Poroelastic fracture, with broken permeability tensor #4

kmokstad · 2016-05-03T05:15:32Z

This one replaces #3. Contains in addition a method for calculation of the anisotropic permeability of the cracked material, based on Miehe (2015), Part III. Please take a look at it.

The PoroElasticity class now needs a virtual version of evalPermeabilityMatrix in which the new method is used.

akva2 · 2016-05-24T17:40:19Z

PoroFracture.C

+
+
+/*!
+  This method calculates the anisotropic peremeability for the broken solid


permeabilty

akva2 · 2016-05-25T15:07:48Z

vennligst inkluder en fiks for

In file included from /var/lib/jenkins/jobs/IFEM-PR-builder/workspace/deps/IFEM-OpenFrac/FractureElasticityVoigt.C:14:
/var/lib/jenkins/jobs/IFEM-PR-builder/workspace/deps/IFEM-OpenFrac/FractureElasticityVoigt.h:14:9: warning: '_FRACTURE_ELASTICITY_VOIGT_H' is used as a header guard here, followed by #define of a different macro [-Wheader-guard]
#ifndef _FRACTURE_ELASTICITY_VOIGT_H
        ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
/var/lib/jenkins/jobs/IFEM-PR-builder/workspace/deps/IFEM-OpenFrac/FractureElasticityVoigt.h:15:9: note: '_FRACTURE_ELASTICITY_VOIGH_H' is defined here; did you mean '_FRACTURE_ELASTICITY_VOIGT_H'?
#define _FRACTURE_ELASTICITY_VOIGH_H
        ^~~~~~~~~~~~~~~~~~~~~~~~~~~~
        _FRACTURE_ELASTICITY_VOIGT_H

vil ikke røre brancher du holder som gissel..

kmokstad · 2018-01-18T16:22:56Z

jenkins build this please

akva2

Some doxy suggestions

akva2 · 2018-01-18T20:24:22Z

PoroFracture.C

+//!
+//! \author Knut Morten Okstad / SINTEF
+//!
+//! \brief Integrand implementations for elasticity problems with fracture.


akva2 · 2018-01-18T20:24:50Z

PoroFracture.C

+
+
+/*!
+  \brief Class representing the integrand of elasticity problems with fracture.


akva2 · 2018-01-18T20:30:15Z

PoroFracture.h

+  virtual std::string getField2Name(size_t i, const char* prefix) const;
+
+protected:
+  //! \brief Computes the elasticity matrices for a quadrature point.


akva2 · 2018-02-08T15:29:58Z

PoroFracture.h

+  //! \param[in] elem_sizes Size of each basis on the element
+  //! \param[in] basis_sizes Size of each basis on the patch level
+  //! \param elmInt The local integral object for current element
+  virtual bool initElement(const std::vector<int>& MNPC,


som sagt i møtet; initElement har en mixed-overload men det er ingen evalIntMx. men jeg ser nå at det bruker default overload'en (kall non-mixed) så det bør være okay.

Ja, jeg mener å huske at -mixed var operativt når Eivind drev og prøvde for noen år siden.
Jeg har kun en input-file liggende Miehe71.xinp i Test-katalogen, men intet test-opsett forøvring.

akva2 · 2018-02-13T09:19:36Z

SIMExplPhaseField.C

+
+SIMExplPhaseField::SIMExplPhaseField (SIMoutput* gridOwner)
+{
+  myHeading = "Explcit phase field";


akva2 · 2018-02-13T09:25:47Z

SIMExplPhaseField.C

+bool SIMExplPhaseField::parse (const TiXmlElement* elem)
+{
+  const char* value = utl::getValue(elem,"phasefield");
+  if (!value) return false;


this doesn't seem right. we are passed all subtags of <simulation>, not all will match phasefield and returning false here will halt parsing.

Indeed wrong. Should here forward to parent-class method for all other tags. Don't know when that vanished but fixed now. Probably something went wrong during a rebase at one point.

akva2 · 2018-02-13T11:28:13Z

the explicit phase field construct is rather problematic for mixed. it uses the model from the elasticity problem, which is mixed. while i can fix the projection to only project onto basis 1, it's not fixable when extracting the patch dependencies. as we are in a SIMdependency scope there, i do not have access to the additional MADOF required for a single component on first basis (which has to sit in the elasticity SIM). i see no clean way around setting up a model in the phase sim itself.

