quantification of dispersion

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
lpeters
Posts: 6
Joined: Mon Nov 15, 2004 10:32 am
Location: Netherlands

quantification of dispersion

Post by lpeters » Wed Nov 24, 2004 7:59 am

Hello,

Is it possible to quantify the total dispersion generated by Hydrus2D, including the numerical dispersion resulting from the space and time discretisation? Especially for cases when the Peclet number is largrer than 1.

Any comments are welcome!
thanks for your time,
Lies

rvang
Posts: 46
Joined: Wed Apr 24, 2002 7:22 pm
Location: USA
Contact:

Post by rvang » Wed Nov 24, 2004 4:51 pm

Lies,

HYDRUS should produce very little artificial dispersion with any realistic discretization. This unless you select upstream weighting or GFE with artificial dispersion to avoid oscillations if the elements are relatively large. Oscillations are controlled by the Courant and grid Peclet numbers as explained in section 6.4.4 of the manual). To have some fun and see what happens, run HYDRUS-1D a few times for different nodal spacings (solute transport during steady flow) and analyze the results (inverse analysis) with an analytical solution (e.g., CFITM within STANMOD). Thanks.--Rien van G.

Post Reply