Guidance for real-time mesh deformation

Can you please point me towards C++ documentation, tutorials or samples for efficient real-time mesh deformation with integration to the physics engine. I would also be grateful for general background information about how the Unreal engine manages its meshes to ensure I play to its strengths.

I already have the engine for calculating the desired low level mesh deformations and need to work out how best to integrate it. The potential size of mesh is the limit of the given platform (huge) so I need to work out how best to apply the deformations to an Unreal mesh.

My questions:

  1. How to efficiently make a small edit to a large complex mesh?
  2. How to split a large complex mesh into two - do two new meshes need to be created or can we reuse the existing mesh?
  3. How costly is recalculating a collision mesh after a minor edit? Does it have to be recalculated completely or can we recalculate a section?
  4. On some systems, I already fake the huge mesh by breaking it into many smaller pieces to reduce the overhead of recreating sections so I won’t be surprised to need to do the same thing here. The question then would be how to ensure the multiple mesh sections operate as one unit within the physics engine?
  5. I am aware of Unreal’s destructible support but that appears to be edit-time calculated. I essentially need run-time destructible support where I can dictate the location of specific fractures. Can existing destructible infrastructure help with this?

Really wishing somebody answered your question because I am in the exact same situation as you. Why is there no C++ documentation or official tutorials. Like ■■■■. Shouldn’t the experienced programmers get the luxuries.