3d Printed Liquid Piston Stirling Engine V4 reattempt and success!

 After several more attempts I managed to sort out the gasket issue and am delighted to report success!

Here's a video of the engine running.

Several tricks made the difference in successfully printing the gaskets.  First, I changed Prusa to print the parts sequentially.  This eliminated a significant amount of stringing.  Second, I used a dirty print bed.  Instead of my normal practice of cleaning it with isopropyl alcohol I intentionally rubbed it barehanded to distribute skin oil on the surface.  Third, I removed the print immediately after the print head retracted while the build plate was still hot.  This yielded a part that only required trimming off a single blob before installation.  The gasket is a snug fit but will press in place.

Manufacturing the cover requires cutting the Acrylic sheet to size and drilling the mounting moles.  Conveniently the housing provides an excellent drilling guide.  A 1/8" drill bit provides appropriate clearance for the m3 mounting screws and nuts.

The model in the video was powered for ~15 minutes in steady state operation for about 60% of that time with tapwater as the working fluid.  No leaks were noted.  (Yeah!)  Visible in the video, the water starts to discolor and some particulates are visible.  I'm unclear if the source of this is electrolytic corrosion of the resistor leads or the resistor housing material breaking down.  This merits additional investigation.  Near the end of the video clip the water level in the output column appears much higher than earlier in the run.  I believe this is because the cold end is unable to reject heat sufficiently.  In the source book for this model another of the author's engines uses a hypodermic needle as a slow leak to balance the pressure in the displacer chamber and output tube.  I believe that would help with this issue, as would sorting out a heat rejection mechanism.

Suggestions for future work:
  • Understand the cause of the discoloration.
    • A possible approach could include wrapping the suspected components with Kapton Tape.
  • Play with the position of the heater to see if that impacts any operational characteristics.
  • Regenerator and/or active cooling for the hot end.
  • Think about a fluidics mechanism to "cut-off" heat from the heater while displacing the hot air to the cold end without dousing the heater in water and dumping all its heat into the working fluid.
  • Mr. West's book indicates the connection to the output colum, aka the tuning loop, should be significantly longer.  Do the math on this to define "significantly" and create some test models to measure how this impacts performance.
  • Instrumentation: Record pressure, temperature in the hot end and cold end, and input power.
  • Compare the mathematically predicted oscillation rate with the physical experiment.
  • Big stretch goal. Think about a 3d printable hybrid engine with a metal (tin can?  test tube?) hot end, 3d printed liquid cooled housing, liquid piston, hydraulic driven mechanical displacer with an integrated annular regenerator.
A few more things:
A prodigious number of Stirling engine resources, including many of Mr. West's papers, can be found here: National Technical Reports Library - NTIS

A paper describing a similar engine was published here, and the paper is open access. 
Ning YangRobert RickardKevin PluckterTodd Sulchek; Integrated two-cylinder liquid piston Stirling engine. Appl. Phys. Lett. 6 October 2014; 105 (14): 143903. https://doi.org/10.1063/1.4897621