Все выпуски

Статья носит методический характер и посвящена решению трех классических уравнений математической физики (Лапласа, диффузии и волнового) простейшими численными схемами в формулировке клеточных автоматов (КА). Особое внимание уделяется законам сохранения вещества и неприятному эффекту избыточной гексагональной симметрии (ИГС).

Делается вывод о том, что по сравнению с классическими конечно-разностными методами, хотя локальная функция перехода (ЛФП) КА терминологически эквивалентна шаблону вычислительной двухслоевой явной схемы, различие состоит в замене матричных (direct) методов (например, метода прогонки для трехдиагональной матрицы) итерационными. Из этого следуют более жесткие требования к дискретизации условий для граничных КА-ячеек.

Для гексагональной сетки и консервативных граничных условий записана корректная ЛФП для граничных ячеек, справедливая, по крайней мере, для границ прямоугольной и круговой формы. Предложена идея разделения ЛФП на internal, boundary и postfix. На примере этой задачи заново осмыслено значение числа Куранта–Леви как соотношения скорости сходимости КА к решению задачи, данному на фиксированный момент времени, и скорости изменения самого решения в динамике.

The paper has methodical character; it is devoted to three classic partial differential equations (Laplace, Diffusion and Wave) solution using simple numerical methods in terms of Cellular Automata. Special attention was payed to the matter conservation law and the offensive effect of excessive hexagonal symmetry.

It has been shown that in contrary to finite-difference approach, in spite of terminological equivalence of CA local transition function to the pattern of computing double layer explicit method, CA approach contains the replacement of matrix technique by iterative ones (for instance, sweep method for three diagonal matrixes). This suggests that discretization of boundary conditions for CA-cells needs more rigid conditions.

The correct local transition function (LTF) of the boundary cells, which is valid at least for the boundaries of the rectangular and circular shapes have been firstly proposed and empirically given for the hexagonal grid and the conservative boundary conditions. The idea of LTF separation into «internal», «boundary» and «postfix» have been proposed. By the example of this problem the value of the Courant-Levy constant was re-evaluated as the CA convergence speed ratio to the solution, which is given at a fixed time, and to the rate of the solution change over time.

This paper studies electromagnetic fields, frequencies of lasing, and emission thresholds of a drop-shaped microcavity laser. From the mathematical point of view, the original problem is a nonstandard two-parametric eigenvalue problem for the Helmholtz equation on the whole plane. The desired positive parameters are the lasing frequency and the threshold gain, the corresponding eigenfunctions are the amplitudes of the lasing modes. This problem is usually referred to as the lasing eigenvalue problem. In this study, spectral characteristics are calculated numerically, by solving the lasing eigenvalue problem on the basis of the set of Muller boundary integral equations, which is approximated by the Nystr¨om method. The Muller equations have weakly singular kernels, hence the corresponding operator is Fredholm with zero index. The Nyström method is a special modification of the polynomial quadrature method for boundary integral equations with weakly singular kernels. This algorithm is accurate for functions that are well approximated by trigonometric polynomials, for example, for eigenmodes of resonators with smooth boundaries. This approach leads to a characteristic equation for mode frequencies and lasing thresholds. It is a nonlinear algebraic eigenvalue problem, which is solved numerically by the residual inverse iteration method. In this paper, this technique is extended to the numerical modeling of microcavity lasers having a more complicated form. In contrast to the microcavity lasers with smooth contours, which were previously investigated by the Nyström method, the drop has a corner. We propose a special modification of the Nyström method for contours with corners, which takes also the symmetry of the resonator into account. The results of numerical experiments presented in the paper demonstrate the practical effectiveness of the proposed algorithm.

This paper studies electromagnetic fields, frequencies of lasing, and emission thresholds of a drop-shaped microcavity laser. From the mathematical point of view, the original problem is a nonstandard two-parametric eigenvalue problem for the Helmholtz equation on the whole plane. The desired positive parameters are the lasing frequency and the threshold gain, the corresponding eigenfunctions are the amplitudes of the lasing modes. This problem is usually referred to as the lasing eigenvalue problem. In this study, spectral characteristics are calculated numerically, by solving the lasing eigenvalue problem on the basis of the set of Muller boundary integral equations, which is approximated by the Nystr¨om method. The Muller equations have weakly singular kernels, hence the corresponding operator is Fredholm with zero index. The Nyström method is a special modification of the polynomial quadrature method for boundary integral equations with weakly singular kernels. This algorithm is accurate for functions that are well approximated by trigonometric polynomials, for example, for eigenmodes of resonators with smooth boundaries. This approach leads to a characteristic equation for mode frequencies and lasing thresholds. It is a nonlinear algebraic eigenvalue problem, which is solved numerically by the residual inverse iteration method. In this paper, this technique is extended to the numerical modeling of microcavity lasers having a more complicated form. In contrast to the microcavity lasers with smooth contours, which were previously investigated by the Nyström method, the drop has a corner. We propose a special modification of the Nyström method for contours with corners, which takes also the symmetry of the resonator into account. The results of numerical experiments presented in the paper demonstrate the practical effectiveness of the proposed algorithm.

В статье предложена математическая модель фазового перехода на примере гидрирования/дегидрирования порошка металла. Рассматривается одна частица, форма которой обладает некоторой симметрией. Шар, цилиндр и плоская пластина являются частными случаями симметричных форм. Модель описывает как сценарий «сжимающегося ядра» (формирование слоя новой фазы на поверхности частицы с его последующим утолщением), так и сценарий «образования и роста зародышей», при которых сплошной слой не формируется до полного исчезновения старой фазы. Модель представляет собой неклассическую диффузионную краевую задачу со свободной границей и нелинейными граничными условими III рода. Предположения симметрии позволяют свести задачу к одной пространственной переменной. Модель апробирована на серии экспериментальных данных. Показано, что влияние формы частиц на кинетику несущественно. Также показано, что ансамбль частиц различных форм с распределением по размерам может быть аппроксимирован одной частицей «среднего» размера простой формы, что оправдывает использование в моделях упрощающих предположений.

In the paper we construct the model of phase change. Process of hydriding / dehydriding is taken as an example. A single powder particle is considered under the assumption about its symmetry. A ball, a cylinder, and a flat plate are examples of such symmetrical shapes. The model desribes both the "shrinking core"(when the skin of the new phase appears on the surface of the particle) and the "nucleation and growth"(when the skin does not appear till complete vanishing of the old phase) scenarios. The model is the non-classical boundary-value problem with the free boundary and nonlinear Neumann boundary condition. The symmetry assumptions allow to reduce the problem to the single spatial variable. The model was tested on the series of experimental data. We show that the particle shape’s influence on the kinetics is insignificant. We also show that a set of particles of different shapes with size distribution can be approxomated by the single particle of the "average" size and of a simple shape; this justifies using single particle approximation and simple shapes in mathematical models.