I realise this might be quite a complex thing to accomplish and that it might ideally be done with some scripting - but perhaps there are already solutions that can simplify this task, or just simply good ideas... so please bear with me, and let me know if you have any ideas regarding this:
I am working on a spaceship design and I intend to cover most of the ship with tiles similar to those used on the Space Shuttle orbiters.
I might be facing two challenges:
- Generating tens of thousands of relatively irregular tiles as dynamically as possible as individual objects - based on the mesh of my ship. The concept you see in the picture of the orbiter (above) is pretty much what I want: Larger segments made up of mostly rectangular tiles meeting other segments at odd angles. Having them as individual objects is important because the animation requires for most or all of them to come off individually.
- While being individual objects, the tiles need to share a painted texture as they should show decolouring of larger areas and streaks caused by reentry as seen on the orbiter in the image.
I obviously don't expect anyone to solve my problems for me, but any input, hints, pointers etc. from more experienced users are very welcome!
The current mesh is made up of individual metal sheets that are sculpted into shape as seen below.
My first guess would be to model another mesh of the ship where every tile is represented by one polygon. I think that would generally be an okay approach despite the significant time it will probably take. More procedural/dynamic ways of doing this would be appreciated!
But, assuming the above approach:
- I'd assume there is a tool to turn each face into an individual object - are there any things to consider for maintaining the orientation of the texture I will have painted on the entire mesh before separating?
- Every tile will need to be rounded in accordance with the ship. That could be accomplished with a subdivide before separating (not afterwards), but then I'd separate my mesh into groups of faces rather than individual faces. Does anybody have ideas on how to accelerate/automate that? Or perhaps much better approaches to begin with?