I decided to make this engine open source so that anyone could build it. The PE2 design (stands for printed engine 2) is licensed under the Attribution-Share Alike – Creative Commons license. The previous post introduced the Stirling engine with a lot of information about the design. This post and a few more will provide all the detailed information needed to build the engine. The design files in STL format are already posted on Thingiverse for the 3D printed parts. This post provides the part drawings for the non-printed parts and my comments on building each part. The full drawings of all the parts presented here are also available on Thingiverse in PDF format and are easier to read at full size.
The following two assembly drawings show the arrangement of the parts and their names.
The second assembly drawing has various parts made invisible so you can see some of the internal parts. I have left out out the fasteners (machine screws, nuts, and washers) in these drawings, but I will provide a list in a future post.
The cold plate was made from .125” 6061 aluminum. Other thicknesses down to .080 should be adequate. I would not recommend thinner material for two reasons. First, there is one tapped 6-32 hole through the aluminum. For secure threads and good sealing a thicker material is better. Second, a material of .063 thickness will start to have a certain amount of flexing with the pressure changes (according to my calculations) when the engine is running. Flexing will always result in power loss.
The three .196 inch diameter holes around the perimeter are for optional supports to suspend the engine in the inverted position for heating from above. The video of the engine running has some shots of this. They are completely optional.
The six holes on a 2.017 inch radius are are for attaching the cylinder body. Two of the holes plus the tapped hole near the center are for attaching the flywheel support. The CAD program gives the exact dimension but you don’t need these accurate to 0.001 inches.
For those not experienced in laying out reasonably accurate metal parts here are some pointers. You need an adjustable divider that is reasonably rigid, a steel rule with fine graduations, a metal scribe, and a prick punch which is like a light center punch with a sharp point. Use the prick punch with a light hammer to mark the center for what will be the .094” diameter hole. Using the steel rule set the divider to 2.017 inches (2.02), and scribe the circle for the 6 holes. Six holes can be easily and accurately laid out with the divider by starting with a prick punch mark somewhere on the circle you just laid out. Use the divider without changing the setting by placing it in the new prick punch mark and making two short arcs that mark the circle. Prick punch these arcs where they cross the circle and work you way around the circle until you have all 6 holes marked.
The tapped hole and the larger hole need to be found either with a protractor or by geometry. I created a perpendicular from two of the six holes that were just laid out, scribed a line through the center that goes through the location of the tapped hole and the large hole. Then just measure the distance from the center to the two holes with the steel rule.
Always layout everything before you start cutting and drilling. Make sure to layout the perimeter circle and the 3 optional support holes if you plan to use them.
Incidentally, the outer contour of the cold plate is not important. You can make it rectangular if you want. The larger the better for more cooling.
The hot plate also uses 6061 aluminum. Again, anything down to .080 inch thickness will be fine. For simplicity you could just cut the disk out as a 4.41 inch diameter circle. I did that and then later scribed the outer contour using the cylinder body as a guide.
The power piston and the brass power cylinder are a matched pair. I didn’t make a drawing for the power cylinder because it is just a cut to length piece of brass tubing 1.000 outside diameter x .065 wall x 2.0 inches long. It’s important that the brass tube be 1.000 inch diameter because the O-ring used to seal the power cylinder to the cold plate depends on this measurement. As long as you need a lathe to make the power piston, you might as well true up the ends of the power cylinder on the lathe so that it aligns perfectly perpendicular to the cold plate. The inside and outside diameter surfaces of the power cylinder are used as manufactured.
The power piston outside diameter is turned on a lathe to be a slip fit in the brass tube. You want the piston to easily fall under its own weight when the cylinder is oriented vertically. It should be essentially frictionless. When you seal the bottom of the brass power cylinder, the power piston should fall very slowly under it’s own weight (more than 10 seconds to fall one inch). You’ll of course need to attach the power piston insert or plug the hole to test this. The outside diameter of the power piston will be about .001 to .002 inches smaller than the ID of the power cylinder.
The drawing above shows the end result of what the power piston should look like. You can bore it out of a .875 or 1.00 inch aluminum rod. In my case I made it from a .875 diameter x .065 wall aluminum tube and then made a .125 inch thick disk and press fit it into one end. You could also epoxy the disk in. The 0.750 inch inside diameter is needed to fit the piston insert.
I plan to test a new 3D printed piston insert soon that will allow the piston to be made using just the aluminum tube without the end piece. As soon as I get that built and tested I’ll make the new piston insert STL file available. Then all you’ll need to do is turn the OD of the aluminum tube and cut it to length.
Foam displacer parts
First a little information on the design. A Stirling engine of this design works by forcing a gas (in this case air) back and forth through heat exchangers that alternately heat and cool the gas. These alternate heating and cooling cycles result in pressure changes which drive the power piston. The displacer is the part of the engine that forces the gas back and forth. It also experiences the pressure changes that could be as high as +/- 0.75 PSI on this engine. Although that may not sound like a lot of pressure it is a 5 pound load spread over the flat faces of the displacer. If the displacer were hollow it would need internal stiffening not to flex, adding more mass to the reciprocating displacer. I provide this information for those that might want to modify the displacer.
Light-weight closed-cell foam can handle these pressure changes without flexing and weighs much less than even the thinnest printable disk of ABS. I still plan to test an all-printed displacer in the future, probably a two-piece glued together one. Until then I’m using the following design.
The three following drawings probably make these parts look more involved and complex than they are. The printed displacer is a pattern for making the foam pieces. You can ignore these drawings if you want and just make the displacer 1 inch long and make sure the outside diameter doesn’t extend further than the printed displacer part. Also minimize air gaps and spaces between the foam pieces and the printed displacer.
You can probably use any rigid closed-cell foam that can handle temperatures up to 170 degF. Styrofoam or expanded polystyrene tends to be readily available. I use a polyurethane foam called Last-a-foam which I like because it is easy to cut and sand. It is also usable up to 275 degF.
Displacer foam 1
The displacer foam 1 part fits inside the printed displacer part so it is smaller in diameter than the other parts. It also needs to be only .20 inches thick. The other foam pieces are all .25 inches thick.
Displacer foam 2
The displacer foam 2 part fits against the other side of the printed displacer.
Displacer foam 2A
The displacer foam 2A parts (you need two of them) can be mounted against either the foam 1 or foam 2 parts to make the complete displacer stack 1.00 inches thick. Both the foam 2 and 2A drawings show a 2.93 inch diameter. It isn’t really practical to hold tight tolerances with foam parts. The important thing is that they don’t extend out beyond the OD of the printed displacer. If they do they will rub on the cylinder body which will prevent the engine from running.
The remaining parts for this engine are either stock parts such as machine screws and O-rings or cut-to-length parts such as pins and shafts from 1/16 inch music wire. I’ll cover those parts in the next post.