Implementation and validation of a 3D image-based prediction model for the thermal conductivity of cellular and granular porous building blocks

Porous building blocks are increasingly being used in the building sector, as they offer a relatively high thermal resistance because of their porous microstructure. A clear understanding of the relation between the material's microstructural properties and effective thermal conductivity is still lacking though, impeding a correct analysis and design of existing and new porous building blocks. Therefore, this paper presents a 3D model framework to study the heat transfer through both cellular and granular porous building blocks, performing simulations directly at the pore scale. A 3D voxel-grid representation of the pore structure is adopted, allowing to study both virtually generated materials and actually available materials. The accuracy and the impact of the grid parameters is verified on a set of elementary microstructures, demonstrating a good performance when using adequate settings. The thermal conductivity simulation framework is validated against experimental measurements on two highly different porous materials: a low-porosity granular sintered glass filter and a highly-porous cellular-granular acoustic absorber. Comparison of simulation outcomes with experimental results confirms the good performance of the model, indicating its potential to evaluate and optimize the thermal conductivity of porous building blocks.