Averaging in hysteresis loops

[Bassel Alkadour]

Dear Evans,
I am trying to get the average of the magnetization components (x,y,z) in hysteresis loops. Example of the script is at the end.

I collect that the magnetization  components at the output are just the values at the end of each (sim:loop-time-steps).
Changing the value of (sim:time-steps-increment)  has no influence on the output (output:magnetisation).
Also,  increasing the value of "sim:loop-time-steps" does not give any better statistics of Mz as the two figures above show.

Is there a way to get the average of the magnetization components over  each ("sim:loop-time-steps")  or during a portion of it?

Thank you.



dimensions:unit-cell-size = 1 !nm
dimensions:system-size-x = 4 !nm
dimensions:system-size-y = 4 !nm
dimensions:system-size-z = 4 !nm

create:periodic-boundaries-x
create:periodic-boundaries-y
create:periodic-boundaries-z

#------------------------------------------
# Material Files:
#------------------------------------------
material:file="Pr.mat"
#------------------------------------------
# Simulation attributes:
#------------------------------------------
sim:temperature=10

sim:maximum-applied-field-strength=5
sim:applied-field-strength-increment=0.01

sim:loop-time-steps=10000
sim:time-step=1.0E-15
sim:equilibration-time-steps=5000
sim:time-steps-increment = 5


#------------------------------------------
# Program and integrator details
#------------------------------------------

sim:program = hysteresis-loop
sim:integrator=llg-heun
#------------------------------------------
# data output
#------------------------------------------
output:real-time
output:applied-field-strength
output:magnetisation





#Pr.mat file:

#-------------------------------------------------------
materials:num-materials = 2
#-------------------------------------------------------
material[1]:material-name= Ni
material[1]:exchange-matrix[1] = 2.72e-21
material[1]:exchange-matrix[2] = 3.20e-21
material[1]:atomic-spin-moment = 0.63 !muB
material[1]:damping-constant = 0.5
material[1]:uniaxial-anisotropy-constant = 3.3e-26
material[1]:uniaxial-anisotropy-direction = 0 ,0.001,1
material[1]:initial-spin-direction = 0,0,1
#-------------------------------------------------------
material[2]:material-name = Fe
material[2]:exchange-matrix[1] = 3.20e-21
material[2]:exchange-matrix[2] = -9.60e-22
material[2]:atomic-spin-moment = 2.67 !muB
material[2]:damping-constant = 0.5
material[2]:uniaxial-anisotropy-constant = 3.3e-26
material[2]:uniaxial-anisotropy-direction = 0 ,0.001,1
material[2]:initial-spin-direction = 0,0,1
#-------------------------------------------------------

material[1]:alloy-host

material[2]:minimum-height = 0.0
material[2]:maximum-height = 0.0

material[1]:alloy-fraction[2] = 0.2

[Richard Evans]

Dear Bassel,

I have pushed a new update to the develop branch which now enables calculation of the average vector of the magnetization, in total and for different materials. The new keywords are:

output:mean-magnetisation

output:material-mean-magnetisation

These will now calculate an average magnetization over loop time steps, which should reduce the noise for smaller numbers of field steps. Obviously if the number of loop time steps is small then this can still be a noisy quantity. In general the average also does not converge well for small particles, since the transverse fluctuations of the magnetization in time are quite slow compared to the integration time step, and so they will still tend to be noisy even with averaging.

All the best,

Richard 

--
You received this message because you are subscribed to the Google Groups "Vampire Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to vampire-user...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.