Urban planning and big infrastructure designing demand two novel and seemingly contrasting approaches: 1) a continuous description of subsoil nature and behavior under natural hazards, to increase the resilience of urban areas; and 2) a reliable characterization of subsoil hydro-mechanical properties and monitoring their working behavior. Both exigences can be addressed by reconstructing 3D mechanical models at a local scale by extracting from large databases several in-situ testings, already available for several urbanized territories worldwide. In this paper, 182 cone tip resistance qc, sleeve friction fs, and pore pressure u2 profiles, drawn from CPTs performed in the Bologna district (Padania Plain, Italy), have been used. Here, the alluvial deposits are mixtures of silt, clay, and sands, and they locally show gravel lenses where the ancient fans from the Apennines can be detected. Their heterogeneous hydro-mechanical characters cannot be described only through point investigations such as CPTs. Additionally, the variability of these mechanical profiles and the uncertainties due to the limited amount of data must be assessed and used in designing and hazard mapping. Thus, to draw a continuous 3D subsoil mechanical model based on these 182 CPTs, the Partially Heterotopic Co-Kriging technique (PHCK) has been applied. This approach is a multivariate technique that can be used when only some measurements are taken at the same locations. It allows for the estimation of the distribution of qc, fs, and u2 values in the studied domain by considering the spatial variability of the preceding random functions but also their spatial correlations. Differences in variance and spatial resolution between the measurements on the horizontal plane and along the vertical direction were accounted for by considering anisotropic spatial dependence models. As a result, this study has provided horizontal maps and vertical sections of qc, fs, and u2, as well as their 3D solid models.

Densification-type ground improvement has gained wide acceptance as an effective technique to improve the static and cyclic strength and seismic performance of liquefiable soils. However, there has been little attempt to quantify the degree of change in certain measures of soil spatial variability following densification. This study uses the random field model (RFM) framework to quantify changes in the trend, inherent variability, and the autocorrelation of soil following various densification-based ground improvement methods using case history data. Ground improvement technologies investigated include driven displacement piles, vibro-replacement (i.e., stone columns), and deep dynamic compaction. Vertical RFM parameters are shown to change significantly following densification, with reductions in the coefficient of inherent variability, and increases in the scale of fluctuation, the latter of which appears related to the initial relative density. One case history is used to illustrate the utility of the RFM framework to link spatial variability of the post-shaking settlements to the results of a geostatistical model of the subsurface, including the three-dimensional distribution of CPT measurements and fines content. It is shown that the spatial distribution of differential surface settlements is correlated to those spatially-varying subsurface characteristics that are known to lead to larger liquefaction-induced settlement. Normal 0 false false false EN-US X-NONE X-NONE /* Style Definitions */ table.MsoNormalTable {mso-style-name:Table Normal; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-priority:99; mso-style-parent:; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-para-margin-bottom:.0001pt; mso-pagination:widow-orphan; font-size:10.0pt; font-family:Times New Roman,serif;}

This paper showcases novel underground stratification based on sparse machine learning (SML) methods, including sparse Bayesian learning (SBL) and least absolute shrinkage selection operator (Lasso). The SML methods proposed by the authors were applied to two- and three-dimensional underground stratification of actual sites, Odagawa Riverbank (Okayama, Japan) and New Lock (Terneuzen, the Netherlands), to demonstrate their performances. Cone penetration test (CPT) data were available in both sites, and they were converted to the soil behavior type (SBT) index for the underground stratification analysis. The trends of Ic/SBT profiles and their distribution colormaps estimated by the two methods were compared to discuss the methodological characteristics. For the Odagawa Riverbank case, the detection ratio of SBT obtained by the two methods was also compared to investigate the estimation accuracy in terms of stratification ability.

An instrumented trial embankment was constructed on the soft ground with the use of prefabricated vertical drains at Ballina in northern New South Wales (NSW, Australia) as part of the Australian Research Council Centre of Excellence for Geotechnical Science and Engineering (CGSE) research program. Comprehensive geotechnical site investigations were performed and field monitoring data were collected, which make it possible to study and compare various settlement prediction methods. In this paper, the Asaoka method, the Hyperbolic method, and the Bayesian updating approach are employed to predict the settlement of the trial embankment. The predictions of the three methods are compared for different amounts of monitoring data. The results show that the predicted settlements are close to the measurements and the accuracy of the ultimate settlement prediction can be improved by incorporating more monitoring data. The time increment can significantly influence the accuracy of the result predicted by the Asaoka method. The Bayesian updating approach is in agreement with the observed settlement by using only 215 days of the monitoring data. Normal 0 false false false EN-US X-NONE X-NONE /* Style Definitions */ table.MsoNormalTable {mso-style-name:Table Normal; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-priority:99; mso-style-parent:; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-pagination:widow-orphan; font-size:10.0pt; font-family:Times New Roman,serif;}

