Utilising recycled materials, such as construction and demolition waste (C&DW), into soil improvement projects offers a promising solution to reduce the environmental impact of the C&DW industry. This approach helps address issues related to waste generation, resource depletion, and environmental degradation, while enhancing the overall sustainability and resilience of soil stabilisation efforts. This study investigates the effectiveness of incorporating recycled C&DW into cement-treated peat and clayey soils to enhance their strength and stiffness. To achieve this goal, laboratory experiments were conducted on over 296 soil specimens to assess their Unconfined Compressive Strength (UCS), small-strain Young's modulus (E0) and shear modulus (G0). These tests included varying curing times (28, 60, 90, and 120 days), different cement and recycled material content, and water-to-cement ratios. Moreover, laboratory testing methods for determining geotechnical parameters are often time-consuming and prone to challenges. In this context, reliable predictive models, such as artificial neural networks (ANNs), offer an efficient alternative for accurately assessing these parameters. The findings of this research reveal that, along with cement content, the water-to-cement ratio (w/c) and curing time are key factors influencing the strength and stiffness of treated soft soils, underscoring their critical role in soil stabilisation. Additionally, while minimizing cement content and increasing RM yield improvements in both peat and clay, the effect is more pronounced in peat due to the time-dependent nature of pozzolanic reactions. This suggests that achieving optimal performance with increased strength and stiffness requires a carefully balanced RM content. Finally, the study demonstrates that ANN-based models can accurately predict the strength and mechanical properties of soft soils, offering a viable alternative to traditional UCS and FFR tests.

Investigations on the changes in pore water pressures and stress during the construction of stiffened deep cement mixing (SDCM) piles are scarce, resulting in an unsatisfactory understanding of the bearing capacity formation process. Thus, this paper presents a preliminary field study to investigate the variation characteristics of pore water pressures, total stress and effective stress during the construction of SDCM piles derived from field tests. In the meantime, cone penetration tests (CPTs) were conducted before and after the construction of SDCM piles. The results show that the variation ranges of pore water pressure, total stress and effective stress of soils around piles decreased with increasing distance between the measuring point and piles when the depths of the measuring points were the same. During the piling process, the effective stress increased by approximately 53-103%, and the pile side frictions increased accordingly, while the tip resistance and side resistance values of soils around piles increased by 27-106% and 2-145%, respectively. Additionally, SDCM piles successively formed different load-bearing components with decreasing bearing capacity along the pile diameter direction, which realized a better bearing efficiency than conventional piles made with homogeneous materials. In essence, they were also the source of significant economic advantages of SDCM piles. Through this study, we expect to provide a reference for further studies on the bearing mechanism of SDCM piles in soft soil regions.