Convergence problem

A discussion forum for users of the new HYDRUS 2D/3D. HYDRUS is a software package for simulating water, heat and solute movement in two- and three-dimensional variably saturated media. Happy Posting!
Post Reply
PhamTien
Posts: 1
Joined: Sat May 19, 2018 4:17 pm
Location: Japan

Convergence problem

Post by PhamTien » Thu Jun 07, 2018 6:09 am

Dear Professor Jirka and everyone,

Hello,
I am doing a project on simulating the water flow infiltration within a 2D real-scale hillslope with two different soil layers. The upper layer is clay and another is hard bedrock with a very clay material. The river is located on the left side of the hillslope as the real on-site condition.
In the model, time variable boundary condition was specified with hourly rainfall data during 93 days (a total of 2232 hours). I set the boundary conditions as follows: the top and bottom boundary condition are Atmospheric and No Flux, respectively; the right side of the slope is No Flux; the left side of the slope near the river is Gradient boundary condition (the gradient parallel with the slope).
Although I tried to modify the model with various initial conditions and boundary conditions, the model crushed in between due to the convergence problem always.

I would be much appreciated if you can help me to check the model and give me suggestions to solve the problem.
Do you think the specified boundary conditions and parameters in the time-variable boundary condition (except for rainfall) are all right?

I attached herewith the file: https://drive.google.com/file/d/1FaQ-AJ ... sp=sharing

Thank you very much in advance.

Best regards,
Phamtien

Jirka
Posts: 4620
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Convergence problem

Post by Jirka » Thu Jun 07, 2018 6:43 pm

I do not have time to look at the project this week. Perhaps next week.

You need to understand that the Richards equation is a highly nonlinear equation and thus there is no guarantee that the code will converge for all situation. The convergence f the code can be increased by sticking with some basic rules for temporal and spatial discretization as described in http://www.pc-progress.com/Documents/No ... zation.pdf. I may also be good to first work with HYDRUS-1D and find out what discretization is required for your particular soil hydraulic properties and initial and boundary conditions.

J.

Post Reply