Dissertations / Theses on the topic 'Technological and non-destructive testing methods'

Thesis (M.S. in Civil Engineering)--University of Maryland, College Park, 1992.
"A scholarly paper submitted to Assistant Professor Ian Flood." Description based on title screen as viewed on April 16, 2009. Includes bibliographical references (p. 49-52). Also available in print.

In this work, unmanned aerial vehicle-based non-destructive testing methods for concrete structures are evaluated and developed. There exists a need for improved infrastructure inspection techniques with increased expediency. Unmanned aerial vehicles (UAVs) are highly mobile and have shown promise towards achieving this directive, but more work is required to adapt traditional NDT methods to be UAV-compatible. To this end, concrete sounding techniques were evaluated with a quantitative acoustic frequency analysis procedure on a series of concrete slabs. One such method was adapted for use with a UAV and was used to detect subsurface voids in one of the concrete samples and offer a means of depth estimation. This work was complemented with experiments concerning UAV-based visual and infrared imaging techniques already in practice for UAV-based concrete inspection. Together, findings indicate the strengths and weaknesses of the NDTs tested and suggest further improvements for UAV-based NDTs and inspection strategies moving forward. Development of a novel sensor platform for UAV-based measurement, as well as results of an actual bridge inspection using infrared and optical methods demonstrate the present capabilities of the UAV-based instrumentation.
Graduate
2019-07-03

Includes bibliographical references (leaves 122-135).
There is currently an increase in the usage of composite materials both in civilian and military aircraft because of their mechanical resistance, high toughness, reduced weight and immunity to corrosion. Any damage in the fibers of these composites requires an evaluation in order to ensure the integrity of aircraft structures. Composite materials are subject to various degrees of damage during their service life and therefore maintenance of these light-weight structures is a new challenge for non-destructive testing (NDT) which requires techniques that respond with a high probability of detection of defects on such high-specific strength materials. In this thesis, Digital shearography and Infrared Thermography (IRT) techniques are employed to test aircraft composite materials. Background information on the techniques has been presented and the literature survey has confirmed the use of these techniques on aircraft structures. Much of the inspection work reviewed in the literature has focused on qualitative evaluation of the defects rather than quantitative.

Bridge structures have an important role in economic, social and environmental aspects of society life. Bridges are also subject to a natural process of deterioration of construction materials, as well as natural and environmental events such as flooding, freezing, thawing etc. Health monitoring and assessment of the structural integrity of bridges have been the focus of engineers and researchers for decades. Currently, the various aspects of bridge health are monitored separately. However, measuring these aspects independently does not give the overall health of the bridge and crucial indicators of structural damage can be neglected. Generally, bridge health assessments take the form of individual NDT (non-destructive techniques) detecting individual defects. However value can be added to these results by combining and comparing the findings of several different NDT surveys. By completing this, a more accurate assessment of bridge health is obtained. This increases confidence in the decision as to whether remedial action is necessary. In this thesis an integrated bridge health monitoring approach is proposed which applies several NDT specifically chosen for bridge health assessments, thus achieving this added value. This method can be used as a part of a comprehensive bridge monitoring strategy as an assessment tool to evaluate the bridges structural health. This approach enables the user of this approach to obtain a detailed structural report on the bridge with all the necessary information pertaining to its’ health, allowing for a fully educated decision to be made regarding whether remedial action is necessary. This research presents the results of the applications of such methods on case studies utilising Ground Penetrating Radar (GPR), IBIS-S technology / system (deflection and vibration detection sensor system with interferometric capability) and Accelerometer sensors. It also evaluates the effectiveness of the adopted methods and technologies by comparing and validating the yielded results with conventional methods (modelling and visual inspection). The research presents and discusses processed data obtained by the above mentioned methods in detail and reports on challenges encountered in setting up and materialising the assessment process. This work also reports on Finite Element Modelling (FEM) of the main case study (Pentagon Road Bridge) using specialist software (SAP2000 and ANSYS) in order to simulate the perceived movement of the bridge under dynamic and static conditions. The analytical results output were compared with results obtained by the applications of the above non-destructive methods. Thus by using these techniques the main aim of this thesis is to develop an integrated model/approach for the assessment and monitoring of the structural integrity and overall functionality of bridges. All the above methods were validated using preliminary case studies (GPR), additional equipment (accelerometers for IBIS-S validation) and additional techniques and information (SAP 2000 and ANSYS were compared to one another and IBIS-S results). All of these techniques were applied on the Pentagon Road Bridge. This bridge was chosen as no information was available regarding its structural composition. Visual inspection showed the external defects of the structure: cracking, moisture ingress and concrete delamination was present in one of the spans of the bridge. The GPR surveys gave the position of the rebars and also signs of moisture ingress at depths of 20cm (confirmed using velocity analysis). IBIS-S gave results for the deflection of the structure. FEM was used to model the behaviour of the bridge assuming no defects. To achieve additional model accuracy the results of the rebar position were input in to the model and it was calibrated using IBIS-S data. The deflection results from the model were then compared to the actual deflection data to identify areas of deterioration. It was found that excessive deflection occurred on one of the spans. It was thus found that all NDT indicated that a particular span was an area of significant deterioration and remedial action should be completed on this section in the near future. Future prediction was also completed by running simulations in ANSYS for increasing crack lengths and dynamic loading. It was found that if there is no remedial action excessive beam bending moments will occur and eventual collapse. The results of this research demonstrated that GPR provided information on the extent of the internal structural defects of the bridge under study (moisture ingress and delamination) whilst IBIS-S technology and Accelerometer sensors permitted measurement of the magnitude of the vibration of the bridge under dynamic and static loading conditions. The results depicted similarities between the FEM results and the adopted non-destructive methods results in location and pattern. This work can potentially contribute towards a better understanding of the mechanical and physical behaviours of bridge structures and ultimately assess their life expectancy and functionality.

EM computation methods for the simulation and reconstruction of the metallic plate properties are investigated in this thesis. Two major computational problems exist in EM NDT: the forward problem and the inverse problem. The forward problem is to calculate the frequency-dependent inductance for steel plates with arbitrary values of permeability, conductivity, thickness and lift-off (i.e. the distance between the sensor and test sample). The inverse problem involves how to determine each parameter, i.e. permeability, conductivity, thickness and lift-off from the frequency-dependent inductance measurements. The purpose of this dissertation is to develop advanced forward and inverse solvers. This work will mainly deal with metallic plate structure in the low-frequency induction scheme. For the forward problem, both edge-element FEM and Dodd and Deeds analytical solution to simulate the eddy current probe-coil problems are developed. The feasibility and accuracy of the proposed forward solvers are verified by experiments and numerical solutions. An example of computation of eddy currents in metallic plates is also carried out to test the performance of the solver. The dissertation then goes further to consider the solution of the inverse problem of determining unique values for the four variables - permeability, conductivity, thickness and lift-off (i.e. the distance between the sensor and test sample) from the multi-frequency inductance spectra. In particular, novel methods on how to compensate lift-off variations are proposed. In addition, CIP is explored to measure the permeability of ferrous plates. These methods are verified by measurement results from EM sensors.

In recent years Terahertz (THz) time domain spectroscopy has emerged as a promising new technology with potential applications in a variety of fields, including industrial manufacturing, security screening and medical imaging. Pulsed THz systems are uniquely suited for non-destructive evaluation (NDE) of the sub-surface layers of dielectric packaging and coating materials, because they provide high dynamic range over a wide bandwidth in the far infrared portion of the electromagnetic spectrum. Often the dielectric materials of the packaging and/or surface coating layers exhibit relatively low loss and abrupt changes in the refractive index at the layer boundaries can be observed as a train of THz pulses in A-scan data. However, many practical applications of THz NDE will require fast signal acquisition to efficiently scan and evaluate many samples. The conventional processing approach shown in much of the published work in the field of THz NDE does not perform well in low signal-to-noise ratio (SNR) conditions. In addition, many samples of interest contain thin film layers and the THz pulses reflecting from the boundaries overlap on top of one another. Thus, it is not always possible to calculate the thickness of thin films from conventional time difference of arrival (TDOA) measurements. In this dissertation physics-based signal processing methods that have been historically used for radar/sonar signal processing are adapted and applied for THz NDE of layered media. Results are demonstrated with measured data from a pulsed THz system in the Northwest Electromagnetic and Acoustics Research Laboratory (NEAR-Lab) at Portland State University (PSU). This research is expected to provide an important link for THz researchers to access and apply the robust methods that have been developed over several decades for other applications. Two key contributions of this work are: 1. Development of a matched filter approach for THz NDE of thick layered media based on the maximum likelihood estimator (MLE). 2. Development of a matched field processing (MFP) approach for THz NDE of thin-film layered media, based on techniques in the underwater acoustics literature.

