Sensitivity analysis of the parameters for assessing a hygrothermal transfer model HAM in bio-based hemp concrete material

Abstract

In order to better predict the hygrothermal behavior of porous building envelopes, it is essential to develop reliable mathematical models which accurately capture the coupled heat, air, and mass (HAM) transport and its influence on the indoor atmosphere. Recently, numerous models have been developed to predict the performance of bio-based building envelopes. However, the large number of input parameters and lack of knowledge of the intrinsic characteristics of the materials mean these models are quite difficult to implement. To determine the most influential hygrothermal parameters for a hemp concrete wall, a sensitivity study was conducted on the input parameters, based on the probability density distribution of Gaussian centered law; the results are presented herein. A reduced model was developed by eliminating the parameters with little or no influence on HAM transfer. Then, a dimensionless study of the reduced model was carried out. This generated an indicative parameter ς, which can be used to classify building materials and identify the most suitable. In MATLAB software, the simulated results for a hemp concrete wall showed satisfactory concordance with the experimental results found in the literature. The temperature profiles are accurately estimated. Note that the model must always take account of the evolution of water content to ensure accurate RH predictions.