Efficiency/affordability.
Wind tunnel testing is expensive.
Comprehensive testing impractical.
Computational fluid dynamics (CFD) is useful.
Adapting grids difficult and time consuming.
Solution: do not mesh the vortex generating device, model its effects via source terms (body forces).
Easy optimisation.

BAY method needs modification for use on this scale.
Immersed boundary methods are better to model VGs directly, but harder to implement in general.
Vortex velocity-profile-based approaches are more versatile and a new version has been formulated to ensure a smooth solution.
Explicit calculation method needs modification for use in complex flows.
Asymmetrical vortices.
Iterative approach for change in boundary conditions.
Penalty and direct forcing work set velocities well, but provide less information about the vortex generation.