A multi-dataset validation study of moisture prediction in Norway spruce wood exposed outdoors
2025
Niklewski, Jonas | Brischke, Christian
Wood used in outdoor construction undergoes continuous cycles of wetting and drying, resulting in fluctuating moisture contents that directly influence its long-term durability. Excess moisture above a critical threshold leads to deterioration by fungal decay, limiting the service life of the structure. Service life models thus rely on accurate predictive models of moisture behaviour. However, capturing the complexities of free water movement in real-world scenarios – where environmental factors such as rainfall and humidity vary unpredictably – remains a major challenge. Numerical approaches, particularly diffusion-based models grounded on Fick’s laws, have been used for this purpose. However, these models have not yet undergone comprehensive validation under relevant outdoor conditions. The present study directly addresses these issues by validating a single numerical model configuration across multiple datasets, including both gravimetric and point-type moisture measurements taken at the wood surface and in the core. This comprehensive validation approach seeks to ensure that the model not only predicts overall moisture fluctuations but also captures the internal distribution of water. Results indicate good agreement between simulated and observed moisture content, though further exploration under more diverse climatic conditions is recommended for further validation. Ultimately, by validating the numerical model across multiple datasets, this study evaluates the model’s predictive performance, reveals the conditions under which higher bias may arise, and examines how that bias translates into service life predictions.
Show more [+] Less [-]AGROVOC Keywords
Bibliographic information
Publisher IRG Secretariat
This bibliographic record has been provided by Thünen-Institut