Problem with results

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
Noemie
Posts: 3
Joined: Thu Jun 10, 2021 9:11 am
Location: France

Problem with results

Post by Noemie » Thu Jun 10, 2021 10:08 am

Dear all,

I a begginer on Hydrus, I have very low experience on 1D and I just start to use Hydrus 2D/3D. Also I'm sorry if I use simple word, I'm not English but I'm doing my best.
I try to study a system of agricultural terrace on Hydrus 3D. I work on windows with Hydrus 2.05.
My system simulate water flow and Root water Uptake for 10 days.
I place water root uptake with the modele Feddes and choose the data of oranges.
I've got two layer material. Precipitation on the first four days and 0 precipitation after (because I want to see the evolution of the simulation).
I've put hydraulic conductivity of gravel on the "walls" (0.001) and of a clay on the sides because I don't want my water to go on the side (10^-9)
I've implant atmospheric BC on the surface, free drainage on the "walls" and below, and no flux on the side.

The simulation run successfully... but there is a problem, at 2 days, the runnig simulation seems to meet a obstacle because the system doesn't evolve after.
In the section "Result _ Graphical display", all the display doesn't work after 2 days like a saturation of the system. But it can't be saturated because after three days, there is no precipitation ! And the model doesn't evolve at all.
3days.PNG
3days.PNG (312.54 KiB) Viewed 164 times
I tried to change the time step because I read on the forum that it could be the problem but it doesn't change anything.

Thanks for the help !
Attachments
HydrusCalculation.PNG
HydrusCalculation.PNG (37.31 KiB) Viewed 161 times
HydrusCalculation.PNG
HydrusCalculation.PNG (37.31 KiB) Viewed 163 times

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Problem with results

Post by Jirka » Thu Jun 10, 2021 4:16 pm

Hi,

If you do not have experience with 3D modeling, you should first start with 2D. It runs faster and you will get some experience.

It seems that your project has many problems.
1. I do not know what you refer to by the terms “walls”, where you give gravel properties and free drainage (gravitational vertical flow). Note that free drainage cannot be given on vertical sides. It can be specified only at the bottom.
2. You get non convergence on day 2 for some reason. I obviously cannot know why from the minimum information I have about the project. However, clearly your service and root zone is fully saturated (zero pressure at hAtm and hRoot). It seem that you are putting more water on the domain than it can handle.
3. Your description of the problem is confusing and contradictory: “Precipitation on the first four days and 0 precipitation after” and “But it can't be saturated because after three days, there is no precipitation”.

J.

Noemie
Posts: 3
Joined: Thu Jun 10, 2021 9:11 am
Location: France

Re: Problem with results

Post by Noemie » Thu Jun 10, 2021 8:10 pm

Hi Jirka,

By walls, I refering to Stone wall that old the ground at the end of some agricultural terrace. So, thanks for tell me that free drainage cannot apply on vertical sides, I will correct that. By applying this, I thougt that the system would understand that water can be drain by wall (The stones wall would act as a filter / drain system).
For my precipitation, I'm sorry for the horrible description, I made mistake, I've put
Days P (m/j)
1 0,0081
2 0,0388
3 0
4 0,0167

And after, it's only 0 m/j for précipitation for the Last 6 days. Only three days has precipitation but on four day so I got mixed up, sorry.
For the water content, do the system not take it on the soil parameter ?
I do not find water content in the soil parameter but It makes sense to me that it's somewhere (or I miss it ? I would check tomorrow at work).
I put Loam and Sandy Loam for soil so I thougt that the system would empty with time (this is why I want to make 10 days of simulation).
Hope I explain better and really thanks for your help !

Noémie

Jirka
Posts: 5268
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Problem with results

Post by Jirka » Thu Jun 10, 2021 11:28 pm

As I stated above, I would still recommend to start with a 2D cross-section, before moving to a fully 3D model (if it is even required).
1. If water can drain through the stone wall, you can use a Seepage Face BC there. However, you need to understand that this water will leave the domain (and not drip on the lower section.
2. The water content is not the soil hydraulic parameter. You can specify initial water contents (using initial conditions).
3. When the rain is 3 cm/d and you have higher Ks than that, you should have no problem to model this system. If Ks is lower than 3 cm/d, which your results seem to indicate since you have surface pressure head equal to zero, you should also not have any problem (if you use adequate spatial and temporal discretization).

J.

PS: This is my last post for a week, as I’m going hiking to Sierra Nevada tomorrow.

Noemie
Posts: 3
Joined: Thu Jun 10, 2021 9:11 am
Location: France

Re: Problem with results

Post by Noemie » Fri Jun 11, 2021 7:18 am

Thanks again for your help, I will use 2D model as you recommended before moved to 3D model.
I had tought Seepage Face BC would be adapted but I was advised to rather put Free drainage, I see that my hunch was true. Too bad that the water leave the domain but that's how it is.

Good hiking !
Noémie

Post Reply