kmokstad · 2018-02-14T05:33:03Z

Have to think about that. I see the program crashes now running with -explcrack and -mixed, while it runs (but diverges) without the -mixed option. I guess these two options have never been combined before, unless @TheBB remembers doing so back in 2016?

akva2 · 2018-08-24T09:30:36Z

fixed explicit phase field plus mixed. it's not entirely pretty as this is not using the established mechanisms to handle this (fields). you need https://github.com/akva2/IFEM-OpenFrac/tree/fix_poro_expl and https://github.com/akva2/IFEM/tree/fix_expl_mixed

…asticity without the phase-field coupling, but still with the strain energy split

… point, and the broken permeability tensor based on phase field and deformation state

Fixed: Removed extraction of element velocity and accelereation in PoroFracture::initElement, it is done by the parent class method.

and new command-line options -explcrack to activate it

akva2 · 2018-09-06T13:39:47Z

jenkins build this ifem=286 ifem-poroelasticity=27 valgrind please

akva2 · 2018-09-06T13:41:27Z

jenkins build this ifem=286 ifem-poroelasticity=27 valgrind please

akva2 · 2018-09-06T13:46:50Z

jenkins build this ifem=286 ifem-poroelasticity=27 valgrind please

kmokstad · 2018-09-07T06:13:07Z

With the last change here (that is, we specify an additional MADOF also when the phase field solver is in the loop, and not only with an explicit phase field), and yes, the phase-field solver has to share grid with the poro-elastic solver, only specifying zero dofs pr node for the second basis and 1 dofs per node for the first, then with the OPM/IFEM#289 fix Miehe71.xinp with -poro -mixed runs through without valgrind issues or divergence. But the solution is totally wrong still (look at the phase field in GLview). But that is probably due to other things. I think the use of Field for the phase-field dependency when mixed should be tried out next, to see if that gives a better solution. In that case the current implementation is still buggy. If not, we have a problem with the formulation as such.

akva2 · 2018-09-07T07:36:19Z

If I

adjust ngauss to 3
adjust Gc to values more aligned to the return energies

it runs and looks sane. of course, i'm not really sure what's supposed to happen in this case though.

kmokstad · 2018-09-08T06:59:40Z

jenkins build this please

kmokstad · 2018-09-08T09:50:52Z

jenkins build this please

Filter out the pressure variables in the getElementSolution method.

such that we don't have to override the getLocalIntegral methods. Using override keyword for reimplemented virtual methods.

kmokstad · 2018-09-08T10:07:38Z

jenkins build this please

kmokstad · 2018-09-08T11:17:03Z

The last comments above (from Sep 2 and Sep 7) applies to commits which now are placed in #37 instead.

kmokstad force-pushed the poro-fracture2 branch from 32aa416 to 1dae5f4 Compare May 4, 2016 05:24

akva2 reviewed May 24, 2016
View reviewed changes

PoroFracture.C Outdated

/*!

This method calculates the anisotropic peremeability for the broken solid

Copy link

Member

akva2 May 24, 2016

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

permeabilty

kmokstad force-pushed the poro-fracture2 branch from 1dae5f4 to 194848e Compare May 25, 2016 05:27

kmokstad force-pushed the poro-fracture2 branch from 194848e to 51f8b42 Compare May 31, 2016 16:04

kmokstad mentioned this pull request Jun 2, 2016

Added: Option to use NonLinSIM and HHTSIM as the time integrator #6

Merged

kmokstad force-pushed the poro-fracture2 branch from 75a4230 to 50ff785 Compare July 7, 2016 13:33

kmokstad force-pushed the poro-fracture2 branch 3 times, most recently from 010d7ee to 749d85c Compare August 3, 2016 06:05

kmokstad force-pushed the poro-fracture2 branch from 749d85c to 89dc524 Compare December 15, 2016 16:31

kmokstad mentioned this pull request Dec 15, 2016

Diverse småtterier... OPM/IFEM#97

Merged

TheBB mentioned this pull request Jan 24, 2017

Poroelastic fracture, with broken permeability tensor #17

Closed

kmokstad force-pushed the poro-fracture2 branch from a51f1e4 to 488ac19 Compare January 27, 2017 06:56

kmokstad force-pushed the poro-fracture2 branch from 488ac19 to 02ce760 Compare January 18, 2018 16:19

akva2 reviewed Jan 18, 2018

View reviewed changes

akva2 reviewed Feb 8, 2018

View reviewed changes

kmokstad force-pushed the poro-fracture2 branch from 02ce760 to 3e05502 Compare February 13, 2018 06:10

akva2 reviewed Feb 13, 2018

View reviewed changes

kmokstad force-pushed the poro-fracture2 branch from 3e05502 to 541ef72 Compare February 14, 2018 05:27

kmokstad mentioned this pull request Aug 26, 2018

Poro fracture WIP #3

Closed

kmokstad force-pushed the poro-fracture2 branch from 541ef72 to 19bf7ae Compare August 30, 2018 15:10

kmokstad mentioned this pull request Sep 2, 2018

Use of additional MADOF in projection and extraction of scalar field on mixed basis OPM/IFEM#286

