The Largest Open Access Portal of Nondestructive Testing (NDT)

An ultrasonic, resonant, pulse-echo, and air-coupled nondestructive testing (NDT) technique is presented. It is intended for components, with regular geometries where it is possible to excite resonant modes, made of materials that have a high acoustic impedance (Z) and low attenuation coefficient (α). Under these conditions, these resonances will present a very large quality factor (Q) and decay time (τ). This feature is used to avoid the dead zone, produced by the echo coming from the first wall, by receiving the resonant echo from the whole specimen over a longer period of time. This echo is analyzed in the frequency domain to determine specimen resonant frequency, which can be further used to determine either velocity or thickness. Using wideband air-coupled transducers, we tested the technique on plates (steel, aluminum, and silicone rubber) by exciting the mode of the first thickness. As expected, the higher the Z and the lower the α, the better the technique performed. Sensitivity to deviations of the angle of incidence away from normal (±2°) and the possibility to generate shear waves were also studied. Then, it was tested on steel cylindrical pipes that had different wall thicknesses and diameters. Finally, the use of this technique to generate C-Scan images of steel plates with different thicknesses was demonstrated.

Pattern recognition, which aims to associate data with a condition of a structure, has been applied successfully in Structural Health Monitoring (SHM) for damage diagnosis. Such association becomes more complex when applied to strain data measured from aerospace structures where the operational conditions produce changes in the strain patterns that are not related to a damage occurrence. Moreover, a damage occurrence only produces subtle changes in such patterns. That is why novelty detection strategies based on unsupervised learning have not demonstrated suitable results for strain-based SHM in operating aerospace structures. For example, imagine that it is possible to have data from all the different operational conditions for an aerostructure and, therefore, construct a model (e.g. statistical) from these data. Such model may be too general and some data from damage conditions may fit into the model and, subsequently, classified as a normal condition. One successfully-proved approach is to use unsupervised-learning, density-based classification to create clusters according to the operational condition and, then, build models for each specific cluster. In previous works, the authors implemented such methodology in an aluminum beam under simulated environmental conditions and subsequently, in the wing’s main beam of an Unmanned Aerial Vehicle (UAV) made of composites. The results for the metallic structure demonstrated a good performance since the changes in the patterns due to the operational condition variations were clear and identifiable. On the other hand, the data acquired from the UAV demonstrated to be fuzzy and without clear transitions among clusters. The aim of this work is to explore fuzzy clustering techniques in order to improve the global performance of the methodology for composite aerospace structures, which exhibit high stiffness. Fuzzy C-Means (FCM) and GustafsonKessel (GK) algorithms were tested using the UAV flight data and their performance was evaluated through Receiver Operating Characteristic (ROC) analysis.

The first approaches to damage detection in the context of Structural Health Monitoring (SHM) focused on identification techniques based on physical models, which imply the need of a model to describe the relationship between actions and responses in the structure. However, in the case of aerospace structures, where complex designs and materials are used to accomplish the demanding requirements, it is difficult to estimate accurate models to be used for diagnosis in aircraft Health and Usage Monitoring Systems (HUMS). This is even more evident with the increased use of composite materials in the aerospace industry. In this way, data-driven models which are based only on experimental data have been successfully implemented. Such techniques use machine learning algorithms to “learn” and evaluate the structural integrity with high accuracy. In previous work, the authors have developed a methodology based on Self-Organizing Maps (SOM) and Principal Component Analysis (PCA) for this kind of structures. Suitable results were obtained for strain data acquired from an Unmanned Aerial Vehicle (SMARP UAV) in flight tests by means of 20 Fiber Bragg Grating (FBG) sensors. However, there are different existing machine learning techniques that can be applied to develop novel methodologies for this kind of structures. One promising approach is Gaussian Process (GP) modeling, which can be seen as a generalization of a Gaussian distribution, adapted to classification problems for diagnosis. The main advantage of using this type of modeling is its well-founded approach to deal with the learning and model selection problem. Therefore, the aim of this work is to evaluate a methodology for damage detection in aerospace structures based on GP modeling by means of discrete strain measurements. The performance of the methodology is evaluated by using Receiver Operating Characteristic (ROC) curve analysis achieving a F1 score of about 0.944 in the best case.

