I am in the process of starting my first 5axis post in post builder where I can use either the generic or the Fanuc 30i as my base post, is there anything I should look out for with either of these? I have mostly created 3axis posts and am looking for any common issue before I go to far.
Any help greatly received,
Have a look at the sample postprocessors that come with the install in the machine tool library.
They are already equipped with some nice 5-axis functionality to support G43.4 (RTCP) and G68.2/G53.1 (work plane).
Take some time in understanding how its built and you only need to add your specific features to it.I don't know in which version you are working, but try to start with the ones that come with NX9.
When you are working with NX8, you might have an issue with those, and i advuise you to use the NX8 ones. This is because there is new functionality added in NX8.5 and in their posts to retrieve the MSYS data.
Which table/table post supports RTCP? We recently purchaced a OKK trunnion with RTCP and I've learned it's about more than just ouputting a G43.4. We're currently using Postworks and the code seems identical to head/head output. I'd rather use an NX post but when I asked someone at GTAC what it would take they said they'd never heard of RTCP for a trunnion. Any help would be greatly appreciated in that Postworks requires a lot of APT formatting.
We use RTCP and what it does is lets you output all your X,Y,Z code from your machine coordinate system. It's basically posting out everything from your MCS coordinate system regardless of the tool's orientation. It is nice in some ways and annoying in others. The main thing I don't like is the tool axis is not always parrallel to Z so when you setup the post to position in X&Y then approach in Z it is sometimes producing some sketchy moves. Also make sure to note that when you turn on G43.4 (for a Fanuc 31i) the line in which you instate RTPC is not turning it on all the way, it's only turning on the height offset portion of the compensation. We got into issues with this as the machine will "catch up" to where you told it to go after it reads in the 5 axis portion of the comp.
I dont have all the OOTB postprocessors in my head, but for instance the ones coming with sim08 (table table) are a very good starting point.
The functionality per controller might be a bit different, but for Fanuc they support:
It will require a basic knowledge of TCL and the NX MOM environment, to modify the postprocessors to your needs.
Thanks Jake, that sounds like what I'm seeing. I'm old school with trunnions (putting MCS at center of A and C) so it took me a bit to wrap my head around this. I've searched through Postbuilder and haven't found the switch (I'm hoping it's that simple) to turn it on. Just for the hell of it I used a gantry style post and the code is "essentially" identical to what I'm seeing from Postworks.
This is a known issue, and it has quite a simple solution to it, that requires a bit of work in the post processor.
The easiest way to do this, is doing the positoning in Swiveling or WCS rotation mode (Fanuc, G68 or G68.2). Rotate table (if still needed), and after that you position in XY based on the rotated workplane, you can cancel the workplane rotation and then activate RTCP, the next motion is a X,Y and Z motion to the RTCP (MCS) coordinate.
Above is best for Table Table.
Other comfigurations, might need a different approach, but the trick is to let the controller do the math, because the conroller only knows where the piece is at the table.
I hope this helps.
Sorry for hijacking your thread Ray!
We have a head table (B over A which isn't a configuration that's in any of the samples or documentation, which is understandable since you don't see too many machines setup this way) and what I ended up doing was taking the Z position and the table rotation; and also taking the X position, head angle, piviot distance, and the gauge length of the tool in the MCS coordinate system and using all of that to trig out a "fake" RTPC position.
I posted out the position where the tool would be after G43.4 was turned on then made the machine go there, turned on RTCP, and then output the "same" position in MCS coordinates. I did basically the opposite of what you suggest by doing all the calculations myself lol
(SEQ-RUF-SURF-1) X-160.3353 Y-.243 A.235 B1.79 G43.4 H110 X-160.7482 Y-.1782 Z15.7988 <-- this is really just a straight Z-axis approach move.
For a Head Table Machine tool this is a good approach :
In case of a multi axis operation, the preceding steps 7 to 9 are are to be omitted and following line to be added:
7. Move (XY)Z to the same initial TCP position as above.
But depending on your preference that could be a different one.