Liakopoulos Experiment_Bench Mark Problem

46 views
Skip to first unread message

Mohammad Islam

unread,
May 9, 2017, 5:55:06 PM5/9/17
to ogs-users
Hi All,
I was trying to run a benchmarks problem on Liakopoulos experiment as follows


I also read following two references on Liakopoulos experiment


Ref2) O. Kolditz et al. (eds.), Thermo-Hydro-Mechanical-Chemical Processes in Fractured Porous Media, Lecture Notes in Computational Science and Engineering 86, DOI 10.1007/978-3-642-27177-9 12, © Springer-Verlag Berlin Heidelberg 2012

247, Chapter 12, Page 247-253


Unfortunately, it is not running. After close look, I found that .msh file is not in correct order so I modified it. But, still same situation. From Ref1, I found that this example problem run well on both ogs5 and ogs6 (Ref 1, Page4, Figure 1). First of all can any one brief me any clue how to run this example problem? Secondly, what is the purpose for following files:

1) .lay

2) .gsp

3) .rfd

4) .fine (under Quad)

5) .coarse (under Quad)


In addition, in .gsp file there is another file extension named .rfi, but I did not find it in the folder.


Any co-operation will be appreciable.


Regards,

Mohammad


Dr. Mohammad Nurul Islam

 

ORISE PostdoctoralResearcher

The National EnergyTechnology Laboratory (NETL)

US Department of Energy (DoE)

NETL-PGH, 84/309

Email:Mohamma...@netl.doe.gov

 

Visiting Scholar

Civil and Environmental Engineering

University of Pittsburgh, USA

736 Benedum Hall,

3700 O'Hara St, Pittsburgh, PA 15213

Mobile: 1-412-3158889

Email: mn...@pitt.edu

Norihiro Watanabe

unread,
May 11, 2017, 6:19:14 PM5/11/17
to ogs-...@googlegroups.com
Hi Mohammad,

(Please correct me if I'm wroing)
Regarding the second question, .rfd is used to run the benchmark but
other files are not needed.

1) .lay is a TECPLOT layout file to create some figure.
2) .gsp collects all input file names used for a simulation. The file
may be still used in some pre-processing GUI (e.g. GINA, GS/RF4 GUI).
3) .rfd is often used to define X-Y curves, e.g. time-dependent
function, etc. In the Liakopoulos project, it seems the file defines
two curves (saturation-capillary pressure, saturation-rel.
permeability). They are referred in the .mmp file.
4, 5) These two are mesh files with different refinement size. I think
someone wanted to keep them for testing. If you want to try them,
simply rename the file name to h2_Liako.msh.


Best,
Nori
> --
> You received this message because you are subscribed to the Google Groups
> "ogs-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to ogs-users+...@googlegroups.com.
> For more options, visit https://groups.google.com/d/optout.



--
Norihiro Watanabe

Mohammad Islam

unread,
May 15, 2017, 9:07:47 AM5/15/17
to ogs-users
Dear Norihiro Watanabe,
Thank you for your kind reply.

Now the problem is running. I plotted them for Height-Capillary Pressure and Height-Air Pressure. Qualitative results are ok, but there are some oscillations compared to the results presented in Reference (Figure 12.2, and PAGE 252).

Ref2) O. Kolditz et al. (eds.), Thermo-Hydro-Mechanical-Chemical Processes in Fractured Porous Media, Lecture Notes in Computational Science and Engineering 86, DOI 10.1007/978-3-642-27177-9 12, © Springer-Verlag Berlin Heidelberg 2012

247, Chapter 12, Page 247-253


Any advice how to minimize the oscillations? For your kind considerations, I attached excel file and data files. I used Newton solver rather than Picard.

Thanks in advance. 

Regards,
Mohammad
L-tri.xlsx
L-Tri.pcs
L-Tri.rfd
L-Tri.bc
L-Tri.gli
L-Tri.ic
L-Tri.mfp
L-Tri.mmp
L-Tri.msh
L-Tri.num
L-Tri.out

Mohammad Islam

unread,
May 15, 2017, 12:13:26 PM5/15/17
to ogs-users
Just for my curiosity, I also used Picard. But, found similar noise in the graph.

Regards,
Mohammad

Mohammad Islam

unread,
May 23, 2017, 10:02:16 AM5/23/17
to ogs-users
Problem solved, and there is no more noise. Couple of elements in *.msh are missing. I attached screen shot.

Mohammad
L-Tri_msh.pdf

Norihiro Watanabe

unread,
May 23, 2017, 5:06:31 PM5/23/17
to ogs-...@googlegroups.com
Do you mean the mesh file in the GitHub repository has that problem?

nori

On 05/23/2017 11:02 PM, Mohammad Islam wrote:
> Problem solved, and there is no more noise. Couple of elements in *.msh
> are missing. I attached screen shot.
>
> Mohammad
>
> On Monday, 15 May 2017 12:13:26 UTC-4, Mohammad Islam wrote:
>
> Just for my curiosity, I also used Picard. But, found similar noise
> in the graph.
>
> Regards,
> Mohammad
>
> On Monday, 15 May 2017 09:07:47 UTC-4, Mohammad Islam wrote:
>
> Dear Norihiro Watanabe,
> Thank you for your kind reply.
>
> Now the problem is running. I plotted them for Height-Capillary
> Pressure and Height-Air Pressure. Qualitative results are ok,
> but there are some oscillations compared to the results
> presented in Reference (Figure 12.2, and PAGE 252).
>
> Ref2) O. Kolditz et al. (eds.),
> /Thermo-Hydro-Mechanical-Chemical Processes in Fractured
> //Porous Media/, Lecture Notes in Computational Science and
> <https://groups.google.com/d/optout>.
>
>
>
> --
> Norihiro Watanabe
>
> --
> You received this message because you are subscribed to the Google
> Groups "ogs-users" group.
> To unsubscribe from this group and stop receiving emails from it, send
> an email to ogs-users+...@googlegroups.com
> <mailto:ogs-users+...@googlegroups.com>.

Mohammad Islam

unread,
May 24, 2017, 4:01:31 PM5/24/17
to ogs-...@googlegroups.com
Hi Nori,
No, the problem in my side. When I converted, h2_Liako.msh to h2_Liako-c_Tri.msh  under  $ELEMENTS it was  68 instead of 76. Therefore, some elements were deleted.
Regards,
Mohammad

-----------------------------------------------------------------

Dr. Mohammad Nurul Islam

ORISE Postdoctoral Researcher

The National Energy Technology Laboratory (NETL)

US Department of Energy (DoE)

NETL-PGH, 84/309

 

Visiting Scholar

Civil and Environmental Engineering

University of Pittsburgh, USA

736 Benedum Hall,

3700 O'Hara St, Pittsburgh, PA 15213

Mobile: 1-412-3158889

Email: mn...@pitt.edu

To unsubscribe from this group and stop receiving emails from it, send an email to ogs-users+unsubscribe@googlegroups.com <mailto:ogs-users+unsubscribe@googlegroups.com>.

For more options, visit https://groups.google.com/d/optout.

--
You received this message because you are subscribed to a topic in the Google Groups "ogs-users" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/ogs-users/QfFwVRx886U/unsubscribe.
To unsubscribe from this group and all its topics, send an email to ogs-users+unsubscribe@googlegroups.com.
h2_Liako.msh
h2_Liako.pdf
h2_Liako-c_Tri.msh
Reply all
Reply to author
Forward
0 new messages