AssafL wrote:Short PI leaves air in the puck. When pressure is applied the air compressed and puck density increases, slowing flow.

This is an interesting hypothesis but where's the evidence? You keep saying that air is like a spring and nobody is disputing that. But you are implying that ground roasted coffee is also like a spring. Are you saying that packed coffee grinds compress/decompress like air? Or that the grinds find a more ordered packing when compressed and then return to a less ordered state when the pressure is removed? It's hard to picture that.

How much of a density increase are you postulating anyway? Somewhere on youtube there's a video of a see-through portafilter. With 9 bar applied there is no noticeable change in the puck. If you're right shouldn't we be seeing the dry puck compress like a spring?

Looking back over some old transparent portafilter videos, the air above the puck seems to be completely displaced with water. E.g., Air may be trapped in the puck when there is channeling / uneven extraction. But that does not appear to be a factor in my tests. The pucks are completely wetted and the extractions are even, with no visual evidence of channeling.

In light of the above, I don't understand how the "springiness" of air explains espresso body and mouthfeel.

jpender wrote: This is an interesting hypothesis but where's the evidence? You keep saying that air is like a spring and nobody is disputing that. But you are implying that ground roasted coffee is also like a spring. Are you saying that packed coffee grinds compress/decompress like air? Or that the grinds find a more ordered packing when compressed and then return to a less ordered state when the pressure is removed? It's hard to picture that.

How much of a density increase are you postulating anyway? Somewhere on youtube there's a video of a see-through portafilter. With 9 bar applied there is no noticeable change in the puck. If you're right shouldn't we be seeing the dry puck compress like a spring?

Well obviously neither water (liquid phase) nor coffee grinds (solid) compress much. So the only volume that can compress down is air between grinds - and air (maybe CO2) trapped inside grinds (trapped during the roast process.

So when the top of the puck gets wet and presses down the air in the puck under the wet layer (ignoring the air above the puck which just compresses using the ideal gas law) presses on the air which is either free to escape the basket filter holes - or compress down in volume (if trapped).

Obviously, the thicker the layer permeated by water during the Pi phase the less compression that will happen (air is displaced by water during PI and water is incompressible). A good pour video will PI enough time show flow so you won't see the phenomenon.

I don't know how the grinds settle when the air is driven out or compresses. Obviously if the voids between grinds are smaller, less water can flow and density increases. At some fine grinds the puck will seal. The only solution would be very slow PI. This actually works well.

I think this is proven since it is fully reversible. Even if you were to compress the puck enough to seal it - just back off the pressure to allow it to spring back - and PI will continue. Wait for drops to form and follow through with the regular pour.

RapidCoffee wrote:Looking back over some old transparent portafilter videos, the air above the puck seems to be completely displaced with water. E.g.,

Air may be trapped in the puck when there is channeling / uneven extraction. But that does not appear to be a factor in my tests. The pucks are completely wetted and the extractions are even, with no visual evidence of channeling.

In light of the above, I don't understand how the "springiness" of air explains espresso body and mouthfeel.

It doesn't. It explains why flow is faster with PI.

But flow does affect body. If it wouldn't than higher EY would have higher mouthfeel and it doesn't. My guess here is that is emulsification which requires agitation (except micro emulsification). That agitation may happen more with flow rate and turbulence.

It may but I have no idea how to test and validate the hypothesis.

AssafL wrote:I don't know how the grinds settle when the air is driven out or compresses. Obviously if the voids between grinds are smaller, less water can flow and density increases. At some fine grinds the puck will seal.

If the voids are smaller? I thought that was your premise, that the density increases. Ballpark figure for the void in a tamped puck is roughly 20% of the total volume (my estimate). So how much would a dry puck have to shrink under pressure to "seal" it?

AssafL wrote:I think this is proven since it is fully reversible. Even if you were to compress the puck enough to seal it - just back off the pressure to allow it to spring back - and PI will continue.

I don't think the reversibility is proof at all. It fits your hypothesis but doesn't provide any evidence.

jpender wrote:If the voids are smaller? I thought that was your premise, that the density increases. Ballpark figure for the void in a tamped puck is roughly 20% of the total volume (my estimate). So how much would a dry puck have to shrink under pressure to "seal" it?

