unstructured mesh generation for geophysical flow simulations


The primary benefit of the unstructured grid over a conventional structured grid lies in its ability to efficiently discretize arbitrary geometries. Smooth and continuous transition from high resolution where needed to low resolution elsewhere can be provided via user-specified adaptation criteria. Figure shows a regional OMEGA grid centered at Washington D.C. with a radius of 10 degrees. The adaptation criteria for mesh generation, was set to refinement around land-water boundaries.



The left panel shows the OMEGA grid structure and the coordinate system. OMEGA grids are unstructured in horizontal and structured in the vertical. The OMEGA model can also simulate global/synoptic scale flows (right panel). The global mesh in the right panel also shows the domain decomposition of the mesh used (via METIS libraries) for parallel computations.



Grid for Mars (right) using MOLA data (left)



Antartica - adaptation to coastline

solution-adaptive mesh generation



Virtual potential temperature isosurface (355K). The virtual potential temperature is a good indicator of the hurricane core. The storm is viewed from the east-southeast and from an altitude of about 15 km. Mountains in Central America can be seen at the upper left corner. Dynamic grid adaptation was used in the simulation. The adaptation criteria, was set to pressure perturbation minima, i.e., to follow the eye of the hurricane.

details

Ahmad, N., D. Bacon, Z. Boybeyi, T. Dunn, S. Gopalakrishnan, M. Hall, P. Lee, D. Mays, A. Sarma, M. Turner, and T. Wait, 2002: Unstructured Adaptive Grid Generation for Geophysical Flow Simulations. Numerical Grid Generation in Computational Field Simulation. Proceedings of the 8th Intl. Conf. held at Honolulu, Hawaii. pp. 457-466. Edited by B. K. Soni, et al.

Ahmad, N., D. Bacon, Z. Boybeyi, T. Dunn, M. Hall, D. Mays, A. Sarma and M. Turner, 1998: A Solution-Adaptive Grid Generation Scheme for Atmospheric Flow Simulations. Numerical Grid Generation in Computational Field Simulation. Proceedings of the 6th Intl. Conf. held at Greenwich, England. pp. 327-335. Edited by M. Cross, et al.

mesh generation links

Professor Rainald Lohner

Mesh Generation on the Web

Meshing Research Corner