Hello All,
Doing a finish pass with a 1/16" tapered ball nose at 50 IPM 4% stepover. I am guessing these lines might be related to the machine being dirty but was hoping to confirm from someone more experienced. I have not cleaned/oiled it thoroughly plan to let if finish. then will be looking for instructions or a tutorial shortly on the cleaning / oiling & lower the z 1/16"-1/8 and rerunning…
Thanks!
I can’t quite answer that but it looks like it moved. What it is, could be alot. Maybe your clamping method wasn’t good, maybe the pine shifted… at least it looks like pine. Could be maple though. Could be a problem in your file. I noticed your going left to right, while i haven’t experienced the closed loop benefit, maybe the 4% stepover was not enough to trip it or not enough to register and it didn’t go or it went to far. I usually do 8% on 1/16.
If your going to do 4%, you might as well do 8% on a 1/32. Detail would be better (doesnt look like a great amount of detail is needed).
Also you gotta bump it up… i did 200/200ipm with 8% on my 1/16 on my shapeoko. My elite is not fully mission capable at this time so i haven’t played with it beyond going thru motions or jogging to corners
How long did the rough take you… i dont do 3d roughing passes like that and they take between 20 to i think the longest one was like an hour and a half…
Hi Chris,
The work piece didn’t shift it was actually a bit of a challenge to release after the crave. I am using the blue tape and superglue method which is working well as a hold down. unfortunately it is maple but am hoping to rerun the file later after I resolve. My roughing pass was around 1.5 hrs. I will try higher IPM on finish in the future thanks for confirming that (still a bit of a new to the cnc world).
Further research keeps pointing to needing to further clean. Especially the Z axis when I first got the machine setup my dust collection was far from ideal so would not be surprised if a bunch of dust is in there. Now is much better with a 2 hp dust collector. So will pull apart this weekend and clean every little piece.
Did want to share oiling it and blowing everything off with a ton of compressed air didnt help in fact the post cleaning / oil test is worse than the first. Picture below.
Oh yeah, wow… it did get kinda screwy there on the right.
I would entertain the z level instead of 3d raster. Same file, different logic. 3d raster takes almost twice as long
Usual preface, I’m with PreciseBits so while I try to only post general information take everything I say with the understanding that I have a bias.
I kind of doubt this assuming that this is posted in the right section and I understand the hardware used. Normally when people worry about this it is due to loosing steps. With the Masso and closed loop you shouldn’t be having that issue. Mechanically the other things that come to mind are something is loose and able to move (spindle mount, tool in the collet, etc.) or deflection. From the pictures there’s sections that look like steps where the height change stays at the new level. That wouldn’t be deflection as the load across the piece wouldn’t account for that.
I’ve also seen where software or the controller does this when the plunge speed doesn’t match the feed. Some of our customers that do production work like this not only set the plunge and feed the same for the finishing pass but also have set their accelerations to match. Not sure how much this would help in this case as both of those depend on hardware, CAM, and controller. For the same reason I won’t go more with this right now.
I’m going to preface this next one as I seriously doubt it being the issue. Wood can warp from cutting it. There’s a bunch of listed reasons for it, moisture content, surface stress, nature of wood grain, etc. You might want to check if your piece is warped or cupped though. The reason I doubt this is the issue is you glued the whole board down. There could be some movement in it still as flexible and inconsistent as wood is. I would be surprised if the results looked like this though. Solutions to this vary from skim cutting the surfaces to pocketing the back. I’m dubious of how effective these are.
Two things I would do regardless.
First change your finishing from raster to anything else. Offset or maze are usually much more consistent (and less obvious when they mess up). If you have to use raster try to offset it by 45°. Both of those should result in a better finish even without the issue you are having.
Second, like Chris said… feed faster… a LOT faster. Any time you are running less than 50% of the tool diameter you run into chip thinning. This is basically where you are cutting a smaller chipload (feed) than you would be at 50% or greater. It gets worse the smaller your stepover gets.
To give an example lets say we’re cutting at 100IPM with a 2 flute tool at 10KRPM. Your chipload for that would be 0.005" (Feed / RPM / flutes). What that means is that the thickest part of the chip that each flute is cutting per rotation is 0.005" thick. Now let’s say that you are only taking a 20% stepover. To get that same chip size you need to cut 1.25x the original or 125IPM. At 8% this becomes 1.84x, 184IPM. At 4% it’s 2.55x, 255IPM.
The reason this is important is that cutting too small of a chip will lead to more heat and a poorer finish as there is a minimum size chip needed to actually CUT and not just rub or grind the material out of the way. Assuming the machine has no issue with the feed, the forces for compensating for chip thinning are about the same as the 50% stepover (some changes from chip form). This is because the forces (and your resulting deflection) are determined by the cubic material removed per rev per flute (chipload and pass depth).
Hope that’s useful. I’ll post back if I remember any other things that have caused this type of issue. Otherwise, let me know if there’s something I can expand on or help with.
Thank you! This is an elite machine with the closed loop motors.
I will check the mount and the collet. There has been a ton of fine dust in the collet after carves so maybe this is the culprit. Will also implement the fast feed and different path calculation method.
Josh
problem seems to have been not fully seated bit/end mill. Thank you everyone!
