No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Damage detection in RCC beam with ANNs
Abstract
Accumulation of damages during the service life of a structure can reduce its safety. Every structure that is constructed has a particular age but these structures can deteriorate before their service life due to various factors such as harsh environmental conditions, fatigue due to service loading, etc. To access the information regarding the health index of structure the need for various unconventional damage assessment practices and dependable structural health monitoring systems is presently high. Structures to perform efficiently damage assessment and appropriate retrofitting are required. Structural health monitoring (SHM) has been verified to be an economical technique for damage assessment in structures over the past several decades. In reinforced concrete beams flexural cracks distribute non-linearly and propagate along with all directions. The crack continues to propagate until the structure or structural component fractures. Due to this complex behavior of cracks, simplified damage simulation techniques such as reductions in the modulus of elasticity or section depth or stiffness of rotational spring elements cannot be applied to simulate flexural cracks in reinforced concrete components. Besides these simplified techniques, dynamic properties have been used extensively in the past. But this methodology has many disadvantages such as dynamic characteristics obtained through experiments can vary due to measurement errors, noise, and environmental changes, which can greatly affect precision. This research will address the above gap in knowledge by developing a model that can represent the complex behavior of cracks and then utilize artificial neural networks to assess damage in RC flexural members. In this study, consideration is given to the fundamental strategy for developing Plasticity Damage Models and ANNs to predict the extent and location of the damage from beam structures' measured detection data. ABAQUS finite element software is used with properly validated models of material throughout the study. A promising approach to evaluate and detect damage makes use of artificial neural networks (ANNs) to solve these two problems. ANNs are a strong artificial intelligence (AI) technique that has been widely accepted in predicting the extent and location of structural damage. ANNs are trained using deflection data obtained from intact and damaged beam structures simulations of finite elements. It is also found that reducing the number of required outputs significantly improved the quality of predictions made by ANN. The findings obtained were fair and they demonstrated strong alignment with the real values. This means that using ANNs is an excellent method for measuring damage and the issue of detecting damage. When using ANNs, some essential problems of traditional approaches to detecting damage can be solved and the precision of identification of damage can be improved.
ResearchGate has not been able to resolve any citations for this publication.
Any structure in presence of crack is susceptible to failure depending on the mode of vibration. Failure is due to the resonance formed by the superposition of frequency of periodic force acting on structure and the natural frequency of the structure. To be alert about resonance due to periodic load, it is important to determine natural frequency. In this study of modal analysis, natural frequency and mode shapes of transverse vibration for both un-cracked and cracked cantilever beam has been extracted for first three modes. The analysis has been extended to investigate the effect of crack opening size and mesh refinement. For cracked beam, analysis is performed for various crack depth and crack location. As structural discontinuity problems are difficult to solve analytically, leading commercial Finite Element Analysis software - “Abaqus” is used to perform all the analysis computationally. For modeling un-cracked beam, hexahedral element is used whereas for cracked beam modeling, both hexahedral and wedge elements are selected for better result. In our study we observed that natural frequency reduces with the presence of crack. The amount of reduction varies depending on crack location, depth and crack opening size. Non-dimensional representation highlights that failure criteria largely depend on mode of vibration. The findings of this study can be applied to predict the structural sustainability under varying loads.
REINFORCED CONCRETE (RC) BEAM-COLUMN CONNECTIONS ESPECIALLY THOSE WITHOUT TRANSVERSE REINFORCEMENT IN JOINT REGION CAN EXHIBIT BRITTLE BEHAVIOR WHEN INTENSIVE DAMAGE IS CONCENTRATED IN THE JOINT REGION DURING AN EARTHQUAKE EVENT. BRITTLE BEHAVIOR IN THE JOINT REGION CAN COMPROMISE THE DUCTILE DESIGN PHILOSOPHY AND THE EXPECTED OVERALL PERFORMANCE OF STRUCTURE WHEN SUBJECTED TO SEIS-MIC LOADING. CONSIDERING THE IMPORTANCE OF JOINT SHEAR FAILURE INFLUENCES ON STRENGTH, DUCTILITY AND STABILITY OF RC MOMENT RESISTING FRAMES, A FINITE ELEMENT MODELING WHICH FOCUSES ON JOINT SHEAR BEHAVIOR IS PRESENTED IN THIS ARTICLE. NONLINEAR FINITE ELEMENT ANALYSIS (FEA) OF RC BEAM-COLUMN CONNECTIONS IS PERFORMED IN ORDER TO INVESTIGATE THE JOINT SHEAR FAILURE MODE IN TERMS OF JOINT SHEAR CAPACITY, DEFORMATIONS AND CRACKING PATTERN. A 3D FINITE ELEMENT MODEL CAPABLE OF APPROPRIATELY MODELING THE CONCRETE STRESS-STRAIN BEHAVIOR, TENSILE CRACKING AND COMPRESSIVE DAMAGE OF CONCRETE AND INDIRECT MODELING OF STEEL-CONCRETE BOND IS USED. IN ORDER TO DEFINE NONLINEAR BEHAVIOR OF CONCRETE MATERIAL, THE CON-CRETE DAMAGE PLASTICITY IS APPLIED TO THE NUMERICAL MODEL AS A DISTRIBUTED PLASTICITY OVER THE WHOLE GEOMETRY. FINITE ELEMENT MODEL IS THEN VERIFIED AGAINST EXPERIMENTAL RESULTS OF TWO NON-DUCTILE BEAM-COLUMN CONNECTIONS (ONE EXTERIOR AND ONE INTERIOR) WHICH ARE VULNERABLE TO JOINT SHEAR FAILURE. THE COMPARISON BE-TWEEN EXPERIMENTAL AND NUMERICAL RESULTS INDICATES THAT THE FE MODEL IS ABLE TO SIMULATE THE PERFORMANCE OF THE BEAM-COLUMN CONNECTIONS AND IS ABLE TO CAPTURE THE JOINT SHEAR FAILURE IN RC BEAM-COLUMN CONNECTIONS.
