Hi Timo,
Sorry, I was not clear in the last email. The 2p simulation is also run with
DuMux using Box method and AMG-Backend.
As I understand, the box method would be the best choice, or at least the most
easy to implement, for simulations in such non-manifold grids. So eventually,
the problems in my 1p and 2p simulations are due to the conditions at the
boundaries. Please let me know if you have any advance in debugging.
Thank you very much and have a nice day.
Best wishes,
Tri Dat
-----Original Message-----
From: "Timo Koch" <timo.k...@iws.uni-stuttgart.de>
Sent: 7/23/2016 1:29 PM
To: "dumux@listserv.uni-stuttgart.de" <dumux@listserv.uni-stuttgart.de>; "NGO
Tri Dat" <tri-dat....@ifpen.fr>
Subject: Re: [DuMuX] Convergence problem with FoamGrid 2D3D + Box method
Hi Tri Dat,
On 22.07.2016 18:41, NGO Tri Dat wrote:
Hi Timo,
Thank you for your feedback!
For information, I have also a simulation of 2p flow in another 2d-in-3d DFN.
I considered the behavior of a LNAPL plume in a medium initially saturated with
water, hydrostatic pressure for initial condition (figure 1 & 2).
A no-flow Neumann BC is prescribed for the top and the bottom faces, and
Dirichlet BC otherwise.
Did you run that with Dumux?
The simulation run correctly (saturation Sn at the final time of simulation,
figure 3) but I got a strange pressure field at a fracture which is always
saturated in water (figure 4).
I’m not sure that this problem is due to boundary condition or the condition at
the intersection between fractures. In addition, the distribution of fluxes at
the intersections between fractures when using the box method is not very
clear for me. Is the mass conservation at these locations ensured with box
method?
Mass conservation should come naturally with the box method at intersections.
As the degrees of freedom are on the intersection you enforce continuity of
pressure.
The fluxes are only computed on the box boundaries with are inside the elements
meeting at the intersection. The box control volume basically spans over all
three (or more)
intersecting elements at the fracture joint. So without doing anything special
the box method should ensure mass conservation here.
If you use cell-centered finite volumes this is different. First, the gradients
in the current dumux master branch would be approximated incorrectly in a 2d in
3d setting (basically cutting over the edge at an intersection). Secondly, the
intersections with neighboring elements of the other two (or more) fracture
will be handled as independent intersections. So to ensure the flow is
spreading right you would have to take care about mass conservation yourself. A
common approach (at least for one-phase flow, I'm not sure about 2-phase flow)
is a star-delta type analogy from electrical engineering. The flow (at an
intersection with three or more fractures) from one element to another then
depends on transmissibilities of all participating fractures.
Best wishes,
Timo
Kind regards,
Tri Dat
De : Timo Koch [mailto:timo.k...@iws.uni-stuttgart.de]
Envoyé : vendredi 22 juillet 2016 17:39
À : DuMuX User Forum; NGO Tri Dat
Objet : Re: [DuMuX] Convergence problem with FoamGrid 2D3D + Box method
Hi Tri Dat,
Bernd was right and you were probably the first one computing on a 2d in 3d
grid with box.
I looked into it today. The box sub control volume face computed the wrong
normal for dim=2 and dimWorld=3.
I fixed this on a branch feature/ discrete-fracture-2d3d-test. This makes
everything converge for my little fracture network. However, something on the
boundaries looks still strange (saturation greater than 1, I tested the 2p
model).
I implemented the same problem on our feature/globalvectors branch that has
some major changes and improvements over the current version. There,
everything looked fine.
We have to look into the issue on the boundaries, so stay tuned.
Best wishes
Timo
On 22.07.2016 14:15, NGO Tri Dat wrote:
Hi Bernd,
You can find, in the attached file, the source files of the test case.
Please let me know if you cannot compile the program.
Thank you for your help.
Kind regards,
Tri Dat
De : Dumux [mailto:dumux-boun...@listserv.uni-stuttgart.de] De la part de Bernd
Flemisch
Envoyé : vendredi 22 juillet 2016 12:09
À : DuMuX User Forum
Objet : Re: [DuMuX] Convergence problem with FoamGrid 2D3D + Box method
Hi Tri Dat,
you might be the first one that tries to run Dumux on a non-manifold 2d-in-3d
grid, so I'm not that surprised that you run into problems. Can you send me the
grid, problem, spatialparams and main files so that we can have a closer look?
