fix insert/rate/region command

Syntax

fix ID group-ID insert/rate/region seed seed_value distributiontemplate dist-ID general_keywords general_values insert_rate_keywords insert_rate_values
  • ID, group-ID are documented in fix command
  • insert/rate/region = style names of these fix commands
  • seed = obligatory keyword
  • seed_value = random # seed (prime number greater 10000)
  • distributiontemplate = obligatory keyword
  • dist-ID = ID of a fix_particledistribution_discrete to be used for particle insertion
  • one or more general keyword/value pairs can be appended
  • general_keywords = verbose or maxattampt or nparticles or mass or particlerate or massrate or insert_every or overlapcheck or all_in or random_distribute or compress_tags or vel constant or vel uniform or vel gaussian or orientation or omega or set_property
verbose = yes or no
maxattempt value = ma
  ma = max # of insertion attempts per atom (positive integer)
nparticles values = np or INF
  np =  number of particles to insert (positive integer)
  INF =  insert as many particles as possible
mass values = mp
  mp = mass of particles to be inserted (positive float)
  INF =  insert as many particles as possible
particlerate values = pr
  pr = particle insertion rate (particles/time units)
massrate values = mr
  mr = mass insertion rate (mass/time units)
insert_every value = ie
  ie = every how many time-steps particles are inserted - insertion happens periodically (positive integer)
start value = ts
  ts = time-step at which insertion should start (positive integer larger than current time-step)
overlapcheck value = yes or no
all_in value = yes or no
random_distribute value = exact or uncorrelated
compress_tags value = yes or no
vel constant values = vx vy vz
  vx = x-velocity at insertion (velocity units)
  vy = y-velocity at insertion (velocity units)
  vz = z-velocity at insertion (velocity units)
vel uniform values = vx vy vz vFluctx vFlucty vFluctz
  vx = mean x-velocity at insertion (velocity units)
  vy = mean y-velocity at insertion (velocity units)
  vz = mean z-velocity at insertion (velocity units)
  vFluctx = amplitude of uniform x-velocity fluctuation at insertion (velocity units)
  vFlucty = amplitude of uniform y-velocity fluctuation at insertion (velocity units)
  vFluctz = amplitude of uniform z-velocity fluctuation at insertion (velocity units)
vel gaussian values = vx vy vz vFluctx vFlucty vFluctz
  vx = mean x-velocity at insertion (velocity units)
  vy = mean y-velocity at insertion (velocity units)
  vz = mean z-velocity at insertion (velocity units)
  vFluctx = standard deviation of Gaussian x-velocity fluctuation at insertion (velocity units)
  vFlucty = standard deviation of Gaussian y-velocity fluctuation at insertion (velocity units)
  vFluctz = standard deviation of Gaussian z-velocity fluctuation at insertion (velocity units)
orientation values = random or template or constant q1 q2 q3 q4
  random = randomize rotational orientation
  template = use orientation from particle template
  constant = use constant quaternion for orientation
  q1 q2 q3 q4 = quaternion values for constant orientation
omega values = constant omegax omegay omegaz
  constant = obligatory word
  omegax = x-component of angular velocity (1/time units)
  omegay = y-component of angular velocity (1/time units)
  omegaz = z-component of angular velocity (1/time units)
set_property values = property-varname val
  property-varname = variable name of a fix property/atom holding a scalar value for each particle
  val = value to initialize the property with upon insertion
  • following the general keyword/value section, one or more rate_region keyword/value pairs can be appended for the fix insert/rate/region command
  • rate_region keywords = region or check_dist_from_subdomain_border or ntry_mc
region value = region-ID
  region-ID = ID of the region where the particles will be generated (positive integer)
check_dist_from_subdomain_border values = yes or no
ntry_mc values = n
  n = number of Monte-Carlo steps for calculating the region's volume  (positive integer)

Examples

fix ins all insert/rate/region seed 123457 distributiontemplate pdd1 nparticles 1000 particlerate 5000 insert_every 1000 overlapcheck yes region mysphere ntry_mc 10000

General description:

Insert particles into a granular run every few timesteps within a specified region at a specified rate.

The verbose keyword controls whether statistics about particle insertion is output to the screen each time particles are inserted.

This command must use the region keyword to define an insertion volume. The specified region must have been previously defined with a region command. Dynamic regions are not supported as insertion region. Each timestep particles are inserted, they are placed randomly inside the insertion volume.

To specify the number of particles to be inserted, you must use either the nparticles or the mass keyword (but not both). In the latter case, the number of particles to be inserted is calculated from the mass expectancy given by the particle distribution.

