add a new geometry: fractal_headlands, inspired...
slightly changed the 'disc' configuration: prop...
- add a new domain in 'grid.py' : smoothsquare....
Merge branch 'master' of ssh://git.code.sf.net/...
- bug fix in param.py that was just commited (n...
- rewrite join_output to speed it up:
- fix the blacklist in do_output.py: add 'read'...
- add a read() function to param.py to read it ...
- rewrite the vortex impinging a wall into dipo...
- remove the last import that caused trouble on...
- adapt the restart to the new fluid2d object, ...
- add the momentum conservation (also in the no...
fix the momentum computation
fix bug in for_advection.f90 / now vorticity is...
added exp_freedecay to the repository
no the no_slip source term does conserves vorti...
improve a little bit the inversion driver: 2 ca...
euler+noslip (monoproc) conserves super well
* version in progress ... in which I never use ...
add a fill_halo after the call to advection
fix minor bug in timescheme.py (call to rhs_adv...
created: a new object (timescheme.py) doing the...
implemented: the three flux splitting schemes. ...
fix bug in for_advection.f90 : mx5,mx3 , my5 an...
creation of a Fortran routine to compute the ad...
previous commit seems to be uncomplete
add the new multigrid (entirely rewritten, no m...
Some minor improvements
Some modifications to find a bug in fill_halo b...
Cleaning and improving code
ommit to change branch to master
Minor modifications in exp_vortex_impining_a_wall
Trouvé bug avec nhalo > 2
EN COURS : travail sur le bug nhalo 3
Merge branch 'master' of ssh://git.code.sf.net/...
Finalisation du mode restart
add copy/pasting of global attributes in join_o...
Restart files = OK
Developpement du Restart
test a nx=4096 / ca tient avec up3
Objetif : enfouissement de MPI dans le code
* add a __main__ section in join_output:
* ajout d'un join_output_monoproc pour pouvoir ...
* cleaned the time module + printing on screen
* no_slip: go a step further, now the NoSlipLap...
* added a new operator in pde.py: NoSlipLaplac
* no slip is working with mpi
* add the function "generate_random_large_scale...
generation of a random large-scale field
YES!
* debugging the inversion with multiproc
modification de geometry.py :
creation de la classe Diag (diagnostics)
output :
Modification des output :
Merge branch 'master' of ssh://git.code.sf.net/...
creation de l'objet Time avec la methode 'refresh'
* euler and pure advection are now running on c...
fix the T2P interpolation matrix: P points on ...
mise en place du noslip / ok
Euler now works fine but only if nx==ny
multigrid en cours de branchement, ne marche pa...
gros check dans tous les sens pour avoir l'adve...
Introduce in fluid2d.py a global umax and mean
debug de fonction de courant
Added a fancy pdf creator to show mpi grid config!
tracer advection
Modification de pde.py
Committer: Clement VIC - Th GR/XC <cvic@aries....
- mpi + advection seems to work pretty well
put back the non corrupted versio of grid.py
* new pde = advection of tracer + its example s...
add:
preliminary work to introduce MPI:
Introduction of Fluid2D module ine the wave2D e...
A param.xsl was created to generate automaticly...
Cleaning Code
Cleaning code
Working on the performance of the code
Working on performance
In validation of vortex impining a wall
Vortex_impinging_a_wall (Euler) experiment work...
Coding vortex_impinging_a_wall (doesn't work yet)
Starting to code the vortex_impinging_a_wall ex...
Merge branch 'master' of ssh://git.code.sf.net/...
Merge param.xml (I hope that is correct !)
start the documentation in the param.xml
Parameters are stored in a XML file (param.xml)
Strating to code parameters in a data base in XML
Wave2D works for a rectangular domain (ex.: nx=...
Clean the src directory
The code was cleaned
19 May 2014 : Wave2D works in python
Add graphic window to visalize results
Wave2d is running but results are false :-(
wave2d version a faire fonctionner
add the dictionary class param for example
An exmaple of param class inherit from diction...
Merge branch 'master' of ssh://git.code.sf.net/...
work in progress