problem when making a hysteresis curve of a small nanostructure

[Arturo]

I have tried to relax a pyramidal nanostructure using a 0.20 nm mesh, to avoid the defects.  The volume of the system is 18x18x18 nm3, and the number of cells in the system is 730,000 approx. When trying to make a hysteresis curve, for large magnetic field (2 T), mumax3 does not modify the magnetization of the nanostructure. Is this due to the size of the mesh?

Bmax: = 2000.0e-3

Bstep: = 5.0e-3

alpha = 0.5

TableAdd (B_ext)

for B: = - Bmax; B <= Bmax; B + = Bstep {

     B_ext = vector (B, 0, 0)

     relax ()

     tablesave ()

}

[GUIDOBETH LUIS NICOLµS SµEZ TORO]

see the example in the page of mumax3 for hysteresis loop. I recommend reducing the error in the calculation of energy minimization wiht minimizerstop = 1e-7. And to reduce possible problems with geometry and mesh, you can set the smoothness with which you work on the edges, I recommend using EdgeSmooth = 8.

MinimizerStop = 1e-7
EdgeSmooth = 8
TableAdd(B_ext)

for B:=0.0; B<=Bmax; B+=Bstep{
    B_ext = vector(B, 0, 0)
    minimize()   // small changes best minimized by minimize()
    tablesave()
}

--
You received this message because you are subscribed to the Google Groups "mumax2" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mumax2+un...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/mumax2/f195c449-2cff-401c-9304-949470b038f6%40googlegroups.com.

--

Guidobeth Saez 

Estudiante de Magister en Ciencias Físicas

Institución: Universidad de La Frontera

Ciudad: Temuco, Chile.

Contacto: +569 6813 3548.

[Arturo]

Thank you, but I had already tried that, the same thing happens. I don't know why the magnetization doesn't change. I have programmed the same structure in oommf with the same mesh and the magnetization changes. it is strange.

To view this discussion on the web visit https://groups.google.com/d/msgid/mumax2/CAO%3DT9%3DBn4QB_%3DfX9n08FbQN5Y5WU_veiK7qD4BfedKeNTan%3Diw%40mail.gmail.com.

[GUIDOBETH LUIS NICOLµS SµEZ TORO]

the problem is that instead of relax () you must use minimize () in the hysteresis cycle. in addition to verifying that the initial condition for magnetization must be previously relaxed before being entered into the hysteresis cycle, something like m = uniform (0,0,1) relax ().

To view this discussion on the web visit https://groups.google.com/d/msgid/mumax2/CACBfG6FX2UdKtNOiyU1ztwXFaRiKfQknREzNXXtGRJ%2BC3nwp%3DQ%40mail.gmail.com.

[Arturo]

No, that's not the problem. I've used minimize() instead of relax(), and the same problems happens.

[Felipe Garcia]

Dear Arturo,

First, a mesh of 0.2 nm is very small. 0.2 nm is of the order of the cell size in many materials. To be consistent, one has to increase that at least to 1nm or close to that. Otherwise, the number of atom per cell is close to one and the continuous approach of micromagnetism is no longer valid.

Second, the problem you have is the problem of nucleation. If one applies the field along the easy axis then the system does not nucleate at the proper value because there is no breaking of the symmetry. The total torque is zero and then it does not switch. The system is in an unstable state but it does now switch like a ball in the top of a mountain. The problem in your case is increased because of the small cell that makes a a large exchange field. This question has been asked before in mumax group several times. Here for example https://groups.google.com/forum/#!topic/mumax2/IJOX93ms21w

To solve this problem, one needs to add some noise before each step or run the simulations with a small angle between the field and the anisotropy axis. In you case, the dipolar field could break the symmetry but it seems that it is not the case.

Best regards,

Felipe

On Sun, Mar 29, 2020 at 11:57 PM Arturo <mari...@gmail.com> wrote:

No, that's not the problem. I've used minimize() instead of relax(), and the same problems happens.

--
You received this message because you are subscribed to the Google Groups "mumax2" group.
To unsubscribe from this group and stop receiving emails from it, send an email to mumax2+un...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/mumax2/1f76080c-ebd3-46aa-9115-1fba4a333f8b%40googlegroups.com.