When fatigue is considered, it is expected that structures and machinery eventually fail. Still, when this damage is unexpected, besides of the negative economic impact that it produces, life of people could be potentially at risk. Thus, nowadays it is imperative that the infrastructure managers, ought to program regular inspection and maintenance for their assets; in addition, designers and materials manufacturers, can access to appropriate diagnostic tools in order to build superior and more reliable materials. In this regard, and for a number of applications, non-destructive evaluation techniques have proven to be an efficient and helpful alternative to traditional destructive assays of materials. Particularly, for the design area of materials, in recent times researchers have exploited the Acoustic Emission (AE) phenomenon as an additional assessing tool with which characterize the mechanical properties of specimens. Nevertheless, several challenges arise when treat said phenomenon, since its intensity, duration and arrival behavior is essentially stochastic for traditional signal processing means, leading to inaccuracies for the outcome assessment. In this dissertation, efforts are focused on assisting in the characterization of the mechanical properties of advanced high strength steels during under uniaxial tensile tests. Particularly of interest, is being able to detect the nucleation and growth of a crack throughout said test. Therefore, the resulting AE waves generated by the specimen during the test are assessed with the aim of characterize their evolution. For this, on the introduction, a brief review about non-destructive methods emphasizing the AE phenomenon is introduced. Next is presented, an exhaustive analysis with regard to the challenge and deficiencies of detecting and segmenting each AE event over a continuous data-stream with the traditional threshold detection method, and additionally, with current state of the art methods. Following, a novel AE event detection method is proposed, with the aim of overcome the aforementioned limitations. Evidence showed that the proposed method (which is based on the short-time features of the waveform of the AE signal), excels the detection capabilities of current state of the art methods, when onset and endtime precision, as well as when quality of detection and computational speed are also considered. Finally, a methodology aimed to analyze the frequency spectrum evolution of the AE phenomenon during the tensile test, is proposed. Results indicate that it is feasible to correlate nucleation and growth of a crack with the frequency content evolution of AE events.
Cuando se considera la fatiga de los materiales, se espera que eventualmente las estructuras y las maquinarias fallen. Sin embargo, cuando este daño es inesperado, además del impacto económico que este produce, la vida de las personas podría estar potencialmente en riesgo. Por lo que hoy en día, es imperativo que los administradores de las infraestructuras deban programar evaluaciones y mantenimientos de manera regular para sus activos. De igual manera, los diseñadores y fabricantes de materiales deberían de poseer herramientas de diagnóstico apropiadas con el propósito de obtener mejores y más confiables materiales. En este sentido, y para un amplio número de aplicaciones, las técnicas de evaluación no destructivas han demostrado ser una útil y eficiente alternativa a los ensayos destructivos tradicionales de materiales. De manera particular, en el área de diseño de materiales, recientemente los investigadores han aprovechado el fenómeno de Emisión Acústica (EA) como una herramienta complementaria de evaluación, con la cual poder caracterizar las propiedades mecánicas de los especímenes. No obstante, una multitud de desafíos emergen al tratar dicho fenómeno, ya que el comportamiento de su intensidad, duración y aparición es esencialmente estocástico desde el punto de vista del procesado de señales tradicional, conllevando a resultados imprecisos de las evaluaciones. Esta disertación se enfoca en colaborar en la caracterización de las propiedades mecánicas de Aceros Avanzados de Alta Resistencia (AAAR), para ensayos de tracción de tensión uniaxiales, con énfasis particular en la detección de fatiga, esto es la nucleación y generación de grietas en dichos componentes metálicos. Para ello, las ondas mecánicas de EA que estos especímenes generan durante los ensayos, son estudiadas con el objetivo de caracterizar su evolución. En la introducción de este documento, se presenta una breve revisión acerca de los métodos existentes no destructivos con énfasis particular al fenómeno de EA. A continuación, se muestra un análisis exhaustivo respecto a los desafíos para la detección de eventos de EA y las y deficiencias del método tradicional de detección; de manera adicional se evalúa el desempeño de los métodos actuales de detección de EA pertenecientes al estado del arte. Después, con el objetivo de superar las limitaciones presentadas por el método tradicional, se propone un nuevo método de detección de actividad de EA; la evidencia demuestra que el método propuesto (basado en el análisis en tiempo corto de la forma de onda), supera las capacidades de detección de los métodos pertenecientes al estado del arte, cuando se evalúa la precisión de la detección de la llegada y conclusión de las ondas de EA; además de, cuando también se consideran la calidad de detección de eventos y la velocidad de cálculo. Finalmente, se propone una metodología con el propósito de evaluar la evolución de la energía del espectro frecuencial del fenómeno de EA durante un ensayo de tracción; los resultados demuestran que es posible correlacionar el contenido de dicha evolución frecuencial con respecto a la nucleación y crecimiento de grietas en AAAR's.

A series of rail image inspection algorithms have been developed for Tata Steels Scunthorpe rail production line. The following thesis describes the contributions made by the author in the design and application of these algorithms. A fully automated rail inspection system that has never been implemented before in any such company or setup has been developed. An industrial computer vision system (JLI) already exists for the image acquisition of rails during production at a rail manufacturing plant in Scunthorpe. An automated inspection system using the same JLI vision system has been developed for the detection of rail‟s surface defects during manufacturing process. This is to complement the human factor by developing a fully automated image processing based system to recognize the faults with an improved efficiency and to allow an exhaustive detection on the entire rail in production. A set of bespoke algorithms has been developed from a plethora of available image processing techniques to extract and identify components in an image of rail in order to detect abnormalities. This has been achieved through offline processing of the rail images using the blended use of different object recognition and image processing techniques, in particular, variation of standard image processing techniques. Several edge detection methods as well as adapted well known Artificial Neural Network and Principal Component Analysis techniques for fault detection on rail have been developed. A combination of customised existing image algorithms and newly developed algorithms have been put together to perform the efficient defect detection. The developed system is fast, reliable and efficient for detection of unique artefacts occurring on the rail surface during production followed by fault classification on the rail imaging system. Extensive testing shows that the defect detection techniques developed for automated rail inspection is capable of detecting more than 90% of the defects present in the available data set of rail images, which has more than 100,000 images under investigation. This demonstrates the efficiency and accuracy of the algorithms developed in this work.

Thesis (PhD(For))--Stellenbosch University, 2014.
ENGLISH ABSTRACT: Pinus patula is the most intensively planted conifer in the tropics and sub‐tropics. In South Africa Pinus patula plantations are the main saw‐log resource for structural lumber production. Improved intensive silvicultural practices and tree breeding have resulted in marked increases in the rate of growth. To reap the financial benefits of the faster growth, plantation managers are more and more inclined to reduce rotation ages, which inevitably results in the production of higher proportions of juvenile wood at final harvest, and lumber which often does not meet the minimum requirements for stiffness for structural lumber. Knowledge of the variation and the accurate prediction of the mechanical properties of the timber of standing trees can have various benefits for growers and processors of trees. It can be used for tree allocation to different processing facilities, for processing production planning, and to assist tree breeders to screen and select for superior breeding material. The objectives of this study were (1), to examine the within‐ and between‐tree variation in wood properties of young South African grown Pinus patula trees known to have important impacts on the suitability of sawn lumber for structural purposes and (2), to develop empirical prediction models for the flexural lumber properties from standing Pinus patula, based on variables that could be assessed non‐destructively from standing trees. Sample material was obtained from 170 trees (16‐20 years old) established in 17 compartments along the Mpumalanga escarpment of South Africa. A large number of variables which could be obtained non‐destructively from the trees while they were still standing, were measured. The trees were subsequently felled and two logs, 2.1 m in length, were extracted from each tree at two height positions. The 340 logs were processed into 1402 pieces of lumber for further measurements and destructive testing. Results showed that the mean modulus of elasticity measured on edge (MOEedge) was far below the limits set for structural grade softwood timber in South Africa. All the desirable properties for structural lumber improved with distance from the pith with the exception of the 5th percentile value for modulus of rupture (MOR), which was higher at the pith than for the boards processed adjacent to the pith. Boards processed from the lower part of the stem were superior in most of the important properties compared to those higher up in the stem. Separate multiple regression models for predicting the average dynamic MOE (MOEdyn) of individual boards, trees and compartments were developed. The models managed to explain 68%, 60% and 95% of the variation in MOEdyn respectively. The models developed for MOR explained 40% and 42% of variability at board and tree level respectively. At compartment level, 80% of the variation in the 5th percentile MOR value could be explained by the model. Sensitivity analyses showed that site index at base age of 10 years, acoustic time‐of‐flight, wood density and ring width were the most influential variables in the MOE models. The models indicated that tree slenderness during early growth seems to play a major role in determining the dynamic MOE and MOR of lumber. This is in agreement with Euler’s buckling theory and the bending stress theory. Microfibril angle (MFA) and density were measured on radial strips taken from a sub‐sample of trees with the Silviscan 3 technology. The mean microfibril angle per year ring in Pinus patula varied between 7o and 29o. In general MFA decreased with distance from the pith and height above ground level. A multiple regression model including microfibril angle, density and ring width explained 71% of the variation in the dynamic MOE of boards. Sensitivity analysis on the model showed that microfibril angle and density had roughly equal influences on predicting the MOEdyn of Pinus patula boards.
AFRIKAANSE OPSOMMING: Pinus patula is die mees aangeplante naaldhoutspesie in die tropiese en sub‐tropiese areas van die wêreld. Dit is die grootste bron van saagblokke vir die produksie van strukturele hout in SA. Intensiewe boskultuurpraktyke en boomteling het gelei tot ‘n merkbare verhoging in die groeitempo van die spesie. Plantasiebestuurders is gevolglik geneig om rotasie‐ouderdomme te verlaag, wat lei tot ‘n groter persentasie jeughout wat nie aan die minimum styfheidvereistes van strukturele hout voldoen nie. Kennis van die variasie en die akkurate voorspelling van die meganiese eienskappe van staande bome kan voordele inhou vir beide die verbouers en verwerkers van bome. Dit kan byvoorbeeld van hulp wees met die toewysing van bome aan verwerkingsfasiliteite, vir produksiebeplanning, en vir ondersteuning met die keuse van teelmateriaal vir boomtelers. Die doelwitte van hierdie studie was (1), om die binne– en tussenboomvariasie in die houteienskappe, wat ‘n bepalende invloed het op die geskiktheid van jong Suid Afrikaanse Pinus patula bome vir strukturele hout produksie, te ondersoek en (2), om empiriese modelle vir die voorspelling van die buigeienskappe van planke te ontwikkel, gebaseer op veranderlikes wat niedestruktief op staande Pinus patula bome ge‐evalueer is. Monsters vir die studie is verkry vanaf 170 bome (16‐20 jaar oud), geplant in 17 vakke op die Mpumalanga platorand van Suid Afrika. ‘n Groot aantal veranderlikes is nie‐destruktief gemeet op die staande bome waarna die bome gevel is en twee saagblokke, 2.1m in lengte, is op twee hoogte posisies uit elke boom verwyder. Die 340 blokke is verwerk tot 1402 planke vir verdere metings en destruktiewe toetse. Resultate het getoon dat die gemiddelde modulus van elastisiteit gemeet op die dwarskant (MOEedge) aansienlik laer was as wat vereis word vir strukturelegraad hout in Suid Afrika. Al die gewenste eienskappe het toegeneem met afstand vanaf die murg behalwe die 5de persentiel breekmodulus (MOR), wat hoër was vir murgplanke as vir aangrensende planke. Planke afkomstig van die laer dele van die stam het oor die algemeen beter eienskappe gehad as planke afkomstig van die hoër dele. Veelvuldige regressiemodelle kon 68%, 60% en 95% van die variasie in die gemiddelde dinamiese MOE (MOEdyn) op die vlak van onderskeidelik individuele planke, bome en vakke verklaar. Die modelle vir MOR kon 40% en 42% van die variasie op onderskeidelik plank‐ en boomvlak verklaar. Die model vir 5de persentiel MOR van vakke kon 80% van die variasie verklaar. ‘n Sensitiwiteitsanalise het aangetoon dat groeiplekindeks op ouderdom 10, akoestiese vlugtyd, digtheid en jaarringwydte die belangrikste veranderlikes was wat MOEdyn beïnvloed het. Die modelle het aangetoon dat die slankheid van bome tydens vroeë groei vermoedelik ‘n belangrike invloed op die MOEdyn en MOR van planke het. Dit is in ooreenstemming met Euler se knikteorie en die buigsterkteteorie. Die mikrofibrilhoek en digtheid van ‘n steekproef van die bome is gemeet met die Silviscan 3 apparaat. Die gemiddelde mikrofibrilhoek per jaarring het tussen 7 o en 29o varieer. Hierdie variasie was hoofsaaklik afhanklik van boomhoogte en aantal jaarringe vanaf die murg. ‘n Veelvuldige regressiemodel wat mikrofibrilhoek, digtheid en jaarringwydte insluit, kon 71% van die variasie in MOEdyn verklaar. ‘n Sensitiwiteitsanalise op die model het aangetoon dat mikrofibrilhoek en digtheid ongeveer ewe belangrik was wat betref hulle invloed op die voorspelde MOEdyn van Pinus patula planke.

