Re: MMU2 feasibility torture test ?
Thanks a lot for your feedback. Yes, you are right, the stl are not totally clean and contain holes / swapped faces etc.
Commercial software that process X-ray microtomography image stacks were used to produce these mesh files (such as Avizo/Amira). And such mesh files usually convey a lot of topological inconsistencies, and need to be cleaned before being 3D printed. I tried to clean them as much as possible with Meshlab, but apparently it is not enough. Would you have recommendations to clean mesh files before feeding Prusa Slic3r?
Also, there is another potential issue: the "connective_material" stl file and the 3 others overlap somehow in space. Could it be the reason why Slic3r can't produce the gcode?
A few ideas. The main goal is to end up with just exterior surfaces, rather than interior voids that the CT scan will pick up.
For an object like this, I tend to
Project point clouds on a grid, rotate, Project point clouds, rotate, etc, etc, until the entire exterior is covered in point clouds. Then rebuild the mesh from the combined clouds. I'd do that in CloudCompare or 3DReshaper
Otherwise, you can also do a lot with filtering based on ambient occlusion. (Meshlab or Cloudcompare).
Re: MMU2 feasibility torture test ?
Thanks a lot for your suggestions, I will give it a try!