Correct calculation of beam space charge in Baxevanis solver

[AngelFP]

The current way of calculating the azimuthal magnetic field had a problem where it did not take into account the whole charge of the beam, especially when the number of radial grid points along the bunch was small.

This PR fixes this issue as well as other related problems. These include:

• Correct calculation of b_theta.
• The use of Ruyten shapes for charge deposition has been reactivated and added as option.
• Several bug fixes in the implementation of the Ruyten shapes.
• The deposition methods now force all the charge to be deposited within the grid boundaries. This avoids having less charge in the edges of the grid, particularly when using 'cubic' particle shapes.
• Fix small bug in the calculation of z_max in the deposition method.