What I mean to do is to model the water proof fault which divided the the research area into two parts.
The research area is simple, a flat confined aquifer (rectangular) , east and west boundary is Dirichlet Boundary and the northern and southern boundary is 0 flux Neuman boundary. No wells and no recharge.
I divided the research area into two parallel parts (the eastern and the western part) using a line. Then I built 3-D mesh separately for east part by chosing the polygon of the east part and building 2-D mesh for it.
Then , I did the same thing to the western part, assigned the same mesh layers as the layers of eastern part, so that each nodes on central boundary has two IDs.
However the computing result is not as what I expected, which might be incorrect. Although the pressure head at the same node on the fault is different for both IDs, There is still flux from west to east.
I checked the node ID, found that although I construct 3-D mesh to each part of the aquifer, each node on the top and bottom of the aquifer at the place of fault still have one ID.
My question is: Is this a bug of GMS or is this method to construct mesh to each part separatedly wrong? Why does the nodes on the top and bottom have one ID while others on the fault have two?
Thank you very much!