Supervised machine learning and its algorithms are a developing trend in the prediction of rockfill material (RFM) mechanical properties. This study investigates supervised learning algorithms—support vector machine (SVM), random forest (RF), AdaBoost, and k-nearest neighbor (KNN) for the prediction of the RFM shear strength. A total of 165 RFM case studies with 13 key material properties for rockfill characterization have been applied to construct and validate the models. The performance of the SVM, RF, AdaBoost, and KNN models are assessed using statistical parameters, including the coefficient of determination (R²), Nash–Sutcliffe efficiency (NSE) coefficient, root mean square error (RMSE), and ratio of the RMSE to the standard deviation of measured data (RSR). The applications for the abovementioned models for predicting the shear strength of RFM are compared and discussed. The analysis of the R² together with NSE, RMSE, and RSR for the RFM shear strength data set demonstrates that the SVM achieved a better prediction performance with (R² = 0.9655, NSE = 0.9639, RMSE = 0.1135, and RSR = 0.1899) succeeded by the RF model with (R² = 0.9545, NSE = 0.9542, RMSE = 0.1279, and RSR = 0.2140), the AdaBoost model with (R² = 0.9390, NSE = 0.9388, RMSE = 0.1478, and RSR = 0.2474), and the KNN with (R² = 0.6233, NSE = 0.6180, RMSE = 0.3693, and RSR = 0.6181). Furthermore, the sensitivity analysis result shows that normal stress was the key parameter affecting the shear strength of RFM.

During severe earthquakes, liquefaction-induced lateral displacement causes significant damage to designed structures. As a result, geotechnical specialists must accurately estimate lateral displacement in liquefaction-prone areas in order to ensure long-term development. This research proposes a Gaussian Process Regression (GPR) model based on 247 post liquefaction in-situ free face ground conditions case studies for analyzing liquefaction-induced lateral displacement. The performance of the GPR model is assessed using statistical parameters, including the coefficient of determination, coefficient of correlation, Nash–Sutcliffe efficiency coefficient, root mean square error (RMSE), and ratio of the RMSE to the standard deviation of measured data. The developed GPR model predictive ability is compared to that of three other known models—evolutionary polynomial regression, artificial neural network, and multi-layer regression available in the literature. The results show that the GPR model can accurately learn complicated nonlinear relationships between lateral displacement and its influencing factors. A sensitivity analysis is also presented in this study to assess the effects of input parameters on lateral displacement.

In a combined piled raft foundation (CPRF) both raft and piles take their share of the total load applied. However, in practice, the contribution of a raft in taking load is usually ignored and the load is assumed to be supported on piles. This way of CPRF becomes excessively conservative and uneconomical. To economize the design, relative load sharing of raft and piles in CPRF has to be found. In this connection, different simplified methods have been developed, each one with some limitations. In this study, three simplified methods have been applied to two cases of pile‐raft systems. The methods include Randolph, Poulos‐Davis‐Randolph, and modified Poulos‐Davis‐Randolph. The first case is a hypothetical case consisting of a 12 m × 12 m raft supported on a square group of nine piles. The second case study is an actual eight‐story building to be constructed in Peshawar, Pakistan. The building is supported on a pile‐raft system, with raft resting on very soft clay underlain by dense sand. The two case studies are also modelled in the finite element program PLAXIS 3D for comparison. The results of all the simplified methods are comparable with PLAXIS 3D. However, the Randolph method is much closer to PLAXIS 3D for the two cases studied. Furthermore, it is also shown that piles in a piled raft system can be used as “stress reducers” as well as a “settlement reducers.” Additionally, the effect of interaction factors is also evaluated with the s/d ratio as well as with varying soil stiffness. It was concluded that ignoring these factors leads to a very unsafe design of the pile‐raft system.

This article presents the experimental and numerical analysis behavior on Mechanically Stabilized Earth Wall (MSE) under applied overburden load performed on the 1.5 m high, 0.9 m width, and 1.2 m length reinforced with deformed steel bars embedded in sand alone and tires shred-sand mixture. The study investigates how deformed bars, strength, and geometry affect the failure mechanism. The top of the wall was laden with additional overburden weight at various stages to explore the pre-failure wall behavior. The horizontal displacements were measured using potentiometers of the wall face and by potentiometers placed at the top of the loading plate. The results of the observations were compared to the analysis results derived from a numerical model created using the Plaxis 3D software. Numerical modeling was also applied to assess the behavior of MSE wall (3D model) on the failure mechanism of the walls. The parameters for the numerical models were derived from independent tests results, which were compared with the experimental observations. A good level of agreement with measurements was confirmed for the 3D model with the experimental data. From the results, it was deduced that at 30 kPa and 40 kPa, the tire shred-sand mixture with reinforcement gave a 36% and 58% reduction in face deflection compared to sand with reinforcement. The difference between numerical and experimental values ranges from 12% to 15%.

Expansive soils are problematic and viewed as a potential hazard for buildings and structures due to swell and shrink phenomena. The damaging effect of these soils is strongly correlated with the soil‐water characteristics of expansive soils present in the shallow depth. The seasonal wetting‐drying cycle is vital in fluctuating moisture content in the surficial soils. As such, soils remain unsaturated most of the time due to high absorption capacity. Therefore, it is crucial to assess them as unsaturated soil, and the soil‐water characteristic curve (SWCC) is an essential tool for measuring unsaturated soils’ mechanical and hydraulic properties. The main objective of this study was to establish both field‐ and lab‐based SWCCs for the expansive soils and compare them for determining the possible difference between them. For this purpose, eight sites of expansive soils were selected for sampling and in situ testing. These sites include three locations of Karak, three locations of Kohat, and two locations of D.I areas. Based on the experimental results, Karak’s expansive soil indicated a high suction value of 705 kPa, while D. I Khan’s soil showed the least suction equal to 595 kPa. The comparison of field and lab SWCCs for the potential sites presented a close agreement in the matric suction values beyond the air entry values (AEVs), particularly in the residual suction zones. It was also concluded that for expansive soils, the field‐ and lab‐based SWCCs are comparable beyond the AEVs. The established curves can be successfully utilized to assess local expansive soils in the framework of unsaturated soils.