Solution of the self-adjoint radiative transfer equation on hybrid computer systems (Q317211)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Solution of the self-adjoint radiative transfer equation on hybrid computer systems
scientific article

    Statements

    Solution of the self-adjoint radiative transfer equation on hybrid computer systems (English)
    0 references
    0 references
    0 references
    30 September 2016
    0 references
    The authors propose a new efficient algorithm for the calculation of radiative energy transfer based on the solution of a second-order equation with a self-adjoint operator. The algorithm takes into account the angular dependence of the radiation intensity and is free of the ray effect. The numerical implementation is greatly reduced by using a multigroup spectral model and highly accurate quadrature formulae for the sphere, the result is a concentration of the solution of an \(M\times N\) independent elliptic equation, where \(M\) is the number of spectral groups and \(N\) is the number of quadrature points. This is the important key to ensure the required accuracy that can be as high as several tens or even hundreds. The spatial discretization of these equations yields a system of linear algebraic equations forming the discretization of the basic transfer equation between computer processes according to the decomposition of the computational domain. This is an important step to help to simplify the difference operator, to avoid the mixed partial derivatives and to accelerate the convergence of the iterative solution. The result is roughly that a symmetric self-adjoint operator is used instead of the asymmetric operator in the original transport equation, besides other advantages.
    0 references
    0 references
    0 references
    0 references
    0 references
    0 references
    radiative energy transfer
    0 references
    high-temperature plasma
    0 references
    numerical simulation
    0 references
    unstructured grids
    0 references
    high-performance computations
    0 references
    0 references