Professor Marios Chryssanthopoulos

As building structures in different parts of the world become increasingly exposed to extreme events, there has been a notable research and professional effort to ensure the design of more robust buildings which are insensitive to local failures. At the same time, several works performed in the field of forensic structural engineering have contributed to advancing knowledge on causes and risk factors of structural failure. This includes the creation of several collapse databases, most of which focus mainly on the underlying hazards causing failure. While such databases have provided invaluable insights for preventing structural failures, they do not lend themselves well to the analysis of how failure propagates, which can be useful for improving the progressive collapse resistance of buildings. To this end, a novel database of building collapses is presented in this article which systematically collects information on the hazards, initial failures and their corresponding propagation mechanisms. In addition, key information related to the context and the consequences of collapse is also gathered. Based on the information compiled in the database, this article provides an in-depth analysis of the most commonly occurring initial failures and propagation mechanisms, with significant conclusions extracted from the study of past collapses. The application of different consequence models for estimating fatalities and reconstruction costs is also presented, leading to recommendations for improving such models and related data collection strategies. •Novel database of building collapses.•Systematic collection of information on initial failures and propagation mechanisms.•Analysis of most commonly occurring hazards, initial failures, and propagation phenomena.•Application and evaluation of consequence models for estimating casualties.•Estimation of reconstruction and replacement costs.

Attempts to formalize inspection and monitoring strategies in industry have struggled to combine evidence from multiple sources (including subject matter expertise) in a mathematically coherent way. The perceived requirement for large amounts of data are often cited as the reason that quantitative risk-based inspection is incompatible with the sparse and imperfect information that is typically available to structural integrity engineers. Current industrial guidance is also limited in its methods of distinguishing quality of inspections, as this is typically based on simplified (qualitative) heuristics. In this paper, Bayesian multi-level (partial pooling) models are proposed as a flexible and transparent method of combining imperfect and incomplete information, to support decision-making regarding the integrity management of in-service structures. This work builds on the established theoretical framework for computing the expected value of information, by allowing for partial pooling between inspection measurements (or groups of measurements). This method is demonstrated for a simulated example of a structure with active corrosion in multiple locations, which acknowledges that the data will be associated with some precision, bias, and reliability. Quantifying the extent to which an inspection of one location can reduce uncertainty in damage models at remote locations has been shown to influence many aspects of the expected value of an inspection. These results are considered in the context of the current challenges in risk based structural integrity management.

•Surface roughness of corroded wires, subjected to an accelerated corrosion method, was measured using a 3D laser scanner.•Spatial characteristics of area loss and pit depth were quantified on nominally identical corroded specimens.•Different corrosion indicators were specified in order to explore correlations with fatigue performance.•The maximum area loss was better correlated, compared to maximum pit depth, with the breakage position.•Effect of corrosion was quantified through the estimation of the coefficients in the S-N relationships. This study investigates the fatigue performance of corroded bridge wires by providing characterisation of the corrosion effects through non-contact surface mapping and by undertaking fatigue tests under constant amplitude loading. For the corrosion level investigated, the breakage position in corroded wires was found to be better correlated with maximum area loss rather than pit depth. Using also test results for un-corroded wires, S-N relationships were determined for both that set and the corroded wire set created in this study. The effect of corrosion was quantified through the estimation of the coefficients for both the mean and the design S-N curves.

Engineers perform fatigue assessments to support structural integrity management. Given that the purpose of these calculations is linked to problems of decision making under various sources of uncertainty, probabilistic methods are often more useful than deterministic alternatives. Guidance on the direct probabilistic application of procedures in existing industrial standards is currently limited and dependencies between marginal probabilistic models are generally not considered, despite their potential significance being acknowledged. This paper proposes the use of Bayesian data analysis as a flexible and intuitive approach to coherently and consistently account for uncertainty and dependency in fatigue crack growth rate models. Various Bayesian models are established and presented, based on the same data as the existing models in BS 7910 (a widely used industrial standard). The models are compared in terms of their out of sample predictive accuracy, using methods with a basis in information theory and cross-validation. The Bayesian models exhibit an improved performance, with the most accurate predictions resulting from multi-level (hierarchical) models, which account for variation between constituent test datasets and partially pool information.