Since the MINER Act of 2006, the minimum static load of in-situ underground mine seals has been increased from 20-psi to either 50-psi if monitoring is conducted or 120-psi if left unmonitored. These minimum strength requirements in seals must be designed, built, and maintained throughout the lifetime of the seal. Due to this, it has become necessary to assess the effectiveness of non-destructive testing (NDT) technologies to determine seal integrity, which in this case, are explored using sonic waves and tracer gases. Through both small and large scale testing, two NDT methods were evaluated on their abilities to determine integrity of the seal. A sonic wave technique to observe a change in wave velocity to identify faults within the seal material. As a NDT method, tracer gases may be used as a potential indicator of a connection between both sides of the seal material through a series of faults and cracks within the material itself. This paper reviews the history of underground mine seals and discusses the overall assessment of sonic waves and tracer gases to serve as NDT methods for estimating the integrity of these seals.
Master of Science

Non-Destructive Testing of industrial components carries vital importance, both financially and safety-wise. Among all Non-Destructive techniques, Long Range Ultrasonic Testing utilizing the guided wave phenomena is a young technology proven to be commercially valid. Owing to its well-documented analytical models, Ultrasonic Guided Waves has been successfully applied to cylindrical and plate-like structures. Its applications to complex structures such as multi-wire cables are fairly immature, mainly due to the high complexity of wave propagation. Research performed by the author approaches the long range inspection of overhead transmission line cables using ultrasonic guided waves. Existing studies focusing on guided wave application on power cables are extremely limited in inspection range, which dramatically degrades its chances of commercialization. This thesis consists of three main chapters, all of which approaches different problems associated with the inspection of power cables. In the first chapter, a thorough analysis of wave propagation in ACSR (most widely used power cable) cables is conducted. It is shown that high frequency guided waves, by concentrating the energy on the surface layers, can travel much further in the form of fundamental longitudinal wave mode, than previous studies have shown. Defect detection studies proved the system’s capability of detecting defects which introduce either increase or decrease in cross sectional area of the cable. Results of the chapter indicate the detectability of defects as small as 4.5% of the cross sectional area through a 26.5 meter long cable without any post-processing. In the second chapter, several algorithms are proposed to increase the inspection range and signal quality. Well-documented wavelet-denoising algorithm is optimized for power cables and up to 24% signal-to-noise ratio improvement is achieved. By introducing an attenuation correction framework, a theoretical inspection range of 75 meters is presented. A new framework combining dispersion compensation and attenuation correction is proposed and verified, which shows an inspection range of 130 meters and SNR improvement up to 8 dBs. Last chapter addresses the accurate localization of structural defects. Having proven the optimum excitation and related wave propagation in ACSR cables, a system having a more complex wave propagation characteristics is studied. A new algorithm combining pulse compression using Maximal Length Sequences and dispersion compensation is applied to multi-modal signals obtained from a solid aluminum rod. The algorithm proved to be able to improve signal quality and extract an accurate location for defects. Maximal Length Sequences are compared to chirp signals in terms of SNR improvement and localization, which produced favourable results for MLS in terms of localization and for chirp in terms of SNR improvement.

