Understanding soil characteristics can help determine how the ground will respond under certain loads. The finer fraction – the smaller, finer particles (e.g., fine sand, silt, and clay) – plays a large role in how soil will behave. When erosion occurs, it is usually those finer particles that are affected, and so there is no visible change; however, the shear strength of the soil has been compromised. Some examples of consequent failure include dam collapse, ground subsidence, and sinkholes. So how can engineers prevent this problem? What is the right mix of soils?
There is widespread understanding that stability is closely aligned with soil particle size distribution, but there have been different interpretations of the constriction size distribution. Researchers Kuang Cheng, Mengsi Duan, Buddhima Indraratna, and Thanh T. Nguyen explored soil internal instability for both prediction accuracy and robustness using particle-size and constriction-size based criteria. In their study, “Critical Examination of Internal Stability Criteria for Granular Soils and Development of a Coupled PSD-CSD Approach,” the authors developed a database of 232 different soils and then evaluated different metrics, including the mass of eroded finer particles and the change in the hydraulic conductivity of soil, to identify the internal stability of soils. Learn more about this study and how it can predict internal stability with greater reliability in the Journal of Geotechnical and Geoenvironmental Engineering at https://ascelibrary.org/doi/10.1061/JGGEFK.GTENG-13479. The abstract is below.
Abstract
The internal instability of soils involving the loss of soil mass and consequent failures under seepage has received immense attention in the past decades, resulting in various assessment methods. Empirical criteria based on the particle size distribution (PSD) and the constriction size distribution (CSD) are some of the most preferable approaches; however, their past investigations often focus on only the prediction accuracy, while underestimating the importance of prediction reliability (robustness). The current study aims to address this issue by carrying out an extensive and novel assessment of the prediction performance of the five most typical internal stability criteria, including four PSD-based criteria and one CSD-based criterion, using a large data set of 232 continuously and gap-graded soils through various experiments during 1985–2023. A novel assessment approach based on two newly proposed parameters, i.e., the percentage of deviation and the robustness index is developed and applied consistently to the selected criteria and soils. The results show that the CSD- and the Kenney and Lau PSD-based methods by incorporating the width of the gradation gap exhibit the most accurate prediction with an accuracy that can exceed 94% for gap-graded soils. However, their accuracy can drop significantly to less than 70% in continuously graded soils when closer to the borderlines (i.e., the percentage of deviation <20%). A coupled PSD-CSD guideline that can effectively employ key advantages of both PSD- and CSD-based criteria is thus proposed. The novel guideline can significantly enhance the prediction accuracy and robustness while minimizing the excessive computational cost often required in the conventional CSD-based method.
Learn more about this formula and how it could be applied to enhance your soil’s stability in the ASCE Library: https://ascelibrary.org/doi/10.1061/JGGEFK.GTENG-13479.