Although the majority of creep models are comprehensive and up-to-date, there is a lack of consensus in their utilisation due to substantial scatter in their predictions, even when comparisons are made under well-controlled conditions. On one hand, creep entails complex phenomena that depend on several factors and, on the other hand, these models are typically utilised on a deterministic basis without fully incorporating information related to random input variability. In this paper, a methodology is proposed, based on Bayesian updating methods, for creep deformation prediction by combining prior model distributions obtained through Monte Carlo simulation with in-situ measurements obtained from concrete specimens. Both single point-in-time and sequential updating approaches are formulated and contrasted in the context of site data collected over a period of about six years. For the specific structure examined, the sequential updating method offers advantages in terms of the estimated variability of future predictions. The proposed methodology is suitable for quantifying the value of monitoring information, as demonstrated by considering the change in prediction variability against the length of observation period.

The offshore wind turbines (OWTs) are dynamically sensitive, whose fundamental frequency can be very close to the forcing frequencies activated by the environmental and turbine loads. Minor changes of support conditions may lead to the shift of natural frequencies, and this could be disastrous if resonance happens. To monitor the support conditions and thus to enhance the safety of OWTs, a model updating method is developed in this study. A hybrid sensing system was fabricated and set up in the laboratory to investigate the long-term dynamic behaviour of the OWT system with monopile foundation in sandy deposits. A finite element (FE) model was constructed to simulate structural behaviours of the OWT system. Distributed nonlinear springs and a roller boundary condition are used to model the soil-structure-interaction (SSI) properties. The FE model and the test results were used to analyze the variation of the support condition of the monopile, through an FE model updating process using Estimation of Distribution Algorithms (EDAs). The results show that the fundamental frequency of the test model increases after a period under cyclic loading, which is attributed to the compaction of the surrounding sand instead of local damage of the structure. The hybrid sensing system is reliable to detect both the acceleration and strain responses of the OWT model and can be potentially applied to the remote monitoring of real OWTs. The EDAs based model updating technique is demonstrated to be successful for the support condition monitoring of the OWT system, which is potentially useful for other model updating and condition monitoring applications.

A novel methodology is presented for probabilistic fatigue life prediction of welded joints in orthotropic bridge steel decks. Monitoring data were used to specify time-series model parameters for the main drivers of fatigue damage in such structures, namely pavement temperatures and heavy traffic intensities, which influence the stress range distributions at critical locations. Polynomial regression models were developed to quantify the relationship between fatigue loading, derived using S-N principles from strain measurements at welded joints, with pavement temperatures and heavy traffic counts. The different models were integrated within a fatigue reliability framework, in which the uncertainties arising from material properties and fatigue damage at failure were modelled via random variables. A Monte Carlo scheme was then deployed to predict S-N fatigue damage using the fatigue loading regression models and simulated time-series of heavy traffic and pavement temperatures. Thus, fatigue reliability profiles were generated, which account for different scenarios in terms of future changes in traffic and pavement temperature. The proposed methodology was illustrated considering actual monitoring outcomes from the Great Belt Bridge (Denmark) with reliability profiles developed for both ‘baseline’ and ‘adverse’ scenarios in the context of asset integrity management. The combined effect of higher temperature and heavy traffic levels was shown to result in considerable reductions in fatigue reliability, with a commonly used threshold being reached up to 40 years earlier compared to the baseline ‘no change’ scenario. However, this reduction was not uniform for all the fatigue details considered, emphasizing the importance of monitoring different locations, based on a thorough understanding of the fatigue behaviour of the orthotropic steel deck.

An increasing number of tubular steel structures have exceeded their design service lives; hence, monitoring of these structures is crucial in preventing any unforeseen failures and corresponding catastrophic consequences - safety or economic. As is well known, vibration-based structural health monitoring (SHM) presents non-destructive methods for damage identification, though their application in corrosion problems appears somewhat limited. Furthermore, majority of the SHM techniques reported in literature deal with prismatic or beam-like members; tubular structures have received less research attention. In this paper, numerical models of a pipe in its intact and corroded conditions are built and analysed using ABAQUS®. Modal parameters extracted from analyses results are utilised to detect, locate and quantify corrosion. A potential indicator “Normalised Displacement Modeshape (NDM)” is introduced and tested alongside existing damage indicators. Results from the employed indicators are compared and the capabilities of each indicator in identification of the investigated corrosion patterns are discussed.

