Incorporating hysteresis in hygrothermal modeling of cross-laminated timber walls: Model development, validation and moisture risk analysis

While interest in the hygrothermal performance of mass timber construction continues to grow, a critical gap remains in accounting for moisture hysteresis, particularly under conditions of rainwater leakage. Existing hygrothermal simulation tools such as WUFI and DELPHIN often neglect hysteresis effects, potentially underestimating moisture risks in building envelopes. This study addresses this limitation by implementing a one-dimensional coupled heat and moisture transfer model in Python that incorporates moisture hysteresis. The model was first validated using an experimental dataset from a date palm concrete wall reported in the literature, achieving mean absolute error (MAEs) of relative humidity (RH) ranging from 0.41 % to 4.51 % across different depths. It was further validated against 18-month field measurements of a full-scale cross-laminated timber (CLT) test wall exposed to Vancouver's climate. Hysteresis effects were assessed in two sections of the wall: one without water injection and another with controlled water injections simulating wind-driven rain (WDR) leakage. Simulation results were compared with field measurements of temperature, relative humidity (RH), and moisture content (MC) for model validation. Following validation, the model was used to simulate the CLT wall under historical and projected future climate conditions using extreme moisture reference years and including WDR exposure across four orientations. Mold growth index and wood decay (mass loss) are used as the moisture-induced risk indicators. Results showed that including hysteresis improved agreement with measured MC in the leakage-affected section and decreased the MAE during the second water-injection phase by 1.42 % MC. Hysteresis also reduced the maximum absolute difference by 3.12 % and producing more conservative drying behavior of CLT. Under long-term climate scenarios, hysteresis effects led to minor variations in predicted moisture risks depending on orientation but did not significantly alter overall performance trends. These findings suggest that while hysteresis can meaningfully influence local moisture dynamics, its exclusion in standard design simulations may be acceptable for most practical applications, though care should be taken under high moisture loading scenarios. • A hysteresis-inclusive heat and moisture model was developed in Python. • Model was validated against CLT field measurements with and without water injection. • Including hysteresis did not increase predicted mold growth risk. • Excluding hysteresis can overestimate decay under high WDR due to slower drying. • Excluding hysteresis may underestimate wood decay during repeated RH spikes.