MyTrunnion5 - Fanuc 16iMB - TCP

[Josh Kinney -Elite Engineering]

Does anyone have experience setting the correct parameters on a 16i for TCP on a table/table machine? I'm getting killed here and no help from Fanuc or Kitamura after paying for the option.

 

-Josh

[Guest MTB Technical Services]

Question #1  Why do you want to use TCP on a table-table machine?

 

It's actually more work for a table-table post and as a practical matter creates more headaches related to rotary velocity.

[Josh Kinney -Elite Engineering]

IPM vs Inverse time feed is one. Should I just use inverse and skip the headaches with TCP?

[Guest MTB Technical Services]

IPM vs Inverse time feed is one. Should I just use inverse and skip the headaches with TCP?

 

If that's the only reason then I would argue against it.

It's easier to get the proper Inverse time feedrate for dual rotaries with different max velocities.

You can also switch to Degree Per Minute where it's possible.

RTCP for a table-table rotary makes writing the post that much more difficult.

You also need to increase your look-ahead on the control as you can easily have a situation

where the slowest rotary axis instantly goes to max velocity.

 

Watch some videos of dual-rotaries running RTCP at high-speed.

It will scare the hell out of you because the tool/workpiece relationship is maintained throughout positioning.

 

RTCP was originally designed for head-head machines to account for tool length and maintain the positional relationship to the workpiece.

I always recommend against RTCP for table-table configurations.

[Josh Kinney -Elite Engineering]

That sounds like a good plan. It's turned into a headache for sure. Mazak is easy to use G43.4. Fanuc not so much. It is never fun to buy options for a used machine....

 

Any advise for setting up the proper inverse feed time for my machine? Or should it be good with the standard 5axis post. Excuse my lack of understanding.

 

Thanks for the help.

 

Josh-

[Guest MTB Technical Services]

That sounds like a good plan. It's turned into a headache for sure. Mazak is easy to use G43.4. Fanuc not so much. It is never fun to buy options for a used machine....

 

Any advise for setting up the proper inverse feed time for my machine? Or should it be good with the standard 5axis post. Excuse my lack of understanding.

 

Thanks for the help.

 

Josh-

 

 

You're trying to do it with a very old Fanuc Control. Circa 1998.

It's not a problem on a 31i.

Add to that the post would need to be configured to output table coordinates BEFORE rotation.

You'd actually be outputting the global CL locations.

 

The standard post should work fine.

Just make sure you set the max velocity for the rotary axes in the Machine/Control Def.

I haven't looked at that the X8 configurations but if you can't set separate velocity ranges for each rotary

set the max to the slowest axis.

[Josh Kinney -Elite Engineering]

How do I find max velocity for my machines rotary axes? Should Kitamura be able to supply that?

 

Thanks again,

 

Josh-

[Guest MTB Technical Services]

How do I find max velocity for my machines rotary axes? Should Kitamura be able to supply that?

 

Thanks again,

 

Josh-

 

IKitamura should be able to tell you.

[Colin Gilchrist]

If you are using the Generic Fanuc 5X Mill Post, all the rotary settings are controlled from within the post itself. So just set the max inverse time value in there. I think there is only a single velocity value for inverse time, in the Gen Fan 5X post.

[Josh Kinney -Elite Engineering]

Thanks Colin.

 

This is what I got from Kitamura:

 

rapids

Prm 1420 A=2250 deg/min

Prm 1420 C=6000 deg/min

feedrate

Prm 1422 = 24000 mm/min max feedrate all axis

 

Post has this line:

 

maxfrinv : 99999.99 #Limit for feed inverse time

 

What do i set it to?

 

-Josh

[Josh Kinney -Elite Engineering]

Guys, I seen this on my mazak but can't remember what the fix was. This should be a flat square toolpath but instead it's digging arched trenches. Any ideas?

 

Thanks,

 

-Josh

 

 

[Guest]

Generally I linearize the rapid motion to get things to arrive where you want when you expect it to be there. Dogleg is a REAL byotch on rotary retracts.

 

CAMplete handles this for me when I need it to.

[Josh Kinney -Elite Engineering]

Those are feed moves.

[Colin Gilchrist]

Are you only getting vector positions in the corners of the toolpath? It looks like the machine is trying to move from one corner position, to the next, without vectors in-between.

 

Try setting the "max step" to .02. Also, which toolpath?

[Josh Kinney -Elite Engineering]

Colin,

I think that might be the problem. It has some pretty large rotations to get the motion and I believe that first move I one line. I believe some of the others have multiple moves but maybe not enough. I also seem to recall this being a problem when I first started programming pur variaxis. Chalk it up to rookie 5 axis programming.... I believe this toolpath is a curve 5axis.

 

I'll head into work and try cranking up the resolution. I hope that works.

 

Thanks!

 

Josh-