[EN] This thesis considers new applications of non-Gaussian mixtures in the framework of statistical signal processing and pattern recognition. The non-Gaussian mixtures were implemented by mixtures of independent component analyzers (ICA). The fundamental hypothesis of ICA is that the observed signals can be expressed as a linear transformation of a set of hidden variables, usually referred to as sources, which are statistically independent. This independence allows factoring the original M-dimensional probability density function (PDF) of the data as a product of one-dimensional probability densities, greatly simplifying the modeling of the data. ICA mixture models (ICAMM) provide further flexibility by alleviating the independency requirement of ICA, thus allowing the model to obtain local projections of the data without compromising its generalization capabilities. Here are explored new possibilities of ICAMM for the purposes of estimation and classification of signals. The thesis makes several contributions to the research in non-Gaussian mixtures: (i) a method for maximum-likelihood estimation of missing data, based on the maximization of the PDF of the data given the ICAMM; (ii) a method for Bayesian estimation of missing data that minimizes the mean squared error and can obtain the confidence interval of the prediction; (iii) a generalization of the sequential dependence model for ICAMM to semi-supervised or supervised learning and multiple chains of dependence, thus allowing the use of multimodal data; and (iv) introduction of ICAMM in diverse novel applications, both for estimation and for classification. The developed methods were validated via an extensive number of simulations that covered multiple scenarios. These tested the sensitivity of the proposed methods with respect to the following parameters: number of values to estimate; kinds of source distributions; correspondence of the data with respect to the assumptions of the model; number of classes in the mixture model; and unsupervised, semi-supervised, and supervised learning. The performance of the proposed methods was evaluated using several figures of merit, and compared with the performance of multiple classical and state-of-the-art techniques for estimation and classification. Aside from the simulations, the methods were also tested on several sets of real data from different types: data from seismic exploration studies; ground penetrating radar surveys; and biomedical data. These data correspond to the following applications: reconstruction of damaged or missing data from ground-penetrating radar surveys of historical walls; reconstruction of damaged or missing data from a seismic exploration survey; reconstruction of artifacted or missing electroencephalographic (EEG) data; diagnosis of sleep disorders; modeling of the brain response during memory tasks; and exploration of EEG data from subjects performing a battery of neuropsychological tests. The obtained results demonstrate the capability of the proposed methods to work on problems with real data. Furthermore, the proposed methods are general-purpose and can be used in many signal processing fields.
[ES] Esta tesis considera nuevas aplicaciones de las mezclas no Gaussianas dentro del marco de trabajo del procesado estadístico de señal y del reconocimiento de patrones. Las mezclas no Gaussianas fueron implementadas mediante mezclas de analizadores de componentes independientes (ICA). La hipótesis fundamental de ICA es que las señales observadas pueden expresarse como una transformación lineal de un grupo de variables ocultas, normalmente llamadas fuentes, que son estadísticamente independientes. Esta independencia permite factorizar la función de densidad de probabilidad (PDF) original M-dimensional de los datos como un producto de densidades unidimensionales, simplificando ampliamente el modelado de los datos. Los modelos de mezclas ICA (ICAMM) aportan una mayor flexibilidad al relajar el requisito de independencia de ICA, permitiendo que el modelo obtenga proyecciones locales de los datos sin comprometer su capacidad de generalización. Aquí se exploran nuevas posibilidades de ICAMM para los propósitos de estimación y clasificación de señales. La tesis realiza varias contribuciones a la investigación en mezclas no Gaussianas: (i) un método de estimación de datos faltantes por máxima verosimilitud, basado en la maximización de la PDF de los datos dado el ICAMM; (ii) un método de estimación Bayesiana de datos faltantes que minimiza el error cuadrático medio y puede obtener el intervalo de confianza de la predicción; (iii) una generalización del modelo de dependencia secuencial de ICAMM para aprendizaje supervisado o semi-supervisado y múltiples cadenas de dependencia, permitiendo así el uso de datos multimodales; y (iv) introducción de ICAMM en varias aplicaciones novedosas, tanto para estimación como para clasificación. Los métodos desarrollados fueron validados mediante un número extenso de simulaciones que cubrieron múltiples escenarios. Éstos comprobaron la sensibilidad de los métodos propuestos con respecto a los siguientes parámetros: número de valores a estimar; tipo de distribuciones de las fuentes; correspondencia de los datos con respecto a las suposiciones del modelo; número de clases en el modelo de mezclas; y aprendizaje supervisado, semi-supervisado y no supervisado. El rendimiento de los métodos propuestos fue evaluado usando varias figuras de mérito, y comparado con el rendimiento de múltiples técnicas clásicas y del estado del arte para estimación y clasificación. Además de las simulaciones, los métodos también fueron probados sobre varios grupos de datos de diferente tipo: datos de estudios de exploración sísmica; exploraciones por radar de penetración terrestre; y datos biomédicos. Estos datos corresponden a las siguientes aplicaciones: reconstrucción de datos dañados o faltantes de exploraciones de radar de penetración terrestre de muros históricos; reconstrucción de datos dañados o faltantes de un estudio de exploración sísmica; reconstrucción de datos electroencefalográficos (EEG) dañados o artefactados; diagnóstico de desórdenes del sueño; modelado de la respuesta del cerebro durante tareas de memoria; y exploración de datos EEG de sujetos durante la realización de una batería de pruebas neuropsicológicas. Los resultados obtenidos demuestran la capacidad de los métodos propuestos para trabajar en problemas con datos reales. Además, los métodos propuestos son de propósito general y pueden utilizarse en muchos campos del procesado de señal.
[CAT] Aquesta tesi considera noves aplicacions de barreges no Gaussianes dins del marc de treball del processament estadístic de senyal i del reconeixement de patrons. Les barreges no Gaussianes van ser implementades mitjançant barreges d'analitzadors de components independents (ICA). La hipòtesi fonamental d'ICA és que els senyals observats poden ser expressats com una transformació lineal d'un grup de variables ocultes, comunament anomenades fonts, que són estadísticament independents. Aquesta independència permet factoritzar la funció de densitat de probabilitat (PDF) original M-dimensional de les dades com un producte de densitats de probabilitat unidimensionals, simplificant àmpliament la modelització de les dades. Els models de barreges ICA (ICAMM) aporten una major flexibilitat en alleugerar el requeriment d'independència d'ICA, permetent així que el model obtinga projeccions locals de les dades sense comprometre la seva capacitat de generalització. Ací s'exploren noves possibilitats d'ICAMM pels propòsits d'estimació i classificació de senyals. Aquesta tesi aporta diverses contribucions a la recerca en barreges no Gaussianes: (i) un mètode d'estimació de dades faltants per màxima versemblança, basat en la maximització de la PDF de les dades donat l'ICAMM; (ii) un mètode d'estimació Bayesiana de dades faltants que minimitza l'error quadràtic mitjà i pot obtenir l'interval de confiança de la predicció; (iii) una generalització del model de dependència seqüencial d'ICAMM per entrenament supervisat o semi-supervisat i múltiples cadenes de dependència, permetent així l'ús de dades multimodals; i (iv) introducció d'ICAMM en diverses noves aplicacions, tant per a estimació com per a classificació. Els mètodes desenvolupats van ser validats mitjançant una extensa quantitat de simulacions que cobriren múltiples situacions. Aquestes van verificar la sensibilitat dels mètodes proposats amb respecte als següents paràmetres: nombre de valors per estimar; mena de distribucions de les fonts; correspondència de les dades amb respecte a les suposicions del model; nombre de classes del model de barreges; i aprenentatge supervisat, semi-supervisat i no-supervisat. El rendiment dels mètodes proposats va ser avaluat mitjançant diverses figures de mèrit, i comparat amb el rendiments de múltiples tècniques clàssiques i de l'estat de l'art per a estimació i classificació. A banda de les simulacions, els mètodes van ser verificats també sobre diversos grups de dades reals de diferents tipus: dades d'estudis d'exploració sísmica; exploracions de radars de penetració de terra; i dades biomèdiques. Aquestes dades corresponen a les següents aplicacions: reconstrucció de dades danyades o faltants d'estudis d'exploracions de radar de penetració de terra sobre murs històrics; reconstrucció de dades danyades o faltants en un estudi d'exploració sísmica; reconstrucció de dades electroencefalogràfiques (EEG) artefactuades o faltants; diagnosi de desordres de la son; modelització de la resposta del cervell durant tasques de memòria; i exploració de dades EEG de subjectes realitzant una bateria de tests neuropsicològics. Els resultats obtinguts han demostrat la capacitat dels mètodes proposats per treballar en problemes amb dades reals. A més, els mètodes proposats són de propòsit general i poden fer-se servir en molts camps del processament de senyal.
Safont Armero, G. (2015). New Insights in Prediction and Dynamic Modeling from Non-Gaussian Mixture Processing Methods [Tesis doctoral no publicada]. Universitat Politècnica de València. https://doi.org/10.4995/Thesis/10251/53913
TESIS

The purpose of this work is to contribute to a process formaking buildings with good function and to avoid prematurefaults.

The design, construction and installation of low-slopedroofs are important parts of creating a durable building. Mostof the leakages in low-sloped roofs occur where materials withdifferent thermomechanical properties are joined together. Withbetter knowledge about these joints, the expected service lifecould better be estimated. Common roofing materials onlow-sloped roofs are roof membranes.

To avoid damages and to minimise energy consumption thedetection of air and water leaks is essential. It can bedifficult to localise a leak in e.g. a roof since water canflow far within the construction. Leakage detection can beapplied both as a quality assurance method after installationof low-sloped roofs and as field inspection methods. Theleakage detection can also be extended to terrace slabs and thewhole building envelope.

To investigate the strength of jointsbetween sheet metaland roofing membranes, several small-scale tests and somelarge-scale tests were performed. The test methods weredeveloped to match the loads that can be expected on this kindof joints.

A number of water leak-detection methods were evaluatedthrough application on test roofs. Some of the methods todetect leaks on low-sloped roofs can also be used to detect airleakage in other parts of the building envelope. To develop andevaluate air leak-detection procedures, selected methods wereused in two case studies.

The circumstances regarding welding of the material jointswere found to have great impact on the strength. The roofshould be designed so no long-term strain will appear since acomparatively low stress may damage the joint over time.

The performance of the leak-detection methods depends on theroofing material. All methods tested were an improvementcompared to visual inspections. Different recommendedapproaches for leakage detection and quality control is given.The case studies show that air leakage detection could beperformed with good accuracy. The potential difference methodcould without doubt be a tool for leakage localisation inwaterproofing layers both on roofs and in terrace slabs.

Keywords:Roofing, roof membrane, durability,waterproofing, leakage, wind-load, non-destructive testing,NDT, BSL4, BSL3, air leakage, building envelope

As applications for thermoplastic composites increase, the understanding of their properties become more important. Fabrication methods for thermoplastic composites continually improve to match designs specifications. These advanced thermoplastics have begun to show an improvement in mechanical properties over those found in thermoset composites commonly used in industry. Polyaryletherketones (PEK) have high service temperatures, good mechanical properties, and improved processing capabilities compared to thermoplastics used in the past making them important to the aerospace industry. The wide range of types of PEK make them suitable for a variety of applications, but selection of specific chemistries, processing parameters, and composite stack-ups determine the mechanical properties produced. Differential scanning calorimetry (DSC), and Fourier transform infrared spectroscopy (FTIR) were used to determine crystallinity and chemical properties of several polyaryletherketones. Tensile, compressive, and Mode I interlaminar fracture toughness tests were conducted to analyze mechanical properties of these advanced thermoplastics. Several fabrication processes were also tested to determine optimal consolidation and aesthetic appearance of structural members. All testing was conducted at The Boeing Company in Seattle, Washington. Because all testing and conclusions are proprietary a general synopsis of the experience will be presented.

In this thesis, two ultrasonic timing methods are used in order to investigate the accuracy and reliability of measurements for surface breaking cracks having different orientations and heights. Also the best applicable measurement technique is searched by comparing the received test results. These methods are the Time of Flight Diffraction (TOFD) Method using diffraction of longitudinal waves and another method using the reflection of shear waves from the crack tips. In order to simulate and measure the height of surface breaking cracks three sets of test blocks from steel, and two sets of wedges from plexiglas material are manufactured. Also several probes having frequencies of 2Mhz, 4Mhz, 5Mhz and angles of 45o and 70o are used. Some test procedures are created to make realistic comparisons between the test results and the ones found by previous studies in literature. The results are compared according to the standard deviations of errors in crack height measurements and it is found that the depth, orientation of defects and the frequency of probes have considerable affect on the results. With wider probe angles and higher frequencies of probes to some extent the errors are observed to be running low and the height of cracks could be measured closer to the original size. The amount of the errors is increased in measurements with the increasing angle of cracks. The results of both methods are found to be very satisfactory. A range of ±
0.5 mm for means of error from the original vertical crack heights is determined. The results agree with the previous studies.