Merged

kmokstad force-pushed the poro-fracture2 branch from 19bf7ae to f3882e6 Compare September 2, 2018 06:39

kmokstad added 6 commits September 4, 2018 22:50

Added: Command-line argument -mixed and the option to run pure poroel…

c63b935

…asticity without the phase-field coupling, but still with the strain energy split

Added: Methods for evaluating the phase field and gradient at current…

656f8ba

… point, and the broken permeability tensor based on phase field and deformation state

Added: virtual method for computing the total permeability tensor

3c2131c

Added: VTF-output of crack opening and permeability coefficients

f525bfe

Fixed: Ignore crack stretch slightly less than zero.

5d79cf2

Fixed: Removed extraction of element velocity and accelereation in PoroFracture::initElement, it is done by the parent class method.

Added: class SIMExplPhaseField representing an explicit phase field,

908cd7a

and new command-line options -explcrack to activate it

kmokstad force-pushed the poro-fracture2 branch from 9828b4f to 8e7f195 Compare September 4, 2018 20:52

kmokstad mentioned this pull request Sep 6, 2018

Changed: Renamed evalSolCommon to evalSol and made it protected OPM/IFEM-PoroElasticity#27

Merged

kmokstad force-pushed the poro-fracture2 branch 2 times, most recently from 039bc48 to 5ee0407 Compare September 6, 2018 21:35

kmokstad force-pushed the poro-fracture2 branch from 372347c to 088dc66 Compare September 8, 2018 06:59

Added: Command-line option -stopTime overriding the input file setting

7cd00cf

kmokstad force-pushed the poro-fracture2 branch from 088dc66 to 0811794 Compare September 8, 2018 09:50

kmokstad and others added 5 commits September 8, 2018 12:06

Added: Test case from Miehe2015pfm3

637b1aa

Added: Projection of explicit phase field function

1857113

Fixed: Computation of cracked permeability tensor.

7aa4cb2

Filter out the pressure variables in the getElementSolution method.

Changed: Set npv correctly for mixed problems in the constructor,

1645c9a

such that we don't have to override the getLocalIntegral methods. Using override keyword for reimplemented virtual methods.

Added: ifem-poroelasticity to jenkins upstreams

0ba13d8

kmokstad force-pushed the poro-fracture2 branch from 0811794 to 0ba13d8 Compare September 8, 2018 10:07

kmokstad merged commit 0ba13d8 into OPM:master Sep 8, 2018

kmokstad mentioned this pull request Sep 8, 2018

Poro fracture mixed #37

Open



		/*!
		This method calculates the anisotropic peremeability for the broken solid



		/*!
		\brief Class representing the integrand of elasticity problems with fracture.

Poroelastic fracture, with broken permeability tensor #4

Poroelastic fracture, with broken permeability tensor #4

Uh oh!

Conversation

kmokstad commented May 3, 2016

Uh oh!

Choose a reason for hiding this comment

Uh oh!

akva2 commented May 25, 2016

Uh oh!

kmokstad commented Jan 18, 2018

Uh oh!

akva2 left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

akva2 Feb 13, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

akva2 commented Feb 13, 2018

Uh oh!

kmokstad commented Feb 14, 2018

Uh oh!

akva2 commented Aug 24, 2018

Uh oh!

akva2 commented Sep 6, 2018

Uh oh!

akva2 commented Sep 6, 2018

Uh oh!

akva2 commented Sep 6, 2018

Uh oh!

kmokstad commented Sep 7, 2018 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

akva2 commented Sep 7, 2018

Uh oh!

kmokstad commented Sep 8, 2018

Uh oh!

kmokstad commented Sep 8, 2018

Uh oh!

kmokstad commented Sep 8, 2018

Uh oh!

kmokstad commented Sep 8, 2018

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

akva2 Feb 13, 2018 •

edited

Loading

kmokstad commented Sep 7, 2018 •

edited

Loading