Tensile behaviour of compacted bentonite assessed by a Brazilian splitting test: role of particle size and water content
摘要
In deep geological disposal of high-level radioactive waste, the engineered barrier system typically employs compacted bentonite as a buffer material, whose tensile strength is a key parameter controlling the initiation and propagation of tensile cracks. This paper first discusses the applicability of different loading methods in traditional Brazilian splitting test for compacted bentonite and selects flat platen loading as the most suitable research method. Subsequently, using the flat platen Brazilian splitting test, the splitting characteristics of compacted bentonite under different raw material particle sizes and water contents were investigated. The fractal dimension was used to characterize the roughness of the fracture surfaces, and correlations between the fractal dimension, peak load, and failure characteristics were proposed. Finally, the internal mechanisms of splitting failure in different specimens were analyzed based on pore structure characteristics. The study shows that under the same water content conditions, as the raw material particle size decreases, crack initiation in compacted bentonite gradually becomes less predictable, brittle characteristics progressively enhance, and the peak load tends to increase. When the water content is below 15%, the peak load of specimens with larger particle sizes gradually increases with rising water content, while the opposite trend is observed for smaller particle sizes. A significant negative correlation exists between the fractal dimension and the peak load. All specimens exhibited bimodal pore characteristics, with larger pores showing gradually reduced fluctuation in response to water content changes as the raw material particle size decreases. Finally, a predictive model for peak load based on average pore size density and water content was proposed. This work enhances the understanding of the tensile properties of compacted bentonite and provides valuable insights for optimizing buffer materials in deep geological disposal repositories.