Structural health monitoring (SHM) is concerned with the continuous, long-term assessment of structural integrity. One commonly investigated SHM technique uses guided ultrasonic waves, which travel through the structure and interact with damage. Measured signals are then analyzed in software for detection, estimation, and characterization of damage. One common configuration for such a system uses a spatially-distributed array of fixed piezoelectric transducers, which is inexpensive and can cover large areas. Typically, one or more sets of prerecorded baseline signals are measured when the structure is in a known state, with imaging methods operating on differences between follow-up measurements and these baselines. Presented here is a new class of SHM spatially-distributed array algorithms that rely on sparse reconstruction. For this problem, damage over a region of interest (ROI) is considered to be sparse. Two different techniques are demonstrated here. The first, which relies on sparse reconstruction, uses an a priori assumption of scattering behavior to generate a redundant dictionary where each column corresponds to a pixel in the ROI. The second method extends this concept by using multidimensional models for each pixel, with each pixel corresponding to a "block" in the dictionary matrix; this method does not require advance knowledge of scattering behavior. Analysis and experimental results presented demonstrate the validity of the sparsity assumption. Experiments show that images generated with sparse methods are superior to those created with delay-and-sum methods; the techniques here are shown to be tolerant of propagation model mismatch. The block-sparse method described here also allows the extraction of scattering patterns, which can be used for damage characterization.

Le principe du contrôle non destructif par Flux de Fuite Magnétique (FFM) consiste à magnétiser une pièce à contrôler par un champ magnétique intense et à détecter à l'aide d'un capteur magnétique les fuites des lignes de champ qui résultent de la présence d'un défaut dans la pièce. Les méthodes de contrôle FFM sont très employées notamment lors du processus de fabrication des tubes ferromagnétiques par la société Vallourec, le leader mondial des fabricants de tube. Dans le but d'améliorer les performances des systèmes de contrôle installés en usine, le CEA LIST et le centre de recherches de Vallourec (VRA) collaborent pour développer des outils de simulation rapides dédiés au contrôle virtuel des tubes ferromagnétiques. Le système expérimental existant concerne plus particulièrement la détection des défauts longitudinaux. Le problème de modélisation se pose en termes de modélisation d'un système électromagnétique à géométrie complexe en régime magnétostatique non-linéaire. Les courants de Foucault induits par le mouvement relatif entre la pièce et le circuit magnétique sont négligés. Dans ce contexte, une approche semi-analytique reposant sur le formalisme des équations intégrales (EI) a été choisie. Les travaux effectués dans cette thèse ont pour but de traiter des géométries complexes 3D mais limitées dans une première étape aux matériaux linéaires. Toutefois, le caractère non-linéaire de la relation liant l'induction magnétique et le champ magnétique dans un matériau ferromagnétique doit être envisageable lors du choix de la formulation du problème. Après une étude des paramètres influents du système expérimental existant, menée par des simulations par éléments finis, nous avons considéré deux stratégies de modélisation. La première consiste à proposer un schéma de résolution qui combine un module de calcul 2D et un module d'extension du 2D vers le 3D. Le manque de généralisation de cette première approche simplifiée nous a conduits à proposer une deuxième stratégie qui résout le problème complet de magnétostatique 3D. La formulation par équations intégrales porte sur une quantité scalaire auxiliaire : la densité surfacique de charges magnétiques. Afin de pouvoir résoudre à terme un problème 3D non-linéaire, le schéma numérique proposé considère deux hypothèses : la pièce ferromagnétique est divisée en un ensemble de cellules hexaédriques dans lesquelles la perméabilité magnétique est constante et les inconnues du problème, les densités surfaciques de charge sur les faces de chaque cellule sont projetées sur des fonctions de base d'ordre 0. Le calcul numérique des intégrales singulières s'effectue de manière analytique. Plusieurs résultats de simulation confirment la validité du modèle numérique présenté. Même si le modèle présente encore aujourd'hui quelques limitations notamment sur le manque de précision des calculs en présence de défaut, celui-ci donne satisfaction en absence de défaut. Diverses configurations géométriques ont été traitées grâce à l'emploi du mailleur libre Gmsh. Le travail réalisé débouche sur un modèle 3D linéaire intégrable dans un procédé itératif pour effectuer une simulation en régime non-linéaire. Les inconvénients liés au formalisme des équations intégrales sont aujourd'hui contournables grâce aux méthodes de compression de matrices. Ce modèle est un bon candidat pour servir d'outil de simulation pour le contrôle virtuel des matériaux plans ou cylindriques par flux de fuite.

La méconnaissance de la géométrie réelle d'une structure mène à une évaluation incorrecte de son état. Par conséquent, une estimation imprécise de sa capacité portante, sa durabilité, sa stabilité et la nécessité de mettre en place une réparation ou un renforcement. En outre, l'optimisation du temps requis pour le processus de réparation a besoin de bien connaître les différentes parties de la structure à évaluer et également pour éviter les zones critiques telles que les aciers, les câbles, etc., lors de la réparation. Par conséquent, il est nécessaire d'utiliser des techniques d'évaluation non destructive (END) afin de connaître la géométrie réelle de la structure, notamment l'emplacement des armatures dans les structures en béton armé. Le GPR est considéré comme une technique non-destructive idéale pour détecter et localiser les renforts. Cependant, sa précision de localisation est limitée. Le but de ce projet de recherche a donc été d'accroître la précision du GPR en matière de reconnaissance géométrique interne de structures en béton armé. L'objectif principal de cette étude est de localiser précisément le positionnement des armatures dans le plan ausculté ainsi qu'en profondeur. Pour atteindre cet objectif, une nouvelle méthodologie de mesures et du traitement des signaux GPR a été proposée dans cette étude. Plusieurs configurations d'acquisition de données en utilisant des signaux simulés sont testées pour proposer et développer un algorithme d'imagerie du milieu de propagation afin de définir sa géométrie interne et de localiser précisément les barres de renforcement. Des traitements supplémentaires sont appliqués pour améliorer la précision de la détection et pour identifier les différentes interfaces dans le milieu testé. L'algorithme et le traitement sont appliqués aux signaux simulés. Des validations expérimentales ont ensuite été appliquées aux signaux réels acquis sur différentes dalles en béton armé. L'objectif est de tester la capacité de l'algorithme d'imagerie proposé pour localiser différents objets enfouis. Les résultats encourageants montrent que cet algorithme est capable d'estimer la position de différents objets enfouis et pas uniquement les armatures avec une erreur d'estimation de (0-1) mm. Les performances de l'algorithme ont été comparées à celles d'une méthode de migration et aux résultats de mesure obtenus avec un pachomètre. Ces comparaisons ont systématiquement révélé une meilleure précision de la localisation avec l'algorithme développé.Une autre étude a été proposée dans ce travail en testant l'algorithme avec des signaux réels modifiés. Ces signaux sont produits en réduisant le gain le moins possible. La conclusion la plus évidente de cette étude est que l'algorithme proposé est capable de localiser les différents objets même si les signaux réfléchis par eux sont de faible amplitude
Lack of acquaintance in the real geometry of a structure leads to incorrect evaluation of its state. Consequently, this will lead to inaccurate estimation of bearing capacity, durability, stability and moreover, the need for repair or strengthening. Furthermore, optimization of the required time for repair process needs to well recognize the parts of structure to be assessed and also to avoid the critical zones such as reinforcing bars, cables, etc., during repairing. Therefore; it becomes necessary to use a non-destructive testing (NDT) method in order to know the real geometry of structure in particular, the location of reinforcements in reinforced concrete structures. GPR is considered as an ideal non-invasive technique in detecting and locating these reinforcements. However, its accuracy in localization is limited. The aim of this research project has therefore been to increase the accuracy of GPR in recognizing the internal geometry of reinforced concrete structures. The main objective of this study is to locate accurately the position of reinforcements into three dimensions. To achieve this purpose, a new methodology for GPR measurement and processing is proposed in this study.Several configurations of data acquisition using simulated signals are tested to propose and develop an appropriate imaging algorithm for the propagation medium to imagine its internal geometry and to locate accurately the reinforcing bars. Further processing are applied to improve the accuracy of detection and to identify the different interfaces in the tested medium. Both algorithm and processing are applied on simulated signals. Subsequent experimental validations have been applied using real signals acquired from different real reinforced concrete slabs. The goal is to test the ability of proposed imaging algorithm for the localization of different targets. The encouraging results indicate that this algorithm is able to estimate the position of different buried targets and not only the reinforcing bars with an estimation error of (0-1)mm.The performance of proposed algorithm has compared to those of migration method and to the results obtained from pachometer. These comparisons have systematically revealed a better localization accuracy using the developed algorithm.Another study has been proposed in this work by testing the algorithm using modified real signals. These signals are produced by reducing the gain as less as possible. The most obvious finding to emerge from this study is that the proposed algorithm is able to localize the different goals even if the signals reflected by them are of low amplitude

