Moving Boundary Simulations

Check the grid distribution at the interfaces. If the grid is too coarse or too fine-grained then the fluxes may not be defined well.

Use activities instead of potentials when the content of one or more of the elements becomes very low. If you do this, then the activity approaches zero, which is easier to handle than a potential that approaches infinity.

In Console Mode, use SET_SIMULATION_CONDITION to set the parameter Vary potentials or activities to Activities. 

In Graphical Mode the setting is on the Diffusion Calculator in the section Classic model specific. From the list, choose activity [Use the activity of a component in order to find the correct tieline at the phase interface].

Use the automatic starting values when you run the simulation the first time. If the first time step does not converge try decreasing the starting values for the velocities. Alternatively, increase the duration of the first time step.

If the simulation fails during the first time step, then check how the starting values affect the residuals.

In Console Mode, use the command SET_SIMULATION_CONDITION and set the parameter NS01A PRINT CONTROL to 1.

When you adjust this setting, the software then prints the tried unknown value and the residual for each flux balance equation in each iteration. The equations are organized so that the first unknown value is always the migration rate of the phase interfaces followed by the unknown potentials. All unknown values are scaled by the starting value or the calculated value in the previous time step.