Posted Fri, 24 Feb 2017 08:21:47 GMT by Alex Costall PhD Student
Hi again,

Sorry Bjorn, I have managed to confuse myself once again.

Given a 2D system with a Richards-style equation, I want to model a known inflow of fresh water from one side of the model.

Do I apply the known value of recharge to each cell on the border?
i.e. The same number regardless of number of cells or total length?


Do I need to account for the number of cells, and the depth range which those cells span?

For example, in the picture below - do I need to apply -0.5 m/d for each of the nodes in the target area, or do I need to apply -(0.5/8) to each of the nodes, to make a total of -0.5 m/d?

I suspect this is a fundamentally obvious question, and I apologise in advance.
Posted Fri, 24 Feb 2017 10:35:45 GMT by Björn Kaiser
If your model has a cross-sectional projection you may use NeumannBC's. As you already indicated correctly, NeumannBC's have the unit of a velocity. The amount of water entering the system is the integral along the element edges and a unit depth of one meter. You are also right by using a negative algebraic sign for the NeumannBC's if you want to have an inflow.
Posted Fri, 31 Aug 2018 10:47:47 GMT by LIONEL SCHAPER
Hi Bjorn,

You said the inflow is computed by multiplying the flux by the length and a depth of 1 meter. How does it work when we are working on an axisymetric model : how is the inflow calculated ?

Thank you
Posted Mon, 03 Sep 2018 12:10:27 GMT by Björn Kaiser
In 2D axisymmetric projection the flow is not integrated along the length and unit depth of one meter. Instead, the flow is integrated over the area spanned over the rotation axis.

You must be signed in to post in this forum.