The voids are smaller means density is higher. The Maor's air in the puck the lower the density.

As to how much? I don't know. Obviously very coarse grinds will have too many voids and even when packed as tight as possible will not seal at 9 bar.

Finer grinds will seal at 9 bar. That means that the delta density is enough to seal it. Remember that seal isn't water tight. It has to slow down so much that you give up. With the special LM firmware I can do 120 seconds. So maybe there will be drips at 130? Who knows.

Compress it with a hydraulic press under a few tons and you'll get (almost) a solid. No flow whatsoever.

I don't think the reversibility is proof at all. It fits your hypothesis but doesn't provide any evidence.

That is philosophy. In science we don't prove anything. We disprove hypothesis.

So take fines migration. If the fines seal the puck - would reducing the pressure redistribute them in the puck? Unlikely. For a 3-way you could hypothesize that the pressure drop breaks the seal. That is why you should test with a lever or myopically controlled gear pump.

Or take a coffee butter or hydrocolloid hypothesis. Both will seal the puck due to a layer of butter or a layer of dough. Well - reducing pressure won't affect them. So the seal would be permanent. You'd have to discard the puck.

Air springiness (compliance) isn't new. We use it in acoustic suspension loudspeakers for years.

Is your hypothesis that air in the dry part of the puck (premature PI) doesn't compress under the ideal gas theorem? Do you expect the volume of air to stay?

AssafL wrote:As to how much? I don't know. Obviously very coarse grinds will have too many voids and even when packed as tight as possible will not seal at 9 bar.

You're saying you don't know what actually happens. So how do you even know they shrink at all? It isn't observable in that youtube video.

AssafL wrote:That is philosophy. In science we don't prove anything. We disprove hypothesis...

...Is your hypothesis that air in the dry part of the puck (premature PI) doesn't compress under the ideal gas theorem? Do you expect the volume of air to stay?

A hypothesis is just a proposed idea. It isn't more than that until you have evidence. All I said was that you have indirect evidence that doesn't contradict your hypothesis but no direct evidence in support of it.

I don't have a hypothesis. Ignoring air escaping out the bottom, I would expect the air to compress if the puck volume decreased. But does it decrease? The only direct evidence we have is that video and I can't see any change. Can you?

I don't even know where to start. In the case of hypothesis you are confusing mathematical proof with the hypothesis in the scientific method. No relation between the two. Is the drift equation for excess electrons okay for a semi conductor? Or do you add corrections for newer subatomic corrections? Depends on your need. But the current hypothesis will never be proven and at some point add corrections. Unlike math.

As for "compress" all I can say is yes it compresses. If you ever used a tamper at 30# you'd see the initial pile is severely compressed by tamp end. At 9 bar and a 58mm basket you are at 400#. Why would it not compress more?

The settling of the grinds obviously is more noticeable at the lower end (first few #) and it tapers off until you end up with a solid puck (at which point it won't compress anymore). You may surmise that a sealed puck is a solid. Which I don't believe. All it needs is to slow the water enough to block for 60 seconds or so.

The hypothesis we raised about air accounts for and predicts all observations we are aware of. The best way to disprove it is to (easily) find one it doesn't.

AssafL wrote:The hypothesis we raised about air accounts for and predicts all observations we are aware of. The best way to disprove it is to (easily) find one it doesn't.

Apologies, but I'm finding your posts hard to follow. Are you now referring to higher flow rate with PI? If so, please explain how compressibility of air "accounts for and predicts" this.

AssafL wrote:As for "compress" all I can say is yes it compresses. If you ever used a tamper at 30# you'd see the initial pile is severely compressed by tamp end. At 9 bar and a 58mm basket you are at 400#. Why would it not compress more?

The transparent portafilter video shows no obvious puck compression when water is introduced at 9 bar pressure.

AssafL wrote:So take fines migration.

No, let's not. IMHO fines migration is an urban myth, a meme-worthy catchphrase with no supporting evidence. It was first proposed by Petracco in Illy's classic Espresso Coffee book (Ch 7 Percolation), based on the observation that "flow increases when the percolation chamber is inverted". I have the utmost respect for Petracco, but clearly there are other possible mechanisms. The only experimental evidence I can find argues against fines migration.

