Syntax:
compute ID boundary mix-ID value1 value2 ...
compute ID boundary/kk mix-ID value1 value2 ...
n = count of particles hitting boundary nwt = weighted count of particles hitting boundary nflux = flux of particles on boundary mflux = flux of mass on boundary press = magnitude of normal pressure on boundary shx,shy,shz = components of shear stress on boundary ke = flux of particle kinetic energy on boundary erot = flux of particle rotational energy on boundary evib = flux of particle vibrational energy on boundary etot = flux of particle total energy on boundary
Examples:
compute 1 boundary all n press eng compute mine boundary species press shx shy shz
These commands will print values for the current timestep for the xlo and xhi boundaryies, as part of statistical output:
compute 1 boundary all n press stats_style step np c_1[1][1] c_1[1][2] c_1[2][1] c_1[2][2]
These commands will dump time averages for each species and each boundary to a file every 1000 steps:
compute 1 boundary species n press shx shy shz fix 1 ave/time 10 100 1000 c_1[*] mode vector file tmp.boundary
Description:
Define a computation that calculates one or more values for each boundary (i.e. face) of the simulation box, based on the particles that cross or collide with the boundary. The values are summed for each group of species in the specified mixture. See the mixture command for how a set of species can be partitioned into groups.
Note that depending on the settings for the boundary command, when a particle collides with a boundary, it can exit the simulation box (outflow), re-enter from the other side (periodic), reflect specularly from the boundary, or interact with it as if it were a surface. In the surface case, the incident particle may bounce off (possibly as a different species), be captured by the boundary (vanish), or a 2nd particle can also be emitted. The formulas below account for all these possible scenarios. As an example, the pressure exerted on an outflow boundary versus a specularly reflecting boundary is different, since in the former case there is no net momentum flux back into the simulation box by reflected particles.
Also note that all values for a boundary collision are tallied based on the species group of the incident particle. Quantities associated with outgoing particles are part of the same tally, even if they are in different species groups.
The results of this compute can be used by different commands in different ways. The values for a single timestep can be output by the stats_style command.
The values over many sampling timesteps can be averaged by the fix ave/time command. It does its averaging as if the particles striking the boundary at each sampling timestep were combined together into one large set to compute the formulas below. The answer is then divided by the number of sampling timesteps if it is not otherwise normalized by the number of particles. Note that in general this is a different normalization than taking the values produced by the formulas below for a single timestep, summing them over the sampling timesteps, and then dividing by the number of sampling steps. However for the current values listed below, the two normalization methods are the same.
NOTE: If particle weighting is enabled via the global weight command, then all of the values below are scaled by the weight assigned to the grid cell in which the particle collision with the boundary occurs. The only exception is the the n value, which is NOT scaled by the weight; it is a simple count of particle crossings or collisions with the boundary.
The n value counts the number of particles in the group crossing or colliding with the boundary.
The nwt value counts the number of particles in the group crossing or colliding with the boundary and weights the count by the weight assigned to the grid cell in which the particle collision with the boundary occurs. The nwt quantity will only be different than n if particle weighting is enabled via the global weight command.
The nflux value calculates the number flux imparted to the boundary by particles in the group. This is computed as
Nflux = N / (A * dt / fnum)
where N is the number of all contributing particles, normalized by A = the area of the surface element, dt = the timestep, and fnum = the real/simulated particle ratio set by the global fnum command.
The mflux value calculates the mass flux imparted to the boundary by particles in the group. This is computed as
Mflux = Sum_i (mass_i) / (A * dt / fnum)
where the sum is over all contributing particle masses, normalized by the area of the surface element, dt and fnum as defined before.
The press value calculates the pressure P exerted on the boundary in the normal direction by particles in the group, such that outward pressure is positive. This is computed as
p_delta = mass * (V_post - V_pre) P = Sum_i (p_delta_i dot N) / (A * dt / fnum)
where A, dt, fnum are defined as before. P_delta is the change in momentum of a particle, whose velocity changes from V_pre to V_post when colliding with the boundary. The pressure exerted on the boundary is the sum over all contributing p_delta dotted into the normal N of the boundary which is directed into the box, normalized by A = the area of the boundary face and dt = the timestep and fnum = the real/simulated particle ratio set by the global fnum command.
