Reliability analysis of rammed earth structures

Abstract

Rammed earth (RE) structures are widely used for more sustainable and environment-friendly buildings. Due to lack of design standards, the engineering decisions often rely on rule-of-thumb method which may lead to quite conservative or unsafe designs. In this study, load and resistance parameters were treated as random variables in reliability analysis. The reliability index and failure probability of RE structures were evaluated using First-Order-Reliability-Method (FORM) and then compared with Second-Order-Reliability-Method (SORM) and Monte Carlo Sampling method. The analysis was performed based on the different a) load combinations, b) wall geometry, c) material type (unstabilized or cement stabilized) and d) mechanical properties of the materials. Based on the results, the RE wall under moderate loading conditions require smaller wall thickness than recommended wall thickness by various guidelines such as New Mexico-USA, New Zealand and Zimbabwe Codes. However, larger wall thickness is needed under severe loadings conditions, especially when unstabilized materials are used. The compressive strength of unstabilized materials under severe loading conditions should be more than minimum recommended. The sensitivity analysis was performed by calculating different importance and sensitivity vectors. The results show that the compressive strength and the environmental loads factors are the most important random variables that contribute to reliability of the structures. The recommended wall thickness and compressive strength for different conditions are presented.