MercuryMonth 2020

Dates: April 20-May 20th

Due to COVID-19 this year the event will be completely  online.

Update meetings

We will have update meeting at 11am everyday

Add to your agenda

Updated link (To get added to the calendar, send an email to t.weinhart@utwente.nl):

https://calendar.google.com/event?action=TEMPLATE&tmeid=M2hwZjE5NGQ4ZzE4cHQwbTdjdGRiYWthc2RfMjAyMDA0MjBUMDkwMDAwWiB3ZWluaGFydEBt&tmsrc=weinhart%40gmail.com&scp=ALL

Joining

Join the 11am meeting (webex):

https://meetingsemea5.webex.com/meet/timoplath

PW: MDPM

Slack account

 

For this month we will be using slack:https://app.slack.com/client/TJ0BZKBNC/C012B175Z28 or the normal mercrydpm channel at  https://mercurydpm.slack.com/archives/CHZQ527RA

Topics

  1. JIRA Jobs
    1. Documentation - New branch: Documentation.  Anthony will change the root of this branch to the trunk.
      Please use svn checkout https://svn.mercurydpm.org/SourceCode/Branches/NewDocumentation/
      T
      his does not have Juan changes at the moment as they did not merge cleanly i.e. this is based on the current trunk and fully passes all selftest for me.
    2. Advanced tutorials (main developers: sit dow and make a list of tutorials necessary; link to youtube videos)
    3. New tutorials

      ---------------------------------
    • Interaction::computeNormalForce - Juan has started. - Timo will improve Hertzian and Mindlin http://docs.mercurydpm.org/Trunk/d7/dc4/ContactModel.html by himself because his info is in German. 
    • Link UnitTests to each contact model. Plot the interactions. 

      Normal contact models:
    • Linear spring dashpot model - UnitTest 
    • Hertzian spring dashpot model - UnitTest 
    • Elastoplastic contact forces - UnitTest 
    • Sintering (time-temperature dependent) - UnitTest 
    • Enhanced sintering model interaction - UnitTest 

      Friction contact models:
    • Sliding friction interaction  - UnitTest 
    • Friction interaction - Tutorial - SelftTest
    • Mindlin interaction - ---

      Adhesive contact models:
    • Dry cohesion for elastic contact forces  - Linear reversible cohesive force UnitTest 
    • Wet cohesion - Irreversible adhesive interaction - UnitTest  - SelftTest
    • Liquid migration - Irreversible adhesive interaction SelftTest
    • Charged bonded UnitTest SelftTest - Do we have any reference for this model? - A// Thomas will send me a paper
    • Pharmi McMeking sinter interaction SelftTest. This self-test is in a USER directory. Maybe could we change it to SelfTests directory ? A// Move to 

      Include check function in unit tests.
      --------------------------------- 
  2. Tutorials
  3. Anthony Thornton take look at UnitTest and see why there is selftest data
  4. Convert demos to tutorial;
  5. Consolidation of coupling (merge/make compatible Mitchel and Hong; parallelise)
    1. Oomph: Non-dimensionalisation is needed, we need to find out why. Hong uses the output function to produce dimensional output.
  6. Release
  7. Resurrect Marnix mu(I) code [Hong]
  8. STEP Files (file format for curved geometries)
  9. Sydney demo programs for teaching [Juan, Ant]
  10. Check naming of files/dir's/variables;
    1. make cleanup branch
  11. Restore old codes

Done:

  1. How to create a video from Paraview (this contains only the initial setting ".pvsm" for particles, walls setting is not there)
  2. Add logo into videos (done-one example has been updated at the end of this page)
  3. Setup Confluence questions

Missing days

Please list the days you cannot attend the daily meeting.

Thomas: May 18

Timo: April 22, May 11-13

Julius: May 12, 19 (Equipment Design Course)

Ant: : April 29.

Hao: May 5.

Mitchel: May 4-5

Juan: May 13-14

Hongyang: May 12

Yousef: May 15 / 18-19

Discussion

Everyone, please curate your own section below yourself.

  
Teach

Do you still need skills to do you tasks?

SVN, MercuryDPM basics, etc.

SlackDo you have access to slack
CalendarDid you get teh calendar invite? Anyone to add?
Juan

Documentation; Sydney demo programs for teaching - Convert MercurySimpleDemos to tutorials.

MercurySimpleDemos:

  • CubicCell.cpp
  • FreeCooling2DPeriodicDemo.cpp
  • FreeCooling3Demo.cpp
  • FreeCoolingDemo.cpp
  • FreeCoolingTryplePeriodic.cpp
  • HeaterBoundaryTest.cpp
  • HourGlass2DDemo.cpp
  • HourGlass2DDemoMPI.cpp
  • HourGlass3DDemo.cpp
  • MinimalExampleDrum.cpp
  • ParticleInclusion.cpp
  • REV_IsotropicCompression.cpp
  • REV_PurShear.cpp
  • REV_SimpleShear.cpp
  • HourGlassSelfTest.cpp Warning: This file cannot be moved

Could I move these self tests to SelfTests directory?

A// Yes

  • FreeFallSelfTest.cpp
  • InsertionBoundarySelfTestMPI.cpp
  • MasterChuteUnitTestMPI.cpp
  • ParticleCreationSelfTest.cpp
  • FreeFallInteractionSelfTest.cpp : Warning: interaction file will take up a lot of disk space! 

Could I move these unit tests to UnitTests directory?