The progressive collapse of the World Trade Centre showed the devastating consequences of pancake-type collapses, triggering significant research on failed floor impacts for different forms of construction. In current high-rise construction, concrete flat slabs supported on columns are widely used and, in this case, the fall of the slab could be prevented depending on the detailing and the horizontal propagation of the collapse from one support to adjacent supports. The activation and interaction of different phenomena during horizontal propagation governed by the slab-column response is investigated in this study for cases of flat slabs with and without integrity reinforcement. The paper focuses on slabs without punching reinforcement which are more critical. A slab system analytical model is presented based on a column removal scenario considering the dynamic response of the column-slab connections before and after punching including membrane effects. The model was verified using finite element models with solid elements at the connections. The results highlighted the key role of integrity reinforcement towards preventing slabs from falling by means of activating tensile membrane action concentrated around the columns.

Vibration-based condition identification of bolted connections can benefit the effective maintenance and operation of steel structures. Existing studies show that modal parameters are not sensitive to such damage as loss of preload. In contrast, structural responses in the time domain contain all the information regarding a structural system. Therefore, this study aims to exploit time-domain data directly for condition identification of bolted connection. Finite element (FE) model updating is carried out based on the vibration test data of a steel frame, with various combinations of bolts with loss of preload, representing different damage scenarios. It is shown that the match between the numerically simulated and measured acceleration responses of the steel frame cannot be achieved. The reason is that time-dependent nonlinearity is generated in bolted connections during dynamic excitation of the steel frame. To capture the nonlinearity, a virtual viscous damper is proposed. By using the proposed damper alongside the updated system matrices of the FE model, the time domain acceleration responses are estimated with great consistency with the measured responses. The results demonstrate that the proposed virtual damper is not only effective in estimating the time domain acceleration responses in each damage case, but also has the potential for condition identification of bolted connections with such small damage as just one bolt with loss of preload. It can also be applied to other challenging scenarios of condition identification, where modal parameters are not sensitive to the damage.

Tubular structures are widely used in infrastructure, especially for pipes and marine installations. Corrosion is a key damage type for these structures, and if undetected may cause structural deficiency or even collapse. This paper investigates the effects of localised corrosion on the vibration properties of axisymmetric tubular structures and proposes a novel baseline-free damage identification method to overcome the fact that the difference between undamaged and damaged properties is usually small and contaminated by noise. The proposed vibration-based method exploits the phenomenon of repeated mode pairs which is peculiar to axisymmetric tubular structures with non-axisymmetric damage. Unlike intact tubular structures, it is found that each vibration mode of a locally damaged tubular structure comprises of two components: active and passive. The active component is sensitive to the local damage and can be used to infer damage existence while the passive component is retentive of the undamaged pipe's properties. Therefore, the passive component of a damaged pipe’s local vibration mode pair (LVMP) can be used to predict its baseline (undamaged) modal properties, which in this research was achieved within a margin of ±2%, whereas the active component confirms damage existence. The work also establishes the criteria for validation of LVMPs through experimental tests involving various corrosion scenarios ranging from longitudinal strips to through-thickness pits. Based on the obtained results, it is proposed that the properties of LVMPs can be linked to corrosion-like local damage in axisymmetric tubular structures, underpinning a new damage detection and identification approach.

Failure assessment diagrams are an integral component of asset integrity management in a variety of industrial sectors. They allow for the assessment of the significance of cracks in structures, between the domains of brittle fracture and plastic collapse. Numerical modelling, or empirically determined assessment lines across this continuum, are the basis of the guidance in current industrial standards. However, as the requirement of such assessments progresses from demonstrating safety to optimising resilience and resource allocation, it will become necessary to compute more informative (probabilistic) estimates, which are compatible with decision analysis. In this paper, Bayesian regression models are proposed as a suitable method of quantifying (aleatory and epistemic) uncertainty in the limit state on a failure assessment diagram. The data considered in this study consists of laboratory tests completed on wide plate fracture specimens. This work is intended to address the inconsistencies in current editions of industrial standards, and limitations (regarding flexibility and application) of existing scientific literature on the topic. Potential applications are discussed, including the use of fracture mechanics in meaningful reliability analysis, quantitative risk management, and optimising experimental design for future material tests. •A probabilistic failure assessment diagram is compared to methods used in industry.•The proposed gaussian process model is flexible, non-linear and probabilistic.•It can therefore approximate the behaviour between brittle and plastic domains.•Applications of reliability analysis and experimental design are demonstrated.

