I am a very lazy Aeropress brewer, but this very laziness resulted in an unexpected observation and coffee connection. Why was it that the bubbles that formed on the top of my coffee were always on one side of the Aeropress cylinder?
First some background. I had been brewing the coffee using an adaptation of the “inversion method“, perhaps it should be called the Bean Thinking inversion method, or BTiM. You can read a good brew guide for preparing coffee properly using the inversion method here. The method I used follows.
After the usual steps of rinsing the filter and blooming the grounds, I filled the Aeropress cylinder using (just boiled) water poured from a standard kettle. Owing to the configuration of the kitchen and the fact that I am right hand dominant, I used my right hand to do this. This generally results in a fantastic bubble pattern on the surface of the coffee, as you can see in the picture, but one that is generally asymmetric – the bubbles appear on the left hand side of the well.
Subsequent repetition of the technique showed that the adverb “always” was a bit of an over statement. However, without recording the bubble patterns after each brew, it is easier to remember the unexpected results, which tend to be asymmetric, and not notice the more even distributions. Michael Polanyi comments on this at length in his book on science and scientific theory/practise “Personal Knowledge” (and while it seems a strange title at the start of the book, the reasoning behind the title becomes clearer throughout the text).
Nonetheless, out of 11 brews, 7 had bubbles primarily on the left hand side of the cylinder, with 3 brews of a more even distribution and 1 with a right hand side distribution.
What could be happening? A discussion on Twitter led to a theory by @baristapierre that air was effectively being pulled into the Aeropress during the pour. Because of the way that water from a kettle spout would flow into the Aeropress, this trapped air, the bubbles, would appear primarily on the opposite side of the Aeropress cylinder from the pour. What would happen if I poured using my left hand?
Here I need to put a note on the method. While I am right hand dominant, I write with both hands and use each hand fairly interchangeably (while tending to favour the right). There may be some effect of my handedness on the pour, but this should not necessarily be too significant. What happened as I changed the handedness of the pour?
Sure enough, the bubbles appeared on the right hand side of the well. And again, but then a run of more ambiguous results. Finally, after 11 brews with the left hand, 5 had bubbles on the right hand side but 4 had bubbles on the left hand side and 2 had a more even distribution (You can see a graph of the final results below).
And so, it appears quite likely that the air-trapping on pouring mechanism suggested by @baristapierre is a good explanation of the bubble distribution seen. But why do the bubbles form at all? The question of how bubbles form in air entrapment caused by turbulence is one that we may ask at the beach while watching (and listening) to the waves crashing in towards the seashore. Each wave generates thousands, millions of bubbles (hence the white caps and froth on the waves as they come in). These bubbles are not only responsible for the sounds you hear as the waves come in, but as they burst they release aerosols of salt and organic matter into the atmosphere that in turn affect cloud formation and can even influence hurricane dynamics.
And yet astonishingly, it was not until 2002 that a theory was developed (and experiments performed) to measure the bubble formation in water waves. Using high speed video, the study photographed simulated waves in a laboratory and found that the waves formed in two phases. A first involved the collapse of the air cavity formed as the wave folded in on itself. These bubbles are larger and responsible for the low frequency “crash” of the wave that you can hear. The second, concomitant, phase, involved the impact of the water with itself, much like the bubbles in the Aeropress. These bubbles were much smaller (< 1mm radius) and so the sound associated with them was a higher hiss. Together these bubbles create the crash-hiss sound that is so familiar with the breaking of the waves.
By carefully analysing 225 images over 14 wave breaking events, the study found that the number of bubbles per unit area decreased with bubble radius but with a different dependence depending on whether the bubbles were smaller, or larger, than approximately 1mm.
These results were checked against the Aeropress brew with the highest number of bubbles (1 May 2020). On the positives, it was clear that the bubble size observed was similar to the ocean waves study of 2002. The bubble density also decreased with increasing bubble radius (see the graph below). However the rate of decrease with radius was not as observed in the ocean study. There could be many reasons for this, including the fact that counting from a steamy photograph taken 30 seconds after the pour was not the most accurate method of analysis! Nonetheless, it does emphasise, that, though there are many connections between the physics of a coffee and the physics of the world, and though there are even more connections if we use the coffee as a prompt to let our minds wander into the wonders of the universe, the Aeropress is not an ocean and we can stretch analogy too far.
And so, while we may not learn much about ocean dynamics while brewing Aeropress coffee, it turns out we can learn a fair bit about experimental technique and how science is actually done, including what Polanyi meant about noticing unusual distributions. This will be the subject of the next post. In the meantime, what have you seen while brewing coffee? Do let me know in the comments, on Twitter or on Facebook.