zhyang-dev/openfoam
1
0
Fork 0
mirror of https://develop.openfoam.com/Development/openfoam.git synced 2025-11-28 03:28:01 +00:00
Code Issues Actions Packages Projects Releases Wiki Activity

ENH: ReleaseNotes : update

Browse Source
This commit is contained in:
mattijs
2010-11-22 10:47:25 +00:00
parent a4d0118859
commit 455b88683b
1 changed files with 17 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
ReleaseNotes-dev
View File

@ -35,6 +35,9 @@
*** Core library *** Core library
+ Large number of code refinements and consistency improvements to support + Large number of code refinements and consistency improvements to support
other developments. other developments.
Parallel : using non-blocking communications wherever possible. This will
lead to lower requirements for MPI_BUFFER_SIZE and possibly better
start-up performance on larger numbers of processors.
*** Turbulence modelling *** Turbulence modelling
*** *Updated* thermophysical libraries *** *Updated* thermophysical libraries
+ Now use the more descriptive dictionary input format, e.g. for a single + Now use the more descriptive dictionary input format, e.g. for a single
@ -102,15 +105,28 @@
+ Parallel aware + Parallel aware
*** *New* ptscotch decomposition method *** *New* ptscotch decomposition method
*** *Updated* particle tracking algorithm *** *Updated* particle tracking algorithm
+ uses non-blocking parallel transfers
+ does 'minimum-tet' decomposition of face to work with warped faces (snappyHexMesh!)
+ reuses tracking state in interpolation - improves consistency and speed
*** *Updated* split cyclics into two separate patches. See doc/changed/splitCyclics.txt *** *Updated* split cyclics into two separate patches. See doc/changed/splitCyclics.txt
* *Updated* interpolation (volPointInterpolation) now works without the * *Updated* interpolation (volPointInterpolation) now works without the
globalPointPatch. Moving mesh cases can now be run non-parallel and globalPointPatch. Moving mesh cases can now be run non-parallel and
continued in parallel and reconstructed without any limitation. continued in parallel and reconstructed without any limitation.
*** *New* compact binary I/O for faces and cells. This speeds up reading/writing meshes in binary. *** *New* compact binary I/O for faces and cells. This speeds up reading/writing meshes in binary.
*** *Updated* runTimeModifiable *** *Updated* runTimeModifiable
+ on linux uses inotify instead of time stamps - more efficient for large + user selectable model to use for checking run-time modifiable
dictionaries:
+ timeStamp : use time stamp as before
+ inotify : use inotify instead of time stamp
+ timeStampMaster : only master checks and reads file. Slaves get file
contents from master.
+ inotifyMaster : only master checks and reads file. Slaves get file
contents from master.
+ on linux optionally uses inotify instead of time stamps - more efficient for large
numbers of monitored files. No more fileModificationSkew needed. numbers of monitored files. No more fileModificationSkew needed.
+ single integer reduction instead of one reduction per monitored file. + single integer reduction instead of one reduction per monitored file.
+ only files that can be re-read are being checked. Drastic reduction of
number of files to check.
* Solvers * Solvers
A number of new solvers have been developed for a range of engineering A number of new solvers have been developed for a range of engineering
applications. There has been a set of improvements to certain classes of applications. There has been a set of improvements to certain classes of