Search for: All records

Drones are increasingly used during routine inspections of bridges to improve data consistency, work efficiency, inspector safety, and cost effectiveness. Most drones, however, are operated manually within a visual line of sight and thus unable to inspect long-span bridges that are not completely visible to operators. In this paper, aerial nondestructive evaluation (aNDE) will be envisioned for elevated structures such as bridges, buildings, dams, nuclear power plants, and tunnels. To enable aerial nondestructive testing (aNDT), a human-robot system will be created to integrate haptic sensing and dexterous manipulation into a drone or a structural crawler in augmented/virtual reality (AR/VR) for beyond-visual-line-of-sight (BVLOS) inspection of bridges. Some of the technical challenges and potential solutions associated with aNDT&E will be presented. Example applications of the advanced technologies will be demonstrated in simulated bridge decks with stipulated conditions. The developed human-robot system can transform current on-site inspection to future tele-inspection, minimizing impact to traffic passing over the bridges. The automated tele-inspection can save as much as 75% in time and 95% in cost.

 

Bridge inspection is an important step in preserving and rehabilitating transportation infrastructure for extending their service lives. The advancement of mobile robotic technology allows the rapid collection of a large amount of inspection video data. However, the data are mainly the images of complex scenes, wherein a bridge of various structural elements mix with a cluttered background. Assisting bridge inspectors in extracting structural elements of bridges from the big complex video data, and sorting them out by classes, will prepare inspectors for the element-wise inspection to determine the condition of bridges. This article is motivated to develop an assistive intelligence model for segmenting multiclass bridge elements from the inspection videos captured by an aerial inspection platform. With a small initial training dataset labeled by inspectors, a Mask Region-based Convolutional Neural Network pre-trained on a large public dataset was transferred to the new task of multiclass bridge element segmentation. Besides, the temporal coherence analysis attempts to recover false negatives and identify the weakness that the neural network can learn to improve. Furthermore, a semi-supervised self-training method was developed to engage experienced inspectors in refining the network iteratively. Quantitative and qualitative results from evaluating the developed deep neural network demonstrate that the proposed method can utilize a small amount of time and guidance from experienced inspectors (3.58 h for labeling 66 images) to build the network of excellent performance (91.8% precision, 93.6% recall, and 92.7% f1-score). Importantly, the article illustrates an approach to leveraging the domain knowledge and experiences of bridge professionals into computational intelligence models to efficiently adapt the models to varied bridges in the National Bridge Inventory. 

In this Letter, a high-accuracy, two-dimensional displacement sensor is proposed, designed, and demonstrated based on the concept of an extrinsic Fabry–Perot Interferometer. The sensor is composed of two bundled single-mode optic fibers in parallel and two plasmonic metasurface resonators inscribed on a gold substrate via a focused ion beam. The fiber end surface and the metasurface are in parallel with a small cavity between. The cavity change or$Z$-component displacement is determined from the pattern of interference fringes. The$X$-component displacement, perpendicular to the$Z$component, is identified from wavelength-selective metasurface resonators, which possess unique resonant wavelengths due to different nanostructure designs. The sensor was calibrated with six displacements applied through a three-axis precision linear stage. Test results indicated that the proposed interferometer can measure displacements with a maximum error of 5.4 µm or 2.2%.

 

As-grown graphene via chemical vapor deposition (CVD) has potential defects, cracks, and disordered grain boundaries induced by the synthesis and transfer process. Graphene/silver nanowire/graphene (Gr/AgNW/Gr) sandwich composite has been proposed to overcome these drawbacks significantly as the AgNW network can provide extra connections on graphene layers to enhance the stiffness and electrical conductivity. However, the existing substrate (polyethylene terephthalate (PET), glass, silicon, and so on) for composite production limits its application and mechanics behavior study. In this work, a vacuum annealing method is proposed and validated to synthesize the free-stand Gr/AgNW/Gr nanocomposite film on transmission electron microscopy (TEM) grids. AgNW average spacing, optical transmittance, and electrical conductivity are characterized and correlated with different AgNW concentrations. Atomic force microscope (AFM) indentation on the free-stand composite indicates that the AgNW network can increase the composite film stiffness by approximately 460% with the AgNW concentration higher than 0.6 mg/mL. Raman spectroscopy shows the existence of a graphene layer and the disturbance of the AgNW network. The proposed method provides a robust way to synthesize free-stand Gr/AgNW/Gr nanocomposite and the characterization results can be utilized to optimize the nanocomposite design for future applications. 

In this study, the strain transfer rate from an axially loaded, inelastic concrete tube to a glass fiber reinforced polymer (GFRP) packaged optical fiber with Bragg gratings is derived when the radial deformation of an “equivalent elastic” concrete tube is constrained by the packaged fiber. The concrete strains, both undisturbed and disturbed by the presence of the fiber Bragg gratings sensor, are analytically evaluated, and their difference (up to over 30%) is related to the development length at two free ends of the GFRP package. The mechanism of strain transfer is dominated by a ratio of average fiber and concrete strains in elastic range and by the averaging effect and a ratio of disturbed and undisturbed concrete strains in inelastic range. The analytical strain transfer rate was significantly reduced from 0.95, when concrete behaved elastically, to less than 0.4, when concrete damaged severely. This result was experimentally validated with less than 10% difference prior to concrete fracture. The validated model is applicable to fiber optic sensors that are embedded into concrete structures by a concrete cover of at least 10 times of the radius of the optic fiber.