Paint Box (with MAGNETS!)

A little 3D printing project, making inserts for a small plastic case to hold small paint pots used for model painting. Also an experiment with embedding magnets in the 3D print.

The paint pots being held are 14ml Humbrol enamel paints. I mainly use these paints for colouring the dots on my custom die, or other very-tiny paint jobs.

While I usually print in ABS (mainly because I acquired a huge number of old rolls of the stuff second-hand), I am using PLA this time as it is a bit more resistant to the kind of paint-thinner-like chemicals it is going to be around in this usage. PLA is (supposedly) a bit easier to print with, and is also plant-oil-based rather than petrolium-based as most other plastics are.

Closed box.

Starting with the same type of plastic box as I used for my custom die case, I re-used the basic 3D model of the top and bottom inserts I had developed for that. I have several of these cases bought cheap in bulk years ago which I am gradually finding uses for. Apparently they are actually fishing-tackle boxes and are still readily available from various online stores - there are two similar-looking sizes, I am using the larger one here. They come with compatmentalised trays glued in, which I remove by placing them in near-boiling water for a few minutes to soften the glue, then carefully pry out with a small spatula.

Box open with paints and brush.

I had space for 10 paint pots, and a bit of extra space that (almost - I had to trim the handles down 20mm) fit two of the tiny brushes I use, so I made slots for those too, rather than completely waste the space.

The Bottom tray is to hold the paint pots in place, and - when closed - the top section clamps down on them just firmly enough to (hopefully) stop any transport-accidents with not-quite-on lids.

Box with top tray out.

I also had about 8mm of spare space in the lid of the box. I had hoped to put extra brush-storage in here, but it wasn't quite deep enough. Still not wanting to waste the space, I instead made a rectangular recess to hold a laminated safety-data-sheet (SDS) for the paints.

To achieve this, I split the top insert into two parts, a thin one that would be glued into the top of the box lid, and a thicker part that will be removable.

Rather than have the removable insert falling out of the lid every time I opened the the box, I embedded strong magnets in both top and botom pieces. I left cavities the size of the magnets in the 3D prints and paused the print at the top of these cavities, super-gluing in the magnets before resuming the print. Super-glue was necessary to hold the magnets in place, otherwise they would jump off the print and stick to the print-head rails as they got close! I printed two layers - about 0.6mm - over the tops of the magnets to make it look nice and neat. I smeared and finger-tacked some superglue across the tops of the magnets too, prior to restarting the print, so the layers printed over them would have a rough surface to better stick to.

So now the removable panel stays in the lid, but is easy enough to pull out if one wants to get to the SDS.



You can download the Blender 3 and STL files I used here.

These files are provided as-is and without warranty or any restrictions on usage or distribution. You will probably need the blender file to generate new STL files, since you will need to resize the magnet cavities to match whatever magnets you have - leave an extra 0.5mm space on sides (not top and bottom) to account for side-expansion in the 3D print process.


Some things I learned.

The manufacturer's 3D slicing software is no longer optimised for my old printer (I have a MkI and the software is presumably optimised for the current Mk3 model these days). I had to slow it down to half-speed, from the driver defaults, to stop the stepper motors over-running themselves and printing at a slant.

There is also a bug (probably only with the manufacturer's slicing software's interaction with my MkI) that was causing layer drift on the 'top' layers of the print (6 layers, 2mm, thick in this case as I wanted a bit more strength on these faces of the inserts). This was an issue even at half-speed printing. Backing off to quarter speed for these specific layers fixed this. Over subsequent prints, I gradually bumped back up to one-third speed without issue and decided that was good enough!

The long strait sides of the print are a bit finicky and have a bad habit of corner-lifting off the print bed during early stages of printing, ruining the later layers. Online suggests a 'thin smear' of glue stick. I found that a 'thin smear' was not enough, but layering it on thick as lip-gloss worked effectively for providing good bed-adhesion. Long strait edges are apparently a bit of an issue for this type of 3D-print technology, as I didn't have any trouble of this nature with some similarly-sized organically-shaped PLA prints I did for a student just prior.

Initially, I slightly under-sized the magnet cavities with the intention of using a soldering iron to heat the magnets and press them into the holes. Heating the magnets above the softening temperature of the plastic de-magnetised them (not a huge surprise - heating iron de-magnitises it, so the same obviously applies to neodymium magnets - I just wasn't sure at what temperature the effect would kick in). I re-sized the cavities and super-glued the magnets in instead. Neither the lower levels of heat coming up from the heated print bed (through several layers of printed plastic, too), or the spot-heat around the in-print plastic as it comes from the print nozzle, seem to have been high enough to damage the magnets, in the latter case, the heat probably soaks into the metal of the magnet, and 'averages down' quickly enough to avoid any such issues except, possibly, on the top fraction-of-a-milimeter of the magnet, if even that.

The magnets are still visible through the 0.6mm (2 layers) of plastic covering them. Burying them deeper would have reduced their magnetic attraction (further appart) and as these magnets are normally concealed, I don't really care. Using a darker plastic colour would likely have also concealed them better (if that actually mattered to me in this instance). I specifically did these inserts in white to contrast with the paint pot colours.

The two-layer over-print on the magnets does, however, ensure the surfaces are properly smooth (and easy to lightly sandpaper even smoother), which having the magnets flush with the surface would have made more difficult. Plus I would have had to use removable supports, and the extra clean-up that entails, in one of the trays if I had done it this way.

If I wasn't doing this as a learning exercise (and manually embedding magnets mid-print, in this case), it would likely have been cheaper, probably have been quicker, and definately have been less frustrating to send my 3D files off to a commercial 3D printer. Though through the multiple (twelve!) day-long print failures I had to go through to get satisfatory prints, I did also have the opportunity to make unrelated small adjustments to the model that became apparent from examining the mis-print parts: just minor tweaks, and nothing I couldn't have lived without, but small improvements nonetheless. So for iterative prototyping, obviously a cheap 3D printer can still be a very useful tool.


upBack to Glenn's Project page.