This paper investigates the swelling behaviour of compacted MX80 bentonite under hydration with the presence of initial voids that simulate the technological voids in real nuclear waste repositories. With water introduced into samples from the initial voids, the swollen bentonite first filled up the initial voids, and then swelling pressure was generated. At different hydration times, the microstructure was investigated at different heights (initial void, top, middle and bottom) using mercury intrusion porosimetry, together with the determination of dry density, water content and suction. It was observed that the microstructure was characterised by the presence of a large- and a small-pore families at bottom and middle layers, but by the creation of a new medium-pore family with dominant pore mode around 0·04–2 μm at the top and initial void layers after the void space was filled by swollen bentonite. From bottom to void, the significant increase in medium pores was accompanied by an increase in inaccessible pores and a decrease in large and small pores. Further examination showed that the sample could be divided into compression and swelling zones based on the variation of void ratio with swelling pressure: at the compression zone, the soil was compressed over time, characterised by a significant reduction of medium and large pores; whereas at the swelling zone, the soil was still undergoing swelling, represented by the increase of inaccessible pores and medium pores. In addition, the changes in microstructure along the sample were found to be well correlated with changes in dry density and water content: the water content was decreasing and the dry density was increasing from the initial void to the bottom positions; over time the dry density of the upper part was increasing due to compression, while the dry density of the lower part was decreasing due to swelling. This provided evidence for the main mechanism of compacted bentonite swelling with the presence of technological voids: the swollen bentonite fills the initial voids first and then undergoes compression by the swelling of the bentonite behind.

Isotropically consolidated undrained triaxial compression shear tests were performed on three reconstituted clays to investigate the effect of initial water content w₀on undrained strength behaviour. The values of w₀were adjusted within the range of 1·0–2·0 times the liquid limit. The predominant clay mineral is identified as illite for the considered clays, based on a semi-quantitative analysis of the X-ray diffraction patterns. The laboratory tests show that the stress–strain curve in terms of deviator stress against axial strain and the effective stress path in terms of deviator stress against effective mean stress are significantly affected by w₀. The undrained strength ratio R*_su, defined as the ratio of the undrained shear strength S*_uto the isotropic consolidation stress, varies with w₀within a wide spectrum, ranging from 0·28 to 0·60 for the three reconstituted clays investigated. The relationship between void index I_vand S*_uchanges with R*_su, and becomes identical to the intrinsic strength line proposed by Chandler when R*_su= 0·33. It is also evidenced from the laboratory tests that the value of S*_udepends on both the water content and the liquid limit.

This paper presents the role of superabsorbent polymer (SAP) on strength and microstructure development in cemented clays with notably high water content. A series of unconfined compressive strength (UCS), scanning electron microscope (SEM) and X-ray diffraction (XRD) tests were performed to identify strength behavior and microstructure. Results showed that SAP significantly influenced the mechanical behavior of cemented clays with notably high water content, characterized by an increase in the unconfined compressive strength and a decrease in the after-curing water content with SAP content. This revealed that the strength increase due to SAP was directly related to the water absorption by SAP. Meanwhile, XRD results showed that the hydration products were controlled by cement and lime content, regardless of SAP content. That meant there was no chemical reaction between SAP particles used in this study and cement or lime. The microstructure analysis by SEM revealed that SAP played an important role in the microstructure of cemented clays. With an increase in SAP content, the water absorbed by SAP increased significantly, leading to a decrease in the pore volume and a denser soil fabric. This behavior indicated that the primary role of SAP on strength increase was to absorb and fix water in cemented clays. Consequently, the clay–cement cluster distance decreased with an increase in solid mass (soil particles and swollen SAP particles) and a decrease in pore water. The corresponding tighter flocculated fabric due to SAP eventually led to the strength increase.