FF: Cartesian grid for AWAE and WD outputs

[ebranlard]

This pull request is ready to be merged.

Feature or improvement description
This pull requests introduces Cartesian grids for wake planes, in preparation for the curled wake model implementation.

• Code revisions that will slightly change the results:
  • AWAE takes as input wake deficits on a Cartesian grid instead of a polar grid, to accommodate future
  • WakeDynamics (WD) outputs wakes on a Cartesian grid.
  • For the baseline axisymmetric/polar wake model, Wake Dynamics converts wake deficits on the the polar grid onto the Cartesian grid
• New features not currently covered by the r-test:
  • WD introduces an option Mod_Wake=3, where wakes are computed on a Cartesian grid directly. The wake dynamics equation are solved using a First-Order Foward Euler scheme, that only accounts for momentum conservation and not mass conservation (similar to the Curled-wake model implementation).
• The r-tests results are slightly affected (see below). New baselines are produced. Since an update is needed, we use this opportunity to reduce the simulation time and storage needed by these tests cases, while keeping them "fairly physical":
  • shorter simulation times
  • slightly closer turbines when needed
  • Change f_c
  • less output channels (e.g. removing the channels that are 0 because wake does reach given diameter downstream)

NOTE: these changes were reviewed by @jjonkman several months ago, prior to the implementation of the curled wake, as part of #774 and an old branch f/ff-split

Impacted areas of the software
FAST.Farm, WD+AWAE

Test results, if applicable

Mostly affects Vx and Vr at some distance downstream (e.g. 3D) due to slightly different convection (AWAE receiving slightly different wakes).

LESinflow test case:

TSinflow tests case: