Pagoda Puzzle 51 Piece Combined and Scaled by Shmoee

Last crawled date: 2 years, 11 months ago

I scaled the original 51 piece "Japanese Crystal Puzzle" by NacAlain to 1/3 the original size in the three dimensions and I organized all the pieces into one stl to fit on a common Ender 3/Ender 5 sized print bed. (I've also seen it referred to as Pagoda and Burr puzzle as well)
I have now printed it (successfully the first time!). It took me about 7 hours at 0.28mm layer height at 80mm/s.
For me, all these scaled down pieces are just a little too wide for the channels that interlock the pieces together so each of my pieces needs about 30 seconds of filing with a mini-file to make the channels a little wider so they fit together (a mini-file is useful for making them a hair deeper as well so they can interlock flush). If I were to print it this combined stl again, I would scale the width (of the pieces) down by a percent or two and have a small file around to deepen a channel up as needed so everything fits flushly. After a few seconds of filing the channels wider and deeper, the assembly is still tight enough to hold itself together as it is being constructed which makes it easy to put together.
To reduce the amount of filing, I advise making the width of every piece a little narrower.
Narrowing the width should be easy to do in the slicer since all the parts in the combined stl are oriented in the same direction so you might make all the parts a little narrower by adjusting the width axis. The width of the channels would stay the same. That would result in for less post-processing (by filing of the channels wider (or the width thinner) I believe.
The three longest pieces that create the initial cross section I think could be improved a little as the tips of all three pieces have some sort of disconnect causing the tips to print as their own squares. They still appear to work though. I would make those starting three pieces about 1mm to 2mm longer as well so that the other pieces have a little more room as they stack up. If you scale down the width of all pieces by maybe 1%, the length of these longest three parts will probably be fine as they are as the pieces would crowd less as they stack up.
I found a video solution for a similar puzzle is here if interested which looks to be correct for this model (as I currently have it half assembled):
https://www.youtube.com/watch?v=nWlco82aSnY
If you have good bed adhesion (I used a Creality glass bed which held the parts securely). I used eSun PLA at 218C and a 65C bed.
It is probably a good idea to print a couple of the short pieces to test that the pieces fit into the channels properly. I didn't but that might have saved me a little filing. Regardless, the model as is can be assembled if you don't mind a little filing (assuming your printer performance is similar to mine).