O.A. Chubykalo

Brownian dynamics approach to interacting magnetic moments

The question of how to introduce thermal fluctuations in the equation of motion of a magnetic system is addressed. Using the approach of the fluctuation-dissipation theorem we calculate the properties of the noise for both, the fluctuating field and the additive fluctuating torque (force) representation. In contrast to earlier calculations we consider the general case of a system of interacting magnetic moments. We show that the interactions do not result in any correlations of thermal fluctuations in the field representation and that the same widely used formula can be used in the most general case. We further prove that close to the equilibrium where the fluctuation-dissipation theorem is valid, both, field and additive torque (force) representations coincide, being different far away from it. We also show that the uncorrelated character of the noise is due to the form of the Landau-Lifshitz (or Gilbert) damping and under different damping formalisms, the normal mode analysis is proper.

A micromagnetic approach, based on the Monte Carlo algorithm, to the thermally activated magnetization reversal processes

The applicability of the Monte Carlo algorithm to the solution of different micromagnetic problems is discussed in terms of the results obtained from the simulation of thermally activated reversal processes. Our simulational results, supported by experimental data obtained in amorphous thin films and polycrystalline multilayers evidence the occurrence of non-Arrhenius relaxational behaviour characterized by a very slow variation of the macroscopic state of the system for times immediately after the application of the measuring field.

Real time quantification of Monte Carlo steps for different time scales

Time quantification of Monte Carlo steps is studied by the implementation of a new technique which takes into account the realistic size of thermal fluctuations of magnetization along with Landau–Lifshitz–Gilbert dynamic correlations. The computational model has been specifically developed for an ensemble of isolated single-domain particles. The numerical results have been compared with Langevin dynamics calculations and theoretically predicted Brown’s asymptotes for relaxation time of single spin system. In addition we demonstrated that real time quantification of Monte Carlo steps is also possible for different time scales. Implementation of real time scales into Monte Carlo calculations for different sizes of time steps is shown to be convergent to the expected value if the Monte Carlo acceptance rate is taken into account.

Influence of exchange on signal-to-noise ratio in longitudinal recording media

In this work a systematic study of the influence of exchange on media signal-to-noise ratio, (SNR) was undertaken. Log-normal grain distributions were utilized and four models of the intergranular exchange interaction were studied. In addition, three models of media SNR were examined. In all of our studies media SNR was found to plateau and then exhibit a small maximum as exchange is increased. This maximum occurs because remanent magnetization at the center of the bit increases faster than the noise grows at the bit transition in this low coupling regime. For larger values of exchange, the signal-to-noise ratio decreases due to the rapid increase of transition noise. These results suggest that this plateau region should be exploited in studies of the effect of exchange on thermal stability.

Field and thermally activated demagnetization processes in ultra-thin films with in-plane anisotropy: occurrence of non-equivalent reversal modes

We present a micromagnetic analysis of the field and thermally induced magnetization reversal occurring in ultra-thin films having in-plane anisotropy. We used for that purpose a Monte Carlo algorithm which allowed us to work at finite temperature and to simulate the effect on the demagnetization of the thermal fluctuations. Our results show that, for the examined system sizes, demagnetization is ruled by the ratio of the lateral dimension of the system to the magnetostatic correlation length associated with magnetization. When that ratio is small the coherent rotation mechanism is prefered by both the hysteretic and relaxational reversal. The increase of that ratio is accompanied by the occurrence of different incoherent collective reversal mechanisms (which try to minimize the magnetostatic energy). We evidence that those mechanisms can coexist in a given system due to the very similar involved energies. Finally, when the magnetization correlation length becomes small compared to the system size a nucleation-propagation mechanism rules the reversal processes.

Micromagnetic modelling of thermal decay in interacting systems

A computational approach is presented which allows the simulation of time-dependent magnetisation processes over long time scales. The method is based on Monte Carlo methods and allows continuous magnetization changes, thereby allowing cooperative reversal mechanisms, whilst retaining a quantified time step. The method is applied to studies of the decay of information in written bits.

Simulation of magnetic relaxation by a Monte Carlo technique with correlations and quantified time steps

A new integrated numerical approach for simulation of fast and slow relaxation in magnets has been developed. It is based on Monte Carlo calculations which additionally take into account important dynamic information provided by the Langevin dynamics method. Real time quantification of Monte Carlo steps has been achieved by the new technique for a simple one-dimensional modelled system of interacting spins.

Barkhausen jump distributions in a micromagnetic model

Abstract We examine and qualify the collective behavior of a micromagnetic model, which allowed us to distinguish between local events (avalanches) and global ones (Barkhausen jumps). We show that it is possible to scale, in limited system size ranges, the Barkhausen jump distribution functions and discuss the applicability to the description of the collective behavior of the self-organized criticality ideas.

Micromagnetic simulation of transverse biased initial susceptibility measurements

Abstract By means of the hybrid finite element–boundary element method (FEM–BEM) formulated by T.R. Koehler and D.R. Fredkin, we have simulated transverse biased initial susceptibility (TBIS) experiments in various thin film systems. We have varied system size, anisotropy constant, exchange constant and saturation magnetization value. According to the theory developed by Hoffmann, the TBIS measurements provide the values for two fields, the effective anisotropy field and the so-called dispersion field. In the present paper we explore the limits of the validity of TBIS experiments. Our simulations clearly show a good agreement between the anisotropy field obtained by means of the Hoffmanns theory (according to TBIS experiments) and the average anisotropy field for all studied systems. The relation of the Hoffmanns dispersion field with other different fields, calculated from the equilibrium moment configuration, as well as with microstructural characteristic features of the system has also been analysed.

Longitudinal recording media performance as a function of exchange, linear recording density and media texture

The traditional view of the longitudinal magnetic recording media is that optimal signal-to-noise (SNR) is obtained by diminishing intergranular exchange coupling. In the present work the influence of the exchange has been investigated in a realistic media with log-normal distribution of columnar grains (modeled by optimized Voronoi polyhedral) for different linear recording densities and media texture. A more realistic exchange model (surface exchange between the neighboring grains and volume exchange inside a grain) has been employed. The data suggest that some exchange can be present in the system without sacrificing SNR.