Path drilling operation - feature not recognised as a hole

[nmw01223]

Another beginner question.

I've got a model that is created by designing half of it, then mirroring it to create the other half - it is a symmetrical flat plate, basically. Finally a fusion of the two is created.

It has six holes in it, two of which are on the the centre line. The centre line holes are created as semi-circles, that, once mirrored, come out as holes as expected. The other four are just normal holes.

When I try to create a path drilling operation, it only automatically recognises the four holes that are not on the centre line. The other two it does not see as 'holes' at all. In addition to that, in base location I cannot select the edges or faces of those 'holes' - I have to enter the X/Y locations manually.

That does work, and everything comes out right, but it does seem a non-ideal way of doing it, having to enter the location manually. Is there a way of making it see those mirrored semi-circular pockets as holes? (I suppose I could add them later as holes to the 'fusioned' part, but that doesn't seem ideal either).

[chrisb]

Not sure if it's a workaround or the way to go: did you try to refine the last boolean?

[nmw01223]

Yes, I set refine to true. It certainly looks like a hole.

The only thing I found that worked was to make the original semi-circles in the 'half' part, taking care the diameter was smaller than the hole needed to be, mirror and create a fusion, create a new body and pull the fusion in as a base feature, then pocket the holes to the full size. Then they all show up.

I can see the sense of excluding anything in the base geometry list that isn't obviously a round hole, or all sorts of curves would show up causing a muddle, but in the base location UI, it would be really useful to be able to select any vertex.

[GeneFC]

If you include a small sample file it might be possible to get a better answer.

One item that can occur is that the drilling operation will not recognize any hole that is smaller than the drill size. It is possible that the mirrored shape is a very tiny bit smaller than the ordinary holes.

Simple to check by using a smaller bit size.

Since I operate in a purely manual tool change mode I always drill with a very small bit in FreeCAD and use what I need in the shop.

Gene

[nmw01223]

I have created a small project and uploaded it at https://drive.google.com/file/d/1nprkHa ... sp=sharing.

It's a 10x10mm block with four holes, all 1.6mm diam. It was created as 5x10mm, with the top hole as a real hole, the centre one created from a 180 degree arc, and the bottom a circle pocketed at 1.6mm diameter in a second operation. Then it was mirrored and fusioned (refine=true) to create the 10x10mm block with now four holes, all 1.6mm. Call the top two holes A and B - these are the ones that are real holes, the middle one C - created from two arcs and the bottom one D - pocketed into a semi-circle.

In the path job I am using a 1.5mm endmill - not a drill. Note it is 1.5mm, not 1.6mm to see if what you suggested about slight differences in size could be true. If I add a drilling operation, only A and B are detected automatically, and C and D cannot be selected in base location either.

So I conclude:
1. Making the tool noticeably smaller than the hole does not help in this case.
2. Whether a 'half' hole is created that way, or pocketed, makes no difference.
3. refine=true on the fusion makes no difference.

As before, if I were to create another body, bring in the fusion into that, and then create a real 360 deg. hole that way, it would be fine.

I wonder if this is the sort of thing I should report as a possible bug or feature enhancement to make such holes able to be picked up automatically?

drill_path.FCStd

(36.21 KiB) Downloaded 4 times

[GeneFC]

For convenience of others I have attached the file to your post.

Gene

[GeneFC]

I took a look.

Just select one of the curved arcs that define the top of the hole and the drill operation should work.

HOWEVER, I recall that this capability was added fairly recently. Works fine in the latest development versions (0.22), but may not work in earlier versions.

The drill selection has gone back and forth a bit from extremely liberal (every possible arc in the model) to extremely conservative (pure circles only). The latest incarnation keeps the automation quite conservative but allow the user to add shapes that are not automatically selected.

The current approach is very good, in my opinion.

Gene

[nmw01223]

OK, thanks for that information.

I'm using 0.20.1 I think - downloaded earlier this year, and I see the latest stable release is now 0.21.1. The only thing selectable in 0.20.1 is a full circle hole.

I'll try that. Not over keen on nightly builds, but I'll keep an eye on progress.

Thanks for your input, also for attaching the file to the earlier reply.

[GeneFC]

I'll try that. Not over keen on nightly builds

I would certainly recommend an upgrade to the released version 0.21.1.

The "nightly builds" are not quite the same as they are in some other software projects. They are almost always "stable" in terms of being free of serious bugs. Bugs happen occasionally, but not often. The "instability" comes from changes that literally occur hour by hour. However, there is no need to update on any particular schedule.

If you have the computer space and internet connectivity (500 MB download) I recommend giving the development versions a try. If it works correctly you can keep it as long as you like.

There are many new features and corrections that can be quite useful.

Gene

[nmw01223]

Yes, thanks for that. I did actually use 0.19 when 0.18 was the stable release and it was fine, so there's a precedent. However, from the release notes it seems 0.21 may do what I need for now. I need to take a look.

I'm really waiting for the release to address, at least in part, the topological naming problem. Not sure if that is 0.22 or 1.0 (as originally proposed, I think).