MercuryDPM May 2017

Tasks for MercuryDPM:

WhatWhoPriorityComments
Profile some example codes and optimise MercuryDPM accordinglyIranaLow

Marnix has some preliminary result

 

Change output file-directory structure to a nicer one2 peopleMedium

Binary output, in directory structure, with descriptor, binary-to-?? reader, logger tool

see MDPM-118 MDPM-137 MDPM-122 MDPM-121

 

Resurrect MercuryED   
Check all comments in Alpha if they are complete and correctBertHighAlso code that should be merged from Trunk to Alpha
Make a new alpha and betaAnthonyHigh 
Check all \todo and \bugAnthony or Thomas?High 
Write tutorial for coarse graining and add to documentationThomas & Deepak?  

MDPM-177 - Neighbourhood detection of walls 

MDPM-176 TriangulatedWall

Thomas and LucaHigh 
MDPM-109 Add more tutorials

Jonny/Deepak

High 

MDPM-123 size of BaseParticle/BaseInteractable/BaseObject

ThomasLow 

MDPM-185 Make a restartable structure for Lambda functions

AnthonyLow

 

Change license from 2013-2014 to 2013-2017BertLow

JMFT: Done using

s/Copyright (c) \(20..\)-20../Copyright (c) \1-2017/g
s/Copyright (c) \(20..\) /Copyright (c) \1-2017/g

Have included script to update this for future years. 

Rethink inheritance structure around Files, FilesAndRunNumber, DPMBase.

Also think about what should be possible, and what not.

ThomasMedium 
Consider replacing all Vec3D and Matrix3D by SmallVector<3> and SmallMatrix<3,3> in the kernelThomas and IranaMedium

Or even consider a typedef for Vec3D and Matrix3D and use the new, templated classes everywhere

TW: Can this also do the Matrix3DSymmetric?

IFCD: Not yet, and thinking about it, this might be a bad idea anyway.
Currently, all the entries in Vec3D and Matrix3D are public;
we cannot change the data-names without breaking everyone's code.
(This is one of the main reasons why public variables are evil.)

Rewrite RoughBottom s.t. it can be used in other applications than the chuteAnthony + ThomasLow 
Add a species that keeps restitution coefficient and bulk/Young's modulus constant instead of stiffness and dampingAnthonyMediumAlso needed for non-spherical particles, might be implemented before May.
Finish implementation of non-spherical particles and merge to trunkIrana & DeepakHigh 
Consider to template Mercury on particle-type: spherical, non-sphericalAnthonyLow 

Parallel code

Needs list of global functions + documentating

Marnix + other helping.High 
Check particle for interaction needs to be an order n algorithmMarnixMedium 
Full distance travelled displacementDeepak + talk to KasperLow 
Write coupling interface for oomph-libAnthony + MarnixMedium 
Review Kits commentsJonny  
Git-SVN bridgeDeepak, AnthonyHigh 
Mergy Jonny's add-one in betaAnthony, Thomas, Jonny  

Applications based paper

Kasper + Marnix

Luca

Kit

Deepak

Irana

Thomas

Ant

Jonny

 

Thomas  
WebsiteAnthonyHigh