Inside the crucible - what IS going on in there?

Seems like a while since we've discussed what we think, speculate, or "know" happens inside the portafilter basket during the various stages of extraction.  A bit of this has come up in another discussion:
http://www.baristaexchange.com/forum/topics/ideal-tamper-base-shape...
however it seems like a more thorough discussion can occur if we break away from that original question.  Some good stuff put up over there, so you should definitely read it if you haven't already.

So, what do you think is going on in there?


(note: I will ask one favor... let's approach this neutrally, scientifically, and skeptically.  I'm pretty sure that we do not actually know exactly what is happening in there... we have developed some really good theories to explain what is happening, have made observations that support them, and have made other observations that may not be fully explained.  So let's approach the subject in that way - submitting what you suspect is happening in there and sharing any supporting evidence you may have.)

Also, if you happen to have any links to other discussions that you feel have covered this well, please share them.  I don't get out as much as I should...

Ok, geek on...

Views: 198

Reply to This

Replies to This Discussion

I'm a newbie...1st post...so go easy on me!

I saw a similar experiment being done with engines once. The combustion chamber is a similar "black box" that's hard to understand. Anyway, they actually drew a grid in the gases inside the chamber with lasers, and were able to photograph the grid as it distorted. Pretty cool.

Brady said:
Mike Sabol said:
... I think someone at MIT should set up a super-computer simulation of espresso extraction.

You know, we could probably get a little bit of a look into what is happening by mixing some ground-size "marker particles" (glitter?) in with the coffee. See where they end up once the whole thing is done?

Though this would be a great excuse to play with nano-bots, so I'm going to recommend that route instead.
Layers of bleached coffee? Would be a real bitch to get the tamping right without messing it up.

I like this topic. Always wanted to have a better idea of what was going on in there! Seems like the issue gets clouded a bit by changes in dose to affect a head space change, when you'd need to change the gasket thickness to limit it to a single variable.

Brady said:
Mike Sabol said:
... I think someone at MIT should set up a super-computer simulation of espresso extraction.

You know, we could probably get a little bit of a look into what is happening by mixing some ground-size "marker particles" (glitter?) in with the coffee. See where they end up once the whole thing is done?

Though this would be a great excuse to play with nano-bots, so I'm going to recommend that route instead.
Well, that brings up the question of how much headspace really matters, and how it effects the extraction.

Does anyone disagree with the fines migration theory? What about the puck matrix concept?

If the pressure is building above the puck to force the pressure through, at what point does the pressure switch to omnidirectional?

Here's an interesting idea to toss into the mix. When I had a Gaggia with no three-way-valve, I would on rare occasion end up with the puck stuck to the screen, with the basket coming away from it almost perfectly clean.

Is this a demonstration of the omnidirectional pressure working together with the puck matrix? (the sticking is owed to the same attribute as the puck solidity.. curly coffee particles)

Just some more thoughts before I head out. I'm still not done, but I thought headspace was an interesting idea to explore.
Now that I've been thinking about it more I have some further thoughts. The holes in the filter basket do not provide a lot of resistance to incoming water so there is no reason to think they would provide additional resistance to the the extracted liquid when it hits the bottom. If this is the case then the water flow would be mostly one way. Down. I think there might be some omni directional pressure stuff going in the saturated layer while the water is making it's way through the puck. I can imagine convection currents getting set up in that saturated layer before the water makes it all the way to the bottom of the basket but after the bottom is hit I think the flow would change to directional. This would explain the fines migration. The top layer of coffee would act like a viscous fluid while the lower portion of the puck was still acting to create resistance. This would put the particles into suspension while there was still omni-directional pressure. It this case I think it would resemble a french press for a time. The fine, (more dense), particles would fall to the bottom and the bigger chunks would float to the top. I think the head space at the top of puck would be so the water could provide uniform pressure across the top of the puck which would lead to a more uniform extraction. If the puck were in contact with the top screen before the extraction started the water path would end up much more complicated and make channeling more likely.

For an experiment we could try placing some very small pieces of rock or a tiny quantity of sand at the top of the puck after the tamp. I bet you would find the sand at the bottom of the basket after the shot was pulled. Provided, of course, no damage to the machine would result from such a thing. Perhaps some mineral powder like that used to adjust tds in fish tanks.
Don't think I can contribute to any advancement here, but I've certainly experienced too little head space. This was when I changed gaskets to get a LM potafilter to fit in my E61 type machine. Is the conventional wisdom that as long as there's enough, it's not critical?

