Merge IBPSA into IDEAS

[lucasverleyen]

I have merged the latest IPBSA master branch of today (ibpsa/modelica-ibpsa@8ed71ca) into this branch.

[lucasverleyen]

I'll list here some issues that popped up while updating the reference results

• IDEAS-only models included in the unit tests: I have changed the use_inputFilter into use_strokeTime or use_riseTime in pump and valve models, according to the new IBPSA implementation.
• IDEAS.Airflow.AHU: This is an IDEAS-only package. Some parameters in the pump data record in IBPSA have been changed. I have updated the Adsolair pump performance record accordingly.
• IDEAS.Fluid.Actuators.Valves.TwoWayTRV: is not in IBPSA. Hence, I have manually changed use_inputFilter to use_strokeTime, according to the new IBPSA implementation.
• IDEAS.Fluid.Actuators.Valves.Simplified.Examples: These models used a pump with a dp_nominal = 0. However, with the new IBPSA version, this gives a division by 0 error. Hence, I have changed this parameter to dp_nominal = 1 Pa.
• IDEAS.Fluid.HeatExchangers.Examples.EmbeddedPipeDp: This model used a pump with a dp_nominal = 0. However, with the new IBPSA version, this gives a division by 0 error. Hence, I have changed this parameter to dp_nominal = 1 bar.
• IDEAS.Utilities: 3 examples were giving an error because a .txt file was not found. I noticed that those files were removed after merging the IBPSA updates in IDEAS. Hence, I have checked all the removed files. I have restored a few .txt files, and all other removed files are intentionally removed, according to the changes in the IBPSA library.
• IDEAS.Examples.PPD12: The reference results have been updated, because the dynamics of the valves have been changed from a 2nd order to a linear transition. When the use_strokeTime is set to false, the model gives identical results. Also, the pump data FanPPD12.mo has been updated according to the new record parameters in IBPSA.
• IDEAS.Examples.Tutorial: Reference results of a few examples have been updated, as some of the models use valves and pumps with their transient behaviour which has changed in IBPSA.
• RunUnitTests.py: This is a python script in the main folder of the IDEAS repository. This file has been removed by the merge script, but it seems to me that this file is outdated and not used anymore. @jelgerjansen, can you please check this?
• Reference results: There are new reference results, due to 1/ new reference results in IBPSA, 2/ updates in IBPSA models (mentioned above) 3/ some numerical deviations in a few cases (e.g. IDEAS.Examples.TwinHouses, there was a difference smaller than 0.02 on a thermal power of 5 kW...)

[lucasverleyen]

@jelgerjansen, a few hints as to why we have some very small deviations in the new reference results:

• Diffuse radiation error when simulating over 1 year ibpsa/modelica-ibpsa#1716: this has led to minor changes in the reference results in the Buildings library (https://github.com/lbl-srg/modelica-buildings/pull/3374/files). We also use these models in our SimInfoManager and this might explain the small
• there have been made changes in the PartialMixingVolume, but should not give deviations (Volume model not working with MSL mixture gasses ibpsa/modelica-ibpsa#1650)
• MassFlowSource_WeatherData has also been modified, but is not used as far as I can see...

[jelgerjansen]

@lucasverleyen I would think that the 'Diffuse radiation error' only affects models with a simulation end time surpassing one year, which is not the case for the unit test models, right?

[lucasverleyen]

@jelgerjansen that sounds indeed logical. However, I found reference results that were changed slightly, even if they took less than 1 year. e.g. https://github.com/lbl-srg/modelica-buildings/blob/master/Buildings/Resources/Scripts/Dymola/Controls/Predictors/Examples/BESTEST.mos. with reference results in pull request https://github.com/lbl-srg/modelica-buildings/pull/3374/files

[jelgerjansen]

The small changes in reference results for IDEAS models might be due to several reasons:

• Used version of BuildingsPy and its dependencies (e.g. pyfunnel)
• This IBPSA PR might have a minor effect on the event handling for models using the weather data reader

[jelgerjansen]

I added a couple of missing unit tests, more specifically:

• IDEAS.Electrical
• IDEAS.Fluid.FMI
• IDEAS.Fluid.Humidifiers
• IDEAS.Fluid.Production
• IDEAS.Fluid.SolarCollectors

Some of these unit test models show changes in translation statistics and minor differences in reference results which is due to the used Dymola version (currently 2022x for IDEAS, but 2024x for IBPSA).