08:00
Departure of the Shuttles from Novotel Paris-Saclay & from Mercure Massy TGV to Saclay, NDT Aerospace 2019 Symposium

Session chairs :
Room :
10:15
We.2.A.1
Simulation of a fiber-optic based SHM system. Application to a bonded repair patch

A. Güemes1, M.E. Reyes1, J. Garcia-Ramirez1, C. Aguilar1, A. Fernandez-Lopez1

1 Universidad Politecnica de Madrid, Spain

imulation tools are very useful to understand and optimize NDT systems, and particularly the SHM systems. An algorithm has been proposed to quantify the damage detection capability of a fiber optic sensor network, with a finite number of strain measurements. A damage index is obtained by simulation, dependent of the damage size and position, and the number and distribution of sensors, and also of the noise of the signals; by repeating the simulation under random noise, a POD (Probability of Detection) can be calculated. The analysis is applied to a simple representative experiment, a bonded repair patch, under mechanical and thermal loads.
10:20
We.2.A.2
Development of multi-physics multi-scale modelling platform for CFRP composites using inductive thermography techniques.

A. Ba1, B. Yilmaz2, H.K. Bui1, E. Jasiuniene2, G. Berthiau1

1 IREENA, Saint Nazaire, France 2 Kaunas University of Technology, Kaunas, Lithuania

Due to their excellent mechanical performance, the use of carbon fiber composites has been growing in recent decades. However, the large-scale development of these materials depends on the improvements of the processes during the various stages of their whole life cycle (producing, forming, assembly, inspection, recycling). At various stages of the life cycle of the material, non-destructive testing (NDT) methods can be used to characterize the health state of the material. They play a vital role in the quality control and risk management. The Induction Thermography is significantly promising NDT technique for inspection of composites and it based on the measurement of eddy current thermal effects. The development of these methods requires multiphysics electromagnetic – thermal modelling. The developed models will deal with some numerical issues concerning thin regions of strong anisotropy and the multiscale geometries. In this paper, we develop a FEM coupled electromagnetic-thermal model in order to investigate the detectability of metal foil incorporate on the interface of three layer composite-adhesive bond.
10:25
We.2.A.3
Simulation-supported Analysis of Complex Radiographs

N. Brierley1

1 The Manufacturing Technology Centre, Coventry, UK

The analysis of radiographic images acquired for samples of complex geometry (such as those enabled by additive manufacturing techniques), where the imprint of the sample's geometric complexity in the radiograph is likely to undermine the ability to identify defect indications, is very challenging. An algorithm has been developed to assist in such circumstances, exploiting knowledge of the component geometry and approximate inspection configuration to generate a simulated baseline image with which a quantitative comparison can be conducted – in principle, allowing the geometric complexity to be subtracted out. This technique however requires the simulation to be accurately calibrated to the experimental configuration, necessitating the use of a numerical optimisation to fit the simulation parameters. As a by-product, the parameters of the imperfectly specified experimental set-up are recovered. The algorithm design is described, followed by results from increasingly complex test problems that illustrate the value of the approach.
10:30
We.2.A.4
NOVI-SIM: A new fast simulation tool for X-ray computed tomography testing

A. Autret1, G. Bravais1, O. Guiraud1, B. Fayard1

1 NOVITOM, Grenoble, France

X-ray computed tomography (CT) is a powerful imaging technique that has proven to be an effective tool for non-destructive 3D inspection. Nowadays, it is more and more employed as a standard characterisation method in R&D applications as well as a mean of control in production line. However, since X-ray CT acquisitions are time consuming and expensive, when compared to more classical NDT tools, its usage is still rather limited. The image quality of a CT scan depends on many acquisition parameters and should be adapted to each object analyzed. The parameters are often estimated based on the experience of the operator and then adjusted if the experimental result is not fully satisfactory. The procedure can be time consuming for non-experienced users and unusual or complex samples. In this context a simulation tool for tomography (and radiography) control is a key asset. It allows for calculating the best experimental parameters and reducing the acquisition time while optimizing the data quality. Based on its experience of synchrotron tomography and interaction of X-rays with matter, Novitom has developed a new simulation tool for X-ray tomography: NOVI-SIM. It uses a mixed approach (ray-tracing and wave optics) for fast execution and simulation of a large variety of sources, several detectors and contrast modes: absorption, phase contrast, single distance phase estimation …. The calculated images include a model for noise propagation and are fully representative of those experimentally obtained. The graphical interface allows for an intuitive use of the software and an easy adjustment of the parameters: exposure time, source type and power, type and thickness of filters, sample composition and geometry, noise level in the detector … The software is thus very valuable for operators of laboratory CT scanners and R&D engineers using tomography for NDT or material evaluation. After a brief description of the physical approaches used to perform fast simulation of image formation in radiography and tomography, the presentation will focus on the comparison of the output results with respect to experimental results acquired on phantom objects and challenging parts.
10:35
We.2.A.5
Semi-analytical modeling of eddy current inspection of stratified and homogenized composite materials
Composite materials are increasingly used in the aeronautic industry to reduce the weight of aeronautical structures. Therefore, the inspection of these complex materials for the detection of potential defects is a major challenge due to their complex physical behavior, to ensure their good performance in different operations. Within the Department « Imagerie Simulation pour le Contrôle » (DISC) from the institute CEA LIST, this thesis work is part of the European project NDTonAir (European Training Network in NDT and SHM of Aircraft structures). This project addresses several issues of NDT in the aeronautic field involving different methods of control based on several testing technics as ultrasonic waves, electromagnetic waves or infrared thermography. We are interested in the interaction of an eddy current sensor with a composite material affected by a defect such as a delamination or a fiber breakage. The goal is to develop a robust, fast and accurate numerical model based on a semi-analytic approach to optimize electromagnetic testing methods for this type of material. This model will be implemented into the commercial software CIVA developed by CEA LIST. In the first part of this thesis, we consider the computation of the EM field within a material without any defect. Therefore we assume in a preliminary step that each ply in the stratified material is homogenized. Thus, the modeling problem amounts to consider a homogeneous anisotropic media considering that an eddy current sensor is scanning the top surface of a planar specimen. The material is characterized by a conductivity tensor (axial or biaxial form) in each layer corresponding to the preferred orientation of the fibers. The semi-analytic model is based on writing Maxwell’s equations in the Fourier domain in order to obtain a modal expansion of the EM field in each layer. The unknown coefficients are evaluated by applying boundary conditions at each interface. To address the multi-layered structure, a global S-matrix algorithm is implemented since it provides a stable behavior. Considering numerical experiments, the numerical data provided by the model will be compared to other simulated data obtained by a FE commercial software.
10:40
We.2.A.6
Guided Waves Simulation Capabilities in Isotropic and Anisotropic Structures in View of Structural Health Monitoring Applications

R. Sridaran Venkat1, C. Boller1, A. Ernst1, M. El Bakouri1

1 Universität des Saarlandes, Saarbrücken, Germany

Guided waves is a means very much sought in the context of Structural Health Monitoring (SHM). In view of a structure’s damage tolerance and hence light weight potential a reliable SHM system has to be configured. Compared to classical NDT where the options of data sampling are virtually endless, SHM is fixed to a network of sensors and possibly also actuators, which hence has to be optimum. To determine this optimum, numerical simulation is possibly the best option. Specifically when a structure is designed damage tolerant the allowable size and location of a damage must be known. As a consequence guided wave propagation in a damage tolerant structure can be simulated and since the difference in the signals between a damaged and an undamaged condition is the key information to be ‘harvested’ by a sensing system it is this difference which has to be identified in view of configuring the optimum sensor network for monitoring. There are various bases on how such differential signals can be determined, being not just limited to the guided waves’ amplitude and phase only but being also extended to so called damage indices, which have been continuously developed over the years to specifically retrieve features from guided wave signals. The paper to be presented will initially provide examples on how to identify such optimum sensor networks through simulation with regard to isotropic materials based on what is called ‘differential imaging’ before moving to anisotropic materials such as CFRP, where numerical simulation capabilities are quickly reaching limitations in case no specific measures can be taken, which will be briefly discussed as well.
08:15
Opening of the Symposium, Welcome Coffee and exhibition

Session chairs :
Room :
08:45
Opening Ceremony

Opening Ceremony

Session chairs :
Room :
08:45
Welcome & Introduction
09:00
Plenary Sessions

Plenary Sessions

Session chairs :
Room :
09:00
NDE Challenges at Airbus
Mr Alvaro Espada Tejedor - Head of NDT at Airbus
The aerospace industry is one of the engines of the global economy. The need of commercial aircrafts, satellites and helicopters, to name some, has a direct impact in the Non-Destructive Testing (NDT). Other sectors, such as the medical, nuclear or piping, rely as well on non-destructive techniques, from ‘day 1’ to end of life of the structures/systems. The aerospace sector is not different: the NDT has been one of the key enablers. NDT (also known as NDI or NDE) is probably unknown for the majority of the global popula-tion. However, when we name “SAFETY”, we talk about NDT. When we refer to “SAVING LIVES”, we talk about NDT. We, the NDT community, understand the high impact that our work has in the society. This understanding implies responsibility, accuracy and knowledge. Definitely, for us, “failure is not an option”. Since the first applications in the aerospace sector, and in many cases due to incidents, the Authorities have been more and more requesting inspections. Methods such as visual in-spection, eddy currents, ultrasonics, to name some of them, have been further customized for the aerospace needs. And it has been in the last 40 years when the equipment and appli-cations have evolved to the level of what we have nowadays. However the demands in the aerospace sector are increasing, and the current NDT ecosystem is not ready to answer to them. Faster, more-efficient and same-reliability are not enough. Additionally, the aerospace industry, including the operators and Authorities, are showing their concerns about the knowledge and experience that will leave after the experienced NDT personnel retires. Those 40 years of development have been possible with the engagement of many NDT fellows. Is the new wave of NDT inspectors and engineers ready to take over? New competences are required, and the transition will be fast. Airbus, as well other actors in the NDT world, is working on automation, digitalization and connectivity. These three factors, when combined, bring a new set of tools that will form the next chapter of NDT. An overview of the needs and developments will be presented. Looking into the (near) future, the society is demanding more options to travel around the world, either within a city, or between continents, with environmentally friendly aircrafts. Air-bus is committed to exceed the needs and requirements by using new materials, designs and manufacturing processes. The development of new concepts goes hand in hand with the development of the next generation of NDT equipment, and the NDT personnel. We, the global NDT Community, have great challenges in front of us to demonstrate again that NDT is a key enabler.
Télécharger le document.
09:45
Structural Health Monitoring : Taking your NDT system along for the ride
Seth Kessler - Metis Design Corp
Damage tolerant design for aircraft traditionally requires periodic tear-down and manual NDT by trained inspectors at fixed flight-hour intervals to determine if flaws have grown to a detectable size. These intervals are set to minimize risk, and ensure multiple NDT opportunities to detect a flaw before it could achieve a critical size. While this approach is extremely safe, its conservatism can be quite costly to the life-cycle costs of the aircraft due to the inspection process and associated time out of service for preventative maintenance. For several decades, researchers have been investigating structural health monitoring (SHM) systems as a cost-effective compliment to manual NDT, leading to condition based maintenance (CBM) where targeted manual inspection would be triggered “as needed” based on frequent automated inspections rather than on a fixed schedule. While SHM sensors, hardware and algorithms have matured sufficiently to be taken seriously in the past few years, several roadblocks still remain before these systems can be introduced into military or commercial service. This presentation will highlight the key factors that need to be addressed before an SHM system can be fielded. First, since the SHM system is permanently installed within the aircraft, hardware and sensors much meet environmental and electrical airworthiness standards. Second, since sensors generally need to be intimately bonded to the structure being inspected, mechanical loading and durability requirements must be met. Finally, the detection performance of the SHM system must able to be quantified not only at the time of installation, but throughout its design life. While testing for the first two items are relatively straightforward, SHM system detection sensitivity assessment could prove to be quite challenging. NDT typically uses statistical metrics like probability of detection (PoD) to express the detection capabilities, however traditional approaches for measuring PoD are impractical for SHM due to fact that their sensors are permanently bonded thus many sensors must be used once then “wasted” to achieve statistical significance. Alternative proposed approaches to establishing statistical equiveillance to PoD for SHM will be described. Finally, an overview of the activities of the SAE International Steering Committee on Structural Health Monitoring (AISC-SHM) will be provided.
Télécharger le document.
10:30
Coffee Break

