USER FORUM
(you are viewing a thread; or go back to list of threads)
Trouble with Subtractive Extrude of Hexagonal Pattern (by Thovthe)
In the attached .slvs I get the red outline even when I extrude the hexagon pattern far past the upper face which I thought was the workaround for this error. Is there a way to achieve the same effect that's robust enough to let me keep NURBS enabled? I'd really appreciate some help on this.
Hexagonal pattern problem solved (by frank)
Hi Thovthe,
I've redrawn your piece to purge all the degrees of freedom (DoF) - to be sure that these does not interfere with NURBS.
I also change the process used to make the object : First using solid piece with the exterior shape, and then ablate the material in two steps (the hexagons first, then the interior of the dish).
Don't mind the construction hexagon (in green). I wanted to use the step and repeat function to avoid the tedious process of drawing all of them without reference.
I've also modified some Solvespace parameters in the configuration panel to be able to avoid red outline :
- chord tolerance : 0.005
- max piecewise linear segments : 32 (24 seems to be OK too)
Others parameters:
I've used dimensions in millimeters
Solvespace version : 3.1~30e42ca8
I've redrawn your piece to purge all the degrees of freedom (DoF) - to be sure that these does not interfere with NURBS.
I also change the process used to make the object : First using solid piece with the exterior shape, and then ablate the material in two steps (the hexagons first, then the interior of the dish).
Don't mind the construction hexagon (in green). I wanted to use the step and repeat function to avoid the tedious process of drawing all of them without reference.
I've also modified some Solvespace parameters in the configuration panel to be able to avoid red outline :
- chord tolerance : 0.005
- max piecewise linear segments : 32 (24 seems to be OK too)
Others parameters:
I've used dimensions in millimeters
Solvespace version : 3.1~30e42ca8
Thanks and a Couple Qs (by Thovthe)
Thank you so much frank!
It's very nice to have a working part. Using the transform tool for more complex arrays like that was a good trick for me to learn.
Is there a concept I could take away to apply in future? Why does this particular way of doing it by making the rough shape, extruding the pattern, then subtracting the inside seem to work when that part of the geometry looks to be exactly the same?
I am also wondering, "What DOF was unconstrained on your computer?", because each step showed 'OK' on my end.
Thanks again.
It's very nice to have a working part. Using the transform tool for more complex arrays like that was a good trick for me to learn.
Is there a concept I could take away to apply in future? Why does this particular way of doing it by making the rough shape, extruding the pattern, then subtracting the inside seem to work when that part of the geometry looks to be exactly the same?
I am also wondering, "What DOF was unconstrained on your computer?", because each step showed 'OK' on my end.
Thanks again.
Some explanations (by frank)
Hi Thovthe. You're welcome.
Solvespace has some discrepancies in his NURBS and MESH components : https://github.com/solvespace/solvespace/issues/738
Then, as a rule of thumb, it's better to extract polygons first and then circles and/or arcs. This is why I preferably choose to use a solid piece to extract from.
But there are also exception ;)
For instance in these drawings :
In the first one (Part_of_Octagonal_Medieval_Tower_v1.slvs) there are naked edges (red outline). Have a look at g008 (NURBS problem). In the property browser window, on the top, choose the tenth icon (show triangle mesh of solid model), and you will see where the MESH subroutine has issues.
In the other one (Part_of_Octagonal_Medieval_Tower_v2.slvs) the are none : Have a look at g008 and g010 (clean 3d rendering). The way you draw the piece and the chord tolerance are important (as always, trial and error ...)
Parameters used in the configuration panel :
- Chord tolerance : 0.01
- Max piecewise linear segments : 24
Parameter in the drawings :
- Dimensions in millimeters
------------------
In overlap.slvs I will show you why there is a problem when there is one (ore more) extrusion(s) using circle and arcs.
Parameters used in the configuration panel :
- Chord tolerance : 0.001
- Max piecewise linear segments : 24
Parameters in the drawing :
- Dimensions in millimeters
- Angle of the line segment relying the center of the drawing to the center of the small circle : 150.0035142494
When you zoom in as much as possible - in g002 for instance - you can see two things :
- A circle is composed of a closed serie of straight lines,
- The big circle and the small one are overlapping because of that. Then, when the MESH/NURBS can't draw the corresponding triangle to match the resulting surface, a red outline appears.
With the angle I chose there are no errors. But if you change the angle of the line in g002, some NURBS errors will pop up.
------------------
The DoF is linked to the construction line (line close to the angle definition in g002. This line is not constrained in length).
Hope that helps.
Solvespace has some discrepancies in his NURBS and MESH components : https://github.com/solvespace/solvespace/issues/738
Then, as a rule of thumb, it's better to extract polygons first and then circles and/or arcs. This is why I preferably choose to use a solid piece to extract from.
But there are also exception ;)
For instance in these drawings :
In the first one (Part_of_Octagonal_Medieval_Tower_v1.slvs) there are naked edges (red outline). Have a look at g008 (NURBS problem). In the property browser window, on the top, choose the tenth icon (show triangle mesh of solid model), and you will see where the MESH subroutine has issues.
In the other one (Part_of_Octagonal_Medieval_Tower_v2.slvs) the are none : Have a look at g008 and g010 (clean 3d rendering). The way you draw the piece and the chord tolerance are important (as always, trial and error ...)
Parameters used in the configuration panel :
- Chord tolerance : 0.01
- Max piecewise linear segments : 24
Parameter in the drawings :
- Dimensions in millimeters
------------------
In overlap.slvs I will show you why there is a problem when there is one (ore more) extrusion(s) using circle and arcs.
Parameters used in the configuration panel :
- Chord tolerance : 0.001
- Max piecewise linear segments : 24
Parameters in the drawing :
- Dimensions in millimeters
- Angle of the line segment relying the center of the drawing to the center of the small circle : 150.0035142494
When you zoom in as much as possible - in g002 for instance - you can see two things :
- A circle is composed of a closed serie of straight lines,
- The big circle and the small one are overlapping because of that. Then, when the MESH/NURBS can't draw the corresponding triangle to match the resulting surface, a red outline appears.
With the angle I chose there are no errors. But if you change the angle of the line in g002, some NURBS errors will pop up.
------------------
The DoF is linked to the construction line (line close to the angle definition in g002. This line is not constrained in length).
Hope that helps.
Post a reply to this comment: