Unconsistance results with the same parameters

[Tang Wan-Sheng]

Dear all,

I am working on the FM-AFM system. However, I met some problems on the MH-loop simulation.

I used exactly the same parameters, but only the range of the simulated magnetic field was different, I had tried ±0.2 T and ±0.3 T by changing Bstart in the following code (seeing the following figure 1). I had also tried to change relax() to run(1e-8). These results are all quite different to each other. I think these unconsistance results are very nonsense, the MH-loop should not depand on the range of the external field as long as the start point of the external field is large enough. However, I have no idea about the problems of the codes. I am hoping someone can give me some help.

Figure 1.

Best regards,

==================================

===== Follwoing is the source code =====

//============================

//----- Structure Hint -----

/*

Structure := FM_AFM

*/

//============================

//----- Setting the parameter -----//

Num_AFM_b  := 1

Msat_AFM_b := 700e3

Ku1_AFM_b  := 2.5e6

EA_AFM_b  := vector(1, 0, 0) // x-axis

m_init_AFM_b := RandomMag()

Num_AFM  := 3

Msat_AFM := 700e3

Ku1_AFM  := 0.1e6

EA_AFM   := vector(1, 0, 0)

m_init_AFM := RandomMag()

Num_FM  := 1

Msat_FM := 1000e4

Ku1_FM  := 7.5e4

EA_FM  := vector(0, 0, 1)

m_init_FM := uniform(0.01, 0.01, 1)

//============================

// Exchange coupling

Aex_FM        :=  1.0e-11

Aex_FM_AFM    :=  5.0e-13

Aex_AFM       := -1.0e-12

Aex_AFM_AFM_b := -5.0e-13

Aex_AFM_b     := -1.0e-12

//============================

//============================

//----- Setting the number of cell in xyz -----//

thick := Num_FM + Num_AFM + Num_AFM_b

setgridsize(2, 2, thick)

//----- Setting the size of each cell in xyz -----//

setcellsize(2e-9, 2e-9, 1e-9)

//----- Setting the measurement -----//

TableAdd(m_full)

TableAdd(B_ext)

TableAdd(m.region(0))

//----- Setting the subregion -----//

// AFM_b

defregion(2, layer(4))

// AFM

defregion(1, layers(1, 4))

// FM

defregion(0, layer(0))

//----- Enable/Disable the demagnetization field -----//

EnableDemag = False

x_PBC := 20

y_PBC := 20

z_PBC := 0

setPBC(x_PBC, y_PBC, z_PBC)

//----- Setting material parameter -----

// Material property

alpha.setregion(2, 0.1)

alpha.setregion(1, 0.1)

alpha.setregion(0, 0.02)

// ======================================================= //

/* >>>>> AFM_b <<<<< */

Msat.setregion(2, Msat_AFM_b)

Ku1.setregion(2, Ku1_AFM_b)

anisU.setregion(2, EA_AFM_b)

m.setregion(2, m_init_AFM_b)

Aex.setregion(2, Aex_AFM_b)

// ------------------------------------------------------------------

ext_InterExchange(2, 1, Aex_AFM_AFM_b)

// ------------------------------------------------------------------

/* >>>>> AFM <<<<< */

Msat.setregion(1, Msat_AFM)

Ku1.setregion(1, Ku1_AFM)

anisU.setregion(1, EA_AFM)

m.setregion(1, m_init_AFM)

Aex.setregion(1, Aex_AFM)

// ------------------------------------------------------------------

ext_InterExchange(1, 0, Aex_FM_AFM)

// ------------------------------------------------------------------

/* >>>>> FM <<<<< */

Msat.setregion(0, Msat_FM)

Ku1.setregion(0, Ku1_FM)

anisU.setregion(0, EA_FM)

m.setregion(0, m_init_FM)

Aex.setregion(0, Aex_FM)

//======================================================================

// frozenspins.setregion(0, 1)

saveAs(m, "Initial") // store m as .ovf

relax()

saveAs(m, "InitialRelaxed") // store m as .ovf

//======================================================================

//----- Setting external field variation (T) -----

Bstart := 0.30

B_step := 1e-3

for B := Bstart ; B > -Bstart; B-=B_step{

    B_ext = vector(0, 0, B)

   

    relax()

   

    save(m) // store m as .ovf

    tablesave()

}

for B := -Bstart ; B < Bstart; B+=B_step{

    B_ext = vector(0, 0, B)

   

    relax()

    save(m) // store m as .ovf

    tablesave()

}

[Tang Wan-Sheng]

Dear all,

Hoping someone can help me solving this problem. I also want to ask how to determine to use relax, minimize or run to do the simulation?

Best regards,

Tang Wan-Sheng 在 2025年9月2日 星期二下午6:00:12 [UTC+8] 的信中寫道：