AW: Bugs fixed --> there are still some

Norihiro_Watanabe2 · January 17, 2014, 10:10am

Hi Hua,

I tested your xz example, but am not sure what you mean "the velocity changes direction, especially around tunnel in the late time". For me it looks ok. please check an animation here

https://dl.dropboxusercontent.com/u/331205/xz.avi

I used OGS ver 5.4.2 and PVD for output which correctly output ux,uz according to your coordinate system.

best,
nori

···

On 01/17/2014 09:47 AM, Shao, Hua wrote:

My BC definition using GRADIENT is the same as Georg, as said that it is not correct if I use it in xy section.

Both Version 5.4 and 5.5 doesn't work if I used the function described by Marc. I attached the .BC in this mail.

We have not understood the function described by Nori and Wenqing. What are a, b, c ....?

If I need only a distribution
  at y=400 300000 Pa
  at y=700 100000 Pa

How can I define them?

Best
Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com] Im Auftrag von Wenqing Wang
Gesendet: Donnerstag, 16. Januar 2014 16:57
An: ogs-users@googlegroups.com
Betreff: Re: AW: AW: [ogs-users] Bugs fixed --> there are still some

typo in 'For IC':
$DIS_TYPE
    FUNCTION
   $GEO_TYPE
    SUBDOMAIN
    2
     0 1.693567e5+4000*x+-11559.45*y
     2 1.693567e5+4000*x+-11559.45*y

The integer before the expression is the material index of a domain.

On 16.01.2014 16:52, Wenqing Wang wrote:

Yes, you can do like this but keep in mind that only + and * are
operators, and x, y, z denote coordinates. For BC, the syntax is that
in Nori's email. For IC: it takes the form:

$DIS_TYPE
   FUNCTION
  $GEO_TYPE
   SUBDOMAIN
   2
    0 1.693567e5+4000*x+-11559.45*y

The expression should be a+b*x+c*y+d*z, where a, b, c, and d are the
constant to be input.

Wenqing

On 16.01.2014 15:11, Norihiro Watanabe wrote:

you can also try a distribution type FUNCTION, which
f(x,y,z)=ax+by+cz. Example is

  $DIS_TYPE
   FUNCTION
   0+-11453*y

best,
nori

On 01/16/2014 03:07 PM, Georg Kosakowski wrote:

If you only need a gradient in Z and would like to account for
increase of pressure with depth ...you can try this:

#BOUNDARY_CONDITION
   $PCS_TYPE
    RICHARDS_FLOW
   $PRIMARY_VARIABLE
    PRESSURE1
   $GEO_TYPE
    POLYLINE LEFT_BOUNDARY
   $DIS_TYPE
    GRADIENT 0.0 0.0 9.81e+3 ; at z=0 m we have Pressure =0 -> fully
saturated ...saturation is decreasing with increasing z

GRADIENT [reference depth] [reference depth value] [Gradient]

Greetings, Georg

On 01/16/2014 02:50 PM, Marc Walther wrote:

Hi Hua,

yes, I remember Georg having some problems with high gradients of
permeability some time ago.

Concerning the BC, you might try to use this (.bc):

$DIS_TYPE
LINEAR [number of points defined below] [point_ID1] [PrimVar_value]
[name of point with ID1] [point_ID2] [PrimVar_value] [name of point
with ID2]

for this, you will have to give names to the points in the .gli like:

[point_ID1] x y z *$NAME *[name of point with ID1]

Maybe this helps you with applying a gradient.

Best,
Marc

On Do 16 Jan 2014 14:24:30 CET, Georg Kosakowski wrote:

Hoi Hua,
you are correct regarding the boundary conditions. Gradients are
only defined for z-direction. I remember I looked this up in the
code, because I wondered how to define the direction of gradients!
I also found, that it is necessary to change a bit the numerical
settings, as otherwise the solution for the pressure field is not
accurate enough in low-permeability media and the flow field might
get "wrong".
Try setting the solver tolerance to 1e-20 or even smaller ...
Greetings, Georg

On 01/16/2014 02:18 PM, Shao, Hua wrote:

Thank you all for lots tips. I don’t know where is my branch, I
attached my two simplified cases in the zip-form.

The xy coordinate system doesn’t allow me to used gradient as
boundary conditions. The calculation of velocity, at least
direction looks properly.

The xz coordinate system runs well, however, the velocity changes
direction, especially around tunnel in the late time.

Hua

*Von:*ogs-users@googlegroups.com
[mailto:ogs-users@googlegroups.com]
*Im Auftrag von *Marc Walther
*Gesendet:* Donnerstag, 16. Januar 2014 12:26
*An:* ogs-users@googlegroups.com
*Betreff:* Re: AW: [ogs-users] Bugs fixed --> there are still
some

Dear Hua,

Could you upload the setup to your branch and give a link, that
we can take a look?

I don't understand what you mean by 1. "it doesn't work" - can
you give some details?

2. For 2D, I think OGS will always write results in the first two
components of the vectors. You can easily transform the vector in
ParaView or TecPlot.

3. The setup would help to understand why this happens

Best,
Marc

Am 16.01.2014 11:56 schrieb "Shao, Hua" <Hua.Shao@bgr.de
<mailto:Hua.Shao@bgr.de>>:

Dear all

We are currently using Richards-model for one project. Based on
some Benchmark-example, e.g. Foryth, we got a very quick start.
However, we have also problems as:
1. 2D problem doesn't work if the problem in y-z coordinate
system 2. Definition of coordinate system is quite confuse, i.e.
2D problem in x-z section, output of velocity is ux, uy 3. The
calculation of velocity is with great uncertainty: in some test
cases, we got even vx=vy (???) in the output file.

Does anyone have an idea or experience on the 'RICHARD's flow'?

Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>
[mailto:ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>] Im Auftrag von Wenqing Wang
Gesendet: Dienstag, 14. Januar 2014 12:55
An: ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>
Betreff: [ogs-users] Bugs fixed

Dear all,

If you solve THM coupling problem with deactivated domains,
please patch your source code with my today's commits , which
contain the bug fixing for the boundary condition assignment.

Wenqing

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users%2Bunsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out
<https://groups.google.com/groups/opt_out>\.

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users%2Bunsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out
<https://groups.google.com/groups/opt_out>\.

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users+unsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul
Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email:georg.kosakowski@psi.ch h323:+41563104743
homepage:Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul
Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email:georg.kosakowski@psi.ch h323:+41563104743
homepage:Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
Dr.-Ing. Norihiro Watanabe
Department of Environmental Informatics (ENVINF)
Wissenschaftler

Helmholtz-Zentrum für Umweltforschung GmbH - UFZ
Helmholtz Centre for Environmental Research GmbH - UFZ
Permoserstraße 15 / 04318 Leipzig / Germany

norihiro.watanabe@ufz.de / http://www.ufz.de
Telefon +49 341 235 1806

Sitz der Gesellschaft: Leipzig
Registergericht: Amtsgericht Leipzig, Handelsregister Nr. B 4703
Vorsitzender des Aufsichtsrats: MinDirig Wilfried Kraus
Wissenschaftlicher Geschäftsführer: Prof. Dr. Georg Teutsch
Administrativer Geschäftsführer: Dr. Heike Graßmann

Marc_Walther · January 17, 2014, 10:21am

Hi Hua,

Nori was faster, but I also think, the results look reasonable. I extended the simulation time with the xz-example and came to this quasi-steady-state.

Marc

PS: in the mmp, under PERMEABILITY_SATURATION, there is a line twice (0 3) - don't know, if that belongs there

···

On 17.01.2014 11:10, Norihiro Watanabe wrote:

Hi Hua,

I tested your xz example, but am not sure what you mean "the velocity changes direction, especially around tunnel in the late time". For me it looks ok. please check an animation here

https://dl.dropboxusercontent.com/u/331205/xz.avi

I used OGS ver 5.4.2 and PVD for output which correctly output ux,uz according to your coordinate system.

best,
nori

On 01/17/2014 09:47 AM, Shao, Hua wrote:

My BC definition using GRADIENT is the same as Georg, as said that it is not correct if I use it in xy section.

Both Version 5.4 and 5.5 doesn't work if I used the function described by Marc. I attached the .BC in this mail.

We have not understood the function described by Nori and Wenqing. What are a, b, c ....?

If I need only a distribution
    at y=400 300000 Pa
    at y=700 100000 Pa

How can I define them?

Best
Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com] Im Auftrag von Wenqing Wang
Gesendet: Donnerstag, 16. Januar 2014 16:57
An: ogs-users@googlegroups.com
Betreff: Re: AW: AW: [ogs-users] Bugs fixed --> there are still some

typo in 'For IC':
$DIS_TYPE
    FUNCTION
   $GEO_TYPE
    SUBDOMAIN
    2
     0 1.693567e5+4000*x+-11559.45*y
     2 1.693567e5+4000*x+-11559.45*y

The integer before the expression is the material index of a domain.

On 16.01.2014 16:52, Wenqing Wang wrote:

Yes, you can do like this but keep in mind that only + and * are
operators, and x, y, z denote coordinates. For BC, the syntax is that
in Nori's email. For IC: it takes the form:

$DIS_TYPE
   FUNCTION
  $GEO_TYPE
   SUBDOMAIN
   2
    0 1.693567e5+4000*x+-11559.45*y

The expression should be a+b*x+c*y+d*z, where a, b, c, and d are the
constant to be input.

Wenqing

On 16.01.2014 15:11, Norihiro Watanabe wrote:

you can also try a distribution type FUNCTION, which
f(x,y,z)=ax+by+cz. Example is

  $DIS_TYPE
   FUNCTION
   0+-11453*y

best,
nori

On 01/16/2014 03:07 PM, Georg Kosakowski wrote:

If you only need a gradient in Z and would like to account for
increase of pressure with depth ...you can try this:

#BOUNDARY_CONDITION
   $PCS_TYPE
    RICHARDS_FLOW
   $PRIMARY_VARIABLE
    PRESSURE1
   $GEO_TYPE
    POLYLINE LEFT_BOUNDARY
   $DIS_TYPE
    GRADIENT 0.0 0.0 9.81e+3 ; at z=0 m we have Pressure =0 -> fully
saturated ...saturation is decreasing with increasing z

GRADIENT [reference depth] [reference depth value] [Gradient]

Greetings, Georg

On 01/16/2014 02:50 PM, Marc Walther wrote:

Hi Hua,

yes, I remember Georg having some problems with high gradients of
permeability some time ago.

