floating point error when doing tutorial

A discussion forum for old Hydrus-2D & Meshgen. Happy Posting!
Download PDF document with all topics (February 8, 2008, 1.9MB):
http://www.pc-progress.com//Downloads/F ... on_H2D.pdf
Post Reply
hkiela
Posts: 2
Joined: Tue Jan 26, 2010 9:44 am
Location: Netherlands

floating point error when doing tutorial

Post by hkiela » Tue Jan 26, 2010 11:17 am

hello,

I am only starting with hydrus, so I started on doing the tutorials on the site; especially with session 4a water flow to a stream, since it is the most similar to what I intend to model. I opened a new project and filled everything in as it is stated.

when I now try to run the model I get the following (error):
vertical plane flow, V = L*L
reading nodal information
reading element information
reading geometric information
reading material information
generating materials
Floating Point Error - Invalid number!

Am I right to conclude that it is to do with the soil characteristics, since it stops afterwards? Like in the tutorial I used loam as my soil species (from drop down menu).

When in the soil hydraulic parameters window, I always need to press 'ok' and then go manually to the geometry and mesh editor, or the program will shut down. Might this be caused by the same problem?

I am running hydrus2.05 under Vista (since I'm a student I cannot upgrade the edition, I'll have to do with what university has licenced...)

thanks for the help,
cheers,
Heleen

Jirka
Posts: 4790
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Post by Jirka » Tue Jan 26, 2010 2:53 pm

Can you run examples that come with HYDRUS? J.

hkiela
Posts: 2
Joined: Tue Jan 26, 2010 9:44 am
Location: Netherlands

Post by hkiela » Wed Jan 27, 2010 6:27 am

hi,

yes the original version example works fine. I've opened them now simultaneously (in 2 different hydrus windows) and looked for any differences between them. The only one I can find is perhaps in the number of total nodes (but of the same order) and perhaps an extra node used for the seepage face, but otherwise none.

Can I perhaps send it to you to check?

thanks for your help,

heleen

Jirka
Posts: 4790
Joined: Sat Mar 16, 2002 3:47 pm
Location: USA
Location: Riverside, CA

Post by Jirka » Wed Jan 27, 2010 2:41 pm

heleen,

If the original projects work, then problem is clearly in your input. Make sure that your initial and boundary conditions are physically reasonable. Make sure that you allow small time steps. For some reason, many users believe that if they do not allow small time steps the program will run faster, while that generally leads to nonconvergence.

Jirka

Post Reply