Freeze-thaw modeling in soi

Moderator: CODE_BRIGHT Team

JornSearch
Posts: 1
Joined: Sun Nov 03, 2019 6:40 am

Freeze-thaw modeling in soi

Postby JornSearch » Sun Nov 03, 2019 6:47 am

Freeze-thaw modeling in soil (THM case)
Postby Alireza » Mon Aug 12, 2019 5:02 pm

Dear CODE-BRIGHT users, CODE-BRIGHT team,
I am using CODE-BRIGHT to model freeze-thaw phenomenon in soil with application in tunneling.
I am trying to first check and simulate freeze-thaw in Oedometer test sample and then model its application on tunneling.
I have simply modeled freeze-thaw in Oedometer test sample using the Oedometer test example (considering the THM case). The sample surface temperature is changing from 15 degree to -7 degree and PRETENSION CURVE ITYCL = 18 and Liquid phase relative permeability ITYCL = 12 for freezing model, however the model can't run to the end and stops when surface temperature reaches to 0 degree.
I have attached all files.
The final error in output file says:
********************************************************************
Stress : 1 0.000E+00 124 2
Force balance : 2 0.501E-03 89
Water balance : 2 0.897E-02 2
Energy balance : 2 0.152E+04 2
Error estimation(rel): NaN NaN NaN NaN
Time step factor: 1.3999999999999999
Displacement : 2 0.110+217 111
Time step is reduced to: 0.503-133
because displacement correction: 0.110+217
is greater than dumx: 0.100E+00 at node: 111

Convergence is not possible, dtime
has been reduced 100 times because
inadmisible corrections or
large error estimation, Stop
User avatar
code_bright
Posts: 81
Joined: Thu Dec 04, 2014 2:19 pm
Location: Universitat Politècnica de Catalunya, Barcelona
Contact:

Re: Freeze-thaw modeling in soi

Postby code_bright » Thu Nov 28, 2019 7:49 am

Hi!

Convergence problems are usual in numerical modelling; using Code_bright or whichever software. Even advanced users have to deal with this. The most typical cause of this convergence problems is an incorrect set of boundary conditions. However, there might be another reasons.

In order to understand how to deal with numerical issues in Code_Bright, you can check the "time step control" document or the "description of the <root>_gen.out file" in the following link:

https://deca.upc.edu/en/projects/code_b ... /documents

Moreover, I suggest to simplify the model till it works. For example, try first to solve only the mechanical problem or only the hydro-thermal problem, so you may find where are the numerical issues.

We don't have a tutorial for the freezing model yet, but I would say it your problem is not specific of the freezing model.

Hope it helps,

Alfonso

Return to “CODE_BRIGHT”

Who is online

Users browsing this forum: No registered users and 1 guest