µMAG Standard Problem #4 results

See the problem specification.
µMAG could not succeed without the contributions of colleagues working outside NIST. Although we value the contributions made by these colleagues, NIST does not necessarily endorse the views expressed or the data presented in the submitted solutions shown below.

Solution directory

Submitted Solution: José L. Martins and Tania Rocha

December 16, 2001.
José L. Martins and Tania Rocha
INESC, Lisbon, Portugal
José Martins

These calculations were done using a finite-difference scheme, always with two cells across the thickness of the sample. Three mesh sizes were used: approx. 1 nm cells, approx. 2 nm cells and approx. 4 nm cells (in the two larger axis of the sample).

The LLG equations were solved using 2nd order or 4th order Runge-Kutta. time step of 1e-14 s and adaptive time step (if torque surpasses a specific value, time step is decreased to half its value, if torque is considered very low, time step is doubled to speed up calculations).

The demagnetization field is calculated solving the Poisson equation numerically using a Multigrid method. The boundary conditions are obtained calculating multipole expansion of magnetic charge inside bigger cells and then integrating this expansion.

Exchange interactions were calculated for the 6 neighbours (well, in this case, 5 neighbours) of each cell.



Average magnetization vs t Magnetization data when first crosses zero
Please send comments to rmcmichael@nist.gov and join the µMAG discussion e-mail list.

Site Directory

µMAG organization / NIST CTCMS / rmcmichael@nist.gov