Plenary Sessions

Session chairs :
Room :
10:50
WE.1.A
Advances in NDT Methods I
AUDIT A

Advances in NDT Methods I

Session chairs :
Room : AUDIT A
10:50
WE.1.A.1
Millimeter waves Radar : A way to see through the airplane covering ?
Adrien Chopard (Laboratoire IMS - UMR 5218 CNRS, Université Bordeaux)
The usability of pulsed broadband terahertz radiation for the inspection of polymer materials from the aeronautics industry is investigated, with the goal of developing a mobile Frequency modulated continuous wave system that operates in reflection geometry. Frequency-modulated continuous-wave (FMCW) technique was originally investigated in radar for locating a distant object by means of a reflected radio wave. Contrarily to the continuous wave systems, FMCW has the possibility to provide information on the time delay. By modulating the frequency, the concept of the technique is equivalent to putting a unique "time stamp" on the emitted wave at every instant. First, the frequency is modulated with a triangle waveform by using a voltage-controlled oscillator (VCO), which varies the frequency between a maximum and a minimum value while changing the voltage. Then, the emitted signal is frequency-multiplied and amplified to reach the working frequency. This signal is considered as a reference signal and is emitted by a horn antenna. After being transmitted or reflected by the sample, the signal is detected and mixed with the reference signal using a multiplexer . The mixing produces a beat signal at a frequency fb, which is directly proportional to the time delay τ. This signal contains amplitude and phase information. A wide range of samples based on glass and reinforced plastics with various types of geometry is examined using an imaging system; the results are evaluated in frequency domain at 100GHz and 300GHz. The conductivity of carbon fibers prevents penetration of the respective samples but also allows analysis of coatings from the reflected THz pulses. Glass fiber composites are, in principle, transparent for millimeter wave radiation, but commonly with significant absorption for wavelengths >1 THz. Depending on depth, matrix material, and size, defects like foreign material, object or moisture contamination can be visualized through a covering structure of aeronautics. If a defect is not too deep in the sample, its location and its shape can be correctly identified from the delay between partial reflections at the surface and the defect itself. Acknowledgement: The authors thank Nouvelle Région d’Aquitaine and BPI France for Athermo project funding
Télécharger le document.
11:10
WE.1.A.2
Standardization, Influences, Possibilities and Risks of UV LED Technology to Fluorescent Magnetic Particle Inspection (MPI) and Penetrant Inspection (FPI)
Marc Breit (SECU-CHEK GmbH)
UV LED Technology is about to completely substitute discharge bulb-based UV-Sources like Mercury Vapour Lamps. The endless design possibilities of LED Technology are both a blessing and curse. A blessing because it allows to allows to build UV Sources that can make the inspection easier, better and faster than ever before and a curse because the technology itself is not, as the bulb-based UV-A sources, secure by design and it needs high efforts to achieve at least similar process security and performance with UV LED Technology. The wide range of very different lamps and the missed characterisation challenge responsible and users. All the possibilities and their influences to this security-critical process also challenge the standardization. The presentation will outline the process-relevant characteristics of UV sources, the possibilities the technology offers to Fluorescent Penetrant Inspection (FPI) Magnetic Particle Inspection (MPI) and how standardization is about to handle to describe the issues of LED based UV Sources that didn’t exist in the previous used technology.
Télécharger le document.
11:30
WE.1.A.3
New method for the evaluation of the Small Reflectors in the Ultrasonic Inspection of the Forging Titanium Alloy
Teodor Tranca (Dial Servicii srl)
Titanium’s accelerating usage in global markets is attributable to its distinctive combination of physical and metallurgical properties. The key to best utilizing titanium is to exploit these characteristics, especially as they complement one another in a given application, rather than to just directly substitute titanium for another metal. The current essay highlights some of the most important aspects concerning the industrial use of Titanium and its alloys. Starting with those regarding its properties, continuing with some technological processes involved in developing those materials and ending with the issues regarding recycling processes, titanium challenges scientists and engineers as well because of its remarkable properties that allow special applications in various human activities. Commercial Ti-6Al-4V forgings are widely used in the rotating components of aircraft engines. The failure of such parts can be quite catastrophic because of the large amount of kinetic energy. To ensure the safety and longer lifetime of these critical parts working in the hostile environments of high temperature and high stress, the need to detect smaller defects becomes more and more important. Ultrasonic inspection is one of the non-destructive evaluation (NDE) methods widely used by the titanium forgings manufacturers because of its capability to penetrate to the interior of a component. Over the last decades sizing methods were established like DGS (Distance Gain Size) or DAC (Distance Amplitude Correction) for defects smaller than the beam profile. Those methods utilize the echo amplitude and provide results which are proportional to the defect area. With the progress of material technology and ultrasonic inspection the need to detect and size smaller defects is growing. Therefore, both for flat bottom holes and disc shaped reflectors the usability for small defects needs to be checked.
Télécharger le document.
11:50
WE.1.A.4
Study of the damage evaluation method using the frequency change in the tensile test of the CFRP
Hiraku Kawasaki (IIC IHI Japan)
We have been studying the damage evaluation method during the tensile test of CFRP (Carbon Fiber Reinforced Plastic) materials by AE (Acoustic Emission) method. We reported previously that there are some possibilities to detect a sign of delamination occurred by the variation of F.C.O.G (Frequency Center Of Gravity) concentrated part of AE waves during the tensile test of CFRP materials by AE method. In this study, we confirmed that if the material is soundness, the concentrated part of F.C.O.G of AE wave will be risen with increasing the load stress. On the other hand, it was also confirmed that the concentrated part of F. C. O. G was decreased when damage has occurred. In this time, we investigate the AE signals generated by CFRP damage from tensile test of test pieces by only the carbon fiber and the ultrasonic propagation test during the tensile test. As a result, we were able to extract only the frequency of AE generated by damage of the carbon fiber from the AE signal generated during tensile test.
Télécharger le document.
12:10
WE.1.A.5
Contactless Terahertz Paint Thickness Measurements : specificity of aeronautics industry
Adrien Chopard (Laboratoire IMS - UMR 5218 CNRS, Université Bordeaux)
The thickness of paint films is one of the most critical quality parameters in the aeronautics paint process. Numerous measurement techniques exist for quantifying coated film thickness but many are unsuited to deployment in industrial environments or non metallis subtrate. With the advent of robust, turnkey systems in recent years, time-resolved terahertz pulsed sensing has matured sufficiently to find application in providing quantitative analysis of physical properties to a range of industries, including semiconductor package inspection and non-destructive testing of larger scale composites For many applications, layer thickness determination can be accomplished using a simple time-of-flight approach, whereby the thickness of individual layers can be estimated from the successive reflections from front and back interfaces of each layer[3]. This may be applicable to thick paint layers where the interface reflections are well separated, as in marine paint systems [4], but modern paint systems such as aeronautics painting are comprised of many thin paint layers, making accurate identification of individual interface reflections difficult. Terahertz time-domain spectroscopy is a specifically appropriated technique to analyze layered structure composition and geometry and dimension. Inverse electromagnetic problems are commonly performed to extract, from a recorded reflection THz-signal, the distinctive layer dielectric properties and thicknesses. However, main origins and formation routes of the signal are left unassessed while it could be of great value to deepen, to enlarge and to optimize stratified material THz-probing possibilities, control and understanding. In this work, an iterative algorithm implements a connected propagation tree where each node denotes a subsequent pulse subdivision. For each interface, nodes are pre-filled with the respective dielectric interface transmitted and reflected power coefficients. Descendant pulses are thus individually monitored and their carried proportion of the incident power remains at any time accessible. By constraining the algorithm to a given number of internal subdivisions, the original signal is recovered and the power contribution of each optical path is assessable. On the basis of this knowledge, the transfer function to be calculated within inverse electromagnetic problems can be simplified to enhance algorithm convergence speed. For instance, convergence speed-up would benefit the automation thickness chain control in industry. The algorithm robustness is demonstrated through the reconstruction of a THz-signal from a micro-metric layered structure used in aeronautics. The power proportions of different optical paths contributing to each part of the signal were explored. The algorithm allowed to reduce by about 99\% the number of calculated pulses compared to a standard stratified model. Acknowledgement: The authors thank Nouvelle Région d’Aquitaine and BPI France for Athermo project funding
Télécharger le document.
WE.1.B
Structural Health Monitoring I
AUDIT B

Structural Health Monitoring I

Session chairs :
Room : AUDIT B
10:50
WE.1.B.1
Damage Quantification in Aluminium-CFRP Composite Structures using Guided Wave Wavenumber Mapping
Yevgeniya Lugovtsova (BAM, Berlin)
The use of composite materials is associated not only with the advantages of weight reduction and improved structural performance but also with the risk of barely visible impacts or manufacturing damages. One of the promising techniques for the detection and characterisation of such damages is based on ultrasonic guided wave propagation and analysis. However, the multimodal nature and dispersive behaviour of these waves make their analysis difficult. Various signal processing techniques have been proposed for easier interpretation of guided wave signals and extraction of the necessary information about the damage. One of them is the wavenumber mapping which consists of creating a cartography of the wavenumber of a propagating mode over an inspected area, using a dense wavefield acquisition measured for example with a scanning laser Doppler vibrometer. This technique allows both the quantification of the in-plane size and the depth of damage, for example, impact- induced delamination in composite laminates. In this contribution, this imaging technique is applied to delaminated aluminium-CFRP composite structures which correspond to composite-overwrapped pressure vessels used for storing gases in aerospace and automotive industries. At first, the numerical investigations for the simple delamination positioned at different depth across the thickness are presented. Next, the analysis of experimental data obtained from the measurements of guided waves propagating in an aluminium-CFRP composite plate with impact-induced damage is performed. The output of the imaging is a three-dimensional representation of the delamination induced by the impact. Good agreement between the numerical and experimental residual thickness map can be found.
Télécharger le document.
11:10
WE.1.B.2
Machine learning based temperature compensation for Guided Wave Imaging in Structural Health Monitoring
Florian Le Bourdais (CEA-list)
In Guided Wave (GW) Structural Health Monitoring (SHM), a baseline, i.e a set of measurements taken on the inspected structure in a pristine state, is often required to separate the contributions of the defect(s) from the other propagating wavepackets. Due to the sensitivity of GWs to Environmental and Operational Conditions (OEC), most GW-SHM techniques are limited in terms of range of applicability. More specifically, the inspection of the unknown state must be conducted under the same EOCs as the ones of the pristine state. Besides measuring baselines on the structure under all the EOCs of interest, which is prohibitively expensive, a potential solution is to compensate the EOC effects on the measured signals. Several solutions in the literature, such as Baseline Signal Stretch and Dynamic Time Warping, have been proposed to solve this problem, but are somewhat limited in terms of amplitude of compensation or range of application. In this paper, a model-based machine learning procedure to compensate the measured signals in an unknown state and known EOCs is presented. The compensation model is trained on experimental data acquired at various temperatures on a structure representative of the one of interest. In other words, an experiment under various EOC must be conducted on a simplified version of the structure with at least two transducers. Material and transducers of the simplified experiment must be identical to the ones of the real structure, but the actual geometry might differ. The compensated signals are then used to conduct guided wave imaging, allowing immediate defect detection and localization. Results are shown for both aluminum and composite panels.
Télécharger le document.
11:30
WE.1.B.3
Hybrid Structural Health Monitoring on composite plates with embedded and secondary bonded Fiber Bragg Gratings arrays and piezoelectric patches
E. Monaco (Univ of Napoli)
Secondary bonded or embedded sensors are usually adopted in Structural Health Monitoring of composite aerospace structures. Each type of sensor has advantages and drawbacks when used separately although their proper integrated combination may improve the overall performance of a SHM system. The aim of the present work is to evaluate the feasibility of an efficient hybrid SHM system based on elastic wave propagation. The structure is sensorized with both piezoelectric patches (PZT) able to generate and receive strain waves and Fiber Bragg Gratings (FBG) working as receivers only. A dedicated software implementing acquisition strategies and data analysis allows to the SHM system to sense and locate low velocity impacts (in this mode both PZTs and FBGs work as receivers in a passive way) and to localize eventual impact damages at their early stage with few sensors: in this active mode PZTs work as sources and receivers, FBGs as receivers only. Many studies have been carried out by the scientific community to show SHM applications with a variety of bonded sensors integrated in to the structure. FBG sensors are among the most promising ones due to their lightweight, small dimensions, EMI immunity and good resistance to chemical environments. Generally, FBGs are used as static sensors for temperature and strain measurements but in the SHM system proposed herein the acquisition from FBGs are managed through a 4-channel acoustic emission sensor interrogator developed by Redondo Optics Inc. with a sampling rate slightly higher than 1MHz per channel. Ultrasonic pitch-catch technique has been used to perform the impact detection and hybrid guided wave (GuW) propagation. The technique has been applied to two CFRP panels. In the first one a FBGs network was embedded betweenthe composite layers during the manufacturing process, performed by Automatic Fiber Placement with Liquid Resin Infusion. The second one was exploited for GuW propagation and impact detection tests, using a secondary bonded PZT-FBG network of sensors and FBG arrays. A dedicated software in LABVIEW environment has been developed starting from the source code provided by the interrogator producer. The preliminary tests on the FBG's sensitivity and on the impact detection confirmed the capability of the FBG technology for this specific application, evidencing a good accuracy in the signals time delay measurements for both passive event detection and active guided waves propagation methodologies implementation. The proposed hybrid SHM system has been successfully applied to the two different panels, achieving good results in both low velocity impact tests and guided wave propagation experiments. The signals acquired through the FBG sensors are compared with those gathered from PZTs, demonstrating a very satisfactory matching. As a conclusion this paper demonstrated the feasibility of the set-up of the array of FBGs sensors in terms of hardware and software, matching the needed signal acquisition accuracy and parameters for both passive and active SHM. Some results will be presented for aircraft wings skin panels.
Télécharger le document.
11:50
WE.1.B.4
Improving Lamb wave detection for SHM using a dedicated LWDS electronics
Gladys Jaussaud (Cedrat Technologies)
Ultrasonic waves generated by piezo-electric patches offer an efficient way to perform Structural Health Monitoring (SHM). Permanently attached sensor arrays are interesting for on-demand interrogation of the structure in order to localize defects (pits), or follow defects progression (cracks). In this paper, improvements of the piezo-electronic system efficiency and robustness are proposed as required in H2020 REMAP project. SHM based on Lamb wave detection needs a dedicated electronic system enable to emit and detect Lamb waves. That is ensured by the Lamb Wave Detection System (LWDS), an electronic device developed by Cedrat Technologies. This module can drive a lot of piezo sensor (up to 36), and allows to work either in pitch-catch mode (each sensor is in emission or in detection) or pulse-echo mode (the same sensor can be in emission and in reception using a commutation time). This feature can drastically improve detection time, and defect localization. The integration of the sensors is another very influent feature for Lamb wave detection. That has been improved by focusing on the coupling efficiency between the piezo patch and Lamb wave propagation. Coupling dispersion curves, FEM modelling, and experimental testing have been investigated to define sensors and integration process favorable to Lamb wave detection. [1] E. Balmes, M. Guskov, and J.-P Bianchi, « Validation and verification of FE models of piezo based SHM systems, », ISMA International Conference on Noise and Vibration Engineering, Sep. 2016 [2] M. Rébillat, R. Hajrya, and N. Mechbal, « Nonlinear structural damage detection based on cascade of Hammerstein models, » Mechanical Systems and Signal Processing, vol. 48, no. 1, pp. 247-259, Oct. 2014.
Télécharger le document.
12:10
WE.1.B.5
Thermal aging evaluation of composite plates through electromechanical impedance
Pierre Marechal (University of Le Havre)
Electromechanical impedance monitoring is investigated to quantify changes in physical and mechanical properties during the ageing of composite plates. In this context, an experimental measurement protocol is proposed, on the basis of a broadband 1 MHz center frequency piezoelectric transducer. After a preliminary characterization of the parameters of the transducer itself, the acoustical impedance of the front medium is deduced. More particularly, the acoustical properties such as longitudinal wave velocity and attenuation are identified in the studied composite carbon/epoxy plates. From the electrical measurement results in the MHz frequency range, the acoustical impedance of the plate is plotted in the complex plane, showing characteristic signatures corresponding to the actually monitored composite plate. The ageing is quantified, showing the effectiveness of this non-destructive evaluation method. First, the acoustical parameters of an carbon/epoxy composite plate are modeled in order to study their sensitivity. During the ageing, those acoustical properties are known to evolve significantly, i.e. a decrease of the wave velocity and an increase of the attenuation, which can both be related to the increase of the porosity level in the plate. Secondly, these acoustical properties are evaluated on a set of samples which were submitted to a range of ageing duration from 500 to 5000 h performed at a thermostated temperature of 180°C. The feasibility of the evaluation of the ageing is demonstrated, and the precision of this measurement is discussed both in terms of temperature dependency and reproducibility.
Télécharger le document.
WE.1.C
Modelling & Simulation I
AUDIT C

