Apparent TNP with RT FreeCAD (sample file attached)

Discussion about the development of the Assembly workbench.
Forum rules
Be nice to others! Respect the FreeCAD code of conduct!
exsolvespacer
Posts: 169
Joined: Sat Jan 23, 2021 1:15 pm

Apparent TNP with RT FreeCAD (sample file attached)

Post by exsolvespacer »

I am using this branch:

Code: Select all

Version: 2021.1015.24301 +3962 (Git) AppImage
Build type: Release
Branch: LinkStage3
Hash: c6cb39697eb439caf7e5bd0208ca3082a7820547
And if I edit one of the angles in Sketch001 (e.g. change 45 to 44), and close it I get the dreaded TNP in Sketch004 which I thought was fixed in the RT branch. It appears to be because the external geometry goes "wrong" because the line it was based on was moved from 45 to 44 degrees.
problem.png
problem.png (48.25 KiB) Viewed 2994 times

Look at the green line in the bottom left of the screenshot above.

Does anyone know if this is my fault or FreeCAD's? Thanks.
Attachments
Prop glue jig.FCStd
(134.35 KiB) Downloaded 40 times
drmacro
Veteran
Posts: 8865
Joined: Sun Mar 02, 2014 4:35 pm

Re: Apparent TNP with RT FreeCAD (sample file attached)

Post by drmacro »

This is a perfect candidate for quadrant symmetry. 1/4 the work, 1/4 the constraints...

I'm not at a pc so can't look ATM.

Are you actually deleting any geometry in the sketch? Deletes are more likely than changes in dimensions to trigger tnp.

But, if the angle change causes an edge to go away or be added, possibly.
Star Trek II: The Wrath of Khan: Spock: "...His pattern indicates two-dimensional thinking."
exsolvespacer
Posts: 169
Joined: Sat Jan 23, 2021 1:15 pm

Re: Apparent TNP with RT FreeCAD (sample file attached)

Post by exsolvespacer »

drmacro wrote: Wed Dec 08, 2021 9:39 pm This is a perfect candidate for quadrant symmetry. 1/4 the work, 1/4 the constraints...

Are you actually deleting any geometry in the sketch? Deletes are more likely than changes in dimensions to trigger tnp.

But, if the angle change causes an edge to go away or be added, possibly.
No geometry is deleted. Simply a small angle change.

Do you mean quadrant symmetry using the TechDraw Workbench?
TheMarkster
Veteran
Posts: 5505
Joined: Thu Apr 05, 2018 1:53 am

Re: Apparent TNP with RT FreeCAD (sample file attached)

Post by TheMarkster »

If I open the file I get lots of xy_plane for origin001 not found or some such. Since the Part container is superfluous I deleted it and kept the content.

Then I right click on the document and select mark for recompute, and recompute. Then Pad gives an error about a broken face. This is without modifying any angle values. Here is the sketch in question:
Snip macro screenshot-b5ec19.png
Snip macro screenshot-b5ec19.png (84.5 KiB) Viewed 2930 times
It is known that Pad facemaker cheese cannot manage such sketches with nested wires within nested wires.
chrisb
Veteran
Posts: 53922
Joined: Tue Mar 17, 2015 9:14 am

Re: Apparent TNP with RT FreeCAD (sample file attached)

Post by chrisb »

Moved to Assembly forum. This branch is discussed here.
A Sketcher Lecture with in-depth information is available in English, auf Deutsch, en français, en español.
exsolvespacer
Posts: 169
Joined: Sat Jan 23, 2021 1:15 pm

Re: Apparent TNP with RT FreeCAD (sample file attached)

Post by exsolvespacer »

TheMarkster wrote: Wed Dec 08, 2021 10:19 pm If I open the file I get lots of xy_plane for origin001 not found or some such. Since the Part container is superfluous I deleted it and kept the content.