The past several years have witnessed an increase in research on the nonlinear analysis of the structures made from reinforced concrete. Several mathematical models were created to analyze the behavior of concrete and the reinforcements. Factors including inelasticity, time dependence, cracking and the interactive effects between reinforcement and concrete were considered. The crushing of the concrete in compression and the cracking of the concrete in tension
are the two common failure modes of concrete. Material models were
introduced for analyzing the behavior of unconfined concrete, and a possible constitutive model was the concrete damage plasticity (CDP) model. Due to the complexity of the CDP theory, the procedure was simplified and a simplified concrete damage plasticity (SCDP) model was developed in this paper. The SCDP model was further characterized in tabular forms to simulate the behavior of unconfined concrete. The parameters of the concrete damage plasticity model, including a damage parameter, strain hardening/softening rules,
and certain other elements, were presented through the tables shown in the paper for concrete grades B20, B30, B40 and B50. All the aspects were discussed in relation to the effective application of a finite element method in the analysis. Finally, a simply supported prestressed beam was analyzed with respect to four different concrete grades through the finite element program.
The results showed that the proposed model had good correlation with prior arts and empirical formulations.
The nonlinear analysis of a reinforced concrete beam was conducted based on the finite element analysis software ABAQUS. In this simply supported beam analysis, the plasticity model of concrete damage in ABAQUS has been introduced thoroughly. Finally, the results of the experimentation and the ABAQUS analysis were compared in a diagram, accordingly reasons of the result difference between the two methods were discussed, which can be a useful reference for the further study of the nonlinear analysis of reinforced concrete.
This study aims to investigate the capacity of simplified and advanced numerical models to reproduce the nonlinear monotonic behavior of reinforced concrete framed structures. The considered simplified models are commonly used in earthquake engineering and are based either on concentrated or distributed plasticity. The considered advanced models are based on continuum mechanics and use the Damaged Plasticity Model for concrete simulation. The simplified and advanced models are used to simulate pushing experiments of a 2D RC frame. The results from the models are compared with the experiments results. That comparison shows the higher ability of the continuum mechanics models to represent the actual behavior of the frame until the collapse; the simplified models lack of the capacity to represent accurately all the phases of the behavior. This paper includes an explanation of the damage plasticity models and their implementation in the FEM package ABAQUS. A strategy to avoid the mesh sensitivity in the advanced numerical simulation of reinforced concrete structures is also described.
Bending test of seven reinforced concrete beams are modeled in finite element program to validate the modeling strategies by comparing the structural response of the beams. Three beams in the set are pre-damaged and strengthened with fiber reinforced composites before the bending tests. Cracks are implemented into the model by inserting geometrical discontinuities to represent the pre-damaged beams. Parametric variables such as crack width, length and interval are chosen to simulate different pre-damage levels. Once the proposed modeling strategies are validated by real experimental tests then 196 finite element models are created to study the effects of pre-damage levels on the moment capacity of reinforced concrete beams repaired with CFRP. Results indicate that inclusion of pre-damage levels by means of cracks into the cross sections have significant effect on beams moment capacity.
A technique to localize damage in structures that can be treated as linear in the pre and post-damage state is presented. Central to the approach is the computation of a set of vectors, designated as Damage Locating Vectors (DLVs) that have the property of inducing stress fields whose magnitude is zero in the damaged elements (small in the presence of truncations and approximations). The DLVs are associated with sensor coordinates and are computed systematically as the null space of the change in measured flexibility. The approach is not structure-type dependent and can be applied to single or multiple element damage scenarios. Knowledge about the system is restricted to that needed for a static analysis in the undamaged state, namely, the undamaged topology and, if the structure is indeterminate, the relative stiffness characteristics. Results from numerical simulations suggest that the method can operate successfully under realistic conditions.
