|
This article is cited in 46 scientific papers (total in 47 papers)
Predictive modelling of coastal hydrophysical processes in a multiprocession system based on explicit schemes
A. I. Sukhinova, A. E. Chistyakova, A. V. Shishenyab, E. F. Timofeevac a Don State Technical University, Rostov-on-Don
b Open Joint Stock Company "SiVizhnLab", Taganrog
c North-Caucasus Federal University
Abstract:
The paper has been devoted to comparison of computational efficiency of implicit and explicit regularized schemes for modeling of two hydrophysics problems — suspension transportation and storm surge, which may be represented as systems of nonlinear diffusion-convection equations. Regularized schemes have been applied for numerical solution of 2 problems. One of them is bottom surface reformation resulting from sedimentation of suspension on the bottom of the water body at ground dumping; the second one is the task of modeling storm surge in Taganrog gulf of the Azov Sea on 24–25 September 2014, when under on influence of hurricane wind with velocity 40 m/sec during 30 hours, water level raised in comparison of normal level near 420 cm. Computational experiments demonstrate time expenditure advantage of the explicit scheme over the implicit one. The explicit regularized scheme requires in 71.5 times less time processing in comparison of implicit one at 512 cores of supercomputing system of Southern Federal University in Taganrog at computational grid with 5001$\times$5001$\times$101 nodes.
Keywords:
coastal systems, suspension transportation, storm surge, implicit and explicit regularized schemes, systems of diffusion-convection equations.
Received: 14.03.2017
Citation:
A. I. Sukhinov, A. E. Chistyakov, A. V. Shishenya, E. F. Timofeeva, “Predictive modelling of coastal hydrophysical processes in a multiprocession system based on explicit schemes”, Matem. Mod., 30:3 (2018), 83–100; Math. Models Comput. Simul., 10:5 (2018), 648–658
Linking options:
https://www.mathnet.ru/eng/mm3950 https://www.mathnet.ru/eng/mm/v30/i3/p83
|
Statistics & downloads: |
Abstract page: | 555 | Full-text PDF : | 116 | References: | 57 | First page: | 15 |
|