Hey team I’ve done a few iterations of the tool.
Upon pressing it into the basket the grinds are creating so much friction that the last 5mm doesn’t want to slice any more. It’s completely stopped by the compressed grinds at the bottom.
We may have to use razor blades, which is (technically) a much bigger endeavour. I’ll give the fully 3d printed model one more crack. Here’s hoping!
Maybe try to stick with the original design with only 4 pieces (instead of 8 or even 16). The less blades, the less friction.
You could also think of a second design to only measure inner/outer density, where the blades are fixed on top of the tool, so there are only two pieces. If you combine the results of both measurements, it’s like you have a tool with 8 pieces, but much less friction.
Also, even with razor blades, the problem remains: You’re trying to add the volume of the blade to an already very compressed puck. There’s no space for it. Maybe you need a basket, that can expand in some way. Or somehow “fix” the puck, take it out, then cut it.
Edit: Or use very thin needles instead of blades.
Could this then be approached from the other way, turn the puck out and then cut it? An apple wedge cutter style device could be an answer. 
Matt,
While trying this method of a tool for direct measurement, possibly Archimedes one of our greatest early engineers might have an indirect approach 
Check with Hendon to see if he has a contact that can help with this experiment, this is a practical engineering problem that is of the same structure that Archimedes ran into, and we do this all the time. How to measure characteristics of an object with as little direct manipulation as possible.
Ultrasound? A magical pressure pad that records the pressure footprint of a puck? Maybe weight volume displacement of partially submerged puck in fluid at various orientations? Of course extraction of puck and then “sealing” an issue. Just spitballing here, just have the MIT grad students work on it (otherwise known as peasants of the tenured).
Ideally, would have a robust mechanism for measuring density variance in a puck and display as a 3d model. This would complement the more practical “tool” (direct, or indirect measurement) which would be a coarser approximation. Mass produce the tool, then would be a method of QA. Problem has been solved as such at a Macro scale, given petroleum engineering discipline. But small scale? Dunno.
Of course… MIT has become IP protective in some ways… find another University.
Thanks for the brain food everyone!
Funnily enough, I was doing some o-ring and gasket research just now and came across this magic stuff!!!
http://www.sensorprod.com/pressurex-micro-green.php?app=Heat%20Sink
We’d be able to put some of it on top of (or below) the coffee before we tamp. Tamp it. Then see whether the coffee had hot spots or not.
Any thoughts?!
(update: emailed a local supplier and will try to pick up samples on Monday next week 
)
Something interesting on Insta: https://instagram.com/p/BYgTQ7wj_x6/
Need the right psi film, Fuji?
Sensitivities To Accommodate A Wide Range Of Pressures
Film Type Pressure Range
Extreme Low 7.2 - 28 PSI (0.5 - 2 kg/cm²)
Can’t really determine the resolution, don’t know size of ink dot from info. Oh well, try it and see what happens.
Don’t nutate, or tap with hand, or anything. Have the worst possible distribution. Then apply the film and tamp. See what you get.
Then do the usual manual approach, then tamp. Compare.
For such a small area not sure the resolution will be meaningful, but definitely worth a try.
I would assume parts of the puck that have air pockets would have a smaller relative pressure recorded.
Non helpful results would be same color between two tests. Or if the different colors, and then the last is a solid color.
How did you think about the insta?
It looks good but since I’ve learnt the particle size distribution map,
I won’t measure something like this by only seeing or touching it any more.
So the pressure film arrived!!
But it was sent to 9 Yarra Street South Yarra. Not 12-18 Yarra Place South Melbourne as per the address label.
Here’s the kicker - that’s the address of a friend of mine from high school who just rented an office there whom I haven’t spoken to in years.
The universe doesn’t want this test to happen!
Hey Matt!! just chiming in. Have you ever heard about the Toyota Production System? It’s pretty cool. The Lean methodology really can give you a lot of insight to this problem.
I dare to say it without reading through all comments, but one good way of working through any problem is to use the five whys. Like, why am I needing to distribute my coffee? Well, asking that on each subsequent answer to the original question you can usually get to th root cause of the problem. I’m sure you know it as it’s quiet a mainstream thing.
Solve the root cause and you don’t need to rely so heavily on distribution (a “band-aid fix” to the problem) or at least you’ll arrive at something more feasible in an operating environment… Like I mean to say, sort of control the output of the grinder so that a palm tap is sufficient for final distribution.
Put the problem to the grinder manufacturers as if these super automatic machines are going to be any good as you say (and we all know), they probably already are nearing an answer. or maybe they don’t want to give away their IP just yet
Keep up the great work!
Hey gang
The pressure film is a bust 
It just doesn’t have the resolution we need for this kind of analysis. We did have fun cutting circles and sliding them into a basket for the puqpress to devour!
Meanwhile, I’ve arranged to have the old tool 3D printed in solid stainless steel. It’ll have razor thin 1mm walls and 0.2mm blades, and be built like a tank. If this can’t do it, I don’t have many other options.
Apart from being super cool, Direct Metal Laser Sintering is pretty expensive and will cost $400AUD (320USD). That said it would’ve been much more as cnc (I also don’t think cnc could’ve made the walls that thin).
ONWARDS!!!
Yeah I wish grinder manufacturers operated on something even remotely similar to Toyota’s system. They would’ve fixed it by now!
Re superautos it’ll be way better than what we can do with manipulating grinds manually. Can’t wait to reveal!
Hi Matt
Is it fair to say that the Vlog#3 is well overdue 
i have been home roasting and espresso extracting for over 20 yrs. needless to say, i have gone thru a number of roasters, grinders and espresso machines. my current set up is an hg-1 grinder, a expobar brewtus IV machine and a behmor 1600 plus roaster. i buy all my green beans from sweet marias. i have used a bottomless portafilter for the last year and a half. i added the ONA ocd2 to my workflow three days ago–prior to the ona i was using one of those knock offs from amazon. i have pulled nine shots with the ona and everyone of them has had absolutely no channeling and beautiful extraction right from the center of the filter. this is amazing to me–miraculous even.
Good information thanks for sharing
Fieldengineer