Bug with linear solver?

Now the velocity value is checked for possible nan.

Need to fix the tests! This will be done when adjoint is ok

Tested only for phi influence (same results as dart). The rest is implemented but untested until the bug for dirichlet bc is fixed.

Removed unecessary surface in surface quads mesh of 3D rae2822 (rae_3_visc.geo)

Merged agardTest branch

Validate 2D transonic capabilities.

Driver now computes the average position of the transition

DDicretization now throws a runtime error if the mesh size is < 2 elements.

VTK now supported + pipeline fixed for 3D cases.

Back to blaster docker image and remove MKL

Dart interface is now able to write the surface pressure distribution.

Adding scipy before script

Added test case on the rae2822 3D finite wing thattests the 3D functionnalities of the code with vtk writing