Here's my 'Mk2' Vertical Axis Wind Turbine (VAWT). I've always liked the lines of these, but they really caught my attention a few weeks ago when I realised that I couldn't explain how they work.

I'm now a lot less confused about how a wing works (really good explanation here if you're interested: http://www.av8n.com/how/htm/airfoils.html), and I've learnt a few new tricks with the printer.

This is my second version and it can generate a much bigger turbine regardless of the printer that's used; it's a lot more efficient with filament too.

At the moment I don't have this driving anything, but the plan is to attach a small water pump (printed of course) to lift water into a gravity feed system.

update: I've added a couple of different blades sizes to suit different build volumes. If you're after something different, have a play with the SCAD file or if that's not your thing, let me know and I can generate some.

update 2: Well it's been up for a while now and I've got a better idea of how it behaves in different winds. It takes a fair bit to get it going and unless it's really blowing it doesn't spin too fast. I think the idea of driving something with it was probably a bit hopeful, but as a sculpture, I still quite like it. The next version will probably have a few more blades and a bigger diameter. All this takes is a few different settings int he scad file, but I haven't gotten around to it yet.

Changes from Mk1
You can find a copy of my first go here http://www.thingiverse.com/thing:940911.

1. The blade profile has changed to a NACA0024 which is apparently better suited to a VAWT than my 'that looks about right' effort;
2. The blades are hollow now. This reduces filament use (obviously) and reduces the weight in the blade which in turn reduces the 'centrifugal' force. This is a big deal as the core of the blade adds very little to its strength, but does increase the force acting on it. The other nice thing about this is that there were lots of tiny delaminations along the tail in the Mk1 which are gone now.
3. The blade is now made up of 'n' sections (3 in this print) which slot together. By splitting it up like this, the blade can really be as long as you like regardless of build area.
4. The first version used pins to align the two halves (didn't work all that well). This version prints a socket on one side of the join that slots into the hollow section of the other piece.
5. The blade is now connected to the central shaft via 'n' struts (again, 3 in this case). One of the big problems with the first one was that the joint (weakest point) was in the middle of the blade where the force is greatest and each half was supported from it's end. This arrangement should be much stronger.
6. By getting rid of the base, the diameter of the windmill can be as long the struts are. Between this and the segmented blades, you can probably make quite a big version if you're keen.

Ideas for Mk3

1. A transverse web across the widest part of the blade; this will improve the strength without adding much weight.
2. Apparently a 'solidity' ratio of 88% (i.e. 88% of the space is taken up by blade) is the optimum for a VAWT. This would mean blade about double the size of this one.
3. I've got an idea for how to make the struts printable with a very strong connection to the blade.

https://youtu.be/jbJLZhv7Xt8