.jpg
 
After you cut one, did you leave it set up and sweep it again? With it still set up, are the holes in a straight line to eachother?
 
Can you attach some kind of steady to support the tube midway?

I can't imagine the tube NOT moving with that much length unsupported.
 
I did not, but I'll definitely try that and report back. Another thing I forgot to mention was that I checked the 4-jaw universal chuck for runout. I clamped a piece of TGP in there and measured only about .001".
 
Also, which holes are the worst? If they're at their worst in the middle, and better at the ends, then I think you have your answer.
 
Are you positive your Z axis is calibrated properly and never drags? The capacitive sensors on lasers I've worked with in the past tended to drift over the years.
 
I assume that you've already checked that the tailstock is dead in line with the headstock?

Is it always the same holes that are off? Are the holes in one position of the tube always off the worst?

If the machine can repeat itself, then everything else can be corrected.
 
I don't know enough about the machine to have an answer to that, but we do have Mazak techs PM this machine religiously, so I would hope that would be checked during a service.
 
That's entirely possible, but I eliminated that by sweeping the length of the tube a few times throughout the program and I saw very very little distortion (<.005").
 
Yes, chuck and live center have been checked for alignment. The material is straight (enough), the part is not warping during the program, and the holes are randomly off centerline.

At this point, since all holes are programmed to be on centerline (Y 0.), I am literally going to stick a Noga and indicator up on the y-axis to see if the machine is going back to Y 0. for every hole pierce (which I programmed to be all at the center of each hole). If it's not going back to Y 0., the machine has a mechanical problem. I'll report back on what I find.
 
How do you figure it's not moving while you're cutting?

When you stopped, threw an indicator in there and checked alignment the tube went back to position.

I kinda find all this .001" over 4 feet on small square tube a bit nuts to begin with.

I've machined pieces like that in VMC's and built fixtures to weld them. That shit moves like crazy and it's never that straight. There's a ton of stress in square tubing.
 
Because I have seen two holes, a few inches apart, one will be on center and the other will be off by .060". The material has stress, but there's no way all that could happen in 2-3".
 
Because I have seen two holes, a few inches apart, one will be on center and the other will be off by .060". The material has stress, but there's no way all that could happen in 2-3".

Sheet metal on a laser can warp more than that in less distance if conditions aren't ideal. What's interesting to note is that it returns very near to straight once it cools fully.
 
square tube always, and really is always, has a twist in it. When it is on horses/bunks/floor the twist is gravity straightend. Once you have it in a chuck and live center gravity has no datum to straightn to. It does not take much, 1/16 out is very little twist (size dependent). You can tell because top web and bottom web are out about the same in opposite directions. Tube is also not square, just to add in the fun. I find the only way to hold ok center and cuts on square (plasma not laser) is to have stock as short as possible, level the middle across the tube (across flanges, not length) and then press play.
 
have you tried to cut a hole, then move off from it manually then restart the program if the machine is finding the set point again on the hole you just cut?

aka basically how repeatable is the same hole mulitple times?
 








 
Back
Top