Interchangeable Kitty Treat Puzzle

About:

Many years ago, I purchased a very simple version of this kind of treat puzzle made for small dogs. I figured my cats could probably manage it … which they did … a little too easily. I wanted to design a version that was: 

• tempting - they can see and smell the treats.
• doable - the puzzles all have textures/holes the cat can use to manipulate them.
• challenging - no towering covers that are very easy to bat away, or levers that are extremely easy to activate.
• hard to lose pieces - all of the covers are either attached or have holes to tether them to the base.

Let's face it, our kitty overlords are cunning, and we should show that we believe in their abilities.

The Model:

This model includes a base piece and two kinds of interchangeable puzzle tops.

The base has nine bowls in which to place treats and 4 integrated snaps to attach the interchangeable tops. This is one file.

Puzzle Version A has two print-in-place sliding covers, three insets for man-hole-style covers, three insets for hinge-style lids, and sockets for the base snaps. This version has four associated files:

1. a file with round covers that fit into the middle row on the top (these have beveled holes in which either TPU filament or knotted string can be connected and attached to similar holes in the puzzle top)
2. a file with square lids with a hinge component that you attach to hinge components on the puzzle top with regular filament.
3. tops with 44mm rails and 43.5mm print-in-place sliders:  
   1. one with tight sockets (see printing info below)
   2. one with looser sockets

Puzzle Version B has three channels and six print-in-place sliding covers with 0.25mm tolerances and sockets for the base snaps. 

In a  folder, there are also three files for customization: the top, a round cover, and a square lid.

In another folder, there are tops for both Puzzle A and Puzzle B with individual sliders that can be sized for your preferred tolerance and then combined in your modeling software.

A future version puzzle top will have threaded holes to accommodate individual pot-like interchangeable puzzles.

Printing & Assembly:

I highly recommend starting with the included test pieces to test which socket size works best for your printer/filament, whether the 0.25mm (43.5mm slider in a 44mm channel) works for you, and if the hinged lids fit okay. A big THANK YOU to @Nobody'sNightmare for helping me identify initial issues!

What I used: Anycubic Vyper, 75C heated bed, 80mm/s, 20% infill, with Sunlu white PLA.

TEST PIECES: Approximately 3.5 hours. were easy to print. Both the tight and loose sockets fit the snap, but the loose was noticeably easier to get on. The holes for the hinge, which are 2mm, did take a drill bit to open enough for the 1.75 filament, and Inhad to use an Xacto knife to cut through the elephant-foot that connected the bottom of the slider to the channel, but this may depend on how dialed-in your printer settings are (mine aren't).

Puzzle B Top:  Approximately 5 hours. After my initial print, I found that I wrongly calculated the tolerance, which worked but was sloppily loose. That initial file is still included but is in a marked folder. The recommended file, not in a folder, has a tolerance of 0.25mm. PRINT THE TEST FILE TO MAKE SURE THIS WORKS. In another folder, I posted a file with the sliders separate from the top with 0.1 tolerance so that they can be printed separately and popped in (since it seems possible to do so), OR there are individual sliders that you can size to achieve the tolerance that works best for you. You can manually combine them in your modeling software.

Base: Approximately 8 hours. I've been having adhesion issues with large, flat bottoms at 60 and 70C … and apparently now also at 75C) on my printer, and this was no exception. Unless you know for sure you don't have issues printing models with perfectly-flat bottoms, I would use some kind of adhesive aid to keep the first layer down. Hopefully, the test file will help you figure this out.

The uneven bottom affected how well the four snaps lined up with the sockets (the last one was a bear). Additionally, I noticed that the PLA is very stiff, and it's difficult to get the snaps and TIGHT sockets together, even when they're perfectly aligned. I ended up carefully, successfully, using a hammer - but that shouldn't be necessary.  Since I used the snaps and sockets tool on TinkerCAD, which gives an option of a tight fit - which I originally used - or a looser, “spinning" fit, I added tops with LOOSE fitting sockets. Again, use the test file to figure out which will work best for you.

Additional notes forthcoming …