This thesis work focuses mainly on the application of non-destructive testing (NDT) methods on wood-based panels (WBP) in order to estimate the bending properties. To prove the accuracy and applicability of these methods on WBP, their results are correlated with results from a standardized static bending test. The behavior in different climate conditions and the application on panels of larger sizes is also questioned to provide an indication about strong points and boundaries of NDT methods applied on WBP.The bending properties are of major importance, especially for materials suited to bear loads. Bending stiffness, represented by the modulus of elasticity (MOE), is an expression of the deflection rate of a material under load. The bending strength, represented by the modulus of rupture (MOR), is an expression of the maximum load withstood by a material before rupture.Before testing, the material is acclimatized in three climate conditions: dry (20°C, 35% RH), standard (20°C, 65% RH) and wet (20°C, 85% RH), to understand the bending properties variation and how the NDT methods are affected by the variation in moisture content.The materials used are seven types of WBP, in particular four types of particleboards (PB), one type of high-density fiberboard (HDF), one type of dual density PB (with high and low density areas along the production direction) and one type of light-weight panel (Board-on-stiles, a composite panel of HDF, PB and paper honeycomb).To test the bending properties the following NDT methods are considered: transversal resonance vibration and longitudinal resonance vibration with the use of the BING system and the time-of-flight with the use of Fakopp Ultrasonic Timer and Silvatest Trio. The resonance vibration methods, transversal and longitudinal, are based on the relation between resonance vibration properties and bending properties of a material. The relation with bending properties also exists for the stress wave velocity (SWV) through a material, calculated with the time-of-flight method. The dynamic MOE resulting from these tests is then correlated with the static MOE and MOR from the static bending test.The NDT methods resulted to be reliable on WBP, with generally high levels of correlation between dynamic MOE and static MOE and MoR. The highest correlation value for MoE is with the transversal resonance vibration while the highest for MOR is with the longitudinal resonance vibration. The results of the dynamic MOE for all the NDT methods are higher than the static MOE, as confirmed also in the literature; the average ratio between the dynamic and the static MOE is, for example, up to 1,6 for WBP in standard climate condition, tested with Fakopp U.T.. These results are extremely higher than values suggested by previous studies. Moreover, the ratio increases with increasing relative humidity of the climate condition. The results from the tests on larger sizes suggest a possible application in this field. The time-of-flight method is suitable for in-plane uniform materials, like the PB and HDF, while the transversal resonance methods give also a good representation of the properties of the dual density PB and the light-weight panel.

L’inspection thermographique est une technique populaire de contrôle non destructif qui fournit des images de distributions de température sur de grandes étendues aux surfaces des pièces testées. Détecter les délaminations entre couches métalliques est le sujet ici. La simulation de ces inspections contribue en effet à compléter les études expérimentales, à évaluer les performances en termes de détection, et à être support d'algorithmes basés sur modèles. On se focalise sur un modèle semi-analytique basé sur un développement tronqué en fonctions propres par région. Le problème est résolu dans le domaine de Laplace en temps, et la distribution de température approximée par un développement sur une base produit tensoriel. Les sources considérées sont des lampes flash, mais aussi des sources courants de Foucault (conduisant à un couplage électromagnétisme et chaleur). La description des délaminages sous forme de minces couches d'air se révèle équivalente à l'introduction d'une résistance superficielle au flux de chaleur permettant le traitement via l'approche modale sans discrétisation supplémentaire. Des calculs complémentaires par des codes commercial (méthode des éléments finis) et interne (technique d'intégration finie) confirment l'exactitude. Puis une grande attention est donnée à l'imagerie et la détection. Une procédure en deux étapes est conçue : débruitage des signaux bruts et détection de tout éventuel défaut en utilisant une reconstruction de signal thermographique conduisant à une haute résolution spatiale et temporelle dans le plan transverse, complété par une détection de frontière, puis optimisation itérative, les résultats de la première étape étant utilisées pour la régularisation d'un schéma moindres carrés afin de caractériser épaisseurs et profondeurs. Tout ce qui précède est illustré par de nombreuses simulations numériques dans des conditions proches de l'application
Thermographic inspection is a popular nondestructive testing (NdT) technique that provides images of temperature distribution over large areas at surfaces of tested workpieces. Detecting delaminations between metallic layers is the matter here. Simulation of these inspections indeed helps to complement experimental studies, evaluate performance in terms of detection and support model-based algorithms. A semi-analytical model based on a truncated region eigenfunction expansion for simulation of thermographic inspection is focused onto. The problem is solved in the Laplace domain w.r.t time, and the temperature distribution approximated by expanding it on a tensor product basis. Considered sources are lamps providing thermal excitation but may also be eddy current sources (leading to a coupled electromagnetic and heat problem). The description of the delaminations as thin air gaps between the workpiece layers proves to be equivalent with introduction of a surface resistance to the heat flow, enabling treatment via the applied modal approach without additional discretisation. Complementary computations by industry (Finite Element Method) and in-house (Finite Integration Technique) codes confirm the accuracy of the developments. Then, much attention is put on imaging and detection. A two-step procedure is devised, first denoising of raw signals and detection of any possible defect using a thermographic signal reconstruction leading to high spatial and temporal resolution in the transverse plane, completed by proper edge detection, second an iterative optimization being employed, with results of the first step used for regularization of a least-square scheme to characterize thicknesses and depths. All the above is illustrated by comprehensive numerical simulations in conditions close to practice

Cette thèse porte sur la modélisation de la propagation d’ondes élastiques guidées (OG) dans des structures fines - de type plaque - en vue de leur contrôle non-destructif (CND). Ces ondes sont générées par des traducteurs ultrasonores quelconques placés à la surface de la pièce - par exemple une céramique PZT ou un EMAT. On développe un modèle semi-analytique, nommé modèle du pinceau modal d’ondes guidées. Ce modèle se base sur les approximations de l’acoustique géométrique, ce qui le rend particulièrement performant du point de vue numérique et permet de simuler la propagation d’OG sur de longues distances plus rapidement qu’avec des modèles numériques classiques. Le champ de déplacement est alors prédit seulement aux points d’intérêt. Un pinceau peut être utilisé pour traiter de façon générique les réflexions multiples d’une onde sur les bords latéraux de la plaque. Le modèle prend également en compte aussi bien les matériaux métalliques (isotropes) que les matériaux composites (anisotropes) ; la plaque peut également être courbée et avoir des propriétés mécaniques continûment variables. Les formes d’ondes sont obtenues sous la forme de leur décomposition modale, ce qui facilite grandement leur interprétation. Enfin, des comparaisons avec un modèle d’éléments finis permettent de valider l’approche de pinceaux
This thesis focuses on the modelling of the propagation of elastic guided waves (GW) in thin plate like structures for their non-destructive testing (NDT). These waves are generated by any ultrasonic transducer positioned on the surface of the piece - for example, a PZT ceramic or an EMAT. A semianalytical model, the so-called GW modal pencil model, has been developed. It is based on the geometrical acoustic approximation, which makes it particularly efficient from a numerical point-of-view and able to simulate GW propagation over long distances faster than with conventional numerical models. The displacement field is then predicted only at the points of interest. The pencil model can be used to generically deal with multiple reflections of GW on plate edges. It also takes into account plates made of either metallic (isotropic) or composite (anisotropic) materials. The plate can be curved and have continuously variable mechanical properties. The waveforms are obtained in the form of their modal decomposition, which greatly eases their interpretation. Finally, comparisons with a finite element model allow the pencil approach to be validated

Les structures mécaniques utilisées de nos jours ne cessent d’évoluer en utilisant des matériaux composites ou à gradient fonctionnel afin de répondre aux enjeux de résistance accrue, allégement de la structure et amélioration des performances. Ceux-ci nécessitent un contrôle adéquat de leur état de santé afin de s’assurer de l’intégrité de la structure. L’utilisation des ondes guidées ultrasonores fournit un moyen efficace et rapide d’inspection sur de longues distances. Néanmoins, ces ondes présentent certaines caractéristiques complexes qui rendent la tâche très difficile. L’utilisation d’outils d’analyse tels que les modèles numériques constitue un grand atout pour ce type d’application. Dans ce contexte, l’objectif de cette de thèse est le développement d’un outil de modélisation performant, permettant d’étudier la propagation des ondes guidées ultrasonores avec une grande précision et une faible consommation de ressources et de temps de calculs. De ce fait, l’intérêt est porté sur des méthodes numériques d’ordres élevés dont les propriétés de convergence sont beaucoup améliorées que les méthodes classiques. En particulier, la méthode semi-analytique éléments finis de Galerkin discontinue pour la détermination des courbes de dispersion des ondes guidées est développée. La méthode est applicable aux structures planes et cylindriques fabriquées de matériaux isotropes, anisotropes et hétérogènes (à gradient fonctionnel de propriétés). Une étude comparative sur l’analyse des performances de ces méthodes est effectuée. Celle-ci a démontré la capacité de la méthode à modéliser la propagation des ondes guidées ultrasonores dans des guides d’ondes à section arbitraire avec des performances prometteuses par rapport à la méthode des éléments finis classique
The mechanical structures used today are constantly evolving using composite or functionally gradient materials to meet the challenges of increased strength, lightening the structure and improving performance. These require adequate control of their state of health to ensure the integrity of the structure. The use of Ultrasound Guided Waves (UGW) provides an efficient and fast way of inspection over long distances. Nevertheless, these waves have some complex features that make the task very difficult. The use of analysis tools such as numerical models is a great asset for this type of application. In this context, the objective of this thesis is the development of a powerful modeling tool, allowing to study the propagation of UGWs with a great precision, less computational time and consumption of resources. Accordingly, we are interested in higher order numerical methods whose convergence properties are much improved than the classical methods. In particular, a semi-analytical discontinuous Galerkin finite element method (SADG-FE) is developped for the determination of the dispersion properties of guided waves in arbitrary cross-section waveguides. The method is applicable to plates and cylindrical structures made of isotropic, anisotropic heterogeneous (functionally graded) materials. The performance analysis of these methods and their comparisons are performed with respect to the models based on the classical finite element method. The results demonstrated the ability of the proposed method to model the propagation of ultrasounic guided waves in arbitrary section waveguides with promising performance over the conventional finite element method

The aim of my master’s thesis is to develop a brief overview of the historical development of brick, diagnosis methodologies of masonry structures and their application in the practical part for evaluation of structural survey masonry, carried out at the Faculty of Philosophy, Masaryk University in Brno.

