Issues with 4th axis toolpath generation(possible calculation error)

Issues with 4th axis toolpath generation(possible calculation error)

anders
Participant Participant
753 Views
2 Replies
Message 1 of 3

Issues with 4th axis toolpath generation(possible calculation error)

anders
Participant
Participant

I am in the process of getting the 4:th axis on my machine working.(a HMC with rotary table centered at the y axis origin).

 

But I am noticing some very strange behavior when posting code which almost seems like a calculation error of some sort.

 

if machining a line on a cylinder between B0 and B180 Y0 to  (straight line with no lead-ins or tool comp disabled) on a cylinder of diameter 40mm the output I get is over 1000 lines long when the actual output of such a operation should be as simple as:(this is disregarding federate (which my control handles by feeding it the average radius of a move on a cylindrical surface with B5=radius

 

G0 X0 Y0

G0 Z40 B0

G1 Z20 F150

G1 Y20 B180 B5=20 F150

... retract and things here....

 

Not 1k lines for a single open contour with a constant slope.

I have seen others with similar problems and it seems like regardless of post modification(even the dumper post agrees to the output) normally I would simply merge the segments with the same slope in the pp but I would prefer getting it properly fixed(it severely limits 4.th axis use to indexing for a lot of people with slightly older controls)

 

(I notice others are having the same issue).

 

I can post part file and post processor but the behavior is simple enough to observe with the dumper.

 

/Anders

 

 

 

 

 

 

0 Likes
754 Views
2 Replies
Replies (2)
Message 2 of 3

bob.schultz
Alumni
Alumni

The HSM post processors will buffer 4-axis moves when the movement causes only the rotary axis to move, so you will not get 1k of lines output, but it will only buffer the output if only the 4th axis moves, it will not buffer moves when a linear and the rotary axis both move (such as a spiral).  This would take a bit more effort in determining that the axes are actually moving at a constant rate to each other.



Bob Schultz
Sr. Post Processor Developer

0 Likes
Message 3 of 3

anders
Participant
Participant

I see alot of people with the same issue with varying controls please consider implementing it  or adding provisions for disabling the linear5d and making the post post regular moves so it can be handled in the post.

As it sadly makes any old machine useless for full 4 axis machining(my control will do rotary wrapping by itself with a radius value  where you start but as rotary wrap only uses linear moves I can’t program with that function without unwrapping geometry first and that adds all to of a lot of extra work)(it will also do arcs in xyzb but I can’t post that out).

0 Likes