Dogbone dressup problem (again)

Here's the place for discussion related to CAM/CNC and the development of the Path module.
Forum rules
Be nice to others! Respect the FreeCAD code of conduct!
User avatar
sliptonic
Veteran
Posts: 3453
Joined: Tue Oct 25, 2011 10:46 pm
Location: Columbia, Missouri
Contact:

Re: Dogbone dressup problem (again)

Post by sliptonic »

Looking a little closer, it's not just the offsetting. If you set the tool diameter down to even .1 mm, there's still do dogbone.
The real reason is a combination of the offsetting and the curve accuracy.

If you set the tool diameter to 1.5mm, set the path curve accuracy very high and set the GUI to at least 5 decimal places, the bone gets generated. So the issue is that after offsetting the two segments, there's almost no deflection between the curve and the straight piece. It just gets rounded into the G2 move.
2022-01-17_16-58.png
2022-01-17_16-58.png (18.76 KiB) Viewed 1450 times
2022-01-17_16-58.png
2022-01-17_16-58.png (18.76 KiB) Viewed 1450 times
Attachments
2022-01-17_16-57.png
2022-01-17_16-57.png (49.92 KiB) Viewed 1450 times
User avatar
sliptonic
Veteran
Posts: 3453
Joined: Tue Oct 25, 2011 10:46 pm
Location: Columbia, Missouri
Contact:

Re: Dogbone dressup problem (again)

Post by sliptonic »

Just for grins, I started cranking up the accuracy to see if I could force it to generate a bone with a 3mm tool.
I can. You have to set your accuracy to 0.00001000 mm and it will make gcode generation VERY slow.
Giulio Buccini
Posts: 541
Joined: Fri Aug 26, 2016 9:07 pm
Location: Munich - Gemany

Re: Dogbone dressup problem (again)

Post by Giulio Buccini »

But why with the original bracket (the one not mirrored) dogbone-dressup works flawlessy? The length of that short side is the same...
Anyway I will try to make that face longer by modifying the sketch.

NOTE
If you change the direction of the cut (from CW to CCW) also the other two dogbones disappear.
This behaviour is not normal IMO, that sides are rectilinear and very long.
Selection_7705.jpg
Selection_7705.jpg (53.48 KiB) Viewed 1421 times
Giulio Buccini
Posts: 541
Joined: Fri Aug 26, 2016 9:07 pm
Location: Munich - Gemany

Re: Dogbone dressup problem (again)

Post by Giulio Buccini »

sliptonic wrote: Mon Jan 17, 2022 11:08 pm Just for grins, I started cranking up the accuracy to see if I could force it to generate a bone with a 3mm tool.
I can. You have to set your accuracy to 0.00001000 mm and it will make gcode generation VERY slow.
I will repeat me: what surprise me is that the same bracket, with the same sizes, the same geometry, the same tool can be "dogboned" with zero problems.
If I mirror the object into a new one then I have to increase the accuracy to one millionth.
Why?
This makes no sense to me...
herbk
Veteran
Posts: 2655
Joined: Mon Nov 03, 2014 3:45 pm
Location: Windsbach, Bavarya (Germany)

Re: Dogbone dressup problem (again)

Post by herbk »

Hi Giulio,
Giulio Buccini wrote: Tue Jan 18, 2022 12:14 am But why with the original bracket (the one not mirrored) dogbone-dressup works flawlessy? The length of that short side is the same...
in my mind the problem is based at the mirroring tool.
I play'd a bit with your part yesterday evening and get a posible reason: As i put a little box to this corner and take a close look, i can see that this part isn't realy straight.
Attachments
corner.png
corner.png (3 KiB) Viewed 1348 times
corner_01.png
corner_01.png (4.03 KiB) Viewed 1348 times
Gruß Herbert
Giulio Buccini
Posts: 541
Joined: Fri Aug 26, 2016 9:07 pm
Location: Munich - Gemany

Re: Dogbone dressup problem (again)

Post by Giulio Buccini »

I second you Herbk, there is something in the mirror operation that changes the geometry of the object somehow. Maybe the order of the edges/points, or their orientation, or whatever... I have no clue about this.

I tried to use the mirror tool of the Part WB, as well as the mirror tool of the Draft WB. The result is the same: the dressup won't work properly in both cases.


herbk wrote: Tue Jan 18, 2022 7:04 am I play'd a bit with your part yesterday evening and get a posible reason: As i put a little box to this corner and take a close look, i can see that this part isn't realy straight.
That side is rectilinear only in the final part of it. I added a small straight segment of just 2 mm at the end by following the suggestion of sliptonic. This trick has worked well for the original bracket, but it seems to not work for the mirrored part.

Here below the (fully constrained) sketch I use for the original part. As you can see, there is a small segment before the curved edge join the big protruding tab. The angle between the short segment and the side of the big tab is 90°.
Selection_7706.jpg
Selection_7706.jpg (52.6 KiB) Viewed 1300 times
Giulio Buccini
Posts: 541
Joined: Fri Aug 26, 2016 9:07 pm
Location: Munich - Gemany

Re: Dogbone dressup problem (again)

Post by Giulio Buccini »

I tried to create the bracket from the scratch by starting from the a pure 2D geometry:

1. I selected my bracket and generated a planar geometry by applying the Shape-2D-View command in the Draft-WB;
2. I converted the 2D shape to a sketch by using the Draft-to-Sketch command of the Path-WB;
3. I erased four segments in sketch to allow in the two big tabs to allow the step #4 below;
4. I created a new body by using the Part-Design-WB;
5. I moved the sketch inside the body by usinf the model-three;
6. I issued the Pad command of the Part-Design-WB (pad height = 4 mm)
7. I switched to the Path-WB and...
7.1 I created a new job;
7.2 I created a new profile operation for the outline/contour of the bracket
7.3 I applied the dogbone-dressup on the above profile

Result: it still not working.
The dressup operation is ignoring one corner again...

So I think that the mirror-operation is not the cause of the problem, because it persists even if the object is created from the scratch by using a simple 2D polyline made of segments and arcs.

Please check the attached file.
Attachments
dogbone test - by shape2DView.FCStd
(196.93 KiB) Downloaded 19 times
Giulio Buccini
Posts: 541
Joined: Fri Aug 26, 2016 9:07 pm
Location: Munich - Gemany

Re: Dogbone dressup problem (again)

Post by Giulio Buccini »

Again:
there is not a valid reason why changing the mill-direction from CW to CCW even the two correctly done dogbones disappear... :shock:

This leads me to think that the problem is not caused by the liearity and the length of the segments in the corners: if the corners was valid in CW direction then they should stay valid even if I switch to CCW direction... I'm right?

I suspect that there is some bug in the algorithm used for calculation. :?
GeneFC
Veteran
Posts: 5373
Joined: Sat Mar 19, 2016 3:36 pm
Location: Punta Gorda, FL

Re: Dogbone dressup problem (again)

Post by GeneFC »

As Sliptonic pointed out above this seems to be all about precision and tolerances.

The mirror operation and switching from CW to CCW are not precise to 15 decimal places. Nothing in FreeCAD has that sort of ultimate precision.

It may be possible to expand the allowed operation range for the dressups, but it is likely that the current behavior is not actually a "bug".

Gene
Giulio Buccini
Posts: 541
Joined: Fri Aug 26, 2016 9:07 pm
Location: Munich - Gemany

Re: Dogbone dressup problem (again)

Post by Giulio Buccini »

:(
Post Reply