Wrong size/ dimension of hole

[lugi]

Hi Forum,

I've got a very strange kind of problem. A programmed hole with diam 8mm is always milled with +/- 7.3mm.

Machine Stepcraft D2 600
UCCNC
G-code with Fusion 360 and CAM of UCCNC.

I've checked the Fusion G-code with NC Corrector and the dimensions of the toolpath are OK, as well as the G-code generated with the UCCNC CAM.

Today I've corrected the Backlash and the steps per unit in UCCNC.

Unfortunately the hole is still milled with 7.3mm.

Has anybody got an idea of what can be checked, wrong or whatever to solve this problem?

Thanks a lot!

[ger21]

Try cutting the hole in the opposite direction, and see what size it is? Most likely issue is deflection. I've always found small holes to be the most difficult thing to get accurate sizes.

[lugi]

Hi ger21,

Thanks a lot for your answer.

Since your comment I spent a few hours testing futher.

For the moment I have the following conclusion.

In general, my backlash on X and Y is 0.008 and 0.006, so very minimal and after various tests, it seems to me that I've got more consistant results without correcting it.

Concerning the 8mm hole.
During my tests before my initial post, I varied speed between 1000 and 100, always with the same result, but mainly in plywood.
For more precise analysis I changed to Plexi.
As it's programmed as a pocket in Fusion 360, I can choose between lateral correction left or right, which finally decides if the tool follows the path clock- or counterclockwise - left = ccw, right = cw.

I now manage the have an 8mm hole , but only with a speedrate of 100. At higher speed the hole is milled conically towards the bottom. I use a 3mm cutter and I guess the problem as well comes from the toolpath created in Fusion. The main helix is laterally displaced from the center of the hole and the final dimension is done at full depth of the hole - guess there's deflection happening.

Another point might be my use of the caliper - I don't force it usually, but let it find it's "resting" position when measuring.
As an example, I measure a hole to be 7.8mm, but a 7.95 miller shaft fits. Only by applying force on the caliper, I reach the 7.95 inside the hole.

As a conclusion for the hole - for me best setting is speed 100, milled CCW (lateral correction left).
Milling it CW makes the hole too big.

On the other hand, I had problems with a 47 x 47mm square.
As this was programmed as a contour in Fusion 360, I can choose between climbmilling (CW) or conventional milling (CCW).
First I always used climbmilling as it was described to be advantageous for ball screws.
But after my tests, I see better results by setting it to conventional milling. With climbmilling the square gets too big.

That's it for the moment.

Thanks for your help!

[ger21]

I forgot to mention that speed plays a large role with small holes.

Also, conventional cutting will almost always give a better cut quality with woods.

[lugi]

Hi Gerry,

thanks for your confirmation - I'm now back working with plywood and there's really a huge difference between the two directions.

Best
Lukas