This is a personal project which began in 2009, and has been slowly approaching completion since. The design is the popular "Low Temperature Differential Stirling" or "Miser" proudly purchased from Jerry Howell at http://www.model-engine-plans.com.
Engines are devices which convert a supplied energy (usually in the form of combustible
fuel) into useful mechanical work. In the case of Stirling engines, the supplied energy is in the
form of heat, rather than fuel! It is a heat engine, operating between two different temperature reservoirs. The implication is that a Stirling engine must be physically placed between a heat source, and another low-temperature sink. Because the heat is not produced within the machine, it is called an external combustion engine. Potential heat sources include the flame from a candle, a stream of hot water, or light from the sun.
Since there is no combustion within the machine, it may be sealed
such that it does not exchange any mass with its surroundings. This
means that the ‘working fluid,’ usually a finite quantity of gas such as
air or hydrogen, is continually reused within the cycle. For this
reason it is also considered to be a closed vapor cycle (gas and diesel
engines are open cycles). During operation of a Stirling engine, the
working fluid progresses through various temperature and pressure
states, but always returns to its initial conditions at the start of
Up to this point, approximately 50% of the machine has been completed. The primary components which remain to be completed include the power piston (graphite) and another cylinder. Otherwise, only relatively trivial parts remain!
The most fun part to make so far has been the flywheel. Originally I intended to produce this manually but decided that due to time constraints it was more practical to produce on the CNC mill. The process of making parts by this method is summarized below:
First, the paper drawing for the part itself is studied and it is redrawn in a CAD program. I take this time to consider how the part will be most effectively made, using a minimum of tools, with the least amount of material, and in the least amount of time. I decide to mill the majority of the part from the top face of the wheel, then flipped over. At this point, the excess material can be faced off and any features added in order to match the other side. This will be only two machining operations.
The CAD drawing is then imported into my CAM (computer-aided manufacturing) program such that I may decide the manner in which the CNC mill will physically cut away the part. The CNC mill is able to move a cutter within a volume which is described by an X-Y-Z location. So, in this program, I declare my cutting tool diameter, speed, feed rate, and physical length, as well as theelements of the drawing I would like to cut or keep.
After programming is complete, I run a 'virtualization' of the machining process. This allows me to view a rendered graphic of the finished part.
A 1/2" thick piece of 6061 aluminum was nearly a perfect size for producing the 5" diameter flywheel. I held it in a vice, and marked out its approximate location in layout ink (red), with scribe marks showing the outside diameter and the center point for my convenience.
After facing off the entire piece of aluminum (to make it both flat and give it an attractive finish), I begin to execute the CAM program which was created on the computer earlier. The first cut shown above creates the recess into which the spokes will later be cut. The tool is a 1/2" endmill.
Here, the spokes have been cut with the same tool, and with some of the tighter corners cleaned up using a 1/4" endmill. The white fluid is a water-based oil lubricant which aids in both lubricating and cooling the hot cutting tool, as well as flushing away chips.
The 1/2" endmill again has been used to generate the outer diameter of the flywheel. The coolant nozzles are visible in the upper right. Without the coolant, the tool might not survive!
Some of the finer details have been created using an 1/8" radius ball end mill. This include both the fluting along the circumference of the flywheel, as well as the subtle fillets found nearby.
The first side being completed, the part is removed from the vice. I use the vertical bandsaw to remove the unnecessary material from the finished part. Only the other side remains to be completed.
Taking advantage of the easily-machined aluminum jaws, I write another program to cut a 5" diameter recess into them. This will allow my to grip the flywheel by its circumference so that I may finish the other side.
Here the flywheel is prepared to be face off, and the remaining features added.
Facing off all the extra material that remained on the back, all that is left are the recessed cuts and the fluting using the ball mill. This is done next.
Here is an action shot of the ball mill putting the fluted along the outer ring. The finished flywheel waiting to be removed is shown below.
This part would have required much more time and effort on a manual mill. The nice thing about the CNC mill is that success is more or less guaranteed assuming you planned and programmed properly! Also, the part looks great, tooling marks and all!
THE POWER CYLINDER
This part is one of the most important as far as attention to detail. In particular the bore of the cylinder must be a nearly air-tight sliding fit with the mating graphite piston. Because the piston will be made last, the exact bore diameter of the piston is not critical, merely that it does not have any taper and should be very smooth. 1.5" O.D. Round 304 stainless steel stock is selected and the rough shape is turned on the Monarch 10EE lathe.
The external features on this part are mainly aesthetic: fillets, flutes, and chamfers. A 20 degree diamond insert is perfect for removing most of the material between shoulders.
A 1/2" drill removes the bulk of material from the cylinder's bore. A boring bar will be used to bring it to its finish size.
A corner rounder is used to generate the external fillets.
Similarly, a grooving tool is used to produce the internal fillets.
It is important that the bore is finished last. Had it been finished first, any subsequent removal of material could affect the straightness or roundness of the bar. Internal stresses in the stainless steel can warp a part as it is cut.
This boring operation was tedious. A target I.D. of 0.625" was easily approached by taking light finish passes at a low feed rate. A .0004" taper along the length of the part was remediated by some light sanding and attention with a stick of craytex. Telescoping gages and a tenths reading Mitutoyo mic were used to comparatively measure along the length of the bore.
The part was removed from the lathe and separated from the stock in the bandsaw.
Placed into another lathe, the back was cleaned up and a slight chamfer is tuned onto the bore entrance.
With the remaining work to be done on the mill, a good method for holding the part was needed. Because indexing was necessary for the fluting operation, a mandrel was made so that the cylinder could be held in the indexing fixture. A mandrel is quickly made:
Aluminum stock is first turned to a snug slip fit on the cylinder (approx. 0623" O.D.)
It is drilled to 7/16" in preparation for a 1/4 NPT tap.
Approx 13 full turns brings an NPT tap to proper depth.
A vertical bandsaw is used to slit the mandrel lengthwise. As a piece of NPT pipe is threaded in, the tapered threads, the mandrel expands to capture the part by the bore. This mandrel was transferred to the indexing fixture on the Bridgeport to receive twelve equally spaced flutes using a 3/16" ball endmill.
A final flat feature is cut using a flat endmill. The part is deburred and declared complete.
... under construction