Coupled Two phase flow (PS type) and Deformation flow

Mohammad Islam

unread,

Jan 8, 2019, 11:51:23 AM1/8/19

to ogs-users

Hi All,

I am trying to couple two-phase flow (PS) with deformation flow for a fractured composite medium. If it is individual two-phase flow (PS) it is working fine. But, the deformation flow is not satisfactory.

Let me explain the problem details.

(1) At first, I did Kueper problem-PS type benchmark solution.

(2) Before my coupled problem, I have attempted to solve coupled problem (H2M considering PS) for the KETZIN field case (Kolditz et al. (2012-a), Page 619).

After reading Kolditz et al (2012, Chapter 12, Page 254), I understood that "global scheme is sensitive to matrix solvers. The LIS solver (BiCG with Jacobi preconditioned) works well on Windows". Therefore, I used BiCG solver and Jacobi preconditioner.

In general, the distribution of CO2 with respect to time should be similar to the distribution of stress with respect to time. However, in my case they are not similar. Any idea to couple two-phase flow (PS) with the defromation flow is appreciable. Is there any simialr example problem?

Thanks in advance.

Regards,

Mohammad

Relevant input files are as follows

GeoSys-BC: Boundary Conditions ------------------------------------------------

#BOUNDARY_CONDITION

$PCS_TYPE

PS_GLOBAL

$PRIMARY_VARIABLE

PRESSURE1

$GEO_TYPE

POLYLINE D_BC

$DIS_TYPE

CONSTANT 6.49e6

#BOUNDARY_CONDITION

$PCS_TYPE

PS_GLOBAL

$PRIMARY_VARIABLE

SATURATION2

$GEO_TYPE

POLYLINE D_BC

$DIS_TYPE

CONSTANT 0.05

#BOUNDARY_CONDITION

$PCS_TYPE

DEFORMATION

$PRIMARY_VARIABLE

DISPLACEMENT_X1

$GEO_TYPE

POLYLINE D_BC

$DIS_TYPE

CONSTANT 0

#BOUNDARY_CONDITION

$PCS_TYPE

DEFORMATION

$PRIMARY_VARIABLE

DISPLACEMENT_Y1

$GEO_TYPE

POLYLINE OUT

$DIS_TYPE

CONSTANT 0

#BOUNDARY_CONDITION

$PCS_TYPE

DEFORMATION

$PRIMARY_VARIABLE

DISPLACEMENT_Y1

$GEO_TYPE

POLYLINE TOP

$DIS_TYPE

CONSTANT 0

#STOP

GeoSys-IC Initial Conditions ------------------------------------------------

#INITIAL_CONDITION

$PCS_TYPE

PS_GLOBAL

$PRIMARY_VARIABLE

PRESSURE1

$GEO_TYPE

DOMAIN

$DIS_TYPE

CONSTANT 6.49e6

#INITIAL_CONDITION

$PCS_TYPE

PS_GLOBAL

$PRIMARY_VARIABLE

SATURATION2

$GEO_TYPE

DOMAIN

$DIS_TYPE

CONSTANT 0.05

#INITIAL_CONDITION

$PCS_TYPE

PS_GLOBAL

$PRIMARY_VARIABLE

SATURATION2

$GEO_TYPE

POLYLINE U_BC

$DIS_TYPE

CONSTANT 0.65

#INITIAL_CONDITION

$PCS_TYPE

DEFORMATION

$PRIMARY_VARIABLE

STRESS_XX

$GEO_TYPE

SUB_DOMAIN

3

0 -6005191.5+7798.95*z

1 -6005191.5+7798.95*z

2 -6005191.5+7798.95*z

#INITIAL_CONDITION

$PCS_TYPE

DEFORMATION

$PRIMARY_VARIABLE

STRESS_YY

$GEO_TYPE

SUB_DOMAIN

3

0 -6005191.5+7798.95*z

1 -6005191.5+7798.95*z

2 -6005191.5+7798.95*z

#INITIAL_CONDITION

$PCS_TYPE

DEFORMATION

$PRIMARY_VARIABLE

STRESS_ZZ

$GEO_TYPE

SUB_DOMAIN

3

0 -20017305+25996.5*z

1 -20017305+25996.5*z

2 -20017305+25996.5*z

#STOP

GeoSys-MAT-FP: Fluid Properties ------------------------------------

#FLUID_PROPERTIES

$FLUID_TYPE

LIQUID

$DENSITY

1 1.17300e+003

$VISCOSITY

1 1.25200e-003

#FLUID_PROPERTIES

$FLUID_TYPE

GAS

$DENSITY

1 0.84800e+003

$VISCOSITY

1 8.1000e-005

#STOP

GeoSys-MMP Material Medium Properties ------------------------------------

#MEDIUM_PROPERTIES

$GEOMETRY_DIMENSION

2

$GEOMETRY_AREA

1.000000e+000

$POROSITY

1 0.26

$PERMEABILITY_TENSOR

ISOTROPIC 3.e-13

$PERMEABILITY_SATURATION

6 0.35 0.95 2.0

66 0.0 1.0 2.0 1e-7

$CAPILLARY_PRESSURE

6 1.0e4

#STOP

GeoSys-MAT-SP: Solid Properties ------------------------------------

#SOLID_PROPERTIES

$DENSITY

1 2650.0

$ELASTICITY

POISSION 0.3

YOUNGS_MODULUS:

1 2.0e+11

$BIOT_CONSTANT

1.

