dcpm model gives wrong results for 'wMechanical'

[christian.potthast@…]

Please find attached a test model, where the dcpm model (Electrical.Machines.BasicMachines.DCMachines.DC_PermanentMagnet) gives a wrong result for the angular velocity (dcpm.wMechanical).

(I assume that OpenModelica is wrong here, because Dymola and SimulationX produce the same result which significantly differs from OpenModelica.)

MSL version: 3.2.1
Dymola version: 2014 FD01
SimulationX version: 3.6.1.26028
OMEdit version: 1.9.1 (r18938)

Attached files:
testmotor_3.mo
Dymola_result.png
OM_result.png
OM_OutputMessages.png

[anonymous]

Testmodel for dcpm model

[anonymous]

result plot from Dymola

[anonymous]

result plot from OM(Edit)

[anonymous]

Output messages from OM(Edit)

[sjoelund.se]

This might be because of one of the many warnings triggered. First of all, many variables have start-value out of bounds.

<ScalarVariable name = "dcpm.coreParameters.wRef" useStart="true" start="0.0" min="1e-60" unit="rad/s" displayUnit="rad/s" />

There seems to be a front-end/back-end warning missing regarding this, and could certainly cause unexpected results...

[wbraun]

There are im Dymola events, which even not occur for us as zero-crossings.

[wbraun]

It's really strange Dymola has the following events:

Expression dcpm.la.i > dcpm.brush.brushParameters.ILinear became true ( (dcpm.la.i) - (dcpm.brush.brushParameters.ILinear) = 1.38045e-12 )
Iterating to find consistent restart conditions.
      during event at Time :  0.1019447655784616
Expression dcpm.la.i > dcpm.brush.brushParameters.ILinear became false ( (dcpm.la.i) - (dcpm.brush.brushParameters.ILinear) = -1.29292e-10 )
Iterating to find consistent restart conditions.

But this expressions are inside a function:

function Modelica.Electrical.Machines.Losses.DCMachines.brushVoltageDrop \"Voltage drop of carbon brushes\"
  input Modelica.Electrical.Machines.Losses.BrushParameters brushParameters \"Brush loss parameters\";
  input Real i(quantity = \"ElectricCurrent\", unit = \"A\") \"Actual current\";
  output Real v(quantity = \"ElectricPotential\", unit = \"V\") \"Voltage drop\";
algorithm
  if brushParameters.V <= 0.0 then
    v := 0.0;
  else
    v := if i > brushParameters.ILinear then brushParameters.V else if i < (-brushParameters.ILinear) then -brushParameters.V else brushParameters.V * i / brushParameters.ILinear;
  end if;
end Modelica.Electrical.Machines.Losses.DCMachines.brushVoltageDrop;

So they should never generate events!?

[sjoelund.se]

Functions may only generate events if they are inlined and annotation GenerateEvents=true. This function uses neither.

[wbraun]

Okay, we end-up in different flat equation. Dymola has:

  dcpm.brush.v := smooth(1, (if dcpm.la.i > dcpm.brush.brushParameters.ILinear
     then 0.7 else (if dcpm.la.i <  -dcpm.brush.brushParameters.ILinear then 
    -0.7 else 0.7*dcpm.la.i/dcpm.brush.brushParameters.ILinear)));

and we get:

dcpm.brush.v = 0.0;

That's also the reason why we haven't that events.

[sjoelund.se]

This ticket was not closed for 1.9.1, which has now been released. It was batch modified for milestone 1.9.2 (but maybe an empty milestone was more appropriate; feel free to change it).

[sjoelund.se]

Milestone changed to 1.9.3 since 1.9.2 was released.

[sjoelund.se]

Moved to new milestone 1.9.4

[sjoelund.se]

Milestone pushed to 1.9.5

[sjoelund.se]

Milestone renamed

[wbraun]

It seems that this has been fixed, perhaps by #3791.

[sjoelund.se]

Milestone deleted