Inverse methods are often central to many approaches to structural health monitoring, for example the detection and location of damage in bridge structures from measured vibration data. This paper reviews the general principals of inverse problems useful in damage assessment. Of particular importance are model based parameter estimation and damage localisation. Models of structures and their parameterisation are vital for the successful application of inverse problems. The problems of ill-conditioning and the use of subset selection is highlighted. A simple example is used to demonstrate the benefits of subset selection and the effect of measurement noise and modelling errors.
A new two-step approach for probabilistic structural health monitoring is presented, which involves modal identification followed by damage assessment using the pre- and post-damage modal parameters based on a new Bayesian model updating algorithm. The new approach aims to attack the structural health monitoring problems with incomplete modeshape information by including the underlying full modeshapes of the system as extra random variables, and by employing the Expectation-Maximisation algorithm to determine the most probable parameter values. The non-concave non-linear optimisation problem associated with incomplete modeshape cases is converted into two coupled quadratic optimisation problems, so that the computation becomes simpler and more robust. We illustrate the new approach by analysing the Phase II Simulated Benchmark problems sponsored by the IASC-ASCE Task Group on Structural Health Monitoring. The results of the analysis show that the probabilistic approach is able to successfully detect and locate the simulated damage involving stiffness loss in the braces of the analytical benchmark model based on simulated ambient-vibration data.
Modal-based damage-detection algorithms were used to identify the location of defects commonly found in timber and to estimate
their severities. In this study, the authors propose modifications to an existing damage-detection algorithm for locating
and evaluating damage by comparing the modal strain energy before and after damage using the first two flexural modes of vibration.
Experimental verification was performed on pin-pin supported timber beams by employing the algorithms with extracted modal
parameters using experimental modal analysis. Single and multiple cases of damage used to simulate pocket(s) of rot with various
severities were inflicted by removing sections of timber beam specimens. The proposed damage indicator, computed from the
first two flexural modes, was capable of detecting all damage locations. It was also able to estimate, with reasonable accuracy,
the severity of damage in term of loss of sectional moment of inertia. The modified damage index method is generally reliable
in detecting the location and estimating the severity of simulated defects in timber beams.
The present paper reports the results obtained by a methodology of damage detection in which combined EMA and FEM data are
used to localise the damage on mechanical components. The utilized method uses the eigenvalues and eigenvectors obtained from
FE modelling, compared with the eigenvalues obtained experimentally on damaged specimens. The method assumes a linear behaviour
of the materials. Firstly it is applied to rectangular plates in order to test its reliability in discovering the damage located
on very simply shaped specimens that are made of homogeneous and isotropic material. The methodology was subsequently applied
to mechanical components of complex shape, allowing the location of the damage to be accurately identified.
A practical method to non-destructively locate and estimate size of a crack by using changes in natural frequencies of a structure is presented. First, a crack detection algorithm to locate and size cracks in beam-type structures using a few natural frequencies is outlined. A crack location model and a crack size model are formulated by relating fractional changes in modal energy to changes in natural frequencies due to damage such as cracks or other geometrical changes. Next, the feasibility and practicality of the crack detection scheme are evaluated for several damage scenarios by locating and sizing cracks in test beams for which a few natural frequencies are available. By applying the approach to the test beams, it is observed that crack can be confidently located with a relatively small localization error. It is also observed that crack size can be estimated with a relatively small size error.
Dynamic modal parameters are appealing features for a prompt diagnosis of structural conditions, since they are relatively simple to measure and utilize. The feasibility of using them for damage detection, however, is limited when their changes go undisclosed due to uncertain temperature conditions, particularly for large structures. In this paper, a vibration-based damage monitoring scheme to give warning of the occurrence, the location, and the severity of damage under temperature-induced uncertainty conditions is proposed. Firstly, experiments on a model plate-girder bridge, for which a set of modal parameters were measured under uncertain temperature conditions, are described. Secondly, a damage warning model is selected to statistically identify the occurrence of damage, by recognizing the patterns of damage-driven changes in natural frequencies of the test structure, and by distinguishing temperature-induced off-limits. Thirdly, a frequency-based damage index method based on the concept of modal strain energy is implemented into the test structure to predict the location and the severity of damage. In order to adjust the temperature-induced changes in natural frequencies that are used for damage detection, a set of empirical frequency correction formulae are derived from the relationship between temperature and frequency ratio.
In this paper, a newly derived algorithm to predict locations and severities of damage in structures using changes in modal characteristics is presented. First, two existing algorithms of damage detection are reviewed and the new algorithm is formulated in order to improve the accuracy of damage localization and severity estimation by eliminating erratic assumptions and limits in the existing algorithms. Next, the damage prediction accuracy is numerically assessed for each algorithm when applied to a two-span continuous beam for which pre- and post-damage modal parameters are available for only a few modes of vibration. Compared to the existing damage detection algorithms, the new algorithm improved the accuracy of damage localization and severity estimation results in the test beam.
