Mesh Lines and confirming cumulative boundary flux rates

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!
mortsde
Posts: 7
Joined: Mon Sep 08, 2014 10:38 pm
Location: USA

Mesh Lines and confirming cumulative boundary flux rates

Post by mortsde » Fri May 24, 2019 5:37 am

Jirka and Mirek,

Thanks in advance and apologies for asking a question that probably has been asked and answered many times.

I'm interested in applying unit gradient boundary conditions at multiple locations and I'm wondering if there is a way to separate flux rates at the boundary. I'm simulating a lysimeter with lateral drainage measurements and deep percolation measurements.

I've tried using mesh-lines at each boundary location - however - I have noticed other users and Jirka has replied that mesh-lines are calculated with less precision than boundary conditions. I also recall that boundary condition types (free drainage - seepage faces - etc.) cannot be calculated separately, even if multiple conditions exist.

I'm hopeful you have suggestions for creative solutions to this problem and/or can correct my misconception. For the record, WatBalR is low at approximately 0.3%.

Again - thanks for your time.
Alex

Jirka
Posts: 4824
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Jirka » Fri May 24, 2019 1:09 pm

First, I think you can use mesh-lines for this (especially if you want to have both actual and cumulative fluxes). I do not think that the fluxes across meshlines are calculated with less precision. The same approach is used for those calculations as for calculations of boundary fluxes. You can define up to four meshlines.

Second, in the Boundary.out file, you can find nodal fluxes for all boundary nodes at a particular print time. If you add fluxes for selected nodes, you should get your actual boundary flux. However, in this way you will not get the cumulative flux.

J.

mortsde
Posts: 7
Joined: Mon Sep 08, 2014 10:38 pm
Location: USA

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by mortsde » Fri May 31, 2019 11:32 pm

Thanks for the reply Jirka.

Otthman
Posts: 6
Joined: Thu Nov 07, 2019 8:04 pm
Location: Canada

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Otthman » Mon Nov 11, 2019 7:52 pm

Hi dear Jirka,

Is there a difference between tracing the meshline horizontally from left to right and from right to left ?

Jirka
Posts: 4824
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Jirka » Mon Nov 11, 2019 8:56 pm

Meshlines are oriented depending on how you define them. The fluxes should be the same and independent of orientation except for the sign (+/-). J.

Otthman
Posts: 6
Joined: Thu Nov 07, 2019 8:04 pm
Location: Canada

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Otthman » Mon Nov 11, 2019 9:14 pm

ok very well thank you. I have to do 05 meshline and the whole thing does not work in the same. When I change the third mesheline in another direction it works. But this does not suit me in the rest because I do have sign problems.

Jirka
Posts: 4824
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Jirka » Tue Nov 12, 2019 5:01 am

I do not see why that would be a problem. If you reverse the points defining the meshline, you change the sign (although I do not see why you would do that). J.

Otthman
Posts: 6
Joined: Thu Nov 07, 2019 8:04 pm
Location: Canada

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Otthman » Tue Dec 17, 2019 10:13 pm

Good evening Mr. Jirka

In my modeling, I have 05 mesh lines. Qunad I run the simulation with the 05 mesh lines, an error is displayed.
But when I delete the fourth mesh line the calculation is done correctly.

Thank you

Jirka
Posts: 4824
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Jirka » Tue Dec 17, 2019 11:05 pm

Indeed, there seems to be an error in GUI when 5 or more meshlines are used. The program works well when there are up to 4 meshlines. We need to check this bug. Sorry for that. In the meantime, use only up to 4 meshlines. J.

Mirek
Posts: 414
Joined: Sat Mar 16, 2002 4:00 pm
Location: Czech Republic
Contact:

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Mirek » Wed Dec 18, 2019 11:35 am

This error has been fixed. New installation programs Hydrus3D_3.02.0520_x86.exe and Hydrus3D_3.02.0530_x64.exe are available in ELIS.

Otthman
Posts: 6
Joined: Thu Nov 07, 2019 8:04 pm
Location: Canada

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Otthman » Wed Dec 18, 2019 11:05 pm

Thank you all for your different answers.
I have attached an image of the error message.
Maybe this will bring a lot of clarity. I use version 2.04 of hydrus.
Mr. Mirek means ELIS
Attachments
Nouveau document 2019-12-18 13.44.07_1.jpg
Nouveau document 2019-12-18 13.44.07_1.jpg (958.66 KiB) Viewed 168 times

Jirka
Posts: 4824
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Jirka » Thu Dec 19, 2019 12:07 am

I should clarify. I found a bug in Version 3.02, and that is the bug Mirek fixed. I have not seen the error in version 2.05. I do not have anymore the version 2.04. J.

Mirek
Posts: 414
Joined: Sat Mar 16, 2002 4:00 pm
Location: Czech Republic
Contact:

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Mirek » Thu Dec 19, 2019 7:29 am

Yes, this error was only fixed in the latest version of HYDRUS (which is 3.02.xxxx). HYDRUS 2.x is no longer maintained.

Jirka
Posts: 4824
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Jirka » Thu Dec 19, 2019 5:42 pm

I have tested both version 2.04 and 2.05 with more than 6 meshlines, and with 3 solutes, and the code was working fine for me. J.

Otthman
Posts: 6
Joined: Thu Nov 07, 2019 8:04 pm
Location: Canada

Re: Mesh Lines and confirming cumulative boundary flux rates

Post by Otthman » Tue Jan 14, 2020 2:41 pm

Hello Jirka and Mirek,
I think I understand the source of the problem. I have five models (lysimeter in a cell). The first four works. The five cases are differentiated by the dimensions of the lysimeter. The fifth case has the largest lysimeter. I think it doesn’t work because of the amount of data.
Is the finding made a possibility?

Thank.
Last edited by Otthman on Thu Jan 16, 2020 3:57 pm, edited 1 time in total.

Post Reply