Evaluation of collapsibility characteristics in deep loess strata using in-situ sand well immersion test

The significant collapse settlement of deep loess upon water infiltration poses a serious threat to engineering construction safety. Accurate evaluation of collapsibility characteristics is a critical issue in geotechnical engineering. Evaluation methods relying on laboratory tests often diverge from field results. This study focuses on the Shenheyuan loess. It conducts in-situ sand well immersion tests to monitor water migration in the loess stratum and settlement of soil layers at various burial depths during immersion. The results are compared with laboratory collapse test data. The findings indicate that during the in-situ sand well immersion test, the monitored settlements at all points were small, well below the 70 mm criterion for self-weight collapsibility in loess. Therefore, the site is classified as a non-self-weight-collapsible site. In contrast, laboratory tests classify the site as self-weight collapsible with a collapsibility level of grade II. A comparative analysis with the adjacent site confirms that evaluating collapsibility characteristics using an in-situ sand well immersion test is more reliable. The discrepancy between the laboratory test and the in-situ sand well immersion test primarily stems from the regional correction coefficient specified in the code, which fails to account for site-specific factors, such as high groundwater levels, climatic conditions, and soils with high clay content. Finally, a preliminary site-specific correction coefficient of 0.03 is back-calculated based on the measured field response. This study demonstrates the effectiveness of the in-situ sand well immersion test for evaluating the collapsibility of deep loess. It provides an important reference for geotechnical design under similar special geological conditions.