Part I've been making forever is now coming out with gouges from the machine - WTF?

I have a VMC that I've owned for 6 years and a part I've been making for years also out of acrylic.

I machine the back of the part, then mount on a fixture and 3D profile the front to something of an egg shape. The last pass is with a 1/4" ball endmill and a 1/16" stepover to give sort of a crosshatch pattern on the front. I do this back-and-forth operation in 2 directions offset 90 degrees to each other leaving the surface of the part looking almost knurled.

I run this job occasionally, and sometimes one of the parts will work loose on the fixture and get a crappy finish - no biggie, I run 20+ at a time and if I lose one it's not a huge problem. I have 10 different programs with different shapes/sizes of these parts, all very similarly made. But lately, on some of the parts, I am getting horrible finishes on 10 or 12 of the parts from the finishing pass.

It looks like 1 or 2 of the cross-hatch finishing moves with the ball endmill will be 1/16" lower than the rest, leaving a groove in the part. And somtimes it looks like the tool only touched the part in the X direction crosshatch cut and didn't even touch it in the Y - so instead of a crosshatch I get just parallel lines. Sometimes half of one part will have the correct crosshatch and the other half will just be grooves in one direction - looks like the ball mill stopped making contact with the part half way through the second crosshatch cutting portion of the program.

The weird thing is that it appears to happen repeatably in the same places on the same parts in the fixture. I can run 10 parts and #3 (for example) will always have maybe a groove that's deeper than the rest always in the same place on the part. I have other parts that work fine and always give a beautiful finish to every part on the fixture (and we're talking the SAME fixture!).

I am trying to figure out what could be going on to cause this. Given it's the same fixture and same feed/speed, I thought it has to be the programs - but those haven't changed at all in years. No settings have been changed on the VMC either, nor have any coordinate systems been changed. I wouldn't think it's anything mechanical on the machine because it happens in the same places on the same parts but other parts in the same vise in the same spot on the machine will come out perfectly fine. I've made sure the vise is tight and secure, changed out the finishing tool, made sure my fixture is good, checked backlash in the Z axis on the machine. All were fine.

Any ideas what else to check or what the hell might be going on here? Tool coming loose in the holder maybe? Drawbar issue? Doesn't seem like those since it's so repeatable.

If your using a cam system try reposting it. Maybe something did happen to the program. Had to reload my plc before and no way I could have edited it but reloading it from a backup sure fixed it.

Checked backlash in the X any Y ???

Jeff

Maybe check machine level. It's been cold here in the northeast, if the machine is near an outside wall the floor may have moved a little.

Thanks for the suggestions.

-I'd hate to re-post the program because I've made little tweaks here and there to improve cycle time - but I am 100% sure nothing has happened to the programs. I even pulled a backup from a couple years ago and did a verify/compare and the programs were identical.

-I've checked backlash in X and Y and they look fine

-I will check machine level but I am not sure how that would affect just some parts and not others? I'm going about 0.040 deep on the final cut to get the crosshatch pattern.


I've watched the machine run and there is nothing obvious to see visually - it moves back and forth over the parts fine and doesn't look like it's out at all in the Z axis... just when I blow the parts off I notice the problem.

I would repost for shits n giggles and give it a temp name and just see if it replicates

also no knowing what control but, I know on my siemens 840c the more programs I get loaded on the the harddrive
would cause strange problems. And the tech said he had a customer that would have issues that the parts catcher
would just drop un-annouced. After deleting some programs problems went away. So something else to look at.

One thing that will have changed is the material is it possible that there are some kind of variation in hardness that could cause this.

I would look at the fixture and how the fixture is mounted to the machine.

always in the same fixture location?

Bad fixture location [comes loose from bad screw or mounting boss]

bad entry move into the part for that fixture location[tool nudges part on the way in knocking it out of position]

wrong tool tool speed or feed from some inadvertent change in the program[say for instance accidentally removing a line that changed spindle speed after the last location]

Same material but different density can be a bugger to track down. Try running at 75% speed and see if there is any difference. Is it the exact same cutter? Rake angle can pull material into the cutter.

Are you running this process dry or wet? Sounds like the tool is occasionally getting a plugged flute.

I'm running wet.

