Thermo Special Interest Group met on February 14, 2023 from 9:30 am till 11:30 am.
- Tests on implementations of ICapeThermoCompounds and ICapeThermoPropertyRoutine
AmsterCHEM (represented by Jasper van BATEN), BASF (represented by Sergej BLAGOV), KBC (represented by Richard SZCZEPANSKI), University of Cape Town (represented by Klaus MÖLLER), Michel PONS (Consultant as Chief Technology Officer).
ICapeThermoCompounds: Thermo SIG previously validated the strategy to perform almost any test both with and without a Material Context set on the Property Package, since the Material Context set on the Property Package should have no impact on the outcome of the call.
Today, Thermo SIG developed a strategy to test that a Property Package, when any of the methods of ICapeThermoCompounds is called, returns information pertaining solely to the Property Package, and not to the Material Object defined as Material Context. This new strategy calls for additional information in the configuration of a Test, such as stating which subset of the Compounds supported by the Property Package can be considered on the Material Object, while still making the Material Object suitable for use by the Property Package.
ICapeThermoPropertyRoutine: Thermo SIG defined general guidelines for tests to be carried on implementations of ICapeThermoPropertyRoutine. Among these guidelines, there is the need to verify that the order of Compounds defined on the Material Object has no impact on the values returned by the Property Package where Compounds are organized in a different order. Adequate mapping of one list onto the other must be tested.
This guideline raised the issue of how to compare values returned by different calls on the same method. Compliance tests are never testing the engineering relevance of a particular value, but may need to verify that two values are the “same”. Indeed, calculation of values may rely on a resolution path that may slightly differ between calls, because of a different sequence of calls, of a different arrangement (compound order for instance) of the conditions at which the calculation is requested. Hence small differences in values, supposed to be strictly equal, are possible. Equality condition should not be strict but with some tolerance that needs to be defined. This point will be addressed further.
The next meeting of the Thermo SIG is scheduled for February 21, 2023 at 9:30 am.
Any CO-LaN Member interested in the Thermo SIG activities is welcome to join this Special Interest Group. Contact the co-leaders of the SIG for further information: Sergej BLAGOV at BASF () and Jasper van BATEN at AmsterCHEM (). The Thermo SIG is looking for additional parties, well versed into any aspect of thermodynamics applied to process simulation and willing to contribute to the maintenance and development of CAPE-OPEN interface specifications related to thermodynamical aspects.