From: "andy pugh" <[email protected]>
>
> On 4 March 2011 17:29, Igor Chudov <[email protected]> wrote:
>
>> Everyone knows that milling out concave corners with a round end mill 
>> leaves
>> round corners, why warn about it?
>
> <Devils Advocaat>
> Because you are asking for a square corner, and not getting it.
> You should know that you are going to have a rounded corner, and the
> G-code should determine what the corner radius should be, not the
> interpreter.
>
> However, it is so long since I used cutter comp that I have forgotten
> what the actual behaviour is. I can see arguments both ways, but
> anything which stops folk with braindead postprocessors blaming the
> EMC2 devs gets my vote :)
>
>
> -- 
> atp

Perhaps speaking from ignorance, but my belief was that an inside corner 
being rounded to tool radius is not "warned" about, it certainly isn't a 
gouge - i.e. a cut beyond the programmed face - but that the warning relates 
to the case of a sequence of short line segments in a corner; when, with 
limited lookahead, steps would be done which crash the "outgoing" face of 
the corner before the planner even knows there is a corner coming up.

One probably does not hand code this except by accident or in very odd 
geometry but CAM is very prone to do it when approximating curves. In the 
CAM case if one makes the CAM output tool centreline data it has the benefit 
of knowing all about the geometry so will turn early enough to avoid a 
gouge. One can still apply wear compensation - the crude requirement becomes 
that the wear value is small compared with the line segments.

John Prentice



------------------------------------------------------------------------------
What You Don't Know About Data Connectivity CAN Hurt You
This paper provides an overview of data connectivity, details
its effect on application quality, and explores various alternative
solutions. http://p.sf.net/sfu/progress-d2d
_______________________________________________
Emc-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/emc-users

Reply via email to