lp:~jdu/fluidity/reducedmodel
- Get this branch:
- bzr branch lp:~jdu/fluidity/reducedmodel
Branch merges
Recent revisions
- 3575. By Lawrence Mitchell
-
Fix help output for --log option
Turning on logging doesn't set the verbosity level to 2, so don't
claim that it does. - 3574. By Stephan Kramer
-
Bug-fix for https:/
/bugs.launchpad .net/fluidity/ +bug/842975, for flredecomp on flmls with extrusion. This will be tested in periodic_
2plus1_ adaptivity_ parallel (once it works :-)). - 3572. By Jon Hill
-
New option to allow the ocean forcing routines to use instantaneous values. This makes it a lot easier to construct our own ERA40-like netcdf files without having to worry about accumulating values.
- 3571. By Adam Candy
-
Correction to version number update on initial and *subsequent* builds - merge request approved.
This forces the version number to be updated on every make. Previously, version.h was only updated (created) in a fresh checkout and the version number was not updated on subsequent make calls.
Further comments to note (from https:/
/code.launchpad .net/~asc/ fluidity/ fixup-version- number2/ +merge/ 71887):
Note this is fine if the new make call is made in the root Fluidity folder. This change should be propagated to the following which also depend on version.h:
femtools/Makefile. dependencies
qg_strat/Makefile. in
main/Makefile.inUsing 'cd ..; make update_
version_ number' is a rather ugly way to achieve this - could possibly use a 'make -C ../ update_ version_ number' here? The important change is the one contained in this merge, and if make is called in the root folder first (as it is in most cases) this will achieve an update of the version number as desired.
- 3570. By Adam Candy
-
Conditions on running with mg relaxed. It is now possible to run without vertical_lumping and in this case a warning is given suggesting it is used.
- 3569. By Patrick Farrell
-
Clean up the configure.in; a debugging print statement
got left behind somehow. - 3568. By Christian Jacobs
-
Merging ctjacobs-multiphase branch revisions (from r3452 to r3532 inclusive) into trunk. Branch queue passes all unit, short and medium tests on buildbot.
Summary of changes:
===================- Implemented the fluid-particle drag term in Multiphase.F90. The mphase_stokes_law and mphase_
tephra_ settling test cases have been updated to use this new implementation, rather than a Python diagnostic Source field. An MMS test which includes this fluid-particle drag term is available in longtests/ mphase_ mms_p1dgp2_ fpdrag, and converges at the expected order. - Implemented a new boundary condition type, called 'flux', for control volume discretisations. Updated the mphase_
tephra_ settling test case to use this, which allows to flux in through the boundary at the correct rate. Also added documentation regarding this boundary condition type to the manual. - When checking the tfield_bc_type in Field_Equations
_CV.F90, we now compare against the more meaningful integer parameters BC_TYPE_ WEAKDIRICHLET (=1), BC_TYPE_NEUMANN (=2), etc. - Added the 'tephra_settling' example to the examples directory, along with some documentation (background, simulation setup, results, pretty pictures) in manual/
examples. tex. - In manual/
model_equations .tex and manual/ configuring_ fluidity. tex: Added documentation on the multi-phase flow model equations, setting up multi-phase flow simulations, available inter-phase interaction terms, and current limitations of the model. - Corrected typos in various places in the manual.
- Made the mphase_
tephra_ settling test more realistic by adding in random perturbations to the PhaseVolumeFraction influx. Also made a few minor tweaks (e.g. finish_time, timestep), and added pass_tests. - Added a three-phase test called mphase_
three_layers. This checks that our model can simulate more than one dispersed phase, allowing us to use several different particle sizes in our tephra settling simulation later on. - Moved the mphase_
tephra_ settling_ 3d test case to the longtests directory. - 3567. By Patrick Farrell
-
I think this should fix Tim's BAS unit test fail.
The line was padded with whitespace at the end to position 157,
running over the 132 character limit, and so the compiler didn't
see the end of the line ... - 3566. By Patrick Farrell
-
Possible fix for lp:832197. It seems that the old version of scipy
installed on cx1 doesn't take in the 'shape' argument for scipy.sparse.csr_matrix. __init_ _.
Fail gracefully in that case, rather than printing an exception which may
confuse users.
Branch metadata
- Branch format:
- Branch format 7
- Repository format:
- Bazaar repository format 2a (needs bzr 1.16 or later)
- Stacked on:
- lp:fluidity