bug in radiationconverter?

[Mathadon]

IDEAS.BoundaryConditions.SolarIrradiation.RadiationConvertor
The computation of HDirNor seems wrong.

[jelgerjansen]

@Mathadon can you be more specific here? Which part of the code seems to be wrong and why?

[Mathadon]

  Modelica.Blocks.Sources.RealExpression HDirMax(y=
        if zenAng.zen > Modelica.Constants.pi/2 then 0 else
        IDEAS.Utilities.Math.Functions.spliceFunction(
          x=min(incSouth.incAng, incEast.incAng) - incWest.incAng,
          pos=HDirWest.y,
          neg=IDEAS.Utilities.Math.Functions.spliceFunction(
            x=incEast.incAng - incSouth.incAng + 0.25,
            pos=HDirSouth.y,
            neg=HDirEast.y,
            deltax=0.3),
          deltax=0.3))

this is indeed the direct solar irradiation but not in the normal direction.

So it requires a division by a cosine of some angle.

[jelgerjansen]

The component is only used within IDEAS.Examples.TwinHouses.BaseClasses.TwinHousesSimInManager so I would not put this one very high on our priority list. Let's at least wait for the UnitTests to be up and running (issue #1316) before solving this issue.

When doing the update, the following changes should also be considered:

• Update the documentation, mentioning that this only holds for sensors measuring the total solar irradiation.
• Apply the same changes to IDEAS.BoundaryConditions.SolarIrradiation.RadiationConvertorSimplified
• Update the documentation of the latter model (it refers to IDEAS.Examples.TwinHouses.BaseClasses.RadiationConvertorPerez which does not exist any more. @Mathadon I think the latter was replaced by the RadiationConvertor model mentioned at the start of this issue?)

[Mathadon]

The simplified model does not generate an algebraic loop if I recall correctly. And consequently it cannot lead to a set of equations for which no solution exists. Which the other model sometimes had problems with. Perhaps that's even caused by this bug. That's all I remember 😅