Machine learning models have been shown to be useful for predicting and assessing structural performance, identifying structural condition and informing preemptive and recovery decisions by ...extracting patterns from data collected via various sources and media. This paper presents a review of the historical development and recent advances in the application of machine learning to the area of building structural design and performance assessment. To this end, an overview of machine learning theory and the most relevant algorithms is provided with the goal of identifying problems suitable for machine learning and the appropriate models to use. The machine learning applications in building structural design and performance assessment are then reviewed in four main categories: (1) predicting structural response and performance, (2) interpreting experimental data and formulating models to predict component-level structural properties, (3) information retrieval using images and written text and (4) recognizing patterns in structural health monitoring data. The challenges of bringing machine learning into structural engineering practice are identified, and future research opportunities are discussed.
•Provides formulation of machine learning (ML) algorithms that are relevant to building structural engineering.•Synthesizes the state of practice and research for ML applications in building structural engineering.•Discusses the challenges and opportunities in bringing ML applications into practice.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UILJ, UL, UM, UPCLJ, UPUK, ZAGLJ
•A framework for assessing the post-earthquake structural safety of damaged buildings is presented.•The concepts of response and damage patterns are introduced and incorporated into a systematic ...methodology integrating probabilistic seismic demand analysis, component-level damage simulation and robust assessments of the residual collapse capacity.•Machine learning algorithms are used to explicitly link the response and damage patterns to residual collapse capacity of a damaged structure, and are able to probabilistically predict the structural safety states given any available information.•A series of predictive models including Classification and Regression Trees and Random Forests are developed and examined in detail to achieve the optimal model which balance multiple performance measurements.•In contrast to previously judgement-based methods for the tagging process, this innovative approach provides a solid statistical support for structural safety assessment.•High prediction accuracies are observed based on either response and damage patterns.
A machine learning framework is presented to assess post-earthquake structural safety. The concepts of response and damage patterns are introduced and incorporated into a systematic methodology for generating a robust dataset for any damaged building. Incremental dynamic analysis using sequential ground motions is used to evaluate the residual collapse capacity of the damaged structure. Machine learning algorithms are used to map response and damage patterns to the structural safety state (safe or unsafe to occupy) of the building based on an acceptable threshold of residual collapse capacity. Predictive models including classification and regression tree and Random Forests are used to probabilistically identify the structural safety state of an earthquake-damaged building. The proposed framework is applied to a 4-story reinforced concrete special moment frame building. Distinct yet partially overlapping response and damage patterns are found for the damaged building classified as safe and unsafe. High prediction accuracies of 91% and 88% are achieved when the safety state is assessed using response and damage patterns respectively. The proposed framework could be used to rapidly evaluate whether a damaged building remains structurally safe to occupy after a seismic event and can be implemented as a subroutine in community resilience evaluation or building lifecycle performance assessment and optimization.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UL, UM, UPCLJ, UPUK
AbstractA framework is presented for incorporating probabilistic building performance limit states in the assessment of community resilience to earthquakes. The limit states are defined on the basis ...of their implications to postearthquake functionality and recovery. They include damage triggering inspection, occupiable damage with loss of functionality, unoccupiable damage, irreparable damage, and collapse. Fragility curves are developed linking earthquake ground motion intensity to the probability of exceedance for each of the limit states. A characteristic recovery path is defined for each limit state on the basis of discrete functioning states, the time spent within each state, and the level of functionality associated with each state. A building recovery function is computed accounting for the uncertainty in the occurrence of each recovery path and its associated limit state. The outcome is a probabilistic assessment of recovery of functionality at the building level for a given ground motion intensity. The effects of externalities and other socioeconomic factors on building-level recovery and ways to incorporate these in the framework are discussed. A case study is presented to demonstrate the application of the proposed framework to model the postearthquake recovery of the shelter-in-place housing capacity of an inventory of residential buildings. This type of assessment can inform planning and policy decisions to manage the earthquake risk to residential housing capacity of communities.
