NC Pros and Cons

Subscribed!

I have ZERO NC experience, Siemens controls DMG MORI machine. I am also interested.

This is a very short forum, I've read through it all today.

Is there anyone out there that any recent experiences/issues with NC?

I got some info from Hexagon last week and we did a webex demo today. I invited 2 of our machine programmers, they were excited. It looks like it can help them eliminate some manual work as well. Seems promising. Just to answer some of my original questions.

No you don't have to be a machinist. NC actually turns your CMM code into G-code that gets input into the machining program.
If the machine crashes, recalibrate the probes.
You can have more than one probe in the machine, just like a CMM.
NC can use renishaw probes, or upgrade to hexagons.

You can inspect a feature like a circle and based on the deviation, automatically offset the machine cutting tool for the next part to be closer to nominal.
No manual alignments, use the CNC machine coordinates until you dcc align to the part.
Its works like a server with up to 10 machines on 1 seat. We are probably going to start with 1 and then work our way up. A tech has to install pcdmis on each machine to get the machine parameters and such.
You can use variables, assignments, and print reports just like a CMM.
No annual calibration. all accuracy relies on calibration artifact.

A have a few more technical questions and I'll post as soon as I get them answered.

Good stuff acgarcia, thanks for posting!

NC-DMIS is literally PC-DMIS. If you know how to do something in PC-DMIS, you can do it for NC-DMIS assuming the CNC machine is capable of doing it.

On the CNC machine, if you want to use different "tip angles" on a 5 axis machine, you must rotate the part to the probe, instead of the probe to the part.

For starting alignments, instead of using "STARTUP" you will use "G55" or "G54" (I think that's what it as. This was on Haas VMCs with 5 axis trunions).

Don't listen to Hexagon, you better have your machines calibrated (especially if you are ISO). That's absurd to assert that they don't require it. Especially with 5 axis trunions. I'm struggling to remember who we used as our machine calibration service, I want to say it was "Lasers Inc".

You will definitely want to prove any part program by verifying it on the CMM, for at least as long as it takes to build trust in the results. We were required, by our customer (DoD mainly) to match 1 for 1 NC versus CMM for something like the first 20 pieces, and then critical features only for the next 20 pieces, and then once every whatever number of pieces. And if anything ever mismatched (CNC says good, part not CMMed, failed in the field) then we would have to start over. YMMV.

Also, it's been nearly 6 years since I worked around NC-DMIS, so my recollection may be fuzzy, or no longer relevant.

VinniUSMC , I am concerned with the machines not being calibrated. We are ISO as well and I'm not sure how to handle this. CNC maintenance has always been problem area here. I'm not sure if I go through with NC if I'll be opening a can of worms. I've seen some videos on renishaw's ball bar tech but like I said, machine maintenance has always been a weak point for us. I'll check in with our auditors on this.

VinniUSMC , I am concerned with the machines not being calibrated. We are ISO as well and I'm not sure how to handle this. CNC maintenance has always been problem area here. I'm not sure if I go through with NC if I'll be opening a can of worms. I've seen some videos on renishaw's ball bar tech but like I said, machine maintenance has always been a weak point for us. I'll check in with our auditors on this.

If you are using them as a measurement device and using those results to accept/reject product, it better be calibrated.

If you are using them only as process control, and all of your product goes through a secondary process, like the CMM, then you probably don't. But then you have to make sure on setup to first run the NC program, then verify the product on the CMM to verify your setup.

I don't use NC sensors, but when I see them, I always have a doubt about the accuracy. I feel there are "judge and party" : if the machine has a geometrical defect, the part is out, but the measurement should be OK, because rails are the same...
Just a thought...

A good reason for calibration and for part validation on a CMM.