September 21, 2025
Description
A card game box design that creates boxes very similar to the deck boxes that cards are usually purchased with. Instead of making heavy use of the Z axis like most 3D prints, it is designed to be glued and folded after printing.
Demo and assembly at https://youtu.be/s4gXLKx4FPg
A very different design from my previous 3D printed flatpack card box in that this tries to emulate the design of traditional cardboard flatpack deck boxes as much as possible.
There are several types of boxes and designs. The smaller boxes fit within the largest box type to form a 2-player game kit that can hold 2 standard deck boxes and additional small accessories.
All designs expect a 0.20mm layer height.
All models should already be oriented correctly, so that the correct side is lying on the print bed.
There are 3 different box types, with different design options.
CardBoxPlainHexWindowOrnateDoubleDeckSheetStdWideTabAccessDoubleDeckSheetInnerPlainToplessFor a full set, print 2 deck boxes, one double box, and one inner box. The card boxes should match the size of a commercial MTG card box (which I assume is a standardized size that works with other collectible cards). This means the 3D printed card boxes can replace cardboard ones. And also, cardboard boxes can fit in the double boxes.
Tested with PLA and PETG. Both work fine. In theory, you'd want the most ductile plastic for the living hinges to last the most bending cycles. So PETG should be preferred over PLA, and Nylon should be preferred over PETG. But in practice, PLA has worked fine for me, and I haven't had a living hinge break on me yet, so it's durable enough to be used for a while. This observation also applies to silk filaments. Also, I've never printed with Nylon, that's just "in theory" ramblings.
When the print is finished, make sure that the printed plastic is cooler than its glass transition temperature. This is the temperature at which the material is no longer molten, but will take on permanent physical changes if deformed. If the plastic is too hot when peeled off the bed, the sheet can have a curl from peeling it permanently in its structure. You can let it cool by waiting, maybe with a fan - or if it's on a bedsheet, lift the sheet and fan it for a short while.
The printed sheet is designed to bend one way, where the side facing the bed bends inward (an origami mountain fold).
Instructions may not be the most clearly written, but I'm assuming too much detail isn't needed for this (essentially, glue the tab, and fold everything else 90 degrees). See project video for details.
For any given box type:
When printing the top of the sheet, which is the part of the box that's visible from the outside, different fill patterns can be applied in the slicer. For patterns that create ringed patterns, such as Concentric, Archimedean Chord, and Octagram Spiral, their surface finish can create streaking highlight effects when printed in silk filament.
Holographic transfer sheets can kind of be used. The surface finish for the outside of the box isn't on the bed, so the hologram will be transferred to the inside.
Be careful if you use super glue, eye safety, accidental adhesion, and whatnot.
Make 100% sure adhesives are dry before putting cards into the boxes.
If you have a print that has stubborn adhesion to the bed, be careful picking up corners with your fingernail. It's rare, but not impossible for the sharp-and-thin feature to slip into the nail bed.
Source Blender files are included. The main file is CardBoxSheet_*.blend, which is a non-specific box, but most closely resembles the card deck box. The "Params" object has Custom Properties that can be modified, which drive shapekey changes. Although the mesh (mainly for the tabs) gets corrupted if the box is too small.
The parameters don't have logical zero defaults. I was trying to make the zero-ed parameters the size of the deck box, but in practice, the exact measurements come out a little different when assembled. So I printed a default box, and then modified the parameters based on how different the values were from the target inner box dimensions I wanted. A very trial-and-error process. After test prints were verified that the dimensions were dialed in, a duplicate of the Blender object was made, the modifiers and shapekeys were applied (destructive process), and extra features were modeled in. This was done for each box type. Not elegant, but it got me results.
License:
Creative Commons - Attribution