Adding cement to soft soils may lead to brittle behavior and the occurrence of sudden damage. Methods to further improve the tensile and flexural properties of cemented clay are noteworthy topics. This paper mainly focuses on the effect of cement and moisture content on the strength and flexural properties of cemented clay reinforced by PVA fiber. The selected clayey soil was a kaolin with cement content of 5%, 10%, and 15% and moisture content of 50%, 56%, 63%, and 70%. The results show that the incorporation of 0.6% fiber can effectively improve the deformability of cemented clay in unconfined compression tests (UCS). The strengthening effect of fiber, as seen in the peak strength and post-peak strength of UCS, was significantly related to cement content. As the water content increased, the compressive strength of the fiber-reinforced cemented clay decreased, but its load-bearing capacity enhanced. When the cement content was 15%, the splitting tensile strength of fiber-reinforced cemented specimens increased by 11% compared to cemented soil, but the deformability of the specimens became poor. In the cement-content interval from 5% to 10%, the bending toughness was significantly improved. Sufficient cement addition ensures the enhancement of PVA fibers on strength and flexural properties of cement-stabilized clayey soil.

Cement-admixed clay (CAC) is a widely-used soil stabilization technique for enhancing the strength and stiffness of soft clay. However, the stress-strain behavior of CAC is complex and nonlinear, and also depends on various factors such as mixing proportion, confining pressure, stress path, and shearing condition. In this study, we propose a novel approach for modeling the stress-strain behavior of CAC using recurrent neural networks (RNNs), which are a type of deep learning (DL) technique that can well capture the temporal dependencies and nonlinearities in sequential data. We compare three types of RNNs: traditional RNN, long short-term memory (LSTM) neural network, and gated recurrent unit (GRU) neural network, and evaluate their performance in simulating the strain- and stress-controlled triaxial test results of 25 CAC specimens with different mixing proportions and confining pressures. The results demonstrate that the LSTM model, incorporating a 2-time step backward, exhibits superior prediction accuracy and generalization capability compared to other evaluated models, achieving a mean absolute percentage error (MAPE) of 4%. This LSTM model is capable of capturing the stress-strain behavior of CACs across various loading conditions and mixing proportions within a unified framework. Therefore, we suggest that the LSTM model is a promising tool for modeling and analyzing the mechanical behavior of CAC in geotechnical engineering applications.