CI Tests fail for some models

[DaJansenGit]

Some tests currently fail after the newest IBPSA merge, like here due to divation from reference results.

First @MichaMans and me thought this was due to the update from reference results in the IBPSA to dymola 2020

But even after using the dymola 2020 container the deviation changes but is still there, this can be seen here

I simulated one of the failing thermalzone examples locally, both with dymola 2020 and the deviation is still there, I used the same solver with same tolerances. The screenshot shows the timestamp with the highest deviation recording to the regression test.

@MichaMans @PMehrfeld do you have any idea where this deviation comes from?

[MichaMans]

@DaJansenGit I have no idea where the deviations come from. The marked area/time at comparison is the time with the highest deviation? If so, we should maybe check:

• if we are still in the VDI 6007 deviation range
• why the check does not fail in the ibpsa --> open a discussion branch there?
• check again if the reference results from the ibpsa and our are 100% the same

[DaJansenGit]

I updated the screenshot, it shows now the timestamp of the highest deviation.

• I will check the VDI6007 range
• I will open an discussion branch in IBPSA
• the reference results are 100 % the same, I checked this already

[mlauster]

I can't check your example results, since they are not accessible for external, but the VDI range is 0.1 K for hourly means.

[DaJansenGit]

We solved this issue with:

• a new merge of the current master
• the usage of dymola 2020 container
• the creation of a new CI runner server with better performance

Thanks for you effort.