The present paper describes an assessment method for risks brought on by heavy rain in order to determine the priority of countermeasures for small earth-fill dams. To evaluate the failure risks of an earth-fill dam, defined as the product of failure probability and damage costs, an unsaturated seepage flow analysis is carried out based on the measured rainfall conditions. The damage costs are calculated using the correlation between the results of a flood analysis after breaching and land use and asset data. Subsequently, an evaluation to prioritize the countermeasures is conducted based on the failure risks. The larger of the two comparative failure probabilities—namely, the probability of overflow and the probability of slip failure—will showcase the dominant factor in the failure of small earth-fill dams, and the failure risk evaluation will determine the priority of the countermeasures for repairing and/or reinforcing decrepit earth-fill dams.

The slope displacement prediction is crucial for the development of an early warning system, which can help prevent or reduce losses of lives, properties, and the local environment. This problem is particularly important in the Three Gorges Reservoir (TGR) area, where the influence of geological, weather, and hydraulic conditions on landslides is significant. It is generally acknowledged that reservoir landslides are complex nonlinear systems with dynamic and inter-related features. However, most studies focus on how to express the static relationships between triggering factors and the landside displacement. In this paper, a long short-term memory (LSTM) neural network model was applied for predicting the total displacement of the Bazimen landslide, based on the decomposition of displacement time series. The accumulated displacement can be divided into two main parts: the trend and the periodic terms. The long-term trend was fit with a cubic nonlinear regression model; the residual one (the periodic displacement) was predicted via the LSTM model. By analyzing historical information and the Pearson correlation coefficient, a dynamic model was developed using six controlling factors. The good consistency between the predicted and monitored data proves the superiority of the model in predicting dynamic time-series problems. Compared with conventional static methods (i.e., MARS and SVM), the LSTM model can make full use of historical information due to its special “memory” structure. However, all these three methods can only perform well in forecasting one-step problems. To meet the requirement of multi-step forecasting, the Facebook Prophet model was also used in this study to predict landslide displacements with a longer period. The predicted results demonstrate the model’s superiority in efficiency and practice, at a cost of prediction accuracy.

This paper reports the results of calculations for the margins of safety for internal stability limit states for two as-built geosynthetic MSE walls constructed in 2006 in Washington State, U.S.A. The walls are unique because a) the internal stability design for the reinforcement was based on a working stress design method which includes the stiffness of the reinforcement as a key parameter, and b) the walls were instrumented and monitored during and after construction. Margins of safety are quantified deterministically using factors of safety, and probabilistically using reliability indices. The calculations are performed using the properties of the as-built materials as opposed to the properties assumed at the time of original design. The results show that the margins of safety for tensile and pullout strength are huge and thus these limit states are not of practical concern for these structures. The critical limit state is the soil failure limit state. Margins of safety for this limit state are shown to be closer to recommended minimums but nevertheless adequate for all reinforcement layers in both walls.

Normal 0 false false false EN-US ZH-TW AR-SA /* Style Definitions */ table.MsoNormalTable {mso-style-name:Table Normal; mso-tstyle-rowband-size:0; mso-tstyle-colband-size:0; mso-style-noshow:yes; mso-style-priority:99; mso-style-parent:; mso-padding-alt:0in 5.4pt 0in 5.4pt; mso-para-margin:0in; mso-pagination:widow-orphan; font-size:10.0pt; font-family:Times New Roman,serif;} The question to be answered is: how well are current P-y curves predicting the behavior of large diameter piles subjected to monotonic lateral loading? Current P-y curves were developed about 60 years ago based on lateral load tests on piles which ranged from 0.3 to 0.6 m in diameter. Today’s pile diameters can reach 4 m or more. This significant difference in scale brings into question the applicability of these early P-y curves to today’s large diameter piles. A horizontal load tests database of 46 piles with diameters larger than 1.5 m (up to 3.0 m) and 64 piles with diameters less than 1.5 m both in sand and in clay was assembled. Predictions of load and displacement were carried out using commonly used P-y curves and the software LPILE. Within this paper, these predictions are compared to the measured loads and measured displacements. The ratio of predicted over measured quantities is plotted against pile diameter, and trends are noted. Modifications to the P-y curves are then proposed so that the ratio of predicted over measured displacement remains approximately independent of diameter. Finally, the probability that the predictions will be unsafe is evaluated.