Tracking offers the possibility to extract data from a simulation for given subsections of the mesh and specific points in time.
When the data is given in form of polynomials, it usually is required to
perform some kind of sampling to obtain a visualization.
The ply_sampled_tracking
implements this functionality.
See ply_sampling_module for details on the configuration of the sampling
strategy and tem_tracking_module for details on the general configuration
of trackings (description of mesh subsections to extract data for).
An important special case is the tracking of a single point.
In this case no sampling has to be done, instead the polynomial representing
the state in the element that contains the requested point needs to be
evaluated at this point. This can be configured by setting
use_get_point=true
in the output subtable of the tracking object.
A complete example looks like this:
tracking = {
label = 'track_shearhat2D',
folder = './',
variable = {'momentum','density','energy'},
shape = {
kind = 'canoND',
object= {
origin ={0.01*dx, 0., 0.}
}
},
time_control = {
min = 0,
max = sim_control.time_control.max,
interval = {iter = 10}
},
output = {
format = 'ascii',
use_get_point = true
}
}
This tracks the state variables (momentum, density and energy) in a single
point (0.01*dx, 0, 0)
and writes them every ten iteration to an ASCII file.
Each point in time gets written on a new line in the same file.
If you do not use the use_get_point
option (its false
by default), and no
sampling is active, all degrees of freedom of the field in the element, that
contains the point, will be written.
You can limit the number of degrees of freedom by setting ndofs
to some
value. The first mode of the Legendre series is the integral mean, so this is
usually the value you want to get. Thus, setting ndofs=1
gives you the
averaged value of the element the point is found in.
The according output table would then look as follows:
output = {
format = 'ascii',
ndofs = 1
}
Of course, more complex shapes may be tracked, in that case it usually is
not sensible to use ascii
output anymore. Instead you are than likely to
want to view data later on and accordingly write it in VTK format.
For this, a sampling table (see ply_sampling_module) should be
considered.
Type | Visibility | Attributes | Name | Initial | |||
---|---|---|---|---|---|---|---|
type(tem_tracking_type), | public | :: | tracking |
Contains all tracking headers, control and tracking entities active on local process |
|||
type(treelmesh_type), | public, | allocatable | :: | mesh(:) |
Subsampled mesh for each tracking. |
||
type(tem_varSys_type), | public, | allocatable | :: | varsys(:) |
Variable system description after subsampling. |
||
type(ply_sampling_type), | public | :: | sampling |
Configuration of the subsampling (applied to all trackings). |
|||
integer, | public | :: | ndims |
Dimensionality of the data to sample. |
Load the configuration of sampled tracking objects.
Type | Intent | Optional | Attributes | Name | ||
---|---|---|---|---|---|---|
type(ply_sampled_tracking_type), | intent(out) | :: | me |
Sampled tracking data to load from the config |
||
type(flu_State) | :: | conf |
Lua config to load the tracking from |
Initialize the sampled tracking entities.
Type | Intent | Optional | Attributes | Name | ||
---|---|---|---|---|---|---|
type(ply_sampled_tracking_type), | intent(inout) | :: | me |
Sampled tracking variable to initialize. It has to be configured by ply_sampled_tracking_load beforehand. |
||
type(treelmesh_type), | intent(in) | :: | mesh |
The global mesh. |
||
type(tem_solveHead_type), | intent(in) | :: | solver |
Information about the solver (used to construct file name strings). |
||
type(tem_varSys_type), | intent(in) | :: | varSys |
Global variable system with description of the data to get the tracking variables from. |
||
type(tem_BC_prop_type), | intent(in) | :: | bc |
Boundary condition properties, used to identify elements close to the boundary. |
||
type(tem_stencilHeader_type), | intent(in), | optional | :: | stencil |
Description of the stencil in the numerical scheme. |
|
type(tem_comm_env_type), | intent(in) | :: | proc |
General communication environment |
||
integer, | intent(in) | :: | nDofs |
Number of degrees of freedom to use in the output. |
||
integer, | intent(in) | :: | nDims |
Number of dimensions in the polynomial representations. |
Output sampled tracking data.
Type | Intent | Optional | Attributes | Name | ||
---|---|---|---|---|---|---|
type(ply_sampled_tracking_type), | intent(inout) | :: | me |
Sampled tracking instances. |
||
type(treelmesh_type), | intent(in) | :: | mesh |
Global mesh, required for the sampling. |
||
type(tem_BC_prop_type), | intent(in) | :: | bc |
Boundary properties, needed to inherit boundary information to refined meshes and allow the extraction of boundary shape geometries. |
||
type(tem_solveHead_type), | intent(in) | :: | solver |
Information about the solver, needed for the output file name. |
||
type(tem_comm_env_type), | intent(in) | :: | proc |
General communication environment |
||
type(tem_varSys_type), | intent(in) | :: | varSys |
Original variable system |
||
integer, | intent(in) | :: | var_degree(:) |
Maximal polynomial degree for each variable |
||
integer, | intent(in) | :: | lvl_degree(:) |
Maximal polynomial degree for each level |
||
integer, | intent(in) | :: | var_space(:) |
Maximal polynomial space for each variable |
||
type(tem_simControl_type), | intent(in), | optional | :: | simControl |
Simulation control to determine, whether trackings should be written |
|
type(tem_time_type), | intent(in), | optional | :: | time |
Provide a time for the current data set to write in tracking. |