The shx, shy, shz values calculate the shear pressure components Sx, Sy, Sz extered on the boundary in the tangential direction to its normal by particles in the group, with respect to the x, y, z coordinate axes. These are computed as
p_delta = mass * (V_post - V_pre) p_delta_t = p_delta - (p_delta dot N) N Sx = - Sum_i (p_delta_t_x) / (A * dt / fnum) Sy = - Sum_i (p_delta_t_y) / (A * dt / fnum) Sz = - Sum_i (p_delta_t_z) / (A * dt / fnum)
where p_delta, V_pre, V_post, N, A, dt, and fnum are defined as before. P_delta_t is the tangential component of the change in momentum vector p_delta of a particle. P_delta_t_x (and y,z) are its x, y, z components.
The ke value calculates the kinetic energy flux Eflux imparted to the boundary by particles in the group, such that energy lost by a particle is a positive flux. This is computed as
e_delta = 1/2 mass (V_post^2 - V_pre^2) Eflux = - Sum_i (e_delta) / (A * dt / fnum)
where e_delta is the kinetic energy change in a particle, whose velocity changes from V_pre to V_post when colliding with the boundary. The energy flux imparted to the boundary is the sum over all contributing e_delta, normalized by A = the area of the boundary face and dt = the timestep and fnum = the real/simulated particle ratio set by the global fnum command.
The erot value calculates the rotational energy flux Eflux imparted to the boundary by particles in the group, such that energy lost by a particle is a positive flux. This is computed as
e_delta = Erot_post - Erot_pre Eflux = - Sum_i (e_delta) / (A * dt / fnum)
where e_delta is the rotational energy change in a particle, whose internal rotational energy changes from Erot_pre to Erot_post when colliding with the boundary. The flux equation is the same as for the ke value.
The evib value calculates the vibrational energy flux Eflux imparted to the boundary by particles in the group, such that energy lost by a particle is a positive flux. This is computed as
e_delta = Evib_post - Evib_pre Eflux = - Sum_i (e_delta) / (A * dt / fnum)
where e_delta is the vibrational energy change in a particle, whose internal vibrational energy changes from Evib_pre to Evib_post when colliding with the boundary. The flux equation is the same as for the ke value.
The etot value calculates the total energy flux imparted to the boundary by particles in the group, such that energy lost by a particle is a positive flux. This is simply the sum of kinetic, rotational, and vibrational energies. Thus the total energy flux is the sum of what is computed by the ke, erot, and evib values.
Output info:
This compute calculates a global array, with the number of columns equal to the number of values times the number of groups. The ordering of columns is first by values, then by groups. I.e. if the n and u values were specified as keywords, then the first two columns would be n and u for the first group, the 3rd and 4th columns would be n and u for the second group, etc. The number of rows is 4 for a 2d simulation for the 4 faces (xlo, xhi, ylo, yhi), and it is 6 for a 3d simulation (xlo, xhi, ylo, yhi, zlo, zhi).
The array can be accessed by any command that uses global array values from a compute as input. See Section 6.4 for an overview of SPARTA output options.
The array values will be in the units appropriate to the individual values as described above. N is unitless. Press, shx, shy, shz are in pressure units. Ke, erot, evib, and etot are in energy/area-time units for 3d simulations and energy/length-time units for 2d simulations.
Styles with a kk suffix are functionally the same as the corresponding style without the suffix. They have been optimized to run faster, depending on your available hardware, as discussed in the Accelerating SPARTA section of the manual. The accelerated styles take the same arguments and should produce the same results, except for different random number, round-off and precision issues.
These accelerated styles are part of the KOKKOS package. They are only enabled if SPARTA was built with that package. See the Making SPARTA section for more info.
You can specify the accelerated styles explicitly in your input script by including their suffix, or you can use the -suffix command-line switch when you invoke SPARTA, or you can use the suffix command in your input script.
See the Accelerating SPARTA section of the manual for more instructions on how to use the accelerated styles effectively.
Restrictions:
If specified with a kk suffix, this compute can be used no more than twice in the same input script (active at the same time).
Related commands:
Default: none