Journal of Performance of Constructed Facilities

Abstract

Segmental tunnel linings have a complex structural behavior that is difficult to predict. Full-scale in situ structural monitoring allows us to gain access to these structures, observe the acting mechanisms, and better understand the actual structural behavior under realistic boundary conditions. In a recent structural monitoring project in Frankfurt, Germany, strains, deformations, and temperatures were measured at the segmental lining of a twin-tube tunnel boring machine (TBM) tunnel. Hereby, internal forces were back-calculated from the strain measurements. The results of the monitoring are presented and discussed in this paper. The measurements began immediately after the assembly of the segments and have been performed continuously over several years. Because of that, it was possible to observe that the internal forces differ significantly between the construction stage and the final state. During the construction stage, the internal forces develop quickly within hours after ring assembly, showing a distribution along the lining’s circumference that is hard to predict. Almost all radial deformations occur during this period and remain unchanged afterward. After this first early phase, a period of stress redistribution follows, including significant changes in the internal forces. At its end, the stable, final state is reached. In this state, the internal forces match better with expectations. In the monitored areas, the stresses during the construction stage slightly exceeded those of the final state, which should be avoided for economic reasons. Additional investigations regard the temperature development within the lining, the influence of the second tube construction on the already built tube, and the development of longitudinal ring interaction over time. The latter significantly influences the structural behavior, but its development is often subject to uncertainties. In this case, it was observed that the longitudinal interaction decreases considerably over time.

Practical Applications

In the structural design of segmental tunnel linings, structural or geotechnical engineers have various options regarding the available calculation methods and modeling strategies. All of these options are part of the state of the art, and the selection is up to the engineer. However, the results can be quite different depending on the chosen option. Consequently, uncertainties remain regarding the real linings’ internal forces and structural reliability. The actual mechanisms can be identified by investigating the real structural behavior of segmental tunnel linings with in situ structural monitoring. Then, with back-analyses, on the one hand, it can be tested to determine how well different calculation methods and modeling strategies correspond to reality. On the other hand, the monitoring data can contribute to improving existing calculation methods, developing new modeling approaches, and establishing more detailed recommendations or optimizations regarding the design and construction of segmental tunnel linings. This paper presents findings derived from a structural monitoring project on a segmental lining. Different mechanisms and aspects, sometimes uncertain or debatable in tunnel design, are discussed regarding their real in situ appearance. The gained data will also be used for further back-analyses and investigations.

Crack detection methods of high-rise building walls based on traditional computer vision (CV) heavily rely on manual selection and extraction of design features. Convolutional neural network (CNN)-based CV can actively learn the features of cracks and adapt to complex backgrounds, solving the limitations of traditional crack detection methods. This paper explores faster region-CNN, single shot multibox detector (SSD), You Only Look Once for crack detection, and Mask R-CNN for crack segmentation and proposes a novel automatic crack geometric quantification method by combining CNN-based object detection and segmentation. The contents include (1) crack detection and bounding box extraction, exploring a variety of models, selecting the best model to detect the image taken by an unmanned aerial vehicle (UAV), and extracting the crack region; (2) crack segmentation, using the detection results of the first part as input for more accurate detection and segmentation of cracks; and (3) a novel pixel-level geometric quantization method of crack based on Hough straight-line detection, mainly including crack length and width. Then, the pixel level is transformed into the actual geometric quantization to simply determine the crack severity. The three models generated in these three parts can be used for managing exterior wall cracks in high-rise buildings for different inspection purposes.

This study addresses the limitations of existing indirect bridge damage identification methods that are based on the vehicle–bridge coupled vibration theory of highway bridges. To overcome these shortcomings, we propose an extended approach that incorporates various types of deep-learning models with vehicle–bridge coupled vibration responses. The proposed method is demonstrated using a three-span continuous beam bridge as a case study. First, a vehicle and bridge analysis model is established, and bridge damage is simulated using unit stiffness reduction, considering different damage scenarios. Next, to account for road roughness randomness, vehicle–bridge coupling vibration analysis is performed under various road roughness conditions, yielding the vertical acceleration vibration signal of the vehicle. Subsequently, we employ an end-to-end damage recognition method, utilizing the vehicle acceleration response as the network input, to construct two types of deep-learning models: one-dimensional convolutional neural network (1D-CNN) and convolutional long short-term memory neural network (CNN-LSTM). The recognition performance of both models is compared and analyzed. Taking Zhengzhou Taohuayu Self-Anchored Suspension Bridge in China as an example, this study delves into the capability of bridge damage identification using deep learning. The results demonstrate that the one-dimensional convolutional neural network achieves excellent recognition performance in terms of both damage location and severity.

Given that PVC membranes are a current solution for waterproofing flat roofs, and their use has been increasing in Portugal, it is important to analyze the main reasons for the occurrence of unusual anomalies in membranes mechanically fastened to the roof (not protected, directly exposed to the normal atmospheric agents) such as the premature occurrence (after less than 8 years old) of very localized dark spots on its surface that are longitudinally along all the rolls of the membrane. An experimental study was carried out and essentially it showed a clear decrease in the thickness of the top layer of the membrane and its total mass in these areas—already with the presence of the knots of the reinforcement mesh on the surface of the membrane, seen through an electron microscope—which is associated with very low levels of plasticizer for the service life of this membrane. From the analysis done, it can be concluded that the defect in question is associated with a less careful formulation of the PVC compound (or mix) during the manufacturing process for unprotected membranes on the roof.

Increasing demand for railway transportation combined with more severe climate events, such as extreme heat, leads to an increased risk of degradation of track support and failure due to rail buckling. In this paper, distributed fiber optic sensing (DFOS) was used, for the first time, to assess track support degradation and the likelihood of rail dynamic buckling of curved rail sections. A monitoring campaign was conducted to measure the dynamic distributed strain response of a 9-m-long section of curved track during the passage of a passenger train. The distributed strain data were used to assess the axial strain and vertical bending curvature response during the passage of the train, and the distributed vertical curvature profile was then used to evaluate the wheel forces and track modulus of the monitored site using the Bayesian inference approach. The estimated wheel forces showed good agreement with the expected values, and the estimated track modulus was comparable to that measured using conventional techniques at other similar rail sites. With the estimated wheel forces and track modulus as inputs, a finite-element model developed in a commercial software package (i.e., ABAQUS) was used to assess the dynamic buckling capacity of the rail by considering the reduced rail lateral resistance due to the passing train. The results indicate that for this site, the passage of locomotives reduces the thermal buckling capacity by several degrees Celsius depending on the initial geometric imperfections in the rail, whereas passenger cars have negligible impact on the capacity.