Material is acrylic and it's from the same supplier (Arcylite), but of course it could be bad material... I buy it a full sheet at a time cut into blocks and it takes me a month or two to burn through a block. I don't remember exactly when this started happening but it's likely it's all from the same batch of material.

Fixture location is always the same - it's an aluminum plate that I hold in soft jaws on the vise. I've made sure the vise and jaws are tight. When the parts are on the fixture, I can't move them by hand... it used to be that because there are multiple pieces cut from one solid block... on the 2nd operation when I was cutting them out, some would wind up slightly loose and get a crap finish. So I put heavy duty springs on the back of each screw which solved that problem. I won't rule out fixture issues but it seems unlikely.

But just to be sure, I unbolted the vise and cleaned everything, took the jaws out and cleaned all that. I also adjusted backlash settings again for all 3 axes... previously I just left the settings what they were when measuring, then added in the difference. This time I set the backlash to zero and measured it then input that number... although the numbers weren't much different (tenth or two, maybe less).

I've triple checked the program, I am beyond 100% sure nothing has changed. It's line-for-line the same as it's always been and I've even checked several backups of the program going back months and years - no change (and this problem started happening in the last few weeks).

I just ran OP1 on a few parts, and bolted them up to the fixtures and will run them and see if I've managed to solve the issue

I'm also going to clean out the spindle taper (already did but I'll do it again) and clean off all the holders in case it's a stray chip. Open to any other suggestions too! Thanks.

What type of machine? What year? Does it have resolvers? This totally smells like bad resolver(s) on a Fadal.

Matt

dksoba said:

What type of machine? What year? Does it have resolvers? This totally smells like bad resolver(s) on a Fadal.

Matt

Click to expand...

It's a 97 (I think, maybe 98) Leadwell V40 with a Mistubishi Meldas 50M control.



I ran some more of the parts last night and same issue. I'm going to try changing out the tool holder and tool as well as taking apart the vise and cleaning it all up (haven't rebuilt vise in 8 years). I'm also going to try another part at 50% speed and see if there's any difference. Well report back.

try temporarily moving the part to the left or right then front / back to eliminate a bad spot in one of the ball screws
if the problem follows its probably in the program
if the problem shifts, look at the ball screws.....or ways

SRT Mike said:

It's a 97 (I think, maybe 98) Leadwell V40 with a Mistubishi Meldas 50M control.



I ran some more of the parts last night and same issue. I'm going to try changing out the tool holder and tool as well as taking apart the vise and cleaning it all up (haven't rebuilt vise in 8 years). I'm also going to try another part at 50% speed and see if there's any difference. Well report back.

Click to expand...

Tool holder, collet.......maybe a bad cutter? Spindle bearings???? loose draw stud?

SRT Mike said:

It's a 97 (I think, maybe 98) Leadwell V40 with a Mistubishi Meldas 50M control.



I ran some more of the parts last night and same issue. I'm going to try changing out the tool holder and tool as well as taking apart the vise and cleaning it all up (haven't rebuilt vise in 8 years). I'm also going to try another part at 50% speed and see if there's any difference. Well report back.

Click to expand...

I don't see how the tool holder and vise could cause this to happen sometimes, and not other times. I could see how this happens w/a bad resolver (basically the bearings go bad and get rough spots, but you don't hit those rough spots at the same place all the time because the balls in the bearings move around so much).

What type of servo motors do you have?

Matt

Sounds machine related. Backlash or your servos or steppers are skipping positions.

Contour gouges can be backlash or motion errors. What machine is it?

This is too funny. I just went through this about 30 minutes ago. A single line on the side of a part. The strange things is it was a straight side and it was about in the middle. I have 3 vises so it is the same program in each and it was only on one part. I first thought about something with ballscrew but the other side is parallel and when it comes back the other side there is nothing. I ran several runs and it was there every time. It looked just like the deflection from a chip being recut. I ran for about an hour and it was on everyone. If I lined up the part ends you could see the lines in the side were not exactly in the same spot. I was thinking about a way cover binding or something. Could not figure it out. I was running aluminum in talon grips. I turned the coolant down from a blast to a stream and went away. Only thing I can figure is at that coolant nozzle angles and talon grip pockets it was blasting chips back into the cutter at that point.
True story