Seem like too much would allow for currents of flow across the face of the puck? Kinda the opposite of what you get with an old lever machine, with the column of water descending thru the grounds?

Jason Haeger said:
Well, that brings up the question of how much headspace really matters, and how it effects the extraction. Does anyone disagree with the fines migration theory? What about the puck matrix concept?

If the pressure is building above the puck to force the pressure through, at what point does the pressure switch to omnidirectional?

Here's an interesting idea to toss into the mix. When I had a Gaggia with no three-way-valve, I would on rare occasion end up with the puck stuck to the screen, with the basket coming away from it almost perfectly clean.

Is this a demonstration of the omnidirectional pressure working together with the puck matrix? (the sticking is owed to the same attribute as the puck solidity.. curly coffee particles)

Just some more thoughts before I head out. I'm still not done, but I thought headspace was an interesting idea to explore.
I think you're definitely on the right track.

I think head space determines the size of the "water cushion", as it is known.

I agree with you that the pressure is mostly downward since the puck BELOW the building water cushion is what is providing the resistance.

Your description of what happens to the puck during extraction reminds me of the Sabados dosing theory. I don't have a link to the article, but this thread on CoffeeGeek discusses it a bit.

I have found that I tend to usually prefer the Wendelboe method, as it was what I was already doing before I read the Dosing Debate articles as a result of my own experiments and tastes. However, I also agree that it depends on the roast as well as the coffee.

The real question, is why does it matter which coffee or roast style is used, and why do different methods end up in different cup results even with the same bean?

I go by the "coffee is a flavor container" theory more often than not. The more desirable extractables the coffee contains, the lower the dose required to get the most out of it. The inverse is the rule in the other direction. It works.. most of the time.

But sometimes I find other scenarios. This espresso blend that I've been working with lately REALLY prefers a high dose. It also prefers a substantial headspace. Triple Basket to the rescue.

This, however is not always the case. Some coffees don't like headspace at all. I have developed a bit of a formulaic chart in my subconscious that correlates what goes in with what comes out, but nothing at all about the actual reasons as to why these things are. Of course, there is some philosophy involved in that. (I can assume causes, but I have zero proof, and it may not be true the next time)

At this point, I think I may have confused the direction further. Not at all what I intended.

Mike Sabol said:
Now that I've been thinking about it more I have some further thoughts. The holes in the filter basket do not provide a lot of resistance to incoming water so there is no reason to think they would provide additional resistance to the the extracted liquid when it hits the bottom. If this is the case then the water flow would be mostly one way. Down. I think there might be some omni directional pressure stuff going in the saturated layer while the water is making it's way through the puck. I can imagine convection currents getting set up in that saturated layer before the water makes it all the way to the bottom of the basket but after the bottom is hit I think the flow would change to directional. This would explain the fines migration. The top layer of coffee would act like a viscous fluid while the lower portion of the puck was still acting to create resistance. This would put the particles into suspension while there was still omni-directional pressure. It this case I think it would resemble a french press for a time. The fine, (more dense), particles would fall to the bottom and the bigger chunks would float to the top. I think the head space at the top of puck would be so the water could provide uniform pressure across the top of the puck which would lead to a more uniform extraction. If the puck were in contact with the top screen before the extraction started the water path would end up much more complicated and make channeling more likely.

For an experiment we could try placing some very small pieces of rock or a tiny quantity of sand at the top of the puck after the tamp. I bet you would find the sand at the bottom of the basket after the shot was pulled. Provided, of course, no damage to the machine would result from such a thing. Perhaps some mineral powder like that used to adjust tds in fish tanks.
I'll submit these to the discussion. They seem relevant since it seems like there are 3 different models for water and puck behavior that have been mentioned so far (well, 4 if you count the hot tub angels). One where the puck fuses and becomes a porous medium, one where the puck behaves more like a packed bed, and one where it acts like a fluidized bed. I'll dig around a bit more (and read the threads that Jason attached) later, but in the meantime - here:

http://en.wikipedia.org/wiki/Fluidized_bed
http://en.wikipedia.org/wiki/Packed_bed
(some good overviews).

