A comprehensive benchmark dataset for the validation of building component heat, air, and moisture (HAM) models

Numerical heat, air and moisture (HAM) modeling allows predicting hygrothermal responses of building components with higher efficiency and less effort than laboratory experiments and field measurements. However, inaccuracy and/or incorrectness may appear in the predictions for the same case through different HAM models, primarily due to limitations or deviations in the description of physical phenomena and/or the implementation of mathematical algorithms. User preferences, biases, and/or mistakes with respect to implementing material properties, boundary conditions and other factors may also yield disparity. While a correct implementation of the numerical models is typically verified by the developers, the validity of the HAM models may remain questionable without the confrontation with experimental datasets. However, well-determined criteria and well-documented datasets for establishing the correct prediction of the transient hygrothermal responses of building components by HAM models remain very scarce. To address this issue, a dedicated benchmark experiment was conducted in the hot box-cold box (HB-CB) setup at KU Leuven, Belgium, on four wall assemblies composed of calcium silicate board, mineral wool, wood fiber board, and vapour barrier in different orders. Temperature, relative humidity, heat fluxes and moisture masses, as hygrothermal responses, were monitored under quasi-steady state boundary conditions. Full-scale characterization of the materials from the same batch was performed, along with a determination of the surface transport coefficients within the HB-CB setup. This comprehensive dataset allows a proper model validation by incorporating experimental datasets of material properties and surface transport coefficients and by confronting simulated hygrothermal responses with experimental evidence. In addition, sensitivity analysis can be performed to obtain insights into the impact of uncertainties in characterizing material properties on hygrothermal simulation predictions.