Application of MCNP to estimate the radial neutron flux distribution of the training and research reactor AKR-II

Abstract

Consulta en la Biblioteca ETSI Industriales (9055)

[EN] The rapid development of computer systems and informatics enables increasingly comprehensive and complex computer simulation in science, humanities, medicine, economy and engineering. Such investigations allow gaining insight into basic fundamentals of complex systems. Computer simulations do not necessarily require an objective facility representing the complex system under interest. A typical question in simulations is which condition leads to undesirable system states. In particular, catastrophic events can be investigated in detail without human and technical damage. A typical application of computer simulations is the transport theory. This theory describes the physical mechanism of non-equilibrium systems with linear and nonlinear transport equations. Well known linear transport processes are heat conduction and diffusion problems. In contrast, radiation transport processes like neutron transport, is in general a nonlinear transport problem. Only in case of a sufficient thin particle gas without any self-interaction can be described by a linear transport equation. The classical description of a linear transport problem is the Boltzmann transport equation. This is a seven-dimensional, partial, integro-differential equation, where three dimensions correspond to the location in space, three dimensions to the momentum and one to the time [1-3]. The Boltzmann transport equation is a balance equation of a density function which is changing with time due to sourceand scattering emission, absorption and movement of the particles. An analytical solution is only available for few special cases. Hence, the first task of a transport simulation is to apply appropriate simplification technics using conservation laws, symmetries, and minimal principles to transform the original physical problem into an adequate mathematical model. For this purpose, there are so-called deterministic (like finite element methods, FEM) and stochastic methods (like Monte-Carlo methods) applicable.