Refactor damper models

[AntoineGautier]

This issue is to integrate into modelica-ibpsa the developments related to issue #1298 in modelica-buildings.
See the presentation at Aachen and the updated version describing the latest implementation.

[AntoineGautier]

(I include here some comments made in the PR.)

To clarify the motivation for renaming the parameters for nominal pressure drop:

In the master branch:

• Exponential: dp_nominal is the pressure drop between the 2 fluid ports. This is consistent with the super class PartialTwoPort which defines dp = port_a.p - port_b.p. This super class is inherited from by all damper models.

• VAVBoxExponential: dp_nominal is either the pressure drop of the fixed resistance only (if not dp_nominalIncludesDamper) or the total pressure drop between the 2 fluid ports (if dp_nominalIncludesDamper).
  This class extends PartialDamperExponential which uses dp_nominal as the total pressure drop between the 2 fluid ports in the call to homotopy(..., simplified=m_flow_nominal_pos*dp/dp_nominal_pos).

• PressureIndependent: dp_nominal is the pressure drop of the damper only (which is the exact opposite of VAVBoxExponential with dp_nominalIncludesDamper=false), dpFixed_nominal is the pressure drop of the fixed resistance.

The modifications we discussed lead to:

• PartialDamperExponential

  • New parameter dpDamper_nominal with default value 10 Pa. EDIT Better set no default value to be consistent with current implementation?
  • New parameter dpFixed_nominal with default value 0 Pa.
  • v_nominal with default value 1 m/s changed into final v_nominal = (2 / rho_default / k1 * dpDamper_nominal)^0.5.
  • dp_nominal changed into final dp_nominal = dpDamper_nominal + dpFixed_nominal.
  • kFixed changed into final kFixed = if dpFixed_nominal > Modelica.Constants.eps then m_flow_nominal / sqrt(dpFixed_nominal) else Modelica.Constants.inf .

  Note that this will make PartialDamperExponential similar to PartialTwoWayValve which contains the declaration of dpValve_nominal and dpFixed_nominal.

• Exponential

  • No changes in the exposed parameters. The model in master uses protected parameters as intermediary steps to compute dp_nominal: those are no longer needed with the modifications made in PartialDamperExponential.

• VAVBoxExponential changed into Exponential

  • dp_nominalIncludesDamper is removed.

• PressureIndependent

  • Now extends directly Exponential.

  Remaining point: we could include the leakage coefficient declaration in Exponential.

• MixingBox

  • To be modified in consistency with the refactoring of PartialDamperExponential.

@mwetter @Mathadon please review the proposed changes. Do you need additional elements before our next iteration on Thursday?

[mwetter]

@AntoineGautier : These changes look good to me.

[Mathadon]

@AntoineGautier thanks for the summary, looks good. I agree with no nominal for dpDamper_nominal.

[AntoineGautier]

• PartialDamperExponential can be retired if strictly identical to Exponential.
• Adopt the leakage coefficient parameter for all damper models.
• Check the parameters description of k0 and k1: those are loss coefficients, not flow coefficients.

Remark: For future reference I also include here the elements regarding the nominal pressure drop as no default will be provided.

10 to 40 Pa seems the typical range for balancing ducted dampers see p. 6:
http://www.swegon.lt/wp-content/uploads/2012/03/REACT.pdf
ASHRAE Guideline 16 INFORMATIVE ANNEX C shows pressure drop around 10 Pa for relief, outdoor and return air damper. The guideline also mentions typical range of face velocity for relief (5 to 8 m/s), outdoor (2 to 8 m/s) and return air damper (4 to 8 m/s).
Adopting 10 Pa as default in the library leads to a nominal face velocity of 6.1 m/s which seems reasonable (note that 1 m/s is the current default value). Also with a VAV box of typically 100-150 Pa total pressure drop, the damper's authority would be 7-10% which falls into the optimal range.

Remark: ASHRAE Dampers and Air Flow Control p. 44 shows that the application has a major effect on the pressure drop: depending on the configuration the same damper with the same face velocity can have a total pressure drop ranging from 6.5 Pa to 42.5 Pa.

[AntoineGautier]

Closed by #1267.
The feature enhancement related to implementing an interpolation between a fully closed and a fully open characteristic shall be addressed in a separate issue.