Programming for 5-axis TCP

[Bob W.]

I know a bit about TCP and what it does but I have heard there is a number of things that need to be done differently from a programming perspective. What makes programming for TCP different that standard programming without TCP? I have heard about the difficulties in getting a good, edit-free post as well. What are the typical program edits that are needed with TCP programs posted from Mastercam?

[Guest MTB Technical Services]

The biggest thing to get used to are the use of 'table' coordinates.

 

Your X,Y,Z values will be the actual CL position in Mastercam.

These are the positions before rotation.

Your post will only need to calculate the actual rotary angles.

TCP on Fanuc will use IPM instead of Inverse Time.

[Mick]

Prior to our machines delivery (Okuma MU500), I took a post that didn't support TCPC, and edited it accordingly. Got the code looking right (well, as right as I figured, as I didn't have a machine to test it on). I also built up a Vericut machine model as well. Everything seemed to be working fine.

 

Took it with me to the Okuma Thinc facility in Charlotte, where they have a MU500 on the floor. Tested the code, and, well... it sucked

 

It turned out there was a lot that had to be changed, and to be honest, the post has some workarounds in it. Whilst the machine is a table/table, I had to configure as a different machine type to get working TCPC code out. On top of that, at a null toolchange in 5 axis operations, results in some weird code (linear moves in all 5 axes) between the operations. I gave up trying to get that fixed, and made some mods to utilise the force toolchange function.

 

The biggest issue I have found, is getting the approach and retract right. That can cause some grief. We're still considering purchasing an aftermarket post though.

[Guest]

CAMplete gives me perfect stuff every time.

[Niezingerly]

CAMplete: Yes...