Re: Optimizing cutting paths, was Traveling Salesman Joke

--- In CAD_CAM_EDM_DRO@y..., "IMService" <imserv@v...> wrote:

>
> We make some parts here that are about 2 feet long. While this is

a bit larger than some hobby milling machines can cut, it really
dramatizes the effects of how you make your cutting strategies. (It
is small for a CNC router too.)

>
> These parts have 6 counterbored mounting through holes, a long T-

slot up the middle and they get a 45 degree chamfer on all top
surfaces, and a final flycut surface. The holes are centered(spot
drill), drilled, the counterbores are circular interpolated, the slot
is roughed with one tool T-slotted with another and finished with
another. I first programmed this by selecting the holes from left to
right and generating the G-code. When I watched it cut, it was
really good, while it was cutting. Because I had not paid any
attention to the fact that each pass was with a different tool, each
operation did a rapid for 2 feet before the next tool could start to
cut. Each individual tool's cutting strategy was perfect, but they
were all wrong when observed for all the processes. I finally
reversed the cutting order between the spotting and the drilling, and
the milling, etc, so that each tool alternated between left to right
and then right to left. (Some tools did a complete circuit and
returned to their start point while cutting) Because the part was so
long this reduced the total running time for the part by about 30%,
even though the actual cut time was the same.

>
> I have also found that long runs **generally** are more productive

than short ones. If you have 2 rows of holes, like the pin holes in
a bookcase, it will usually have a shorter cut time if you follow
first one row then the other, rather than trying to zig-zag between
them.

>

Fred.
This is something I have previously stated in earlier posts.
If you spend a little time either before or after generating the tool
paths depending on how your software runs, it can pay dividend on
runs or long pieces.
If it's a one off then just let it run. It isn't worth the hassle of
spending 30 minutes cleaning up and to save ten machining.
I did a large logo a while ago by bringing a DXF into Millwrite and
processing it. Programming took literally 2 minutes but machining
took 50 minutes. If I had edited it I could have got machining down
to about 30 minutes but at the expense of probably 3/4 of an hour
editing
I do a lot of sub contract drilling. The jobs are often large and can
run into the 1,000's off.
Just by altering the sequence of holes and tuning the peck cycles I
can save valuable minutes. Multiplied by 1,000 this is well worth the
effort.
I can do all this dry by editing the operations before I press the
post process button so I know I have the fastest time.