Message Boards

Boundary conditions

Karla Contreras, modified 4 Years ago.

Boundary conditions

Youngling Posts: 1 Join Date: 3/16/14 Recent Posts
Hello everyone
I am modeling a river estuary area. As boundary condition entry downstream water level in a time series and as upstream boundary condition in a total discharge time series. But as a result I see that the river is empty, it's like I was not taking on water.
Change the sign of the inflow, by the increasing number of nodes, but the result is the same

If I could help I would appreciate it.



attached model


Richard Measures, modified 4 Years ago.

RE: Boundary conditions

Jedi Knight Posts: 178 Join Date: 3/23/11 Recent Posts
Hi Karla,

Do you have any water at the river boundary in your initial condition. Total discharge boundary conditions only work when the boundary is wet at some point along its length. When simulating rivers I usually create an *.ini initial conditions file with a sloping water level or even just a small puddle of water at the upstream boundary to get the boundary going. You can generate *.ini files to do this quite easily using quickin and a text editor by manually specifying water elevation and then setting zero velocity everywhere (see definition of ini file in Appendix A of the manual).

It would be a useful improvement to the total discharge boundary condition if it was modified (in the source code) to allow starting from a dry initial condition and probably not too hard to implement if you or anyone else feels inspired to make the change :-)

meddy danial, modified 1 Year ago.

RE: Boundary conditions

Youngling Posts: 19 Join Date: 7/25/18 Recent Posts

Dear Richard

Just a short question. Can I only use Initial condition of uniform value instead of *.ini files. What is the impact of numerical result if I use initial condition with uniform value compared to using *.ini file?

Thanks in advance


Arjen Markus, modified 1 Year ago.

RE: Boundary conditions

Jedi Knight Posts: 223 Join Date: 1/26/11 Recent Posts

In general - within the sort of problems we solve with Delft3D - you do not want the initial conditions to influence the result over the long run. The initial condition is mostly a nuisance, especially if you have a tidal flow or some other form of external forcing. (It is different if you want to model a transient phenomenon like a dam break or a tsunami). For a parameter like the water level, the transient part dies out within a few tidal cycles. For salinity and temperature this is likely to take much longer.

That said, the numerical difference between a uniform value and a value that may vary over space is of impact only during the time of transition.




meddy danial, modified 1 Year ago.

RE: Boundary conditions

Youngling Posts: 19 Join Date: 7/25/18 Recent Posts

Dear Arjen Markus

Thank you for your reply. Nice answer.