A// Yes

  • PeriodicBoundaryUnitTestMPI.cpp
  • PeriodicChuteUnitTestMPI.cpp
  • PeriodicWallsUnitTest.cpp
  • ParallelUnitTest.cpp
  • HGridUnitTest.cpp

Which demo tests does Anthony Thornton want for teaching?

R//

  • Job MDPM 301: Update Hour Glass Demo. The description of this job is: This code is written in a weird way with respect to what N.
    To update the structure, some changes need to be done. Maybe I could rewrite this code using the tutorial format.
  • Job MDPM 282: Write a tutorial on how to create your own code
  • Documentation: Write your own code. Updated.
  • Job 306: Add a review doc of what files can be turned on
  • Job 304: Hourglass does not seem able to write multiple data files.

I've created a new beginner tutorial, tutorial11. This tutorial shows how to work with axisymmetric walls inside a cylindrical domain. Furthermore, it is based on the hourglass demo. This tutorial can write multiple data files (Job 304).

Thomas
  • Resurrect old Driver code directories (SerializedDriver, Leidenfrost, GranularJet, Avalanching, CartesianShearCell, Segregation), see r4549
  • Thomas Weinhart Contraction directory cannot be resurrected, make a new Contraction driver.
  • Anthony Thornton look at Chute directory, Codes are extremely old (#include "scr/Chute.h")
  • Send Juan paper on Chargedbonded.

  • Merge completed features from my Branches (MercuryLab, ResurrectOldDriverCodes/) into Trunk, see r4522, r4549
  • Julius Mbaziira Check whether *SelfTestMPI.cpp/*UnitTestMPI.cpp is executed if MPI=ON
    No, they are never executed. They need to be renamed *MPI2Test.cpp
  • Fixed "Mathjax not displayed" issue in documentation, see r4554
  • Add Test rules to coding style
  • Clean up Doxygen section in coding style, add /ref
  • Update license
  • Discuss: HertzMindlin tied together, or shear modulus updated from Hertz.
Ant
  1. Change settings of MercuryDPM website [done]
  2. Make mercurydpm forward-only email adresses work (info, personal, etc)
  3. Backup script for web pages
  • Finish off oomph-lib testing branch, so we have an oomphlib branch we can commit to
Timo
  • Documentation Mindlin
  • Enhance Mindlin
  • Implement a collisionselftest for mindlin
  • Walls doc

Doxygen infinite loop script (now works in background): DoxyLoop.sh

Mitchel
  1. Coupling;
  2. look at open jobs at JIRA (started)
  • Tutorial on 2D and 3D parameter studies (JIRA MDPM - 307)
    • Code
    • Documentation
    • Validation
Hong
  • Make the structure of the coupled codes consistent
    • Clean up my branches to be merged with Mitchel's
    • CoupledProblem class and how to inherit
  • Volume coupling
    • Match the current velocity
    • (Augmented) Lagrange multiplier
    • Merge implicit and explicit branches into the main coupling branch
  • Surface coupling
    • Smooth particle-finite element contacts
    • Triangulated walls with prescribed velocity from Oomph-lib
    • Coarse-Grained coupling force
  • Marnix \mu(I) rheology
  • Works on a CG converter tool (e.g. stat2vtu.py). In Tools
Yousef
  • Clean up melting, add test cases (PS: My branch should not be merged back till the work is published).
  • Coding style and format
  • Error checks
  • Test Cases:
    • ....
  • Update ThermalParticleVtkWriter: to avoid dynamic_cast!
  • JIRA jobs:
  • I noticed some jobs are repeated: Do not delete, just link them (e.g. latex pb in documentations).
  • Test failing on cluster and Windows 10 [MDPM-311] : I ran fullTest on Win10 and cluster, all passed.
  • Add Doxygen section to the "Coding style of the kernel" on the website [MDPM-95]: The section already on the website, do we need to add it in the documentation?
  • make separate VTKWriter class for Wall/Interaction VTK writer [MDPM-240]: There is also done, right? -> a unit test is needed.
  • Adhesive particle contacts increases simulation time [MDPM-323]: which adhesive species? -> we need to contact Nikzad Falahati, Anthony knows him.
  • ..

 

Hao

Check the StressStrainControlBoundary:

  • Check the comment of each function implemented in this boundary and modify the documentation.
  • Add a detailed version of the tutorial of one deformation mode. (New page on Isotropic compression.)
  • Update the detailed pdf and it to the additional document.
  • Maybe add more videos and tutorials on different deformation modes:
    • PureShear,
    • SimpleShear, (Volume conserving, no stress control activated)
    • SimpleShear, (Changing volume, activate the stress control)
  • How to deal with the input data folder? Reference it? Or make them separate folders but then need to check all the other folders also.

Please don't copy input data, instead refer to the source directory using getMercurySourceDir in CMakeDefinitions.

Include the path in the cpp file and read the input data from there, so avoid copying.

 

Here is the issue of MPI without c++ binding issue. I think it is general in all the new version of openmpi, not only on mac.

can someone test this with newest openmpi on either windows or linux?

Error rendering macro 'jira' : Unable to locate Jira server for this macro. It may be due to Application Link configuration.

 

Added the logo with slogan to the NewDocumentation branch.

JuliusConverting demos to tutorial (Homogeneous and Inhomogeneous Cooling States, F-C-Demos)

 

Todos

  • Make a separate discussion on the coupling branch [Hong, Mitchel, Thomas, Anthony]; Tuesday  after 11am meeting