A conventional method of damage modeling by a reduction in stiffness is insufficient to model the complex non-linear damage characteristics of concrete material accurately. In this research, the concrete damage plasticity constitutive model is used to develop the numerical model of a deck beam on a berthing jetty in the Abaqus finite element package. The model constitutes a solid section of 3D hexahedral brick elements for concrete material embedded with 2D quadrilateral surface elements as reinforcements. The model was validated against experimental results of a beam of comparable dimensions in a cited literature. The validated beam model is then used in a three-point load test configuration to demonstrate its applicability for preliminary numerical evaluation of damage detection strategy in marine concrete structural health monitoring. The natural frequency was identified to detect the presence of damage and mode shape curvature was found sensitive to the location of damage.
Currently, visual inspection is performed in order to evaluate damage in structures. This approach is affected by the constraints of time and the availability of qualified personnel. Thus, new approaches to damage identification that provide faster and more accurate results are pursued. A promising approach to damage evaluation and detection utilizes artificial neural networks (ANNs) in solving these two problems. ANNs are a powerful artificial intelligence (AI) technique that have received wide acceptance in predicting the extent and location of damage in structures. In this study, the fundamental strategy for developing ANNs to predict the severity and location of double-point damage cases from the measured data of the dynamic behavior of the structure in I-beam structures is considered. ANNs are trained using vibration data consisting of natural frequencies and mode shapes obtained from experimental modal analysis and finite element simulations of intact and damaged I-beam structures. By using ANNs, some significant problems of conventional damage identification approaches can be overcome and damage detection accuracy can be improved.
The capabilities of the human brain have always fascinated scientists and led them to investigate its inner workings. Over the past 50 years a number of models have been developed which have attempted to replicate the brain's various functions. At the same time the development of computers was taking a totally different direction. As a result, today's computer architectures, operating systems, and programming have very little in common with information processing as performed by the brain. Currently we are experiencing a reevaluation of the brain's abilities, and models of information processing in the brain have been translated into algorithms and made widely available. The basic building-block of these brain models (neural networks) is an information processing unit that is a model of a neuron. An artificial neuron of this kind performs only rather simple mathematical operations; its effectiveness is derived solely from the way in which large numbers of neurons may be connected to form a network. Just as the various neural models replicate different abilities of the brain, they can be used to solve different types of problem: the classification of objects, the modeling of functional relationships, the storage and retrieval of information, and the representation of large amounts of data. This potential suggests many possibilities for the processing of chemical data, and already applications cover a wide area: spectroscopic analysis, prediction of reactions, chemical process control, and the analysis of electrostatic potentials. All these are just a small sample of the great many possibilities.
Nonlinear finite element analyses of reinforced concrete slab-column connections under static and pseudo-dynamic loadings were conducted to investigate their failures modes in terms of ultimate load and cracking patterns. The 3D finite element analyses (FEA) were performed with the appropriate modeling of element size and mesh, and the constitutive modeling of concrete. The material parameters of the damaged plasticity model in ABAQUS were calibrated based on the test results of an interior slab-column connection. The predictive capability of the calibrated model was demonstrated by simulating different slab-column connections without shear reinforcement. Interior slab-column specimens under static loading, interior specimens under static and reversed cyclic loadings, and edge specimens under static and horizontal loadings were examined. The comparison between experimental and numerical results indicates that the calibrated model properly predicts the punching shear response of the slabs.
The use of steel and fiber reinforced polymers (FRPs) for strengthening RC beams can significantly improve the flexural strength, fatigue life and the serviceability of the beams compared to un-strengthened beams. Prestressing materials enable the material to become more efficient since a greater portion of its tensile capacity is employed. Investigations have shown that prestressed FRPs are effective materials for strengthening deteriorated structures. This paper presents a comprehensive review on the flexural behavior of strengthened RC beams using prestressed FRPs. The review covers the near surface mounted (NSM), externally bonded reinforcement (EBR) and externally post-tensioned techniques (EPT) and the corresponding advantages and disadvantages are highlighted. Anchorage systems and the effect of prestressing levels on the ductility, deformability and bond behavior of prestressed FRPs are also addressed. Recommendations for the future research are also presented.
This paper presents an improved crack model incorporating non-linearity of flexural damage in concrete to reproduce changes in vibration properties of cracked reinforced concrete beams. A reinforced concrete beam model with multiple-distributed flexural cracks is developed, in which the cracked regions are modelled using the fictitious crack approach and the undamaged parts are treated in a linear-elastic manner. The model is subject to incremental static four-point bending, and its dynamic behaviour is examined using different sinusoidal excitations including swept sine and harmonic signals. From the swept sine excitations, the model simulates changes in resonant frequency with increasing damage. The harmonic excitations are utilised to investigate changes in modal stiffness extracted from the restoring force surfaces, and changes in the level of non-linearity are deduced from the appearance of super-harmonics in the frequency domain. The simulation results are compared with experimental data of reinforced concrete beams subject to incremental static four-point bending. The comparisons revealed that the proposed crack model is able to quantitatively predict changes in vibration characteristics of cracked reinforced concrete beams. Changes are sensitive to support stiffness, where the sensitivity increases with stiffer support conditions. Changes in the level of non-linearity with damage are not suitable for damage detection in reinforced concrete structures because they do not follow a monotonic trend.