Likewise, you can use the particlerate or the massrate keyword (but not both) to control the insertion rate.

The frequency of the particle insertion is controlled by the keyword insert_every, which defines the number of time-steps between two insertions. The number of particles to be inserted at each insertion event is calculated from the insertion rate and insert_every. The start keyword can be used to set the time-step at which the insertion should start.

This command must use the distributiontemplate keyword to refer to a fix_particledistribution_discrete (defined by dist-fix-ID) that defines the properties of the inserted particles.

Inserted particles are assigned the atom type specified by the particledistribution defined via the fix_particledistribution_discrete and are assigned to 4 groups: the default group “all” and the group specified in the fix insert command, as well as the groups specified in the fix_particledistribution_discrete and fix_particletemplate_sphere command (all of which can also be “all”).

The keyword overlapcheck controls if overlap is checked for at insertion, both within the inserted particle package and with other existing particles. If this option is turned off, insertion will scale very well in parallel, otherwise not. Be aware that in case of no overlap check, highly overlapping configurations will be produced, so you will have to relax these configurations.

With keyword check_dist_from_subdomain_border, the user can specify if in case of overlapcheck = yes particles are inserted only at least one bounding radius away from processor domain borders to avoid possible overlaps with particles inserted on another process.

If overlapcheck if performed, the number of insertion attempts per particle can be specified via the maxattempt keyword. Each timestep particles are inserted, the command will make up to a total of M tries to insert the new particles without overlaps, where M = # of inserted particles * ma. If unsuccessful at completing all insertions, a warning will be printed.

The all_in flag determines if the particle is completely contained in the insertion region (all_in yes) or only the particle center (all_in no). Currently all_in yes is not yet supported for all types of insertion.

Keyword random_distribute controls how the number of particles to be inserted is distributed among parallel processors and among the particle templates in the particle distribution. For style exact, the number of particles to be inserted each step is matched exactly. For style uncorrelated, the number of particles to be inserted for each particle template will be rounded in an uncorrelated way, so the total number of inserted particles may vary for each insertion step. However, statistically both ways should produce the same result. Style uncorrelated may be faster in parallel since it does not need global MPI operations. Please note that if the # of particles to be inserted is calculated e.g. from a particle mass to be inserted, the number of particles to be inserted each insertion step will vary by 1, irrespective of the random_distribute settings. This is because in this case the # of particles to insert in each step will be a floating point number, and applying a simple floor/ceil rounding operation would lead to a statistical bias.

If keyword compress_tags is set to ‘yes’, then atom IDs are re-assigned after each insertion procedure. The default is ‘no’. This is typically only recommended if some models internally store arrays which have the length defined via the max ID of any atom.

Warning

Setting compress_tags to ‘yes’ will result in all contact history to be lost at that point in time. External post-processing tools will give wrong results because they typically track atoms via their IDs.

The initial velocity and rotational velocity can be controlled via the vel and omega keywords. vel constant simply patches a constant velocity to the inserted particles, vel uniform sets uniformly distributed velocities with mean and amplitude. vel gaussian sets Gaussian distributed particle velocities with mean and std. deviation.

The set_property option can be used to initialize scalar per-particle properties such as temperatures, which are stored in a a fix property/atom.

Restart, fix_modify, output, run start/stop, minimize info

Information about this fix is written to binary restart files. This means you can restart a simulation while inserting particles, when the restart file was written during the insertion operation.

None of the fix_modify options are relevant to this fix. A global vector is stored by this fix for access by various output commands. The first component of the vector is the number of particles already inserted, the second component is the mass of particles already inserted. No parameter of this fix can be used with the start/stop keywords of the run command. This fix is not invoked during energy minimization.

Restrictions

The overlapcheck = ‘yes’ option performs an inherently serial operation and will thus not scale well in parallel. For this reason, if you want to generate large systems, you are advised to turn overlapcheck off and let the packing relax afterwards to generate a valid packing.

Option all_in = ‘yes’ will not work if the region used is a tet mesh region.

Keywords duration and extrude_length can not be used together.

Currently all_in yes is not yet supported for all types of insertion.

Dynamic regions are not supported as insertion region.

Default

The defaults are maxattempt = 50, all_in = no, overlapcheck = yes vel = 0.0 0.0 0.0, omega = 0.0 0.0 0.0, start = next time-step, duration = insert_every, ntry_mc = 100000, random_distribute = exact, compress_tags = no, check_dist_from_subdomain_border = yes