The load may be changing when it is doing the lift and travel above the tab. Are you using a finishing pass? The rest might be leaving too much material because things are flexing. Coming back with a full depth finishing pass might remove more from the rest of the side and it would be more flush.
The bit may not be perpendicular to the XY plane. If the bit is slanted, it would cut a different depth as it raised and lowered
You can try full depth tabs. They won’t work with your flush trim router but anymore. But they wouldn’t leave this gauge.
Yeah, the tab is 6mm. The workpiece is 18mm thick. Maybe this is a good thing to change. I only need ~3mm for the bearing of my flush bit. I was trying to make the tabs as short as possible so I only had a z movement for the last two passes (ex 3mm DOC). Maybe I should do it the other way around 18mm-3mm and leave it as a 15mm tab
Yes, doing a full depth finishing pass. 0.20mm allowance (1/8" endmill)
Hmm, maybe but I would expect my 5% round nose bit stepover to really show that and it is very consistent.
It did it with full depth tabs too. I’ll have to try and find a picture or run a test.
What depth of cut are you using in the picture? I don’t think you would see a difference in XY stepover. But if you took that cut in 5 passes, one of the sides would have steps going down.
Machine/bit flex, dull tool, deep cuts, fast X/Y, slow Z and possibly acceleration. The tool only moving in Z to plunge/retract releases the tension on the side of the tool and allows it to dig into the sidewall. I imagine there is also a divot at the path start/end point. Ramping the plunge helps to minimize the effect. Full tabs typically makes it worse, more pauses and Z only moves.
The picture shows the part function with a 1/8" endmill, 3mm DOC.
The round end nose bit comment was when I was doing the pocket on the inside (not shown). I was thinking if my Z wasn’t perpendicular, I would have higher ridges on one side. Sorry if my comment was confusing
Yup. Tried many things, but stiil had the marks. What i do now is leave less than .5mm on the bottom of piece with no tabs. Easily separate it with a razor knife then lightly sand the whole outer edge for a clean edge.
My two cents from what works for me.
Painters tape and CA glue and don’t use a holding tab at all. Or if I’m doing something bit I’ll shut off dust collection on the last pass. The dust in the grove will hold it from moving. Mostly I screw down where I can to not have to worry about it. I hate dealing with holding tabs. The flush trim bit on the router is a good idea though I’ll have to remember that next time
I do the same, I was sick of having to remove the tab marks and often having to remove quite a bit of the area around them… I have stopped using tabs in most cases, not worth the hassle.
I think this is due to deflection, either tool, belts, or machine in general. When you are cutting at a constant feed rate, the load is constant so the deflection is constant.
When you stop feeding to raise up for the tab, the load changes, and in the dwell time, the mill has more time to bite in (think “finishing pass” with federate of 0)
Try a triangular tab. This way the machine is still feeding in the x/y plane while the z moves occur. Keep the feed rates constant and you should be able to smooth out that area.
At one point I had contemplated making a postprocessor that would detect holding tabs and move the cutter a tiny bit outward to avoid those marks. I don’t know of a solution.
For me the last one is something I’ve often done — when there were already screw holes milled into a piece, I use those to add hold down screws, and CNC the profile with no tabs at all.