The ability to rapidly assess the spatial distribution and severity of building damage is essential to post-event emergency response and recovery. Visually identifying and classifying individual ...building damage requires significant time and personnel resources and can last for months after the event. This article evaluates the feasibility of using machine learning techniques such as discriminant analysis, k-nearest neighbors, decision trees, and random forests, to rapidly predict earthquake-induced building damage. Data from the 2014 South Napa earthquake are used for the study where building damage is classified based on the assigned Applied Technology Council (ATC)-20 tag (red, yellow, and green). Spectral acceleration at a period of 0.3 s, fault distance, and several building specific characteristics (e.g. age, floor area, presence of plan irregularity) are used as features or predictor variables for the machine learning models. A portion of the damage data from the Napa earthquake is used to obtain the forecast model, and the performance of each machine learning technique is evaluated using the remaining (test) data. It is noted that the random forest algorithm can accurately predict the assigned tags for 66% of the buildings in the test dataset.
Full text
Available for:
NUK, OILJ, SAZU, UKNU, UL, UM, UPUK
AbstractReliable analytical and empirical models of the force-deformation parameters used to characterize the nonlinear behavior of masonry panels are essential to simulating the seismic response of ...infilled reinforced concrete and steel frame systems. This paper presents the development of empirical equations to predict the backbone curve parameters of infill panels modeled using equivalent struts. For this purpose, a database of 264 infilled frame experiments is assembled from the existing literature. The experimental data from a subset of 113 specimens is used to calibrate the force-deformation parameters of the infill equivalent struts. Using the results from multivariate regression analyses, empirical equations are proposed for the backbone curve parameters that define the axial response of the infill struts. Discussions and recommendations for the cyclic degradation and pinching effect parameters are also presented.
The increase in seismic activity after a large-magnitude earthquake coupled with the reduction in the lateral load-carrying capacity of the affected structures presents a significant human and ...financial risk to communities. The focus of this paper is placed on quantifying the impact of both the elevated post-mainshock seismic hazard as well as the mainshock-induced structural damage on the seismic risk of three reinforced concrete moment frame structures. The seismic hazard due to sequential earthquakes is examined in both pre- and post-mainshock environments. The time-dependent nature of seismic hazard in the post-mainshock environment is accounted for through the adoption of a Markov risk assessment framework. In the post-mainshock environment, the seismic risk is examined as a function of the time elapsed since the mainshock’s occurrence while in the pre-mainshock environment, the risk is investigated during an assumed lifespan of 50 years for the studied structures. For the buildings and the high-seismicity site used in this study, both the increased post-mainshock seismic hazard as well as the reduction in the structural capacity are found to have a great influence on the seismic risk. The substantial contribution of aftershocks to the collapse risk in the pre-mainshock environment highlights the need for a design procedure that accounts for the additional seismic risk from aftershocks.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UL, UM, UPCLJ, UPUK
AbstractPerformance-based design optimization (PBDO) aims to design safe, resilient, and cost-effective structures. Methods used for PBDO have evolved by integrating numerical modeling, ...performance-based design principles, and optimization algorithms. The PBDO framework enables the design of structures with optimal performance and cost. This paper provides researchers with a comprehensive review of the rapidly growing field of PBDO. The evolution of PBDO methods is discussed with the goal of identifying the challenges that must be addressed in future studies. Knowledge gaps are brought to the forefront to emphasize the need for further investigation that expands on the application of PBDO in structural design. Various deterministic and probabilistic formulations for PBDO of structural systems under seismic and wind loading are reviewed. The formulations encompass one or more objective functions, including the upfront, life-cycle, and repair costs. Furthermore, the paper reviews retrofit design studies that have used PBDO methods. The high computational demand in performing PBDO is identified as the major challenge. Possible approaches to alleviate this and other challenges are discussed.
