3D MHD Simulations of Disk Accretion onto Magnetized Stars: Numerical Approach and Sample Simulations

We present results of global 3D MHD simulations of disk accretion to a rotating star with dipole and more complex magnetic fields using a Godunov-type code based on the “cubed sphere” grid developed earlier in our group. We describe the code and the grid and show examples of simulation results.

💡 Research Summary

The paper presents a comprehensive study of three‑dimensional magnetohydrodynamic (MHD) simulations of accretion from a circumstellar disk onto a rotating, magnetized star. Recognizing the limitations of earlier two‑dimensional axisymmetric models—particularly their inability to capture non‑axisymmetric magnetic geometries, tilted dipoles, and higher‑order multipole fields—the authors develop a new numerical framework based on a Godunov‑type scheme implemented on a “cubed‑sphere” grid. The cubed‑sphere geometry divides the spherical domain into six quasi‑Cartesian patches, allowing high‑resolution, low‑distortion coverage of both polar and equatorial regions while preserving the advantages of structured grids.

The governing MHD equations are cast in conservative form and solved using an HLLD Riemann solver, which accurately resolves fast and Alfvénic shocks as well as contact discontinuities. Temporal integration employs a second‑order TVD Runge‑Kutta method, and the divergence‑free condition (∇·B = 0) is enforced via a Constrained Transport (CT) algorithm, preventing the buildup of numerical magnetic monopoles. Parallelization is achieved through MPI domain decomposition; each of the six patches is further split among many processors, and non‑blocking communication across overlapping buffer zones yields scaling efficiencies above 80 % on several thousand cores.

Physical setup: the star is prescribed a solid‑body rotation and a magnetic field that can be a pure dipole, a tilted dipole, or a superposition of dipole and higher‑order multipoles (e.g., quadrupole). The magnetic axis orientation (inclination θ and azimuth φ) is freely adjustable, enabling systematic exploration of misaligned configurations. The surrounding accretion disk is initialized as a geometrically thin, Keplerian structure with an α‑type viscosity term to drive inflow. At the inner boundary a sponge layer absorbs inflowing material, minimizing artificial reflections.

Key simulation results fall into two categories. First, for tilted dipole fields (e.g., 30° inclination) the disk material is funneled along magnetic field lines into “accretion curtains” that strike the stellar surface at latitudes offset from the rotation pole. This produces asymmetric hot spots and periodic X‑ray pulses, in agreement with observations of many young stellar objects and accreting pulsars. Second, when a quadrupole component is added to the dipole, the magnetic topology becomes considerably more complex. Some field lines thread the disk at smaller radii than the canonical magnetospheric radius, allowing a fraction of the gas to accrete directly onto the star without first forming a well‑defined funnel. Consequently, both traditional funnel‑flow accretion and direct impact coexist, leading to highly variable accretion rates and spot distributions.

The authors validate the code by comparing 1‑D MHD shock tube solutions and by reproducing benchmark results from earlier 2‑D axisymmetric studies. Convergence tests using grid resolutions of 128³, 256³, and 512³ cells demonstrate that key diagnostics—magnetospheric radius, funnel opening angle, and mass‑accretion rate—converge to within 5 % at the 256³ level. Energy conservation remains within 0.1 % over 10 stellar rotation periods, and the ∇·B error stays below 10⁻⁶ throughout the runs.

Overall, the paper delivers a robust, scalable tool for exploring star‑disk magnetic interactions in full three dimensions. By accommodating arbitrary magnetic multipole structures and by providing high‑fidelity treatment of shocks and divergence cleaning, the framework bridges the gap between idealized theoretical models and the complex, time‑variable phenomena observed in T Tauri stars, accreting X‑ray pulsars, and other magnetically active astrophysical systems. The authors outline future extensions that will incorporate radiative transfer, realistic thermodynamics, and magnetic field evolution, paving the way for truly comprehensive simulations of magnetospheric accretion.