The input line flows can also be edited on the desktop with the help of active objects.
Flow boundary conditions must be defined on all boundary lines. The following boundary conditions can be specified:
Pore pressure on a given line is equal to zero
c) Pore pressure
- distribution of pore pressure p can be specified numerically
- distribution of pore pressure can be specified by inputting the location of the groundwater table (by prescribing the total head h)
d) Inflow/outflow on a line q - it is specified in velocity units e.g. [m/day, ft/day] - the flow velocity into/out of the region is specified. The default setting corresponds to an impermeable boundary for which q = 0.
e) Seepage surface - this boundary condition is introduced providing the boundary cannot be uniquely divided into the part with prescribed pore pressure and the part with prescribed inflow/outflow (the exit point is not known). In such a case the analysis is performed in two steps. In the first step, the program locates the exit point. The actual flow analysis with known boundary conditions is then carried out in the second step. In some cases, both steps must be repeated several times. When enhanced input is considered the program requires entering a fictitious permeability kv in units [m/day]. This is essentially a penalty term, a sufficiently large number in general, ensuring that along an impermeable boundary the value of total h will be equal to the y-coordinate of a given point (q = 0). For a part of the boundary with no flow condition, we have kv = 0. Variables q and h are then related by:
if h > 0 (S = 1) inside soil body
if h < 0 (S < 1) inside soil body
Note: If in case of transient flow we directly define in the first calculation stage the location of the groundwater table (phreatic surface) as an initial condition, we should define in the next calculation step along the boundaries below the water level in the region of fully saturated soil a boundary with the prescribed pore pressure having a corresponding value and not the seepage surface. In the case of seepage surface, the program would immediately label this boundary as a boundary with zero pore pressure and not the originally assumed boundary with the pore pressure distribution in accordance with the expected height of GWT. In such a case, the analysis will not converge, because water, instead of flowing out of the domain at p = 0, will have a tendency to flow in, which not realistic.