Concerning the BC, you might try to use this (.bc):

$DIS_TYPE
LINEAR [number of points defined below] [point_ID1] [PrimVar_value]
[name of point with ID1] [point_ID2] [PrimVar_value] [name of point
with ID2]

for this, you will have to give names to the points in the .gli like:

[point_ID1] x y z *$NAME *[name of point with ID1]

Maybe this helps you with applying a gradient.

Best,
Marc

On Do 16 Jan 2014 14:24:30 CET, Georg Kosakowski wrote:

Hoi Hua,
you are correct regarding the boundary conditions. Gradients are
only defined for z-direction. I remember I looked this up in the
code, because I wondered how to define the direction of gradients!
I also found, that it is necessary to change a bit the numerical
settings, as otherwise the solution for the pressure field is not
accurate enough in low-permeability media and the flow field might
get "wrong".
Try setting the solver tolerance to 1e-20 or even smaller ...
Greetings, Georg

On 01/16/2014 02:18 PM, Shao, Hua wrote:

Thank you all for lots tips. I don’t know where is my branch, I
attached my two simplified cases in the zip-form.

The xy coordinate system doesn’t allow me to used gradient as
boundary conditions. The calculation of velocity, at least
direction looks properly.

The xz coordinate system runs well, however, the velocity changes
direction, especially around tunnel in the late time.

Hua

*Von:*ogs-users@googlegroups.com
[mailto:ogs-users@googlegroups.com]
*Im Auftrag von *Marc Walther
*Gesendet:* Donnerstag, 16. Januar 2014 12:26
*An:* ogs-users@googlegroups.com
*Betreff:* Re: AW: [ogs-users] Bugs fixed --> there are still
some

Dear Hua,

Could you upload the setup to your branch and give a link, that
we can take a look?

I don't understand what you mean by 1. "it doesn't work" - can
you give some details?

2. For 2D, I think OGS will always write results in the first two
components of the vectors. You can easily transform the vector in
ParaView or TecPlot.

3. The setup would help to understand why this happens

Best,
Marc

Am 16.01.2014 11:56 schrieb "Shao, Hua" <Hua.Shao@bgr.de
<mailto:Hua.Shao@bgr.de>>:

Dear all

We are currently using Richards-model for one project. Based on
some Benchmark-example, e.g. Foryth, we got a very quick start.
However, we have also problems as:
1. 2D problem doesn't work if the problem in y-z coordinate
system 2. Definition of coordinate system is quite confuse, i.e.
2D problem in x-z section, output of velocity is ux, uy 3. The
calculation of velocity is with great uncertainty: in some test
cases, we got even vx=vy (???) in the output file.

Does anyone have an idea or experience on the 'RICHARD's flow'?

Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>
[mailto:ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>] Im Auftrag von Wenqing Wang
Gesendet: Dienstag, 14. Januar 2014 12:55
An: ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>
Betreff: [ogs-users] Bugs fixed

Dear all,

If you solve THM coupling problem with deactivated domains,
please patch your source code with my today's commits , which
contain the bug fixing for the boundary condition assignment.

Wenqing

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users%2Bunsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out
<https://groups.google.com/groups/opt_out>\.

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users%2Bunsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out
<https://groups.google.com/groups/opt_out>\.

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users+unsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul
Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email:georg.kosakowski@psi.ch h323:+41563104743
homepage:Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul
Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email:georg.kosakowski@psi.ch h323:+41563104743
homepage:Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

Shao_Hua · January 17, 2014, 11:29am

Dear all

Thank you very much for your prompt answers.
I would like to summary all discussion topics:
  1. the IC and BC definition as Wenqing mailed to us work very well at least in the xy-plane.
  BC-Definition is the same to Marc mail: note point_ID1 is from 1 not 0!
  But the results are ... see XY_plane_velocity##
  2. If I look the results from Nori and Marc, they look well (Marc: twice definition 0 3 is from a benchmark called H_us in the example). However, in my simulation the velocity are generally upwards (see XZ_plane_velocity.png). How do you get such pretty picture (only extend your simulation time to 13000 years?)
  In summary: 2D can also be in x-z plane!

Hua

···

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com] Im Auftrag von Marc Walther
Gesendet: Freitag, 17. Januar 2014 11:21
An: ogs-users@googlegroups.com
Betreff: Re: AW: AW: AW: [ogs-users] Bugs fixed --> there are still some

Hi Hua,

Nori was faster, but I also think, the results look reasonable. I extended the simulation time with the xz-example and came to this quasi-steady-state.

Marc

PS: in the mmp, under PERMEABILITY_SATURATION, there is a line twice (0
3) - don't know, if that belongs there

On 17.01.2014 11:10, Norihiro Watanabe wrote:

Hi Hua,

I tested your xz example, but am not sure what you mean "the velocity
changes direction, especially around tunnel in the late time". For me
it looks ok. please check an animation here

https://dl.dropboxusercontent.com/u/331205/xz.avi

I used OGS ver 5.4.2 and PVD for output which correctly output ux,uz
according to your coordinate system.

best,
nori

On 01/17/2014 09:47 AM, Shao, Hua wrote:

My BC definition using GRADIENT is the same as Georg, as said that it
is not correct if I use it in xy section.

Both Version 5.4 and 5.5 doesn't work if I used the function
described by Marc. I attached the .BC in this mail.

We have not understood the function described by Nori and Wenqing.
What are a, b, c ....?

If I need only a distribution
    at y=400 300000 Pa
    at y=700 100000 Pa

How can I define them?

Best
Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com]
Im Auftrag von Wenqing Wang
Gesendet: Donnerstag, 16. Januar 2014 16:57
An: ogs-users@googlegroups.com
Betreff: Re: AW: AW: [ogs-users] Bugs fixed --> there are still some

typo in 'For IC':
$DIS_TYPE
    FUNCTION
   $GEO_TYPE
    SUBDOMAIN
    2
     0 1.693567e5+4000*x+-11559.45*y
     2 1.693567e5+4000*x+-11559.45*y

The integer before the expression is the material index of a domain.

On 16.01.2014 16:52, Wenqing Wang wrote:

Yes, you can do like this but keep in mind that only + and * are
operators, and x, y, z denote coordinates. For BC, the syntax is that
in Nori's email. For IC: it takes the form:

$DIS_TYPE
   FUNCTION
  $GEO_TYPE
   SUBDOMAIN
   2
    0 1.693567e5+4000*x+-11559.45*y
    0 1.693567e5+4000*x+-11559.45*y

The expression should be a+b*x+c*y+d*z, where a, b, c, and d are the
constant to be input.

Wenqing

On 16.01.2014 15:11, Norihiro Watanabe wrote:

you can also try a distribution type FUNCTION, which
f(x,y,z)=ax+by+cz. Example is

  $DIS_TYPE
   FUNCTION
   0+-11453*y

best,
nori

On 01/16/2014 03:07 PM, Georg Kosakowski wrote:

If you only need a gradient in Z and would like to account for
increase of pressure with depth ...you can try this:

#BOUNDARY_CONDITION
   $PCS_TYPE
    RICHARDS_FLOW
   $PRIMARY_VARIABLE
    PRESSURE1
   $GEO_TYPE
    POLYLINE LEFT_BOUNDARY
   $DIS_TYPE
    GRADIENT 0.0 0.0 9.81e+3 ; at z=0 m we have Pressure =0 ->
fully
saturated ...saturation is decreasing with increasing z

GRADIENT [reference depth] [reference depth value] [Gradient]

Greetings, Georg

On 01/16/2014 02:50 PM, Marc Walther wrote:

Hi Hua,

yes, I remember Georg having some problems with high gradients of
permeability some time ago.

Concerning the BC, you might try to use this (.bc):

$DIS_TYPE
LINEAR [number of points defined below] [point_ID1] [PrimVar_value]
[name of point with ID1] [point_ID2] [PrimVar_value] [name of point
with ID2]

for this, you will have to give names to the points in the .gli
like:

[point_ID1] x y z *$NAME *[name of point with ID1]

Maybe this helps you with applying a gradient.

Best,
Marc

On Do 16 Jan 2014 14:24:30 CET, Georg Kosakowski wrote:

Hoi Hua,
you are correct regarding the boundary conditions. Gradients are
only defined for z-direction. I remember I looked this up in the
code, because I wondered how to define the direction of gradients!
I also found, that it is necessary to change a bit the numerical
settings, as otherwise the solution for the pressure field is not
accurate enough in low-permeability media and the flow field might
get "wrong".
Try setting the solver tolerance to 1e-20 or even smaller ...
Greetings, Georg

On 01/16/2014 02:18 PM, Shao, Hua wrote:

Thank you all for lots tips. I don’t know where is my branch, I
attached my two simplified cases in the zip-form.

The xy coordinate system doesn’t allow me to used gradient as
boundary conditions. The calculation of velocity, at least
direction looks properly.

The xz coordinate system runs well, however, the velocity changes
direction, especially around tunnel in the late time.

Hua

*Von:*ogs-users@googlegroups.com
[mailto:ogs-users@googlegroups.com]
*Im Auftrag von *Marc Walther
*Gesendet:* Donnerstag, 16. Januar 2014 12:26
*An:* ogs-users@googlegroups.com
*Betreff:* Re: AW: [ogs-users] Bugs fixed --> there are still
some

Dear Hua,

Could you upload the setup to your branch and give a link, that
we can take a look?

I don't understand what you mean by 1. "it doesn't work" - can
you give some details?

2. For 2D, I think OGS will always write results in the first two
components of the vectors. You can easily transform the vector in
ParaView or TecPlot.

3. The setup would help to understand why this happens

Best,
Marc

Am 16.01.2014 11:56 schrieb "Shao, Hua" <Hua.Shao@bgr.de
<mailto:Hua.Shao@bgr.de>>:

Dear all

We are currently using Richards-model for one project. Based on
some Benchmark-example, e.g. Foryth, we got a very quick start.
However, we have also problems as:
1. 2D problem doesn't work if the problem in y-z coordinate
system 2. Definition of coordinate system is quite confuse, i.e.
2D problem in x-z section, output of velocity is ux, uy 3. The
calculation of velocity is with great uncertainty: in some test
cases, we got even vx=vy (???) in the output file.

Does anyone have an idea or experience on the 'RICHARD's flow'?

Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>
[mailto:ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>] Im Auftrag von Wenqing Wang
Gesendet: Dienstag, 14. Januar 2014 12:55
An: ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>
Betreff: [ogs-users] Bugs fixed

Dear all,

If you solve THM coupling problem with deactivated domains,
please patch your source code with my today's commits , which
contain the bug fixing for the boundary condition assignment.

