Recreating welding weave sine

Hello,

I have a Fanuc robot with RJ2 controller, and with HandlingTool. I want to use it for a welding project, and I suppose I can't have the weave program on it, only if I recreate it.

I've bee thinking on the logic for this program, but I'm a beginner in Fanuc programming.

Can somebody recommend a documentation or tutorial for doing complex math with positions in TP program?

I need to make position operations (calculate the distance between 2 points) and to generate points between 2 points. Than generate move commands between these points.

Basically this is the logic how I want to make the weave pattern.

Thank you, I would like to have it.

Please send it to arnold_sz_noel@yahoo.com

Thank you very much

I didn't get yet.

hello,

I'm forwarding with this project.

I think I'm almost done, but I have a problem.

My code is weaving, but not in the right coordinate system, and I really don't know how to solve this.

I attach the code.

I use a TP program and a Karel program

Pre-conditions to run the KAREL code:

Save some data to:

R[1] - we can call it half frequency (the distance between passes trough the X axis of the path)

R[2] - amplitude

PR[2] - starting point

PR[3] - end point (this should be only an X offset in the TOOL coordinate system, for testing, I'm not sure if it works with 2D displacement yet)

I run the KAREL code, and the robot makes a weaving motion, but I want it to make it in TOOL coordinate system, not in WORLD system.

I tried with OFFSET PR[] in the TP program, but no luck.

Any suggestion for optimising things is welcome.

this is the TP program

My code generates coordinates like X: 67 Y:-35 Z; 0 W:0 P: 0 R: 0

So theoretically I just need to move incrementally in TOOL Coordinate System.

I tried to use tool offset PR , but I think I didn't define correctly the offset PR.

See attached picture

For example let's say that the tool is at 200 mm from the faceplate in Z direction.

If I only input 200 in Z, it handles like a WORLD coordinate and gives error that it can not reach the position.

I don;t know how to define PR for tool offset correctly.

OK,

Thanks for clearing me the tool offset register concept

Now it moves in the right plane, but I have trouble with the calculated values.

I think the calculations are giving wrong results, because I calculate with WORLD coordinates. I need to calculate in TOOL coordinates.

But the problem is that when I record a position in PR, it writes WORLD coordinates.

How can I make that PR contain a position relative to TOOL system, not WORLD system?

party !!

I've succeed to re-create the weave algorithm, at least the zig-zag pattern for straight path. I think I won't need the circle pattern, I'm totally happy with the zig-zag pattern for now. Next mission will be to think on how to implement it on arc path.

I created the algorithm to be able to interpolate the pitch angle between the start and end point, to be able to use weaving from an open point to a closed corner point, where end point needs to be at a different orientation.

I can share it if somebody wants it.

In the next step I would like to use it in TP program like as follows:

- User defines first line segment of weld by defining start ana end point

- Start WELD (simple RO[]=TRUE)

- Call WEAVE(P[1], P[2])

- END WELD (RO[]=FALSE

- Define next line segment of weld by defining next start and next end point

- Start WELD (simple RO[]=TRUE)

- Call WEAVE(P[1], P[2])

- END WELD (RO[]=FALSE

- .....

My question is, if I can achieve this kind of TP program.

Can I record just P[] positions, without moving command and without using PR? If I could record P[], it would be saved in TP, and no risk of accidental overwrite PR-s by other programs.

Any other suggestion how to set up TP program will be welcomed.

I suppose I can't create a new motion command type with POSITION, which I can record, to use it like:

1: J P[1] 100% FINE
2: Weave P[2]