#STOP

GeoSys-NUM: Numerical Parameter ----------------------------------------

$OVERALL_COUPLING

;min_iter -- max_iter

1 1

#NUMERICS

$PCS_TYPE

PS_GLOBAL

$ELE_MASS_LUMPING

1

$ELE_UPWINDING

0 1.0

$LINEAR_SOLVER

; method error_tolerance max_iterations theta precond storage

3 6 1.e-10 1000 1.0 1 4

$NON_LINEAR_SOLVER

; method error_tolerance max_iterations relaxation

PICARD 1e-5 30 1.0

$COUPLING_CONTROL

;error method -- tolerances

;LMAX 1.e-3 1.e-3

LMAX 10 1.e-3

#NUMERICS

$PCS_TYPE

DEFORMATION

$LINEAR_SOLVER

; method error_tolerance max_iterations theta precond storage

2 2 1.e-12 2000 1.0 100 4

$COUPLING_CONTROL

;error method -- tolerances

LMAX 1.e-3

#STOP

GeoSys-PCS: Processes ------------------------------------------------

#PROCESS

$PCS_TYPE

PS_GLOBAL

$NUM_TYPE

dPcdSwGradSnw

#PROCESS

$PCS_TYPE

DEFORMATION

#STOP

GeoSys-ST: Source Terms ------------------------------------------------

#SOURCE_TERM

$PCS_TYPE

PS_GLOBAL

$PRIMARY_VARIABLE

SATURATION2

$GEO_TYPE

POLYLINE U_BC

$DIS_TYPE

;CONSTANT_NEUMANN 0.4475e-5

CONSTANT_NEUMANN 1710.e-5

#STOP

Kolditz, O., S. Bauer, C. Beyer, N. Böttcher, P. Dietrich, U.-J. Görke, T. Kalbacher, C.-H. Park, U. Sauer, C. Schütze, H. Shao, A. Singh, J. Taron, W. Wang and N. Watanabe (2012-a). "A systematic benchmarking approach for geologic CO2 injection and storage." Environmental Earth Sciences 67(2): 613-632.

Kolditz, O., U.-J. Gorke, H. Shao and W. Wang (2012). Thermo-Hydro-Mechanical-Chemical Processes in Fractured Porous Media: Benchmarks and Examples. Heidelberg, Springer.

Mohammad Islam

unread,

Jan 9, 2019, 2:30:43 PM1/9/19

to ogs-users

Hi All,

Can any one explain little more about the keyword "NEGLECT_H_INI_EFFECT". In which case it will be 2 or 0? It looks this keyword has some impact in the modeling.

Regards,

Mohammad

Wenqing Wang

unread,

Jan 10, 2019, 3:57:03 AM1/10/19

to ogs-...@googlegroups.com

If NEGLECT_H_INI_EFFECT=1, the initial pore pressure and the initial effective stress are excluded from the momentum balance equation, i.e. in the momentum balance equation the total stress is defined as

sigma = sigma^e-sigma^e_0-\alpha_B(p-p_0)

where ^e stands for effective, \alpha_B is the Biot's constant.

In other words, if NEGLECT_H_INI_EFFECT=1, the initial total stress does not give displacement change.

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

Mohammad Islam

unread,

Jan 10, 2019, 10:02:01 AM1/10/19

to ogs-users

Dear Wenqing Wang,

Thank you for your explanation.

I checked rf_pcs.cpp

Line # 383 : Neglect_H_ini = -1; // WX

Line # 1736: dm_pcs->Neglect_H_ini = Neglect_H_ini; // WX:08.2011

dm_pcs->UpdateIniState = UpdateIniState; // WX:10.2011

Line # 2141 if (line_string.find("$NEGLECT_H_INI_EFFECT") == 0) // WX:10.2011

Line # 2146 if (line_string.find("$UPDATE_INI_STATE") == 0) // WX:10.2011

My question is what is the function of UPDATE_INI_STATE. If I don't use NEGLECT_H_INI_EFFECT, fluid flow direction is ok, but deformation flow direction has some issues.

On the other hand, with the same geometry and properties, H2M for the PP scheme is working well, but in PS sceme there is no change in the DEFORMATION flow. Is there any example problem for the H2M (PS scheme)?

Thnaks in advance.

Regards,

Mohammad

Wenqing Wang

unread,

Jan 11, 2019, 4:36:08 AM1/11/19

to ogs-...@googlegroups.com

* UPDATE_INI_STATE is used to update initial value data of variables. It can be used for time stepping excavation modelling.

* As explained before, the use of EGLECT_H_INI_EFFECT = (1 or 2) avoids the displacement change by the initial stress. Better to use EGLECT_H_INI_EFFECT=2

* H2M (PS scheme) has not tested so far.

Mohammad Islam

unread,

Jan 11, 2019, 9:21:21 AM1/11/19

to ogs-users

Dear Wenqing Wang,

Thank you again.

Regards,

Mohammad

Reply all

Reply to author

Forward