Wenqing

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users%2Bunsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out
<https://groups.google.com/groups/opt_out>\.

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users%2Bunsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out
<https://groups.google.com/groups/opt_out>\.

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users+unsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul
Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email:georg.kosakowski@psi.ch h323:+41563104743
homepage:Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul
Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email:georg.kosakowski@psi.ch h323:+41563104743
homepage:Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

Marc_Walther2 · January 17, 2014, 11:59am

Hi Hua,

I think, you can remove the second "0 3".

Velocities are upwards in the beginning for my sims, too. I just noticed, that this is not the final state and extended the sim time in .tim and .rfd files.

As your setup seems to resemble a vertical 2D slice, I would go with the working x-z setup. This would also make it easier to implement something like density-dependent flow or other gravity-related things.

If you REALLY need the horizontal case, can you sent the updated x-y setup, please?

Marc

···

On 17.01.2014 12:29, Shao, Hua wrote:

Dear all

Thank you very much for your prompt answers.
I would like to summary all discussion topics:
  1. the IC and BC definition as Wenqing mailed to us work very well at least in the xy-plane.
  BC-Definition is the same to Marc mail: note point_ID1 is from 1 not 0!
  But the results are ... see XY_plane_velocity##
  2. If I look the results from Nori and Marc, they look well (Marc: twice definition 0 3 is from a benchmark called H_us in the example). However, in my simulation the velocity are generally upwards (see XZ_plane_velocity.png). How do you get such pretty picture (only extend your simulation time to 13000 years?)
  In summary: 2D can also be in x-z plane!

Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com] Im Auftrag von Marc Walther
Gesendet: Freitag, 17. Januar 2014 11:21
An: ogs-users@googlegroups.com
Betreff: Re: AW: AW: AW: [ogs-users] Bugs fixed --> there are still some

Hi Hua,

Nori was faster, but I also think, the results look reasonable. I extended the simulation time with the xz-example and came to this quasi-steady-state.

Marc

PS: in the mmp, under PERMEABILITY_SATURATION, there is a line twice (0
3) - don't know, if that belongs there

On 17.01.2014 11:10, Norihiro Watanabe wrote:

Hi Hua,

I tested your xz example, but am not sure what you mean "the velocity
changes direction, especially around tunnel in the late time". For me
it looks ok. please check an animation here

https://dl.dropboxusercontent.com/u/331205/xz.avi

I used OGS ver 5.4.2 and PVD for output which correctly output ux,uz
according to your coordinate system.

best,
nori

On 01/17/2014 09:47 AM, Shao, Hua wrote:

My BC definition using GRADIENT is the same as Georg, as said that it
is not correct if I use it in xy section.

Both Version 5.4 and 5.5 doesn't work if I used the function
described by Marc. I attached the .BC in this mail.

We have not understood the function described by Nori and Wenqing.
What are a, b, c ....?

If I need only a distribution
     at y=400 300000 Pa
     at y=700 100000 Pa

How can I define them?

Best
Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com]
Im Auftrag von Wenqing Wang
Gesendet: Donnerstag, 16. Januar 2014 16:57
An: ogs-users@googlegroups.com
Betreff: Re: AW: AW: [ogs-users] Bugs fixed --> there are still some

typo in 'For IC':
$DIS_TYPE
     FUNCTION
    $GEO_TYPE
     SUBDOMAIN
     2
      0 1.693567e5+4000*x+-11559.45*y
      2 1.693567e5+4000*x+-11559.45*y

The integer before the expression is the material index of a domain.

On 16.01.2014 16:52, Wenqing Wang wrote:

Yes, you can do like this but keep in mind that only + and * are
operators, and x, y, z denote coordinates. For BC, the syntax is that
in Nori's email. For IC: it takes the form:

$DIS_TYPE
    FUNCTION
   $GEO_TYPE
    SUBDOMAIN
    2
     0 1.693567e5+4000*x+-11559.45*y

The expression should be a+b*x+c*y+d*z, where a, b, c, and d are the
constant to be input.

Wenqing

On 16.01.2014 15:11, Norihiro Watanabe wrote:

you can also try a distribution type FUNCTION, which
f(x,y,z)=ax+by+cz. Example is

   $DIS_TYPE
    FUNCTION
    0+-11453*y

best,
nori

On 01/16/2014 03:07 PM, Georg Kosakowski wrote:

If you only need a gradient in Z and would like to account for
increase of pressure with depth ...you can try this:

#BOUNDARY_CONDITION
    $PCS_TYPE
     RICHARDS_FLOW
    $PRIMARY_VARIABLE
     PRESSURE1
    $GEO_TYPE
     POLYLINE LEFT_BOUNDARY
    $DIS_TYPE
     GRADIENT 0.0 0.0 9.81e+3 ; at z=0 m we have Pressure =0 ->
fully
saturated ...saturation is decreasing with increasing z

GRADIENT [reference depth] [reference depth value] [Gradient]

Greetings, Georg

On 01/16/2014 02:50 PM, Marc Walther wrote:

Hi Hua,

yes, I remember Georg having some problems with high gradients of
permeability some time ago.

Concerning the BC, you might try to use this (.bc):

$DIS_TYPE
LINEAR [number of points defined below] [point_ID1] [PrimVar_value]
[name of point with ID1] [point_ID2] [PrimVar_value] [name of point
with ID2]

for this, you will have to give names to the points in the .gli
like:

[point_ID1] x y z *$NAME *[name of point with ID1]

Maybe this helps you with applying a gradient.

Best,
Marc

On Do 16 Jan 2014 14:24:30 CET, Georg Kosakowski wrote:

Hoi Hua,
you are correct regarding the boundary conditions. Gradients are
only defined for z-direction. I remember I looked this up in the
code, because I wondered how to define the direction of gradients!
I also found, that it is necessary to change a bit the numerical
settings, as otherwise the solution for the pressure field is not
accurate enough in low-permeability media and the flow field might
get "wrong".
Try setting the solver tolerance to 1e-20 or even smaller ...
Greetings, Georg

On 01/16/2014 02:18 PM, Shao, Hua wrote:

Thank you all for lots tips. I don’t know where is my branch, I
attached my two simplified cases in the zip-form.

The xy coordinate system doesn’t allow me to used gradient as
boundary conditions. The calculation of velocity, at least
direction looks properly.

The xz coordinate system runs well, however, the velocity changes
direction, especially around tunnel in the late time.

Hua

*Von:*ogs-users@googlegroups.com
[mailto:ogs-users@googlegroups.com]
*Im Auftrag von *Marc Walther
*Gesendet:* Donnerstag, 16. Januar 2014 12:26
*An:* ogs-users@googlegroups.com
*Betreff:* Re: AW: [ogs-users] Bugs fixed --> there are still
some

Dear Hua,

Could you upload the setup to your branch and give a link, that
we can take a look?

I don't understand what you mean by 1. "it doesn't work" - can
you give some details?

2. For 2D, I think OGS will always write results in the first two
components of the vectors. You can easily transform the vector in
ParaView or TecPlot.

3. The setup would help to understand why this happens

Best,
Marc

Am 16.01.2014 11:56 schrieb "Shao, Hua" <Hua.Shao@bgr.de
<mailto:Hua.Shao@bgr.de>>:

Dear all

We are currently using Richards-model for one project. Based on
some Benchmark-example, e.g. Foryth, we got a very quick start.
However, we have also problems as:
1. 2D problem doesn't work if the problem in y-z coordinate
system 2. Definition of coordinate system is quite confuse, i.e.
2D problem in x-z section, output of velocity is ux, uy 3. The
calculation of velocity is with great uncertainty: in some test
cases, we got even vx=vy (???) in the output file.

Does anyone have an idea or experience on the 'RICHARD's flow'?

Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>
[mailto:ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>] Im Auftrag von Wenqing Wang
Gesendet: Dienstag, 14. Januar 2014 12:55
An: ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>
Betreff: [ogs-users] Bugs fixed

Dear all,

If you solve THM coupling problem with deactivated domains,
please patch your source code with my today's commits , which
contain the bug fixing for the boundary condition assignment.

Wenqing

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users%2Bunsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out
<https://groups.google.com/groups/opt_out>\.

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users%2Bunsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out
<https://groups.google.com/groups/opt_out>\.

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users+unsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul
Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email:georg.kosakowski@psi.ch h323:+41563104743
homepage:Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul
Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email:georg.kosakowski@psi.ch h323:+41563104743
homepage:Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

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

Dipl.-Hydrol. Marc Walther
Wissenschaftlicher Mitarbeiter

Technische Universität Dresden
Fakultät Umweltwissenschaften
Fachrichtung Hydrowissenschaften
Institut für Grundwasserwirtschaft
01062 Dresden

Besucheradresse
Bergstr. 66, 01069 Dresden
Neubau Chemie 2. Bauabschnitt, CHE-BA2, Raum E59

Tel +49 351 463 425 56
Fax +49 351 463 425 52
Mobil +49 178 334 18 19

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

Shao_Hua · January 17, 2014, 12:22pm

Thanks Marc,

my problem is a three dimensional one. I just checked a 2D case in vertical plane for simplification. I noticed a y-z plane case didn't work, I changed to x-z plane. Following the suggestion from Chris, I use now a x-y plane.

I will now move to 3D case.

Hua

···

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com] Im Auftrag von Marc Walther
Gesendet: Freitag, 17. Januar 2014 12:59
An: ogs-users@googlegroups.com
Betreff: Re: AW: AW: AW: AW: [ogs-users] Bugs fixed --> there are still some

Hi Hua,

I think, you can remove the second "0 3".

Velocities are upwards in the beginning for my sims, too. I just noticed, that this is not the final state and extended the sim time in .tim and .rfd files.

As your setup seems to resemble a vertical 2D slice, I would go with the working x-z setup. This would also make it easier to implement something like density-dependent flow or other gravity-related things.

If you REALLY need the horizontal case, can you sent the updated x-y setup, please?

Marc

On 17.01.2014 12:29, Shao, Hua wrote:

Dear all

Thank you very much for your prompt answers.
I would like to summary all discussion topics:
  1. the IC and BC definition as Wenqing mailed to us work very well at least in the xy-plane.
  BC-Definition is the same to Marc mail: note point_ID1 is from 1 not 0!
  But the results are ... see XY_plane_velocity##
  2. If I look the results from Nori and Marc, they look well (Marc: twice definition 0 3 is from a benchmark called H_us in the example). However, in my simulation the velocity are generally upwards (see XZ_plane_velocity.png). How do you get such pretty picture (only extend your simulation time to 13000 years?)
  In summary: 2D can also be in x-z plane!

Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com] Im Auftrag von Marc Walther
Gesendet: Freitag, 17. Januar 2014 11:21
An: ogs-users@googlegroups.com
Betreff: Re: AW: AW: AW: [ogs-users] Bugs fixed --> there are still some

Hi Hua,

Nori was faster, but I also think, the results look reasonable. I extended the simulation time with the xz-example and came to this quasi-steady-state.

Marc

PS: in the mmp, under PERMEABILITY_SATURATION, there is a line twice (0
3) - don't know, if that belongs there

On 17.01.2014 11:10, Norihiro Watanabe wrote:

Hi Hua,

I tested your xz example, but am not sure what you mean "the velocity
changes direction, especially around tunnel in the late time". For me
it looks ok. please check an animation here

https://dl.dropboxusercontent.com/u/331205/xz.avi

I used OGS ver 5.4.2 and PVD for output which correctly output ux,uz
according to your coordinate system.

best,
nori

On 01/17/2014 09:47 AM, Shao, Hua wrote:

My BC definition using GRADIENT is the same as Georg, as said that it
is not correct if I use it in xy section.

Both Version 5.4 and 5.5 doesn't work if I used the function
described by Marc. I attached the .BC in this mail.

We have not understood the function described by Nori and Wenqing.
What are a, b, c ....?

If I need only a distribution
     at y=400 300000 Pa
     at y=700 100000 Pa

How can I define them?

Best
Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com]
Im Auftrag von Wenqing Wang
Gesendet: Donnerstag, 16. Januar 2014 16:57
An: ogs-users@googlegroups.com
Betreff: Re: AW: AW: [ogs-users] Bugs fixed --> there are still some

typo in 'For IC':
$DIS_TYPE
     FUNCTION
    $GEO_TYPE
     SUBDOMAIN
     2
      0 1.693567e5+4000*x+-11559.45*y
      2 1.693567e5+4000*x+-11559.45*y

The integer before the expression is the material index of a domain.

On 16.01.2014 16:52, Wenqing Wang wrote:

Yes, you can do like this but keep in mind that only + and * are
operators, and x, y, z denote coordinates. For BC, the syntax is that
in Nori's email. For IC: it takes the form:

$DIS_TYPE
    FUNCTION
   $GEO_TYPE
    SUBDOMAIN
    2
     0 1.693567e5+4000*x+-11559.45*y
     0 1.693567e5+4000*x+-11559.45*y

The expression should be a+b*x+c*y+d*z, where a, b, c, and d are the
constant to be input.

Wenqing

On 16.01.2014 15:11, Norihiro Watanabe wrote:

you can also try a distribution type FUNCTION, which
f(x,y,z)=ax+by+cz. Example is

   $DIS_TYPE
    FUNCTION
    0+-11453*y

best,
nori

On 01/16/2014 03:07 PM, Georg Kosakowski wrote:

If you only need a gradient in Z and would like to account for
increase of pressure with depth ...you can try this:

#BOUNDARY_CONDITION
    $PCS_TYPE
     RICHARDS_FLOW
    $PRIMARY_VARIABLE
     PRESSURE1
    $GEO_TYPE
     POLYLINE LEFT_BOUNDARY
    $DIS_TYPE
     GRADIENT 0.0 0.0 9.81e+3 ; at z=0 m we have Pressure =0 ->
fully
saturated ...saturation is decreasing with increasing z

GRADIENT [reference depth] [reference depth value] [Gradient]

Greetings, Georg

On 01/16/2014 02:50 PM, Marc Walther wrote:

Hi Hua,

yes, I remember Georg having some problems with high gradients of
permeability some time ago.

Concerning the BC, you might try to use this (.bc):

$DIS_TYPE
LINEAR [number of points defined below] [point_ID1] [PrimVar_value]
[name of point with ID1] [point_ID2] [PrimVar_value] [name of point
with ID2]

for this, you will have to give names to the points in the .gli
like:

[point_ID1] x y z *$NAME *[name of point with ID1]

Maybe this helps you with applying a gradient.

Best,
Marc

On Do 16 Jan 2014 14:24:30 CET, Georg Kosakowski wrote:

Hoi Hua,
you are correct regarding the boundary conditions. Gradients are
only defined for z-direction. I remember I looked this up in the
code, because I wondered how to define the direction of gradients!
I also found, that it is necessary to change a bit the numerical
settings, as otherwise the solution for the pressure field is not
accurate enough in low-permeability media and the flow field might
get "wrong".
Try setting the solver tolerance to 1e-20 or even smaller ...
Greetings, Georg

On 01/16/2014 02:18 PM, Shao, Hua wrote:

Thank you all for lots tips. I don’t know where is my branch, I
attached my two simplified cases in the zip-form.

The xy coordinate system doesn’t allow me to used gradient as
boundary conditions. The calculation of velocity, at least
direction looks properly.

The xz coordinate system runs well, however, the velocity changes
direction, especially around tunnel in the late time.

Hua

*Von:*ogs-users@googlegroups.com
[mailto:ogs-users@googlegroups.com]
*Im Auftrag von *Marc Walther
*Gesendet:* Donnerstag, 16. Januar 2014 12:26
*An:* ogs-users@googlegroups.com
*Betreff:* Re: AW: [ogs-users] Bugs fixed --> there are still
some

Dear Hua,

Could you upload the setup to your branch and give a link, that
we can take a look?

I don't understand what you mean by 1. "it doesn't work" - can
you give some details?

2. For 2D, I think OGS will always write results in the first two
components of the vectors. You can easily transform the vector in
ParaView or TecPlot.

3. The setup would help to understand why this happens

Best,
Marc

Am 16.01.2014 11:56 schrieb "Shao, Hua" <Hua.Shao@bgr.de
<mailto:Hua.Shao@bgr.de>>:

Dear all

We are currently using Richards-model for one project. Based on
some Benchmark-example, e.g. Foryth, we got a very quick start.
However, we have also problems as:
1. 2D problem doesn't work if the problem in y-z coordinate
system 2. Definition of coordinate system is quite confuse, i.e.
2D problem in x-z section, output of velocity is ux, uy 3. The
calculation of velocity is with great uncertainty: in some test
cases, we got even vx=vy (???) in the output file.

Does anyone have an idea or experience on the 'RICHARD's flow'?

Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>
[mailto:ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>] Im Auftrag von Wenqing Wang
Gesendet: Dienstag, 14. Januar 2014 12:55
An: ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>
Betreff: [ogs-users] Bugs fixed

Dear all,

If you solve THM coupling problem with deactivated domains,
please patch your source code with my today's commits , which
contain the bug fixing for the boundary condition assignment.

Wenqing

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users%2Bunsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out
<https://groups.google.com/groups/opt_out>\.

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users%2Bunsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out
<https://groups.google.com/groups/opt_out>\.

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users+unsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul
Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email:georg.kosakowski@psi.ch h323:+41563104743
homepage:Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul
Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email:georg.kosakowski@psi.ch h323:+41563104743
homepage:Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

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

Dipl.-Hydrol. Marc Walther
Wissenschaftlicher Mitarbeiter

Technische Universität Dresden
Fakultät Umweltwissenschaften
Fachrichtung Hydrowissenschaften
Institut für Grundwasserwirtschaft
01062 Dresden

Besucheradresse
Bergstr. 66, 01069 Dresden
Neubau Chemie 2. Bauabschnitt, CHE-BA2, Raum E59

Tel +49 351 463 425 56
Fax +49 351 463 425 52
Mobil +49 178 334 18 19

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

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

Shao_Hua · January 21, 2014, 1:16pm

Dear all

Now we have got quite reasonable results about the velocity (direction and value) in a 3D example. So we can claim the Richards model works properly. For the application in 2D, one should use correct coordinate system.

Thank you very much. It was very excited to get so much support and tips from you.
Hua

···

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com] Im Auftrag von Shao, Hua
Gesendet: Freitag, 17. Januar 2014 12:29
An: ogs-users@googlegroups.com
Betreff: AW: AW: AW: AW: [ogs-users] Bugs fixed --> there are still some

Dear all

Thank you very much for your prompt answers.
I would like to summary all discussion topics:
  1. the IC and BC definition as Wenqing mailed to us work very well at least in the xy-plane.
  BC-Definition is the same to Marc mail: note point_ID1 is from 1 not 0!
  But the results are ... see XY_plane_velocity##
  2. If I look the results from Nori and Marc, they look well (Marc: twice definition 0 3 is from a benchmark called H_us in the example). However, in my simulation the velocity are generally upwards (see XZ_plane_velocity.png). How do you get such pretty picture (only extend your simulation time to 13000 years?)
  In summary: 2D can also be in x-z plane!

Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com] Im Auftrag von Marc Walther
Gesendet: Freitag, 17. Januar 2014 11:21
An: ogs-users@googlegroups.com
Betreff: Re: AW: AW: AW: [ogs-users] Bugs fixed --> there are still some

Hi Hua,

Nori was faster, but I also think, the results look reasonable. I extended the simulation time with the xz-example and came to this quasi-steady-state.

Marc

PS: in the mmp, under PERMEABILITY_SATURATION, there is a line twice (0
3) - don't know, if that belongs there

On 17.01.2014 11:10, Norihiro Watanabe wrote:

Hi Hua,

I tested your xz example, but am not sure what you mean "the velocity
changes direction, especially around tunnel in the late time". For me
it looks ok. please check an animation here

https://dl.dropboxusercontent.com/u/331205/xz.avi

I used OGS ver 5.4.2 and PVD for output which correctly output ux,uz
according to your coordinate system.

best,
nori

On 01/17/2014 09:47 AM, Shao, Hua wrote:

My BC definition using GRADIENT is the same as Georg, as said that it
is not correct if I use it in xy section.

Both Version 5.4 and 5.5 doesn't work if I used the function
described by Marc. I attached the .BC in this mail.

We have not understood the function described by Nori and Wenqing.
What are a, b, c ....?

If I need only a distribution
    at y=400 300000 Pa
    at y=700 100000 Pa

How can I define them?

Best
Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com]
Im Auftrag von Wenqing Wang
Gesendet: Donnerstag, 16. Januar 2014 16:57
An: ogs-users@googlegroups.com
Betreff: Re: AW: AW: [ogs-users] Bugs fixed --> there are still some

typo in 'For IC':
$DIS_TYPE
    FUNCTION
   $GEO_TYPE
    SUBDOMAIN
    2
     0 1.693567e5+4000*x+-11559.45*y
     2 1.693567e5+4000*x+-11559.45*y

The integer before the expression is the material index of a domain.

