This is the old version - here is the new: https://makerworld.com/en/models/209267 

Shaking coffee grounds is fun and, with practice, makes delicious espresso. Doing the whole thing with a 3D printed tool doesn't make much sense at first. 
 

But I have optimized this shaker from scratch with new ideas especially for 3D printing and after the update I solved the two biggest problems (coffee getting stuck and filling into the portafilter). 

• It is also designed to enhance donut extraction (that's experimental), more on that later.

Fluffy coffee grounds, no additional funnel necessary, no mess. The donut espresso shaker fulfills all these requirements, one solution just to pick up from the printer and one to sand after printing.

(Scroll through the images to see examples after use)

Please read the "How to print" and "How to reduce sticking to a minimum" part carefully to achieve this result (the video still shows the old version, the first GIF shows the new one but it need some time to load, it's not a picture :D).

Also RDT helps a lot (especially with the no sanding solution), however the GIF is without RDT. (Do RDT in another cup or clean the shaker after the RDT process.)

The versions with handles fit into holders for portafilters.

Versions for smaller grinders like the Varia VS3?

In the folder "Versions for smaller grinders" you will find smaller versions, but these are usable with a maximum of 20g (well usable up to about 16g).

You can see a short demonstration with the old version (with the sticking problems) here (new video coming soon):

Why would you want a donut extraction? In Stéphan Ribes paper "Radial uniformity of espresso extractions" from 2020, he found that even with a donut extraction, the sides of the espresso puck are under-extracted. Also nicely illustrated here again by Lance Hedrick. 

I am currently testing whether the radial extraction yield can be improved for the Flair Pro 2 and later for other sizes.

I've managed to get some great shots with it so far. 

Note: With use, I had to change the grinder settings for my recipes

There is also a subreddit specifically for 3D printed coffee tools where test results of this tool will be discussed in the future. We are happy about everyone interested and about like-minded people who would also like to share their experiences!: https://www.reddit.com/r/3Dprintingcoffee/ 
 

How to print:

Food safe PETG or PLA (With PETG, less usually sticks)

• Use the finest layer height your printer can afford relative to the time you want to spend printing (I used 0.12 mm layer height, tolerances are optimized for that but there are also more tolerance versions)
• No support
• Enable Ironing for lid and “bolt flat for ironing” to reduce sticking problems to a minimum without sanding
  • Use Rectilinear Top surface pattern if you enable Rectilinear ironing
• Check that it is positioned as shown in the picture
• 100% Infill for the bolts
  • Watch out by removing the bolt from the print plate. Wait until the plate has cooled down and use a scraper. (The bolts are quite stable after removal from the plate, as you can see in the video but you should not forget the leverage effect when removing it).

Feel free too experiment with the settings but here are mine for a „high quality“, less sticky result (enable advanced settings):

For all parts:

• 0.12 mm layer heights (high quality preset if you have one)
• Perimeters / Wall loops: 2
• Top shell layers: 5
• Bottom shell layers: 5
• (In the future or with OrcaSlicer PR-3839: Enable scarf joint to minimize seam visibility)

Funnel:

• 15 % Gyroid Infill
• Top surface pattern: Concentric
• Bottom surface pattern: Concentric
• Ironing: Top surfaces, Concentric (This is just for aesthetic)
• Fuzzy Skin: Outside Walls (better Contour if your slicer already support it). This is to get this finish which you see on my pictures but if you disable it you should get a faster printing time.
  • Fuzzy Skin settings (enable advanced mode):
    • Point distance 0.05 mm
    • Thickness: 0.1 mm

Lid: 

• I used a textured instead of a smooth print plate.
• 15 % Gyroid Infill
• Top surface pattern: Rectilinear
• Bottom surface pattern: Archimedean Chords (You don’t get a seam with this setting)
• Ironing: Top surfaces, Rectilinear (This reduces sticking to a minimum)
• Fuzzy Skin: Outside Walls (better Contour if your slicer already support it).
  • Fuzzy Skin settings (enable advanced mode):
    • Point distance 0.05 mm
    • Thickness: 0.1 mm

Bolt:

• 4mm and 6mm bolt available (choose depending on your filament strength)
• 100% Infill (Rectilinear) (To get that heavy „metal“ feeling and a stable one)
• Don’t forget to disable fuzzy skin for this!
• Only for “bolt flat for ironing” (This bolt reduce sticking to a minimum without sanding, however the other bolts are more fun but have to be sanded)
  • Ironing: Top surfaces,  Rectilinear (change Top surface pattern also to Rectilinear for this)
• Watch out by removing the bolt from the print plate. Wait until the plate has cooled down and use a scraper. (The bolts are quite stable after removal from the plate, as you can see in the video but you should not forget the leverage effect when removing it).

How to reduce sticking to a minimum:

If you have the old version and you don't want to sand I recommend to reprint the original bolt and use ironing (still available in the “versions for smaller grinders” folder, they were/are the same bolts for both sizes).

No additional work methods:

• Use RDT
• Use the ironing feature at least for the lid (see “How to print” section)
• Use a bolt for ironing: (“bolt (flat) for ironing”) and the ironing feature (see “How to print” section)
  • Fitting problem with this bolt? Then scale it 1%-3% down

Sanding methods:

• The shaker is easy to sand due to the angles, even with normal files
• First sand coarse (I used a file) and then fine (I used this sandpaper like thing from Bosch). What I sanded is marked on the picture with green:

• You can also use food safe epoxy after that, but I didn't test it yet.

You like what I do?: https://ko-fi.com/dubios 

Tests and experiments:

Future experiment planned to test the extraction yield radial distribution.

How much coffee remains on the walls (Old Version)?

I first put all the parts together, put it on the scale, waited 10 seconds and read off the value. After use, I put all the parts together again, put them back on the scale and waited 10 seconds. Then, I read the value. After, I used a brush to spread everything into a dosing cup and weighed it again.

Here is the result for PLA-Metal like, 0.12 mm layer height and 19g beans:

Clean container: 65.2g, container after use: 65.3g, coffee residue "0g" (according to the manufacturer, the scale starts at 0.1g, but that doesn't seem to have worked here).

I also thought the result must be more but I think the blue color is very unflattering. When I painted the residue into the dosing cup, it stuck to the wall of the dosing cup and I had to paint it down first to make it look like the picture.

Feel free to contribute!