This article investigates the performance of finite element model updating to identify the induced damage in a two-story reinforced concrete masonry-infilled building using vibration data as well as lidar (light detection and ranging) scans. The building, located in El Centro, California, was severely damaged due to the 2010 El Mayor–Cucapah (Baja California, Mexico) Earthquake, and it was planned to be demolished following a number of ambient and forced vibration tests. The forced vibration tests were performed using an eccentric mass shaker. During the testing sequence, damage was induced to the building by removing four exterior walls. The modal parameters of the structure are estimated using the ambient vibration and forced vibration measurements at the reference state and damaged state. Lidar data are also used to detect surface defects and quantify the temporal changes of surface defects caused by the wall removal and forced vibration tests. Based on site inspections, geometry measurements, and material test data, two initial finite element models are built, namely the un-tuned initial model and the tuned initial model. The tuned initial model implements stiffness reduction factors to account for the observed damage in the building at its reference state while the un-tuned model does not. Two sets of reference models are calibrated to represent the structure at the reference state using the un-tuned and tuned initial models. The reference models are then updated to fit the measured data at the damaged state of the building with damage being estimated as the loss of stiffness in updating substructures. The estimated damage is compared to the nominal value of induced damage and surface defects detected by lidar scans. The analysis of the results indicates that the un-tuned and tuned initial models provide similar updated models and damage identification results which are in good agreement with the nominal values of damage and lidar detection results.

Few studies have been conducted to systematically assess post-earthquake condition of structures using vibration measurements. This paper presents system identification and finite element (FE) modeling of an 18-story apartment building that was damaged during the 2015 Gorkha earthquake and its aftershocks in Nepal. In June 2015, a few months after the earthquake, the authors visited the building and recorded the building's ambient acceleration response. The recorded data are analyzed, and the modal parameters of the structure are identified using an output-only system identification method. A linear FE model of the building is also developed to estimate numerically its dynamic properties. The identified modal parameters are compared to those of the model to identify possible shortcomings of the modeling and identification approaches. The identified natural frequencies and mode shapes for two of the three closely spaced vibration modes in the lower frequency range of interest (0.2-1.0 Hz) are in good agreement with the numerical model. The model is used to estimate the response of the building to the nearby recorded ground motion due to earthquake and the main aftershock. The maximum drift ratios are compared to the observed damage in the building and surface defects detected and quantified by the lidar scans as the research team performed a series of light detection and ranging (lidar) scans from interior of selected floors to document the damage patterns along the height of the building.

Aerial data collection is well known as an efficient method to study the impact following extreme events. While datasets predominately include images for post-disaster remote sensing analyses, images alone cannot provide detailed geometric information due to a lack of depth or the complexity required to extract geometric details. However, geometric and color information can easily be mined from three-dimensional (3D) point clouds. Scene classification is commonly studied within the field of machine learning, where a workflow follows a pipeline operation to compute a series of engineered features for each point and then points are classified based on these features using a learning algorithm. However, these workflows cannot be directly applied to an aerial 3D point cloud due to a large number of points, density variation, and object appearance. In this study, the point cloud datasets are transferred into a volumetric grid model to be used in the training and testing of 3D fully convolutional network models. The goal of these models is to semantically segment two areas that sustained damage after Hurricane Harvey, which occurred in 2017, into six classes, including damaged structures, undamaged structures, debris, roadways, terrain, and vehicles. These classes are selected to understand the distribution and intensity of the damage. The point clouds consist of two distinct areas assembled using aerial Structure-from-Motion from a camera mounted on an unmanned aerial system. The two datasets contain approximately 5000 and 8000 unique instances, and the developed methods are assessed quantitatively using precision, accuracy, recall, and intersection over union metrics.

To achieve risk-based engineered structural designs that provide safety for life and property from tornadoes, sufficient knowledge of tornado wind speeds and wind flow characteristics is needed. Currently, sufficient understanding of the magnitude, frequency, and velocity structure of tornado winds remain elusive. Direct measurements of tornado winds are rare and nearly impossible to acquire, and the pursuit of in situ wind measurements can be precarious, dangerous, and even necessitating the development of safer and more reliable means to understand tornado actions. Remote-sensing technologies including satellite, aerial, lidar, and photogrammetric platforms, have demonstrated an ever-increasing efficiency for collecting, storing, organizing, and communicating tornado hazards information at a multitude of geospatial scales. Current remote-sensing technologies enable wind-engineering researchers to examine tornado effects on the built environment at various spatial scales ranging from the overall path to the neighborhood, building, and ultimately member and/or connection level. Each spatial resolution contains a unique set of challenges for efficiency, ease, and cost of data acquisition and dissemination, as well as contributions to the body of knowledge that help engineers and atmospheric scientists better understand tornado wind speeds. This paper examines the use of remote sensing technologies at four scales in recent tornado investigations, demonstrating the challenges of data collection and processing at each level as well as the utility of the information gleaned from each level in advancing the understanding of tornado effects.

Assessment and evaluation of damage in civil infrastructure is most often conducted visually, despite its subjectivity and qualitative nature in locating and verifying damaged areas. This study aims to present a new workflow to analyze non-temporal point clouds to objectively identify surface damage, defects, cracks, and other anomalies based solely on geometric surface descriptors that are irrespective of point clouds’ underlying geometry. Non-temporal, in this case, refers to a single dataset, which is not relying on a change detection approach. The developed method utilizes vertex normal, surface variation, and curvature as three distinct surface descriptors to locate the likely damaged areas. Two synthetic datasets with planar and cylindrical geometries with known ground truth damage were created and used to test the developed workflow. In addition, the developed method was further validated on three real-world point cloud datasets using lidar and structure-from-motion techniques, which represented different underlying geometries and exhibited varying severity and mechanisms of damage. The analysis of the synthetic datasets demonstrated the robustness of the proposed damage detection method to classify vertices as surface damage with high recall and precision rates and a low false-positive rate. The real-world datasets illustrated the scalability of the damage detection method and its ability to classify areas as damaged and undamaged at the centimeter level. Moreover, the output classification of the damage detection method automatically bins the damaged vertices into different confidence intervals for further classification of detected likely damaged areas. Moving forward, the presented workflow can be used to bolster structural inspections by reducing subjectivity, enhancing reliability, and improving quantification in surface-evident damage.