Fiber Optic Sensors (FOS) have been used for Structural Health Monitoring (SHM) in aerostructures since they offer attractive advantages over standard electric gauges such as electromagnetic immunity, embedded ability, and small size. Damage detection performed directly from data acquired by this kind of sensors (in particular Fiber Bragg Gratings or FBGs) cannot be achieved since such sensors can only measure strain and temperature and are not able to measure damage by themselves. Different alternatives have been proposed in order to develop a global, automatic damage detection technique based on strain data. One of those techniques consists of studying the correlations among different sensors in a sensor network by using pattern recognition techniques, with the aim of unveiling changes in the global stiffness of a structure promoted by damage occurrence. However, in real-world aerostructures, variations in the operations conditions may also change the strain field, affecting the performance of these techniques. In this paper, a novel methodology based on strain and telemetry data fusion is proposed as an improvement of the global damage detection technique based on strain field pattern recognition. The technique was validated by using data a from autopilot telemetry and in-flight strain data of an Unmanned Aerial Vehicle (UAV) instrumented with 20 FBGs embedded in its wing structure. Different artificial damages were induced into the wing’s main beam (made of Carbon Fiber Reinforced Polymer, or CFRP) in order to test the whole methodology using different raw sensor and feature-level data fusion techniques. The results demonstrated the capability of the methodology for detecting damages during UAV operation and aim to provide solutions for a practical implementation of FOS-based SHM in real-world composite aerostructures.

Advanced textile materials are widely used in civil engineering, where recent catastrophic events highlighted the need of efficient methods and technologies for both retrofitting and Structural Health Monitoring (SHM). On the other hand, optical fibers, serving as both sensor and data conduit, are particularly suited to SHM for civil structures. The physical combination of the two mentioned technologies in a single self-sensorized textile led to the concept of “Multifunctional textile”, which can allow both structural reinforcing and monitoring, thanks to the presence of an optical sensor embedded inside the textile during the production process. The present work deals with the development and testing of a geogrid, for application on earthworks like embankments and sustaining walls, sensorized with optical fiber to be used as distributed strain sensor. From the technological point of view, two different approaches were adopted to embed the optical fiber in the textile structure, based, respectively, on one- and two-steps process. The obtained systems were then tested in laboratory according to two different test procedures: the former (tensile test) allowed characterizing the monitoring behaviour in terms of sensor gage factor, the latter (pull-out test) assessed both the reinforcing and monitoring behaviour on a medium scale set-up reproducing the real environment loading conditions. Optical measurements were based on Rayleigh scattering. Finally, the sensorized geogrid was installed and tested in real scale on a prototype embankment, where its technical viability at large scale was eventually evaluated. The size of the embankment was 3 meters high, 12 meters wide and 21 meters long, and sensorized geogrid samples were placed at different positions inside the embankment. In order to provoke known strain patterns to be detected by the embedded sensors, an irrigation system was built inside the embankment, watering it internally and eventually producing “controlled” settlements in the structure. The obtained results allowed setting the conditions, in terms of system performances knowledge and application technology, for its effective use in real geotechnical applications.