This thesis deals with the diagnosis of a residential building from the late nineteen twenties for the needs of reconstruction. The methods generally used in the diagnosis of structures are briefly described in the theoretical part. More deeply this section describes the methods that were used during this particular research. The practical part deals with the diagnosis of the object itself. The first section describes the layout of the villa, its design solution, its faults and defects detected during the inspection. The next section deals with evaluating the current state of construction materials. Static reports of several structural parts were calculated in the last section.

CAPES
A detecção de defeitos de soldagem em imagens radiográficas visa garantir a segurança das estruturas em análise com o objetivo de evitar perdas financeiras e prevenir contra danos ambientais. Atualmente, a inspeção de juntas soldadas é uma atividade essencialmente humana e, portanto, está sujeita a erros relacionados à acuidade visual, à experiência, à fadiga e às distrações do inspetor, afetando a repetitividade e reprodutibilidade deste processo. Nesse sentido, este trabalho apresenta um método para auxiliar na detecção de defeitos em juntas soldadas de tubulações de petróleo, utilizando radiografias computadorizadas adquiridas pela técnica de exposição Parede Dupla Vista Dupla (PDVD). O método desenvolvido compreendeu a aplicação do realce das imagens tratadas, a segmentação de descontinuidades e a redução do espaço de busca pela eliminação da região central da junta soldada PDVD. Dessa maneira, os referidos procedimentos contribuíram para que as descontinuidades segmentadas que correspondiam a regiões de defeito em potencial fossem classificadas por Redes Neurais Artificiais (RNA) Multilayer Perceptron (MLP), realizando a detecção de defeitos de soldagem.
Detection of weld defects in radiographic images aims to ensure the safety of analyzed structures in order to avoiding financial losses and prevent against environmental damage. Nowadays, the inspection of welded joints is essentially a human activity and, therefore, it is subject to errors related to the inspector visual acuity, experience, fatigue and distractions, affecting the repeatability and reproducibility of this process. In this sense, this work presents a method to assist the detection of weld defects in welded joints of petroleum pipelines in computed radiography acquired by Double Wall Double Image (DWDI) technique. The developed method involved the application of contrast enhancement of treated images, segmentation of discontinuities and, the search space reduction by eliminating the central region of the DWDI weld. Thus, these procedures contributed to that segmented discontinuities which correspond to potential weld defects regions were classified by Multilayer Perceptron Neural Networks, performing the detection of weld defects.

CNPQ; LASCA
Este trabalho propõe um algoritmo para identificar descontinuidades internas em uma peça metálica a partir de sinais A-scan provenientes de um ensaio não-destrutivo por ultrassom. As descontinuidades são identificadas por seu formato, dimensões, orientação e localização. Esse algoritmo é baseado na resolução de problemas inversos. O problema da identificação de descontinuidades é modelado utilizando uma abordagem inédita, em que a descontinuidade é caracterizada por dois elementos distintos: a sua amplitude de espalhamento e a localização de seu centro. Com essa nova forma de representação, o problema de identificação da descontinuidade passa a ser dividido em dois subproblemas. O primeiro é o problema de reconstrução para encontrar a localização do centro da descontinuidade e o segundo é o problema de caracterização para estimar os parâmetros geométricos da descontinuidade. Os ensaios realizados comprovam que o problema de localização é resolvido de forma satisfatória utilizando o algoritmo de reconstrução de imagens esparsas UTSR (ultrasonic sparse reconstruction). A caracterização da descontinuidade é realizada por um algoritmo baseado no método dos mínimos quadrados não-lineares, com a inclusão de um termo de regularização não-quadrático utilizando norma l1. Os resultados obtidos tanto com dados simulados como com dados experimentais mostram que esse algoritmo estima as descontinuidades de forma satisfatória.
The present work proposes an algorithm to identify internal discontinuities in metallic specimen from A-scan ultrasonic signals. Discontinuities are identified by their shape, size, orientation and location. This algorithm is based on inverse problems. The identification of discontinuities is modeled using a novel approach. The discontinuity is modeled by two distinct elements: its scattering amplitude and its center location. Through this approach, the identification problem is split into two sub problems. The first is a reconstruction problem to find the location of the discontinuity. The second is a model identification problem to estimate geometric parameters of the discontinuity. Simulations and experimental validation show that the localization problem is satisfactorily solved using the UTSR (ultrasonic sparse reconstruction) algorithm. The characterization of discontinuities is carried out by an algorithm based on non-linear least squares with l1 norm regularization. The results obtained with simulated data as well with experimental data show a fine estimation of discontinuities.

The dissertation thesis deals with the study of non-destructive acoustic methods as instruments for monitoring and analysing corrosion of reinforcing steel in reinforced concrete elements. Four acoustic methods were selected for this task - the impact-echo method, the nonlinear acoustic spectroscopy method, the acoustic emission method, and the ultrasonic pulse velocity method. To verify the functionality of these methods, testing was carried out on three sets of reinforced concrete samples that had been exposed to the effects of sodium chloride, which corroded the embedded steel reinforcement in these samples. Suitable parameters were proposed for individual acoustic methods to monitor corrosion of the reinforcements. In addition, experiments were designed to demonstrate the ability of the selected acoustic methods to reveal the corrosion of steel reinforcement and its influence on the concrete matrix and to assess the condition of the degraded elements and structures. The analysis of the measurement results based on their comparison shows the advantages and disadvantages of the individual methods and of their practical applications. To verify the results, correlation with common methods that are currently used for the study of corrosion was carried out and included for example the electrical resistivity measurement of the reinforcement and simultaneous monitoring of the sample surface using a confocal microscope to record the development of microcracks during the degradation.

Petrobras
A inspeção radiográfica de juntas soldadas de tubulações é a atividade minuciosa e cuidadosa de observar imagens radiográficas de juntas soldadas em busca de pequenos defeitos e descontinuidades que possam comprometer a resistência mecânica dessas juntas. Como toda atividade que requer atenção constante, a inspeção radiográfica está sujeita a erros principalmente devido a fadiga visual e distrações naturais devido a repetitividade e monotonia inerentes à essa atividade. No presente trabalho, apresentam-se duas metodologias que têm por objetivo o auxílio e a automação da atividade de inspeção: a detecção automática dos cordões de solda nas radiografias e o realce das descontinuidades; compondo entre outras funcionalidades, um aplicativo completo de auxílio na inspeção radiográfica que agrega ainda a possibilidade de automação do processamento dessas imagens através da construção de rotinas e sua posterior aplicação a um conjunto de imagens semelhantes. Os resultados obtidos na detecção automática do cordão de solda são promissores, sendo possível, através da metodologia proposta, detectar cordões provenientes diferentes técnicas de ensaios radiográficos usuais. Quanto aos resultados do realce das descontinuidades, apesar de estes ainda não levarem a uma inspeção completamente autônoma e não supervisionada, apresentam resultados melhores do que aqueles existentes atualmente na literatura, principalmente quanto a correlação entre contraste visual do resultado do realce e a probabilidade de ocorrência de descontinuidades nas regiões demarcadas. Por fim, o realce das descontinuidades em conjunto com um aplicativo completo e iterativo contribui para uma maior leveza na atividade de inspeção, com o que se espera uma expressiva redução das taxas de erro devido à fadiga visual e um aumento considerável da produtividade através da automação das rotinas mais repetitivas de processamento digital a que as imagens radiográficas são submetidas durante sua inspeção.
The weld bead radiographic inspection is the activity of meticulously observe a radiographic image looking for small defects and discontinuities in the welded joints that can compromise the mechanical resistance of that joints. As any other activity than requires constant attention, the weld bead inspection is error prone due to visual fatigue, repetition and others distractions inherent to these activity. In this work, two new methodologies for help in the inspection activities are presented: the automatic detection of the weld bead and the highlighting of the weld bead discontinuities. Those that, among others functionalities, are included in a complete software solution for help in the weld bead inspection. Including the feature of macro programing for automation of the most common image processing routines and further processing bath of images in an automatic way. The results from the automatic weld bead detection is beyond the satisfactory, detecting weld bead from all the usual radiographic techniques. About the results of the highlight of the discontinuities, although that are not suited for a complete non supervised weld bead inspection, their correlation among intensity and the probability of the presence of a discontinuity is very well suited for discontinuities highlighting, a helpful tool in weld bead inspection. In conclusion, the proposed methodologies. combined with a fully featured interactive software solution, a lot contribute for the weld bead inspection activity, a decreased error rate due to visual fatigue and a better overall performance due to the automation of the most common procedures involved in this activity.

