Help with bump in contour toolpath

I hope someone can help me understand and solve this issue with this toolpath. This is an outside contour being cut with a 1/8-inch down-cut bit. It always leaves this “bump” right where the toolpath starts and stops. It’s almost like the path starts and stops as soon as the bit reaches the starting point (touches the start point) leaving an amount uncut equal to the 1/4 the radius of the bit (1/64th).

Is this type of bump common and is there a way to solve this issue?


Leave a roughing clearance and take a finishing pass.


It’s quite possible that the bump is the only ‘accurate’ spot on the profile, and the rest of the path is deflecting & digging in, particularly if you are ‘slotting’ (cutting a 1/4" slot with a 1/4" tool).

I would offset my heart shape by 1/2 the tool dia. (0.125"), and edit the ends so the tool starts off the part.

Use a contour toolpath, with “No Offset”. Analyze the toolpath to see where it starts, and which direction it is travelling. In my case, this one started at the lower left point, and travelled around the heart in a CCW direction. I selected the offset vector and “Mirrored” it left to right. Now the path starts at the lower right point & travels CW (Climb cutting).


Tod was right about the bump being the only ‘accurate’ spot. Although I was not taking very much on each pass, it was enough to cause the 1/8 bit to ‘dig’ into the contour causing some inward deflection. When the motion stopped to plunge, the bit would settle into its correct alignment.

I solved the issue by redesigning the geometry to do all of the cutouts with a 1/4-inch bit. Much less deflection and a near-perfect profile on the wood once fully cut out.

Thanks for the feedback Tod.


FWIW: Here’s another approach to potentially eliminate or reduce tool deflection for shapes with curves.

Tool Deflection reduction via Edit Node.pdf (507.2 KB)

That should not be necessary — could you post a test file w/ feeds and speeds where this technique results in more accurate cutting?

10x16 Fly Fisherman new fly Part A - Board Master.c2d (404 KB)
10x16 Fly Fisherman new fly Part B - Inlay Master.c2d (308 KB)

It shouldn’t be necessary, but it sure helps keep the toolpaths crisp and smooth.

I just tried that technique with a circle (4 nodes), copied & unioned (288 nodes).
The circle actually outputs a few more points than the unioned curve. (NC files were 249 lines -vs- 235 lines respectively). Meaning that the tolerance on a toolpath on a curved vector and the tolerance applied when converting it to a polyline are very close.

At the same feeds & speeds, DOC this should have no effect on deflection whatsoever.

To reduce deflection you need to reduce the forces on the tool. Smaller DOC, Slower feed (smaller chip load), or the most efficient way would be to rough then finish.

{edit} same results with Mark’s heart shape. 4 node heart: 4110 points. Many node heart: 3900 points.


This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.