Within concrete, steel reinforcing bars (rebars) are naturally corrosion-protected (pH ~ 13). Concrete carbonation (“generalized” corrosion) and chloride ion penetration (“pitting” corrosion) both accelerate their corrosion rate. Corrosion products grow in volume and the increase in pressure at steel-concrete interface leads to cracks of the concrete layer and acceleration of degradation. They may escape as well, thus leading to a reduction in rebar diameter and global structure weakening. Corrosion is identified as a major pathology for Civil Engineering structures, with a clear impact over their long-term reliability. Until now, inspections are set on periodical base and involve indirect measurement techniques (e.g. chemical-, impedance-, potential- or ultrasonic-based) that are time-consuming, costly, probabilistic in nature, and of limited range of investigation (limited to accessible surfaces). Since it is impossible to predict where corrosion would start in large infrastructures, distributed monitoring techniques are desirable in order to early detect its onset, particularly in hidden areas, and to reduce both maintenance and downtime costs. We investigated the OFDR technique in the perspective of a Condition-Based Maintenance (CBM). The use of telecom-grade fibers as sensors is motivated by their cost-effectiveness, electrochemical passivity, electromagnetic immunity (thunder) and networking/multiplexing capability. We report on an original fiber-based corrosion sensor design employing usual steel rebars in order to avoid galvanic corrosion to occur. Since the rebar-based sensor is also an extensometer, a dedicated design is proposed to discriminate between global thermomechanical loading and local corrosion mechanical effects. Changes in OFDR signals with respect to reference signals provide localization, identification and direct measurement of corrosion. The sensing device was successfully tested under accelerated pitting corrosion as proof-of-concept.

In the manufacturing processes of the aeronautical sector, the use of composite materials is common practice allowing great advances which is rapidly spreading to other fields. The high-quality standards required for this type of material requires a 100% quality control of the structural elements before these parts are assembled and subsequently used. The increasing use of composite materials in more industries implies finding a suitable way for testing and evaluation. Non-destructive testing using ultrasonics is an appropriate method to detect defects in these types of materials. However, because of their complex structure the most applicable and accurate process for detecting flaws and anomalies has to be found. For conventional ultrasonic testing, single-element probes are used to generate an ultrasonic signal into the material to inspect it. However, when inspections are done on composite materials, phased array probes (PA) are now being used to detect component failures and thereby determine their quality more rapidly. The objective of this work is to apply the PA technique in the development of a new automated system that inspects, not only flat parts, but also for curved geometries following the contour of the part without loss of signal, removing the necessity to delimit different zones of inspection, optimizing the representation of results and reducing the time of evaluation and errors in interpretation. This system has robotized trajectory tracking, applying Pulse-Echo, and Double Transmission techniques with Phased Array technology, as well as TT transmission, with water coupling, so that it is possible to inspect one hundred percent of monolithic structures of solid laminates and sandwich structures. During the last years, great efforts have been made to improve different parts of the inspection systems. The results have been satisfactory for simple inspections, but not for inspections of elements with complex shapes is still proving a major challenge.

One of the most challenging issues for automatic integration in ultrasonic systems is the coupling of the probe with the component to be tested. Most common systems use water or oil in local or total immersion. The following paper will present an innovative dry coupling system, where no coupling media it is needed, that can be applied into different, fully automated, inspection techniques. A particular application of this Dry contact inspection technique is MATVEL UT Nodularity control. The MatVel is an up to 8 channel state of the art digital ultrasonic nodularity control system for industrial applications in the cast iron industry; uniquely offering dry coupled ultrasonic measurement capability. The system is suitable for manual or in-line spot testing of components. This is achieved by using a customized transducer and delay line. The MatVel combines all the functionality of an outstanding flaw detector together with specific add-ons customized for the application. The system can therefore meet the requirements for any set up, with live A-scan signal presentation and direct reading of acoustic velocity, thickness measurements and flaw (in spot) detection. To optimize the performance of the MatVel a specific series of transducers has been developed. Depending on the thickness range of the component at the test area, there are a range of state of the art delays of differing dimensions available: interchangeable spherical, rounded tip cylindrical or thin flat. The specific design of the delay has been developed for optimum coupling, acoustic impedance matching and durability under challenging mechanical and environmental testing conditions on rough casting surfaces.

Train Axles are critical components for safety, they are under a great number of charge/discharge cycles. Existing studies in order to control cracks show that fatigue failures are usually rare, when they occur unexpectedly there are catastrophic consequences. Therefore, it is important to determine the type of failure at an early stage to predict the safe life of the axle. So, as part of preventive maintenance activities, laborious manual ultrasonic testing is used. Thus, in recent years, work to develop a system which allows ultrasonic inspection, in service, for axles so as to: reduce human error, register data and generate detailed reports that guarantees the traceability of inspection, perform a complete inspection without disassembling the train axle. In short, to vastly increase the reliability of the test. In this paper, an ultrasonic NDT train axle inspection system is presented, which is able to test the complete axle from just one side of the train with almost 100% of the axle being inspected in one pass. This is carried out utilising different straight and angular beam transducers (up to 10), mounted in an Ogive, to maximise the probability of detection of both circumferential and longitudinal defects. The Boretest has the ability to inspect a whole axle from one side in less than 25mins. Each Ogive can inspect a number of different diameter holes within a range. Through a user-friendly software package, the Boretest plots the results in real time and records the results, including the calibration, of every inspection for further analysis. Furthermore, a new system has been developed for the inspection of train axles that are undergoing maintenance. This system utilizes the Boretest system and a Phased Array Ultrasonic Inspection system inspecting areas under structure not require to be removed during maintenance. This system will also be presented.

One of the key issues in the fight against the smuggling of goods has been the development of scanners for cargo inspection. X-ray-based radiographic system scanners are the most developed sensing modality. However, they are costly and use bulky sources that emit hazardous, ionizing radiation. Aiming to improve the probability of threat detection, an ultrasonic-based technique, capable of detecting the footprint of metallic containers or compartments concealed within the metallic structure of the inspected cargo, has been proposed. The system consists of an array of acoustic transceivers that is attached to the metallic structure-under-inspection, creating a guided acoustic Lamb wave. Reflections due to discontinuities are detected in the images, provided by an imaging algorithm. Taking into consideration that the majority of those images are sparse, this contribution analyzes the application of Compressed Sensing (CS) techniques in order to reduce the amount of measurements needed, thus achieving faster scanning, without compromising the detection capabilities of the system. A parametric study of the image quality, as a function of the samples needed in spatial and frequency domains, is presented, as well as the dependence on the sampling pattern. For this purpose, realistic cargo inspection scenarios have been simulated.

Usually fatigue cracks are detected by means of nondestructive testing, which involves long examination times and qualified personal. The detection process is then, expensive and time consuming. Recent advances in Structural Health Monitoring (SHM) techniques are very promising for the damage assessment in complex structures; however, fatigue detection is still one of the main challenges and open topics in the research field. Few techniques has proven effective for real time fatigue detection during the operation of structures. This paper presents a novel methodology for fatigue detection in structure under operational conditions. The methodology is based on strain measurements and strain field pattern recognition; in turn, pattern recognition is based on dimensional reduction and feature extraction techniques using neural networks. The advantages of the Fiber Optic Sensors (FOS) for strain measurements are exploited, in particular, Fiber Bragg Gratings (FBGs). Unless the strain sensors are closely located to a local fatigue crack, the change in the strain field caused by the crack is very small. Only when strain readings at several points are compared (pattern recognition), some information about fatigue damage may be unveiled. Robust automated techniques are needed to perform such comparison taking into account the redundant or useless information. Two different techniques are proposed for feature extraction: Principal Component Analysis (PCA) and Hierarchical Nonlinear PCA. An experimental validation of the technique is discussed in this paper.

Train Axles are safety critical components and they undergo a great number of charge/discharge cycles. Although existing studies regarding the control cracking show that fatigue failures are usually rare, when they occur unexpectedly they carry not only a high cost of replacement and repair, but also catastrophic consequences. Therefore, it is important to determine any type of failure at an early stage to predict a safe life. Currently, as part of preventive maintenance activities, manual ultrasonic testing is used, which requires time and exceptional understanding of the ultrasonic signals with the consequent costs associated. In this paper an ultrasonic hollow train axle inspection NDT system is presented, which is able to test the complete axle from just one side of the train with almost 100% of the axle being inspected in one pass with an inspection time of ess than 25 minutes. This is carried out by utilizing different angular and zero degree ultrasonic probes to maximize the probability of detection. The system set-up negates the need to prepare complicated calibration set ups before each inspection. As well as being easy-to-use it is versatile, allowing the testing of different types of axles within a range. One of the main aspects considered has been the development of software, which allows real time visual representation, giving a summary of the most important indicators, as well as the state of the system, the position of the probe, the results of the inspection in real time and the ‘A’ Scan of each probe.

Train Axles are safety critical components and they undergo a great number of charge/discharge cycles. Although existing studies regarding the control cracking show that fatigue failures are usually rare, when they occur unexpectedly they carry not only a high cost of replacement and repair, but also catastrophic consequences. Therefore, it is important to determine any type of failure at an early stage to predict a safe life. Currently, as part of preventive maintenance activities, manual ultrasonic testing is used, which requires time and exceptional understanding of the ultrasonic signals with the consequent costs associated. In this paper an ultrasonic hollow train axle inspection NDT system is presented, which is able to test the complete axle from just one side of the train with almost 100% of the axle being inspected in one pass. This is carried out by utilizing different angular and zero degree ultrasonic probes to maximize the probability of detection. The system set-up negates the need to prepare complicated calibration set ups before each inspection. As well as being easy-to-use it is versatile, allowing the testing of different types of axles within a range. The system display shows the state of the system, the position of the probe, the results of the inspection in real time and the ‘A’ Scan of each probe. The results of every inspection are recorded, including the calibration, for further analysis

La importancia en la inspección en tanques de almacenamiento para determinar las pérdidas de espesor producidas por la corrosión ha llevado al desarrollo de un sistema de inspección por ultrasonidos que permita el estudio del espesor de estructuras metálicas, tales como tanques de almacenamiento, tubos etc., pudiendo además operar de forma manual para el estudio detallado de zonas puntuales. Dicho sistema, transporta un sensor de ultrasonidos a través de la zona a inspeccionar para adquirir, de manera automática, la información del espesor, pudiendo realizarse un análisis de los datos adquiridos y la correspondiente generación de informes.

Eddy current inspection is widely specified for Aircraft Wheel Half inspection as there is no need to remove and replace the protective coating, which is necessary with dye penetrant testing. Manual eddy current inspection is permitted but is very time consuming, laborious and thus subject to human error. Automated inspection enables the inspection to be carried our both rapidly and with recordable consistency. This paper describes the requirements of such an inspection and how the development team met these requirements to produce a machine that is both simple to operate and mechanically robust for use in a Wheel and Brake Shop environment.

Measurements of Magnetic Barkhausen Noise (MBN) were carried out on AISI 420 steel samples of dierent hardness levels that were produced by means of controlled dierent heat treatments. The MBN measurements were treated by a supervised recognition method based on principal components and linear discriminant analysis in order to identify the levels of hardness through the MBN signal. Results showed that MBN can be used as a reliable source of information for statistical classication of steels in terms of hardness degree. Keywords: Magnetic Non Destructive Testing, Magnetic Barkhausen Noise, Multivariate Analysis, Principal Component Analysis, Discriminant Analysis, Supervised pattern recognition

In this paper, Magnetic Barkhausen Noise (MBN) measurements were used to identify hardness levels in AISI 420 steel obtained by means of different heat treatments. Barkhausen noise is a very powerful non-destructive tool for the inspection of hardness in steels. The inspecting technique is very fast, simple, cheap to use and does not need contact with sample surface. A data base of MBN signals were obtained and used in the development of an automatic signal processing procedure to identify and classify hardness. The supervised recognition method is based on principal components and discriminant analysis. The methodology used in order to develop the pattern recognition system is detailed. The results showed that Magnetic Barkhausen Noise can be used as reliable source of information for statistical supervised classification of hardness and heat treatment in steels.

In this paper, a methodology for ROI extraction in INDT using multi-resolution analysis is proposed. Complementary, both local procedures Harris operator and gradient direction are used. The proposed methodology is tested using three CFRP specimens having complex shapes and defects at different depths. Besides, another two specimens are considered, which are made of PlexiglasTM and aluminum with circular flat bottom holes at different depths. The results show that the proposed methodology is invariant to the material or defect shape among considered plates, moreover the methodology only has two parameters with no dependency of the variable features of the inspected object.