USER FORUM
(you are viewing a thread; or go back to list of threads)
Partial Lathe and Variable (by Nicholas Seward)
I just found this project and it is amazing. Before I dig through the source code to make it do what I want I wanted to make sure I haven't overlooked some features.
1) Can you lathe by any angle less than 360 degrees? I know I can get by with some clever differences of assemblies but that pushes the usability over the edge.
2) Is there a way to avoid using hard coded numbers for dimensions? For good parametric design you need to be able to change a variable in one place and use it everywhere.
This is by far the best FOSS MCAD solution so far. With the above features, I will be able to leave Autodesk Inventor. For me it is all about productivity. I think with practice this will allow me to be around 50% as productive. That is pretty remarkable.
1) Can you lathe by any angle less than 360 degrees? I know I can get by with some clever differences of assemblies but that pushes the usability over the edge.
2) Is there a way to avoid using hard coded numbers for dimensions? For good parametric design you need to be able to change a variable in one place and use it everywhere.
This is by far the best FOSS MCAD solution so far. With the above features, I will be able to leave Autodesk Inventor. For me it is all about productivity. I think with practice this will allow me to be around 50% as productive. That is pretty remarkable.
(no subject) (by Jonathan Westhues)
(1) The lathe operation, as currently implemented, always goes through a full revolution. You could modify the software to change that, but the NURBS Boolean operations are not particularly robust for surfaces other than surfaces of extrusion (plane, cylinder, etc.). So that's a small fraction of the effort required to really work robustly with the range of surfaces that lathe operation would create.
(2) A dimension is always a constant. I've intended the parametric structure of a model to be entirely geometric. It's possible to express an arithmetic expression in terms of most elementary functions and operations (sum, difference, scale by constant, trigonometry, etc.) in geometric terms, in some cases using exotic constraints offered by SolveSpace, like length ratio or arc length equals line length.
I've expected that approach to be generally more intuitive than user-entered arithmetic expressions in terms of solver variables. In some cases, construction geometry may be necessary. Let me know if you're having difficulty expressing a particular relationship in terms of those geometric constraints, and I may be able to suggest a solution.
(2) A dimension is always a constant. I've intended the parametric structure of a model to be entirely geometric. It's possible to express an arithmetic expression in terms of most elementary functions and operations (sum, difference, scale by constant, trigonometry, etc.) in geometric terms, in some cases using exotic constraints offered by SolveSpace, like length ratio or arc length equals line length.
I've expected that approach to be generally more intuitive than user-entered arithmetic expressions in terms of solver variables. In some cases, construction geometry may be necessary. Let me know if you're having difficulty expressing a particular relationship in terms of those geometric constraints, and I may be able to suggest a solution.
(no subject) (by Uli Franke)
First of all I concur in Nicholas' opinion: great project! I find it much more intuitive than FreeCAD for instance and I really like the minimalistic interface. Modular design can be easily done without even noticing it.
I stumbled over this post while searching for lathe less than 360 degree. I would vote for such a feature as well. For instance if you have to simulate sheet metal bending with a distinct radius.
The second point would be really useful too. For instance if you define a global variable for screw hole diameter like "shd=6.1mm" which then could be referenced in any constraint with a syntax similar to "$shd".
Regards
Uli
I stumbled over this post while searching for lathe less than 360 degree. I would vote for such a feature as well. For instance if you have to simulate sheet metal bending with a distinct radius.
The second point would be really useful too. For instance if you define a global variable for screw hole diameter like "shd=6.1mm" which then could be referenced in any constraint with a syntax similar to "$shd".
Regards
Uli
(no subject) (by Jonathan Westhues)
Agreed that those features would be useful, but what kind of sheet metal bends are you drawing? A bend along a straight line should require only cylindrical surfaces, which you can draw as extrusions.
(no subject) (by Uli Franke)
Yeah, I realized that too some time after posting ;)
Perhaps we should open another thread regarding sheet metal bending. Perhaps we could brainstorm how bending could be a packed into a nice function and how an "unbended"/unfolded floor plan could be exported. Perhaps you already thought about something similar or have some ideas how this could be done with the existing functionality.
After 12 hours playing with solvespace I'm already in love with it. It's one of the best programs a came across since a long time. The solver is a bit picky sometimes but I experienced not a single crash or anything. Crazy stable piece of software! And I really like the reduced interface! Can't stop praising...
Regards
Uli
Perhaps we should open another thread regarding sheet metal bending. Perhaps we could brainstorm how bending could be a packed into a nice function and how an "unbended"/unfolded floor plan could be exported. Perhaps you already thought about something similar or have some ideas how this could be done with the existing functionality.
After 12 hours playing with solvespace I'm already in love with it. It's one of the best programs a came across since a long time. The solver is a bit picky sometimes but I experienced not a single crash or anything. Crazy stable piece of software! And I really like the reduced interface! Can't stop praising...
Regards
Uli
Post a reply to this comment: