Work

Multiscale Modelling of Heat and Mass Transfer in Hemp Wools

Year: 2025
Type: article
Abstract: This paper deals with the development of a multi-scale model of coupled heat and mass transfer for hemp wool insulating materials. Natural fibrous insulating materials from renewable resources like hemp offer an attractive environmental option for conventional insulation materials. These materials exhibit excellent thermal performances in terms of thermal conductivity and thermal capacity but are highly sensitive to moisture variations. To fully describe their behavior, we developed a multi-scale approach founded upon three characteristic scales: the heterogeneous fiber scale, the homogenized fiber network scale, and the final product scale. Using the method of volume averaging, we established transfer equations at each scale and obtained expressions for effective properties (thermal conductivity, permeability, vapor diffusivity) as a function of microstructure and water content. The initial upscaling stage between fiber and homogenized fiber provides constitutive relations between water content and properties, and the second upscaling stage provides the macroscopic hygrothermal insulation behavior. X-ray tomography 3D images enabled direct computation of the effective properties on the actual microstructure. Numerical values of effective thermal conductivity at decreased dry conditions compare favorably with experiments; for example, the calculated average thermal conductivity (0.046 W·m⁻¹·K⁻¹) is almost identical to the measured value (0.047 W·m⁻¹·K⁻¹). The calculated through-thickness vapor diffusivity (1.9 ± 0.09 × 10⁻⁵ m²·s⁻¹) also compares well with experimental data (1.9 ± 0.09 × 10⁻⁵ m²·s⁻¹). This double-scale approach offers a rigorous framework for simulating and predicting the hygrothermal behavior of hemp-based insulation under realistic environmental conditions, contributing to the optimization of these sustainable building materials.
Language: English
Open Access status: bronze
OpenAlex ID: W4411969964
FWCI:
Cites: 16
Cited by:
Related to: 10