Hi there,

I work with the user subroutine SORSCA and use SINK and SOURCE terms for a scalar(1). With some grids everything is fine, but sometimes when I use a different grid NEGATIVE SCALAR VALUES occur and the whole calculation result makes no sense. I already tried different time step sizes for the transient analysis and different residual tolerance values for the scalar. Does anybody know why this happens or did anyone observe the same?? Thanks in advance Mario

With the version v3100 I observed negative scalars in transient runs with the MARS discretization scheme (sometimes the negative value could be of the same order as the maximum positive values !!). The remedy for this was to use GAMMA (bf=0.1) for the scalars.

With the new v3150 there has been no such problems with MARS scheme anymore (excluding very small negative values for the scalars, maximum of the order of -1E-7).