list of things that need to be done

Potential tasks/subjects

• Meshing

• MOOSE functionalities

• Learn how to place a mutli-app in a box (rather than single point) -Redback software architecture:

• memory blow out (using so many material properties). Need to run profiler (Instruments?) Currently we are duplicating a lot of matrix variables in our calculation. Need optimizing (but only after we get the right results, of course)

• clever memory allocation based on AuxVars turned on -move towards using TensorMechanics in its new form (action, userobjects, etc.)

• we would like to be able to use the Moose TensorMechanics module with API calls and also (eventually) be able to contribute our stuff upstream; current state is copy-paste based.

• Bugs

• sure to be plenty

• Action to ask for blocks

• Clean-up code

• (trivial) Remove yield in RedbackMechElastic

• (trivial) Remove “problems” folder

• default values for RedbackMaterial so we can use ElasticMaterial on its own (for Eleni)

• ElasticMaterial clean-up, don’t ask for yield and other useless stuff…

• Static analyser; syntax checker

• Atom.io

• investigate the Moose de-linter; we would like to get our code conformant to Moose standards with minimal effort on our part

• New features

• setting T0 (and P0) for density reference temperature (and pore pressure) as fields from initial step rather than constants

• Navier-Stokes incompressible

• parameter optimization: investigate parameter optimization frameworks like RAVEN (a MOOSE-based code) or Dakota (from Sandia labs). See http://www.diegomandelli.com/papers/2013/MC/MC2013RavenCode/ravenMC2013.pdf for example.

• Documentation

• General document (English, conventions, etc)

• Doxygen

• this is working, but there's nothing on it yet. we would like our comments to conform.

• Benchmarks

• memory and cpu

• Footing problem

• apply stress on part of a matrix block, fluid comes out; what's happening in detail?

• Clean up tests (and document that some only exercise some redback objects, don’t have analytical or solution or anything to compare…)

• unit tests

• Check OpenGeoSys for other benchmarks we could implement

• Mech only (e.g. MCC, https://www.rocscience.com/help/phase2/webhelp9/pdf_files/verification/Phase2_StressVerification_Part1.pdf)

• HPC

• based in UNSW, we have access to http://leonardi.unsw.wikispaces.net, which is an extremely powerful supercomputer. We just need to figure out how to take advantage of it.

• Install MOOSE on Leonardi

• Theory

• Navier-Stokes boundary conditions

• Pore collapse

• Matching sandstone