On 16.01.2014 16:52, Wenqing Wang wrote:

Yes, you can do like this but keep in mind that only + and * are
operators, and x, y, z denote coordinates. For BC, the syntax is
that in Nori's email. For IC: it takes the form:

$DIS_TYPE
   FUNCTION
  $GEO_TYPE
   SUBDOMAIN
   2
    0 1.693567e5+4000*x+-11559.45*y
    0 1.693567e5+4000*x+-11559.45*y

The expression should be a+b*x+c*y+d*z, where a, b, c, and d are the
constant to be input.

Wenqing

On 16.01.2014 15:11, Norihiro Watanabe wrote:

you can also try a distribution type FUNCTION, which
f(x,y,z)=ax+by+cz. Example is

  $DIS_TYPE
   FUNCTION
   0+-11453*y

best,
nori

On 01/16/2014 03:07 PM, Georg Kosakowski wrote:

If you only need a gradient in Z and would like to account for
increase of pressure with depth ...you can try this:

#BOUNDARY_CONDITION
   $PCS_TYPE
    RICHARDS_FLOW
   $PRIMARY_VARIABLE
    PRESSURE1
   $GEO_TYPE
    POLYLINE LEFT_BOUNDARY
   $DIS_TYPE
    GRADIENT 0.0 0.0 9.81e+3 ; at z=0 m we have Pressure =0 ->
fully
saturated ...saturation is decreasing with increasing z

GRADIENT [reference depth] [reference depth value] [Gradient]

Greetings, Georg

On 01/16/2014 02:50 PM, Marc Walther wrote:

Hi Hua,

yes, I remember Georg having some problems with high gradients of
permeability some time ago.

Concerning the BC, you might try to use this (.bc):

$DIS_TYPE
LINEAR [number of points defined below] [point_ID1]
[PrimVar_value] [name of point with ID1] [point_ID2]
[PrimVar_value] [name of point with ID2]

for this, you will have to give names to the points in the .gli
like:

[point_ID1] x y z *$NAME *[name of point with ID1]

Maybe this helps you with applying a gradient.

Best,
Marc

On Do 16 Jan 2014 14:24:30 CET, Georg Kosakowski wrote:

Hoi Hua,
you are correct regarding the boundary conditions. Gradients are
only defined for z-direction. I remember I looked this up in the
code, because I wondered how to define the direction of gradients!
I also found, that it is necessary to change a bit the numerical
settings, as otherwise the solution for the pressure field is
not accurate enough in low-permeability media and the flow field
might get "wrong".
Try setting the solver tolerance to 1e-20 or even smaller ...
Greetings, Georg

On 01/16/2014 02:18 PM, Shao, Hua wrote:

Thank you all for lots tips. I don’t know where is my branch, I
attached my two simplified cases in the zip-form.

The xy coordinate system doesn’t allow me to used gradient as
boundary conditions. The calculation of velocity, at least
direction looks properly.

The xz coordinate system runs well, however, the velocity
changes direction, especially around tunnel in the late time.

Hua

*Von:*ogs-users@googlegroups.com
[mailto:ogs-users@googlegroups.com]
*Im Auftrag von *Marc Walther
*Gesendet:* Donnerstag, 16. Januar 2014 12:26
*An:* ogs-users@googlegroups.com
*Betreff:* Re: AW: [ogs-users] Bugs fixed --> there are still
some

Dear Hua,

Could you upload the setup to your branch and give a link, that
we can take a look?

I don't understand what you mean by 1. "it doesn't work" - can
you give some details?

2. For 2D, I think OGS will always write results in the first
two components of the vectors. You can easily transform the
vector in ParaView or TecPlot.

3. The setup would help to understand why this happens

Best,
Marc

Am 16.01.2014 11:56 schrieb "Shao, Hua" <Hua.Shao@bgr.de
<mailto:Hua.Shao@bgr.de>>:

Dear all

We are currently using Richards-model for one project. Based on
some Benchmark-example, e.g. Foryth, we got a very quick start.
However, we have also problems as:
1. 2D problem doesn't work if the problem in y-z coordinate
system 2. Definition of coordinate system is quite confuse, i.e.
2D problem in x-z section, output of velocity is ux, uy 3. The
calculation of velocity is with great uncertainty: in some test
cases, we got even vx=vy (???) in the output file.

Does anyone have an idea or experience on the 'RICHARD's flow'?

Hua

-----Ursprüngliche Nachricht-----
Von: ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>
[mailto:ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>] Im Auftrag von Wenqing
Wang
Gesendet: Dienstag, 14. Januar 2014 12:55
An: ogs-users@googlegroups.com
<mailto:ogs-users@googlegroups.com>
Betreff: [ogs-users] Bugs fixed

Dear all,

If you solve THM coupling problem with deactivated domains,
please patch your source code with my today's commits , which
contain the bug fixing for the boundary condition assignment.

Wenqing

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users%2Bunsubscribe@googlegroups.com>.
For more options, visit
https://groups.google.com/groups/opt_out
<https://groups.google.com/groups/opt_out>\.

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users%2Bunsubscribe@googlegroups.com>.
For more options, visit
https://groups.google.com/groups/opt_out
<https://groups.google.com/groups/opt_out>\.

--
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+unsubscribe@googlegroups.com
<mailto:ogs-users+unsubscribe@googlegroups.com>.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul
Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email:georg.kosakowski@psi.ch h323:+41563104743
homepage:Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul
Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email:georg.kosakowski@psi.ch h323:+41563104743
homepage:Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

--
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+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/groups/opt_out\.

nbeyond · January 21, 2014, 1:22pm

Good to hear it work in 3D, Hua. Richards model is still limited to single phase, though based on the assumption that the air pressure is atmospheric. Anyhow, it’s very interesting in reading the forum developing.

···

Cheers,

Chan-Hee

On Tue, Jan 21, 2014 at 10:16 PM, Shao, Hua Hua.Shao@bgr.de wrote:

Dear all

Now we have got quite reasonable results about the velocity (direction and value) in a 3D example. So we can claim the Richards model works properly. For the application in 2D, one should use correct coordinate system.

Thank you very much. It was very excited to get so much support and tips from you.

Hua

-----Ursprüngliche Nachricht-----

Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com] Im Auftrag von Shao, Hua

Gesendet: Freitag, 17. Januar 2014 12:29

An: ogs-users@googlegroups.com

Betreff: AW: AW: AW: AW: [ogs-users] Bugs fixed → there are still some

Dear all

Thank you very much for your prompt answers.

I would like to summary all discussion topics:
    1. the IC and BC definition as Wenqing mailed to us work very well at least in the xy-plane.

    BC-Definition is the same to Marc mail: note point_ID1 is from 1 not 0!

    But the results are ... see XY_plane_velocity##

    2. If I look the results from Nori and Marc, they look well (Marc: twice definition 0 3 is from a benchmark called H_us in the example). However, in my simulation the velocity are generally upwards (see XZ_plane_velocity.png). How do you get such pretty picture (only extend your simulation time to 13000 years?)



    In summary: 2D can also be in x-z plane!
Hua

-----Ursprüngliche Nachricht-----Park, Ph.D.

Phone: +82 42 868 3502
Fax: +82 42 868 3358

Geothermal Resources Team
Korea Institute of Geoscience and Mineral Resources (KIGAM)
124 Gwahang-no, Yuseong-gu, Daejeon 305-350, KOREA

Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com] Im Auftrag von Marc Walther

Gesendet: Freitag, 17. Januar 2014 11:21

An: ogs-users@googlegroups.com

Betreff: Re: AW: AW: AW: [ogs-users] Bugs fixed → there are still some

Hi Hua,

Nori was faster, but I also think, the results look reasonable. I extended the simulation time with the xz-example and came to this quasi-steady-state.

Marc

PS: in the mmp, under PERMEABILITY_SATURATION, there is a line twice (0

don’t know, if that belongs there

On 17.01.2014 11:10, Norihiro Watanabe wrote:

Hi Hua,

I tested your xz example, but am not sure what you mean "the velocity

changes direction, especially around tunnel in the late time". For me

it looks ok. please check an animation here

https://dl.dropboxusercontent.com/u/331205/xz.avi

I used OGS ver 5.4.2 and PVD for output which correctly output ux,uz

according to your coordinate system.

best,

nori

On 01/17/2014 09:47 AM, Shao, Hua wrote:

My BC definition using GRADIENT is the same as Georg, as said that it

is not correct if I use it in xy section.

Both Version 5.4 and 5.5 doesn’t work if I used the function

described by Marc. I attached the .BC in this mail.

We have not understood the function described by Nori and Wenqing.

What are a, b, c …?

If I need only a distribution
at y=400    300000 Pa
at y=700     100000 Pa
How can I define them?

Best

Hua

-----Ursprüngliche Nachricht-----

Von: ogs-users@googlegroups.com [mailto:ogs-users@googlegroups.com]

Im Auftrag von Wenqing Wang

Gesendet: Donnerstag, 16. Januar 2014 16:57

An: ogs-users@googlegroups.com

Betreff: Re: AW: AW: [ogs-users] Bugs fixed → there are still some

typo in ‘For IC’:

$DIS_TYPE
FUNCTION
$GEO_TYPE
SUBDOMAIN
2
 0 1.693567e5+4000*x+-11559.45*y
 2 1.693567e5+4000*x+-11559.45*y
The integer before the expression is the material index of a domain.

On 16.01.2014 16:52, Wenqing Wang wrote:

Yes, you can do like this but keep in mind that only + and * are

operators, and x, y, z denote coordinates. For BC, the syntax is

that in Nori’s email. For IC: it takes the form:

$DIS_TYPE

FUNCTION

$GEO_TYPE

SUBDOMAIN

2
0 1.693567e5+4000*x+-11559.45*y
0 1.693567e5+4000*x+-11559.45*y
The expression should be a+bx+cy+d*z, where a, b, c, and d are the

constant to be input.

Wenqing

On 16.01.2014 15:11, Norihiro Watanabe wrote:

you can also try a distribution type FUNCTION, which

f(x,y,z)=ax+by+cz. Example is

$DIS_TYPE

FUNCTION

0±11453*y

best,

nori

On 01/16/2014 03:07 PM, Georg Kosakowski wrote:

If you only need a gradient in Z and would like to account for

increase of pressure with depth …you can try this:

#BOUNDARY_CONDITION

$PCS_TYPE
RICHARDS_FLOW
$PRIMARY_VARIABLE
PRESSURE1
$GEO_TYPE
POLYLINE LEFT_BOUNDARY
$DIS_TYPE
GRADIENT 0.0 0.0 9.81e+3   ;  at z=0 m we have Pressure =0 ->
fully

saturated …saturation is decreasing with increasing z

GRADIENT [reference depth] [reference depth value] [Gradient]

Greetings, Georg

On 01/16/2014 02:50 PM, Marc Walther wrote:

Hi Hua,

yes, I remember Georg having some problems with high gradients of

permeability some time ago.

