Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                
Skip to main content
Sergio Elaskar

    Sergio Elaskar

    W HEN solving numerically the fluid mechanic equations using finite volume techniques, the necessity of computing convective fluxes arises. Traditional approaches for computing such fluxes give good results if the variables undergo smooth... more
    W HEN solving numerically the fluid mechanic equations using finite volume techniques, the necessity of computing convective fluxes arises. Traditional approaches for computing such fluxes give good results if the variables undergo smooth variations, however, they have serious difficulties if the solution contains discontinuities. In these cases, the numerical schemes that use secondor higher-order approximations develop convergence problems and the solution has oscillations next to discontinuities. On the other hand, the schemes that use first-order approximations generate solutions without oscillations, but the discontinuities may poorly be resolved. To deal with this problem, flux limiter functions were built as linear combinations of firstand second-order approximations [1,2]. In this lineal combination, if the first-order approach has more weight than that of the second order, the scheme becomes diffusive and, reciprocally, if the second-order approach has more weight, the scheme becomes compressive. In robust schemes, the number of limiter functions is equal to the number of equations and a spectral decomposition is used [3]. For the three-dimensional Euler equations system, the spectral decomposition leads to the appearance of three lineally degenerate families of waves [4]. Discontinuities associated with these waves are very difficult to resolve except for schemes that use higher compressive limiters, however, schemes with these limiters are not very robust in solving discontinuities associated with the nonlineal wave families [2]. In this work, a scheme is described which has the capacity to satisfactorily solve discontinuities associated with lineally degenerate wave families without losing robustness. It was implemented in a computational code that, using a nonstructured finite volume technique, solves the three-dimensional Euler equations. Results obtained in some applications are presented. Description of the Scheme
    In later papers we have shown that sunward, generally dark, plasma features originated above posteruption flare arcades are consistent with a scenario where plasma voids are generated by the bouncing and interfering of shocks and... more
    In later papers we have shown that sunward, generally dark, plasma features originated above posteruption flare arcades are consistent with a scenario where plasma voids are generated by the bouncing and interfering of shocks and expansion waves upstream of an initial localized deposition of energy which is collimated in the magnetic field direction. In this paper we analyze the multiple production and interaction of supra--arcade downflows (SAD) and the structure of individual SADs that make them relatively stable features while moving. We compare our results with observations and with the scenarios proposed by other authors.
    En el marco de la teoría cinética de plasmas, cuando la dinámica de las partículas esta gobernada por interacciones electromagnéticas de largo alcance, entonces la evolución temporal de la función de distribución de cada especie está dada... more
    En el marco de la teoría cinética de plasmas, cuando la dinámica de las partículas esta gobernada por interacciones electromagnéticas de largo alcance, entonces la evolución temporal de la función de distribución de cada especie está dada por la ecuación de Vlasov, la cual provee una descriptión cinética completa del plasma. Si los campos magnéticos auto generados y externos son despreciables, la fuerza de Lorentz se debe sólo al campo eléctrico, el cual puede computarse a partir de la ecuación de Poisson en el caso no relativista. En este artículo, se presenta y compara una extensa selección de métodos numéricos usados cotidianamente para la resolución del sistema Vlasov-Poisson sobre un espacio de fases bidimensional x − v. Estos métodos son los basadas en diferencias finitas, volúmenes finitos, y semi-Lagrangianos advectivos y conservativos. La precisión de los esquemas se evalúa y compara a través de los problemas de prueba clásicos del amortiguamiento lineal y no lineal de Landau.
    Se presenta un análisis de la implementación en la distribución de OpenFOAM™ de ESI del método de la entalpia para la simulación del proceso de cambio de fase en metales. Se describe el modelo matemático y aspectos cruciales de la... more
    Se presenta un análisis de la implementación en la distribución de OpenFOAM™ de ESI del método de la entalpia para la simulación del proceso de cambio de fase en metales. Se describe el modelo matemático y aspectos cruciales de la implementación numérica. Para validar las capacidades del solver y su funcionamiento, se resuelven dos problemas típicos de cambio de fase: el problema de Stefan de dos fases y el de fusión de galio en una cavidad bidimensional.

    And 190 more