In this paper an extensive research campaign aimed to define the mechanical performance of earth block masonry panels is presented. Uniaxial compression and diagonal compression tests were performed. The test results confirmed the brittle behaviour of earth block masonry under uniaxial compressive load and showed that the failure of earth block masonry under shear load occurs by sliding of the earth blocks along the mortar joints after initial cracking in mortar joints and earth blocks. For diagonal compression test results showed that building technique practice is one of the key factors affecting the structural performances. Experimental behaviour was modelled with a non-linear model capable of describing cracking behaviour. Both micro-modelling and macro-modelling implementing isotropic or orthotropic material laws were used to assess the reliability of different modelling strategies. The model calibration was carried out by sensibility analysis of the input parameters to understand the influence of unit strength on the shear behaviour of masonry.
The paper presents a comparison between different numerical modelling approaches aiming to simulate the in-plain behaviour of three types of earthen materials, namely adobe masonry, rammed earth and cob. For this purpose, uniaxial and diagonal compression tests were carried out, which allowed determining important mechanical parameters, such as compressive strength, Young’s modulus, Poisson’s ratio, shear strength and shear modulus. Furthermore, the tests allowed assessing the level of non-linear behaviour of the respective stress–strain relationships as well as the failure modes. The experimental results were then used for the calibration of numerical models (based on the finite element method) for simulating the non-linear behaviour of the earth materials under in-plane shear loading. Both macro- and micro-modelling approaches were considered for this purpose. The procedures adopted for model calibration established the reliability of various modelling strategies for the different loading conditions. The simplified Approach based on macromodelling shows a satisfactory accuracy and low computational costs. The results reproducing the uniaxial compression are in good correspondence with the post-elastic behaviour observed in the experimental campaign. The micro-modelling approach adopted to reproduce the shear behaviour, even with higher computational cost, represents a suitable tool to predict the adobe masonry and rammed earth collapse mechanisms.