fix ID field/particle axvar ayvar azvar
ID is documented in fix command field/particle = style name of this fix command axvar,ayvar,azvar = names of particle-style variables for acceleration components:ul
fix 1 field/particle gradBx gradBy NULL
Specify the formulas used to calculate the acceleration effect of an external field on particle motion. The ID of this fix can be used by the global field particle command which applies the field when particles are advected during a simulation run. This is done by invoking a method in this fix every timestep, which evaluates the specified particle-style variables.
Each of the axvar, ayvar, and azvar arguments is the name of a particle-style variable. The variables should compute the x,y,z components of acceleration applied to each particle by the field. Any of the three variables can be specified as NULL, which means there is no acceleration in that dimension.
Each timestep when a particle is advected the acceleration vector (a) acts as a perturbation on straight-line motion which affects both the end-of-timestep position (x) and velocity (v) vectors of the particle:
xnew = x + dt*v + 0.5*a*dt^2 vnew = v + dt*a
Note that the formulas encoded by the axvar, ayvar, and azvar variables should produce values that are in units of acceleration (distance/time^2, see the units command), not force. And they should not include the timestep (dt) value in the formulas above. That is applied by SPARTA during advection.
See the variable doc page for a description of the formula syntax allowed for particle-style variables. They can include the particle position, thus enabling a spatially-dependent field. They can include the current timestep and timestep size (dt) to enable a time-dependent field. And they can include properties of the particle, such as its mass or magnetic moment.
Note that the global field command provides three alternatives for specifying an external field:
global field constant ... # field is constant in space and time global field particle ... # field is applied on a per particle basis global field grid ... # field is applied on a per grid cell basis
This fix is only used for per-particle fields. It should only be used for fields which vary spatially or in time; otherwise use the constant option which will be much more efficient. The use of per-particle variables allows the field to vary spatially as a function of particle position. It also allows the field to vary in time by having the variables use the current timestep. The field can also depend on particle attributes, such as its mass and magnetic moment (for a B field).
Note that use of the global field particle command with this fix will evaluate the specified particle-style variables every timestep on all particles. Thus a simulation will typically run several times slower than it would without the external field.
The fix field/grid command is an alternative which should run faster but be more approximate. When used with the global field grid command, the grid-style variables it uses are only invoked once as a pre-calculation (for static fields) or once every N timesteps (for time-varying fields). And the field calculation is done for each grid cell, not for each particle. The trade-off is that the fields it calculates are based on the grid cell center pint, and thus are not as accurate as the calucations performed by this fix.
Restart, output info:
No information about this fix is written to binary restart files.
This fix stores a per-particle array of values which can be accessed by various output commands on any timestep, e.g. by the dump particle command. The values are those produced by evaluating the particle-style variables. The number of rows in the array is the number of particles this processor owns. The number of columns in the array is the number of non-NULL variables specified.
fix field/grid, global field