The Black and Tan (or Half and Half in some countries) is a fascinating beverage, and an excellent demonstration of density stratification in fluids.  Recently we have started to explore the possibility of using 3D printing to create a device that can make a Black and Tan by displacing a less dense beer with a denser one using a siphon rather than the traditional method of carefully pouring the less dense beer on top of the denser beer.

Background:

As we have previously demonstrated, fluids of different densities can easily be layered via a wave tank.  After stumbling across a video on 3D modeling software, I decided to try my hand at Fusion 360.  A few tutorials later, it seemed like designing a Black and Tan maker would be a fun thing to try.

If a denser fluid can be injected underneath another fluid, it should displace the less dense fluid and form a stable layer under it.  Therefore it should be possible to create a funnel or siphon to transport a dense fluid to the bottom of a container and have it form a stable layer under a less dense layer.  To make things more interesting, why not use a Pythagorean cup (i.e., a self starting siphon) as the basis for the siphoning?

Design:

The basic design dubbed the Beer Siphon Mk I, as shown below, is a largish reservoir with a pipe that runs from just above the bottom of the reservoir, up a bit, then down through the floor of the reservoir to just above the bottom of the pint glass.  The bottom of the reservoir also has small feet to hold it slightly off of the top of the glass to allow air to escape as the glass fills.  When a fluid is poured into the reservoir, it will move up the short end of the pipe to the bend.  Once the fluid reaches the bend it will can flow down the longer bit of pipe into the pint glass, starting the siphon action which should continue until the fluid level reaches the inlet just above the floor of the reservoir.

Cross section showing the Beer Siphon Mk I sitting atop a pint glass.

The 3D model for the Beer Siphon Mk I is available on Thingiverse.

One problem with using a self starting siphon to inject a fluid under another fluid is that there will be air trapped in the pipe that can prevent the siphon from starting properly.  To fix this issue, a small hole was placed in the pipe just below the main reservoir, to allow air to escape.

Being very new to both Fusion 360 and 3D printing, the first attempt was a very simple design that, in theory should work, but would probably have major defects.

The model was then sliced using Cura LulzBot Edition rotated 180º around the X axis, with support everywhere.  It was then printed on a LulzBot TAZ 6 using PLA and a 1.2mm print head.

Testing:

It was obvious, just by looking at the 3D printing results, that there were large holes, especially around the edge of the reservoir and around the pipe.  Less obvious were holes along the curved section of the pipe.

Underside of the reservoir. Holes visible along the front edge.

Reservoir interior and upper portion of the siphon pipe.

To fix the holes we melted a block of paraffin wax and poured it over the bottom of the reservoir to fill in the holes. We then filled a narrow glass container with melted wax and dipped the curved bit of pipe to patch its holes as well.

Wax patched underside of reservoir.

Wax patched reservoir interior and siphon pipe.

Once the holes were filled, the siphon was placed over a standard pint glass and 8 ounces of water was poured into the reservoir.  Unfortunately this revealed that the upper portion of the pipe was mostly blocked, leading to just a slow trickle over the curve in the pipe.  With just a slow trickle, a proper siphon was never started and the flow stopped as soon as the level in the reservoir reached the bottom of the curve in the pipe.

What went wrong:

1. Insufficient thicknesses led to holes in the print, which needed to be patched with paraffin wax.
2. The feet were a bit rough and difficult to line up with the lip of a pint glass.
3. Slicing mistakes placed support material inside the siphon pipe which could not be removed.

Layer view showing support material blocking the siphon pipe.

What went right:

1. Once the holes in the reservoir and pipe were patched, some siphoning did occur through small gaps around the support material.
2. The small gap in the inlet managed to print correctly.
3. The small hole in the pipe below the reservoir also printed mostly correctly.

Possible Improvements:

1. Eliminate need for the tiny gap on upper end of the pipe, by looping the pipe back around to the floor of the main reservoir.
2. Thicken certain surfaces to prevent holes when printing.
3. Use chamfers or fillets to overlap layers near edges to prevent leaking.
4. Replace the round feet with radial ridges extending the the edge of the reservoir.
5. When slicing, rotate the model so supports don’t run across the pipe.
6. Use a golden spiral for a more elegant pipe shape.

Conclusion:

While this attempt failed to produce the desired result of being able to siphon a dense fluid under a less dense fluid, it did provide much needed insight into the nature of 3D printing.  With a few modifications we are hopeful that a fully functional Pythagorean cup can be printed.