The present paper addresses the subject of damage detection in structures. A series of numerical simulations on a simple beam are made in order to compare various damage detection methods based on mode shape changes. A generalisation of these methods to the whole frequency ranges of measurement is proposed, i.e. methods based on mode shape changes become based on operational mode shapes. The objective of such a study is to ascertain the possibility of using various damage detection methods without the need for modal identification. In recent years, the authors have developed some simple methods and tools based upon the use of frequency response functions (FRFs), which seem promising and have given good results in some practical applications. Here, some of that work is applied to the new proposed FRF-based methods which are compared and discussed. Numerical simulated examples will be used, as well as an experimental test case, where damage is inflicted in a free–free beam.
Structural damage detection has gained increasing attention from the scientific community since unpredicted major hazards, most with human losses, have been reported. Aircraft crashes and the catastrophic bridge failures are some examples. Security and economy aspects are the important motivations for increasing research on structural health monitoring. Since damage alters the dynamic characteristics of a structure, namely its eigenproperties (natural frequencies, modal damping and modes of vibration), several techniques based on experimental modal analysis have been developed in recent years. A method that covers the four steps of the process of damage detection—existence, localization, extent and prediction—has not yet been recognized or reported. The frequency-response-function (FRF) curvature method encompasses the first three referred steps being based on only the measured data without the need for any modal identification. In this paper, the method is described theoretically and compared with two of the most referenced methods on literature. Numerically generated data, from a lumped-mass system, and experimental data, from a real bridge, are used for better illustration.
Structural health monitoring has become an important research topic in conjunction with damage assessment and safety evaluation of structures. The use of system identification approaches for damage detection has been expanded in recent years owing to the advancements in signal analysis and information processing techniques. Soft computing techniques such as neural networks and genetic algorithm have been utilized increasingly for this end due to their excellent pattern recognition capability. In this study, a neural networks-based damage detection method using the modal properties is presented, which can effectively consider the modelling errors in the baseline finite element model from which the training patterns are to be generated. The differences or the ratios of the mode shape components between before and after damage are used as the input to the neural networks in this method, since they are found to be less sensitive to the modelling errors than the mode shapes themselves. Two numerical example analyses on a simple beam and a multi-girder bridge are presented to demonstrate the effectiveness of the proposed method. Results of laboratory test on a simply supported bridge model and field test on a bridge with multiple girders confirm the applicability of the present method.
A new model for the bending of a Bernoulli-Euler beam is developed using a modified couple stress theory. A variational formulation based on the principle of minimum total potential energy is employed. The new model contains an internal material length scale parameter and can capture the size effect, unlike the classical Bernoulli-Euler beam model. The former reduces to the latter in the absence of the material length scale parameter. As a direct application of the new model, a cantilever beam problem is solved. It is found that the bending rigidity of the cantilever beam predicted by the newly developed model is larger than that predicted by the classical beam model. The difference between the deflections predicted by the two models is very significant when the beam thickness is small, but is diminishing with the increase of the beam thickness. A comparison shows that the predicted size effect agrees fairly well with that observed experimentally.
This work proposes a vibration-based damage evaluation method that can detect, locate, and size damage utilizing only a few of the lower mode shapes. The proposed method is particularly advantageous for beam-like structures with uncertain applied axial load, mass density, and foundation stiffness. Based on a small damage assumption, a linear relationship between damaged and undamaged curvatures is revealed in the context of elasticity. It turns out that the resulting damage index equation inherently suffers from singularities near inflection nodes. The transformation of the problem into the multi-resolution wavelet domain provides a set of coupled linear equations. With the aid of the singular value decomposition technique, the solution to the damage index equation is achieved in the wavelet space. Next, the desired physical solution to the damage index equation is reconstructed from the one in the wavelet space. The performance of the proposed method is compared with two existing damage detection methods using a set of numerical simulations. The proposed method attempts to resolve the mode selection problem, the singularity problem, the axial force problem, and the absolute severity estimation problem, all of which remained unsolved by earlier attempts.
The use of changes in dynamic system characteristics to detect damage has received considerable attention during the last years. This paper presents experimental results obtained within the framework of the development of a health monitoring system for civil engineering structures, based on the changes of dynamic characteristics. As a part of this research, reinforced concrete beams of 6 meters length are subjected to progressing cracking introduced in different steps. The damaged sections are located in symmetrical or asymmetrical positions according to the beam tested. The damage assessment consists in relating the changes observed in the dynamic characteristics and the level of the crack damage introduced in the beams.It appears from this analysis that eigenfrequencies are affected by accumulation of cracks in the beams and that their evolutions are not influenced by the crack damage locations; they decrease with the crack damage accumulation. The MAC factors are less sensitive to crack damage compared to eigenfrequencies, but give an indication of the symmetrical or asymmetrical nature of the induced crack damage. Next to this, the COMAC factors, the strain energy evolution and the changes in flexibility matrices are also examined as to their capability for detection and location of damage in the RC beams, the strain energy method appears to be more precise than the others.