Modelling & Simulation I

Session chairs :
Room : AUDIT C
10:50
WE.1.C.1
A Macro-element strategy dedicated to the transient modelling of the ultrasonic testing in carbon reinforced composite laminates : Efficient incorporation of local fiber orientation, visco-elastic behavior and thin intermediate layers
Edouard Demaldent (CEA list)
Modelling high-frequency wave propagation phenomena in anisotropic laminated materials such as carbon fiber reinforced composite structures is a major asset in numerous advanced industrial fields, such as for NdT in aerospace. Traditional homogenization procedures of each single ply (from 100 to 500 μm) lead to stratified isotropic transverse material properties whose anisotropic orientations depend on both the position of the layer in the stacking and the geometric deformation of the specimen (e.g. from a flat to a curved media). Both resin viscosity and multiple diffractions at the fiber scale attenuate the ultrasonic beam while intermediate epoxy layers (whose thickness is about 10 to 25 μm) reproduce the structural noise that arises around specific cutting frequencies of the bulk wave. Providing numerical solution to these problems requires complex and time-consuming calculation procedures in 3D, sometimes even hardly parameterized with a generic finite element software. It results in a drastic limitation of the computational performance, thus making it difficult to access any parametric study based on modelling. For a few years now, CEA LIST proposes a transient numerical computation strategy based upon a specific block- structured domain decomposition approach, denoted as the macro-element (or macro-mesh) strategy. In the context of curved composite laminates, this solution exploits both the continuity of the curvature as well as the regularity of the stacking to lighten the simulation with an efficient on-the-fly reconstruction of the local fiber orientation [QNDE 18, ARTICLE]. This communication will first recall this strategy and then focus on recent advances in the consideration of visco- elasticity and structural noise phenomena. Concerning the visco-elasticity, after recapping the attenuation behaviours of standard models (Kelvin-Voigt, Maxwell and Zener), we will propose a simple calibration procedure and analyze relevant time discretization strategies for each model, leading to efficient explicit numerical schemes. Concerning the structural noise, we will propose to incorporate the effect of epoxy layers using spring-mass transmission conditions between plies, adequately embedded in the macro-mesh strategy.
Télécharger le document.
11:10
WE.1.C.2
Investigations of the Potential of Ultrasonic Guided Waves Testing for Flaw Detection in Representative Aerospace Structures using CIVA.
Fatma Sellami  (Technical University of Munich)
In Non-Destructive Evaluation and Structural Health Monitoring, simulations tools are needed to model energy-damage interactions. Aircraft structures such as wing skins are natural waveguides that allow energy propagation. The latter property can be used to apply the Guided Waves Testing method to analyze the interactions of ultrasonic waves with different defects geometries, assess material losses and adhesion problems. Using CIVA SP3 Guided Waves Testing Module we generate inspection procedures for the following sample problems: fuselage wall thinning and corrosion which provokes small thickness changes, debonding of lap-splice joints, hidden defects in multilayered-aircraft structures, ice detection, and coating delamination inspections, integrity of stringers, spars and load-carrying paths’ parts susceptible to high stresses and fatigue cracks. Within the simulation platform, CIVA, the Semi-Analytical Finite Element method which relies on modal decomposition is applied for the wave propagation analysis and reciprocity relations are used to compute the elastic wave scattering coefficients. Mode computations, beam computations, inspection simulations and, parametric studies are performed to monitor the designed simplified aircraft components. The fuselage wall thinning case is addressed by assessing the frequency thickness dependence that can be demonstrated by analyzing the frequency content of the signal and the dispersion curves. In the case of lap splice joints and multi-layered structures, we analyze the in- and out- of planes modes that compose the guided waves and monitor energy leaks by varying the transducer parameters. Other theoretical investigations focus on enhancement methods based on time-reversal invariance with the aim of optimally focusing on a defect, resolving the problem of the dispersive character of guided waves, the coexistence of multiple modes at high frequencies and thus improving sensors readings reliability. The time-reversed iteration approach is used to precisely focus elastic waves in time and space in complex mediums by time-reversing and retransmitting echoes until the amplitude signal remains unchanged and a compact time-domain signal from an original multi-mode waveguide is obtained. Forward and inverse simulation approaches are investigated and additionally Acousto-Ultrasonics’ experiments are performed for defect detections (cracks and shrinkage cavities) in Dornier 328 Al2024 stringers and in the upper part of the Telescope Spectrograph Assembly Support Box of the Sentinel-4 instrument.
Télécharger le document.
11:30
WE.1.C.3
Improving Ultrasonic Inspection of Tapered CFRP Structures Using Cloud Based FEA Simulation
Jeff Dobson (OnScale)
The ultrasonic inspection of tapered Carbon Fiber Reinforced Polymer (CFRP) structures is challenging due to their internal ply layer structure often leading to high levels of refraction of the incident beam. This work demonstrates how time domain finite element analysis modelling can be utilized to evaluate and optimize the inspection of these components. Traditionally, fully numerical simulation has been too computationally expensive for many modelling scenarios. However, when deployed on cloud computing, 1,000s of simulations can be executed in parallel to enable parametric studies and full ultrasonic scans to be completed in economical timeframes. The value of modelling to aid the understanding of ultrasonic inspections of CFRP components is demonstrated through ultrasonic B-scan simulations of tapered structures. Model geometries are constructed directly from ply layup specifications, simplifying the generation of complex 3D structures. Each generated model considers the entire composite geometry, including the thin resin layers and variations of out-of-plane ply orientations. 2D Simulations were executed in parallel, allowing a B-scan inspection of over 100 individual measurements to be constructed in under 10 minutes. The inspection performance for varying tapered component geometries is evaluated, along with investigating the ability to detect defects, such as delamination’s. Additional simulations were performed to highlight the capability of simulation to optimize inspection techniques for these challenging structures.
Télécharger le document.
11:50
WE.1.C.4
Ultrasonic NDT and numerical simulation of adhesively bonded aluminium to CFRP materials
Damira Smagulova (Ultrasound Research Institute of Kaunas University of Technology)
The use of bonded aluminium alloy to carbon fiber reinforced plastics (CFRP) has become very popular, particularly in aviation and automotive industries due to increased mechanical performance, high fatigue strength, better damage tolerant and structure lightweight in order to increase reliability, save energy consumption and reduces pollutant emission. A junction between two dissimilar materials is the weakest part of the structure. Delaminations are common defects in adhesively bonded materials which affect the integrity of the entire structure. However, detection of delaminations is a challenging task because of different properties of dissimilar materials and possible complex geometry of the object. The objective of this investigation is delamination detection using ultrasonic non-destructive methods and numerical study of probability of defect detection (POD). Numerical simulation was performed to examine the object using various ultrasonic inspection methods and get the expected behavior of ultrasonic wave and results. POD estimation was done from the both sides, one is metal and other is composite. The adhesively bonded aluminium to CFRP was inspected experimentally using immersion and contact pulse echo techniques. Transducer frequencies were selected based on wavelength of the signal in different materials, their thickness, velocity, propagation distance, attenuation characteristics and expecting defects. A-scans and spectrum of received signals of different frequency transducers were analysed. Multiple reflections from not defected and defected bonding zones, transmission and reflections coefficients were studied and estimated. Post-processing of the data collected from the inspection was performed in order to increase the quality of defect imaging and it’s distinguishing on ultrasonic scans. Results of inspections from both metal and composite sides were compared. As a result phased array and focused transducers of 10 to 15 MHz were used for the investigation as well as CIVA software was used for numerical simulations. The percent of energy reflected from defected bonding zone is greater only by 1% of the energy reflected from not defected bonding zone. All delaminations were detected in immersion and contact method with a different probability. After post-processing increased image quality of ultrasonic scans was achieved and more suitable technique configuration was determined.
Télécharger le document.
12:10
WE.1.C.5
Implicit-explicit scheme for the elastodynamic wave equation in plates
Hajer Methenni (CEA-list)
The objective of our work is to provide an efficient simulation tool for the propagation of elastic waves generated by piezoelectric sensors in thin stratified plates or shells, in the context of Structural Health Monitoring (SHM). This method of Non-Destructive Testing (NDT), aiming at monitoring in-situ and periodically a structure, is increasingly present in many industrial fields, particularly in the aeronautics sector. Efficient simulation tools will be therefore increasingly needed for the transient simulations of elastic wave propagation. Such simulations can be used to improve the interpretations of the reception signals, to calibrate optimal control configuration or as building blocks of imaging processes. Natural discretization procedures, for instance, based on low-order finite elements and explicit schemes can be really costly especially because of the relative small thickness of the plate. Based on high-Order spectral finite elements, we propose an efficient time discretization that treats explicitly the propagative phenomena in the in- plane direction and implicitly the phenomena occurring through the thickness. In our algorithm, the time step can be chosen independently of the discretization parameters along the thickness (this specific point is the bottleneck of explicit methods). The price to pay is to solve at each iteration a linear system which is decoupled into several small linear systems (that can be solved efficiently in parallel) for each interpolation point on the in-plane surface of the plate. We will present numerical results showing the efficiency and accuracy of our approach on realistic applications. Finally we will discuss the potential extension of our method to plates with smoothly varying thickness or shells with small curvature.
Télécharger le document.
12:30
Lunch - Buffet

