Magnetization (domain wall) dynamics with VAMPIRE

[Rabindra Nepal]

Hi,

Few questions about magnetization dynamics with VAMPIRE:
a) Can we get the x, y, z, mx, my, mz data at different times to make a movie of magnetization dynamics? If yes, how?
b) Can we implement an ansatz (eg Walker ansatz for domain wall) as an initial spin configuration? It seems it takes only uniform initial spin configurations in vector form.
c) How is spin dynamics with VAMPIRE superior to spin dynamics with other micromagnetics packages like OOMMF and mumax3, other than defining the exact mesh using crystal structure in VAMPIRE? If VAMPIRE also solves LLG equation, I guess defining a rectangular mesh instead of an actual crystal structure mesh should not affect any result.

Thanks
Rabindra

[richard....@gmail.com]

OOMMF comment

Hi Rabindra,

> a) Can we get the x, y, z, mx, my, mz data at different times to make a movie of magnetization dynamics? If yes, how?

Yes - you can either get the total magnetization from the "output" file, or you can generate atomic configurations which can be rendered using povray or converted to VTK format.

> b) Can we implement an ansatz (eg Walker ansatz for domain wall) as an initial spin configuration? It seems it takes only uniform initial spin configurations in vector form.

At the moment the only options are uniform for each material or random. However, you can use checkpoint files to save a spin configuration at the end of an equilibration simulation and then read it into a different simulation.

> c) How is spin dynamics with VAMPIRE superior to spin dynamics with other micromagnetics packages like OOMMF and mumax3, other than defining the exact mesh using crystal structure in VAMPIRE? If VAMPIRE also solves LLG equation, I guess defining a rectangular mesh instead of an actual crystal structure mesh should not affect any result.

VAMPIRE is somewhat different from OOMMF and mumax, focusing on the atomic scale. Here the exchange formulation is different i.e. Heisenberg rather than continuous. The main difference is the ability to simulate high temperatures or ferri/antiferromagnets at the atomic scale - you can't do this correctly with oommf. However, if you have larger cells vs atoms then the differences with oommf are likely to be small if the demag calculation is included correctly (1 moment/box).

All the best,

Richard

[Md Ahsanul Abeed]

[Md Ahsanul Abeed]

Hello Richard,

I am MD AHSANUL ABEED from VCU. I am a Mumax 3 user. I just want to know that will SLI enabled dual card work for Mumax 3 or OOMMF simulation? I saw recently that SLI enabled dual GTX TITAN X PASCAL AND DUAL GTX 1080 is available in the market.

Thanks..

--Abeed

On Friday, September 30, 2016 at 5:35:08 PM UTC-4, richard....@gmail.com wrote: