RapidCoffee wrote: you have only one data point that shows mass decreasing at the end. Do you think the volume-mass ratio changes significantly throughout the extraction? I'd guess espresso weighs just about the same per unit volume as plain water

I haven't measured, but I believe the thin stuff at the end probably has a specific gravity somewhat greater than water. I expect the thick stuff at the beginning is a lot heavier than plain water (the effect of dissolved solids outweighing the effect of extracted oils).

RapidCoffee wrote:and would expect the mass flow to level off for a constant pump pressure.

Sorry, I'm not sure I understand what you're getting at. It's possible that if we let the extraction run and run, yes, the mass flow will level out, if the puck doesn't choke off the flow for some reason and the specific gravity of the extract approaches 1.

AndyS wrote: It would be great to superimpose the data from the flow meter on top of this graph (which Bill C could do, but not I).

You don't need a flowmeter Andy, you already have one...you know the volume of the pump...just measure displacement over time.

you could use a simple setup to get extremely precise flow measurements.


Sean

lennoncs wrote:You don't need a flowmeter Andy, you already have one...you know the volume of the pump...just measure displacement over time.
you could use a simple setup to get extremely precise flow measurements.

Uh, yeah, but on the GS3, it's just a rotary pump, not some fancy-shmancy piston.

AndyS wrote:I haven't measured, but I believe the thin stuff at the end probably has a specific gravity somewhat greater than water. I expect the thick stuff at the beginning is a lot heavier than plain water (the effect of dissolved solids outweighing the effect of extracted oils).
...
Sorry, I'm not sure I understand what you're getting at. It's possible that if we let the extraction run and run, yes, the mass flow will level out, if the puck doesn't choke off the flow for some reason and the specific gravity of the extract approaches 1.

Sorry if I wasn't clear. If espresso has a specific gravity significantly different from water, that answers everything. If they are similar, I figured the mass flow should level off as the puck saturates and the flow stabilizes. Um, just for kicks - could you weigh an ounce of espresso and an ounce of water? I'd do it myself, but my scale isn't accurate to less than a few pounds...

- John

I hadn't thought of the difference between mass and volume (and their respective .../dt flows). The two issues are the different masses of espresso versus water, and the absorption of the puck over time.

One ultrasimple experiment -- run the water for the shot out of a measuring cup and start writing. Second ultrasimple experiment, let the crema settle for an hour, then measure the density of the coffee from the first five seconds of flow and the last 5 seconds. No fancy equipment required, just old fashioned archimedian* science.

* They've finally cleaned up a palimpset with his writings underneath and some medieval drivel on top -- turns out he wasn't so old fashioned, but had gotten all the way to calculus by the time the romans killed him.

RapidCoffee wrote:Sorry if I wasn't clear. If espresso has a specific gravity significantly different from water, that answers everything. If they are similar, I figured the mass flow should level off as the puck saturates and the flow stabilizes. Um, just for kicks - could you weigh an ounce of espresso and an ounce of water? I'd do it myself, but my scale isn't accurate to less than a few pounds...

- John

John:

The espresso I made this morning measured about 10 brix with a refractometer. If it were a sugar solution, that would correlate to a specific gravity of about 1.04. But there's lots of emulsified oil (and other stuff) in there. The oils are lighter than water, so when I measured the specific gravity, it came out as only 1.02. That surprised me. I thought it would be higher, but I guess I was confusing viscosity with specific gravity.

I haven't tried to measure the brix/specific gravity of the early espresso vs the late espresso. I imagine, from tasting the stuff, that the difference is significant -- but nowhere near significant enough to cause the late dropoff on the graph.

In any case, you are correct, the specific gravity of espresso is fairly close to that of plain water.

Hi.
It is my first time I write in your site.
I ask if there is some news about the setting of this machine, and the possibility of comparison with a GS3.
I wish you a Happy New Year.
Regards,
Theodore,
Athens Greece.

AndyS wrote:Looking at the graph makes me think that a declining pressure profile (DPP) might keep the mass flow rate fairly even through the middle portion of the extraction. At the end, the puck is nearly spent, so there is less material being extracted and the mass flow drops. With a DPP the mass flow would drop even faster, but this is when you'd typically cut off the shot anyway.

I would like to underscore Jim's remarks about preinfusion because preinfusion is critical to the pressure profiling.

One thing that Illy talks about is the migration of fines and its relationship to flow. According to the experiments discussed in the Illy book, as fines migrate from the top of the puck, they collect downstream and slow the flow. Excess pressure can cause fines "dams" and a resulting flow restriction well before the puck has been "spent". As a corollary, in the lever machine pressure mod made by Alchemist, we see that by easing off on the lever, a dammed flow can be made to flow freely again.

There are three very different kinds of "resistance" offered by the puck:

1) The good resistance when preinfusion has caused the puck to swell
2) The bad resistance caused by dammed fines
3) The bad resistance from overtamping and/or too fine a grind

The latter two are actually inhibiting the extraction.

Yet to an eye trained on the the numbers and the graphs and plots alone, the three extractions might appear "to present" the same; the difference might not be apparent in the numbers. A brew pressure of, say, 8 bar when the resistance is caused by dammed fines or by too fine a grind|too dense a tamp is different from a brew pressure of 8 bar that results from a properly preinfused and swollen puck.

If the numbers included preinfusion duration and pressure, as well as some number representing the fineness of the grind|weight of the tamp, the data would be completer, but that could be a hassle. Elsewhere Jim has remarked about doing postmortems on pucks preinfused to varying degrees, and this would be good to bear in mind for these pressure profilings. The profiler should determine the threshold point where the puck has been thoroughly preinfused. A very gentle preinfusion pressure (too gentle for channeling to occur) and then looking for the first drops is a good seat-of-the-pants approach if you want to avoid the hassle of post-mortems. But gentle preinfusion pressure is paramount in that case.

Depth of roast and the bean itself also can have significant effects on the porosity of the puck. Since different beans have different makeups (e.g. robusta more solids than arabica has) would it be advisable to run these tests with an S.O. arabica bean instead of a blend?

Regards
Timo

P.S. Judging from the spring characteristics and piston diameters of those lever machines we have taken apart and measured here on H-B.com, they produce maximum brew pressures of 6 to 7 bar. I haven't had the pleasure of reading a photo essay on a heftier spring lever.