NX4-5 ISV G98/G99 Problem

[Dawson70]

When setting rtrcto Auto/None in a drilling operation I get the same results running machine simulations. Both act as G99 and look to the minimum clearance in the operation which is .100 by default.

My post puts out the G98 which looks to the clearance plane and allows for clearing of clamps and such, but the simulation doesn't match up. I am using a 3-axis mill fitted with a rotary/trunion for the simulations.

I'm assuming the post is working properly, and the kinematics for everything other than drilling also work as expected.

My question is:

Is there something that I may not have set in the machine tool builder application that will fix this? Or, do I need to change something else to get this to work?

Ron Dawson
Manufacturing Engineer
Ducommun AeroStructures-Parsons

http://www.ducommunaero.com/index.html

 

[jelmerra]

What version (NX and PB) are you on and what IS&V method are you using?

The latest nx5 implementation leaves you with calling a bunch of Germans (former conmatix people). The vnc method is very open but requires a bit of validation and additional programming. The special tool drivers for siemens 840D and so on are a call the help desk job.

The machine you refer to might be suitable for the greatest and latest nx5 simulation (very fast).

Jelmer
Rouwe Analysis

http://www.rouwe-analysis.info

 

[Dawson70]

I'm using both NX-5.0.2.2 MP2 and NX4-4.0.4.2 MP4 which are the latest version of each release. The PB I'm using now is 5.2 I think. It's the one that comes with NX5. I'm not sure what you mean by the method I'm using though. I bring in the machine through the tool order view. It's the only way I know how to do it. I've never tryed bringing it in as a regular assembly for fear of not retaining the kinematics.

I ran a test earlier only using 3 axis mode. It was a fairly simple test, and the output .NC code ran as expected on the mill. In this case, a drill path with 2 holes going .20 deep

In the test I set the rtrcto auto which output a G98 making the mill refer to the clearance plane between each hole.
Example:
%
O1000
S3500 M3
G0 G90 G54 X0. Y0.
G43 H1 Z1.<---Tool retracts to Z1. between holes
G81 G98 Z-.2 R.1 F10.<---G98 tells it to look back to previous Z
X10.
G80
M30
%


The regular 3D path replay showed it doing just that. I then set the rtrcto none and a G99 was then output. The regular 3D path replay once again showed it doing what it should.
Example:
%
O1000
S3500 M3
G0 G90 G54 X0. Y0.
G43 H1 Z1.<---Tool does this line once
G81 G99 Z-.2 R.1 F10.<---G99 tells it to look to the R value between holes. This is the min clearance number from the operation dialog.
X10.
G80
M30
%

I then ran the posted code on the mill and got what the regular 3D path displayed. This was also brought into Vericut to rule out any over-sight on my part.

In both cases the IS&V showed the same result. The tool only lifts to the min. clearance (.100 by default) and could not be coaxed into compliance. For some reason it isn't displaying what the 3d replay in the operation is, or what the mill actually does.

If I program each hole seperately I can get it to display what I want, but that takes a lot longer to do only for piece of mind during simulation.

Ron Dawson
Manufacturing Engineer
Ducommun AeroStructures-Parsons

http://www.ducommunaero.com/index.html

 

[jelmerra]

Well it looks like there could be two things going on either the postprocessor ignores G99 and G98 or there is a flaw in the software.

Anyway there is a move cycle procedure in the vnc file you will find that the postprocessor uses a simulation variable for the retract value. Track this simulation variable and find out where it is set. It is likely that there is code for cycles in the preprocess block procedure or in the other commands procedure. Change the simulation variable for the retracts depending upon G99 or G98. That should do the trick.

You could consider using the latest and greatest for NX5 installation. Then there is nothing to worry about coding just use the fanuc control files from one of the simulation examples. In the whats new section you can find how to set it up. The location of the files is somewhat changed in that case. According to the help files it is more intuitive.

The nx5 implementation is supposed to save time validating machine moves against simulation.