Concerning the BC, you might try to use this (.bc):

$DIS_TYPE

LINEAR [number of points defined below] [point_ID1]

[PrimVar_value] [name of point with ID1] [point_ID2]

[PrimVar_value] [name of point with ID2]

for this, you will have to give names to the points in the .gli

like:

[point_ID1] x y z *$NAME *[name of point with ID1]

Maybe this helps you with applying a gradient.

Best,

Marc

On Do 16 Jan 2014 14:24:30 CET, Georg Kosakowski wrote:

Hoi Hua,

you are correct regarding the boundary conditions. Gradients are

only defined for z-direction. I remember I looked this up in the

code, because I wondered how to define the direction of gradients!

I also found, that it is necessary to change a bit the numerical

settings, as otherwise the solution for the pressure field is

not accurate enough in low-permeability media and the flow field

might get “wrong”.

Try setting the solver tolerance to 1e-20 or even smaller …

Greetings, Georg

On 01/16/2014 02:18 PM, Shao, Hua wrote:

Thank you all for lots tips. I don’t know where is my branch, I

attached my two simplified cases in the zip-form.

The xy coordinate system doesn’t allow me to used gradient as

boundary conditions. The calculation of velocity, at least

direction looks properly.

The xz coordinate system runs well, however, the velocity

changes direction, especially around tunnel in the late time.

Hua

*Von:*ogs-users@googlegroups.com

[mailto:ogs-users@googlegroups.com]

*Im Auftrag von *Marc Walther

Gesendet: Donnerstag, 16. Januar 2014 12:26

An: ogs-users@googlegroups.com

Betreff: Re: AW: [ogs-users] Bugs fixed → there are still

some

Dear Hua,

Could you upload the setup to your branch and give a link, that

we can take a look?

I don’t understand what you mean by 1. “it doesn’t work” - can

you give some details?

For 2D, I think OGS will always write results in the first

two components of the vectors. You can easily transform the

vector in ParaView or TecPlot.

The setup would help to understand why this happens

Best,

Marc

Am 16.01.2014 11:56 schrieb “Shao, Hua” <Hua.Shao@bgr.de

mailto:Hua.Shao@bgr.de>:

Dear all

We are currently using Richards-model for one project. Based on

some Benchmark-example, e.g. Foryth, we got a very quick start.

However, we have also problems as:

2D problem doesn’t work if the problem in y-z coordinate

system 2. Definition of coordinate system is quite confuse, i.e.

2D problem in x-z section, output of velocity is ux, uy 3. The

calculation of velocity is with great uncertainty: in some test

cases, we got even vx=vy (???) in the output file.

Does anyone have an idea or experience on the ‘RICHARD’s flow’?

Hua

-----Ursprüngliche Nachricht-----

Von: ogs-users@googlegroups.com

mailto:ogs-users@googlegroups.com

[mailto:ogs-users@googlegroups.com

mailto:ogs-users@googlegroups.com] Im Auftrag von Wenqing

Wang

Gesendet: Dienstag, 14. Januar 2014 12:55

An: ogs-users@googlegroups.com

mailto:ogs-users@googlegroups.com

Betreff: [ogs-users] Bugs fixed

Dear all,

If you solve THM coupling problem with deactivated domains,

please patch your source code with my today’s commits , which

contain the bug fixing for the boundary condition assignment.

Wenqing

–

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+unsubscribe@googlegroups.com

mailto:ogs-users%2Bunsubscribe@googlegroups.com.

For more options, visit

https://groups.google.com/groups/opt_out

<https://groups.google.com/groups/opt_out>.

–

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+unsubscribe@googlegroups.com

mailto:ogs-users%2Bunsubscribe@googlegroups.com.

For more options, visit

https://groups.google.com/groups/opt_out

<https://groups.google.com/groups/opt_out>.

–

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+unsubscribe@googlegroups.com

mailto:ogs-users+unsubscribe@googlegroups.com.

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

–

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+unsubscribe@googlegroups.com.

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

–

PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul

Scherrer Institut * 5232 Villigen/PSI * Switzerland

Phone: +41-56-3104743 * Fax: +41-56-3102821

Email:georg.kosakowski@psi.ch h323:+41563104743

homepage:http://people.web.psi.ch/kosakowski/

–

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+unsubscribe@googlegroups.com.

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

–

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+unsubscribe@googlegroups.com.

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

–

PD Dr. Georg Kosakowski * Laboratory for Waste Management Paul

Scherrer Institut * 5232 Villigen/PSI * Switzerland

Phone: +41-56-3104743 * Fax: +41-56-3102821

Email:georg.kosakowski@psi.ch h323:+41563104743

homepage:http://people.web.psi.ch/kosakowski/

–

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+unsubscribe@googlegroups.com.

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

–

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+unsubscribe@googlegroups.com.

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

–

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+unsubscribe@googlegroups.com.

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

–

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+unsubscribe@googlegroups.com.

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

–

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+unsubscribe@googlegroups.com.

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

wenqing.wang · February 3, 2014, 4:51pm

Since the source code of the partition has been changed in last year (https://github.com/ufz/mesh_partition), which is set to generate partitioned mesh in binary files by default, I'd better to put a link to the corresponding source code here: https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc
It is my branch based on revision 12678 of the trunk, which is very close to the head of the trunk and benchmarked. In this branch, the major changes are for PETSc routines:
1. Parallel mesh data input, and parallel domain data output. For a large mesh, this improves I/O significantly.
2. Deformation

I am not sure when I can merge it back. If you are developing with PETSc or want to explore, you can take a look and merge it. Other information can be found on the updated page: https://svn.ufz.de/ogs/wiki/PETScPage

wenqing.wang · February 3, 2014, 4:56pm

With that changes, a whole domain output of variables can be generated for OGS#PETSc

···

On 03.02.2014 17:51, Wenqing Wang wrote:

Since the source code of the partition has been changed in last year (https://github.com/ufz/mesh_partition\), which is set to generate partitioned mesh in binary files by default, I'd better to put a link to the corresponding source code here: https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc
It is my branch based on revision 12678 of the trunk, which is very close to the head of the trunk and benchmarked. In this branch, the major changes are for PETSc routines:
1. Parallel mesh data input, and parallel domain data output. For a large mesh, this improves I/O significantly.
2. Deformation

I am not sure when I can merge it back. If you are developing with PETSc or want to explore, you can take a look and merge it. Other information can be found on the updated page: https://svn.ufz.de/ogs/wiki/PETScPage

wenqing.wang · March 7, 2014, 9:54am

Dear Nori,

It is in my previous email attached here. I did not change my branch since my last commit.

Shall we merge your branch before, or you do merging after?  I want to do this because that Tom and me have to do something on it.

···

Cheers,

  Wenqing

  -------- Original Message --------

Subject:
PETSC version in my branch
Date:
Mon, 03 Feb 2014 17:51:30 +0100
From:
Wenqing Wang
To:
wenqing.wang@ufz.de ogs-users@googlegroups.com


Since the source code of the partition has been changed in last year (), which is set to generate partitioned mesh in binary files by default, I'd better to put a link to the corresponding source code here: It is my branch based on revision 12678 of the trunk, which is very close to the head of the trunk and benchmarked. In this branch, the major changes are for PETSc routines: 1. Parallel mesh data input, and parallel domain data output. For a large mesh, this improves I/O significantly. 2. Deformation I am not sure when I can merge it back. If you are developing with PETSc or want to explore, you can take a look and merge it. Other information can be found on the updated page:

https://github.com/ufz/mesh_partition https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc https://svn.ufz.de/ogs/wiki/PETScPage
On 07.03.2014 10:44, Norihiro Watanabe wrote:

    Dear Wenqing,



    Could you please tell us a summary of what you have changed there?





    Best,

    Nori



    On 03/07/2014 10:32 AM, Wenqing Wang wrote:

      Dose anybody mind me to merge my branch

       to the trunk? Wenqing

https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc

Norihiro_Watanabe2 · March 7, 2014, 9:56am

Thanks. You don't need to merge my branch. I'll do it after you release a version.

Best,
Nori

···

On 03/07/2014 10:54 AM, Wenqing Wang wrote:

Dear Nori,

It is in my previous email attached here. I did not change my branch
since my last commit.

Shall we merge your branch before, or you do merging after? I want to
do this because that Tom and me have to do something on it.

Cheers,

Wenqing

-------- Original Message --------
Subject: PETSC version in my branch
Date: Mon, 03 Feb 2014 17:51:30 +0100
From: Wenqing Wang <wenqing.wang@ufz.de>
To: ogs-users@googlegroups.com

Since the source code of the partition has been changed in last year
(https://github.com/ufz/mesh_partition\), which is set to generate
partitioned mesh in binary files by default, I'd better to put a link to
the corresponding source code here:
https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc
It is my branch based on revision 12678 of the trunk, which is very
close to the head of the trunk and benchmarked. In this branch, the
major changes are for PETSc routines:
1. Parallel mesh data input, and parallel domain data output. For a
large mesh, this improves I/O significantly.
2. Deformation

I am not sure when I can merge it back. If you are developing with PETSc
or want to explore, you can take a look and merge it. Other information
can be found on the updated page:https://svn.ufz.de/ogs/wiki/PETScPage

On 07.03.2014 10:44, Norihiro Watanabe wrote:

Dear Wenqing,

Could you please tell us a summary of what you have changed there?

Best,
Nori

On 03/07/2014 10:32 AM, Wenqing Wang wrote:

Dose anybody mind me to merge my branch
https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc to the trunk?

