Effective conductivity for DFN

[jbrezmorf]

Goals in 2d:

• determine mean conductivity tensor
• determine variance of log of main components
• determine variogram (using random sub squares of fixed size

TODO:

• fracture generation in 2d:
  • [x] von Mises
  • [ ] general basic shapes (disc, square, line)
  • [ ] more flexible composition of distributions (no shape rotation, no anisotropy, basic shape can specify defaults)
• [ ] 1d in 2d mesh generation
• [ ] connected fracture set in 2d
• [ ] setup flow problem (two pressure BC gradients)
• [ ] scale number of fractures to at least scale range (L/100, L) where L is domain size
• [ ] MLMC script, use MC to get mean and variances
• [ ] MLMC script for getting variogram:
  • [ ] cut a subsquare from a full fractured square
    • [ ] cut and solve a lot of these sub problems for different sizes and positions of sub squares
  • [ ] empirical variogram for a fixed size
• [ ] determine is the variogram is a power law and if the power coeff much the power law of fractures