Bits to atoms with the 3D printer, and atoms to bits with the 3D scanner
Design a 3D model that couldn’t be fabricated using machines we’ve already covered
I’ve never really printed anything much in a 3D printer, so this is all new territory for me. I figure, as this is a test print, and I’ve never done this before, the model itself isn’t that important.
So I used the built-in object types in Fusion 360 to create a coil about 8cm ling, and 3cm diameter. I thought this might have a gentle enough slope that I could print it without support, and if successful, I could get a feel for the mechanical properties of the print: strength, flexibility, springiness, etc.
Then I exported the STL file and opened it in Cura to prepare it for printing.
I used a Creality 3D printer. For the first print, I used a 0.6mm nozzle, and printed with PLA.
Layer height: 0.3mm This is about half the diameter of the nozzle. Gives a good speed/quality trade off. A smaller layer height gives higher quality – 0.1mm is lower limit – but slow speeds can be inconveniently slow without much quality difference - maybe even worse as you’re depositing so little material.
Wall thickness: 1.8mm The width of my part is a consistent 6.5mm, and the nozzle is 0.6mm.
Infill: 0% My walls are thick, so I need very litle infill. Usually you’d use more like 20-30%, and if you were going to put the part under pressure or fit a screw in, you’d make this 100%.
Printing temperature: 200°C On the material packaging, they suggest 195-230°C
Build plate: 70°C So that the model will stick to the glass bed. I turned the printer on so the bed could heat to the right temperature before printing.
Diameter: 1.75mm Check the actual material with calipers
Flow: 100% Print speed: 60mm/s Travel speed: 100mm/s Acceleration (Print/travel: 400/4000mm/s2 Cooling: enabled
Support: none Even though this model has overhangs, I want to test printing it without support.
Build plate adhesion: Brim 8mm A ’skirt’ just goes around the part, not in contact. As this model is quite top-heavy, brim seems like a more supportive option.
I set up the machine again with a smaller 0.4mm nozzle, and edited a few of the settings in Cura:
Smaller nozzle size: 0.4mm Lower print temperature: 190°C Larger brim: 12mm
This attempt faired even worse.
A few weeks later, I tried again, with a smaller model (to save time). And to minimise movement while printing, I taped down the brim with masking tape after the first layer had been printed. This worked better, though there was still some movement part-way thrugh printing:
I also tried downloading a model from Thingiverse, that is closer to the sort of thing I might actually need to print for my project. In this case, an Arduino enclosure with mating parts:
This needs some work to make it viable - the parts don’t currently mate:
3D scan an object
I’ve had a small amount of time to play with 3D scanning techniques. Previously I’ve tried using photogrammetry to create 3D models from photos.
Here’s a bike I found on the street – in retrospect, this was an incredibly challenging thing to scan - with so many voids!
I took 50 photos, and used Agisoft Photoscan to stitch together the images into a 3D model.
I also tried using the iPad-mounted Occipital Structure scanner to try and scan an object in the lab: a set of ratchet crimpers. This is my first attempt. I haven’t had time yet to try varying the scan technique to get a better result, or play more with the software.
“Ready, fire, aim!”
OK, so now I’ve had a chance to look at the instructions on the Structure website, so I can try again…
I tried scanning something simpler and larger in the lab – a footstool: