Flávio de Souza Barbosa

A two-step technique for damage assessment using numerical and experimental vibration data

The degradation process of structural systems is usually due to a combination of reasons such as design or constructive problems, unexpected loading or even natural causes. Such deterioration process results in damaged regions whose main characteristics are localized stiffness decreases in the structure. In this article, the efficiency of a damage detection technique is analyzed through experimental tests and numerical simulations performed in a cantilever beam involving several damage scenarios. The methodology consists of two stages: (i) damage location — determined by means of the strain energy deviation between damaged and undamaged structural vibration modes; and (ii) damage quantification — developed through the analysis of measured natural frequencies of a damaged structure and its respective undamaged numerical model. In general, this two-step strategy shows good results although more robust results are obtained by using numerical data rather than by experimental data.

An artificial multilayer perceptron neural network for diagnosis of proximal dental caries

OBJECTIVE To evaluate if the application of an artificial intelligence model, a multilayer perceptron neural network, improves the radiographic diagnosis of proximal caries. STUDY DESIGN One hundred sixty radiographic images of proximal surfaces of extracted human teeth were assessed regarding the presence of caries by 25 examiners. Examination of the radiographs was used to feed the neural network, and the corresponding teeth were sectioned and assessed under optical microscope (gold standard). This gold standard served to teach the neural network to diagnose caries on the basis of the radiographic exams. To gauge the networks capacity for generalization, i.e., its performance with new cases, data were divided into 3 subgroups for training, test, and cross-validation. The area under the receiver operating characteristic (ROC) curve allowed comparison of efficacy between network and examiner diagnosis. RESULTS For the best of the 25 examiners, the ROC curve area was 0.717, whereas network diagnosis achieved an ROC curve area of 0.884, indicating a sizeable improvement in proximal caries diagnosis. CONCLUSION Considering all examiners, the diagnostic improvement using the neural network was 39.4%.

USING A NEURAL NETWORK FOR SUPPORTING RADIOGRAPHIC DIAGNOSIS OF DENTAL CARIES

This study uses an artificial intelligent model, a radial basis function neural network (RBF), to support radiography diagnosis of dental caries. One hundred and sixty radiography images of proximal faces of extracted human teeth were analyzed by 25 examiners, which diagnosed the presence or absence of dental caries. The same teeth were then subjected to optical microscope analysis, which allowed the verification of their actual conditions. Such information was classified as gold standards, and was employed to training a neural network to diagnose caries by means of radiography images. In order to verify the networks ability to diagnose new cases, data were organized in two subgroups: a training subgroup and a test subgroup. Receiver operating characteristics (ROC) curves allowed the comparison between diagnosis efficacy with or without the use of a neural network, showing that the adopted artificial intelligent model significantly improved diagnosis qualities.

A clustering-based strategy for automated structural modal identification

Structural health monitoring of civil infrastructures has great practical importance for engineers, owners and stakeholders. Numerous researches have been carried out using long-term monitoring, such as the Rio–Niterói Bridge in Brazil, the former Z24 Bridge in Switzerland and the Millau Bridge in France. In fact, some structures are continuously monitored to supply dynamic measurements that can be used for the identification of structural problems such as the presence of cracks, excessive vibration or even to perform a quite extensive structural evaluation concerning its reliability and life cycle. The outputs of such an analysis, commonly entitled modal identification, are the so-called modal parameters, that is, natural frequencies, damping rations and mode shapes. Therefore, the development and validation of tools for the automatic modal identification during normal operation is fundamental, as the success of subsequent damage detection algorithms depends on the accuracy of the modal parameters’ estimates. This work proposes a novel methodology to perform, automatically, the modal identification based on the modes’ estimates data generated by any parametric system identification method. To assess the proposed methodology, several tests are conducted using numerically generated signals, as well as experimental data obtained from a simply supported beam and from a motorway bridge.

Experimental and numerical evaluation of viscoelastic sandwich beams

Viscoelastic materials can dissipate a large amount of energy when subjected to cyclic shear deformations, but they have low bearing capacity. Therefore they are often employed as a damping layer in sandwich structures. These sandwich structures present a high damping ratio and simple application. In order to design sandwich structures, many aspects ranging from computer modeling to laboratory testing should be considered. In this study, a test set of experiments were performed and results are compared with a numerical GHM (Golla, Hughes and Mc Tavish method) based model, in order to establish a method to support viscoelastic sandwich beam design. In this way, starting from the dynamic properties of a viscoelastic material, a numerical model is used to evaluate the behavior of these structures. Comparisons with uncontrolled structures are also presented, showing the dissipative characteristics of this passive control.

Advanced Statistical Techniques Applied to Raw Data for Structural Damage Detection

Structural Health Monitoring is one of the most promising and challenging areas of research in the field of Civil Engineering. Over the last decades, researchers have focused on the development of consistent and reliable indicators aiming to detect, locate, quantify or even predict damage. More recently, some researchers are focusing on the use of raw time histories extracted from structural dynamic monitoring to build damage indicators. In this sense, this paper has as its main interest the use of high-order statistics (HOS) coupled with clustering techniques i.e. the k-means and c-means algorithms to detect structural modification (damage). The approach is applied directly to dynamic measurements (accelerations) obtained on site. The efficiency of such methodology is attested by means of a numerical study performed on a model of a simply supported beam and a study based on a real case railway bridge, in France. Results show that HOS coupled with clustering techniques are able to differentiate damage scenarios with adequate classification rates.

Modelling of the mechanical behavior of concrete affected by alkali-aggregate reaction

Alkali-Aggregate Reaction (AAR) is a major cause of chemical damage in concrete structures, especially those exposed to moisture, leading to cracking, excessive strains and stress formation. This work presents the results of a numerical study where the mechanical behaviour of AAR attained structures is simulated under the assumption of coupling between confinement stresses and the reaction.

Caractérisation dynamique d'un pont en bois exceptionnel : Exemple du pont de Merle

ABSTRACT They are very few road timber bridges in France. Among them, the Merle bridge, built in 1999, is no doubt the most exceptional one. The architectural choice was to avoid any reference to usual bridge types (arch, suspension), due to the localisation of the Merle bridge in a timbered site near a medieval castle. In the framework of a RGCU project, the Merle bridge was instrumented to assess its static and dynamic behaviour, and to get a better understanding of the structural behaviour. Indeed, during the design and construction process, several assumptions were made concerning the concrete slab connection to the timber structure and the general stiffness of the timber parts. This paper will present the different results of the three measurement campaigns for dynamic assessment and will compare them with analyses by the finite element method.