Modelling & Simulation I

Session chairs :
Room : AUDIT C
14:00
WE.2.A.1
Half-Space NDT/NDE through Multi-Frequency Bayesian Compressive Sensing
Marco Salucci (ELEDIA Research Center)
Microwave radiation can be successfully exploited to perform contactless inspection of dielectric structures for NDT/NDE purposes. As a matter of fact, microwave imaging (MI) approaches can retrieve both qualitative (i.e., number, location, and shape) and quantitative (i.e., material composition) information on unknown targets embedded inside an inaccessible domain by means of low-cost, portable, and safe (i.e., non-ionizing) equipment, representing a valid alternative to well-established technologies such as ultrasound (UT) and eddy current testing (ECT). However, solving the arising inverse scattering (IS) problem is a challenging task, requiring the development of suitable countermeasures to paramount issues such as ill-posedness and non-linearity. It is worth observing that typical NDT/NDE probing scenarios are characterized by strongly aspect-limited measurement setups (being often mathematically modelled as half-space configurations), allowing to collect a limited amount of non-redundant data on the imaged domain. To overcome such a lack of information, wide- band electromagnetic sources can be effectively exploited to enable the exploitation of multi-frequency (MF) data and yield regularized solutions thanks to the arising frequency diversity. Moreover, additional a-priori information on the solution can be enforced once the type/class of targets is known (e.g., cracks, delaminations, rust, ...). Indeed, sparseness-promoting approaches formulated within the compressed sensing (CS) paradigm proved to yield stable solutions of the IS problem once a suitable expansion basis is selected to represent the unknowns with few non-null coefficients. Within this context, Bayesian CS (BCS) approaches recently emerged as a computationally efficient solution strategy not requiring the compliance of the restricted isometry property (RIP) of the involved kernel/measurement operator. Accordingly, this work presents a novel MI methodology to perform NDT/NDE inspections of dielectric materials based on the effective integration of a BCS solution strategy and a MF scheme. Thanks to the BCS, sparseness priors are effectively exploited to regularize the IS problem, which is formulated within the contrast source inversion (CSI) framework to avoid the use of linear approximations (e.g., Born) and enable the reconstruction of non-weak scatterers. Moreover, the existing correlation among several spectral components of the collected wide-band data is efficiently and effectively exploited thanks to a multi-task (MT-BCS) inversion approach. Numerical experiments are presented to validate the potentialities as well as the current limitations of the proposed approach.
Télécharger le document.
14:20
WE.2.A.2
Endoscopic spatial phase shift shearography with the interferoskop for turbine blade flaw inspection
M. Eng. Christopher Petry (University of Applied Sciences Trier)
Shearography has already established for NDT applications in the field of aerospace. Compared to thermogra-phy, which is commonly used in this area, the material-independent working shearography offers several ad-vantages, such as for testing impact damage, delamination, crack, void, and undulation in composites. Howev-er, the major disadvantage of shearography is its susceptibility to failure in the presence of vibrations and envi-ronmental disturbances. In the OGKB laboratory of the University of Applied Sciences in Trier, the method is now being stabilized in a way that allows it to be used in raw industrial applications for the first time. The basis for this is the method of spatial phase shifting (SPS) according to the carrier frequency method, which, thanks to the further developed camera technology, firstly delivers representable results. The implementation of this method is presented in the patented measuring device Interferoskop 5.1, which combines shearography with endoscopy for difficult to access areas. The Interferoskop consists of a housing, which contains the camera and the optical setup. The housing is designed for hand-held inspection and can be equipped with different endoscope tips. The laser illumination as well as the object monitoring is performed through the endoscope. The aim of this development project is to create a measuring device for the testing of turbines by spyholes, thus a high time and cost effort by disassembling the turbines can be avoided. The implemented spatial phase shift method allows deformation measurements to be made in the video clock of the camera, which enormously increases process stability and at the same time permits dynamic component excitations. The Interferoskops performance for NDT is demonstrated by the inspection of defective turbine blades. Hereby, simply the cooling behaviour of the turbine blades can be used as active object excitation for the flaw check, while no additional excitation system is required. The new optical miniature design in 3D print construction enables maximum light efficiency and full shear-sensitivity adjustment while maintaining optimal result quality. In addition, the optical setup can be extended by a simple modification for holographic defor- mation measurement. Furthermore, first tests were made in hand-held operation of this micrometre accurate measuring method.
Télécharger le document.
14:40
WE.2.A.3
Neutron and synchrotron X-ray measurements : unique tools in the non-destructive toolbox
Caroline Boudou (ILL)
Neutron and X-ray beams provide unique, powerful and non-destructive access into the heart of materials and components, over length scales ranging from the centimetre to the atom. The Institut Laue-Langevin ILL – Europe’s leading neutron source - and the ESRF – Europe’s leading synchrotron light source, combine 80 experimental stations providing neutron and X-ray beams with exceptional properties to overcome the limitations of standard laboratory characterisation techniques. Regarding challenges in the aeronautic and space sectors, we will detail some advantages obtained by synchrotron and neutron beams in stress mapping and computed tomography. The ESRF’s high-energy synchrotron X-rays can map efficiently the strain and stress at the surface and up to several hundreds of microns below to give depth profiling with high spatial resolution. Neutrons give access to the full stress tensor from 60μm below the surface up to the interior of the component (e.g. 30 cm in Al, 6 to 7 cm in Ti, Ni or steel). Both techniques are able to measure in real time and non-destructively the stress distribution during mechanical loading, or during heating/cooling processes. A current exciting development is the investigation of residual stress build-up during additive manufacturing. Here, we will show examples of measurements taken from aircraft and satellite components. Note that a combination of neutron and X-ray datasets can be offered by the institutes. With regards to computed tomography, synchrotron X-rays and neutrons provide complementary information. Synchrotron X-rays provide remarkable insights into materials up to the nanoscale with a resolution as high as 30 nm. At the microscale, synchrotron X-rays enable vastly more accurate (eg. with phase contrast imaging and lower beam hardening effect) and faster data collection than lab equipment – 2D radiography can be performed at MHz rate for monitoring rapid phenomena and 3D CT can be done at ~Hz speed. With these capabilities, it is possible to monitor dynamic processes in real time (e.g. crack propagation, exploding fuse or bed fusion in additive manufacturing). Neutron tomography makes use of the specific properties of thermal neutrons to penetrate deeply a large majority of materials (e.g. most metals) whilst being highly sensitive to other elements, notably hydrogen and lithium, enabling the imaging of water, polymers or explosive charges embedded in metallic structures, without metal induced image artefacts neither radiation damage. We will conclude our presentation by briefly showing how further neutron and synchrotron X-rays techniques can address issues in composites, electrical components and radiation
Télécharger le document.
15:00
WE.2.A.4
Non-destructive quality evaluation of additively-manufactured components
Sam Yang (CSIRO)
Additive metal manufacturing (AM) offers distinctive advantages over conventional subtractive manufacturing processes such as machining from metal castings. AM is capable of producing complex high-performance components with shapes that are practically impossible to achieve with conventional processes. The design flexibility offered by AM is particularly important for high-value applications such as aerospace, where the performance-to-weight ratio is critical. AM components require little or no additional machining or other processing steps. This reduces both the manufacturing cost and environmental impact through reduced wastage. However, due to the specific conditions of the AM process, AM components often have microscopic structural defects such as porosity that are difficult to detect non-destructively with current off-the-shelf technology. Microscopic structural defects affect the fit-for-purpose performance of the components. These defects typically manifest as microscopic porosity distributions. Available non-destructive testing (NDT) technologies have major shortcomings. For instance, the BET (Brunauer-Emmett-Teller) analysis is unable to determine the location of defects, and the conventional X-ray CT (computed tomography) imaging analysis cannot resolve the microscopic porosity defects in macro-sized parts, due to limitations in resolving the multiple length scales. In the last ten years, CSIRO has developed a data-constrained modelling technology. Together with quantitative X- ray CT, it allows the resolution of fine defects in AM metal components. This has been demonstrated using test samples and synchrotron-based X-ray CT imaging. The technology is being further developed NDT of with macro- sized AM metal components using industrial X-ray CT facilities.
Télécharger le document.
15:20
WE.2.A.5
Real-time bolt looseness characterization using EMI-based PZT sensing network : a GCN perspective
Lu Zhou - National Engineering Research Center on Rail Transit Electrification and Automation (Hong Kong Branch)
Bolted joints are the most vital parts in civil, mechanical and aerospace engineering connecting assemblies and transferring loads. Although subtle and frequently seen, the loosening of bolts can potentially lead to critical consequences, particular in transportation sector. Real-time monitoring of bolted joint conditions and identification of bolt looseness is therefore highly desired. Previous research has demonstrated that the bolt looseness can be identified through a variety of Non-destructive Testing (NDT) techniques including ultrasounds, optic-fiber sensors, vison-based approaches, etc. The Electro-mechanical Impedance measuring method has been proved as an effective NDT technique in detecting the bolt looseness with a single PZT sensor. The method is conducted by exciting the PZT element implemented on the host bolt structure with a series of periodic voltages at the ultrasonic frequency band and collect dynamic responses (electrical-mechanical impedance) from the PZT patch. Despite of the simplicity, the EMI method has been criticized of lacking exact correspondence between the change of EMI data and the structural variation, and therefore has been treated as more of a qualitative validation rather than a quantitative choice. Machine learning has been incorporated in some studies for more accurate EMI-based monitoring but the how a machine learning algorithm can be introduced to best reveal the value of the obtained EMI data is worth further investigating. This paper describes an experimental and data- driven study in quantitatively characterization of the residue torque and the loosening of a bolted joint with a pre- designed and implemented PZT sensing network. A Graph Convolutional Network (GCN) model is established for the PZT sensors. The GCN model is trained using the relationship between data from neighboring sensors and the topology structure of the sensing network rather than data from individual sensors, in a way to more precisely interpret the EMI signatures and correspond the variations to the bolt looseness with a quantitative confidence.
Télécharger le document.
14:00
WE.2.A
Advances in NDT Methods II
AUDIT A

Advances in NDT Methods II

Session chairs :
Room : AUDIT A
14:00
WE.2.A.1
Half-Space NDT/NDE through Multi-Frequency Bayesian Compressive Sensing
Marco Salucci (ELEDIA Research Center)
Microwave radiation can be successfully exploited to perform contactless inspection of dielectric structures for NDT/NDE purposes. As a matter of fact, microwave imaging (MI) approaches can retrieve both qualitative (i.e., number, location, and shape) and quantitative (i.e., material composition) information on unknown targets embedded inside an inaccessible domain by means of low-cost, portable, and safe (i.e., non-ionizing) equipment, representing a valid alternative to well-established technologies such as ultrasound (UT) and eddy current testing (ECT). However, solving the arising inverse scattering (IS) problem is a challenging task, requiring the development of suitable countermeasures to paramount issues such as ill-posedness and non-linearity. It is worth observing that typical NDT/NDE probing scenarios are characterized by strongly aspect-limited measurement setups (being often mathematically modelled as half-space configurations), allowing to collect a limited amount of non-redundant data on the imaged domain. To overcome such a lack of information, wide- band electromagnetic sources can be effectively exploited to enable the exploitation of multi-frequency (MF) data and yield regularized solutions thanks to the arising frequency diversity. Moreover, additional a-priori information on the solution can be enforced once the type/class of targets is known (e.g., cracks, delaminations, rust, ...). Indeed, sparseness-promoting approaches formulated within the compressed sensing (CS) paradigm proved to yield stable solutions of the IS problem once a suitable expansion basis is selected to represent the unknowns with few non-null coefficients. Within this context, Bayesian CS (BCS) approaches recently emerged as a computationally efficient solution strategy not requiring the compliance of the restricted isometry property (RIP) of the involved kernel/measurement operator. Accordingly, this work presents a novel MI methodology to perform NDT/NDE inspections of dielectric materials based on the effective integration of a BCS solution strategy and a MF scheme. Thanks to the BCS, sparseness priors are effectively exploited to regularize the IS problem, which is formulated within the contrast source inversion (CSI) framework to avoid the use of linear approximations (e.g., Born) and enable the reconstruction of non-weak scatterers. Moreover, the existing correlation among several spectral components of the collected wide-band data is efficiently and effectively exploited thanks to a multi-task (MT-BCS) inversion approach. Numerical experiments are presented to validate the potentialities as well as the current limitations of the proposed approach.
Télécharger le document.
14:20
WE.2.A.2
Endoscopic spatial phase shift shearography with the interferoskop for turbine blade flaw inspection
M. Eng. Christopher Petry (University of Applied Sciences Trier)
Shearography has already established for NDT applications in the field of aerospace. Compared to thermogra-phy, which is commonly used in this area, the material-independent working shearography offers several ad-vantages, such as for testing impact damage, delamination, crack, void, and undulation in composites. Howev-er, the major disadvantage of shearography is its susceptibility to failure in the presence of vibrations and envi-ronmental disturbances. In the OGKB laboratory of the University of Applied Sciences in Trier, the method is now being stabilized in a way that allows it to be used in raw industrial applications for the first time. The basis for this is the method of spatial phase shifting (SPS) according to the carrier frequency method, which, thanks to the further developed camera technology, firstly delivers representable results. The implementation of this method is presented in the patented measuring device Interferoskop 5.1, which combines shearography with endoscopy for difficult to access areas. The Interferoskop consists of a housing, which contains the camera and the optical setup. The housing is designed for hand-held inspection and can be equipped with different endoscope tips. The laser illumination as well as the object monitoring is performed through the endoscope. The aim of this development project is to create a measuring device for the testing of turbines by spyholes, thus a high time and cost effort by disassembling the turbines can be avoided. The implemented spatial phase shift method allows deformation measurements to be made in the video clock of the camera, which enormously increases process stability and at the same time permits dynamic component excitations. The Interferoskops performance for NDT is demonstrated by the inspection of defective turbine blades. Hereby, simply the cooling behaviour of the turbine blades can be used as active object excitation for the flaw check, while no additional excitation system is required. The new optical miniature design in 3D print construction enables maximum light efficiency and full shear-sensitivity adjustment while maintaining optimal result quality. In addition, the optical setup can be extended by a simple modification for holographic defor- mation measurement. Furthermore, first tests were made in hand-held operation of this micrometre accurate measuring method.
Télécharger le document.
14:40
WE.2.A.3
Neutron and synchrotron X-ray measurements : unique tools in the non-destructive toolbox
Caroline Boudou (ILL)
Neutron and X-ray beams provide unique, powerful and non-destructive access into the heart of materials and components, over length scales ranging from the centimetre to the atom. The Institut Laue-Langevin ILL – Europe’s leading neutron source - and the ESRF – Europe’s leading synchrotron light source, combine 80 experimental stations providing neutron and X-ray beams with exceptional properties to overcome the limitations of standard laboratory characterisation techniques. Regarding challenges in the aeronautic and space sectors, we will detail some advantages obtained by synchrotron and neutron beams in stress mapping and computed tomography. The ESRF’s high-energy synchrotron X-rays can map efficiently the strain and stress at the surface and up to several hundreds of microns below to give depth profiling with high spatial resolution. Neutrons give access to the full stress tensor from 60μm below the surface up to the interior of the component (e.g. 30 cm in Al, 6 to 7 cm in Ti, Ni or steel). Both techniques are able to measure in real time and non-destructively the stress distribution during mechanical loading, or during heating/cooling processes. A current exciting development is the investigation of residual stress build-up during additive manufacturing. Here, we will show examples of measurements taken from aircraft and satellite components. Note that a combination of neutron and X-ray datasets can be offered by the institutes. With regards to computed tomography, synchrotron X-rays and neutrons provide complementary information. Synchrotron X-rays provide remarkable insights into materials up to the nanoscale with a resolution as high as 30 nm. At the microscale, synchrotron X-rays enable vastly more accurate (eg. with phase contrast imaging and lower beam hardening effect) and faster data collection than lab equipment – 2D radiography can be performed at MHz rate for monitoring rapid phenomena and 3D CT can be done at ~Hz speed. With these capabilities, it is possible to monitor dynamic processes in real time (e.g. crack propagation, exploding fuse or bed fusion in additive manufacturing). Neutron tomography makes use of the specific properties of thermal neutrons to penetrate deeply a large majority of materials (e.g. most metals) whilst being highly sensitive to other elements, notably hydrogen and lithium, enabling the imaging of water, polymers or explosive charges embedded in metallic structures, without metal induced image artefacts neither radiation damage. We will conclude our presentation by briefly showing how further neutron and synchrotron X-rays techniques can address issues in composites, electrical components and radiation
Télécharger le document.
15:00
WE.2.A.4
Non-destructive quality evaluation of additively-manufactured components
Sam Yang (CSIRO)
Additive metal manufacturing (AM) offers distinctive advantages over conventional subtractive manufacturing processes such as machining from metal castings. AM is capable of producing complex high-performance components with shapes that are practically impossible to achieve with conventional processes. The design flexibility offered by AM is particularly important for high-value applications such as aerospace, where the performance-to-weight ratio is critical. AM components require little or no additional machining or other processing steps. This reduces both the manufacturing cost and environmental impact through reduced wastage. However, due to the specific conditions of the AM process, AM components often have microscopic structural defects such as porosity that are difficult to detect non-destructively with current off-the-shelf technology. Microscopic structural defects affect the fit-for-purpose performance of the components. These defects typically manifest as microscopic porosity distributions. Available non-destructive testing (NDT) technologies have major shortcomings. For instance, the BET (Brunauer-Emmett-Teller) analysis is unable to determine the location of defects, and the conventional X-ray CT (computed tomography) imaging analysis cannot resolve the microscopic porosity defects in macro-sized parts, due to limitations in resolving the multiple length scales. In the last ten years, CSIRO has developed a data-constrained modelling technology. Together with quantitative X- ray CT, it allows the resolution of fine defects in AM metal components. This has been demonstrated using test samples and synchrotron-based X-ray CT imaging. The technology is being further developed NDT of with macro- sized AM metal components using industrial X-ray CT facilities.
Télécharger le document.
15:20
WE.2.A.5
Real-time bolt looseness characterization using EMI-based PZT sensing network : a GCN perspective
Lu Zhou  (National Engineering Research Center on Rail Transit Electrification and Automation - Hong Kong)
Bolted joints are the most vital parts in civil, mechanical and aerospace engineering connecting assemblies and transferring loads. Although subtle and frequently seen, the loosening of bolts can potentially lead to critical consequences, particular in transportation sector. Real-time monitoring of bolted joint conditions and identification of bolt looseness is therefore highly desired. Previous research has demonstrated that the bolt looseness can be identified through a variety of Non-destructive Testing (NDT) techniques including ultrasounds, optic-fiber sensors, vison-based approaches, etc. The Electro-mechanical Impedance measuring method has been proved as an effective NDT technique in detecting the bolt looseness with a single PZT sensor. The method is conducted by exciting the PZT element implemented on the host bolt structure with a series of periodic voltages at the ultrasonic frequency band and collect dynamic responses (electrical-mechanical impedance) from the PZT patch. Despite of the simplicity, the EMI method has been criticized of lacking exact correspondence between the change of EMI data and the structural variation, and therefore has been treated as more of a qualitative validation rather than a quantitative choice. Machine learning has been incorporated in some studies for more accurate EMI-based monitoring but the how a machine learning algorithm can be introduced to best reveal the value of the obtained EMI data is worth further investigating. This paper describes an experimental and data- driven study in quantitatively characterization of the residue torque and the loosening of a bolted joint with a pre- designed and implemented PZT sensing network. A Graph Convolutional Network (GCN) model is established for the PZT sensors. The GCN model is trained using the relationship between data from neighboring sensors and the topology structure of the sensing network rather than data from individual sensors, in a way to more precisely interpret the EMI signatures and correspond the variations to the bolt looseness with a quantitative confidence.
Télécharger le document.
WE.2.B
Structural Health Monitoring I (cont.) / STUDENTS CHALLENGE I
AUDIT B

