Please use this identifier to cite or link to this item:
|Title:||Numerical study of finite element method based solutions for propagation of wetting fronts in unsaturated soil|
|Authors:||Tan, T.-S. |
|Source:||Tan, T.-S., Phoon, K.-K., Chong, P.-C. (2004-03). Numerical study of finite element method based solutions for propagation of wetting fronts in unsaturated soil. Journal of Geotechnical and Geoenvironmental Engineering 130 (3) : 254-263. ScholarBank@NUS Repository. https://doi.org/10.1061/(ASCE)1090-0241(2004)130:3(254)|
|Abstract:||The accurate prediction of the propagation of a wetting front in an unsaturated soil subjected to surficial infiltration is of practical importance to many geotechnical and geoenvironmental problems. The finite element method is the most common solution technique as the hydraulic soil properties are highly nonlinear. Two important issues are often found to create difficulties in such analyses. First, numerical oscillations are usually observed in the calculated pore pressures at the wetting front. Second, when a reasonable mesh size and time step are used, the elevation of the wetting front may be seriously overpredicted. This paper is focused on the second issue. The under-relaxation (UR) technique used in the iterative process within each time step is found to have a serious impact on rate of convergence with refinement in mesh size and time step. Two different techniques are typically used; the first evaluates the hydraulic conductivity using an average of heads calculated from the preceding time node and the most recent iteration of the current time node (UR1), and the second evaluates the hydraulic conductivity using the average of heads calculated from the two most recent iterations of the current time nodes (UR2). The study shows that UR1, which is adopted in programs such as SEEP/W, ensures that the solution converges rapidly to a stable solution within a time step, but may converge to the wrong wetting front at a given elapsed time unless a sufficiently refined mesh is used. UR2 converges much more slowly within a time step, but the error in the wetting front is smaller than that generated by UR1. © ASCE/MARCH 2004.|
|Source Title:||Journal of Geotechnical and Geoenvironmental Engineering|
|Appears in Collections:||Staff Publications|
Show full item record
Files in This Item:
There are no files associated with this item.
checked on Feb 14, 2018
WEB OF SCIENCETM
checked on Feb 14, 2018
checked on Feb 20, 2018
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.