classVersion_1_2_0
Extends from TRANSFORM.Icons.ReleaseNotes (Icon for release notes in documentation).
Information
Conversion: none needed. Models written for TRANSFORM 1.1.0, 1.0 or 0.5 load without change, except in the rare case described under Added. A few changes can give different results in your own models; each is marked upgrade note in CHANGELOG.md and says how to get the old behaviour back where that makes sense.
OpenModelica: with this release many more of the library's examples run in OpenModelica and match Dymola. Most of the fixes are in the library's components, media and functions, so they also help your own models; some are in the examples' start values only. The examples that still do not run in OpenModelica are listed in CHANGELOG.md.
Results that change:
- Conduction walls (
TRANSFORM.HeatAndMassTransferconduction models and the walls built on them) now start atTs_start. Their geometry was computed in an algorithm section, which a tool cannot differentiate, so some walls started at the material's reference temperature (298.15 K for most solids) instead, in OpenModelica always and in Dymola in some models. Upgrade note inCHANGELOG.md. TRANSFORM.Nuclear.ReactorKinetics.Functions.Initial_powerBased_powerHistory: the precursor start values from a power history are now right when the history's power changes; the start valuesCs_0andEs_0are no longer ignored without decay heat.- Solid media: OFHC copper conductivity and resistivity follow their table instead of a poor polynomial fit;
Cr225Mo1conductivity is held above its table. Conductivity fits that reach zero below 0 K are no longer evaluated below 0 K, which removes a false steady state; no value at a real temperature changes. MolybdenumandTungsten: in Dymola, their enthalpy at a fixed temperature came out as 0; it is now right.- Examples with corrected starts or data:
Pressurizer_withWall_Test,NaturalCirculation,NaturalCirculation2andCIET_nureth(its expansion tank was far too large).DRACS,nParallel_StateBoundary,nParallel_FlowBoundaryandnParallel_FlowBoundary_MultiTransferSurfacealso have corrected starts, which leave their Dymola results unchanged; so doesPressurizer_Test, within its test tolerance. TRANSFORM.Fluid.Volumes.ExpansionTank_1Port: the port pressure's start value includesdheight. The tank documentation now says how to start a tank whose port joins a component that holds a pressure.
Added: outputs re and im of TRANSFORM.Math.ComplexMath.ccubicRoots (the roots as Real arrays). If your own function replaces ccubicRoots through a replaceable declaration constrained by it, it must now also declare these two outputs; one that extends ccubicRoots must not declare its own re or im. The library now declares SDF 0.4.5 in its uses annotation; only the critical-heat-flux and film-boiling look-up tables need it.
Compatibility, results unchanged: several components and functions are rewritten in standard Modelica that OpenModelica accepts (solid-media functions, trace separators, distributed heat transfer, linspace_2D functions, reading the power-history table), and two examples no longer use constructs that only Dymola accepts.
The full list is in CHANGELOG.md in the TRANSFORM-Library repository.