An inverse method for the estimation of the vapor and liquid diffusivity coefficient of conventional and phase change material based hemp concrete

This study proposes a method for fast estimation of the hydric properties, including liquid and vapor diffusivities, of three distinct materials: a conventional biosourced insulating material (hemp concrete) and two innovative hybrid composites (hemp concrete with capric acid and hemp concrete with lauric acid). To estimate these parameters, an approach based on an inverse method has been developed. It involves minimizing the quadratic difference between numerical and experimental values of relative humidity, using the Nelder-Mead algorithm by assuming an exponential evolution of the liquid and vapor diffusivities. An unsteady -state experiment has been conducted in a climatic chamber during 18 days to provide experimental data for parameters ' estimation. The obtained maximum root mean square error (RMSE) for all the samples is 4.4% that is a satisfactory approximation given the model assumptions and the experimental uncertainties ( +/- 2%). The estimated parameters allow calculating all investigated materials ' liquid and vapor diffusivities. The validation of the approach is achieved by comparing its results with those from the existing literature.