Structural Health Monitoring I (cont.) / STUDENTS CHALLENGE I

Session chairs :
Room : AUDIT B
14:00
WE.2.B.1
Strain and damage sensing by CNT modified adhesive films and fiber optic distributed sensing. Comparison of performances in a double lap bonded joint
Alfredo Guemes (UPM)
Bonded joints show a complex stress field, with large strain gradients, difficult to acquire experimentally, which would be very useful to validate numerical models; normally DIC is used for it. Two experimental techniques are proposed at this paper, and their feasibility and performances are compared. The properties of CNT doped adhesive films are well known for uniform strain field; by plotting conductive patterns on the adhesive film, the responses at the high and low shear strain regions are obtained, giving a much better insight on the changes in electrical conductivity under complex strain field. An optical fibre embedded into the adhesive film, aligned to the loads, experience the strong strain gradients that are predicted by the models. Experimental results are compared to the numerical predictions given by Volkorsen model. The double lap joint is selected to avoid the bending moments caused by load misalignments, with strong out-of-plane displacements that frequently mask strains caused by the pure shear loads. With the double lap, the symmetry of loads simplify the response, only the shear stresses are present, with a minor component of peeling stresses at the discontinuities, Strain measurements were done at several load levels and at two different temperatures, to validate the temperature compensation algorithms. Partial damage are then be inflicted at the adhesive joint, and the ability of the two techniques to detect early damages is quantified. It has been reported previously that partial damages cause very slight changes at the strain field, and consequently they would be undetected by an strain measurement technique. Nevertheless, we have found experimental evidence of changes in the electrical response in case of partial debonding, and even in case of improperly treated surfaces. More tests are on progress, and results will be reported.
Télécharger le document.
14:20
WE.2.B.2
Interpreting results of NDT/SHM methods based on elastic guided waves in composite plate-like structures by a semi-analytic modal simulation tool
Jordan Barras (CEA list)
Elastic guided waves (GW) are valuably used for NDT/SHM of metallic or composite plate-like structures. They allow fast scanning of large parts for the detection of all kinds of inner or surface defects and flaws. However, their complex behaviour is quite challenging as far as solving the inverse problem of defect characterization from a set of measured signals is concerned. Their modal nature and dispersion properties lead to strong variations of signal time-history all along the wavepaths, even more since reflections (at boundaries with possible mode conversion) occur between a source and a receiver. Furthermore, transducers diffraction effects add some complexity to the formation of GW field. Consequently, designing optimal NDT/SHM configurations for a given part, accurately interpreting measured signals, make it necessary to grasp the various phenomena that arise. Simulation obviously constitutes an efficient and quantitative way of doing it. Here, results obtained by means of a newly developed model are presented, to demonstrate how the understanding of complex waveforms typically measured in NDT/SHM experiments is greatly simplified as soon as the modal nature of the various wavepackets can be exhibited and understood. The presentation briefly reviews the semi-analytic model developed, which accounts for multiple-reflections at boundaries of arbitrary shape and transducer diffraction effects. Then, predictions of typical complex wavefields in realistic configurations are shown. They are first computed by means of a time-dependent in-house finite element code. Then, they are fully explained thanks to the modal interpretation gained in reproducing them by means of the semi-analytic model.
Télécharger le document.
14:40
Team 1
STUDENTS CHALLENGE: NDT Characterization of a CFRP Multi-curved component.
Co-authors : Hena Bata Nkirda (Team Leader) | Koh, Myungjun (Team Member) | Labanie, MHD. Fateh (Team Member) | Yuan, Mingchun (Team Member) |
15:00
Team 2
STUDENTS CHALLENGE: Non-contact damage detection in composite laminates using laser-generated ultrasounds
Co-authors : Jakub Spytek (1) | Patrycja Pyzik (1) | Jakub Mrówka (1) | Łukasz Pieczonka (1) | Łukasz Ambroziński (1) 1. AGH University of Science and Technology in Kraków
15:20
Team 3
STUDENTS CHALLENGE: Modelling and Detection of Defects in CFRP Using Electromagnetic and Thermographic Techniques
Co-authors : Junzhen Zhu (1) | Abdoulaye Laye Ba (2) | Qiuji Yi (1) | Adi Marindra (1) | Gui Yun Tian (1) Gerard Berthiau (2) | Huu Kien Bui (1) 1.School of Engineering, Newcastle University 2.Institute of Research in Electrical Energy of Nantes-Atlantique
WE.2.C
Modelling & Simulation II
AUDIT C

Modelling & Simulation II

Session chairs :
Room : AUDIT C
14:00
WE.2.C.1
Parametric study by simulation of a thermography control by induction heating. Application to the detection and characterization of cracks
Olivier Ghibaudo (Safran Tech)
In the last few years, induction thermography has been identified as a non-destructive testing method for detecting and characterizing surface cracks in metals. The sample to be inspected is heated with a short induced electrical current pulse, and the infrared camera records the temperature distribution and transient temporal behavior at the surface during and after the heating pulse. In this work, 3D Finite element simulations, performed with the software FLUX®, were carried out to investigate how the contrast depends on parameters such as excitation frequency, pulse duration, material parameters, crack depth and length. To compare the simulations with the experimental results, the modeled inductor is a realistic U-ferritic yoke surmounted by a copper winding. The two block samples involve two materials massively used in aeronautics: the first one is paramagnetic (nickel based superalloy INCONEL 718), the second one is ferromagnetic (low carbon steel 16CND13), which implies very different skin depths. The simulations take into account the non-linearity of the sample magnetic permeabilities, which makes it possible to calculate finely the total impedance of the inductor. For a current imposed excitation through the inductor, this allows to normalize the observations with respect to the active and reactive power consumed by the inductor. From the temperature distributions at each time step, the analysis of the differential thermal contrast between a defect situation and a defect-free situation, emphasizes the detectability of the defect as a function of the introduced parametric variations. The thermal extractions provided are used firstly to find the optimal excitation parameters (frequency of the induced currents, heating duration) which maximizes the thermal contrast, and on the other hand to define a methodology for sizing defects according to their lengths and depths. The optimum contrast calculated on the phase of the Fourier Transform (Pulse Phase Thermography PPT) provides credible but sensitive information related to the excitation parameters and the geometric parameters of the defects. In the final paper, two methods of analysis will be compared: the PPT and the wavelet pulsed phased thermography.
Télécharger le document.
14:20
WE.2.C.2
Semi-analytical modeling of eddy current inspection for anisotropic material presenting rough interfaces
Houssem Chebbi (CEA list)
Within our department for NDT in the CEA LIST institute, this PhD work is a part of the European project NDTonAir (H2020 Marie Sklodowski-Curie European Training Network) which adresses several issues of NDT in the aeronautic field involving different methods of NDT based on several testing technics as ultrasonic waves, electromagnetic waves or infrared thermography. As part of the project, we are interested in the interaction of a composite material with an eddy current sensor. The goal is to develop a robust, fast and accurate numerical model based on a semi-analytic approach to optimize ECNDT methods for this type of material. This model will be implemented in the commercial platform CIVA developed by CEA LIST. The first part of this work was published in the previous proceedings of the 10th international symposium of NDT in aerospace held in Dresden. Here we propose the second part in which we deal with the fast computation of a quasi-static fields induced by a 3D eddy current probe in a stratified and homogenized anisotropic material with a local 2D5/3D perturbation in the geometry representing a kind of delamination in the material. Subsequently, the original part of the thesis lies in taking into account local deformations of the geometry by applying an original approach called Curvilinear Coordinates Method or CCM which consists in considering a change of coordinates to fit the deformation of the interface and in solving the covariant form of Maxwell’s equations in a new transformed coordinates system using a modal approach. Hence, the complex stratified structure is transformed into a plane structure in the new coordinate system in which the boundary conditions are described in an analytical way. Therefore, no mesh is necessary to maintain local variations of the fields in the structure. Finally, the validation of the model yields in comparing FE data to simulated data provided by the numerical model. Further details and more results will be presented in the full paper.
Télécharger le document.
14:40
WE.2.C.3
Reducing NDT Effort by Coupled Monitoring and Simulation of Liquid Composite Molding Processes
Nico Liebers (German Aerospace Center, Institute of Composite Structures and Adaptive Systems)
Liquid Composite Molding (LCM) is a manufacturing technique which could reduce material and process costs in comparison to the widely used prepreg processes. The raw materials are less expensive and easier to store. Also relatively high layup rates of dry fiber textiles can be achieved. But LCM requires more sophisticated process design due to their sensitivity to deviations in the raw materials or process setup. For example a larger gap around the fiber preform can lead to trapped gas in the middle of the part. Such gas inclusions can lead to porosity or even larger dry spots and thus an unusable part. At the DLR Institute of Composite Structures and Adaptive Systems adapted and low cost ultrasonic sensors for LCM monitoring were developed. The sensors do not require direct contact to the part and therefore do not affect the part surface or mold. Due to enhanced signal analysis not only the flowfront arrival can be detected, but also its flow speed. Further the moments of wetting of the mold surface and the infiltration through the part thickness can be distinguished. The sensor system is also able to perform cure monitoring including the detection of gelation and vitrification points. These sensor inputs are coupled with a fast flow simulation where the model parameters are adapted to match the measurement by a sophisticated optimization algorithm. The most important model parameters are the permeability and gap sizes. By adapting the simulation the pointwise monitoring results are transformed into a two-dimensional representation of the flowfront and enables the prediction of its further course. The results are used to detect gas inclusions and areas of high porosity. If in a LCM process none of these defects are detected the NDT process could be omitted. In the case of detected defects only these areas would have to be tested potentially.
Télécharger le document.
15:00
WE.2.C.4
The use of free-space microwave non-destructive techniques: simulation of damage detection in carbon fibre reinforced composites.
Daniela Munalli (The University of Nottingham)
Microwave non-destructive testing (MNDT) methods represent an effective solution in detecting defects within composite structures with relatively low electrical conductivity. They offer the advantage to overcome the problems of traditional NDT techniques such as coupling, danger coming from ionizing radiation, limited depth of operation, large wavelengths, time consuming post processing. Near-field microwave and millimetre non-invasive inspections have been successfully used for detecting defects such as disbond and delamination in complex structures. In dielectric materials, they can be used for dielectric properties characterization, degree of porosity evaluation, degree of ageing, anisotropy, dielectric mixture constituents determination, state of cure. When it comes to the analysis of carbon fibre reinforced polymers (CFRPs), the use of these non-destructive techniques is restricted by the composite relatively high conductivity of about 104 S/m. In this paper, the investigation of a free-space microwave method for non-destructive testing of unidirectional CFRPs has been carried out by means of a pair of standard gain horn antennas, covering a frequency range from 26.5 GHz to 40 GHz. With the simulations, experimental results related to the presence/severity of the analysed defects are linked to the variations of the measured scattering parameters Sij. The approach is based on the comparison between the electromagnetic signal reflected and transmitted through a healthy sheet material under test, when a radio frequency (RF) wave is incident on it, and the one reflected and transmitted by a damaged sheet. The eventual presence of the defect is revealed by measuring the mismatch between the two transmitted waveforms. The performance of this radio wave technique is investigated in relation to surface defects and also in relation to those types of defects that are less detectable with this method, such as delaminations, cavities and inclusions. The simulations make use of the finite integration technique (FIT) and the finite element method (FEM).
Télécharger le document.
15:20
WE.2.C.5
Non-Destructive Component Inspection and Numerical Simulation : Applications to Additive Manufacturing
David Harman (Synopsys)
Additive manufacturing (AM) gives aerospace part manufactures advantages in terms of flexibility when designing components. In addition, AM methods help with traditional challenges such as prototyping or testing multiple design iterations, as well as for optimizing structures with lattices for weight-saving. However, it can be difficult to quantify defects in AM parts, for example during the printing stage. Non-destructive inspection and numerical simulation methods have therefore been developed to reduce uncertainty by comparing as-designed CAD models to models generated from as-manufactured parts scanned using computed tomography (CT). Synopsys, ANSYS, and North Star Imaging collaborated on a workflow for going from a 3D CT scan of an AM part to a Finite Element (FE) simulation. In this presentation we will discuss the key steps in this process, from scanning of a part through to image segmentation and processing in Synopsys Simpleware™ software. We will show how surface deviation tools in Simpleware software allowed for comparison of the CAD and scanned model to identify and inspect defects. The presentation will also look at how FE simulation in ANSYS software made it possible to evaluate how the CAD and CT-based models perform under specific conditions. To understand the results and the success of this workflow, we will give an example of its use by Moog, Inc., for designing an impulse pressure manifold. This workflow involves inspection of residual problems from the AM process, and how the CT to simulation method identified unforeseen defects and enabled Moog to assess the usability of an AM-produced part. The overall goal of this presentation is to consider the benefits of an integrated workflow for CT-based inspection and simulation of AM parts, with broader applications to traditional manufacturing and design challenges. We will also discuss the value of this method for aerospace companies in terms of reducing the risk of design uncertainty and manufacturing errors, as well as potential future developments in using these techniques.
Télécharger le document.
15:40
Coffee Break