Over the past 30 years detecting damage in a structure from changes in global dynamic parameters has received considerable attention from the civil, aerospace and mechanical engineering communities. The basis for this approach to damage detection is that changes in the structure's physical properties (i.e., boundary conditions, stiffness, mass and/or damping) will, in turn, alter the dynamic characteristics (i.e., resonant frequencies, modal damping and mode shapes) of the structure. Changes in properties such as the flexibility or stiffness matrices derived from measured modal properties and changes in mode shape curvature have shown promise for locating structural damage. However, to date there has not been a study reported in the technical literature that directly compares these various methods. The experimental results reported in this paper and the results of a numerical study reported in an accompanying paper attempt to fill this void in the study of damage detection methods. Five methods for damage assessment that have been reported in the technical literature are summarized and compared using experimental modal data from an undamaged and damaged bridge. For the most severe damage case investigated, all methods can accurately locate the damage. The methods show varying levels of success when applied to less severe damage cases. This paper concludes by summarizing some areas of the damage identification process that require further study.
This paper describes a procedure for damage assessment in a two-storey steel frame and steel-concrete composite floors structure. The procedure is based on a multi-layer perceptron (MLP). A simplified finite element model is used to generate the training data. This model was previously updated through another MLP using two natural frequencies as inputs and the stiffness of the beams and masses as updating parameters. The different combinations of damage at the ends of the longitudinal beams are used as damage scenarios. The training data for the MLP are generated by varying at random the stiffness of the longitudinal beams. Two natural frequencies and mode shapes are used as inputs, and three different definitions of damage (sections, bars and floors) are tried as outputs. MLPs are trained through the error back-propagation algorithm. Finally, the performance of the procedure is evaluated through the experimental data. Only the approach of damage at floor level gives reasonable results.
The effects of high-cycle fatigue on the dynamic modal properties of a full-scale bridge test section are examined in this
study. A span 30.5 m (100 ft) in length was salvaged from an interstate bridge for experimental testing to establish the fatigue
performance of the retrofit procedures which could be used for future field repairs of the fatigue cracks. Horizontal cracks
had formed in the webs of the twin bridge girders in the vicinity of the web-flange weld at or near the connections of the
girders to the floor beams. An electrohydraulic vibration shaker was used to excite the test span in the torsional and bending
modes to promote fatigue crack growth. In addition to the experimental evaluation of the retrofit procedures, the dynamic
response characteristics of the test span during the high-cycle fatigue test were monitored. Experimental data have been used
to establish mode shapes and modal damping ratios during the four-million-cycle fatigue test. Damping ratios ranged from 1.09
percent to 0.53 percent. Modal stiffnesses were calculated based on the experimental mode shapes. There is a modest correlation
between the changes in stiffness and the observed cracking of the steel girders and the subsequent repairs.
Instances of local damage in timber such as knots, decay, and cracks can be translated into a reduction of service life due
to mechanical and environmental loadings. In wood construction, it is very important to evaluate the weakest location and
to detect damage at the earliest possible stage to avoid future catastrophic failure. In this study, modal testing was used
on wood beams to generate the first two mode shapes. A novel statistical algorithm was proposed to extract a damage indicator
by computing mode shapes of vibration testing before and after damage in timbers. The different damage severities, damage
locations, and damage counts were simulated by removing mass from intact beams to verify the algorithm. The results showed
that the proposed statistical algorithm is effective and suitable for the designed damage scenarios. It is reliable for the
detection and location of local damage of different severities, location, and number. The peak values of the damage indicators
computed from the first two mode shapes were sensitive to different damage severities and locations. They were also reliable
for the detection of multiple cases of damage.
This paper describes the determination of the location of damage due to single cracks and due to honeycombs in RC beams using mode shape derivatives from modal testing. The cracks were induced by application of point loads at predetermined locations on the RC beams. The load was increased in stages to obtain different crack heights to simulate the extent and severity of damage. Experimental modal analysis was performed on the beams with cracks prior to and after each load cycle, on a control beam, and beams with honeycombs. The mode shapes and the eigenvectors were used to determine the location of damage.The indicator |λ4| was obtained by rearranging the equation for free transverse vibration of a uniform beam, and applying the fourth order centered finite-divided difference formula to the regressed mode shape data. The equation is an eigenvalue problem, and the value of |λ4| will be a constant. Differences in the values indicate stiffness change, and the affected region indicates the general area of damage.Analysis of results using |λ4| was able to indicate the general region of damage, the exact location being around the center of the region. Curve fitting with Chebyshev series rationals onto the mode shape also highlighted points of high residuals around the region of damage. The proposed algorithm on the mode shape can form the basis of a technique for structural health monitoring of damaged reinforced concrete structures.
