[Geoqus] Effects of Contact in a Poroelastic Inclusion Model

[Constantin Haug]

Dear Geoqus member,

I’ve been encountering some issues when modelling contact in a coupled 
pore pressure displacement analysis in Abaqus 2016.

I’m modelling transient pore pressure depletion in a rectangular 2D 
inclusion. Pore pressure remains hydrostatic throughout the analysis in 
the over- under- and sideburden. The model is completely discretized by 
CPE4P elements.

-material is linear (poro-)elastic

-loading is gravitational

-model boundaries are constrained in their normal direction

In theory pore pressure depletion leads to an increase of effective 
stresses and a contraction of the inclusion.

I compare two cases:

1.The inclusion is bounded by an impermeable tie constraint (Figure 1)

2.The inclusion is bounded by contact on its lateral boundary, extending 
into the over- and underburden (Figure 2). The contact should act as a 
flow barrier, hence a very low */Contact Permeability /is implemented.

For the first model version (no contact) stress and displacement changes 
are qualitatively as expected from analytical solutions.

asdfsad

The inclusion contracts and effective horizontal stress increases as a 
function of proximity to the inclusion’s lateral boundary. Outside of 
the reservoir, horizontal stress relaxes as a consequence of the 
inclusion’s contraction.

asdfsd

Introducing contact into the model the stress and displacement changes 
show roughly the opposite pattern. The inclusion expands and outside the 
exclusion horizontal loading occurs which is not at all the expected and 
the considered correct result.

No slip occurs during both analysis steps. The contact status is closed 
(CSTATUS=CLOSED) throughout the whole modelling time. According to 
Abaqus Documentation 37.4.1, the normal contact pressure (CPRESS) is 
calculated from effective stresses if both sides of the contact have 
pore pressure degrees of freedom.

I’m using the following contact definitions:

**SURFACE INTERACTION, NAME = Fault1*

**CONTACT PERMEABILITY*

*1E-50*

**SURFACE BEHAVIOR, PENALTY, NO SEPARATION, PRESSURE-OVERCLOSURE = HARD*

**FRICTION*

*0.7,0.0,0.0,0.0*

**CONTACT PAIR, INTERACTION=Fault1, TYPE= SURFACE TO SURFACE*

*Westfault, Eastfault*

Has anyone encountered similar issues in poroelastic contact modelling? 
Has anyone got any idea what could be wrong with my contact definition, 
leading to my unwanted results?

I’ve already been modifying the following parameters:

-Pore pressure boundary conditions

     all sideburden nodes active vs nodes on models edge active

-*/Contact Permeability/ and permeability of sideburden

-/Augmented Lagrange /contact formulation

-reduction of penetration tolerance

-swapping of master and slave surface

-modification of contact stiffness

Yet stress changes with contact are still not the ones expected from the 
analytical solution and the model without contact.

Thanks in advance for any input or discussion

-- 
Freundliche Grüße / Kind regards

Constantin Haug

..................................................
M.Sc. Constantin Haug
Institut für Angewandte Geowissenschaften / Institue of Applied Geosciences
Technische Universität  Darmstadt

Schnittspahnstraße 9
D-64287 Darmstadt
Germany

Tel.: ++49 - (0) - 6151 - 1622347
Fax: ++49 - (0) - 6151 - 166539
email:haug at geo.tu-darmstadt.de
Homepage:http://www.geo.tu-darmstadt.de

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.ruhr-uni-bochum.de/pipermail/geoqus/attachments/20170208/a9d1d47d/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/png
Size: 663822 bytes
Desc: not available
URL: <http://lists.ruhr-uni-bochum.de/pipermail/geoqus/attachments/20170208/a9d1d47d/attachment.png>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: image/png
Size: 663148 bytes
Desc: not available
URL: <http://lists.ruhr-uni-bochum.de/pipermail/geoqus/attachments/20170208/a9d1d47d/attachment-0001.png>