Dipole fields have no effect?

[Milton Persson]

Hello! I have been running some curie-temperature simulations without dipole fields turned on, for speed. Later I reran some of the simulations with dipole fields turned on (sim:enable-dipole-fields) but there was absolutely no difference in the results. Something must be wrong.

It would be great if someone could take a look at the attached input file and tell me what's going on. The only difference between the two simulations is the addition/removal of the line "sim:enable-dipole-fields". Removing the line should disable the fields, as it says in the manual that they are disabled by default.

[Richard Evans]

Hi Milton,

What makes you think dipole fields would have any effect on the Curie temperature? Even for an infinite film with perpendicular magnetization, the dipole fields are roughly ~ 1T, and only at zero K, going to zero at Tc as the magnetic order decreases to zero. This corresponds to maybe a 1 Kelvin change in Tc, but it is sample size dependent and geometry dependent, and so for your particular inputs, exchange will certainly dominate and so the effect should be almost undetectable (there should be very small numerical differences in the simulated values, but bear in mind for an MC simulation the dipole field is so small that it may not significantly influence when a particular spin flip move is accepted or rejected. 

All the best,

Richard

[milton...@gmail.com]

Hi Richard! Thanks for your reply and sorry for the delay!

I understand now that dipole fields would not affect Tc much, but that is not really what I'm after. I'm running curie-temperature simulations simply because I want a temperature sweep. The simulations are of a tri-layer, in which we compare two different states -- the state with the top and bottom layers having parallel magnetization, and the state where they have anti-parallel magnetization. Specifically, it is the "magnetization profile" across the central layer that is of interest. We want to look at it at different temperatures, both below, at and above Tc. I don't know how much this would be affected by dipole fields but the system is ~5 nm thick and ~35 nm x 35 nm laterally, and I have seen some out-of-plane magnetization.

We started out using hysteresis-loop simulations and repeating them at different temperatures, but it was much much faster to do curie-temperature simulations where instead of sweeping the field to go between the parallel and anti-parallel states we simply do two simulations with the system initialized in the parallel and anti-parallel state, respectively. Do you think we can use curie-temperature simulations this way or is it better to go back to hysteresis-loop simulations or to something else?

Thanks in advance!
/Milton