Долгосрочное прогнозирование миграции загрязняющих веществ в природных дисперсных средах с применением технологий параллельных вычислений

Abstract

The modeling of nonisothermal heat and moisture transfer in natural disperse environment is considered. It is necessary to apply parallel computing techniques for solving these problems in three-dimensional setting. For this purpose was developed an algorithm of parallel computing, which is based on the creation of distributed data array by solving a system of nonlinear equations with the use of finite element method. Comparative studies of the computing model performance of contaminant transport by the proposed algorithm and the standard instruction, showed a significant reduction in computation time using the developed algorithm.