I also found this discussion on pressure drop of fluid through a porous medium that may be useful to some. It just gave me a headache :)
http://www.columbia.edu/~rl268/ChESite/E3110/Handout_15.pdf

Not sure what that adds to the discussion, but perhaps something in there will click for someone.

I do notice a reference to the migration of smaller particles in fluidized beds in the wikipedia article. That would certainly help explain the fines migration...
This may have been asked already, but has there been any move to create a glass basket? Is it even possible to create a portafilter you can see through? How wild would that be to have a portafiler that allows access into its mystery world of extraction? Would this be a new marketing schema? Imagine the impact on the everyday customer who could WATCH their espresso being created! On second thought, do we even know there would be a sight to see? I don't think i contributed anything just then; just more questions...
In another article, they discuss the idea that fluid flow through a porous bed can be modeled as though the water were flowing through a bundle of long, thin capillary tubes. That strikes me as being a good approach if you assume the puck fuses together and the grounds remain stationary - then the water will have to find millions of tiny little pathways around the grounds.

Regarding the omni-directional-ness of pressure... I'm digging around in my Fluids text for something that my brain can process these days. Slim pickings... I'll find something to back that statement up, but have two thoughts right now.

1. If you are talking about fluid flowing through a pipe, the pressure is essentially equal in all directions. A gas bubble in the water flowing through a pipe is still spherical.

2. Thinking about it a bit more, there would have to be a net downward pressure on the puck resulting from all of those friction losses in the channels. It isn't quite as severe as those that describe a 130psi "blast" suspect though... its like the difference between holding the bowl of a spoon under a stream of water vs holding a straw under it and letting the water run through it. The puck is a bunch of straws, not a spoon :).
Brian Thayer said:
This may have been asked already, but has there been any move to create a glass basket?
Illy uses a see-through portafilter for its research, though there is speculation it is some kind of polycarbonate, rather than glass. There has been voluminous scientific research, with peer-reviewed papers published, on all the questions raised here. "Espresso Coffee, Second Edition: The Science of Quality," edited by Illy and Viani, is available on Amazon.com and is usually the best starting place to dive into this subject.
So much for inventing the wheel. I thought we were on to something...

Well, when I get that book, I will bet that it says the puck acts like a packed bed overall but has some fluidized bed properties early in the extraction process. And I bet that the friction equations in the .pdf Brady posted can be used to simulate the pathways of the water through the grounds in somebodies super-computer simulation of espresso extraction. I bet somebody could get a phd with a simulation like that and the resulting disertation.

I've been thinking about Jasons questions about why different coffees act differently in the basket. I am thinking that the roast profile used activates certain chemicals in the beans and gets them ready to extract in a very similar way to how barly malt is boiled at very specific temperatures and periods of time to get it ready to make a specific type of beer. The same malt can yield vastly different beers based on how it is boiled. Different chemicals are more ready to be pulled out than others. I would also think, then, that temperature profiling in an espresso machine might yield more interesting results than pressure profiling, since most of those flavor chemicals/oils directly interact with temperature while pressure is a secondary force that simply speeds up the process of getting the water in contact with the compounds to be extracted. Now if you could couple temp profiling with pressure profiling.... that might be very interesting.
Thanks, Marshall!

Marshall Fuss said:
Brian Thayer said:
This may have been asked already, but has there been any move to create a glass basket?
Illy uses a see-through portafilter for its research, though there is speculation it is some kind of polycarbonate, rather than glass. There has been voluminous scientific research, with peer-reviewed papers published, on all the questions raised here. "Espresso Coffee, Second Edition: The Science of Quality," edited by Illy and Viani, is available on Amazon.com and is usually the best starting place to dive into this subject.

Reply to Discussion

RSS

Barista Exchange Partners

Barista Exchange Friends

Keep Barista Exchange Free

Are you enjoying Barista Exchange? Is it helping you promote your business and helping you network in this great industry? Donate today to keep it free to all members. Supporters can join the "Supporters Group" with a donation. Thanks!

Clicky Web Analytics

© 2022   Created by Matt Milletto.   Powered by

Badges  |  Report an Issue  |  Terms of Service