More Unique Models - with Parallel Processing

The FIDAP 8.5 release provides a host of new problem-solving capabilities and solver enhancements, including parallel processing to speed your calculations and reduce turnaround time. This version's fully-parallelized segregated solver is based on a mesh partitioning/message passing scheme and runs on Windows NT and most UNIX machines. All aspects of the solution have been parallelized, including input, element pass, solution, and output. Take advantage of your multiprocessor machines to gain faster solutions!

A number of enhancements have been made to FIDAP's free surface modeling capability. The segregated solution algorithm for melt inter-face modeling has been improved to ensure faster and more reliable convergence for crystal growth applications. The solute segregation model has been formalized in the user interface, allowing segregation of multiple species and other capabilities. The new elasticity-based remeshing scheme can be applied to all free surface problems, providing more flexibility than spines or mapped mesh approaches.

In FIDAP 8.5, the filling (VOF) model can now be combined with prescribed boundary motion (deforming mesh). This enables the simulation of various mixing, forming, and pressing operations. Marangoni stress and heat/mass transfer at the interface are now included. There are also new features for solving the electrohydrodynamics problems that frequently arise in the biomedical and micro- electromechanical systems (MEMS) fields.

Finally, FIDAP 8.5 introduces the capability to model fluid structure interaction (FSI) problems, in which the stress and deflection in solid regions must be computed and coupled with the fluid flow solution. With FIDAP 8.5 and FSI, a whole new range of problems can be analyzed, such as deforming arteries and grafts, stress and deflection in extrusion dies, deformable roll and blade coating, and numerous thermoelasticity applications (see article, page 10).

Typical parallel processing performance results for various system architectures. 3D laminar and turbulent test problems were used with meshes from 25,000 to 100,000 nodes.

FluentNEWS