Full-scale dynamic testing of structures can provide valuable information on the service behaviour and performance of structures. With the growing interest in the structural condition of highway bridges, dynamic testing can be used as a tool for assessing the integrity of bridges. From the measured dynamic response, induced by ambient or forced excitation, modal parameters (natural frequencies, mode shapes and modal damping values) and system parameters (stiffness, mass and damping matrices) can be obtained. These identified parameters can then be used to characterize and monitor the performance of the structure. Analytical models of the structure can also be validated using these parameters. Reasons for conducting full-scale tests are given, and a discussion of the types of excitation devices used in forced vibration testing of large civil engineering structures is also included in the paper. A detailed review of published full-scale dynamic tests conducted on bridge structures is also given.
A methodology to identify damage in a structure is presented in this paper. The method utilizes a new form of damage index based on the changes in the distribution of the compliance of the structure due to damage. The changes in the compliance distribution are obtained using the mode shapes of the pre-damaged and the post-damaged state of the structure. The validity of the method is demonstrated using numerically generated data from beam structures and experimental data from a free–free beam structure with inflicted damage. In the numerical and experimental examples, the damage identification performance of the proposed method is compared with that of the existing strain-energy-based method. The results of the numerical and experimental studies indicate that the proposed compliance-based damage index method can be used in damage identification of the structure.
This study aims to develop a multi-stage scheme for damage detection for the cable-stayed Kap Shui Mun Bridge (Hong Kong) by using measured modal data from an on-line instrumentation system, and to perform a damage-identification simulation based on a precise three-dimensional finite element model of the bridge. This multi-stage diagnosis strategy aims at successive detection of the occurrence, location and extent of the structural damage. In the first stage, a novelty detection technique based on auto-associative neural networks is proposed for damage alarming. This method needs only a series of measured natural frequencies of the structure in intact and damaged states, and is inherently tolerant of measurement error and uncertainties in ambient conditions. The goal in the second stage is to identify the deck segment or section that contains damaged member(s). For this purpose, the bridge deck is partitioned into 149 segments defined by 150 sections, and normalized index vectors derived from modal curvature and modal flexibility are presented for damage localization. The third stage consists in identifying specific damaged member(s) and damage extent by using a multi-layer perceptron neural network. Only the structural members occurring in the identified segment are considered in the network input, and the combined modal parameters are used as the input vector for damage extent identification.
The underlying principle behind structural damage detection techniques is that vibration signature, e.g. modal properties or frequency response function (FRF) data, is a sensitive indicator of structural physical integrity and thus can be used to detect damage. Since indirectly-measured modal data contain accumulative errors incurred in modal parameter extraction and provide much less information than FRF data, it is more reasonable and reliable to use directly-measured FRF data for structural damage detection. In this paper, a new damage detection algorithm is formulated to utilize an original analytical model and FRF data measured prior and posterior to damage for structural damage detection. Based on nonlinear perturbation equations of FRF data, an algorithm has been derived which can be used to determine a damage vector indicating both location and magnitude of damage from perturbation equations of FRF data. An additional development with respect to the proposed technique is an effective technique introduced for weighting perturbation equations of FRF data at selected locations and frequencies so as to reduce influence of measurement errors on accuracy of damage detection to the minimum. For extension of the proposed algorithm to cases of incomplete measurement in terms of coordinates, an iterative version of the proposed algorithm has been introduced. The validity and applicability of the proposed damage detection algorithm have been demonstrated through numerical and experimental studies on a practical plane 3-bay frame structure, respectively.
This paper presents a damage detection algorithm using a combination of global (changes in natural frequencies) and local (curvature mode shapes) vibration-based analysis data as input in artificial neural networks (ANNs) for location and severity prediction of damage in beam-like structures. A finite element analysis tool has been used to obtain the dynamic characteristics of intact and damaged cantilever steel beams for the first three natural modes. Different damage scenarios have been introduced by reducing the local thickness of the selected elements at different locations along finite element model (FEM) of the beam structure. The necessary features for damage detection have been selected by performing sensitivity analyses and different input–output sets have been introduced to various ANNs. In order to check the robustness of the input used in the analysis and to simulate the experimental uncertainties, artificial random noise has been generated numerically and added to noise-free data during the training of the ANNs. In the experimental analysis, two steel beams with eight distributed surface-bonded electrical strain gauges and an accelerometer mounted at the tip have been used to obtain modal parameters such as resonant frequencies and strain mode shapes. Finally, trained feed-forward backpropagation ANNs have been tested using the data obtained from the experimental damage case for quantification and localisation of the damage.
Nonlinear finite element analysis was applied to various types of reinforced concrete structures using a new set of constitutive models established in the fixed-angle softened-truss model (FA-STM). A computer code FEAPRC was developed specifically for application to reinforced concrete structures by modifying the general-purpose program FEAP. FEAPRC can take care of the four important characteristics of cracked reinforced concrete: (1) the softening effect of concrete in compression, (2) the tension-stiffening effect by concrete in tension, (3) the average (or smeared) stress–strain curve of steel bars embedded in concrete, and (4) the new, rational shear modulus of concrete. The predictions made by FEAPRC are in good agreement with the experimental results of beams, panels, and framed shear walls.
