5-axis and TCP

[Bob W.]

I just had a Koma rotary table installed on my Makino A51nx and I am being told by Makino that getting the TCP options is not the way to go as it is more difficult to program and it has been problematic with the horizontal machines with rotary tables. I plan to do extensive full 5-axis surface machining with this and I want the very best performance I can get with this arrangement. Will TCP help me get there? What are the advantages/ disadvantages of TCP?

 

Thanks,

Bob

[Guest CNC Apps Guy 1]

TCP is ok, I prefer WSEC (G54.2). They are right, it is problematic o nthat configuration. I'll post more later.

[Guest SAMCRO]

TCP was originally designed for nutating head 5-axis configurations, like aerospace profilers, to eliminate the need

for a CAM system post-processor to account for the gauge length of each and every tool, as well as the pivot distance of the head.

TCP maintains the relationship of the tool tip to the workpiece.

 

 

When you apply TCP to a dual-rotary table configuration, you introduce a whole new set of problems.

[Watcher]

They are right, it is problematic on that configuration. I'll post more later.

 

We look forward to hearing from you. Please don't forget us... :geek: :w00t: :animier:

[Guest CNC Apps Guy 1]

One of the problems with that type of kinematic (Rotary on Rotary) and TCP is axis velocity. You can run a part on a nutating head and speeds and feeds are good, put it on a ROtary on Rotary or even a Trunion Style and the trunion axis will be at max velocity while the rotary axis is barely moving. If you don't have some high end High Speed Options you ARE going to gouge parts. Seen it first-hand. I'd post pictures but I'd do time in Leavenworth.

 

Go with WSEC. Set your work offset as normal, then you put the difference of reality vs. what you programmed for in the WSEC Offsets and voila. WSEC supports FULL 5-Axis. DO NOT go with RTDFO if you plan on doing full 5-Axis. It's not designed for that.

 

ANy other questions?

[Bob W.]

Makes sense. I have the dynamic fixture offset option already and there was another option that was added related to 5-axis but I don't recall what it was (coordinated motion? linear something or other...). What is the WSEC option you mentioned?

[Zoober]

For profilers and gantry nutating heads, it's the cat's xxxx. Seimens does it best with TRAORI.

For any pivot tables, fugetaboutit...

[Guest SAMCRO]

James is spot on.

 

Take a good look at the mass and velocity with which that trunnion is moving at the beginning of that video.

Now consider the even larger mass involved on a dual-rotary horizontal.

Then consider the mass of your work piece.

 

Standard 5-axis with Workpiece Setting Error Compensation (G54.4) is the way to go for such a large dual-rotary setup.

Rotary Table Dynamic Fixture Offset (G54.2) was originally developed to deal with the positioning of the B-Axis on a Horizontal.

 

[Watcher]

One of the problems with that type of kinematic (Rotary on Rotary) and TCP is axis velocity. You can run a part on a nutating head and speeds and feeds are good, put it on a ROtary on Rotary or even a Trunion Style and the trunion axis will be at max velocity while the rotary axis is barely moving. If you don't have some high end High Speed Options you ARE going to gouge parts. Seen it first-hand. I'd post pictures but I'd do time in Leavenworth.

 

Go with WSEC. Set your work offset as normal, then you put the difference of reality vs. what you programmed for in the WSEC Offsets and voila. WSEC supports FULL 5-Axis. DO NOT go with RTDFO if you plan on doing full 5-Axis. It's not designed for that.

 

ANy other questions?

 

 

Thanks James! I was very curious because I didn't know about this. Maybe is because I never worked with a table/table / head/table without some sort of HSM option... good to know... I thought it was some mathematical problem related to TCP, such as singularity. Glad to know it's more about a mechanical/physical limitation....

 

Everyday is a school day in this business...

 

Thanks,

 

Daniel

[?Mark]

"James is spot on"

 

It wouldn't be a first time...

[Guest CNC Apps Guy 1]

You guys are welcome. I'm always glad to be of assistance.

[Guest CNC Apps Guy 1]

I thought it was some mathematical problem related to TCP, such as singularity

Singularity, yeah, I didn't even go there. That can be a problem on every 5-Axis machine, but it rears it's ugly head more on Rotary on Rotary configurations.

 

Word of caution when setting up and programming on a Rotary on Rotary that will save you tons of grief and eliminate singuarity; resist the urge to get everything centerd and symmetrical, tilt that sucker a degree of so also. Yeah it's harder, and you don't necessarrily get automatic planes but you'll virtually eliminate singularity issues.

[Watcher]

Singularity, yeah, I didn't even go there. That can be a problem on every 5-Axis machine, but it rears it's ugly head more on Rotary on Rotary configurations.

 

Word of caution when setting up and programming on a Rotary on Rotary that will save you tons of grief and eliminate singuarity; resist the urge to get everything centerd and symmetrical, tilt that sucker a degree of so also. Yeah it's harder, and you don't necessarrily get automatic planes but you'll virtually eliminate singularity issues.

 

+10000

 

Thanks James!

