Core position-dependent gyrotropic and damping contributions to the Thiele equation approach for accurate spin-torque vortex oscillator dynamics

Abstract

Understanding the nonlinear dynamics of magnetic vortices in spin-torque vortex oscillators (STVOs) is essential for their application in neuromorphic computing. Conventional modeling approaches either rely on the standard Thiele equation, which provides only qualitative predictions, or on micromagnetic simulations, which are computationally expensive. In this work, we refine the Thiele approach by incorporating vortex-profile deformations into the evaluation of the gyrotropic and damping terms. By introducing a more realistic ansatz for the vortex magnetization profile, we determine these effective parameters semi-analytically, and we further develop a procedure to extract them directly from micromagnetic simulations. This numerical framework enables systematic benchmarking of existing analytical models and can be readily extended to other magnetic textures. The comparison between both approaches demonstrates that our semi-analytical model captures the essential nonlinearities of the magnetic vortex dynamics with high accuracy and at low computational cost.