The construction of diaphragm wall panels inevitably changes the initial stress condition and causes movements in the surrounding soil mass, which may also cause settlement and damages to adjacent buildings. Majority of current design and analyses of deep excavations assume that the diaphragm wall is wished-in-place, largely because of the complexities involved to consider the detailed wall installation process. Limited studies suggested that neglecting the wall installation effects would reduce the reliability of these analyses for both predictions and validations. This paper analyzes measured ground response and building settlements caused by diaphragm wall panel installation and highlights the importance of considering these installation effects in practical design. A realistic modeling procedure is then developed to incorporate the sequential diaphragm wall panel construction process in braced excavation analyses, to investigate the installation effects on adjacent ground and buildings. The computed results are consistent with those field measurements from different case studies. The benefits of the proposed approach are demonstrated though comparison with the conventional wished-in-place approach in the braced excavation analyses.

The ground movement during the construction of shallow loess tunnels can easily cause deformation damage to surface buildings. Most the current studies focus on the damage soft soil and rock tunnels to independent buildings, and there are few studies on the case of building groups in loess areas. Using the new Xi 'Yan Railway Luochuan Tunnel as a case study, we conducted on -site testing to study building settlement and crack development characteristics. Three-dimensional numerical simulations were carried out to analyze settlement, flexure deformation, and main tensile strain distribution characteristics of the buildings at different buried depths. The study determines the extent of damage resulting from differential settlement and tension cracks. The results show that construction during the upper, middle, and lower bench stages results in significant ground volume loss, leading to a 'wide and steep ' settlement pattern with a maximum settlement value of 567 mm. Building cracks exhibit positive and inverted splayed shapes, with lengths ranging from 0.5 to 6.0 m and widths between 0 to 170 mm. As buried depth increases, maximum settlement, flexure deformation, and main tensile strain of buildings also increase. The severe damage range of buildings initially increases and then stabilizes, with the maximum range caused by differential settlement and tensile cracks being 34 m and 29 m from the tunnel axis, respectively. Based on the analysis of building damage characteristics, it was determined that a combination of surface measures and measures within the tunnel should be used to control building damage caused by tunnel construction. These research findings can serve as valuable references for similar projects.

Engineers often estimate the amount of liquefaction-induced building settlements (LIBS) as a performance proxy to assess the potential of earthquake-induced damage to buildings. The first robust LIBS models were initially developed in 2017 and 2018 using traditional statistical approaches. More recently, machine learning techniques have started to be used in developing LIBS models. These recent efforts are a step forward in realizing the potential of machine learning in liquefaction engineering; however, they have often considered only one ML technique for a given dataset and typically used only held-out test sets for model assessment. In this study, five ML-based LIBS models with varying flexibility (i.e., ridge regression, partial least square regression - PLSR, random forest, gradient boosting decision tree - GBDT, and support vector regression) are developed using a LIBS database generated by soil-structure numerical simulations of different buildings and soil profiles shaken by ground motions with varying intensity measures. The motivation for considering models with different flexibility is to include different bias-variance trade-offs. Feature selection with different ML techniques indicates that cumulative absolute velocity, spectral acceleration at one second, contact pressure, foundation width, the thickness of the liquefiable layer, and a shearing liquefaction index are important features in estimating LIBS. The developed ML-based models are assessed considering prediction accuracy in test sets, performance against centrifuge tests and case histories, and trends. The assessment indicates that the random forest, GBDT, and SVR models perform best, providing standard deviation reductions up to 40% relative to a multi-linear regression. Specifically, the random forest and GBDT models exhibit a root mean square error (RMSE) of 0.29 and a coefficient of determination (R2) of 0.93 on test sets, demonstrating a notable improvement compared to a traditional multi-linear regression model, which yields an RMSE of 0.47 and an R2 of 0.82. Moreover, random forest and GBDT, alongside SVR, show a good performance in centrifuge tests and case histories. Finally, given the scarcity of LIBS models, this study also contributes to treating epistemic uncertainties in estimating LIBS, which is ultimately beneficial for performance-based assessments.