[Guest SAMCRO]

Singularity, yeah, I didn't even go there. That can be a problem on every 5-Axis machine, but it rears it's ugly head more on Rotary on Rotary configurations.

 

Word of caution when setting up and programming on a Rotary on Rotary that will save you tons of grief and eliminate singuarity; resist the urge to get everything centerd and symmetrical, tilt that sucker a degree of so also. Yeah it's harder, and you don't necessarrily get automatic planes but you'll virtually eliminate singularity issues.

 

You can also "assist" yourself in a post-processor by treating all tool vector components with a Zero value as the lowest possible double precision value - 0.0000000000000001 (1.e-16).

You'll avoid division by zero errors and generally avoid singularity points by passing by them.

 

Here a link to a great paper on avoiding singularity in 5-axis machining by using tool path deformation.

http://www.sciencedi...890695503002876

 

It's $39.95 USD but worth every penny, if you're a math geek like me.

[crazy^millman]

Great Topic guys and perfect information for people getting into 5 axis. Amazing how much people take for granted when it comes to 5 axis programming. It seems like so much is taken for granted, but doing the little things to ensure a good program that runs the way it should on the machine is sometimes no easy task. I am on Vacation in Florida be back end of the month and nice to just sit back and relax.

[Jrygus]

Weird! I just put a koma table on my A81 and Makino is basically insisting that I run TCP. They said that getting the table centered on your B is so much of a pain that you do not want to run DFO. Let TCP do the work for you. I was running DFO until last night they convinced me that it was the wrong thing to do. It is a major pain though trying to re position from one rotation to the next in TCP. Make sure you rotate the C axis when you are at B0. otherwise your tool will head toward your part trying to maintain it's position to the rotary tables. I just think that it is weird that they would tell you that. Are you working with Karl Lippert from Makino?

[Bob W.]

They said that getting the table centered on your B is so much of a pain that you do not want to run DFO.

 

I can't imagine that it would take much more than an hour to get the table tapped into perfect position (within .0001"). I bought a pallet for this as well so once it is set it will never change or need to be reset. I am not working with Karl, my local salesman is handling this which is why I am checking into it on the forum. My main concern is having the options that will help run this at max speed and smoothness.

 

Bob

[Guest CNC Apps Guy 1]

Weird! I just put a koma table on my A81 and Makino is basically insisting that I run TCP. They said that getting the table centered on your B is so much of a pain that you do not want to run DFO. Let TCP do the work for you.

DFO is STRICTLY a 3+2 function. So if you're running full 5 you're going to have issues. I still say WSEC is better than TCP for that configuration and for Trunion style machines. Getting a table centered is not that big of a deal I mean you're only concerend with X and Z lining up with the center of C (assuming your rotary is facing the spindle). You don't even have to worry if it's rotated. You can use a fixture offset for that. Once it's done, unless you crash it you're fine.

 

JM2CFWIW

[Jrygus]

I am continuing with TCP I will let you know if we run into issues. I am going to try and attach a PDF that Makino's 5-axis adcanced technology group sent me on the pros and cons of TCP and DFO. Hope it helps.

1-series 5XR pre-apps presentation 9-21-12.pdf

[Bob W.]

DFO is STRICTLY a 3+2 function. So if you're running full 5 you're going to have issues. I still say WSEC is better than TCP for that configuration and for Trunion style machines. Getting a table centered is not that big of a deal I mean you're only concerend with X and Z lining up with the center of C (assuming your rotary is facing the spindle). You don't even have to worry if it's rotated. You can use a fixture offset for that. Once it's done, unless you crash it you're fine.

 

JM2CFWIW

 

So what exactly is WSEC? Is this a Fanuc option?

[Jeremy Herron@DBS Solutions™]

Workpiece Setup Error Compensation, and It is not just a fanuc option. I know Mazak and others have copied it. It is similar to Dynamic fixture offset, but it also gives you the ability to not only have your program zero in a different location than your center of rotation but also have your part skewed about the X,Y, or Z axis and the machine will track correctly during full 5-axis functions. A lot of functions are avialable and several in WSEC that are not in DFO like Highspeed machining funtions and many others. I am sure Makino has something similar, and once you use it you will not look back.

[Guest CNC Apps Guy 1]

In addition to what Jeremy said, you can probe and write to your WSEC offsets so for example you were machining a casting on your 5-Axis machine, you can throw your part up there and probe for skew, "fix" it, probe for location, "fix" it and you're off to the races. Works quite well.

[Mic]

TCP is ok, I prefer WSEC (G54.2). They are right, it is problematic o nthat configuration. I'll post more later.

 

I thought the DFO was G54.2 and WSEC G54.4?

 

By the way, I've used TCPC for table/table machines for years now and can't see any problem using it over WSEC. And with TCPC you don't need to run inversed time feedrate or leveling of the feedrate.

[Guest CNC Apps Guy 1]

I thought the DFO was G54.2 and WSEC G54.4?

Yes, you are correct. Typo. Sorry.

[Reko]

I have a B/C-axis, nutating head on a Gantry mill.

 

From what I have read here, TCPC is the one I need to use.

 

Should I add the WSEC option? Can or should they be used together?