Turbulent forcing cast rounding typo and minor bug fixes.

[efhunt]

A couple of bug fixes which were missed in testing:

• Improper use of std::lround the static_case was found resulting in the wrong number of modes being calculated.
• kz was used where ky should have been resulting in a modified forcing value in non cubic domains.
• The necessary symmetry breaking modification to Lz in high aspect ratio domains was not implemented in the addTurbVelForces only in the setup.

I apologise for any inconvenience caused.