# Designing windmill blades in solidworks

I am trying to design the windmill blades (the green part) in Solidworks but am having trouble since the radius of curvature for the blades varies at different points. I tried bending it as a sheet metal, but am having a lot of difficulty. Any ideas?

• @user287001 well, cad applications have builtin tools for this. (Well designspark does not but soldworks does) Anyway, why calculate? Well you may need to rescale the design. But yes you can also simulate by bending a paper. Just measure how the bend happens in reality, in less than 5-15 minutes. Just bend a paper, photograph it (preferably the shadow from the side) and draw a curve. May 16 at 7:15

I recommend creating a 3d sketch of the outline of one blade, then using the surface modeling tools to create a body in that closed sketch. Then circular pattern the blade.

This case got some history in Graphic Design SE. It got an answer which suggested making it by joining together 3D models of the blades.

The questioner, of course, was not interested in it. It's only an imagined drawing, no matter it's a 3D model. An acceptable 3D model should be based on measuring which form the paper takes after the cusps are pinned to the axis. Or the behaviour of the paper should be simulated in physics capable software.

I do not have advanced CAD software, but Solidworks should have in its widest version simulation tools which can find what form the paper takes.

How to measure the form of a real propeller:

1. get a real propeller or make at least one blade by clipping it from a real paper. Bend it by pinning the cusps. Fasten the middle part on a planar background wheel like it's fastened in the real thing.
2. use a 3D scanner or photogrammetry to make a 3D model.

Both methods are in practice tricky. Good 3D scanner costs money and companies who do scannings can write hefty bills.

Photogrammetry is possible with free software, but the lights and the target should be prepared properly. The surface cannot have any shadows nor glosses, it must have high contrast non-repeating texture and one must take tens of photos from different enough directions. The paper propeller doesn't stand any wind during the session and the camera must be in manual mode to keep the focal length and aperture constant.

3D scanning, photogrammetry and obviously also physics simulation give the surface as polygon mesh model. I guess CAD program users would hate it because

• traditionally a polygon mesh has been uneditable in programs which use spline based surfaces.
• scanned or by photogrammetry generated mesh is often rough. One must very likely fill holes and smooth it in a mesh based 3D program.

The situation has been changed in last few years. Advanced CAD programs have got tools which create a spline based skin over polygon meshes with reasonable little manual intervention. That skin is like it was drawn from scratch with usual surface tools. See this video for a start: https://www.youtube.com/watch?v=E8RDOfXEh-o

To see more videos of what's possible today make searches for reverse engineering scanned or mesh models.

• You can light your model and scan the objects shadow. If you scan the minimal shadow you should know the shape. But yes photogrammetry would work also. May 23 at 14:33
• The form is conical. No light direction will give paper thin shadow. A single point light source would be needed and that's in practice impossible. May 23 at 14:49
• i didnt say it will be nothing i said minimal which implies not perfect. But like all measurements its probably a good estimation, plus it comes with error bars ;) anyway fitting a curve on a polygon is rather trivial. May 23 at 14:56
• Can you bend a piece of paper and show an example of the process? If it works it would be very useful - so useful that it, of course isn't good business to reveal it. May 24 at 11:04