Wenqing

--
Dr.-Ing. Norihiro Watanabe
Department of Environmental Informatics (ENVINF)
Wissenschaftler

Helmholtz-Zentrum für Umweltforschung GmbH - UFZ
Helmholtz Centre for Environmental Research GmbH - UFZ
Permoserstraße 15 / 04318 Leipzig / Germany

norihiro.watanabe@ufz.de / http://www.ufz.de
Telefon +49 341 235 1806

Sitz der Gesellschaft: Leipzig
Registergericht: Amtsgericht Leipzig, Handelsregister Nr. B 4703
Vorsitzender des Aufsichtsrats: MinDirig Wilfried Kraus
Wissenschaftlicher Geschäftsführer: Prof. Dr. Georg Teutsch
Administrativer Geschäftsführer: Dr. Heike Graßmann

wenqing.wang · March 11, 2014, 2:45pm

Dear All,

I have merged my branch to the trunk as ver. 5.5.3, and posted a release notes on wiki. The following benchmark results have to be confirmed due to the changes in calculation of material parameters in the previous version:

Last Successful Artifacts <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/>

AKS_LONG_H_us_Wet_h_us_line_celia.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/AKS_LONG_H_us_Wet_h_us_line_celia.html> 901,55 KB [fingerprint] <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/AKS_LONG_H_us_Wet_h_us_line_celia.html/*fingerprint*/>

BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html> 127,17 KB [fingerprint] <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html/*fingerprint*/>

BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html> 108,80 KB [fingerprint] <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html/*fingerprint*/>

  BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html> 86,19 KB [fingerprint] <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html/*fingerprint*/>
  BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html> 86,19 KB [fingerprint] <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html/*fingerprint*/>
  HS_LONG_C_comedy2d_cement2d.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/HS_LONG_C_comedy2d_cement2d.html>

Cheers,

Wenqing

···

On 07.03.2014 10:54, Wenqing Wang wrote:

Dear Nori,

It is in my previous email attached here. I did not change my branch since my last commit.

Shall we merge your branch before, or you do merging after? I want to do this because that Tom and me have to do something on it.

Cheers,

Wenqing

-------- Original Message --------
Subject: PETSC version in my branch
Date: Mon, 03 Feb 2014 17:51:30 +0100
From: Wenqing Wang <wenqing.wang@ufz.de>
To: ogs-users@googlegroups.com

Since the source code of the partition has been changed in last year
(https://github.com/ufz/mesh_partition\), which is set to generate
partitioned mesh in binary files by default, I'd better to put a link to
the corresponding source code here:
https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc
It is my branch based on revision 12678 of the trunk, which is very
close to the head of the trunk and benchmarked. In this branch, the
major changes are for PETSc routines:
1. Parallel mesh data input, and parallel domain data output. For a
large mesh, this improves I/O significantly.
2. Deformation

I am not sure when I can merge it back. If you are developing with PETSc
or want to explore, you can take a look and merge it. Other information
can be found on the updated page:https://svn.ufz.de/ogs/wiki/PETScPage

On 07.03.2014 10:44, Norihiro Watanabe wrote:

Dear Wenqing,

Could you please tell us a summary of what you have changed there?

Best,
Nori

On 03/07/2014 10:32 AM, Wenqing Wang wrote:

Dose anybody mind me to merge my branch
https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc to the trunk?

Wenqing

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

bastian.graupner · March 11, 2014, 3:01pm

Hi Wenqing,

I had a quick look to the results and it looks like that the water and gas saturation is changed from 0.6/0.4 to 0.75/0.25.

This could be related to a different interpretation of the initial condition. Did you change anything that could be relevant to pc-S-Relations in OGS?

Best regards,

Bastian

···

Von: ogs-devs@googlegroups.com [mailto:ogs-devs@googlegroups.com] Im Auftrag von Wenqing Wang

Gesendet: Dienstag, 11. März 2014 15:45

An: ogs-devs@googlegroups.com

Betreff: v 5.5.3. Re: Fwd: PETSC version in my branch

Dear All,

I have merged my branch to the trunk as ver. 5.5.3, and posted a release notes on wiki. The following benchmark results have to be confirmed due to the changes in calculation of material parameters in the previous version:

Last Successful Artifacts

AKS_LONG_H_us_Wet_h_us_line_celia.html

901,55 KB

BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html

127,17 KB

BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html

108,80 KB

BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html

86,19 KB

BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html

86,19 KB

HS_LONG_C_comedy2d_cement2d.html

Cheers,

Wenqing

On 07.03.2014 10:54, Wenqing Wang wrote:

Dear Nori,

It is in my previous email attached here. I did not change my branch since my last commit.

Shall we merge your branch before, or you do merging after? I want to do this because that Tom and me have to do something on it.

Cheers,

Wenqing

-------- Original Message --------

Subject:

PETSC version in my branch

Date:

Mon, 03 Feb 2014 17:51:30 +0100

From:

Wenqing Wang wenqing.wang@ufz.de

To:

ogs-users@googlegroups.com

Since the source code of the partition has been changed in last year

([https://github.com/ufz/mesh_partition](https://github.com/ufz/mesh_partition)), which is set to generate

partitioned mesh in binary files by default, I'd better to put a link to

the corresponding  source code here:

[https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc](https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc)

It is my branch based on revision 12678 of the trunk, which is very

close to the head of the trunk and benchmarked. In this branch, the

major changes are for PETSc routines:

1. Parallel mesh data input, and parallel domain data output. For a

large mesh, this improves I/O significantly.

2. Deformation

I am not sure when I can merge it back. If you are developing with PETSc

or want to explore, you can take a look and merge it.  Other information

can be  found on the updated page: [https://svn.ufz.de/ogs/wiki/PETScPage](https://svn.ufz.de/ogs/wiki/PETScPage)

On 07.03.2014 10:44, Norihiro Watanabe wrote:

Dear Wenqing,

Could you please tell us a summary of what you have changed there?

Best,

Nori

On 03/07/2014 10:32 AM, Wenqing Wang wrote:

Dose anybody mind me to merge my branch

https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc to the trunk?

Wenqing

–

You received this message because you are subscribed to the Google Groups “ogs-devs” group.

To unsubscribe from this group and stop receiving emails from it, send an email to ogs-devs+unsubscribe@googlegroups.com.

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

–

You received this message because you are subscribed to the Google Groups “ogs-devs” group.

To unsubscribe from this group and stop receiving emails from it, send an email to ogs-devs+unsubscribe@googlegroups.com.

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

wenqing.wang · March 11, 2014, 4:32pm

Sorry, the result changes are not from the previous version but from the bug fixing in ver. 5.5.2 from Marc. You can confirm the results according to the attached diff file.

diff.pdf (19.3 KB)

···

On 11.03.2014 15:45, Wenqing Wang wrote:

Dear All,

I have merged my branch to the trunk as ver. 5.5.3, and posted a release notes on wiki. The following benchmark results have to be confirmed due to the changes in calculation of material parameters in the previous version:

Last Successful Artifacts <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/>

  AKS_LONG_H_us_Wet_h_us_line_celia.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/AKS_LONG_H_us_Wet_h_us_line_celia.html> 901,55 KB [fingerprint] <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/AKS_LONG_H_us_Wet_h_us_line_celia.html/*fingerprint*/>

  BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html> 127,17 KB [fingerprint] <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html/*fingerprint*/>

  BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html> 108,80 KB [fingerprint] <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html/*fingerprint*/>

  BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html> 86,19 KB [fingerprint] <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html/*fingerprint*/>
  BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html> 86,19 KB [fingerprint] <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html/*fingerprint*/>
  HS_LONG_C_comedy2d_cement2d.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/HS_LONG_C_comedy2d_cement2d.html>

Cheers,

Wenqing

On 07.03.2014 10:54, Wenqing Wang wrote:

Dear Nori,

It is in my previous email attached here. I did not change my branch since my last commit.

Shall we merge your branch before, or you do merging after? I want to do this because that Tom and me have to do something on it.

Cheers,

Wenqing

-------- Original Message --------
Subject: PETSC version in my branch
Date: Mon, 03 Feb 2014 17:51:30 +0100
From: Wenqing Wang <wenqing.wang@ufz.de>
To: ogs-users@googlegroups.com

Since the source code of the partition has been changed in last year
(https://github.com/ufz/mesh_partition\), which is set to generate
partitioned mesh in binary files by default, I'd better to put a link to
the corresponding source code here:
https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc
It is my branch based on revision 12678 of the trunk, which is very
close to the head of the trunk and benchmarked. In this branch, the
major changes are for PETSc routines:
1. Parallel mesh data input, and parallel domain data output. For a
large mesh, this improves I/O significantly.
2. Deformation

I am not sure when I can merge it back. If you are developing with PETSc
or want to explore, you can take a look and merge it. Other information
can be found on the updated page:https://svn.ufz.de/ogs/wiki/PETScPage

On 07.03.2014 10:44, Norihiro Watanabe wrote:

Dear Wenqing,

Could you please tell us a summary of what you have changed there?

Best,
Nori

On 03/07/2014 10:32 AM, Wenqing Wang wrote:

Dose anybody mind me to merge my branch
https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc to the trunk?

Wenqing

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

marc.walther · March 11, 2014, 4:53pm

Hi Bastian,

I did only change pc-S relations for a specific mmp model, as can be
seen from the diff Wenqing sent. The values you mentioned, I never
touched.

The only place where I might have messed up your benchmarks seems to be
the first lines, where we have

if(eS > 1.0)
eS = 1.0;
+ if(eS < MediaProp->capillary pressure values[1])
+ eS = MediaProp->capillary pressure values[1];
...

If this effects you, tell me. I thought, that an effective saturation <
0 is not to be expected at all. We could limit this to the mmp model as
the other changes.

Cheers,
Marc

···

On Di, 2014-03-11 at 15:01 +0000, Graupner Bastian wrote:

Hi Wenqing,

I had a quick look to the results and it looks like that the water and
gas saturation is changed from 0.6/0.4 to 0.75/0.25.

This could be related to a different interpretation of the initial
condition. Did you change anything that could be relevant to
pc-S-Relations in OGS?

Best regards,

Bastian

Von: ogs-devs@googlegroups.com [mailto:ogs-devs@googlegroups.com] Im
Auftrag von Wenqing Wang
Gesendet: Dienstag, 11. März 2014 15:45
An: ogs-devs@googlegroups.com
Betreff: v 5.5.3. Re: Fwd: PETSC version in my branch

Dear All,

I have merged my branch to the trunk as ver. 5.5.3, and posted a
release notes on wiki. The following benchmark results have to be
confirmed due to the changes in calculation of material parameters in
the previous version:

Last Successful Artifacts

AKS_LONG_H_us_Wet_h_us_line_celia.html

901,55 KB

[fingerprint]

BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html

127,17 KB

[fingerprint]

BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html

108,80 KB

[fingerprint]

BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html

86,19 KB

[fingerprint]

BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html

86,19 KB

[fingerprint]

HS_LONG_C_comedy2d_cement2d.html

Cheers,

Wenqing

On 07.03.2014 10:54, Wenqing Wang wrote:

        Dear Nori,

        It is in my previous email attached here. I did not change my
        branch since my last commit.

