Coffee Data Science
How Does Sifting Modify Coffee Particles?
After sifting quite a bit, I recalled some of my experiments showed something weird. For sifting finer coffee less than 400um in diameter, the first part of coffee that came through seemed to be light and fluffy. However, towards the end, the grounds seemed to be more coarse. I didn’t think much of it at the time, and I thought it was mostly due to the inside of the bean vs the outside. Now I am starting to wonder if sifting modifies particle shapes and sizes?
This question is important because if sifting modifies the grounds significantly then:
- It is more challenging to use a sifter to measure particle distribution.
- We don’t quite understand how sifting could contribute to the staccato espresso shot beyond reordering particles.
To test this idea, I used spent coffee, and I put it through 500um sifter. I recombined the grounds pre and post sift, and I measured the particle distributions. Here is the wide lens measurement (Grounds, Post Sift) and the tele lens (Grounds Z, Post Sift Z)
The curves are very similar, but do the particle shapes change?
To investigate, I used k-means clustering on top of LBP as I have done previously. I then grouped particles together, and there were shifts in which clusters accounted for what percentage of particles.
I wanted another way to look at this, so I used a cumulative bar chart.
I reordered the previous to this one to see a side-by-side of all the particle bins.
Large particle sizes seemed to have a larger difference in the particle clusterings as the result of sifting.
This study presents some evidence that the process of sifting changes the particle distribution and has some changes on the particle shape. This is part of the puzzle, and even if sifting isn’t regularly used to make staccato espresso, it might have some good purposes otherwise to modify particle shape.
I didn’t see any increase in fines (<100 um), but it is quite possible some modifications are <10 um sizes fines. This method of particle measurement didn’t capture that information. Below 50 um is when this gets a little unreliable even though the smallest measurement is at 28 um.
If you like, follow me on Twitter, YouTube, and Instagram where I post videos of espresso shots on different machines and espresso related stuff. You can also find me on LinkedIn. You can also follow me on Medium and Subscribe.
