USER FORUM
(you are viewing a thread; or go back to list of threads)
tapered extrusion? (by James )
Hi, is it possible to do a tapered extrusion in solve space?
(no subject) (by Andrew)
Quick answer, no. What you can do is use angled difference extrudes to carve the shape you want. To get the angles you want may require sketching in 3d to create a pair of lines to define the new sketch plane. You can dimension from the plane and normals of a blnk drawing.
(no subject) (by Paul)
No. But after version 3 I have some construction tools I'd like to add. Not sure which will make it. I'm curious what kind of taper you'd like.
One option might be a draft angle applied to an extrusion. This would offset the lines and curves on the top of an extrusion by a fixed amount.
Another that is more likely will be "frustum" extrusion, where the top is uniformly scaled by some amount - which you'd be able to fix using constraints. It should also be able to skew rather than being constrained perpendicular to the sketch plane (you can skew now by selecting a different plane for an extrude).
I'm toying with chamfers on extrusions, but they'd be imperfect in some cases and would be difficult to combine with some of the other extrusion types I haven't even mentioned.
So what type taper were you looking for and which of these possibilities might fit the bill?
One option might be a draft angle applied to an extrusion. This would offset the lines and curves on the top of an extrusion by a fixed amount.
Another that is more likely will be "frustum" extrusion, where the top is uniformly scaled by some amount - which you'd be able to fix using constraints. It should also be able to skew rather than being constrained perpendicular to the sketch plane (you can skew now by selecting a different plane for an extrude).
I'm toying with chamfers on extrusions, but they'd be imperfect in some cases and would be difficult to combine with some of the other extrusion types I haven't even mentioned.
So what type taper were you looking for and which of these possibilities might fit the bill?
(no subject) (by Geert Hospers)
To Paul
I want to model things like propellor.
http://navalex.com/downloads/Michigan_Wheel_Propeller_Geometry.pdf
Basic aerodynamic crosss sections can be defined with a few nurbs that need some variable (I need VARIABLE in length ratio), relative to extrusion.
-chord
-thickness (related to chord)
-curvaure (related to chord)
-rotation of chord
As the simplest nurb is a parabola it seems to me that defining points on those parabola just as on lines could bring the wanted possibilities.
I want to model things like propellor.
http://navalex.com/downloads/Michigan_Wheel_Propeller_Geometry.pdf
Basic aerodynamic crosss sections can be defined with a few nurbs that need some variable (I need VARIABLE in length ratio), relative to extrusion.
-chord
-thickness (related to chord)
-curvaure (related to chord)
-rotation of chord
As the simplest nurb is a parabola it seems to me that defining points on those parabola just as on lines could bring the wanted possibilities.
(no subject) (by Andrew)
Have you looked into Openscad, as a quick Google search 'openscad aerofoil' returned several results including 'https://github.com/ErroneousBosch/OpenSCAD_airfoil'.
(no subject) (by Paul)
To Geert Hospers
That is far more than a tapered extrusion. One thing I was hoping to have is multi-segment extrusions where each segment may change size and be uniformly twisted around some point. Even that is not enough to produce those marine props. I was hoping to do some simple wind turbine blades though. It's unlikely there will be an ability to change the actual shape of the extrusion cross-section along the way. So NACA 0015 at the root will be a smaller and twisted NACA 0015 at the tip. But even this is all quite far down the road.
More generalized construction tools might be possible but they become less of a construction tool and more of a generic surface modeler, which is not what SolveSpace is.
That is far more than a tapered extrusion. One thing I was hoping to have is multi-segment extrusions where each segment may change size and be uniformly twisted around some point. Even that is not enough to produce those marine props. I was hoping to do some simple wind turbine blades though. It's unlikely there will be an ability to change the actual shape of the extrusion cross-section along the way. So NACA 0015 at the root will be a smaller and twisted NACA 0015 at the tip. But even this is all quite far down the road.
More generalized construction tools might be possible but they become less of a construction tool and more of a generic surface modeler, which is not what SolveSpace is.
(no subject) (by Geert Hospers)
To Paul, Andrew
Using Openscad this method describes how to make a Bezier surface. Very suitable to model peropellors and wind turbine blades.
https://openhome.cc/eGossip/OpenSCAD/BezierSurface.html
It looks quite simpel, but not implementable in Solvespace?
Using Openscad this method describes how to make a Bezier surface. Very suitable to model peropellors and wind turbine blades.
https://openhome.cc/eGossip/OpenSCAD/BezierSurface.html
It looks quite simpel, but not implementable in Solvespace?
(no subject) (by Paul)
@Geert Everything in solvespace is a bezier surface. Or more specifically a NURBS surface. All sketch curves - including lines - are rational bezier curves. The issue isn't weather these things could be represented internally by solvespace - they absolutely can. Well, not quite. Solvespace requires surfaces to form a closed shell (looks like a solid object) so creating individual surfaces would violate a lot of assumptions inside and outside the code.
The challenge is making the user interface. We need construction tools that allow useful manipulation of things in the sketch weather thats' 2D or 3D. These things also need to play nicely with the constraint solver, which is more than just a nice-to-have, it's integral to everything.
If OpenScad does what you need, feel free to use it. Blender can probably do it as well. Or FreeCAD. But those programs don't offer the UI and workflow that solvespace does.
See the attached fan design. You can probably get closer to what you want with some boolean operations to cut away parts of it. It's not fully constrained so you can drag the top to change blade pitch/sweep for example. Unfortunately if you want fillets or a round-edge tool you might need to export it and take it over to FreeCAD for finishing.
The challenge is making the user interface. We need construction tools that allow useful manipulation of things in the sketch weather thats' 2D or 3D. These things also need to play nicely with the constraint solver, which is more than just a nice-to-have, it's integral to everything.
If OpenScad does what you need, feel free to use it. Blender can probably do it as well. Or FreeCAD. But those programs don't offer the UI and workflow that solvespace does.
See the attached fan design. You can probably get closer to what you want with some boolean operations to cut away parts of it. It's not fully constrained so you can drag the top to change blade pitch/sweep for example. Unfortunately if you want fillets or a round-edge tool you might need to export it and take it over to FreeCAD for finishing.
(no subject) (by Paul)
A picture:
(no subject) (by Geert Hospers)
@Paul
There is the Constraint "length ratio".
It would be nice if this ratio would accept extrusion as a variable, so that the cross section adapts to it, without changing the start cross section.
Why cannot that be made?
There is the Constraint "length ratio".
It would be nice if this ratio would accept extrusion as a variable, so that the cross section adapts to it, without changing the start cross section.
Why cannot that be made?
Post a reply to this comment: