# MIF x.y'', where x.y represents the format revision
number. (There was a MIF 1.0 format, but it was never part of a
released version of OOMMF.)
After the format identifier line, any line ending in a backslash,
`\', is joined to the succeeding line before any other processing
is performed. Lines beginning with a `#' character are comments
and are ignored. Blank lines are also ignored.
All other lines must consist of a Record Identifier followed by a parameter list. The Record Identifier is separated from the parameter list by one or more `:' and/or `=' characters. Whitespace and case is ignored in the Record Identifier field.
The parameter list must be a proper Tcl list. The parameters are parsed (broken into separate elements) following normal Tcl rules; in short, items are separated by whitespace, except as grouped by double quotes and curly braces. The grouping characters are removed during parsing. Any `#' character that is found outside of any grouping mechanism is interpreted as a comment start character. The `#' and all following characters on that line are interpreted as a comment.
Order of the records in a MIF 1.1 file is unimportant, except as explicitly stated below. If two or more lines contain the same Record Identifier, then the last one takes precedence (except for Field Range records, of which there may be several active). All records are required unless listed as optional. Some of these record types are not yet supported by mmProbEd, however your may edit a MIF 1.1 file by hand and supply it to mmSolve2D using FileSource.
For convenience, the Record Identifier tags are organized into several groups; these groups correspond to the buttons presented by mmProbEd. We follow this convention below.
The last is specified by the ``Damp Coef'' entry in the MIF 1.1 file. If not specified, a default value of 0.5 is used, which allows the solver to converge in a reasonable number of iterations. Physical materials will typically have a damping coefficient in the range 0.004 to 0.15. The solver engine mmSolve2D requires a non-zero damping coefficient.is the gyromagnetic ratio (in m/(A.s)),
is the damping coefficient (dimensionless).
:-} All of these algorithms except FastPipe and None require that the Part Thickness (cf. the Part Geometry section) be set. All algorithms use Fast Fourier Transform (FFT) techniques to accelerate the calculations.
In general, bitmap pixel values are converted to a thickness relative to the nominal thickness by the formula 1-(R+G+B)/(3M), where R, G and B are the magnitudes of the red, green and blue components, respectively, and M is the maximum allowed component magnitude. For example, black has R=G=B=0, so the relative thickness is 1, and white has R=G=B=M, so the relative thickness is 0.
Experiment parameters
The following records specify the applied field schedule:
The optional control point specs determine the conditions that cause the applied field to be stepped, or more precisely, ends the simulation of the magnetization evolution for the current applied field. The control point specs are specified as -type value pairs. There are 3 recognized control point types: -torque, -time, and -iteration. If a -torque pair is given, then the simulation at the current applied field is ended when |m×h| (i.e., |M×H|/Ms2) at all spins in the simulation is smaller than the specified -torque value (dimensionless). If a -time pair is given, then the simulation at the current field is ended when the elapsed simulation time for the current field step reaches the specified -time value (in seconds). Similarly, an -iteration pair steps the applied field when the iteration count for the current field step reaches the -iteration value. If multiple control point specs are given, then the applied field is advanced when any one of the specs is met. If no control point specs are given on a range line, then the Default Control Point Spec is used.
For example, consider the following Field Range line:
Field Range: 0 0 0 0.05 0 0 5 -torque 1e-5 -time 1e-9
This specifies 6 applied field values, (0,0,0), (0.01,0,0), (0.02,0,0), ..., (0.05,0,0) (in Tesla), with the advancement from one to the next occurring whenever |m×h| is smaller than 1e-5 for all spins, or when 1 nanosecond (simulation time) has elapsed at the current field. (If -torque was not specified, then the applied field would be stepped at 1, 2, 3 4 and 5 ns in simulation time.)
This Field Range record is optional, with a default value of 0 0 0 0 0 0 0.
\
\
\
\' continuation characters as necessary.
Here routinecount is the number of routines, and param1count
is the number parameters (including name1) needed by the
first routine, etc.
# MIF 1.1 # # All units are SI. # ####################### MATERIAL PARAMETERS ############################ Ms: 800e3 # Saturation magnetization in A/m. A: 13e-12 # Exchange stiffness in J/m. K1: 0.5e3 # Anisotropy constant in J/m^3. Anisotropy Type: uniaxial # One of <uniaxial|cubic>. Anisotropy Dir1: 1 0 0 # Directional cosines wrt to coordinate axes ####################### DEMAG SPECIFICATION ############################ Demag Type: ConstMag # One of <ConstMag|3dSlab|2dSlab|3dCharge|FastPipe|None>. ########################## PART GEOMETRY ############################### Part Width: 0.25e-6 # Nominal part width in m Part Height: 1.0e-6 # Nominal part height in m Part Thickness: 1e-9 # Part thickness in m. Cell Size: 8.1e-9 # Cell size in m. #Part Shape: # One of <Rectangle|Ellipse|Oval|Mask>. Optional. ###################### INITIAL MAGNETIZATION ########################### Init Mag: Uniform 90 45 # Initial magnetization routine and parameters ###################### EXPERIMENT PARAMETERS ########################### Field Range: -.05 -.01 0. .05 .01 0. 100 # Start_field Stop_field Steps Field Range: .05 .01 0. -.05 -.01 0. 100 Field Type: Multi 4 \ 7 Ribbon 1 0 1.0e-6 0.25e-6 1.0e-6 1e-9 \ 7 Ribbon 1 0 0 0.25e-6 0 1e-9 \ 9 Tie 100 0 0 0.12e-6 0.5e-6 0.13e-6 0.5e-6 8.1e-9 \ 1 Uniform # The above positions ribbons of positive charge along the upper # and lower edges with strength Ms, applies a large (100 Ms) field # to the center cell, and also applies a uniform field across the # sample stepped from (-.05,-.01,0.) to (.05,.01,0.) (Tesla), and # back, in approximately 0.001 T steps. Default Control Point Spec: -torque 1e-6 # Assume equilibrium has been # reached, and step the applied field, when the reduced torque |mxh| # drops below 1e-6. ###################### OUTPUT SPECIFICATIONS ########################### Base Output Filename: samplerun Magnetization Output Format: binary 8 # Save magnetization states # in binary format with full (8-byte) precision. ########################## MISCELLANEOUS ############################### Randomizer Seed: 1 # Value to seed random number generator with. User Comment: This is an example MIF 1.1 file, with lots of comments.
A sample MIF 2.0 file is included below. The first line of a MIF file must be of the
form ``# MIF x.y'', where x.y represents the format revision
number, here 2.0. Unlike MIF 1.1 files, the structure of MIF 2.0
files are governed by the requirement that they be valid Tcl scripts.
During processing MIF 2.0 files are evaluated inside a Tcl safe
interpreter. Safe interpreters disable certain commands (for example,
disk input/output), but otherwise the full power of the Tcl scripting
language is available for use inside a MIF 2.0 file. Two special
commands, Specify and Miscellaneous, are used to communicate
to the solver the details of the problem to be solved.
An OXS simulation is built as a collection of Oxs_Ext (OXS Extension) objects. Each Oxs_Ext object is specified and initialized in the input MIF 2.0 file using the Specify command. The Specify command takes two arguments: the name of the Oxs_Ext object to create, and an initialization string which is passed on to the Oxs_Ext object during its construction. The objects are created in the order in which they appear in the MIF file, so order is important in some cases. In particular, if one Oxs_Ext object refers to another in its initialization string, then the referred to object must precede the referrer in the MIF file.
Here is a simple Specify block:
Specify Oxs_EulerEvolve:foo {
alpha 0.5
start_dm 0.01
}
The name of the new Oxs_Ext object is ``Oxs_EulerEvolve:foo.'' The
first part of this name, up to the colon, is the the C++ class
name of the object. Here Oxs_EulerEvolve is a class that
integrates the Landau-Lifshitz ODE using a simple forward Euler method.
This must be a child of the Oxs_Ext class. The second part of the
name (following the colon), is a name for this particular instance of
the object. In general, it is possible to have multiple instances of an
Oxs_Ext child class in a simulation, but each instance must have a
unique name. These names are used for identification by output
routines, and to allow one Specify block to refer to another
Specify block appearing earlier in the MIF file. If the second
part of the name is not given, then as a default the empty string is
appended. (E.g., if instead of ``Oxs_EulerEvolve:foo'' above the name
was specified as just ``Oxs_EulerEvolve'', then the effective full name
of the created object would be ``Oxs_EulerEvolve:''.)
The second argument to the Specify command is an arbitrary string that is interpreted by the new Oxs_Ext (child) object in its constructor. The format of this string is up to the designer of the child class, but it is recommended that the string be structured as a Tcl list with an even number of elements, with each pair consisting of a key + value pair. This is the format followed by all Oxs_Ext classes released by the OOMMF team. (Refer to the Oxsii documentation for more details on the individual Oxs_Ext child classes.)
In the above example, the initialization string consists of two key +
value pairs, ``alpha 0.5'' and ``start_dm 0.01''. The first specifies
that the damping parameter in the Landau-Lifshitz ODE is 0.5.
The second specifies the initial step size for the integration routine.
Interested parties should refer to a Tcl programming reference (e.g.,
[15]) for details on forming a proper Tcl list, but in this
example the list as a whole is set off with curly braces (``{''
and ``}''), and individual elements are white space delimited.
Sometimes the value portion of a key + value pair will itself be a list, as in this next example:
Specify Oxs_RectangularMesh:mesh {
cellsize {10e-9 10e-9 10e-9}
atlas Oxs_SectionAtlas:WorldAtlas
}
Here the value associated with ``cellsize'' is a list of 3 elements (the
sampling rate along each of the coordinate axes, in meters). Notice
also that the ``atlas'' value refers to an earlier Oxs_Ext
object, ``Oxs_SectionAtlas:WorldAtlas''.
A Specify block may also include embedded Oxs_Ext objects. This is frequently used to initialize a spatially varying quantity. For example,
Specify Oxs_UniaxialAnisotropy {
axis { Oxs_RandomVectorFieldInit {
min_norm 1
max_norm 1
}
}
K1 { Oxs_UniformScalarFieldInit { value 530e3 } }
}
This magneto-crystalline anisotropy object has a cell-wise randomly
distributed easy axis. To initialize its internal data structure,
Oxs_UniaxialAnisotropy creates a temporary
Oxs_RandomVectorFieldInit object. This temporary object is also a
child of the Oxs_Ext hierarchy--this allows it to be constructed
using the same name-lookup machinery invoked by the Specify
command--but the temporary is known only to the enclosing
Oxs_UniaxialAnisotropy object, so it cannot be referenced from
other Specify blocks. It also does not need to be given an
instance name. It does need an initialization string, however, which is
given here as the 4-tuple ``min_norm 1 max_norm 1''. Notice how the
curly braces are nested so that this 4-tuple is presented to
Oxs_RandomVectorFieldInit as a single item, while
``Oxs_RandomVectorFieldInit'' and the associated initialization string
are wrapped up in another Tcl list, so that the value associated with
``axis'' is parsed at that level as a single item.
The Oxs_UniaxialAnisotropy class also supports cell-wise varying K1, so the value associated with ``K1'' is another embedded Oxs_Ext object. In this particular example, however, K1 is uniform throughout the simulation region, so the trivial Oxs_UniformScalarFieldInit class is used for initialization (to the value 530e3 J/m^3).
This concludes a brief overview of the Specify block command and structure. Because the interpretation of the initialization string in the Specify block is left to the constructed object, the MIF 2.0 format is freely extensible. This also means that one must refer to the documentation of each Oxs_Ext child class to know how to interpret the corresponding initialization string. Details on the standard Oxs_Ext child classes are included with the Oxsii documentation.
The Miscellaneous block is intended to provide to the solver any information that does not fit naturally into one of the Specify blocks. This is intended mainly for development purposes, and may be deprecated in the future.
The content of the Miscellaneous block is structured as a Tcl list with an even number of elements, consisting of key + value pairs. The only key currently supported is basename; the associated value is used as a base for output name construction by some of the data output routines. There is an example Miscellaneous block in the sample MIF 2.0 file included below.
# MIF 2.0
#
# All units are SI.
#
# This file must be a valid Tcl script.
#
# Individual Oxs_Ext objects are loaded and initialized
# via Specify command blocks. The following block defines
# the extents (in meters) of the volume to be modeled.
# The prefix 'Oxs_SectionAtlas' specifies the type
# of Oxs_Ext object to create, and the suffix ':WorldAtlas' is
# the name assigned to this particular instance. Each object
# created by a Specify command must have a unique full name
# (here 'Oxs_SectionAtlas:WorldAtlas'). If the suffix is
# not explicitly given, then the default ':' is automatically
# assigned. References may be made to either the full name,
# or the shorter suffix instance name (here ':WorldAtlas') if the
# latter is unique. See the Oxs_StandardDriver block for some
# reference examples.
Specify Oxs_SectionAtlas:WorldAtlas {
top { Oxs_RectangularSection {
xrange {0 500e-9}
yrange {0 250e-9}
zrange {3e-9 9e-9}
} }
bottom { Oxs_RectangularSection {
xrange {0 500e-9}
yrange {0 250e-9}
zrange {0 3e-9}
} }
world { Oxs_RectangularSection {
xrange {0 500e-9}
yrange {0 250e-9}
zrange {0 9e-9}
} }
}
# The Oxs_RectangularMesh object is initialized with the
# discretization cell size (in meters).
Specify Oxs_RectangularMesh:mesh {
cellsize {10e-9 10e-9 10e-9}
atlas :WorldAtlas
}
# Magnetocrystalline anisotropy block.
# Oxs_UniformScalarFieldInit and Oxs_UniformVectorFieldInit
# are examples of embedded Oxs_Ext objects used to provide
# internal initialization of the Oxs_UniaxialAnisotropy
# object.
Specify Oxs_UniaxialAnisotropy {
K1 { Oxs_UniformScalarFieldInit { value 530e3 } }
axis { Oxs_RandomVectorFieldInit {
min_norm 1
max_norm 1
}
}
}
# Exchange energy with spatially varying exchange
# coefficient A. Inside the top layer (refer to
# Oxs_SectionAtlas:WorldAtlas above) A = 13e-12 J/m,
# in the bottom layer A = 30e-12 J/m (taken from the
# default_A value), and the interlayer coupling is
# A = 20e-12 J/m.
Specify Oxs_Exchange6Ngbr {
default_A 30e-12
atlas :WorldAtlas
A {
{ top top 13e-12 }
{ top bottom 20e-12 }
}
}
# Define a couple of constants for later use.
set PI [expr {4*atan(1.)}]
set MU0 [expr {4*$PI*1e-7}]
# The Oxs_UZeeman class is initialized with field ranges
# in A/m. The following block uses the Hscale option to
# allow inputs in mT. To make the $mu0 subsitution active,
# we enclose the block with double quotes ("") instead of
# curly braces ({}). (There are other ways to achieve this.)
Specify Oxs_UZeeman:AppliedField "
Hscale [expr 0.001/$MU0]
Hrange {
{ 0 0 0 10 0 0 2 }
{ 10 0 0 -10 0 0 2 }
{ 0 0 0 0 10 0 4 }
{ 1 1 1 5 5 5 0 }
}
"
# Enable demagnetization (stray) field computation.
# This block takes no parameters.
Specify Oxs_Demag {}
# First order Euler ODE solver
Specify Oxs_EulerEvolve {
alpha 0.5
start_dm 0.01
}
# The following procedure is used to initialize the initial
# spin configuration in the Oxs_StandardDriver block.
proc UpDownSpin { x y z xmin ymin zmin xmax ymax zmax } {
if { $x < 0.55*$xmin + 0.45*$xmax } {
return "0 1 0"
} elseif { $x > 0.45*$xmin + 0.55*$xmax } {
return "0 -1 0"
} else {
return "0 0 1"
}
}
Specify Oxs_StandardDriver {
evolver Oxs_EulerEvolve
min_timestep 1e-18
max_timestep 1e-9
stopping_dm_dt 0.01
mesh :mesh
number_of_stages 1
stage_iteration_limit 0
total_iteration_limit 0
Ms { Oxs_UniformScalarFieldInit { value 8e5 } }
m0 { Oxs_ScriptVectorFieldInit {
script {UpDownSpin}
norm 1
} }
}
# Block specifying various miscellaneous data
Miscellaneous {
basename test
}