Parametric modular wireframe drawer box

Quick start

For a single mxn drawer box, the required parts are:

• one back part of dimensions mxn, with all four sides female ("ffff")

• m+1 side parts of size n, with both sides female ("ff")

• n+1 shelf parts of size m, with both sides female ("ff"), or optionally make the top shelf solid ("_ff_s")

Happy tiling !

Long intro

Why spend ten euros into a drawer box when you can make one in just a few tens of hours and a few spools of filament ?

This model tries to achieve this in a material sparing way, using a truss to support the drawers rather than a box with solid separators. This provides the best strength for a given weight, better than a single wall (vase mode) based system, as dimensions increase.

To make it printable in the most straightforward way without supports, the structure is split into three interlocking two-dimensional parts: back, sides, and shelves. And to allow arbitrarily sizes to be assembled, each set of parts is made extensible using tiling.

Tiling system

The tiling mechanism is very flexible: the parts for each plan (back, sides, and shelves) can be tiled independently, using a male/female connecting scheme. Shelve parts can have their left and right edge male or female, and the same holds for the top and bottom edge of side parts. The tiling for back parts is two-dimensional, and the four edges may be independently male or female, with 16 possibilities.

Tiling does not introduce any redundant beams. For example, the leftmost side of a shelve part is only present if it is female, and similarly for right side, and for back parts. Diagonal beams in sides and shelves also belong to only one side of a male/female connection. A complete box must therefore have its four sides female, so that every beam is present and connected.

The tiling for each plan is completely independent: for example it should be possible to cut the back of a box into three parts in some direction, the sides into four, and and shelves into two. But it is sensible to align the tiling, to assemble several smaller boxes in a modular way.

This results in a quite elaborate tenon system, where the different kinds of parts are also connected together in addition to tiling.

Customization

The model is done in OpenScad and is fully parametric. The following can be customized:

• number of drawers in each direction (per tile)

• male / female connections for each edge

• solid option for the top shelf

• dimensions of the drawers

• beam thickness

• tolerances

The provided parameters use 4mm beams which is roughly the minimum to allow the required space for tenons. The model relies on the 0.4mm nozzle to implement small tenons with 0.8mm minimum thickness.

I provide the files for a 4x3 box for twelve drawers of size 55x100x40mm as an example, which has only female connection. You should generate the files you need based of your dimensions and tiling plan.

Solid top shelf

After filling my full-sized 4x3 box with drawers, I realized that only the top drawers are not covered, so I added a solid top shelf option to make it look more finished and protect the contents from dust. The top shelf can easily be replaced on an already assembled drawer box. The new version (20260726) also fixes a minor bug in left-male shelves.

Drawers

The drawers are classical, they feature notches for a label, a handle, and a higher back side to stop them at the entrance of the slots, unless tilted upwards.

Building and assembly

Printing

I printed all parts with only one perimeter (0.4mm), 0.15mm layer height, three top/bottom layers, and 10% infill density (mostly to support the print). A few small parts should be printed first to adjust the tolerances.

I made vase mode drawers, but I think it only makes sense for the smallest sizes. Larger drawers should be printed in a regular way.

Connecting the parts

The shelves are connected to the sides and to the back, there is no direct connection between the back and the sides. Shelves and sides have to be connected first, before adding the back. I suggest connecting the first shelve with all sides for example, and then adding the next shelves. For very small sizes, some kind of tool can be useful to push the tenons into place if fingers cannot find their way.

Gluing is not normally not required. Tenons stay firmly in place if the tolerance are right, and gravity ensures that shelves rest on the sides and the back. The assembly could be suspended to a wall, attached by the (top) back part(s).

Results

I printed a first tiny version, then a more useful one where drawers have dimensions 55x100x40mm. A 4x3 box of this size has a width and height below 190mm, and can be printed without splitting. The box weights less than 80g (without drawers) for a total storage volume of 2.5 dm³, so the "density" is only 32g/dm³, or 0.032g/cm³. And this would be even lower if extending the box with more drawers.

Now, most of the weight is for the drawers themselves, whose weight is around 18g each, with a 0.8mm thickness, or 24g with 2.4mm thickness, 0.4mm shell and 10% infill.

Next

To make it even more flexible, my ideas for improvement are:

• Support omitting separators to allow a few larger drawers (multiple of the base x and z pitches)

• Parameterize tenons, to provide a larger version (two perimeters)

• Add a hanging option to mount the box on a wall

• Optimize drawers Done!

There are also probably a few remaining bugs, please let me know if you find some, or about any other feedback !

New improved vase mode drawer

In this part I describe a new optimized vase mode printed drawer, which achieves very decent stiffness, an order of magnitude better than a naive vase mode print for a "normal" size (see the comparison picture), while only using marginally more material. I believe that this makes 3d printing drawers for organizers a viable approach.

Context

For the contest I focused on the wireframe box, which was a real challenge by itself, and did not put much effort on the drawers. Since the drawers now account for most of the weight, if 3d printed (almost 3 times the frame's weight), I decided to address this.

The most straightforward way to print a sturdy drawer with a reasonable weight is to give its sides some thickness (for example I used 2.4mm in the previous model) and print it in regular mode with one perimeter and slight infill. Even with a modest thickness, this yields excellent stiffness and costs roughly the external walls. So, improving on this is about halving the weight, with vase mode.

Printing a cube shaped drawer in a naive way in vase mode results in a very flimsy output however, not really usable except for very small dimensions. But using a few tricks to strengthen the print yields dramatic improvement.

New design

The new design relies on vase mode and uses hacks to make it behave more like normal mode at some places. This is done by adding extremely thin notches to simulate some skin and infill printing. Using this method, I added thick beams on the drawer's top edges, which provides most of the added stiffness. An independent, modest improvement is adding triangular beading to the sides, which costs virtually nothing (except a slightly reduced inner volume).

I also designed a new handle shape, which is functional, much stronger and very well suited to vase mode printing, though it requires large enough dimensions.

The resulting print takes only 12g, exactly half the previous non-vase drawers for the same size (55x100x40)mm. I loaded this and a "naively" printed vase drawer (10g) with 700g of coins (actually only 630g in the improved drawer due to the reduced capacity). It's obvious on the picture how much stiffness the added 2g provide...