In recent years, significant efforts have been devoted to developing non-destructive techniques for damage identification in structures. The work reported in this paper is part of an ongoing research on the experimental investigations of the effects of cracks and damages on the integrity of structures, with a view to detect, quantify, and determine their extents and locations. Two sets of aluminum beams were used for this experimental study. Each set consisted of seven beams, the first set had fixed ends, and the second set was simply supported. Cracks were initiated at seven different locations from one end to the other end (along the length of the beam) for each set, with crack depth ratios ranging from 0.1d to 0.7d (d is the beam depth) in steps of 0.1, at each crack location. Measurements of the acceleration frequency responses at seven different points on each beam model were taken using a dual channel frequency analyzer.The damage detection schemes used in this study depended on the measured changes in the first three natural frequencies and the corresponding amplitudes of the measured acceleration frequency response functions.
A seismic damage identification method intended for buildings with steel moment-frame structure is presented in this paper. The method has a statistical approach and is based on artificial neural networks and modal variables. It consists of two main stages. The initial one is devoted to the calibration of the undamaged structure and the final one to the identification of the damaged structure after an earthquake. The inputs of the nets are the first flexural modes (frequencies and mode shapes) at each principal direction of the structure and the outputs are the spatial variables (mass and stiffness). A damage index at each storey is determined by comparing the initial and final stiffness. A simplified finite element model was used to generate the data needed to train the nets. This model is consistent with available modal data and damage definition. The method was simulated on a 5-storey office building under conditions as close as possible to reality. The robustness of the method was verified with simulated data. Latter on, a sensitivity analysis of the mass variability was also carried out. Finally, the influence of modal error in the accuracy of damage predictions was statistically studied. Results are successful as concern as the robustness of the method. However, it is found that this approach is quite sensitive to modal errors.
Vibration-based damage identification (VBDI) techniques rely on the fact that damage in a structure reduces its stiffness and alters its global vibration characteristics. Measurement of changes in the vibration characteristics can therefore be used to determine the damage in the structure. Although VBDI offers several advantages, most of the available damage identification algorithms fail when applied to practical structures due to the effect of measurement errors, need to use incomplete mode shapes, mode truncation, and the nonunique nature of the solutions. This article presents a new robust two-step algorithm that uses the modal energy-based damage index to locate the damage and an artificial neural network technique to determine the magnitude of damage. The proposed algorithm is applied to detect simulated damage in a finite element model of a girder and a similar model of a real bridge named Crowchild Bridge located in Alberta, Canada. The results show that the proposed algorithm is quite effective in identifying the location and magnitude of damage, even in the presence of measurement errors in the input data.
Structural damage identification from dynamictest data
- R Juan
- Angel C Casas
- Aparicio
Juan R. Casas and Angel C. Aparicio. Structural damage identification from dynamictest data. Journal of Structural Engineering, 120(8):2437-2450, aug 1994. ISSN
0733-9445. doi: 10.1061/(ASCE)0733-9445(1994)120:8(2437).
Modelling of reinforced concrete structures and composite structures with concrete strength degradation taken into consideration
- P Kmiecik
- M Kamiński
P. Kmiecik and M. Kamiński. Modelling of reinforced concrete structures and composite
structures with concrete strength degradation taken into consideration. Archives of
Civil and Mechanical Engineering, 11(3):623-636, jan 2011. ISSN 16449665. doi:
10.1016/S1644-9665(12)60105-8.
Modal strain energy decomposition method for damage localization in 3D frame structures
- Huajun Li
- Hezhen Yang
- S.-L. James Hu
Huajun Li, Hezhen Yang, and S.-L. James Hu. Modal strain energy decomposition method for damage localization in 3D frame structures. Journal of Engineering Mechanics, 132(9):941-951, sep 2006. ISSN 0733-9399. doi: 10.1061/(ASCE)
0733-9399(2006)132:9(941).
Damage assessment of a gradually damaged RC beam using dynamic system identification
- J Maeck
- G De Roeck
J. Maeck and G. De Roeck. Damage assessment of a gradually damaged RC beam
using dynamic system identification. Proceedings of IMAC 20, International Modal
Analysis Conference, pp.1560-1566, Los Angeles, California, USA, February 2002,
2002.
Crack detection in a simply supported RC beam using simplified Laplacian
- H A Razak
- S L Teng
- Z Ismail
- A G Rahman
H. A. Razak, S. L. Teng, Z. Ismail, and A. G. Abdul Rahman. Crack detection in a
simply supported RC beam using simplified Laplacian. In Proceedings of the International Modal Analysis Conference -IMAC, 2001.
Energy index method: Technique for identification of structural damages
- Ashkan Sharifi
- Mohammad Reza Banan
Ashkan Sharifi and Mohammad Reza Banan. Energy index method: Technique for
identification of structural damages. Journal of Structural Engineering, 134(6):1061-1064, jun 2008. ISSN 0733-9445. doi: 10.1061/(ASCE)0733-9445(2008)134:6(1061).