Scalar and parallel optimized implementation of the direct simulation Monte Carlo method (Q1924698)

The implementation of the direct simulation Monte Carlo (DSMC) method is studied especially for simulation of rarefied gas flows in which molecular effects become important. The development of a large system named MONACO is described. A localized data structure is used. It is based on a new DSMC algorithm formulated for workstation architectures. Parallel use of processors in possible. A cell is the basic unit of the algorithm rather than a particle which enables the use of unstructured data. Separation of geometric, physical, and organizational aspects of the DSMC algorithm is attained. Calculation of many different flow problems without changing the code is possible. As examples simulation of a diverging channel flow, simulation of a reentry planetary probe, and simulation of a neutral contactor are carried out. Execution of MONACO on 400 nodes of an IBM SP2 computer allows the use of 100 million particles.