Earth construction: Field variabilities and laboratory reproducibility

Abstract Building construction is a major polluting sector. As a result, there is increasing global interest in the development of sustainable building materials with low environmental impact. Earth-based materials are among the materials of interest and building with earth-based materials has thus received a particular renewal of attention. Previous research has focused on the physical characteristics and durability of these materials. The aim of this study is to assess the variability of materials made in-situ and their reproducibility in the laboratory using an automatic normal Proctor machine with different compaction energies. Both cob and light earth were investigated. Cylindrical and prismatic specimens were produced on-site and in the laboratory: cob was made of silt, silty clay, sandy silt, and flax straw; and a separate layer of light earth was made of elastic silt and reed fibres. An experimental program was designed to evaluate the properties of the materials in terms of their water content, density, porosity, compressive strength, and thermal conductivity. The results revealed that the in-situ densities could be reproduced in the laboratory with compaction energies of 0.6 MJ/m3 and 0.2 MJ/m3 for cob and light earth, respectively. These compaction energies will allow the research to produce laboratory specimens that were representative of the materials implemented on-site. Regarding the compressive strength, the values obtained in the laboratory were higher than those of the in-situ specimens. Correction factors of 0.88 and 0.67 for cob and light earth. These values should be applied to calibrate the laboratory results in relation to in-situ. Concerning the thermal conductivity, the values obtained in the laboratory were similar for cob and higher for light earth. A correction factor of 0.87 should be applied to calibrate the laboratory results to those obtained in-situ.