Default:
VOLUME_correction_FreeSurface = 0.0 (off)
The given value is the maximum allowed corrected volume per time step, based on the total volume of a chamber.
Note: This parameter can also be set chamberwise for multiphase simulations (see also
KindOfProblem,
CHAMBER). If it is not set for specific chambers, it is automatically set according to the non-chamberwise definition for all chambers.
If the volume correction for multiple chambers shall be different, use
which sets the correction for all chambers first to 0.001, then it changes the values for chambers 3 and 5.
Representative masses are switched OFF
If the
RepresentativeMassAlgorithm is deactivated, for this type of volume correction, we first compute the potential displacement (distance \( D_\text{pot}\)) of the free surface by
\begin{align} D_\text{pot} = min\left( \alpha , \frac{V_\text{target}-V_\text{current}}{V_\text{current}} \right) \cdot \frac{V_\text{current}}{A_\text{FreeSurface}}\end{align}
and then move, in every time cycle, the free surface artificially by the distance
\begin{align} D_\text{move} = \min\left( 0.01 \cdot H , D_\text{pot} \right)\end{align}
Here, \( \alpha\) is equal to the value of
VOLUME_correction_FreeSurface .
The target volume as well as the current volume are given by the values provided through the
real()-functionality, see the keywords
%VOLUME_TARGET% and
%VOLUME_ACTUAL% . Thus, it becomes clear that the correction can be effected chamberwise, only.
Representative masses are switched ON
In case, the
RepresentativeMassAlgorithm is switched on, the limitation is as follows:
\begin{align} D^{k^\text{chamber}}_\text{pot} = \frac{V^{k^\text{chamber}}_\text{target}-V^{k^\text{chamber}}_\text{current}}{V^{k^\text{chamber}}_\text{current}} \cdot \frac{V^{k^\text{chamber}}_\text{current}}{A^{k^\text{chamber}}_\text{FreeSurface}}\end{align}
\begin{align} D^{k^\text{chamber}}_\text{move} = \min\left( \alpha \cdot H , D^{k^\text{chamber}}_\text{pot} \right)\end{align}
If the
RepresentativeMassAlgorithm is activated, the computation of the target volume is straight forward
\begin{align} V^{k^\text{chamber}}_\text{target} = \sum \limits_{i, i \in \Omega(k^\text{chamber})} { \frac{\overset{\scriptscriptstyle\frown}{m}_{i}}{\rho_i} }\end{align}
\begin{align} V^{k^\text{chamber}}_\text{current} = \sum \limits_{i, i \in \Omega(k^\text{chamber})} { V_i }\end{align}
and \( V_i\) is the value found in
%ind_Vi% .
If, moreover, the clustering of the point cloud is activated (see
SCAN_ClustersOfConnectivity), the target volume and also the free surface corrections are computed clusterwise, i.e.
\begin{align} V^{k^\text{cluster}}_\text{target} = \sum \limits_{i, i \in \Omega(k^\text{cluster})} { \frac{\overset{\scriptscriptstyle\frown}{m}_{i}}{\rho_i} }\end{align}
\begin{align} D^{k^\text{cluster}}_{\text{pot}} = \left( \frac{V^{k^\text{cluster}}_\text{target}-V^{k^\text{cluster}}_\text{current}}{V^{k^\text{cluster}}_\text{current}} \right) \cdot \frac{V^{k^\text{cluster}}_\text{current}}{A^{k^\text{cluster}}_\text{FreeSurface}}\end{align}
Additionally using localized representative density
If using version 1 of
DefinitionRepresentativeDensity , then the potential movement of the free surface is enhanced
\begin{align} D^{k^\text{cluster}}_{\text{pot,enhanced},i} = D^{k^\text{cluster}}_{\text{pot}} + \beta \cdot \min \left( D^{k^\text{cluster}}_\text{pot} , \frac{ \overset{\scriptscriptstyle\frown}{\rho }_{i}-\rho}{ \rho } H_i \right)\end{align}
which attempts to equalize representative volume at the free surface. The standard is \( \beta=1\). However, if
VOLUME_correction_FreeSurface is chosen bigger that 1, then we define \( \beta = \frac{ {\text floor}( {\text VOLUME correction FreeSurface} ) } {100}\), and \( \alpha\) is given by the after-comma-digits, as this parameter has to be smaller than 1 anyways.
In general:
The potential movement is displayed in the variable
%ind_BNDfree_defect%, representing \( \frac{D^{k^\text{cluster}}_\text{pot}}{H_i}.\)
See
VolumeCorrection for more information on volume correction.