Infrastructure performance is of great importance for a nation’s economy and its people’s quality of life. For efficient and effective infrastructure asset management, structural health monitoring (SHM) has been researched extensively in the past 20-30 years. With an increasing number of SHM systems being installed, the interpretation of the large volume of monitoring data, i.e. often manifested as condition identification, becomes essential in asset integrity management. This paper provides an appraisal of existing literature reviews on SHM, considering both reviews on different types of structures and those focused on different approaches for data interpretation. It explores the evolution of research interests in this field and identifies the need for an integrated physics-based and data-driven structural condition identification approach.

The most widely used design approaches today for improving the robustness of buildings rely on improving continuity within the structural system to ensure that loads supported by failed components can be redistributed to the rest of the system. Although this is effective for small initial failures, it can increase the risk of disproportionate collapse after larger initial failures due to collapsing elements pulling down parts of the structure that would otherwise be unaffected. This form of continuity-enabled collapse propagation can be avoided by dividing a structure into different segments. However, completely separating parts of a building results in lower performance under operational conditions, against lateral loads, and after small initial failures. In fact, the advantages of both continuity and segmentation can be combined through a fuse-based segmentation approach in which predefined segment borders ensure connectivity after small initial failures but separate to isolate collapse after larger initial failures. To ensure that this approach is used effectively to improve the robustness of building structures, a design framework is proposed in this article to systematically consider relevant structural and geometric criteria in order to define suitable segmentation configurations for reinforced concrete and steel framed building structures. An application to a realistic case study is also presented to demonstrate the effectiveness of the proposed framework in enhancing structural robustness. •Continuity is advantageous after small initial failures•Segmentation is advantageous after large initial failures•Fuse-based segmentation combines the advantages of continuity and segmentation•Framework is proposed to define segmentation configurations that enhance robustness•Obtained configurations can be used in risk-based assessments

Structural Health Monitoring Systems (SHMS) are increasingly present in most modern long-span bridges. Those systems can be used to better assess the performance of structures by reducing the uncertainties associated with deterioration modelling. This can potentially lead to a reduction of the operational costs. Despite their promise and potential, a gap still remains between the outcomes of those systems and practical bridge management decisions. As a result, huge amounts of data can be continuously collected which are not readily useable, thus being of reduced interest in practical terms. Methodologies which integrate SHMS within Bridge Management Systems (BMS), need to be developed to address asset management issues traditionally informed by visual inspections and scarce Non Destructive Tests (NDT). This integration should overcome the shortcomings of current approaches and exploit the advantages offered by modern sensor technologies. The present paper reviews the different uses of SHMS on long-span bridges. The motivation of using SHMS to inform and improve bridge management decisions is presented. The interest of a local monitoring approach targeting selected structural components is highlighted. The need of a combined approach between traditional inspection techniques, NDT and monitoring is justified in terms of spatial and temporal coverage. The relevance of probabilistic approaches to assess and update structural performance indicators is outlined. The case of the Great Belt Bridge (Denmark) is described to illustrate the use of SHMS on long-span bridges, together with an overview of ongoing research.

A vector-valued intensity measure is presented, which incorporates a relative measure represented by the normalized spectral area. The proposed intensity measure is intended to have high correlation with specific relative engineering demand parameters, which collectively can provide information regarding the damage state and collapse potential of the structure. Extensive dynamic analyses are carried out on a single-degreeof- freedom system with a modified Clough–Johnston hysteresis model, using a dataset of 40 ground motions, in order to investigate the proposed intensity measure characteristics. Response is expressed using the displacement ductility, and the normalized hysteretic energy, both of which are relative engineering demand parameters. Through regression analysis the correlation between the proposed intensity measure and the engineering demand parameters is evaluated. Its domain of applicability is investigated through parametric analysis, by varying the period and the strain-hardening stiffness. Desirable characteristics such as efficiency, sufficiency, and statistical independence are examined. The proposed intensity measure is contrasted to another one, with respect to its correlation to the engineering demand parameters. An approximate procedure for estimating the optimum normalized spectral area is also presented. It is demonstrated that the proposed intensity measure can be used in intensity-based assessments, and, with proper selection of ground motions, in scenario-based assessments.

