Experimental and numerical estimation of thermal conductivity of bio-based building composite materials with an enhanced thermal capacity

Abstract

The paper tackles the important problem of estimating and predicting the thermal conductivity of bio-based building materials that can help decarbonize the building sector. Knowledge and the possibility to optimize their insulating properties is the first criterion for determining the ability of their use in the building sector. The novel yet not well-studied hemp shives and magnesium binder composites with improved thermal mass by microencapsulated phase change material (PCM) were considered. The samples of composites without PCM having different densities as well as with different amounts of microencapsulated PCM and the same density were manufactured and tested experimentally in different states, i.e., dry and after conditioning at relative humidity (RH) of 50, 75, and 90 %, and in the case of samples with PCM also at different average measurement temperatures selected with respect to the phase change range of the PCM. A novel method of predicting bio-based composites' effective thermal conductivity tensor was also developed. The method is based on the numerical solution of the heat conduction equation at the micro-scale considering real composite microstructure. The method was tuned using micro-computed tomography (μCT) data with the microstructure of composites without PCM and then applied to predict the thermal conductivities of composites with PCM, utilizing computational domains with an artificially distributed PCM in the binder. The experimental testing of composites without PCM revealed an apparent effect of increasing thermal conductivity with rising sample density and RH applied during conditioning. With an increase in density from 394 to 576 kg/m3, their thermal conductivities varied from 0.105 to 0.159 W/m/K for the dry state and from 0.157 to 0.241 W/m/K for RH 90 %. A similar effect of sample state and RH was observed for composites with microencapsulated PCM, which had the highest thermal conductivity, equal to 0.265 W/m/K, for the lowest PCM amount and RH 90 %. Moreover, a decrease in composites' effective thermal conductivity with an increase in PCM fraction was observed, except for the lowest PCM fraction, for which it increased. It also rose with increasing the average measurement temperature. The developed micro-scale-based numerical method allowed for predicting the thermal conductivities of composites with accuracies below 4.1 and 7.2 % for composites without and with PCM, respectively, except for the composite with the lowest amount of PCM, which behaved differently. Thus, its suitability for designing and optimizing bio-based composites was shown.