Change R-value of Insulation Layer for Construction By a Multiplier/Percent measures fail (and general improvements)

[MatthewSteen]

The Change R-value of Insulation Layer for Construction By a Multiplier and Change R-value of Insulation Layer for Construction By a Specified Percentage measures fail for interior constructions because the corresponding matched/reversed surface construction is not changed.

Failure

 ** Severe  ** GetSurfaceData: Construction FLOORPLNM-CONST of interzone surface FLOOR1-FLEXLAB-X3-PLNMA does not have the same materials in the reverse order as the construction CEILING-CONST of adjacent surface CEILING1-FLEXLAB-X3-ZONEA
   **   ~~~   ** or the properties of the reversed layers are not correct due to differing layer front and back side values
   **   ~~~   ** ...this problem for this pair will not be reported again.

General Improvements

The above failure was for a calibration use case for zone temperatures (occupied and plenum). Additionally, I think the following would be nice improvements.

1. The multiplier and percent arguments should allow for decreased performance, not just an increase. These are after all in the calibration gem not the EE gem
2. Use the arguments to simply set the existing object rather than cloning it, then changing it. Maybe there's a reason for cloning, but it just adds cruft to a model in my opinion.
   • Removing the cloned code will actually fix the failure too.

[brianlball]

you can use a negative % or a multiplier < 1 in some circumstances for decreased performance

[MatthewSteen]

you can use a negative % or a multiplier < 1 in some circumstances for decreased performance

Multipliers < 1 return false...

openstudio-calibration-gem/lib/measures/RValueOfInsulationForConstructionMultiplier/measure.rb

Lines 53 to 58 in 00a8cbf

	def check_multiplier(runner, multiplier)
	if multiplier < 0
	runner.registerError("Multiplier #{multiplier} cannot be negative.")
	false
	end
	end

[brianlball]

thats < 0, not < 1.

[MatthewSteen]

thats < 0, not < 1.

Oops, yes it is. I was looking at the r_value_prct_inc variable name.