PC-DMIS vs. NX for programming offline

I saw someone else asking about this recently and that prompted me to post an update with some more info I learned.

I spend over an hour on the phone with an NX-CMM expert from Siemens and asked a lot of specific questions about what it was capable of (and wasn't) and how it works (and doesn't). After this, I decided I would prefer to stick to a PC-DMIS offline seat.

My overall impression was that NX does a lot of things to reduce programming time, but sometimes at the expense of your program's run time. My philosophy is that a little (or a lot) extra time spent programming will be worth it when it results in a program that is easier for the operator and faster to cycle parts through the machine.

The magic button which automatically writes 80% of your inspection program will to the following: create features to probe, insert probe/angle changes, insert clearance moves, insert dimensions with correct tolerances. If you are looking to save on run time by probing a bunch of vector/surface points to later construct features (say, in the case of sever concentric diameters) you will still do this the long way: create points in one area, copy, PWP, construct.

NX does not support variables/logic/flow controls like PC-DMIS can. Some variables can be built into the model (PMI). NX can search for high/low points.

NX exports programs in pure DMIS format, which must then be imported/translated into PC-DMIS. I am told by the expert that this usually functions seamlessly with older versions of PC-DMIS (although it didn't with our machine) and using touch-trigger probing. It becomes error-prone when using analog scanning. In other words, once you import into PC-DMIS, expect to spend time tweaking your program to get it to function perfectly.

One of the things I was most interested in was the possibility of revision control for programs. NX is able to detect when changes have been made to the model, and inspection programs can update automatically. For us, though, it wouldn't work because we save all revisions as separate files in Teamcenter, meaning NX would not detect a change in the model that was used. To achieve rev control we would have to save the DMIS file in TC, and have PC-DMIS access that file from TC every time the program was run.

On top of these things, the licensing costs for NX are way more expensive than PC-DMIS.

A couple of positive things about NX:

• It does have collision avoidance and collision detection, like PC-DMIS, and can fully model and simulate part, machine, and inspection routine.
  
• It has some features which are not supported by PC-DMIS (e.g.: torus)
  
• It can be used to make associative templates (Sales pitch: "Write a program once for one part. Use that same template for a different part of similar geometry, with no additional work!")
  

All said and done, however, I felt that NX was not the way for our shop to go.

I saw someone else asking about this recently and that prompted me to post an update with some more info I learned.

I spend over an hour on the phone with an NX-CMM expert from Siemens and asked a lot of specific questions about what it was capable of (and wasn't) and how it works (and doesn't). After this, I decided I would prefer to stick to a PC-DMIS offline seat.

My overall impression was that NX does a lot of things to reduce programming time, but sometimes at the expense of your program's run time. My philosophy is that a little (or a lot) extra time spent programming will be worth it when it results in a program that is easier for the operator and faster to cycle parts through the machine.

The magic button which automatically writes 80% of your inspection program will to the following: create features to probe, insert probe/angle changes, insert clearance moves, insert dimensions with correct tolerances. If you are looking to save on run time by probing a bunch of vector/surface points to later construct features (say, in the case of sever concentric diameters) you will still do this the long way: create points in one area, copy, PWP, construct.

NX does not support variables/logic/flow controls like PC-DMIS can. Some variables can be built into the model (PMI). NX can search for high/low points.

NX exports programs in pure DMIS format, which must then be imported/translated into PC-DMIS. I am told by the expert that this usually functions seamlessly with older versions of PC-DMIS (although it didn't with our machine) and using touch-trigger probing. It becomes error-prone when using analog scanning. In other words, once you import into PC-DMIS, expect to spend time tweaking your program to get it to function perfectly.

One of the things I was most interested in was the possibility of revision control for programs. NX is able to detect when changes have been made to the model, and inspection programs can update automatically. For us, though, it wouldn't work because we save all revisions as separate files in Teamcenter, meaning NX would not detect a change in the model that was used. To achieve rev control we would have to save the DMIS file in TC, and have PC-DMIS access that file from TC every time the program was run.

On top of these things, the licensing costs for NX are way more expensive than PC-DMIS.

A couple of positive things about NX:

• It does have collision avoidance and collision detection, like PC-DMIS, and can fully model and simulate part, machine, and inspection routine.
  
• It has some features which are not supported by PC-DMIS (e.g.: torus)
  
• It can be used to make associative templates (Sales pitch: "Write a program once for one part. Use that same template for a different part of similar geometry, with no additional work!")
  

All said and done, however, I felt that NX was not the way for our shop to go.