5C-02, page 1 Computational Micromagnetics: Prediction of time dependent and thermal

Finite element modeling treats magnetization processes on a length scale of several nanometers and thus gives a quantitative correlation between the microstructure and the magnetic properties of ferromagnetic materials. This work presents a novel finite element / boundary element micro-magnetics solver that combines a wavelet-based matrix compression technique for magnetostatic field calculations with a BDF / GMRES method for the time integration of the Gilbert equation of motion. In addition to the hysteresis properties, the numerical solution of the Gilbert equation simulations show that metastable energy minima and nonuniform magnetic states within the grains are important factors in the reversal dynamics at finite temperature.ation shows how reversed domains nucleate and expand. In an array of acicular NiFe elements the switching field varies by about 8 kA/m depending on the magnetic state of neighboring elements. The switching time of submicron magnetic elements depends on the shape of the elements. Elements with slanted ends decrease the overall reversal time, as a transverse demagnetizing field suppresses oscillations of the magnetization. Thermal activated processes can be included adding a random thermal field to the effective magnetic field. Thermally assisted reversal was studied for CoCrPtTa thin film media