Modelling & Simulation II

Session chairs :
Room : AUDIT C
16:10
WE.3.A
NDT Methods - Magnetics & Eddy Current
AUDIT A

NDT Methods - Magnetics & Eddy Current

Session chairs :
Room : AUDIT A
16:10
WE.3.A.1
Micro-magnetic non-destructive testing on extreme condition aged samples.
Benjamin Ducharne (INSA Lyon)
Creep phenomenon is an important feature to assess in extreme condition applications, although the correlations with microstructure and magnetic behavior remain unclear. Conventional Eddy current testing has been extensively used for the ferromagnetic materials characterization but when it comes to creep damage detection, it becomes difficult to distinguish between the changes caused by the actual creep damage and from the signals generated by other sources like cracks, surface roughness, hardness, etc... In this work, 12%Cr-Mo-W-V creep test samples are investigated using three electromagnetic inspection techniques. Magnetic parameters based on the results are then evaluated in comparison to the microstructure. Additionally, a modified Jiles-Atherton model has been used to numerically reproduce experimental results from Magnetic Incremental Permeability (MIP), Magnetic Barkhausen Noise (MBN) and standard B(H) measurements. All the three techniques exhibit different responses in understanding creep and the modeling parameters derived from the adapted Jiles-Atherton model parameters are then correlated to the microstructure information. Some suitable parameters are then shortlisted according to the application technique. Coupling between the stress and magnetic field is the main and important feature of the ferromagnetic materials consisting of various small magnetic domains in its microstructure [1][2]. Magnetic Incremental Permeability (MIP) is used to investigate samples as it is highly sensitive to stress. On the other hand, Magnetic Barkhausen Noise being sensitive to the mechanical changes in the materials is also used to analyze the samples in addition to standard B(H) curve measurements. Finally, ferromagnetic hysteresis models such as dry friction quasi-static model [3], Preisach model [4], Jiles-Atherton model [5], which are based on magnetic induction B versus applied magnetic field strength H, are implemented to get the simulated data based on experiments. The objectives of these simulations are to improve magnetic signatures interpretations in corelation to microstructure. References [1] Schull P.J, New York. Marcel Dekker, Inc., 2002. [2] B. Ducharne and al., IEEE Trans. on. Mag, vol. 99, pp. 1-6, 2018. [3] B. Ducharne and al., J. of Phys. D: Applied Physics, vol. 40, Iss. 2, pp. 551-555, 2007. [4] B. Zhang and al., IEEE Trans. on. Mag, vol. 54 iss. 3, 2017. [5] B. Zhang and al., IEEE Trans. on. Mag, iss. 99, 2018.
Télécharger le document.
16:30
WE.3.A.2
Development of portable wireless non-destructive crack identification method by using GMR sensor array for overhead crane bridges
Arun Kumar Yadav (AGH University of Science &  Technology)
Structural Health monitoring of overhead cranes bridges by the traditional inspection system with wired and bulky instrumental technologies face many challenges during the harsh environment and under working condition. This paper proposes a portable wireless and efficient NDT method using the GMR sensor array technique for the identification of fatigue cracks in the bridges of travelling overhead cranes. In order to enhance the efficiency of overhead cranes by minimizing the inspection time, a portable wireless robot combined with GMR (Giant Magnetoresitivity ) sensor and NDT technique is presented. This novel solution offers mobility, high accuracy and low power consumption. For the detection of cracks and defects in overhead crane bridge eight GMR (Giant Magneto Resistive) sensors of NVE (AA006-02) placed linearly on a PCB board with equal distance. Two magnetic wheel of neodymium N42 located on both side of the GMR sensor array to magnetize the steel surface for accurate defect reorganization. Unlike MPI (Magnetic Particle Inspection) where a global magnetization require for further inspection, this automated detection system only magnetize the surface area under the vicinity of GMR sensor array. The instrumentation circuit including eight high speed multiplexer, Operational amplifier, and one 8 bit analog to digital converter. A PIC 877A microcontroller and raspberry pie was used to perform local data storage, data processing and controlling. This presented testing solution is quick and offers a step towards automated testing of overhead crane bridges. However it improves the work efficiency and can meet the serious challenges within the inspection of overhead crane bridges.
Télécharger le document.
16:50
WE.3.A.3
Sizing method of crack length on carbon steels for eddy current testing using transmitter-receiver probe
Shurui Zhang (Institute of Fluid Science, Tohoku University)
Transmitter-reciver(TR) probe is one of the most common eddy current testing (ECT) probe used for crack detection. The previous study showed that due to the existence of magnetic flux leakage (MFL) in carbon steels, the amplitude of ECT signal can be influenced by the scanning angle of the TR probe. In addition, the MFL works effectively on crack estimation because the effect of MFL makes the signal more sensitive to be detected. Quantitative measurement of the size of the crack is important for evaluating the extent of the damage to the structure and for timely repair. In the previous research, the 12dB down method was proposed for length sizing. Nevertheless, the evaluation of the crack size is not precise especially when the length of crack is shorter than the distance between transmitter and receiver coils, because the 12dB down method only roughly assumes the crack length by obtained ECT signals. In this study, to improve the precision of the crack sizing especially for short cracks, ECT signals were processed by using two functions: Gaussian function and unary quadratic function, and the relationship between ECT signal and crack length was investigated. The ECT signals were Gaussian fitted, and the relationship between the variance of the obtained Gaussian function and the crack length was calculated as the unary quadratic function. To investigate the difference of the precision of the length estimation due to the shape and the length of the cracks, two types of cracks, rectangular and elliptical, with different lengths were prepared on the carbon steel plate. In addition, the angle of TR probe was set to 0° and 90° in order to investigate the effect of the MFL on ECT signal. The precision of new signal processing method is evaluated by comparing the estimated length and the measured length of cracks. As a result, it is clarified that the new signal processing method can estimate the crack length more precisely than the previous 12dB down method especially for the short cracks. The errors between the measured and the estimated lengths of the elliptical cracks were smaller than that of the rectangular cracks because of the signals obtained from the rectangular cracks were too complicated to be fitted on Gaussian curve. Moreover, the error when the scanning angle is 90° is smaller than that of 0°.
Télécharger le document.
17:10
WE.3.A.4
MR sensors arrays for eddy current testing
N. Sergeeva-Chollet (CEA list)
Eddy Current (EC) Technique is a powerful method for detection of surface notches and of buried flaws during inspection of metallic parts. This technique is used for inspection of aerospace components. Classical winding coils are the most commonly used EC sensors. Nevertheless, when the size of flaws decreases or the defect is rather buried deep inside the material, traditional winding coil probes turn out to reach their limits. For this reason, other technologies are investigated to improve this technique. Magnetoresistive sensors present the advantages of flat frequency response and dimensions at the micron size. These sensors are hence very attractive for the detection of buried defects that require low frequencies because of skin depth effect. Also, they are suitable for small surface defects due to high spatial resolution because of their manufacturing down to μm without losing their field sensitivity. In addition, several magnetic field components can be simultaneously detected for a better analysis of defects. We will present last advances of MR based probes containing arrays of GMR- (giant magneto resistance) and TMR-elements (tunnel magneto resistance) with comparison of experimental datas and simulations on Aluminium and Titanium mock-ups. We will discuss the input brought by the development of ASIC components for compact probes and paths to develop probes adapted to complex geometries.
Télécharger le document.
17:30
WE.3.A.5
Eddy current testing of carbon composites in aerospace industry
Marie Bohacova (Indetec NDT)
Carbon composites due to their specific properties find applications in various industries, especially in aerospace industry. Widely used carbon fibre reinforced polymers (CFRP) have already been applied even for the aircraft primary structures. The development of advanced diagnostic techniques that are able to easily detect and identify the degradation of the carbon fibre material is still a challenge for various NDT methods. This paper describes the possibility to apply eddy current (EC) for testing of carbon composite structures. Two types of eddy current probes were developed and tested with excellent results. The new conventional eddy current probes are able to reliably and easily detect surface and subsurface discontinuities such as delamination and thickness variation. The probe setting parameters are described for different types of carbon composites (type of matrix and reinforcement, layup). Precise settings are necessary for the successful eddy current testing. It was determined that reliable detection of a minimum surface defect size is Ø1.5 mm for specimens and that eddy currents are able to penetrate into a thickness up to approximately 4 mm depending on the type of carbon composite. Additionally, this paper describes the comparison of eddy current testing with ultrasonic phased array method (PAUT). Composite aircraft structures are very susceptible to impact damage usually detected using PAUT. Therefore, sensitivity and resolution analysis of impact damage detection was performed by means of these two methods. It was found out that the results are very similar to each other when standard pulse-echo technique of PAUT was used. This carbon composite eddy current testing (CCECT) could fully replace or complement other NDT methods such as visual or ultrasonic inspections. CCECT could guarantee the product quality in production using the robotic system with appropriate visualisation especially for large area testing, also in aircraft maintenance and in the development of new aircraft structures.
Télécharger le document.
Students Challenge II
AUDIT B

Students Challenge II

Session chairs :
Room : AUDIT B
16:10
Team 4
A data-driven methodology for damage detection in a CFRP composite
Co-authors : Bilal Ali Qadri (1) | Artur Movsessian (2) | Callum Roberts (2) | 1.Aalborg Universitet Esbjerg 2.University of Strathclyde
16:30
Team 5
Ultrasonic Inspection and Cost Analysis of Student Challenge Composite
Co-authors : Mastan Raja Papanaboina (1) | Smagulova Damira(1) | Bengisu Yilmaz(1) 1.Prof.Kazimieras barsauskas ultrasoundresearch institute, Barsausko st.59, Kaunas, Lithuania
16:50
Team 6
Structural health monitoring on a multi-curved CFRP plate: detection and characterisation of artificial damages
Co-authors : Guillermo Azuara (1) | María Luisa Rapún (2) | Eduardo Barrera (1) | Rodrigo Alonso (2) 1.Instrumentation and Applied Acoustics Research Group, Universidad Politécnica de Madrid, Campus Sur UPM, Ctra. Valencia km-7, Madrid, Spain 2. E.T.S.I. Aeronáutica y del Espacio, Universidad Politécnica de Madrid, Plaza Cardenal Cisneros 3, 28028 Madrid, Spain
17:10
Team 7
The state-of-the-art and beyond in non-destructive testing of a complex CFRP component.
Co-authors : Gaétan Poelman (1,2) and Joost Segers (1) | Mathias Kersemans (1) | Saeid Hedayatrasa (1,2) |Erik Verboven (1) | Wim Van Paepegem (1) 1. Mechanics of Materials and Structures (UGent-MMS), Department of Materials, Textiles and Chemical Engineering (MaTCh), Ghent University, Technologiepark-Zwijnaarde 46, 9052 Zwijnaarde, Belgium 2.SIM Program M3 DETECT-IV, Technologiepark-Zwijnaarde 48, B-9052 Zwijnaarde, Belgium
17:30
Team 8
Non-destructive localization and characterization of artificial defects in a composite material for application in aerospace technology
Co-authors : Oliver Maurer (1) | Vincent Glößner(1) | Sascha-Dirk Power(1) | Lea Herrmann (1) 1.Saarland University, Department of Material Science and Engineering, 66123 Saarbrücken, Germany
WE.3.C
NDT Design & Production Integration
AUDIT C