This thesis describes the methods and procedures used in designing production lines, focusing on the production line for machining of train wheels. The thesis also describes the basic types of railway wheels, their design and important design elements. It also deals with the complete manufacturing process of train wheel from material selection through the technological processes of forming, heat treatment, mechanical tests, machining and non-destructive testing and measuring. Suitable equipment for the construction of automatic lines for machining, inspection and measurement of train wheels is selected according to the specified parameters of the wheel types, manufacturing productivity and production shop dimensions, These devices include vertical machining center, single-purpose machine for grease outlet drilling, device for ultrasonic inspection, magnetic particle inspecton device, measuring station and train wheel balancing machine. Several variants of the production line including detailed specification of manipulation cycle times is created. These variants are mutually compared and evaluated to find the optimal solution.

CAPES
Em ensaios não destrutivos por ultrassom, o sinal obtido a partir de um sistema de aquisição de dados real podem estar contaminados por ruído e os ecos podem ter atrasos de tempo subamostrais. Em alguns casos, esses aspectos podem comprometer a informação obtida de um sinal por um sistema de aquisição. Para lidar com essas situações, podem ser utilizadas técnicas de estimativa de atraso temporal (Time Delay Estimation ou TDE) e também técnicas de reconstrução de sinais, para realizar aproximações e obter mais informações sobre o conjunto de dados. As técnicas de TDE podem ser utilizadas com diversas finalidades na defectoscopia, como por exemplo, para a localização precisa de defeitos em peças, no monitoramento da taxa de corrosão em peças, na medição da espessura de um determinado material e etc. Já os métodos de reconstrução de dados possuem uma vasta gama de aplicação, como nos NDT, no imageamento médico, em telecomunicações e etc. Em geral, a maioria das técnicas de estimativa de atraso temporal requerem um modelo de sinal com precisão elevada, caso contrário, a localização dessa estimativa pode ter sua qualidade reduzida. Neste trabalho, é proposto um esquema alternado que estima de forma conjunta, uma referência de eco e atrasos de tempo para vários ecos a partir de medições ruidosas. Além disso, reinterpretando as técnicas utilizadas a partir de uma perspectiva probabilística, estendem-se suas funcionalidades através de uma aplicação conjunta de um estimador de máxima verossimilhança (Maximum Likelihood Estimation ou MLE) e um estimador máximo a posteriori (MAP). Finalmente, através de simulações, resultados são apresentados para demonstrar a superioridade do método proposto em relação aos métodos convencionais.
Abstract (parágrafo único): In non-destructive testing (NDT) with ultrasound, the signal obtained from a real data acquisition system may be contaminated by noise and the echoes may have sub-sample time delays. In some cases, these aspects may compromise the information obtained from a signal by an acquisition system. To deal with these situations, Time Delay Estimation (TDE) techniques and signal reconstruction techniques can be used to perform approximations and also to obtain more information about the data set. TDE techniques can be used for a number of purposes in the defectoscopy, for example, for accurate location of defects in parts, monitoring the corrosion rate in pieces, measuring the thickness of a given material, and so on. Data reconstruction methods have a wide range of applications, such as NDT, medical imaging, telecommunications and so on. In general, most time delay estimation techniques require a high precision signal model, otherwise the location of this estimate may have reduced quality. In this work, an alternative scheme is proposed that jointly estimates an echo model and time delays for several echoes from noisy measurements. In addition, by reinterpreting the utilized techniques from a probabilistic perspective, its functionalities are extended through a joint application of a maximum likelihood estimator (MLE) and a maximum a posteriori (MAP) estimator. Finally, through simulations, results are presented to demonstrate the superiority of the proposed method over conventional methods.

碩士
弘光科技大學
環境工程研究所
103
Pipeline is a very important part in chemical process, since Kaohsiung gas explosion occurred which caused heavy casualties and property losses due to neglected pipeline management. Chemical pipeline management become imperative. The purpose of this study is to compare the applicability of four non-destructive testing methods applied on steel plant gas pipelines. Besides, ultrasonic testing method was used to measure the pipeline thickness in order to evaluate the remaining life and the schedule of replacing a new pipeline. In this study, two welding traces were completed on 2-inch oxygen and natural gas supply pipelines, one is a qualified weld, the other is a defective weld. Three non-destructive testing used for analysis include liquid penetrant testing, magnetic particle testing and radiographic testing. The advantages and disadvantages of each testing were discussed in five perspectives include internal functions, pipeline materials, object shape, real-time observation and the harmfulness of the testing methods. Ultrasonic testing method was used to evaluate the remaining life of the pipelines. The main results of this study is 1) internal functions: radiographic testing, ultrasonic testing can be used to detect the internal part of the pipelines; whether liquid penetrant testing, magnetic particle testing cannot be. 2) pipeline material: liquid penetrant testing, radiographic testing and ultrasonic testing can be used for most materials; whether magnetic particle testing cannot be. 3) object shape: liquid penetrant testing can be used for more complex shapes;radiographic testing, ultrasonic testing and magnetic particle testing can only detect relatively simple shapes. 4) real-time observation: liquid penetrant testing, magnetic particle testing, ultrasonic testing can show the testing results immediately; whether radiographic testing cannot. 5) hazards of testing method: liquid penetrant testing, magnetic particle testing and radiographic testing are hazardous themselves; ultrasonic testing is not hazardous. 6) the remaining life of the oxygen pipeline was estimated to be 7.3 years; natural gas pipeline was 8.3 years. 7) From the analysis of ultrasonic testing, it is found that the wearing rate of the pipeline elbow is more serious, therefore thinner thickness was measured relative to the other sampling tests. It is recommended that radiographic testing should be used for highly hazardous pipelines in the steel plant. In the case of flange welding, as the film cannot be placed evenly, penetrant testing should be used instead of radiographic testing. For the welding trace on the oxygen and the natural gas supply pipelines, it is recommended to adopt 100 percent proportional sampling test. Regarding life estimation of the pipeline, as the remaining life and the next testing time of both pipelines are close, and the pipeline position is adjacent to each other. It is recommended to test both pipelines in the meantime. Special attention should be paid for the abrasion in the elbow of pipelines to prevent pipeline leak. Keywords: Non-destructive testing, Liquid penetrant testing, Magnetic particle testing, Radiographic testing, Ultrasonic testing

Ο στόχος της παρούσας διδακτορικής διατριβής είναι διττός: η ανάπτυξη και η εφαρμογή αριθμητικών τεχνικών για την επίλυση προβλημάτων που εμπίπτουν στην περιοχή του Μη-Καταστροφικού Ελέγχου. Συγκεκριμένα αναπτύχθηκαν η Μέθοδος των Συνοριακών Στοιχείων (ΜΣΣ) και η Μέθοδος των Τοπικών Ολοκληρωτικών Εξισώσεων χωρίς Διακριτοποίηση για την αριθμητική ανάλυση στατικών και μεταβατικών προβλημάτων στο πεδίο της ελαστικότητας και της αλληλεπίδρασης ελαστικού με ακουστικό μέσο στις δύο διαστάσεις. Σημαντικό μέρος της διδακτορικής διατριβής αποτέλεσε η ανάπτυξη προγράμματος ηλεκτρονικού υπολογιστή, το οποίο επιλύει τα προβλήματα στα οποία πραγματεύεται το παρόν σύγγραμμα. Η διδακτορική διατριβή αποτελείται από τρεις ενότητες. Στην πρώτη ενότητα γίνεται πλήρης περιγραφή της απαραίτητης θεωρίας για την κάλυψη και κατανόηση των αριθμητικών ΜΣΣ αλλά και των Τοπικών Μεθόδων χωρίς Διακριτοποίηση (ΤΜχΔ). Στη δεύτερη ενότητα εφαρμόζονται οι προαναφερθείσες αριθμητικές μέθοδοι για την επίλυση στατικών και δυναμικών (στο πεδίο συχνοτήτων) διδιάστατων προβλημάτων, ώστε να πιστοποιηθεί η ακρίβεια και η αξιοπιστία των εν λόγω μεθοδολογιών. Τέλος, στην τρίτη ενότητα οι αριθμητικές ΜΣΣ και ΤΜχΔ εφαρμόζονται για την επίλυση προβλημάτων κυματικής διάδοσης που εμπίπτουν στο πεδίο του Μη-Καταστροφικού Ελέγχου. Πιο συγκεκριμένα μελετήθηκε η κυματική διάδοση σε ελεύθερες επίπεδες πλάκες και σε κυλινδρικές δεξαμενές αποθήκευσης υγρών καυσίμων.
The aim of this doctoral thesis is twofold: the development and implementation of numerical techniques for solving wave propagation problems in Non-Destructive Testing applications. Particularly, the Boundary Element Method (BEM) and the Local Boyndary Integral Equation Method are developed, so as to numerically solve static and transient problems on the field of elasticity and fluid-structure interaction in two dimensions. A major part of the present research is the construction of a computer program for solving such kind of problems. This textbook consists of three sections. In the first section, a thorough description on the theory of the BEM and the Local Meshless Methods (LMM) is done. The second section is dedicated for the numerical implementation of the BEM and LMM for solving steady state and time-harmonic two dimensional elastic and acoustic problems, in order to verify the accuracy and the ability of the proposed methodologies to solve the above-mentioned problems. Finally in the third section, the wave propagation problems of traction-free plates and cylindrical fuel storage tanks is studied, from the perspective of Non-Destructive Testing. The numerical methods of BEM and LMM are implemented, as well as spectral methods are utilized, for drawing useful conclusions on the wave propagation phenomena.