Default units for measurements: missing definitions for IfcThermalTransmittance and others

[MatthiasWeise]

Check of values such as "[20<=Value<=100]" requires to know the unit for a given measurement (meter for length, kg for mass etc.). My understanding is that this unit is documented here: https://github.com/buildingSMART/IDS/blob/master/Documentation/units.md Some measurements such as IfcThermalTransmittance are however missing. Is it fair to assume that SI-base units are always used?

[CBenghi]

It's fair, but if you can add the row to the document, there's an automation process that reuses it for the audit tool. PRs are welcome for this kind of issue.

[MatthiasWeise]

We are running into a similar situation and struggle to use this table in our tool. As the datatype is something like IfcLengthMeasurewe tried to show the unit from the "Unit Symbol" column. Beside missing measurments like IfThermalTransmittanceMeasurethere are also measurements like IfcPositiveLengthMeasure where underlying unit is part of the table but there we need to manage such Measurement - Unit relationship in a separate table.

When this table was discussed the idea was (if I remember well) to focus on most frequently used measurements. Meanwhile I doubt that this is really needed (@Pasi and @ Jiri Hietanen: please double check from your side) . Some questions to the group before going ahead:

• Do we still want to limit the scope of supported measurements/units in IDS or should we allow all from IFC? BTW: Current documentation is not really clear about that.
• If we still want to limit the scope, my assumption is that this is for value comparison only (not for stating existence requirements).

I am in favor of supporting all measurements and propose to use the table from the IFC documentation (see this CSV: UnitMeasurement.csv). It does not include the dimensional exponents, but I wonder if we need that information here. One difference I noticed is for Ration measure, which in our table is in [%].

If we want to keep the limitation, I would at least propose to extend IfcThermalTransmittanceMeasure:

| IfcThermalTransmittanceMeasure | Rate at which energy is transmitted through a body | | W / m2 K | (0, 1, -3, 0, 1, 0, 0) | IfcDerivedUnitEnum.THERMALTRANSMITTANCEUNIT |

[CBenghi]

Hello @MatthiasWeise, I've taken a list of measures from the schemas and updated the markdown file of units. Have a look.