Kwaliteitsbeoordeling van warmte, lucht en vochtmodellen (HAM) om het hygrothermisch gedrag van bouwcomponenten te voorspellen

Accurately predicting the hygrothermal behavior of building components remains a major challenge in building physics. This difficulty stems largely from how simulation tools represent material properties, boundary conditions and related factors. The PhD thesis begins by revisiting the fundamental equations governing heat, air and moisture (HAM) transport, then evaluates the limitations of existing model assessment studies in their scopes, methodologies and benchmarks. To address these issues, the study introduces a stepwise and comprehensive methodology for evaluating the quality of HAM models, particularly WUFI and DELPHIN. Uncertainties related to material properties are examined through a benchmark dataset derived from controlled in-lab experiments. This dataset involves detailed measurements of hygrothermal material properties, boundary conditions, surface transport coefficients and hygrothermal responses. It supports a round robin validation campaign, revealing that variations in the implemented hygric properties can result in significant discrepancies in simulated outcomes. Supplementary comparative simulations further quantify these implementation differences and examine their implications under diverse hygrothermal transport dynamics. The study then shifts focus to boundary conditions, explored through in-situ configurations using real-world climate data. It compares semi-empirical methods for calculating wind-driven rain (WDR) as implemented in WUFI and DELPHIN, uncovering how specific assumptions and simplifications can systematically over- or underestimate WDR loads, resulting in substantial discrepancies in the predicted hygrothermal responses. By integrating theoretical analysis with empirical evidence, the research underscores the risks of relying on default parameters and oversimplified representations. It calls for closer alignment between simulation models and observed reality, concluding with recommendations for enhancing the reliability and robustness of HAM modeling practices.