diff --git a/.github/workflows/checkCron.yml b/.github/workflows/checkCron.yml index eb5907eb0a7..d9b10dd08c6 100644 --- a/.github/workflows/checkCron.yml +++ b/.github/workflows/checkCron.yml @@ -25,7 +25,6 @@ jobs: run: python ./.CI/check_html.py checkLinks ./ spelling_checks: timeout-minutes: 5 - if: github.repository_owner == 'modelica' runs-on: ubuntu-latest steps: - uses: actions/checkout@v3 @@ -40,7 +39,7 @@ jobs: - name: Check for spelling errors run: | printf "[codespell]\nenable-colors=\n" >> .codespellrc - codespell -L additionals,inport ./Modelica/package.mo + codespell -L additionals,inport,manuel ./Modelica/package.mo codespell -L bu,nin ./Modelica/Clocked/RealSignals/Periodic/StateSpace.mo codespell -L mesures ./Modelica/Constants.mo codespell -L nin,inport,fo ./Modelica/StateGraph.mo @@ -48,7 +47,10 @@ jobs: codespell -L utput ./Modelica/Electrical/Digital.mo codespell -L ser ./Modelica/Electrical/Polyphase/Examples/PolyphaseRectifier.mo codespell -L ser ./Modelica/Electrical/Polyphase/Examples/Utilities/PolyphaseRectifierData.mo + codespell -L manuel ./Modelica/Fluid/package.mo codespell -L pres ./Modelica/Fluid/Dissipation.mo + codespell -L manuel ./Modelica/Fluid/Vessels.mo + codespell -L manuel ./Modelica/Fluid/Examples/AST_BatchPlant.mo codespell -L fo,alo,als,cacl,ba,nd,ges,caf,inh ./Modelica/Media/IdealGases/package.mo codespell -L fo,alo,als,cacl,ba,nd,alow,ges,caf,inh ./Modelica/Media/IdealGases/Common/SingleGasesData.mo codespell -L alow ./Modelica/Media/IdealGases/Common/Functions.mo @@ -70,7 +72,7 @@ jobs: codespell -L technik ./Modelica/Magnetic/FundamentalWave/UsersGuide/References.mo codespell -L supportt ./ModelicaTest/Rotational.mo codespell -L lamda,inport ./ModelicaTestConversion4.mo - codespell --skip ./ObsoleteModelica4.mo,./Modelica/package.mo,./Modelica/Clocked/RealSignals/Periodic/StateSpace.mo,./Modelica/Constants.mo,./Modelica/Blocks/Continuous.mo,./Modelica/StateGraph.mo,./Modelica/Electrical/Digital.mo,./Modelica/Electrical/Polyphase/Examples/PolyphaseRectifier.mo,./Modelica/Electrical/Polyphase/Examples/Utilities/PolyphaseRectifierData.mo,./Modelica/Fluid/Dissipation.mo,./Modelica/Resources,./Modelica/Media/IdealGases/package.mo,./Modelica/Media/IdealGases/Common/SingleGasesData.mo,./Modelica/Media/IdealGases/Common/Functions.mo,./Modelica/Media/IdealGases/Common/package.mo,./Modelica/Media/Air/ReferenceAir.mo,./Modelica/Media/R134a.mo,./Modelica/Media/Water/IF97_Utilities.mo,./Modelica/Electrical/Polyphase/UsersGuide/References.mo,./Modelica/Electrical/Spice3.mo,./Modelica/Electrical/Machines/BasicMachines/Components/PartialAirGapDC.mo,./Modelica/Electrical/Machines/Examples/ControlledDCDrives,./Modelica/Electrical/Analog/UsersGuide/References.mo,./Modelica/Math/package.mo,./Modelica/Mechanics/Rotational,./Modelica/Mechanics/Translational,./Modelica/Mechanics/MultiBody,./Modelica/Magnetic/FluxTubes/UsersGuide/Contact.mo,./Modelica/Magnetic/FluxTubes/Material/SoftMagnetic/NickelIron,./Modelica/Magnetic/FundamentalWave/UsersGuide/References.mo,./ModelicaTest/Rotational.mo,./ModelicaTestConversion4.mo,./.CI,./.github,./.git,*.png -L nin,inport,te,sinc,vie,commutating + codespell --skip ./ObsoleteModelica4.mo,./Modelica/package.mo,./Modelica/Clocked/RealSignals/Periodic/StateSpace.mo,./Modelica/Constants.mo,./Modelica/Blocks/Continuous.mo,./Modelica/StateGraph.mo,./Modelica/Electrical/Digital.mo,./Modelica/Electrical/Polyphase/Examples/PolyphaseRectifier.mo,./Modelica/Electrical/Polyphase/Examples/Utilities/PolyphaseRectifierData.mo,./Modelica/Fluid/package.mo,./Modelica/Fluid/Dissipation.mo,./Modelica/Fluid/Vessels.mo,./Modelica/Fluid/Examples/AST_BatchPlant.mo,./Modelica/Resources,./Modelica/Media/IdealGases/package.mo,./Modelica/Media/IdealGases/Common/SingleGasesData.mo,./Modelica/Media/IdealGases/Common/Functions.mo,./Modelica/Media/IdealGases/Common/package.mo,./Modelica/Media/Air/ReferenceAir.mo,./Modelica/Media/R134a.mo,./Modelica/Media/Water/IF97_Utilities.mo,./Modelica/Electrical/Polyphase/UsersGuide/References.mo,./Modelica/Electrical/Spice3.mo,./Modelica/Electrical/Machines/BasicMachines/Components/PartialAirGapDC.mo,./Modelica/Electrical/Machines/Examples/ControlledDCDrives,./Modelica/Electrical/Analog/UsersGuide/References.mo,./Modelica/Math/package.mo,./Modelica/Mechanics/Rotational,./Modelica/Mechanics/Translational,./Modelica/Mechanics/MultiBody,./Modelica/Magnetic/FluxTubes/UsersGuide/Contact.mo,./Modelica/Magnetic/FluxTubes/Material/SoftMagnetic/NickelIron,./Modelica/Magnetic/FundamentalWave/UsersGuide/References.mo,./ModelicaTest/Rotational.mo,./ModelicaTestConversion4.mo,./.CI,./.github,./.git,*.png -L nin,inport,te,sinc,vie,commutating clean_workflow_runs: timeout-minutes: 5 if: github.repository_owner == 'modelica' diff --git a/Modelica/Electrical/Analog/Examples/Lines/SmoothStep.mo b/Modelica/Electrical/Analog/Examples/Lines/SmoothStep.mo index 4a6d1b83460..dcdc993c415 100644 --- a/Modelica/Electrical/Analog/Examples/Lines/SmoothStep.mo +++ b/Modelica/Electrical/Analog/Examples/Lines/SmoothStep.mo @@ -14,7 +14,7 @@ model SmoothStep "Compares oLine and tLine behaviour" parameter Integer N50=50 "Number of lumped segments of oline50"; parameter SI.Velocity c=1/sqrt(l1*c1) "Speed of EM wave"; parameter SI.Time td=len/c "Transmission delay"; - parameter SI.Impedance z0=sqrt(l1/c1) "Charateristic impedance for very high frequency"; + parameter SI.Impedance z0=sqrt(l1/c1) "Characteristic impedance for very high frequency"; Modelica.Blocks.Sources.Step step(startTime=200e-6) annotation (Placement( transformation( extent={{-10,-10},{10,10}}, @@ -170,7 +170,7 @@ equation
If resistances and conductances are negligible, with OLine the results should become closer and closer as the number of segments increases.
Suggested tests:
Use the model as it is (it has firstOrder.T=1e-9
s) and comapare R1.v
, R5.v
, R50.v
, Rdistr.v
: Rdistr.v
is the ideal voltage of the receiving-end resistor for a lossless line, while the others are approximations with 1, 5, 50 cascaded segment models.
Use the model as it is (it has firstOrder.T=1e-9
s) and compare R1.v
, R5.v
, R50.v
, Rdistr.v
: Rdistr.v
is the ideal voltage of the receiving-end resistor for a lossless line, while the others are approximations with 1, 5, 50 cascaded segment models.
The step in this case is nearly ideal, and therefore is a very tough test for cascaded segments, and in fact they are not very good in approximating the real line behaviour.
Repeat the simulation using firstOrder.T=20e-6
s. You will see that the OLine compliance with the reference result given by TLine is much better, and 50 segments are probably adequate for several applications; in comparison with, they add the option of evaluating losses.
This model demonstrates the behaviour of a dimmer with phase-angle control with resistive-inductive load. -Note that due to the inductance the current is not zero at the points in time wehre zero-crossing of the voltage occurs, +Note that due to the inductance the current is not zero at the points in time where zero-crossing of the voltage occurs, and the triac stays conducting until the current becomes zero.