Needed a mannequin head to take pics of a mist-generating mask in development. No problem, just find and download an STL file, 3D print it and done, right ?
Well... took me 2 weeks and 2 kilos of PLA to finally get one right 😅
So if it helps somebody else save time, I documented below my stumbles got-it-right-the-first-time-prowess, and uploaded the files to Github: F3D Fusion360 CAD, STEP 3D models and 3MF slicer files for a Bambu Lab P1S.
Feel free to copy / remix / whatever my files, while respecting lehthanis's model's license terms.
Happy modeling and printing to all !
______________________________________________
Model
Many human head models out there, surely several were going to match my need for a head stand: proportions, ears that a breathing mask can hook to, bare skull, female, not behind a paywall. But nope, was surprised to find only... one !
Source model: from lehthanis on Thingiverse It is a medium polycount model. Small facets can still be seen but, even then, Fusion360 sometimes maddeningly crawls. So, good enough.
A few modifications were needed though, like a more stable base, more clearance between the chin and the table, and adjusted dimensions to better match an adult's head size. So the model was imported into Fusion360 and the saga began 😅
Design for 3D Printing
To get a smooth surface a high density infill is needed since most of the surface at the top of the head is way into overhang territory. Also, the outside skin needs internal force-dispersing structures, otherwise dropping the head would easily make the surface cave in.
But the internal volume is so large that tons of time and filament would have been wasted just on the infill: over 11 hours and 600g... Or ~3mm thick walls could have used, with no infill, but that necessitated about a Kilo worth of filament. Not on my watch, am a cheap impatient b*5t*rd 😂
Using Adaptive Cubic would have helped on time but with poor quality as it increases density only near horizontal surfaces, not on the sides. Here it is at 96% (!) infill density:
Apparenty slicers don't have the concept of 'high infill density only near external surfaces/walls' ?
So, instead, I hollowed out the head. Which took a lot of work as the Fusion360 Shell command fails, likely due to folds in the source mesh. That was worked around by creating a scaled down version of the skull, trimming it further in all sorts of ways, and finally substracting it which effectively hollowed out the model.
Now very high density infill can be used between the inside and outside walls. The resulting print is bomber: it can be dropped without fearing large dings.
The head was initially printed in its natural orientation, which required a lot of support under the chin, and invariably led to ripping the skin in that area when removing the supports. Could have used PETG as the interface filament for PLA, but that would have blown up the print time to 40 hours â˜
And anyway, the head was still a few percents smaller than an adult's and any scaling was going over the printer's 250mm max height.
Now, rotating it ~40° back, adding a wedge to avoid bed adhesion issues, and printing the base separately, solved both the scaling and the support issues:
Finally, using modifiers to locally thin out walls and lower infill density where acceptable, shaping the inside top of head as a 60° cone to avoid supports (pic above), and a few other tricks decreased the printing time to 14 hours. Yay !
Despite the manually painted small support areas, there are still more dings and scars left than I expected. But it's all at the back of the head, so time to move on with my life... 😅
The base holding the head is round and has internal ribs to force the slicer to generate walls inside its volume. This makes it much stronger and faster to print than using a ton of grid infill. The initial rib-less base cracked easily when inadvertently pushing sideway on the head.
Final model:
Result
Time to glue the two together:
Total print time: 16.5 hours
Total filament: 630 grams
Main settings for the head (details in the 3MF files):
- 0.4mm nozzle, 0.16 layer height, PLA
- 3 walls, 6 top, 2 bottom, concentric pattern
- 10% rectilinear infill
- Reduced speeds from Bambu's default: 100mm/s outter wall, 200 inner, 200 sparse infill, 200 solid infill, 60 top surface
- Painted supports, default tree style
- Back of the head facing the Aux fan that blows across the bed, to reduce sagging on the steepest overhangs
Almost 2 weeks to get there, only to find out that styrofoam heads are sold on ebay for $10 😂
Anyway, not sure what I missed. Expected to 'just 3D print an existing STL', seeing all the 'just worked' comments on most head models out there. I.e. am a clueless noob !
Comments
Post a Comment