C-Axis Lathe problem

[#Rekd™]

Hi,

 

Having tons of quality issues with our DMG Eco lathe...very long story..

 

I have this part shown here and it has been drawn and programed (Thanks to Jeremy at http://www.dbssolutionsllc.com/) in Mastercam then tested on the machine. Everything is okay except when it mills the flats for the hex. In the middle of the hex I am getting a line that runs the length of the hex (like an endmill dwell mark).

 

I have programed the same part wit the on board Siemens Shop Turn software and I am getting the same results even with a .002" finish pass.

 

Is this normal? What could be causing it?

 

The part is Aluminum.

 

Thanks in advance

 

[Brando]

could be an x axis backlash issue, is at the point where the x axis is reversing?

[#Rekd™]

could be an x axis backlash issue, is at the point where the x axis is reversing?

 

Yes it is. Can this be fixed?

 

Machine only has about 10 hours on it.

[Brando]

yes it can be fixed, but is this a new machine?

[Brando]

didint see the "10hrs on it ", theres no way it should have any backlash with 10 hrs on it, i would do a repeatability test on the x axis with a tenth indicator and see if it is infact a backlash issue,

 

bieng a new machine its most likely a metric to english conversion error in the machines parameters

[#Rekd™]

They did a laser test for repeatability and the guy said it was within microns.

[#Rekd™]

Here is a photo of the problem area.

 

[Brando]

id still check it yourself, a measuring tool is only as good as the person using it, also would try milling a triangle shape and a square shape, see if it does the same thing, does your c axis have a brake? i remember seeing a post no too long ago that the brake was on when somebody was trying to cut and rotate simultaneously, also check the machines programming book if you have one, maybe theres a specific cutting mode you have to be in or something, last time a ran a siemens was on an index abc and it had gaga g&m codes for just about anything

[Brando]

so it does that mark on all sides of the hex?, or just where your start stop point is at

[#Rekd™]

All sides.

[Brando]

sounds to me like a parameter is bunk or a mode is not activated

[#Rekd™]

It was doing it the last time the DMG applications guy was here and he had no clue. He thought programing it in Mastercam might help!

[RTcat]

Our Mori Lathes do that too. But ours don't gouge, it simply looks like a witness line. We just attribute it to the pause for rotation as talked about earlier. Use Y-Axis to eliminate it, but if you have no Y-Axis then you're S.O.L.

[Brando]

hockey guy, what unit of feed do you use in programs that use the c axis, ipm or inverse time?

[#Rekd™]

Here is a sample of the Mastercam code.

 

T="END_MILL_.375"

;MILL HEX

G94 S1500 M3

spos[4]=0

SPCON(4)

G0 X1.0922 Z1. M08

TRANSMIT

STOPRE

G1 X1.091 Y-.0513 F500.

Z.1

Z-1.038 F30.

G41 X.9944 Y-.0772 F15.

G3 X.9337 Y-.1238 CR=.1

G1 X.574 Y-.7468

G2 X.3597 Y-.8705 CR=.2475

G1 X-.3597

G2 X-.574 Y-.7468 CR=.2475

G1 X-.9337 Y-.1238

G2 X-.9669 Y0. CR=.2475

X-.9337 Y.1238 CR=.2475

G1 X-.574 Y.7468

G2 X-.3597 Y.8705 CR=.2475

G1 X.3597

G2 X.574 Y.7468 CR=.2475

G1 X.9337 Y.1238

G2 X.9669 Y0. CR=.2475

X.9337 Y-.1238 CR=.2475

G1 X.8837 Y-.2104

G3 X.8703 Y-.2604 CR=.1

X.8737 Y-.2862 CR=.1

G1 G40 X.8996 Y-.3828

X1.0477 Y-.0263

G41 X.9511 Y-.0522

G3 X.8904 Y-.0988 CR=.1

G1 X.5307 Y-.7218

G2 X.3597 Y-.8205 CR=.1975

G1 X-.3597

G2 X-.5307 Y-.7218 CR=.1975

G1 X-.8904 Y-.0988

G2 X-.9169 Y0. CR=.1975

X-.8904 Y.0988 CR=.1975

G1 X-.5307 Y.7218

[jfranya]

We have an old G & L VTL that was retrofitted to use live tooling. This made the chuck like a rotary table. Whenever we would mill with the live tooling, the chuck would "shudder". They looked into it and determined it to be an amplitude setting on the reader head for the chuck/c-axis. Never really fixed it though..."just deal with it."

[RayD]

This is due to the "delay" where your simulated Y axis (X+C) changes directions, If you put your part under an optical you can actually see that a simulated "Y" axis will never cut a flat, it is impossible, it will always be convex or concave, the bigger the flat the bigger the deviation, Hold a straight-edge on the flat and look at a light you will see light shine through.

 

I am Also 1000% sure that if those are wrench flats it will make no difference as the tolerance on a combination wrench is about 0.05 depending on quality of wrench, and if the part has anything to do with "field work or service" chances are the end user will be using a pipe wrench on it

[Newbeeee™]

As the machine is so new, I wonder if the X axis has too much backlash value (compensation) in the parameter? So when it goes over centre, it jumps?

 

BTW - how do you like the shop turn? Good yes?

 

Won't ask about the made in china ironwork...

[Brando]

as ray said the line is def caused by the delay from both axes reversing, on a lathe those times should be very short on the c and x axis, im thinking the simulated axis cycle is adding to that delay more then it should be, or maybe not calculating the feedrate correctly, can you try a x & c gcode program, maybe it will help, i def dont think you should be seeing a line like that, a witness mark yes but that is a little more then a witness mark imo, and on a brand new machine too id be a little upset about that

[Jeremy Herron@DBS Solutions™]

If you look at the misc integers/reals page in the toolpath parameters turn transmit off and repost and see if you still get it.

[#Rekd™]

As the machine is so new, I wonder if the X axis has too much backlash value (compensation) in the parameter? So when it goes over centre, it jumps?

 

BTW - how do you like the shop turn? Good yes?

 

Won't ask about the made in china ironwork...

 

No it is like it is dwelling or over traveling in the X direction. Sure a witness mark would be fine but not something you can not file out.

 

Shop Turn is great and pretty easy to use. I wish Mastercam would take a few ques from it. You can program 90% of your parts on the machine. The only thing I would say against the control is since it is pc based it seems like there are "bugs", more on that later.

 

The quality of the machine is $hit, they have been into our shop 4 times to realign the turret. So upon delivery the live tooling would not engage, turns out the shims were not ground corectly and the spline rotation was off. So after a couple of months DMG agrees to send a factory service guy from China to re-align the turret. Well the local guy did not let the "factory" guy do much. Heck they did not even lift the entire turret off even though the shims were covered in thick crap (should have seen the burrs on the shims). The pryed the turret up one side at a time with screw drivers and pry bars. After they had the shims ground they tried to align the turret with there VDI-30 Test mandrel. Well it turns out the mandrel was bent, so all the machines that they had repaired or aligned were off. The put another mandrel in and told me it was within 10 microns. They had a local guy come in and laser the x and z travels for repeatability. So when the local DMG service engineer leaves he says the next time I see you will be when you crash the machine. At this point we were 9 months since delivery and we still had not produced any parts. I start going through making some simple programs going over my notes etc, they send the applications engineer back in for a day of refresher given the lag since the original training in June. I was getting these lines that we are talking about but he had no answers, but I was also getting some diameter differences. I order my own VDI-30 test mandrel and .00005" indicator and do some checking. The turret was out more then 0.01" (ten thousand), so I call them back up. I had to tell them the the second coolant hose had rotted through again as well tell them the turret was not correct. The service manager said it couldn't be. They send out a different local DMG service egineer. He puts in his test mandrel and says it is out a lot, he then says the bolts that fasten the revolving disc of the turret to the turret are all loose. He then asks me if I loosened them? I am already pi$$ed off so I let into the guy. He re-aligns the turret with his mandrel and leaves without calibrating the tool probe. On his service report he puts the aligment is within 0.01mm (10 microns). The next day I put my mandrel up on the surface plate and indicate it and it is within 4 microns, I then put it back into the machine and check the alignment. It is out 0.0015". I then call the service manager he tells me that their mandrel is "new" and calibrated from Germany, I phyiscally held their mandrel and it was far from new! So they send the service engineer back the next week. He uses my mandrel as he says his is not new and it is indeed out. He aligns it in one pocket then I make him put it into another pocket 180 degrees and it is out 30microns in the X axis. He then goes back and spilts the difference. He still does not calibrate my tool probe. He takes the coolant hose with him and they agree to get a decent hose made. I showed the service engineer some intermitant keyboard/ control problems that the applications guy wittnessed as well but neither had answers. He leaves and I calibrate the tool probe with the supplied calibration tool, I then run the above part again and the diameters are still off about 0.004". When I am measuring the tools after calibrating the probe I hit the open door button and the entire control shuts off as if I had hit the estop. I re-start the machine and then get an main contactor fault message. I call the applications guy and he has no idea other then hit reset to clear the error.

 

As of today we are 10 months after delivery and the machine is still in no condition to produce accurate parts IMHO.

 

There is more but this is a coles notes version.

 

Sorry for any typos!

[Watcher]

I will never buy a DMG machine anymore. Had my time with them.

[#Rekd™]

I will never buy a DMG machine anymore. Had my time with them.

 

You can put me on that list as well! I would take a HASS anyday ove this crap.

[RayD]

I will never buy a DMG machine anymore. Had my time with them.

 

They seem to be hit and miss, I have run many over the years, the ones that run great are bulletproof, hold tenths all day long, others not so much...

 

I know alot of shops that have them and they run them hard all day every day, never any issues... then there are shops where they sit for months on end.

 

Does not matter what make/model of machine tool you have they are only as good as your local Service tech

[Newbeeee™]

Hockey

Have you got the 828 control? This has been initially very buggy.

The 810 is pretty stable. We had a few issues on ours where you have to estop and re-boot, but things like door open shutting the machine down seems like plc to me.

Can you not throw it back to them as it hasn't been fit for purpose? If you have everything documented I wouldn't have thought they could argue much.

If it's on finance, get the finance guy in and tell him to take it away.