Finding the proper box for my sensors is not easy. But even after finding one that mostly worked, I wanted to explore more sensors and still have a bit of freedom while doing that. Of course, that means 3D printing with the new 3D printer.
There is just one problem …. 3D printed boxes take a long time to 3D print.
Here is the newest sensor module I designed, which can fit quite a few small sensor modules on it while still being compact. The module can take a motion sensor, a light sensor and a temperature sensor while being compact. So the 3D printed box (the simple kind to design) needs a few extra holes for the motion sensor, airflow for the temperature sensor and a hole for the light sensor. The result is below:
Both parts of the box take 1h 50 minutes (110 minutes) to print, using generously low infill and a 0.3 mm layer height. For the final one deployed I would print it at finer settings, probably reaching close to double print time. But make 10-20 of those and that is many many many hours of printer noise and attention. Not feasible.
Can it be better?
There are still a couple of things that can be done to make the print faster:
- Print faster: at double speed (100mm/s) it takes 70 minutes
- Use a wider line width: double (0.8mm) will get me to 70 minutes
- Make the box walls thinner (1mm instead of 2mm)
- Combining 1 & 2 gets to 46 minutes
- Combining 1, 2 & 3 gets down to 35 minutes.
- Combining 2 & 3 gets down to 50 minutes.
But these combinations result in extruding at up to 24 mm³/s which is beyond the extruder limitation of about 15 mm³/s. The result has dropped in quality a lot and has some issues that might need a few iterations to fix. Even if we are looking at 2-3 times faster printing, i would not want to use this box. Surely, there may be a good compromise somewhere in between, but that is not why I am here.
Vase mode (Spiral outer contour)
What if i could change the printing technique? The vase mode or spiralled outer contour is a technique where the printer only prints one outer contour in a single spiral.
Unfortunately that is not feasible for this type of box. Also known as vase mode, this method needs shapes that are feasible to be printed in a spiral.
What if the box could be spiralized?
There are some things that work normally (the bottom holes) and some that don’t work as they are now:
- holes in the side of the box
- vertical holes for the screws
So i tried to design the box in a way that the functionality is there as well (air circulation holes, hole for light sensor) but make it printable in the vase mode.
How to design a vase printable box
I used FreeCAD for this and started from the initial box design.
The outline printable box is below:
The prints in vase mode are exactly one line with wide. Line width is normally chosen the same as nozzle size (0.4 mm in this case), but it can be made larger. How large depends on the particularities of the printer, but for mine everything up to 0.8 mm (2x) is fine. 1.0 mm is doable, but there is a bit of loss in quality. But since I am looking for speed here, that will do.
With 1.0 mm line width, 0.3 mm height and 50 mm/s print speed, the printing volume is at the limit of 15 mm³/s so there is no point in trying to print faster.
The bottom hole
Is designed as before. However, the thickness of the bottom plate will now be determined by the Layer height and Bottom layers number in Cura, not the actual design. This is how vase mode works.
Side wall holes for ventilation
Ventilation holes are needed so the temperature sensor gets ample air circulation so it has a proper response to the actual air temperature (there is probably no circuit heating due to the low duty cycle). So how to change the outline so it results in a hole?
By adding a pocket in the outer well, we can create an outer contour that looks different at different layers and we’ll create holes in the final box. How deep should it be? About Line width + hole depth which should be a 1mm minimum.
Side wall hole for light sensor
Pretty much same logic as the holes for ventilation, but I will cut the inside of the contour with a knife to make a hole for the light.
The box walls
Are very thick, which gives Cura enough material so it properly picks the outline contour. I found that the thickness should be at least 2*Line width + hole depth. I used 4 mm for this design.
Snap fit connections
To fix the box and lid together I used some basic snap fit design. On the box side, a simple extrusion along the long side of the box will create a dimple in the outline contour. I went for 0.6 mm high and 0.5 mm extrusion, but these measurements need to be fine-tuned.
The resulting box is pretty good. Much better than the normal style box with the printer pushed to the limit:
There are definitely some aesthetics to be improved. The 1 mm thick box is also a bit transparent, but would not be in matte white/black (Unfortunately, this has been my draft printing filament and it just about to finish). Also not very sturdy, but that is ok for such sensors. However, the box does do its job and was quick to print.
The pockets are creating proper ventilation holes:
The hole for the light sensor is made by snipping the inside pocket with a diagonal cutter
OK, how fast?
Printed together, the box and lid require 18 minutes to print which is a 6 fold reduction in time. Make sure to use One at a time setting in Cura.
Obviously, there is less material in this box. In fact, by weight (initial is 19.0 g and final is 7.7 g) the box is 2.5 times less material but prints 6X faster, so the technique still offers a 2.5 X improvement in speed per mass.
This is a viable technique to print good project boxes much faster using the vase mode, if the compromises are acceptable: thinner walls and a different design procedure.
There are clearly things that can be improved:
- The box-lid arrangement is not so nice
- The box design has to be made with the intention to print like this from the start (I am looking for ways so that the same design can be printed both normally and in spiral mode and come out properly).
- this technique with a larger nozzle will allow for thicker walls
- it might be possible to trick the slicer into thinking the inside is still an outer contour and practically double the wall thickness.
But overall, for 6 times reduction in print time while producing a viable box in 18 minutes, I’ll take it.