Limited Knuckle Head Head Zig-Zag Swarf?

[crazy^millman]

In all my years of 5 Axis I have always tried to climb cut swarf toolpaths. I have a machine that only has +200/-200 of travel on the C axis of the Head. I keep getting unwinds out of the post which I expect since the nature of the machine make it happen. I got a nice 1st cut getting everything located and sorted the way I need, but when it goes to the next cut here comes the unwind. I then changed the new swarf to do zig-zag and now I get no unwinds and goes all thew way to the floor, but I am getting climb and then conventional cuts on the part. I am going with it., but anyone else ever just do this to get the job done? Do I make 20 individual operations and let the head unwind above the part for each pass or just do it one and call it done? Believe it or not I am doing it all in once since this is a roughing before heat treat operation. Yes goes against my very nature os having it exactly right, but job need to get done and I am behind schedule.

Anyone done this? Pros or cons I am open to all discussion. Thanks for taking the time to read this and reply.

[gcode]

I've always felt this issue could be solved with some post work

If the machine can do the cut zig zagging, it's physically capable of making it  with climb cuts as well,

its just not getting the correct instructions from the posted code.

[mkd]

Corn cobb rougher should leesen the fear of conventional cuts getting pulled into the cut. In my experience they have the least bending forces.

 Like gcode said, if zigzag does it without unwinds, why would old school swarf need unwinds? Been a while since I've battled a knuclehead, so maybe not up to speed.

[gcode]

I think the toolpath makes the repositioning moves in the same direction as the cut, thus forcing a rewind

If it would cut clockwise and reposition ccw, It could make multiple passes without rewinding.

I believe my JOBS Ever7 post from Postability can do this properly, but its been a while since I have used it

 

[TheePres]

If no internal corners, I've never shyed away from conventional cutting. Though try to keep final cut climbing. 

[gcode]

I tried this with my Postability JOBS Ever 7 Post

The machine's C axis is ± 225°

See the attached screen shot... this is  ModuleWorks swarf milling, using a Stock Model

to trim 6 climb passes to the actual stock

The post outputs this toolpath with no rewinds 

kucklehead swarf.pdf

[crazy^millman]

It is full 360 shape not a limited shape. On a limited shape the Postability post has no issues, but on this full shape it does. Mastercam is not kinematic aware and not sure if I went into the machine side of Moduleworks and set that up it would help either. I have wondered if it would help the tool paths if you went through the trouble of defining a machine for the operation. I throw this one back to Mastercam not the post since I have to think the NCI in the linking moves is controlling all of this.

[jlw™]

Ron, with my head/head CNC has told me all retract/rewind is handled in the post.

I wish there was an option to "reverse rewind" where it would lift off the part just enough and unwind in the opposite of the cut.  It would have to be done on the CAM side tho, too complicated for the post.

[crazy^millman]

1 hour ago, jlw™ said:

Ron, with my head/head CNC has told me all retract/rewind is handled in the post.

I wish there was an option to "reverse rewind" where it would lift off the part just enough and unwind in the opposite of the cut.  It would have to be done on the CAM side tho, too complicated for the post.

Yes I could have programmed it if I wanted to in many operations as I have done it many times. This is just one of those times I didn't need to so I made the call to make it easier on myself. I have posted some unwind topics for this very Machine over the years. Just helped the same customer run a 22" long right angle head and was able to do some full 360 degree sweep race track ID o-ring grooves going down in .01 depths of cuts 13 times and finish both walls for a total of 15 passes and not have one retract unwind. I really didn't think we would hit the .002 profile tolerance, but we did. Like this if I really thought I had to do the extra work I would, but since it is a before heat treat operation I didn't. Just be nice to have a way to control this in the operation and not expect a post to do it. Sorry, but I put something like this back on the CAM side of the process not the PP side of it. The PP should take good motion and output that, not have to adjust motion to make it work. 

[jlw™]

I totally agree, the CAM side should put the unwind motion in the NCI so that you have more control.  Put unwind options in the path.

Trust me, Mcam could do way better on table/head and head/head specific paths.