Problem simulating infiltration

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
ari
Posts: 20
Joined: Thu Jul 28, 2016 4:16 pm
Location: USA

Problem simulating infiltration

Post by ari » Sat Sep 16, 2017 10:46 pm

I am trying to simulate simple infiltration in a 10m by 11m (xz) domain with two sub-regions (sand and loamy sand) for 1000 years with a yearly precipitation rate of 0.5m/year with a solute concentration of 500mg/cubic meter, evaporation rate of 2.04m/year and transpiration rate of 0.876m/year. The initial Pressure Head is -100m at the top of the domain (ground surface) and 0m at the bottom at the water table (linear distribution with depth). The initial concentration is 0mg/m3.
I am having problems running the simulation and obtaining a vertical concentration profile. So I would like to ask few questions and your answers and suggestions will be appreciated.

1. When I am trying to assign the Water Flow and Solute Transport in the Boundary Condition on FE-mesh, I am getting the error message “There are differences between values of the current property (or condition) on FE-Mesh and those on Geometric Objects.” When I tried to transfer values from Geometric Objects to FE-Mesh, my Constant Flux boundary switched to No Flux Boundary and then I had to convert them back to Constant Flux boundary manually. What is the best way to fix this problem?

2. If I want to add a solute concentration with my constant flux water flow in this scenario, where should I specify it? Is it at the Boundary Conditions cBnd1 on Reaction parameters for Solute-1 window (assigned 500 mg/m3) or cValue1 (assigned 0.5 mg/L) at the Time Variable Boundary Condition?

3. While I am trying to add observation points in my domain, the cursor doesn’t work when I am trying to add those points on geometric objects, but it works on FE-mesh (but not in a straight line, only according to the position of FE-nodes). What is the best possible way to add observation points along a straight line?

I am attaching the zipped file (Test.zip) which has the .h3d2 file for your convenience. Please take a look if possible and let me know what modifications are required. I will really appreciate the feedback.

Thanks in advance,
A
Attachments
Test.zip
(754.21 KiB) Downloaded 110 times

Jirka
Posts: 4762
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Problem simulating infiltration

Post by Jirka » Mon Sep 18, 2017 7:17 am

1. When starting this project, you have selected (in the Domain Types and Units window) that you want to “Edit Domain Properties, Initial and Boundary Conditions on the Geometric Objects”. However, you do not seem to be doing it. The constant Flux BC is defined on the FE Mesh (not on Geo Objects where no BCs are specified). You should decide how you want to specify these properties and then make the selection in the Domain Types and Units window (either on Geometrical objects, or on FE Mesh). Note that the preference will always be given to properties specified on FE Mesh (unless requesting otherwise).
2. For constant head and flux BC, this should be given in cBnd 1-5 in the Solute Reaction Parameters window. For Time-Variable BCs (including atmospheric BC), this is specified as cBnd1-3 in the Time-Variable BC window.
3. Again, your geometric objects are only boundary lines and boundary nodes and thus you cannot specify any Observation Nodes inside of the domain. For that, you need to switch to the “Edit Properties on FE Mesh” Mode.

JS

ari
Posts: 20
Joined: Thu Jul 28, 2016 4:16 pm
Location: USA

Re: Problem simulating infiltration

Post by ari » Tue Sep 19, 2017 3:29 am

Thank you very much for your response. According to your suggestion, I have specified IC and BC's in the FE Mesh, specified constant flux solute concentration in cBnd1 in the Solute Reaction Parameters window. For observation nodes along a vertical line, I have created a series of points along a line in the domain geometry and this time I was able to add observation nodes in the FE Mesh according to my choice of locations. I did not get any error message this time but still don't see the concentration data in the observation nodes. Any suggestions?
Attachments
Test.zip
(537.04 KiB) Downloaded 106 times

Jirka
Posts: 4762
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Problem simulating infiltration

Post by Jirka » Tue Sep 19, 2017 11:56 am

I do not know whether you realize it, but your setup simply does not make physical sense. You have a constant flux at the surface and no flow everywhere else. You will thus fill the domain in some 6-7 years fully with water, while you want to simulate 1000 years and the first print time is after 100 years. No wonder that you do not get any results. If you make some minor adjustments to make the problem physically reasonable, you will get reasonable results and will see concentrations in the outputs. J.
Attachments
Test 10 by 11m.zip
(875.73 KiB) Downloaded 105 times

ari
Posts: 20
Joined: Thu Jul 28, 2016 4:16 pm
Location: USA

Re: Problem simulating infiltration

Post by ari » Wed Sep 20, 2017 1:03 am

Thank you so much for your help. I really appreciate it.

Post Reply