Skip to content

Latest commit

 

History

History

README.md

Radially symmetric potential flow in spherical coordinates: analytical solution vs dispersive solver

This problem is just like ../potential_solution but uses spherical coordinates with a domain centred at 45 degrees north. The domain is very small, so a local Cartesian approximation is accurate.

This provides an opportunity to check the dispersive solver in spherical coordinates.

Validation approach

The numerical model results are compared against an analytical potential flow solution. Two test cases are run:

  1. Single grid with OpenMP - Tests the dispersive solver on a uniform grid
  2. Nested grid with MPI - Tests the dispersive solver with grid nesting and domain decomposition

Single grid validation (OpenMP)

Transect comparison at final time

Numerical vs analytical solution transects

Comparison of numerical and analytical solutions along east-west and north-south transects at the final time. The SWALS dispersive solver closely matches the potential wave theory solution.

Free surface comparison

Numerical vs analytical free surface

Comparison of the full water surface elevation between the numerical model (left) and analytical solution (right) at the final time.

Nested grid validation (MPI)

Transect comparison at final time

Numerical vs analytical solution transects (nested)

Comparison of numerical and analytical solutions along east-west and north-south transects for the nested grid case. Results from both the coarse and fine grids are shown, demonstrating accurate solution across the nested domain boundary.

Free surface comparison

Numerical vs analytical free surface (nested)

Comparison of the full water surface elevation for the nested grid configuration, showing the numerical model (left) and the analytical solution (right). A dotted line shows the high-resolution priority domain region.