Toolpath "Art"

[mrdovey]

I've been writing external functions for various program tasks; and to verify correct operation I cobbled up a PostScript output option that draws the toolpath in red and outlines the bit at the start/end of each tool movement. Some of the output is rather interesting...

This is the result of an external function that "drilled" a 2" diameter hole with a 1/4" bit (starting at the center of the hole) using the smallest number of moves that would meet the precision requirements (0.001"):


28

[mikejohn]

Great Idea.
One question, if the drawing is orientated looking down on the workpiece, isn't the toolpath moving the wrong way for a normal cutting bit?

[stevem]

Mike, the bit is "climbing", which is preferable for a smoother finish and reduced splintering. I assume the bit is still rotating in the correct direction.

[gerald_d]

Morris, with the spiral all the way to the finished size, part of the final circle will effectively be cut as one full pass, while the rest is cut as multiple fractional passes. If one lumps this with bit deflection and flex in the machine, the circle could assume an interesting shape. Any particular reason you prefer spiral to concentric? How many circle segments do you use for the spiral?

[mrdovey]

Mike...

"Wrong" way is pretty much a matter of objective. This works well for me - but I /did/ make a point of retuning my SB carriages after the first use.

It'd be a simple sign change (in two places) in the code to make it cut the other-way-round.

Steve...

Right on... The bit does seem to be still rotating in the correct direction (I don't smell smoke and there's no char in the hole


Gerald...

The toolpath is a spiral until the cut that would exceed the size of the intended hole. At that point, the sequence changes to make the final (circular) cut at the exact size wanted. The program generating the SBP codes uses a maximum error parameter that's used to determine how long each move can be; so the number of moves will depend on %(20) [bit size], the amount of overlap specified, and the finished hole size.

The program does parameter checking and recognizes two special conditions: [1] when (by my rules) it's necessary to cut the hole iteratively; and [2] when the hole is the same size as the bit and can be cut with a simple plunge operation.

Errors (inverted holes, small holes with large bit, etc.) are handled by writing an error mesage and pause command to the generated temporary SBP file.

I haven't experienced bit deflection problems since I instituted built-in rules to calculate (and iterate) deeper holes based on bit diameter. I've taken the attitude that flex in the machine is a quality of design/maintenance issue. I'll have to admit that I had to break a bit before I was motivated to automate iteration, though ):

...Morris

[mikejohn]

Steve, I had never heard of climbing, but found a good explanation here. (http://home.att.net/~waterfront-woods/Articles/Climb-Chip-Cutting.html)
Although I am commited to buying a shopbot, I have no experience with CNC routers. Working manually,I always believed there was only one correct way to move the router (or material)
So, another day not wasted, I've learned something new.
Thanks for pointing me in the right direction.
Is there an answer to Geralds point re- spiral against concentric?
Mike

[gerald_d]

Mike, your manual experience in pushing a router is extremely valuable. We don't all subscribe to the theory that "climb-cutting" is best - in our shop everything is conventional cut, to minimise chipping of corners (http://www.talkshopbot.com/forum/cgi-bin/discus/show.cgi?tpc=1038&post=8750#POST8750). However, we mostly cut single-pass, with the bit cutting a full-width groove - in this case climb vs conventional is not relevant.

Here (http://www.talkshopbot.com/forum/messages/29/1570.html) is another thread on this topic. (Just enter the word "climb" on this Forum's search screen)

[mrdovey]

Mike...

Because I approach the outer edge of the hole on a spiral path, the SB will experience different levels of stresses as it makes the final (circular) pass around the edge.

If there is any slop in the system, then it can show up here in the form of a non-circular hole. I experienced a bit of this (about 0.01" eccentricity) and took time to snug up the eccentric rollers on all three axes - and then made certain the setscrews on the drive gears of my steppers were really tight. The problem went away and hasn't returned.

On the ShopBot, climb cutting isn't a big deal - using a sharp bit and getting the feed and spindle speeds right /is/ a big deal. I'm using a 5hp spindle (with sharp bits) and it doesn't seem to have any difficulty going exactly where it's told.


...Morris