        Shall we merge your branch before, or you do merging after? I
        want to do this because that Tom and me have to do something
        on it.

        Cheers,

        Wenqing

        -------- Original Message --------

                              Subject:

        PETSC version in my branch

                                 Date:

        Mon, 03 Feb 2014 17:51:30 +0100

                                 From:

        Wenqing Wang
        <wenqing.wang@ufz.de>

                                   To:

        ogs-users@googlegroups.com

        Since the source code of the partition has been changed in last year
        (https://github.com/ufz/mesh_partition\), which is set to generate
        partitioned mesh in binary files by default, I'd better to put a link to
        the corresponding source code here:
        https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc
        It is my branch based on revision 12678 of the trunk, which is very
        close to the head of the trunk and benchmarked. In this branch, the
        major changes are for PETSc routines:
        1. Parallel mesh data input, and parallel domain data output. For a
        large mesh, this improves I/O significantly.
        2. Deformation

        I am not sure when I can merge it back. If you are developing with PETSc
        or want to explore, you can take a look and merge it. Other information
        can be found on the updated page: https://svn.ufz.de/ogs/wiki/PETScPage

        On 07.03.2014 10:44, Norihiro Watanabe wrote:

                Dear Wenqing,

                Could you please tell us a summary of what you have
                changed there?

                Best,
                Nori

                On 03/07/2014 10:32 AM, Wenqing Wang wrote:

                Dose anybody mind me to merge my branch
                https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc to the trunk?

                Wenqing

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

Dipl.-Hydrol. Marc Walther

Department of Environmental Informatics
Helmholtz Centre for Environmental Research - UFZ
Permoserstraße 15, 04318 Leipzig, Germany

Phone +49 341 235 10 29
Fax +49 341 235 19 39
Mobile +49 178 334 18 19

Email marc.walther@ufz.de
Web http://www.ufz.de

Helmholtz-Zentrum für Umweltforschung GmbH - UFZ
Registered Office: Leipzig
Registration Office: Amtsgericht Leipzig
Trade Register Nr. B 4703
Chairman of the Supervisory Board: MinR Wilfried Kraus
Scientific Director: Prof. Dr. Georg Teutsch
Administrative Director: Dr. Heike Graßmann

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

Georg_Kosakowski · March 12, 2014, 9:00am

Hi Thomas,
the benchmark HS_LONG_C_comedy2d_cement2d.html
<https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/HS_LONG_C_comedy2d_cement2d.html>
fails, after you uploaded bugfixes regarding name search and surface
type definition in the trunk about 1 month ago. I changed the gli i file
for that benchmark and it runs again (at least on my machine). As you
know best what you fixed, can you check the changed gli file, if it
works according to your bugfixes?
Greetings, Georg

cement2d.gli (1.48 KB)

···

On 03/11/2014 03:45 PM, Wenqing Wang wrote:

Dear All,

I have merged my branch to the trunk as ver. 5.5.3, and posted a
release notes on wiki. The following benchmark results have to be
confirmed due to the changes in calculation of material parameters in
the previous version:

Last Successful Artifacts
<https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/>

  AKS_LONG_H_us_Wet_h_us_line_celia.html
<https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/AKS_LONG_H_us_Wet_h_us_line_celia.html>
  901,55 KB [fingerprint]
<https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/AKS_LONG_H_us_Wet_h_us_line_celia.html/*fingerprint*/>

  BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html
<https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html>
  127,17 KB [fingerprint]
<https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html/*fingerprint*/>

  BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html
<https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html>
  108,80 KB [fingerprint]
<https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html/*fingerprint*/>

  BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html> 86,19 KB [fingerprint] <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html/*fingerprint*/>
  BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html> 86,19 KB [fingerprint] <https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html/*fingerprint*/>
  HS_LONG_C_comedy2d_cement2d.html
<https://svn.ufz.de/jenkins/job/Benchmarks_Linux/lastSuccessfulBuild/artifact/benchmarks/results/HS_LONG_C_comedy2d_cement2d.html>

Cheers,

Wenqing

On 07.03.2014 10:54, Wenqing Wang wrote:

Dear Nori,

It is in my previous email attached here. I did not change my branch
since my last commit.

Shall we merge your branch before, or you do merging after? I want
to do this because that Tom and me have to do something on it.

Cheers,

Wenqing

-------- Original Message --------
Subject: PETSC version in my branch
Date: Mon, 03 Feb 2014 17:51:30 +0100
From: Wenqing Wang <wenqing.wang@ufz.de>
To: ogs-users@googlegroups.com

Since the source code of the partition has been changed in last year
(https://github.com/ufz/mesh_partition\), which is set to generate
partitioned mesh in binary files by default, I'd better to put a link to
the corresponding source code here:
https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc
It is my branch based on revision 12678 of the trunk, which is very
close to the head of the trunk and benchmarked. In this branch, the
major changes are for PETSc routines:
1. Parallel mesh data input, and parallel domain data output. For a
large mesh, this improves I/O significantly.
2. Deformation

I am not sure when I can merge it back. If you are developing with PETSc
or want to explore, you can take a look and merge it. Other information
can be found on the updated page: https://svn.ufz.de/ogs/wiki/PETScPage

On 07.03.2014 10:44, Norihiro Watanabe wrote:

Dear Wenqing,

Could you please tell us a summary of what you have changed there?

Best,
Nori

On 03/07/2014 10:32 AM, Wenqing Wang wrote:

Dose anybody mind me to merge my branch
https://svn.ufz.de/svn/ogs/branches/Multiphase/new_petsc to the trunk?

Wenqing

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

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

--
PD Dr. Georg Kosakowski * Laboratory for Waste Management
Paul Scherrer Institut * 5232 Villigen/PSI * Switzerland
Phone: +41-56-3104743 * Fax: +41-56-3102821
Email: georg.kosakowski@psi.ch h323:+41563104743
homepage: Telefonbuch & E-Mail | Paul Scherrer Institut (PSI)

wenqing.wang · March 14, 2014, 2:59pm

With the recent bugfuxing, there are still benchmarks fail in the results comparison:

     Workspace
     Last Successful Artifacts
     AKS_LONG_H_us_Wet_h_us_line_celia.html 901,55 KB [fingerprint]
BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html 127,17 KB [fingerprint]
BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html 108,80 KB [fingerprint]
BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html 86,19 KB [fingerprint]
BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html 86,19 KB [fingerprint]
(The newly failed: )
     CB_C_ChemAppCoupling_cation_exchange_cmp9.html 1,01 KB [fingerprint]
     CB_C_ChemAppCoupling_slow_kin_cap_act_precalc_cmp8.html 1,18 KB [fingerprint]
CB_C_ChemAppCoupling_Wagrien_BatchEqui_PVLE_CAP_wagrien_1D.html

bastian.graupner · March 14, 2014, 3:43pm

I will work on it next week.

Best regards,

Bastian

···

-----Ursprüngliche Nachricht-----
Von: ogs-devs@googlegroups.com [mailto:ogs-devs@googlegroups.com] Im Auftrag von Wenqing Wang
Gesendet: Freitag, 14. März 2014 16:00
An: ogs-devs@googlegroups.com
Betreff: Some benchmarks fail

With the recent bugfuxing, there are still benchmarks fail in the results comparison:

     Workspace
     Last Successful Artifacts
     AKS_LONG_H_us_Wet_h_us_line_celia.html 901,55 KB [fingerprint]
BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html 127,17
KB [fingerprint]
BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html 108,80
KB [fingerprint]
BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html
86,19 KB [fingerprint]
BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html
86,19 KB [fingerprint]
(The newly failed: )
     CB_C_ChemAppCoupling_cation_exchange_cmp9.html 1,01 KB [fingerprint]
     CB_C_ChemAppCoupling_slow_kin_cap_act_precalc_cmp8.html 1,18 KB
[fingerprint]
CB_C_ChemAppCoupling_Wagrien_BatchEqui_PVLE_CAP_wagrien_1D.html

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

wenqing.wang · March 14, 2014, 4:31pm

Hi Bastian,

I have just checked your benchmarks again after we talked. After the commit entitled 'bugfix for RICHARDS-FLOW' by Marc, some of your the benchmarks fail to run. With previous commit just before that, things go well. If you think Marc's changes are OK, we can updates the reference results.

Have a nice weekend.

Wenqing

···

On 14.03.2014 16:43, Graupner Bastian wrote:

I will work on it next week.

Best regards,

Bastian

-----Ursprüngliche Nachricht-----
Von: ogs-devs@googlegroups.com [mailto:ogs-devs@googlegroups.com] Im Auftrag von Wenqing Wang
Gesendet: Freitag, 14. März 2014 16:00
An: ogs-devs@googlegroups.com
Betreff: Some benchmarks fail

With the recent bugfuxing, there are still benchmarks fail in the results comparison:

      Workspace
      Last Successful Artifacts
      AKS_LONG_H_us_Wet_h_us_line_celia.html 901,55 KB [fingerprint]
BG_ECLIPSE_DUMUX_2phase_flow_2phase_tracertransport_2pf_2pt.html 127,17
KB [fingerprint]
BG_ECLIPSE_DUMUX_2phase_flow_radialmodel_2pf_radialmodel.html 108,80
KB [fingerprint]
BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E100_CO2phase_gen_E100.html
86,19 KB [fingerprint]
BG_ECLIPSE_DUMUX_kinetic_CO2phase_generation_E300_CO2phase_gen_E300.html
86,19 KB [fingerprint]
(The newly failed: )
      CB_C_ChemAppCoupling_cation_exchange_cmp9.html 1,01 KB [fingerprint]
      CB_C_ChemAppCoupling_slow_kin_cap_act_precalc_cmp8.html 1,18 KB
[fingerprint]
CB_C_ChemAppCoupling_Wagrien_BatchEqui_PVLE_CAP_wagrien_1D.html

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

wenqing.wang · March 14, 2014, 4:34pm

sorry, 'fail to run' --> 'fail in the result comparison'

···

On 14.03.2014 17:31, Wenqing Wang wrote:

Hi Bastian,

I have just checked your benchmarks again after we talked. After the commit entitled 'bugfix for RICHARDS-FLOW' by Marc, some of your the benchmarks fail to run. With previous commit just before that, things go well. If you think Marc's changes are OK, we can updates the reference results.

Have a nice weekend.

Wenqing