wind direction interpolation in IDEAS.BoundaryConditions.SimInfoManager

[flosfeld]

Weather file data has an hourly input. The SimInfoManager calculates the data in between 2 inputs. Wind direction can change from e.g. 10 to 360, resulting in an interpolation from 10 to 360. In fact, you would expect an interpolation between 10 and 0.

For a wall facing N, the CpAct in outsideAir should be around 0.4. However it is fluctuating, from a pos value to a neg value to a pos value again, in only 1 hour. The variable pWin follows this behavior.

This can be observed in e.g. IDEAS.Examples.Tutorial.Example1:

[jelgerjansen]

@kldjonge since this is more your cup of tea, what are your thoughts on this issue? ;)

The value of CpAct follows from the wind direction (OutsideAir model)

  parameter Real table[:,:]=[0,0.4; 45,0.1; 90,-0.3; 135,-0.35; 180,-0.2; 225,-0.35; 270,-0.3; 315,0.1; 360,0.4] "Cp at different angles of attack";

The wind direction interpolation is done in the ReaderTMY3 model, which should thus be checked with the Modelica Working Group. I'm afraid that implementing a solution for this is not straightforward as the "correct" interpolation would result in a discontinuous curve.

[kldjonge]

Wel spotted @flosfeld. Unfortunantly, as @jelgerjansen says, It seems to be the weatherfile-reader thing (CpAct seems to do what it should given the input it receives from sim.). But ok, let's think of a solution.

Any idea how other software handles this kind of interpolation issue?

First idea would be to generate a Vdir_sd (shotest distance), which somehow decides to either be Vdir or Vdir-360 (and can thus be a negative number) based on the previously known input and the current one. Not directly sure how to implement something like that in Modelica but I think (but should check) the OutsideAir model would still handle it correctly if it receives a negative input.

[jelgerjansen]

Will be solved via ibpsa/modelica-ibpsa#2068 and PR

[mwetter]

This is now corrected on the IBPSA master branch.