After lurking about I figure I should add something to HB.
-I ground fresh homeroasted coffee this morning on my Vario which was set to produce a gloppy 30sec pour that came out quite nice in the naked PF.
-I used a LV-950V2 Laser Particle Size Analyzer using DI water as a medium and using the refractive index of cellulose (1.63) and an imaginary index of 1 (its opaque). The analyzer uses both a red and blue laser to span the range from 10nm to 3mm. It is brand new and top of the line and it is freshly calibrated.
-I ran several times to make sure it was stable for both volume-weighted and number-weighted distributions. I avoided using sonication so as not to reduce the size falsely. I also did not use a dispersant. Below is the combined volume and number distribution.



-Volume
Median Size 155.51776(µm)
Mean Size 212.34103(µm)
Variance 39846(µm2)
Std.Dev. 199.6157(µm)
Mode Size 366.9342(µm)
D10 22.73704(µm)
D90 494.35190(µm)
D(v,0.1) 22.73704(µm)
D(v,0.5) 155.51776(µm)
D(v,0.9) 494.35190(µm)


-Number
Median Size 1.31575(µm)
Mean Size 1.56061(µm)
Variance 8.2187E-1(µm2)
Std.Dev. 0.9066(µm)
Mode Size 1.1369(µm)
D10 1.05540(µm)
D90 2.17708(µm)
D(v,0.1) 23.57645(µm)
D(v,0.5) 167.32407(µm)
D(v,0.9) 503.15186(µm)
-as you can see, the number dist emphasizes the fines and masks the bimodal nature (there are relatively few big particles but they make a large volumetric proportion of the powder.
-FESEM images are to come soon, this should be enlightening because I will not have to metal coat the powder like in other studies.

If anyone wants more information about PSA (particle size analysis) I'm happy to share. I've had too much schooling on it!

An ok first post, I guess.

Have you tried measuring any other grinders?

Do you mean to say E(environmental - uses gas to get rid of the charging)SEM. An E-SEM can be an FE-SEM, I suppose.

Thanks!

The FE part is Field Emission, which lets you work at very low voltage (0.1kV) which helps when looking at nonconducting samples. It is also a variable pressure scope, where ionized gas can help dissipate charge, but I dont tend to use it that way (10^6 torr usually). Environmental scopes are another animal but do use low vac (higher pressure) to look at wet samples.

Interesting. The fines peak appears to be quite large for the Vario, compared to analyses of other espresso grinders.

I continue to be frustrated by these experiments using different methods on different coffees, and thereby not producing as much insight as they could.

I have this suggestion for anyone with repeated access to a laser counter: please ask us before experimenting. We will send calibrated grind samples (e.g. a grind sample of say 8 day old Black Cat which produces a 30 gram shot from 17.5 grams of coffee in 30 seconds using an LM basket -- the machine will be a variable, but this should be close) of just about every espresso grinder you can dream of. Then doing the counts with the same method on near identically pulling grind samples will produce results we can meaningfully compare.

If there's a significant cost involved in doing the measures, I'm sure we can pass a hat to defray it, since the topic is so interesting

I don't mean to frustrate anyone with data, but I believe a qualitative assessment can be made based on the shapes of the curves rather than any particular grain size data. Fortunately, I do have unlimited and free access to the LV-950V2 as well as an entire materials characterization lab. The holy grail may be the BET surface area analyzer. Standardization is a way of life in science but I am a bit reticent to receive powder in plastic bags through the mail....so maybe you can get a big hat to put your grinders in then mail them to me (and a standard GS3)!

Anyway, I just got off the FESEM and looked at the same coffee from yesterday.







I characterize powder on a daily basis and coffee is very unique, its cellular structure is way more complicated than the ceramic and metals I normally deal with. SEM tells us little other than to correlate some shape factor to validate the PSA. A flake structure like uniform coins will be bimodal because there is an edge on view and a face view that will produce two peaks. I dont think this is an issue for coffee because they seem to be equiaxed. My conclusion is that the foam like nature is an overarching factor in bean fracture. I feel, grinder design is at best, like herding cats, maybe one design is a little more effective (idiosyncratic is a better word, maybe) but will lead to only vague, qualitative differences.

Awesome work Kendall! Thanks much for doing this. Wish I could send my k10 for some more SEM pics!

-j

Very cool. Thanks.

I would volunteer to test the K10 and the Tanzania. It is impossible to get the same flavors at work, where I use different grinders. The K10 shines at espresso grind, meanwhile the Tanzania shines at large grind settings (Chemex, Clever Dripper). The grinders are heavy and would need palettes in order to be shipped. Do you think the test could be done in one single Saturday? If so, I wouldn't mind the driving to NJ

Cheers

Great pictures. The bimodality of coffee grind distributions, according to the Illy book on espresso, is due to the fracture of cell walls. A coarse particle has several intact cells, fine particles are like bread crumbs, the fragmented cell walls produced by successive breakages of he bean. I think the picture at 500 magnification shows this really well, with one coarse particle surrounded by several fines.

Rafael, my idea was that you dial in the grinders, grind the coffee and ship grind samples, not grinders. For espresso, we can produce functionally identical samples by dialing in the same coffee to get the same flow in the same basket. For brew grinders, we'd need to do something analogous using a cone filter.

In the previous tests, large conicals produced a more dispersed coarse grind distribution, which might explain why they perform more consistently But that is as far as a qualitative assessment went. The ability to overlay several curves that produced identical shots would be more revealing in how the quantities of fines, coarse particles and their dispersion affect the percolation rate.

Come to think of it; we still don't really know what changing the grind setting does. If you take the Vario, and make identical flow rate shots with a 12, 15 and 18 gram dose (or some spread like that) and compare the grinds, it would tell us a lot about how the change in flow rate is created. Is it more fines, finer coarse particles, both?

Kendell, thank you much for these posts.

another_jim wrote:I continue to be frustrated by these experiments using different methods on different coffees, and thereby not producing as much insight as they could.

+1
The TGP graphs I have seen on this site have little resolution in the x axis. The Mahlkonig graphs we have seen have a detailed x axis, but ambiguous y axis designations.

The above graphs are very nicely detailed. Jims offers are generous. It would be great if capacity and willingness exists to do some comparisons - there is great interest here.

For purposes of entertainment and illustration, a little story:
While waiting in a courthouse with my SO on a traffic matter, she asked me to bring in her purse from the car while she waited for her name to be called. Seemed like an innocuous request.

I started becoming self-conscious when I found myself walking back to the courthouse with a womans purse. I put this feeling aside with the intent of brazening it out. Then there was the security line.

"Is this your purse?" asked the guard.
"Certainly not. My purse is much nicer than this one."
I smiled. He did not.
Out of the purse came a baggie of off-white powder. "What's this?" he said.
"Uh..." I said.

After cooling my heels for a time in a holding room, it was revealed that the bag contained chickpea flour. I was sent on my way. Please don't ask why she had chickpea flour in her purse.

kmills wrote: Standardization is a way of life in science but I am a bit reticent to receive powder in plastic bags through the mail....

Why? Bringing off-white powder thru security lines - that's a problem. I'm not sure I understand what the problem would be with bags of ground coffee from Jim Schulman.

kmills wrote:so maybe you can get a big hat to put your grinders in then mail them to me (and a standard GS3)!

Sounds tough. Mailing some form of standardized samples seems a lot easier.

Not trying to be pushy here, just sayin...