A modelling platform based on regression analysis is developed as a novel approach to structural health monitoring of welded joints of orthotropic bridge steel decks. Monitoring outcomes from the Great Belt Bridge (Denmark) are used to develop regression models following a weighted least squares approach to characterize the normal correlation pattern between environmental conditions (daily-averaged pavement temperatures), operational loads (daily-aggregated heavy traffic counts) and a strain-based performance indicator. The developed models can be used within a structural health monitoring–based asset management framework for performance assessment (i.e. diagnosis of structural performance changes) and performance prediction (i.e. prognosis of structural performance leading to service life estimates). The main novelty of the work presented consists of the development of an algorithm based on statistical control charts related to the prediction bands of the regression models. The algorithm enables the interpretation of new monitoring data and the identification of potentially abnormal behaviours via outlier detection, as part of an envisaged ‘real-time’ performance assessment application. The proposed approach to outlier detection through structural health monitoring is finally illustrated considering actual monitoring outcomes from the bridge. This highlights the applicability of the developed modelling platform and contributes to bridging the gap between monitoring data and monitoring-based information that can lead to more effective asset management decisions.

All engineering structures degrade or become damaged in service to some extent. Information collection activities, such as inspection or structural health monitoring can reduce uncertainty in probabilistic models of structural condition. By linking the information that they provide to the improved integrity management strategies that they facilitate, their expected value can be quantified. This value of information can be obtained using Bayesian decision analysis. In this work an extended value of information model is presented that accounts for the risk associated with exposure to a hazardous environment. By evaluating this risk on the same scale as the risk of structural failure, the relationship between the expected quality of information and the number of staff-hours in a hazardous environment (such as an offshore oil and gas platform) is investigated. An example case study identifies the requirements regarding the precision, bias, and reliability of information from autonomous or remote inspection methods, for them to be considered as an optimal risk management strategy.

Engineers use semi-empirical models of complex degradation phenomena to manage the integrity of structural systems. Historically, the intended application of these models has been to perform deterministic calculations to demonstrate that failure is not expected. However, as integrity management systems are increasingly optimised decision makers require quantified likelihoods of events such as failure to help ensure that maintenance investments are completed where and when they are worthwhile. In the probabilistic application of degradation models, inherent and un-quantified conservatism can lead to ineffective risk management. Bayesian analysis can be used to fit semi-empirical degradation models to existing test data (and can be updated if new data is available) such that the variability within, and the inter-dependency between all model parameters is quantified. This includes model uncertainty parameters. In this paper, an SN curve and a two-stage Paris Law model have been fitted using Markov Chain Monte Carlo sampling. The benefits and challenges associated with this approach are discussed in the context of existing standards and guidelines developed for industry. Posterior predictive sampling is used to demonstrate how the models can be used to produce results that are fully compatible with a Bayesian decision analysis.

In pipelines, pressure vessels and various other steel structures, the remaining thickness of a corroding ligament can be measured directly and repeatedly over time. Statistical analysis of these measurements is a common approach for estimating the rate of corrosion growth, where the uncertainties associated with the inspection activity are taken into account. An additional source of variability in such calculations is the epistemic uncertainty associated with the limited number of measurements that are available to engineers at any point in time. Traditional methods face challenges in fitting models to limited or missing datasets. In such cases, deterministic upper bound values, as recommended in industrial guidance, are sometimes assumed for the purpose of integrity management planning. In this paper, Bayesian inference is proposed as a means for representing available information in consistency with evidence. This, in turn, facilitates decision support in the context of risk-informed integrity management. Aggregating inspection data from multiple locations does not account for the possible variability between the locations, and creating fully independent models can result in excessive levels of uncertainty at locations with limited data. Engineers intuitively acknowledge that the areas with more sites of corrosion should, to some extent, inform estimates of growth rates in other locations. Bayesian multi-level (hierarchical) models provide a mathematical basis for achieving this by means of the appropriate pooling of information, based on the homogeneity of the data. Included in this paper is an outline of the process of fitting a Bayesian multi-level model and a discussion of the benefits and challenges of pooling inspection data between distinct locations, using example calculations and simulated data.