I was excited about the thought of a mouth blown, plastic piston motor that I decided to rough out some design details and see how hard it would be for a beginner's class in Solid Works, as well as result in data files that could be converted into CNC or 3D printer command files. I think I have something that should be possible in both plastic and aluminum, a bit of work to define all the parts, but possible to make on either machine, and hopefully fun to play with and show off as a gee-whizz, look what WE can do here at HSL!
The basic idea is a pair of double acting pistons, with the crank offset by 90 degrees so that no matter what position it is left in, there should be enough torque to start it pumping the next time someone blows into it. Although I have sketched in places for rings to improve the seals, efficiency will not be a big part of this design. And I just drew in propellers for output, but anything else simple can be put there instead.
In the overall drawing we have a top view. The mouth piece is at the top, and then the 'cylinder block' section', with a central air intake runs down the middle, two slide valves on either side of the intake, then the double acting pistons. Below that is the 'crank case' section, with the piston rods coming straight out and going to an external piston rod pivot (necessary for double acting pistons) then the pivoting part of the piston rod, then a two part crank bearing. The straight to pivot joint has long projections out each side, one to slide in the frame and help keep the straight part straight, and the other side sliding in a slot in the slide valve to push it into reverse flow at each extreme of motion. Sorry, no overhead cam on this one. I tried to come up with the simplest design to make it better for a first or second project. The frame of the crankcase is U shaped to stiffen the crank shaft area.
In the second drawing I have tried to detail most if the individual pieces. Sorry, but until I take the SolidWorks class, it will hard to make a better, more detailed set of drawings. But this way, there is still a little fine tuning engineering for the class to do as the design is entered. There is a side view of one piston, with the slide valve behind it.The piston rod passes through the back of the cylinder (probably a good place for an o-ring slot) then reaches the pivot. The slide valve with its slot is visible behind that part, and a section of the crankshaft is shown at the right. Above is the piston themselves, will piston rings sketched in. The pivot arm joint should be two-sided so that the pivot is under double shear. The other end needs to be big enough to bolt on the bottom half of the crank bearing because I don't see how to assemble even the plastic pieces without it. The Slide Valve has a tapper cut at the front end to vent the top of the piston when the slide is pulled back , and then plug the exhaust and position a slot over the piston port to let air in from the intake manifold. The opposite is done at the other end so that when one side is venting through a cross hole, the other is passing compressed air, then slide it in the direction of the piston motion at the extreme end of travel to switch it to go the other way.
I have sketched it out to make the cylinder in a constant cross section cylinder head. On the 3D printer, the air plenum can be square, when milled from aluminum round is going to be easier. The two valve slides channels will be the hardest part to mill, but the top and bottoms can be left rounded if the valve slides themselves have rounded edges. The heads at each end contain the notches the will vent air in and out of the cylinder. The remaining pieces should be easier to cut without complex internal channels. In aluminum, I have sketched 4 long thin bolts to hold the heads, cylinder block, and crankcase together. In plastic, this can either be bolts or just glued together.
Well, what do people think about this as a project for learning Solid Works, then learning how to take the files from Solid Works and run them through CAM software to generate g-code to run the mill or printer?
Mike