Posted Wed, 30 Nov 2016 13:37:51 GMT by Dipl. Ing. Kirnbauer
Dear FEFLOW-Users!

I have a 2d unconfined steady state model with a river as transfer boundary. I want to apply a synthetic flood wave, where I rise the water level in the river bed e.g. 3m in 12 hrs.
The water level of the transfer boundary in the model has a high resolution along the river length.
Is there a way to define only ONE Power Function with the relative water level change vs. time and add that value to the starting absolut water level value for every node in the transient simulation?
Can this be done with the calculator?
Otherwise I would have to define one power function per node along the river with absolut water level values.

I use Feflow 6.1.

Thanks for help,
Posted Thu, 01 Dec 2016 15:30:08 GMT by Björn Kaiser
If the time-dependent water level of the  floodwave has been measured at different point-wise locations you may interpolate these curves between the locations of the measurements.
Posted Sat, 03 Dec 2016 10:37:46 GMT by Dipl. Ing. Kirnbauer
The water level of the river is the result of a 1d-hydraulic Simulation with many weirs and transvers structures. So with a linear interpolation along the river length I would get a large error in water level if I use only few points and interpolate over transvers structures.
As the stage-discharge-curve is more or less similar along the river length, I think the better way would be to start with the water level of the 1d-simulation and add the relative change in water level as described.

Is this possible or do I have to generate a large number of power functions at the upstream and downstream side of the transverse structures and linear interpolation between the structures?
Posted Tue, 13 Dec 2016 11:26:20 GMT by Björn Kaiser
In principle, you can define a large number of power functions and assign them to boundary conditions. Then you may decide whether or not you want to interpolate.

You must be signed in to post in this forum.