NDT Design & Production Integration

Session chairs :
Room : AUDIT C
16:10
WE.3.C.1
Local Acoustic Resonance Spectroscopy: An Escalation Approach for efficient NDT of fiber reinforced composites
Philipp Jatzlau (Technical University of Munich)
Local acoustic resonance spectroscopy (LARS) is a non-destructive technique suitable for fast defect detection in large planar parts, especially in fiber composite structures. Applied in the acoustic frequency spectrum LARS is closely related to the coin tap test. In recent years, several applications in mechanical engineering have been established. These include the inspection of GFRP rotor blades as well as defect detection in CFRP parts in the automotive and aeronautic sectors. LARS is suitable for fast and therefore inexpensive measurements, also because it can be automated easily. For this, an automated tapping device for usage with a small industrial robot is developed in a student project. The objective of this system is to be able to scan plates of a maximum area of approx. 1m² using a roving impulse hammer and a capacitive microphone. Alongside experimental data, numerical simulation of the impulse response within the part is particularly useful to optimize the configuration on both the transmitting and receiving side for early identification of possible disturbances. After being detected by LARS, potential defect locations can then be more closely inspected by more complex non-destructive or destructive methods such as ultrasound. The general goal of this escalation approach is to reduce overall inspection effort.
Télécharger le document.
16:30
WE.3.C.2
3D X-ray Inspection System for Helicopter Rotor Blades
Michael Krumm (RayScan Technologies GmbH)
Modern-style helicopter rotor blades have a complex structure consisting of lightweight materials, such as fiberreinforced plastics, foams or honeycomb structures, and metallic components such as titanium, aluminum or steel. Selection, combination and arrangement of the individual materials and structural elements are optimized to ensure the desired behavior for the respective rotor blade areas. The quality assurance of such complex components requires a complete representation of the total component volume in order to detect pores, defects, delaminations, cracks and other discontinuities and to verify the inner structure. A detailed analysis showed that 3D X-ray techniques are best suited for this purpose. The system presented here, was developed for the following purpose: complete horizontal, fully automatic handling and complete three-dimensional testing of helicopter rotor blades with up to 11 m length and 600 mm width and with up to 400 kg mass. Further, generation of 3D volume sectional images from partially incomplete data sets in different quality levels as well as processing and management of terabytes of memory. As part of the development, a novel measurement strategy for acquiring 3D information based on the successful RoboTom technique has been introduced, combined with an exact manipulation system and a compact radiation protection cabin. This paper presents the developed system and the typical measurement results.
Télécharger le document.
16:50
WE.3.C3
Millimeter-sized non-contact ultrasound probe based on fiber-coupled laser excitation and detection
Wolfgang Rohringer (Xarion Laser Acoustics)
For the ultrasonic inspection of confined spaces, a small probe head is desirable. We present a novel fully integrated laser-acoustic inspection head of 7mm by 7mm footprint: the device combines two miniaturized optical interferometers for acoustic signal detection. These two Fabry-Pérot interferometers are arranged in a 90 degree fashion in respect to each other, so that butt joints and corners can be accurately inspected. The ultrasound signal is generated by a short laser pulse, which is also delivered via an optical fiber, hence also being accessible to confined spaces. Both signal generation and signal detection are not requiring physical contact to the workpiece and do not require a coupling liquid. This novel laser ultrasound testing head can be mounted on a robot in a through-transmission and a single-sided echo configuration, and enable a fully automated inspection. The design of the device is presented, results of scans obtained from CFRP sandwich structures are discussed. The performance and the design are compared to a classical liquid-coupled ultrasound inspection setup, the so-called squirter method. It is found that with comparable sensitivity at 1MHz, the laser-based head is much smaller than the squirter head; on the other hand, the dry inspection using the laser requires additional safety measures such as eye protection. However, a strongly diverging laser beam design leads to a beam expansion to a degree that the laser is eye-safe from a distance of 2m, which is a typical distance of a robot safety fence, which can easily be equipped with a laser interlock. In this scenario, no laser safety googles or additional protective measures are required for the operator.
Télécharger le document.
17:10
WE.3.C.4
Efficient implementation of correlation-based tilt filter for imaging of CRFP ply waviness using laser ultrasound
Lukasz Ambrozinski (AGH University of Science and Technology, Krakow, Poland)
Ply waviness that may arise during the manufacturing of carbon-fiber reinforced plastics (CFRP) is extremely dangerous, dramatically reducing the designed performance of a composite structure. Wrinkles belong to the most difficult defects to detect and characterize since they do not produce large echoes like other flaws. Laser ultrasound (LU) provides sub-ply imaging resolution permitting visualization of CFRP details in a non-contact manner. Therefore, the technique seems feasible for wrinkles detection. It appears, however, that the quality of LU images can be reduced from low signal to noise ratio (SNR) obtained at points with low light reflectivity. A standard procedure to improve SNR is repeating the measurements and averaging the resulting signals. In the case of LU operating at high pulse repetition rate, however, this approach may lead to heat accumulation and can damage the inspected surface. As a possible solution, the scanning can be performed at full repetition rate over a relatively dense grid of sensing points and the adjacent signals can be averaged within a weighting spatial window. In the presence of wrinkles however, the signals are shifted; a simple addition of non-aligned waveforms reduces the signal amplitude and removes information about the structure. As a solution, a correlation-based tiltfilter can be used. The method investigates local time-shifts between the signals and performs averaging along ply orientation. Since the technique is based on short-time cross-correlation, processing of large sets of data in the time domain can be challenging. Therefore, an efficient implementation of the technique in the frequency domain is presented in this paper. Replacing numerically expensive time-domain procedures with multiplication in the frequency domain leads to an efficient algorithm for visualization and quantification of wrinkles severity.
Télécharger le document.
17:30
WE.3.C.5
Nondestructive evaluation of thermal barrier coating thickness degradation using infrared thermography and terahertz-tds imaging
Sreedhar Unnikrishnakurup (Center for non-destructive evaluation, IIT Madras, Chennai, India)
Being hot gas path components, the super alloy flaps and spacers of the gas turbine engine are coated with ceramic Thermal barrier coatings to withstand the high temperatures. During prolonged exposure at high service temperatures, the microstructure of the ceramic layer changes and it undergoes sintering. This can lead to an increase in its thermal conductivity during service. Such increase in the conductivity is undesirable because it would reduce the ability of the ceramic coating to provide the required thermal insulation to the underlying substrate. During service the coatings degrade because of erosion by hot gas and also by localised wear due to rubbing of flaps with spacers. It is necessary to assess the condition of the coatings as a function of service life through suitable non-destructive means. Pulse phase thermography and THz-TDS techniques are used to evaluate the degree of degradation of the TBC top coat thickness. Infrared thermography has the advantage of fast inspection of a large area. In this work, we used a numerical simulation aided calibration and development of a regression model to quantitatively analyse the thickness degradation in TBC system undergone variable time of thermal exposure. These measurements were later verified using well detailed THz-TDS imaging. Experimental results of both technique indicate that these two techniques are feasible for the evaluation of thickness degradation in TBC system.
Télécharger le document.
18:00
Group 1: Departure to CEA List for Labs Visit

NDT Design & Production Integration

Session chairs :
Room : AUDIT C
18:00
Group 2: NDT in Aerospace Conference Center : Refreshment & Exhibition

NDT Design & Production Integration

Session chairs :
Room : AUDIT C
18:45
Departure Group 2 to CEA List

NDT Design & Production Integration

Session chairs :
Room : AUDIT C
19:00
CEA List: Cocktail Dinner

NDT Design & Production Integration

Session chairs :
Room : AUDIT C
21:30
Shuttles departure back to : the Symposium at Campanile, Novotel Saclay and Mercure Massy TGV

NDT Design & Production Integration

Session chairs :
Room : AUDIT C
08:00
Departure of the Shuttles from Novotel Saclay & Mercure Massy TGV to Saclay, NDT Aerospace 2019 Symposium

NDT Design & Production Integration

Session chairs :
Room : AUDIT C
08:15
Opening of the Symposium, Welcome Coffee and exhibition

NDT Design & Production Integration

Session chairs :
Room : AUDIT C
08:30
Plenary Sessions

Plenary Sessions

Session chairs :
Room :
08:30
NDE Challenges at NASA
Elliot Cramer - NASA Langley
The use of composite materials continues to increase in the aerospace community due to the potential benefits of reduced weight, increased strength, and manufacturability. The ability to characterize damage in carbon fiber reinforced polymer composite components is required to facilitate damage progression models capable of yielding accurate remaining life predictions. As these composite structures become larger and more complex, nondestructive evaluation (NDE) techniques capable of quantifying and fully characterizing the material state are needed to enable damage progression models capable of yielding accurate remaining life predictions. This paper will present an overview of current NASA inspection challenges for both aircraft and spacecraft as well as some NDE needs unique to NASA. Further, ongoing NDE research activities at NASA for quantitative characterization of aerospace composites along with a discussion of future directions in NDE research will be presented.
Télécharger le document.
09:15
Ultrasonic and Thermal Techniques from CNDE@IITM for Aerospace Applications
Pr Krishnan Balasubramanian - ITT Madras
NDT methods are currently employed in the quality assurance during the manufacturing and for in-service inspection of aerospace components and structures. Most of these techniques are focused on the detection, sizing, and characterisation of flaws such as cracks, at pre-determined critical locations, that lead to fractures and hence failures in the component. Advanced NDE methods are being developed by the Centre for NDE at the Indian Institute of Technology Madras (CNDE@IITM) that may potentially influence the fabrication, inspection, safety, costing, and maintainability of the aerospace components and its fleets in the military as well as the commercial sector. Some of the techniques that will be discussed here includes: (a) Use of ultrasonic guided waves for the improved inspection of complex structures and components including hidden areas, (b) Structural health monitoring of components and structures using attached and embedded sensor networks, (c) Use of new and novel active thermography techniques for thermal barrier coatings on engine components, and (d) Waveguide sensors for process condition measurements. Using the methods discussed here, the operator now has the opportunity to take vital decisions such as component integrity and propose necessary repair/replacement or estimate the remaining life of the component.
Télécharger le document.
10:00
Coffee Break

Plenary Sessions

Session chairs :
Room :
08:30
NDE Challenges at NASA
Elliot Kramer - NASA Langley
09:15
Ultrasonic and Thermal Techniques from CNDE@IITM for Aerospace Applications
Pr Krishnan Balasubramanian - ITT Madras
10:30
THU.1.A
NDT Methods - From Microwave to Thermography
AUDIT A

