In this file noteworthy changes of new releases of pyLife are documented since 2.0.0.
- Fix crash when HCM algorithm was given a named multiindex load series (#210)
- Drop support for python-3.8 ast it is end of life
- The Wöhler analysis slightly changed (#138)
Support for some FKM-non-linear calculations
- HCM rainflow algorithm
- P_RAM and P_RAJ damage calculation
- Notch approximation law (Neuber and Seeger-Beste)
- Connectivity information is now transferred from odbserver to odbclient as ints.
- Fix unnecessary and harmful copy of pandas object in Wöhler classes (#146)
- Don't assume
failure_probability=0.5in Wöhler calculations (#192)
Fix gradient calculation for non continous nodes in mesh.
-
Massive performance improvement when reading element connectivity by
odbclient.Note that alsoodbserverneeds to be updated in its environment. (#121) -
Fix some warnings on import
- New method
LoadCollective.histogram()(#107)
- Sanitize checks for Wöhler analysis (#108)
- Error messages when odbclient gets unsupported element types (#64)
- Improved documentation
MeshSignalnow allows for additional indeces (#111)
-
Change fracture load levels used for slope
k_1and scatterT_Nestimation. Now only fractures in thefinite_zoneare used for estimation and not all fractures (infinite_zoneandinfinite_zone). Change based on findings in DIN50100:2022-12. (#80, #101)(see this discussion)
-
Rename
FatigueData.fatigue_limittofinite_infinite_transition -
The Bayesian Wöhler analyzer has been shutdown (#74) (see this discussion)
- Add option to manually set
fatigue_limit(now renamed) for Woehler curve estimation. (#73)
- Rainflow counters work with
pd.Seriesof all index type (#69) - Improved documentation
- Fixed confusing load matrix after mean stress transformation (#105)
- Massive performance improvements of three point and four point rainflow counters.
- The Wöhler analyzer now ignores irrelevant pure runout levels (#100)
- Support numpy>=2.0.0
-
History output for
odbclient -
Introduce
WoehlerCurve.miner_original()
-
Non-destructive miner modifiers of
WoehlerCurveThe methods
WoehlerCurve.miner_elementary()andWoehlerCurve.miner_haibach()now return modified copies of the original WoehlerCurve object, rather than modifying the original.
Minor bugfix release
- switch to vtk > 9
- fix two index bugs in gradient.py and hotspot.py
- support for python 3.11
- pandas-2.0 compliant
A minor release, mostly dependency related updates and documentation improvements.
Minor bugfix release
- Fix bug for detection of Abaqus binaries in odbserver
A minor release, mostly dependency related updates and documentation improvements.
-
The
LoadCollectiveaccessor class honors thecyclescolumn in the giveDataFrame. This is useful for manually created load collectives. -
Support for python 3.10
-
Fix off-by-one error in meanstress transformation rebinning
-
Correct the index of the detected turns when NaNs have been dropped in rainflow counting
-
Documentation improved
-
Switch from pymc3 to pymc version 4 for Bayesian Wöhler analysis. The extension is not installable via pip as the current bambi release pulls a numpy version that is incompatible with our demandas. If you need it, please install bambi manually from their current git repo and then pymc via pip.
-
Lift numpy version restrictions