BTW, one of these days it would be nice to get back to espresso body and mouthfeel. Just sayin'...

RapidCoffee wrote:Apologies, but I'm finding your posts hard to follow. Are you now referring to higher flow rate with PI? If so, please explain how compressibility of air "accounts for and predicts" this.

The transparent portafilter video shows no obvious puck compression when water is introduced at 9 bar pressure.

It predicts the cases where water doesn't flow or flows exceedingly slow. It doesn't mean it doesn't happen with coarser ground coffee - except that it probably won't fully seal. I use this quite a bit if I grind too coarse - just brickwall the response by going high pressure as fast as you can (PI -> 0).

None of the movies above show the pressure gauge going up to 9 bar before the puck is fully wetted. At that point water can't compress and thus the puck (whose voids are now filled with water) can't compress and PI is fully completed. That is why he was able to grind finer and get good flow.

No, let's not. IMHO fines migration is an urban myth, with a meme-worthy catchphrase with no supporting evidence. It was first proposed by Petracco in Illy's classic Espresso Coffee book (Ch 7 Percolation), based on the observation that "flow increases when the percolation chamber is inverted". I have the utmost respect for Petracco, but clearly there are other possible mechanisms. The only experimental evidence I can find argues against fines migration.

these were still hypothesis that were raised here not ten years ago...

BTW, one of these days it would be nice to get back to espresso body and mouthfeel. Just sayin'...

Sorry about that - this is a rehash of the discussion in the long PI thread. It answered the flow part of the discussion. So we know how to explain the flow - we have a good model (at least I am confident in it until something better comes along).

To look at your tests I'd need a model for the body, texture and flavonoid parts. For now, texture of coffee (mouthfeel) is pretty obtuse. What I think we know (in a nutshell) is this:

TDS and body obviously go together. So extracting more (higher EY) for a given volume leads to higher TDS.

So this is either a non uniformity issue (some % of coffee in the puck evades water and remains under extracted) or a efficiency issue (some flow dynamics extract more - like turbulent flow etc.) - or a mix of both. This is from the 55% or so extractable from coffee (industrial yield) - currently at home I heard of 25-27% EY.

To add to this - there are hydrocolloids involved (think a flour thickener) and emulsions (think the texture of mayonnaise vs constituents oil and vinegar and emulsifier).

The latter (emulsification) is where Illy (figure 8.3 - pg. 294 Droplet size distribution) marks the highest difference between high body and low body espresso. At the low size of droplets the difference is 10 fold.

The question is why some machines and grinder patterns cause more emulsification I don't have any solid idea. But will say that when you make mayonnaise and aiolis and emulsified Ice Creams (like chocolate) the agitation is important. It is a violent act to take the lipids and disperse them into the continuous phase (an emulsion being the droplets of the dispersed phase locked in the continuous phase by the surface tension and helped by the emulsifier which has a hydrophilic and hydrophobic side).

Finding the energy source for the emulsification is easy: The 9 Bar obviously powers the "beating". But what causes it? I think flow does. That is why the most crema comes at a rather narrow flow rate and may decrease even when EY may increase (grind finer and pull really slow). As I think more about it - CO2 emission under pressure maybe helps. Fresh coffee has more crema from CO2 bubbles - does it have more body? Add to this the suggestion that was made that it may be a micro emulsion and we will be deep in the weeds (albeit I don't think it is).

The former - I would assume - is a matter of temp - especially for long chain polysaccharides carbs ("fiber") these need temps - some as high as 90C to hydrate - so that can play a part too. Go under the hydration temp and the coffee will be thinner. I have no proof of this but Petracco does talk about polysaccharides so I have my suspicions. Fueling that suspicion is the fact that coffee seems to lose body when cold. That is the behavior of a certain class of hydrocolloids we know well (CMC) which is a form of Cellulose. Cellulose doesn't degrade much in the roasting process (https://www.researchgate.net/publicatio ... ffee_beans) so I await to see if it is a texturizer in espresso. If so extraction and serving temp would be important for body.

Tying all of this together into a model isn't easy but my hunch is that it will be required in order to answer your questions...