Problems with area clearance inaccuracy
I am having some issues with some very basic operations that I am struggling to resolve. Let's say I add a number to a square block of wood. I want to clear the area around the number so that the number is raised. I use an Area Clearance toolpath to remove the material around the number. In the particular use case, I am using two passes with a .25 End Mill to reach a final depth of .25" and then a second two passes with a 1/16" End Mill to carve in the fine details.
The raster steps with the 1/4" End Mill complete as expected, but when the machine moves on to complete the profile passes, I am getting tiny bumps along some surfaces as the final profile pass is not going as deep as the original raster passes.
Currently, I am treating this as a machine calibration/repeatability issue, but to be honest, I am not sure whether this is caused by the machine itself or Carveco Maker. Why is Carveco Maker going so close to the letter on the raster passes that the profile pass is really not removing any material and leaving small indentations along the profile of the number?
I have tried playing with tolerance/allowance settings, but since these settings affect both the raster pass and the profile pass, it makes no difference. I would suggest that the internal tolerance for the raster passes is increased, to leave more material for the profile pass, or to add a "Raster Tolerance" parameter specific to the calculation of the raster passes to basically allow me to leave a little extra material for the profile pass to remove so that I can achieve a clean profile.
Any insight that you could provide into this issue would be appreciated.
Comments
Hi Leon Palnau,
Please confirm the:
In addition, please save your Carveco Model file (*.art), use the free WeTransfer service to upload your Carveco Model file (*.art) and generate a link, and then share that link when posting again within this topic. That way, the community can download your file and review the toolpath settings saved within it.
Thanks for your time and cooperation.
SainSmart Prover XL 4030,
"Sainsmart 4030 PROver (inch) (*.tap)"
https://we.tl/t-LI6wRHWn76
Hi Leon Palnau,
We've reviewed your shared sign.art file, and your toolpath settings within Carveco Maker seem to be reasonable. Therefore, your machining results could be attributable to either or both of the following reasons:
I've already went over the mechanicals several times without improvement. I've tested all the mechanical coupler connection by blocking the spindle and trying to rotate the stepper handwheel and all is tight. I took steps to lessen/eliminate spindle vibration quite some time ago.
I still get this result, which I think may be related to vibrations in the bit causing the actual cut toolpath diameter to be slightly wider than expected during the raster passes.
You may be onto something in that reducing the stepdown may help to alleviate this, but honestly I should be able to cut the full .25 depth in a single pass with the 1/4" end mill.
This leads me back to the original question of why Carveco Maker is calculating a 0 tolerance raster tool path, relative to the profile tool path??
By this I mean, why does the outside radius of the raster toolpath touch the effective vector outline, when it is always followed by a profile pass?
This could easily be solved in software, versus having to ensure that the machine has to have 100% repeatability to get a good result.
My next steps will be to perform a careful calibration of the stepper Steps/mm to see if perfection in machine calibration can resolve the issue, along with some repeatability tests I am devising.
I would ask that you consider what I have proposed in regards to raster tolerance though, as there is no practical reason the raster has to touch the vector outline when it will always be followed by a profiling pass.
And by the way, the picture you shared is not a prover 4030 XL.
My machine is also using the Carbide 3d router and sweepy V2 dust collection.
I learned very quickly that the stock "spindle" does not have enough torque to push a 1/4" bit....
Hi Leon Palnau,
Please confirm whether you've already tried specifying an Allowance of zero along with a Final Tool Allowance equal to half of your chosen tool's radius when calculating your Area Clear toolpath.
Thanks for your time and consideration.
Now that I am able to get back to this, I have tested the settings you mentioned above, in the simulation and I think that they may produce better quality as it is using the 1.5 endmill to perform the final profiling.
Do not mistake me though. This is a workaround, not a solution. There is no really valid reason why the 1/4 endmill cannot create a clean profile alone. After performing some additional calibration and repeatability tests, I have concluded that there is no issue with the machine and this is a bug in Carveco Maker likely related to having a two-step area clearance toolpath. I say "likely" because I did not experience this issue in the past using single bit clearance operations. It seems like Carveco Maker is expecting some value in "Final Tool Allowance", and when set to 0 does not create a proper profile with the larger bit.
As a test of this theory, I will remove the second bit operation and create a .gcode that only performs the first 1/4" roughing operation and check whether the profile is smooth.
Hi Leon Palnau,
You were previously asked to confirm whether you've already tried specifying an Allowance of zero along with a Final Tool Allowance equal to half of your chosen tool's radius when calculating your Area Clear toolpath.
Thanks again for your time and consideration.
isn't that what I just did?
Hi Leon Palnau,
Unfortunately, it's not clear whether you're confirming that you've tried:
Thanks again for your time and consideration.
"Now that I am able to get back to this, I have tested the settings you mentioned above, in the simulation and I think that they may produce better quality as it is using the 1.5 endmill to perform the final profiling."
As I said, I will not be reducing the feed rate and stepdown anymore as it already takes longer than it should to carve. There is no reason why I should have to continue to slow down the machining process to account for a software issue.
Leon, how long should it take to carve? I think support were suggesting to slow it down to rule out any issues with the software. The simulation looks fine so normally in that case the problem lies on the machining side of things.
Please sign in to leave a comment.