Yet another boolean question

How do I do a “proper” and accurate boolean union of all the parts in the model in this file?

.Boolean question 01.3dm (3.2 MB)

I can easily do a boolean union on all the parts highlighted in green. That is not a problem. When I try to do the same thing with all the parts below that I either get the error message “Boolean union failed”, or I get this error message. I still can’t seem to get away from this dreaded window.

I was able to get everything more or less patched together, but the result is not very elegant. To do this I moved various objects by .01 mm in different horizontal and vertical directions. In the end I have a printable model, because the .01 mm shift here and there is not going to make a difference. However, the result is not very elegant. I wonder if there is a better way to do this.

You can see the overlaps in several areas in this completed model.
Boolean question 02.3dm (4.9 MB)

I enclose a video to explain my approach to run the Boolean. The curve of the circle was divided with the command subdivide (option point)

Bolean question.mp4 (11.9 MB)

Hi Zsimon,

Thank you for that video. I cannot find a command “subdivide”, so I also do not know what to do with “option point”

Also, the way you run the sweep does not work for me. All the pipes in the green section are 1.0 mm in diameter. The ‘half circles’ of the shank of the ring (bottom sections) are 1.3 mm in diameter. The 12 connecting pipes I run between the bottom sections and the small rings taper from 1.3 mm to 1.0 mm on the outside of the small rings.

I do not understand why you would use the “Sweep 1” command, instead of the “Pipe” command, which is what I used.

By dividing the small ring the way you did, the tapering takes place very differently for these than for all the other small rings which are consistent in diameter all over. I think these images explain what I am trying to say.

This is what it ends up looking like on the inside:

This works, because all the diameters are still 1.0 mm. The intersection inside the top red circle is a bit off, but it is still close enough to 1.0 m for this to work.

But this is what it ends up looking like on the outside:

This does not work for me.