Tangential shaving clips an already machined face

[japrenticeuk]

In tangential shaving I tend to cheat by using a "false" gear thickness to minimise cutting air during the to and fro X motion.

Even with a width that makes the cutter fully form the tooth for each tangential pass one hits trouble when moving from the upward facing face to the adjacent downward face. The cutter fouls the tip at one edge. This can be seen in



the relevant code snippet is

Code: Select all

G1 Y-3.964  Z14.158  A30.114  F250 
G1 X7.150  F200  
G1 Y-5.255  Z16.574  A35.717  F250 
G1 X-7.150  F200 
 
G1 Y-1.202  Z-0.406  A-1.200
 
G1 X7.150  
G1 Y-1.392  Z-4.742  A-10.061  F250 
G1 X-7.150  F200  
G1 Y-1.704  Z-7.064  A-14.851  F250 
G1 X7.150  F200  
G1 Y-2.223  Z-9.406  A-19.755  F250 
G1 X-7.150  F200  
G1 Y-2.967  Z-11.769  A-24.821  F250 
G1 X7.150  F200  
G1 Y-3.964  Z-14.158  A-30.114  F250 

The line with blanks before it is the culprit. A and Z are going in a coordinated "straight line" with the centre of the cutter well beyond the blank at X=-7.15 but the cutter still clips the just machined tooth face.

With a given diameter of saw the problem can be avoided by stating the face width is bigger than it really is but this cuts air for both ends of every tangential shave  :(

I think the change of code to

Code: Select all

G1 Y-5.255  Z16.574  A35.717  F250 
G1 X-7.150  F200 
 
G1   Y0.0 
G1   Z-0.406  A-1.200
G1   Y-1.202

G1 X7.150  
G1 Y-1.392  Z-4.742  A-10.061  F250 

would always be safe wherever X ends when switching sides and will have negligible effect on overall machining time.

Sadly I can think of no easy way to identify this context for each tooth in a text editor loop so it's a Gearotic bug or an enhancement request!

John Prentice

[ArtF]

John:

    Now this is why I like it when I see a problem from you... so  concise.

I knew slit saws may have an issue, in fact I had suggested when I first added the axis conversion to allow it
that a person should make the setting of width larger as a function of saw radius so that the blade didnt do as we see above.
  Your point is valid and added to the schedule.

Thx
Art

[ArtF]

Note for Archive:

  This suggestion has been implemented for next version. On switch from one flank to next, a pullout to 0 is performed,
then rotation implemented, and then the tool is moved to depth. This is true for tangental shaving runs. Verification
of this aiding in removal of artifacts will be nmoted here, or further bug reports on this should be posted here.

output GCode then in this example should read

G1  Y0.0  //pull to zeroswitch to other flank
G1  Z-0.406  A-1.200 //rotate and position for cut.
G1  Y-1.202 //move to depth of cut..

(This is for a slit saw, but the change affects all tools as the time change is insignificant imo)

Thx
Art