Abstract When solving the performance‐based earthquake engineering (PBEE) integral, the engineering demand parameters (EDPs) are assumed to be independent of the “upstream” parameters used in ground ...motion models (e.g., magnitude and source‐to‐site distance) after conditioning on the intensity measure (IM). This paper formulates a methodology for evaluating this conditional independence assumption through the lens of causal inference (CI). From a causal perspective, the effect of an upstream parameter on the EDP of interest is partially mediated by the IM with the remainder having a direct influence on the EDP. An IM is judged to be desirable (from a conditional independence standpoint) if the mediated effect of the upstream parameter is maximized, which implies that the direct effect is minimized. In the language of causal inference (CI), the IM, EDP and upstream parameters are described as the “treatment”, “outcome” and “confounding” variables, respectively. It then follows that the best performing IM is the one that maximizes the effect on the EDP after controlling for the upstream parameters. A semi‐parametric model that employs double machine learning is used to estimate the causal effect. The methodology is demonstrated through a case study application utilizing the responses from five special steel moment frames analyzed using 240 site‐agnostic ground motions. Compared to sufficiency‐based assessments, the casual inference method produces findings that are more explainable using structural dynamics principles and invariant to the number of ground motions used in the evaluation.
Full text
Available for:
BFBNIB, FZAB, GIS, IJS, KILJ, NLZOH, NUK, OILJ, SBCE, SBMB, UL, UM, UPUK
Abstract Shape memory alloys (SMAs) have found several applications in earthquake‐resilient structures. However, because of high material costs, their implementation on industry projects is still ...limited. Developing design approaches that minimize the use of expensive SMAs is critical to facilitating their widespread adoption in real structures. This paper proposes a performance‐based seismic design optimization procedure for self‐centering steel moment‐resisting frames (SC‐MRFs) with SMA‐bolted endplate connections. The topology optimization uses a metaheuristic algorithm to minimize the frame's total cost, including the initial construction and expected repair costs. The design variables are the steel beam and column sections, SMA connection properties, and the topology of the SMA connections. Different constraints are considered, such as the constructability of the chosen steel sections, member strengths, performance‐based design, Park‐Ang damage index, and strong‐column weak‐beam requirements. Furthermore, the seismic safety of optimal designs is assessed by calculating adjusted collapse margin ratios according to FEMA‐P695. An illustrative optimization study using three‐ and nine‐story SC‐MRFs is presented. The optimal SC‐MRFs are then assessed in terms of cost and seismic performance. The results confirm the effectiveness of the proposed optimum design, which minimizes the use of SMAs while ensuring improved seismic performance. The case studies show that the optimal placement of SMA connections can reduce the total cost by up to 71% and 61% for the three‐ and nine‐story SC‐MRFs, respectively, compared to nonoptimal frames. Moreover, the optimal SC‐MRFs exhibit more uniform drift distributions, lower residual story drifts by up to 96%, and increase collapse capacity by up to 102%.
Full text
Available for:
BFBNIB, FZAB, GIS, IJS, KILJ, NLZOH, NUK, OILJ, SBCE, SBMB, UL, UM, UPUK
•Pattern recognition used to assess the residual capacity of damaged tall buildings.•Incorporates feature filtering and selection for dimension reduction.•Response patterns mapped to the residual ...capacities with adequate performance.•Incorporates the use of reserved features with tolerable performance loss.
A pattern recognition approach is proposed to quantitatively assess the residual structural capacity of earthquake-damaged tall buildings. Sequential nonlinear response history analyses using as-recorded mainshock-aftershock ground motions are conducted to generate distinct feature patterns comprised of spatially distributed global and local engineering demand parameters (EDP) within the tall building. Residual structural capacity is assessed based on the median spectral intensity corresponding to the collapse prevention performance level. Dispersion-based filtering and feature selection using Least Absolute Shrinkage and Selector Operator (LASSO) are performed to effectively reduce the high dimensional feature space while selecting the most informative ones. The features that survive the filtering but excluded by LASSO are reserved and grouped based on their correlations with those that are selected. These reserved features can be utilized if the selected ones are unavailable. Predictive models using Support Vector Machine are constructed to map the EDP-based features to the residual structural capacity of the tall building, where satisfactory performance is observed as measured by the root mean square errors in the testing dataset. In addition to guiding post-earthquake inspections and residual structural capacity assessments, the proposed framework can inform optimal sensor placement as well as provide time-dependent limit state evaluation in aftershock environments.
Full text
Available for:
GEOZS, IJS, IMTLJ, KILJ, KISLJ, NLZOH, NUK, OILJ, PNG, SAZU, SBCE, SBJE, UL, UM, UPCLJ, UPUK