Soap Film and minimal surfaces

Hy,

I am on a mission to create as close approximation of a minimal surface (mesh), inside Grasshopper as possible. So far I have tested a few different approaches (Kangaroo 1,2 spring relaxation, soap film, Milipede and Minimal Surface Generator plugins). My goal is to create a cubic cell of a Schwartz D surface, because it can be divided into 6 congurent surfaces and thus tested.

Based on the first observations Soap Film gives a very good approximation, and Milipede can achieve it as well, with enough iterations (green and red).

However, I find it hard to understand the behaviour of the Soap Film relaxation. The results that I get are frequently unpredictable, required number of iterations is very high even with low polygon meshes, and most noticeably the resulting mesh is not symetrical, which is evident because of the centeral point of the mesh moving into arbitrary position.

I would be very glad to hear some more thoughts on the topic of minimal surface approximation and maybe some insights into the Soap film relaxation. Am I missing something? 

Thanks! Denis Hitrec