Kind regards
Bernd
On 07/22/2016 11:51 AM, NGO Tri Dat wrote:
Hi Dumuxers,
I have a convergence problem with my simulations of 1p flow in a DFN using
FoamGrid (as in the attached file). The fractures is 2D embedded in a 3D bloc
of 100x100x8 m^3. The matrix is assumed to be impervious. A pressure gradient
is prescribed along the X-axis. Box method and AMG Backend are used.
The error message is:
Dune reported error: Dune::MathError
[timeIntegration:/work/irlin104_1/ngotr/DUNE/dumux-2.9.0/dumux/implicit/problem.hh:662]:
Newton solver didn't converge after 10 time-step divisions. dt=9.76563e-14.
The solutions of the current and the previous time steps have been saved to
restart files.
Any hints will be greatly appreciated.
Kind regards,
Tri Dat Ngo
__________________________
Avant d'imprimer, pensez à l'environnement ! Please consider the environment
before printing !
Ce message et toutes ses pièces jointes sont confidentiels et établis à
l'intention exclusive de ses destinataires. Toute utilisation non conforme à sa
destination, toute diffusion ou toute publication, totale ou partielle, est
interdite, sauf autorisation expresse. IFP Energies nouvelles décline toute
responsabilité au titre de ce message. This message and any attachments are
confidential and intended solely for the addressees. Any unauthorised use or
dissemination is prohibited. IFP Energies nouvelles should not be liable for
this message.
__________________________
_______________________________________________Dumux mailing
listdu...@listserv.uni-stuttgart.dehttps://listserv.uni-stuttgart.de/mailman/listinfo/dumux
-- _______________________________________________________________ Bernd
Flemisch phone: +49 711 685 69162IWS, Universität
Stuttgart fax: +49 711 685 60430Pfaffenwaldring 61
email: be...@iws.uni-stuttgart.ded-70569 Stuttgart url:
www.hydrosys.uni-stuttgart.de_________________________________________________________________________________________
Avant d'imprimer, pensez à l'environnement ! Please consider the environment
before printing !
Ce message et toutes ses pièces jointes sont confidentiels et établis à
l'intention exclusive de ses destinataires. Toute utilisation non conforme à sa
destination, toute diffusion ou toute publication, totale ou partielle, est
interdite, sauf autorisation expresse. IFP Energies nouvelles décline toute
responsabilité au titre de ce message. This message and any attachments are
confidential and intended solely for the addressees. Any unauthorised use or
dissemination is prohibited. IFP Energies nouvelles should not be liable for
this message.
__________________________
_______________________________________________Dumux mailing
listdu...@listserv.uni-stuttgart.dehttps://listserv.uni-stuttgart.de/mailman/listinfo/dumux
-- ____________________________________________________________________ Timo
Koch phone: +49 711 685 64676IWS, Universität Stuttgart
fax: +49 711 685 60430Pfaffenwaldring 61 email:
timo.k...@iws.uni-stuttgart.ded-70569 Stuttgart url:
www.hydrosys.uni-stuttgart.de______________________________________________________________________________________________
Avant d'imprimer, pensez à l'environnement ! Please consider the environment
before printing !
Ce message et toutes ses pièces jointes sont confidentiels et établis à
l'intention exclusive de ses destinataires. Toute utilisation non conforme à sa
destination, toute diffusion ou toute publication, totale ou partielle, est
interdite, sauf autorisation expresse. IFP Energies nouvelles décline toute
responsabilité au titre de ce message. This message and any attachments are
confidential and intended solely for the addressees. Any unauthorised use or
dissemination is prohibited. IFP Energies nouvelles should not be liable for
this message.
__________________________
_______________________________________________
Dumux mailing list
Dumux@listserv.uni-stuttgart.de
https://listserv.uni-stuttgart.de/mailman/listinfo/dumux
--
____________________________________________________________________
Timo Koch phone: +49 711 685 64676
IWS, Universität Stuttgart fax: +49 711 685 60430
Pfaffenwaldring 61 email: timo.k...@iws.uni-stuttgart.de
D-70569 Stuttgart url: www.hydrosys.uni-stuttgart.de
____________________________________________________________________
_______________________________________________
Dumux mailing list
Dumux@listserv.uni-stuttgart.de
https://listserv.uni-stuttgart.de/mailman/listinfo/dumux