Then I right click on the document and select mark for recompute, and recompute. Then Pad gives an error about a broken face. This is without modifying any angle values. Here is the sketch in question:

Snip macro screenshot-b5ec19.png

It is known that Pad facemaker cheese cannot manage such sketches with nested wires within nested wires.
I don't get any errors like that... are you using the RT branch?
User avatar
jpg87
Posts: 809
Joined: Thu Mar 16, 2017 7:16 am
Location: Limoges - France
Contact:

Re: Apparent TNP with RT FreeCAD (sample file attached)

Post by jpg87 »

exsolvespacer wrote: Wed Dec 08, 2021 9:32 pm ...
Hello,
Although the Realthunder branch is very interesting, there are still times when problems arise when modifying.
So I allow myself to suggest some techniques to often avoid these problems :
- there are generally several ways to build an object, but I think it is always interesting to anticipate the choice of functions to use from a search for the geometric peculiarities of the object to be built ;
- try to simplify sketches;
- use relationships between dimensions when possible ;
- the "Extrude to face" function is very tempting, but extruding a length equal to a dimension that created the face is sometimes safer ;
- creating a reference (datum) plane linked to a face by a dimension used to create the face can be useful for attaching a sketch ...
- use rotational or translational copy functions when possible ;
- Realthunder's symmetry copy is very powerful !

My suggested solution is not perfect, but maybe it can give you some ideas ...
Attachments
Prop glue jpg87.FCStd
(161.39 KiB) Downloaded 37 times
Prop glue.jpg
Prop glue.jpg (60.09 KiB) Viewed 2737 times
My website : http://help-freecad-jpg87.fr updated 2023/11/06
User avatar
onekk
Veteran
Posts: 6144
Joined: Sat Jan 17, 2015 7:48 am
Contact:

Re: Apparent TNP with RT FreeCAD (sample file attached)

Post by onekk »

Many of these problem could be solved using scripting, where all is parametric, as you "hardcode" or ask user for input for parameters and do whatever you want.

Some very complex things are more easy this way, but it is not the "usual workflow".

You drive directly FreeCAD with most of his capabilities exposed, and with some advantage in calculations, as you have not to fiddle with preferences and finding GUI command in a menu or toolbar.

Drawback is that you have to code, so you have to write a code that is not an usual task for a "designer".

This way you left almost completely apart TNP as objects are ever the same despite their "names" as a python object has is "unique" reference, at least during program run.

Another advantage is that not matter how complex is the program, you have ever some thing to start even when version changing or other thing will modify something, maybe you have to change some invocation or change name of some methods, but the "relations" are in the program and remain written and readable even if the code will not produce useful design, all the relevant information are here and are not lost.

As said is a "completely different approach", so maybe it is not the "perfect solution" for every task and for every people.

Regards

Carlo D.
GitHub page: https://github.com/onekk/freecad-doc.
- In deep articles on FreeCAD.
- Learning how to model with scripting.
- Various other stuffs.

Blog: https://okkmkblog.wordpress.com/
exsolvespacer
Posts: 169
Joined: Sat Jan 23, 2021 1:15 pm

Re: Apparent TNP with RT FreeCAD (sample file attached)

Post by exsolvespacer »

jpg87 wrote: Thu Dec 09, 2021 9:37 am
exsolvespacer wrote: Wed Dec 08, 2021 9:32 pm ...
Hello,
Although the Realthunder branch is very interesting, there are still times when problems arise when modifying.
So I allow myself to suggest some techniques to often avoid these problems :
- there are generally several ways to build an object, but I think it is always interesting to anticipate the choice of functions to use from a search for the geometric peculiarities of the object to be built ;
- try to simplify sketches;
- use relationships between dimensions when possible ;
- the "Extrude to face" function is very tempting, but extruding a length equal to a dimension that created the face is sometimes safer ;
- creating a reference (datum) plane linked to a face by a dimension used to create the face can be useful for attaching a sketch ...
- use rotational or translational copy functions when possible ;
- Realthunder's symmetry copy is very powerful !

My suggested solution is not perfect, but maybe it can give you some ideas ...
Thanks for all the tips. If I was designing from scratch, your method is how I would have approached it. It turned out as it is because originally it was not symmetric, but it turned out that way as I iterated the design.
exsolvespacer
Posts: 169
Joined: Sat Jan 23, 2021 1:15 pm

Re: Apparent TNP with RT FreeCAD (sample file attached)

Post by exsolvespacer »

onekk wrote: Thu Dec 09, 2021 3:11 pm Many of these problem could be solved using scripting, where all is parametric, as you "hardcode" or ask user for input for parameters and do whatever you want.
ent approach", so maybe it is not the "perfect solution" for every task and for every people.

I understand OpenSCAD is like that.

The sooner the TNP is fixed the better. Sometimes I wonder how many hours of effort have been lost because of this.
Post Reply