I think I did everything you suggested. It did get rid of the mico-hole in the center of the wheel.
I went back to the basic shape of the slot cutter and did a new array, this time picking the center of the
wheel Revolve and not the slightly off center of the slot cutter. I then did an Extrude on one of the new
cutters, then a Boolean Difference. Unfortunately, it still did not cut holes in the wheel.

I have attached a new 3dm file. I could just use the corrected files you sent me, but I need to
understand my mistakes.

Thanks for you help.

Hi David,
I arrayed the cutter at center by nine and it booleaned ok here.

Select the large wheel, then boolean difference and then select all the cutters and right clcik.

Hi David, one of your cutout pieces seems to have a couple of extraneous little tiny micro curves pieces overlapping on top of the main piece. I found them by selecting all the big pieces and then doing a Select > Invert which then displayed "2 curves selected", and turning on their control points shows that there are 2 micro curves here:



So something is a bit odd with at least one of those cutting pieces - it's got some kind of little overlap or mis-alignment thing in it as well. I'll see if I can find which one is the bad one.

But anyway, that's why Burr's method of picking one of them and redoing the circular array works, because that got rid of the bad cutting piece and replicated a good one into its place instead.

Also one note - the booleans know how to use a planar curve as a cutting object directly, so you don't actually need to do extrusions as a separate step if you want to cut a hole all the way through the object, just pick the curves as the cutting objects directly and the booleans will internally extrude them for you. So that also means that if you do make extrusions you want to make sure to avoid selecting the curves also as cutting objects or else you'll be telling the booleans to try and cut those areas twice.

- Michael

Hi David, yeah only one of those cutting curves seems to be ok - the others appear to have a small area of self-intersection in the curve.

Notice what happens here if I turn on control points and pull them around:



See that one area of the curve where it kind of pulls out differently? That's an area where there is a tiny additional segment in the curve.

If you zoom in here and find these points:



Moving that middle point down shows that it's not just a micro segment, but that it actually kind of zig-zags backwards for a tiny amount before continuing forwards - that's the kind of self-intersection that will confuse things that involve intersections:



It's hard to know how that happened without seeing some of the earlier steps in how it was created - stuff like that can happen when 2 curves are barely grazing each other and are close to being tangent to one another where they are closest to each other but not quite - that kind of "almost tangent but not quite" situation can cause little bits left over when you try to trim stuff like that with each other.

- Michael

You guys are awesome.

I see what you describe about the construction of my cutting object. I rebuilt it and it works perfectly.

Thanks for taking the time to analyse this problem. I learned alot.

Regards.