NDT Methods - From Microwave to Thermography

Session chairs :
Room : AUDIT A
10:30
THU.1.A.1
Sizing the width of tilted cracks using laser spot lock-in-thermography
Arantza Mendioroz (University of the Basque Country UPV/EHU)
Laser spot thermography has been successfully applied to characterize the width of narrow (down to 0.5 micron) vertical cracks in both, modulated and pulsed regimes. In addition to the experimental thermographic procedure, the accurate quantification of the properties of the crack involves the calculation of the surface temperature distribution. The solution of this heat diffusion problem can only be accomplished analytically in the case of infinite vertical cracks. For more realistic situations, i.e., when the crack is finite, is tilted, or is not flat, etc., numerical approaches are needed. In this work, we address the characterization of infinite tilted cracks using laser spot lock-in thermography. Continuous finite element models (FEM) have been widely implemented to solve the heat diffusion problem in cracked specimens. These models propose separate domains for the sound material and the crack, which is treated as a volume filled with air. As a consequence, the system is discretized by using conforming meshes that match at the interfaces between the bulk and the crack, leading to continuous temperature distributions in both domains. Even if these methods provide appropriate solutions, they present important shortcomings such as the treatment of narrow cracks, which involves extremely fine meshes leading to very poor computational efficiency. To overcome this difficulty, we have developed a new numerical approach based of discontinuous Galerking finite elements modelling (DG-FEM) which is much more efficient to deal with narrow cracks (below 10 micron). The use of DG-FEM enables non-continuous solutions of the thermal diffusion equations over the crack, which is modeled as a single interface characterized by a thermal resistance (Rth) value. The procedure we propose to evaluate this thermal resistance lies in fitting the experimental surface temperature profile perpendicular to the crack, through the center of the laser spot, to the theoretical model. Once determined, this thermal resistance can be related to an air gap width (L) through the thermal conductivity of air (Kair) by Rth = L / Kair. In order to validate the methodology, we have performed laser spot lock-in thermography experiments on stainless steel (AISI-304) samples containing infinite artificial slanted cracks of different widths at pre-selected angles (30, 45 and 60 degrees) with the sample surface. The experimental temperature amplitude profiles were introduced as input data on the developed inversion algorithm. The results confirm the success of the proposed methodology, combining DG-FEM and laser spot lock-in thermography, in characterizing the width of tilted cracks.
Télécharger le document.
10:50
THU.1.A.2
Radome Inspection with Terahertz Waves
Joachim Jonuscheit (Fraunhofer ITWM, Kaisers-lautern, Germany)
We present a terahertz imaging system for aircraft radome inspection based on a two-frequency FMCW radarn working at center frequencies of 100 and 150 GHz and with 40 and60 GHz bandwidth, respectively. The imaging sensor combines two all-electronic multiplier chains at the respective frequencies and is integrated in the production environment of radomes built of glass fiber-reinforced composite structures. While images acquired at the two distinct working frequencies can yield different information of possible structural defects, at the same time, a data fusion algorithm is applied combining the signals of both terahertz sensors for enhanced depth resolution by a total FMCW bandwidth of 100 GHz. I. INTRODUCTION Fiber-reinforced plastic (FRP) composite structures are widely used in mechanical construction where a combination of light weight, high structural integrity, and – in the case of aircraft radomes – transparency for radio and microwave frequencies is required. Such composite structures commonly consist of multiple functional layers of various materials, e.g., aramide and glass fiber material. Since the structures are built layer by layer, an inline production quality control is desirable to detect possible delaminations, structural imperfections and other defects already during manufacturing. Defect detection is furthermore desired in in-field inspection of complete radome structures during maintenance procedures. Frequency-modulated continuous-wave (FMCW) terahertz imaging systems have shown to present an excellent combination of spatial and depth resolution for typical defect sizes while providing large penetration depths in the respective materials [1,2]. In this contribution we demonstrate a two-frequency FMCW imaging system which was integrated in an aircraft radome production environment [3]. We show first measurements of a test radome structure as well as images of flat FRP panels with embedded artificial defects as well as real radomes. REFERENCES [1] I. Amenabar, F. Lopez, and A. Mendikute, Journal of Infrared, Millimeter, and Terahertz Waves, vol. 34, no. 2, p. 152, 2013. [2] E. Cristofani et al., Optical Engineering, vol. 53, no. 3, p. 031211, 2014. [3] F. Friederich et al., “Terahertz Radome Inspection,” Photonics, vol. 5, no. 1, p. 1, 2018.
Télécharger le document.
11:10
THU.1.A.3
Single-shot pulsed terahertz non-destructive evaluation systems for in-line production control and automated inspection
Uli Schmidhammer - Teratonics
The regulated reduction of greenhouse gas emissions is accelerating the adoption of increasingly lightweight innovative materials - such as composites, plastics and coatings - in the transportation industries. New means of Non-Destructive Testing and Evaluation (NDT&E) are needed to meet the quality and safety requirements for these materials during production and maintenance operations. Supported by the "Industry 4.0", a production integrated NDT&E is often needed during the manufacturing process itself (as opposed to a control of the finished product) in order to minimise the waste of time, raw materials and energy. These new controls require stability, operational simplicity, speed, depth of analysis and characterization of internal defects. These needs are not all met by current methods (X-rays, ultrasound, optical methods...). Teratonics satisfies these industrial requirements by providing non-destructive, fast, non-contact and non-ionizing control solutions based on a breakthrough single-shot terahertz (THz) pulse technology invented initially at the University Paris Sud / CNRS, Paris Saclay. By reducing the acquisition time by more than 10,000,000 times compared to the commercial state of the art, our technology opens up a major field of application for THz radiation analysis. Thanks to its unparalleled speed and stability, our solution can be used as a production integrated NDT&E system for in-line inspection. It can also be used as a mobile NDT&E equipment for maintenance operations. It can perform dimensional control (including thicknesses with micrometric precision), and defect sizing and imaging on the surface and inside products. For example, in composite tapes, it is possible to control the fiber / resin ratio and the spatial homogeneity. During the placement of the tapes, the orientation and the contact of the tapes can be checked. When producing composites by RTM or pultrusion, typical defects that can be detected are fiber misalignment/waviness, fiber density variations, bubbles, inclusions, or delamination. In glued products, our technology allows to characterize the distribution and the thickness of the bead of glue and to reveal bubbles or inclusions. It can also be used to detect weld defects, thickness variations or bubbles in welded plastics or composites. Finally, our technology can control thicknesses of single or multi-layer coatings with accuracies of the order of a few microns on both metal and plastic / composite surfaces. Our presentation will present the major results obtained to date and will suggest future directions, both on the NDT system design and on the data analytics.
Télécharger le document.
11:30
THU.1.A.4
Non-destructive inspection on heat damaged composite materials
Andrew Ngo (IMRE, A*Star)
Over the years, composite has become one of the most widely used materials in the aerospace industry. With higher usage, composite materials will be subjected to damage from external factors, with the primary source being heat. Over exposure to heat or contact to excessive thermal energy will lead to exterior and interior flaws like discoloration, delamination or core damage which all result to degradation in the structural integrity of composite materials. The most common and efficient method to examine the material is by performing non-destructive inspection. Inspections are done to detect and locate damages on the surface, subsurface as well as the internal core without affecting the serviceability of the composite component. However, there is a risk for minor flaw to be left undetected and overtime result in an accumulation of thermal stress. Structural failure of the material could happen. In this study, we evaluate the detection sensitivity of ultrasonic and thermography inspection for composite specimens heated at various temperature. The composite sample consists of 12 layers in orientation sequence of [0/90/+45/-45/90/0]s, and the temperature ranges from 50oC to 300oC in steps of 50oC.
Télécharger le document.
11:50
THU.1.A.5
3D Surface Topography Measurement Using Elastomeric Contact
Thierry Mantel (GelSight Inc.)
GelSight is a 3D surface imaging technology based on an elastomeric sensor that enables high-resolution surface metrology and compliant tactile sensing. Using controlled illumination and advanced computer vision techniques, GelSight can transform multiple 2D images into a detailed 3D surface within seconds. A painted membrane on the gel closely conforms with any surface it touches. Then multiple images of the surface with different lighting conditions are taken in a very short time from a single high-resolution digital camera. A 3D reconstruction algorithm generates a height map allowing precise measurement of surface shapes regardless of the optical properties of the material. Based on these concepts, GelSight, Inc. has developed a portable high-resolution 3D measurement system that can quantify surface geometry down to one micrometer on any rigid material, including metals, composites, plastics, glass, fabric, and leather. The system can be taken to the factory floor or in a maintenance facility for insitu measurements on aerospace components. Working with researchers in the aerospace industry, GelSight, Inc. has developed streamlined algorithms of interest to the NDT community, including automatic measurement of the deepest point of a scratch, hole diameter and circularity, and fastener flushness. The GelSight Mobile system also has basic measurements such as gap widths, offsets, Z height profiles, radius of curvature and roughness. The system provides detailed digital information that can easily be integrated in advanced manufacturing organizations or in advanced on-line maintenance management services. For more advanced applications, the system can export data under formats readable by commercial analysis and metrology software. GelSight systems are currently deployed in various industries such as aerospace, automotive, energy and many others.
Télécharger le document.
12:10
THU.1.A.6
Flying Line Active IR Thermography applied to the inspection of environmental barrier coatings
Ludovic Gaverina (Onera)
In the aerospace industry, Environmental barrier coated (EBC) ceramic matrix composites (CMCs) are considered as an attractive option for use in turbine engines. When the system is subjected to thermal gradients, throughthickness cracks can appear. It is therefore crucial to investigate damage evolutions of the EBC. Twenty years ago, Krapez et al. proposed the “flying spot technique” based on a constant displacement of the laser spot to detect cracks in steel. Later, the Pulsed Flying Spot (P.F.S) was developed to obtain in-plane thermal diffusivity fields on heterogeneous and anisotropic materials. Though the laser spot is a very interesting approach due to its known analytical solution, it has a lower scanning rate than the line scan method. For this reason, the flying line method has been proposed instead to detect cracks in EBCs. Recently, ONERA has developed and improved two laser based setup: (i) one to perform thermal fatigue test and create a cracking network on the EBCs and (ii) a second one dedicated to the non-destructive inspection of the EBCs by the flying spot method. The present paper is based on experimental tests carried out on a coated woven SiC/SiC CMC. First, advanced post-processing methods are developed and applied to the thermal sequences acquired by flying line thermography, in order to improve the detection of the cracks. Then, damage localization characterization approaches are introduced and discussed. Finally, conclusions and perspectives are addressed.
Télécharger le document.
THU.1.B
Structural Health Monitoring II
AUDIT B

Structural Health Monitoring II

Session chairs :
Room : AUDIT B
10:30
THU.1.B.1
A toolbox concept to configure a shm solution on ageing aircraft structures based on acoustic and thermal methods
Ramanan, Sridaran Venkat (Saarland University)
One of the challenging issues in aerospace industries is to configure a SHM solution for ageing aircraft structures. This is because of the lack of information regarding the current state of the material properties and the higher likelihood of damage to occur. Principally, when classical NDT actuators/sensors are embedded onto/into structures to perform real-time monitoring of tolerable damages, an SHM solution is principally obtained. However, in reality this is not that much easy to be realised since most of the NDT techniques are off-line methods meaning they are human based and not implicitly ready for automation. Classical ultrasonic NDT methods like phased array ultrasonics have undergone technological innovations in terms of data acquisition and data representation through the development of Sampling Phased Array (SPA) and Reverse Phase Matching (RPM) methods. These methods have advantages in determining in-situ material properties and result in a comparatively higher probability of damage detection. Unlike the classical ultrasonic methods, where transducers can be virtually moved to any location on a structure, guided waves (GW) based sensing and actuation in SHM is fixed to a ‘static’ actuator/sensor network. In order to detect damages considered to be tolerable efficiently the optimal positions of actua tors/sensor need to be identified in advance. To identify those optimal positions numerical simulation is indispensable, allowing something to be generated called differential imaging, that identifies where damage related sensor signals are to be recorded best. Laser Doppler Vibrometry (LDV) has proven to be another efficient tool for monitoring in the recent years often to be used as a means for validation. In addition to this, linear/nonlinear vibrational acoustics combined with thermography have shown a large potential in identifying particu-lar types of damage where the response has been monitoring representative frequencies based on Local Damage Resonance (LDR). When such LDR frequencies are used to excite the ultrasonic actuators, the signal received from the sensors will be sensitive to the given damage to which the LDR frequency refers. This paper will present how a ‘toolbox’ has to be configured such that approaches used in classical NDT including SPA, RPM and LDR along with numerical simulations can be applied to design SHM system solutions for monitoring isotropic and anisotropic structural parts.
Télécharger le document.
10:50
THU.1.B.2
Structural health monitoring system for moisture ingress detection in sandwich composite structures
Sevilia Sunetchiieva (KU Leuven)
Aircraft innovation includes increasing the performance of air structures by utilizing new materials in order to be more durable and lighter. The percentage composite usage in recent airplane designs has reached more than 50% by weight which goes hand in hand with a reduction of fuel consumption. At the same time, the Maintenance, Repair and Overhaul (MRO) requirements for composites are very different from those applied to metals which lead to new NDT procedures, including a still pending lack of trained staff and certification issues. Honeycomb sandwich materials are widely used in aircraft structures. Some of them are taking an important role in the flight control process, like rudders, elevators, and winglets which in a long run need to be appropriately maintained. Despite the thickness of the core including the metal meshes, sandwich composites are still relatively light and have a relatively high flexural strength. The most common purpose of manufacturing sandwich structures is to obtain the greatest stiffness at minimum total density (average). However, the water ingress problem is very common for these structures with even further risk of corrosion if the core is made of aluminum honeycomb and/or adhesive bond (node and skin-to-core) degradation. Moreover, structural damage (node bond failure) could appear due to repeated freeze-thaw cycles of trapped standing water (in liquid or vapor state) in the structure during normal flight operations. Here, we aim to monitor the water ingression by percolation threshold sensors integrated into monolithic and sandwich composite structures, addressing the difficulties of the sandwich composite manufacturing process, e.g. high pressure and temperature, its quality and data acquisition. The percolation threshold sensor is based on the collapse of a conductive particles network in a randomized lattice structure. The presented SHM system finally is validated on a lab scale and it is robust to environmental changes and sensitive to moisture ingression. Part of the research leading to these results has received funding from the European Community's Seventh Framework Programme [FP7/2007-2013] under grant agreement n°212912 and the “NDTonAIR” project (Training Network in Non-Destructive Testing and Structural Health Monitoring of Aircraft structures) under the action: H2020-MSCA-ITN-2016- GRANT 722134.
Télécharger le document.
11:10
THU.1.B.3
Wireless system for structural health monitoring with Guided Ultrasonic Waves : Detection of defects
Marwen Aouini (Institut de Soudure)
In this paper, a miniaturized, low-power wireless system was developed to provide structural health monitoring (SHM) of a metal structure based on Ultrasonic Guided Waves (UGW) technique. This non-destructive technique has an excellent long-range diagnostic capability and it is effective for detecting material losses, cracks and structural fatigue damages. The developed system makes it possible to send the collected short-range and longrange data to a processing center. It plays both the role of an actuator to excite a piezoelectric patch with UGW as well as the role of a sensor that can detect and remotely send the waves responses through emails and/or via an FTP server. In order to control the energy autonomy of the system, the deep sleep mode was used. In this mode, a microcontroller that has an internal clock was used. This clock allows the system to wake up at a predetermined time interval to send and receive the waves using only one single module and then transmit the data remotely. After the data transmission, it goes into deep sleep mode. To validate and verify the effectiveness of the proposed approach, tests were carried out on two types of databases collected using different artificial damages. In the first database, damages were created by removing the material. While in the second case, damages were simulated by adding magnets on the surface of the structure. The choice of this method of the damage creation by magnets was justified using a simulation study. A graphical interface has been developed for the visualization and the automatic detection of damage by using statistical features sensitive to the presence of damages and setting a detection threshold. This threshold is calculated based on the statistical distribution of the damage indices. The experimental results show that the statistical index that we have proposed is efficient.
Télécharger le document.
11:30
THU.1.B.4
Damage assessment of composite structures using electromechanical impedance method
Tomasz Wandowski (Polish Academy of Sciences, Institute of Fluid–Flow Machinery, Poland)
In this paper results of application of electromechanical impedance (EMI) method for damage assessment in composite materials were presented. Object of investigation was a glass fibre reinforced polymer (GFRP) sandwich structure with honeycomb core, stiffeners and bonded joints. The EMI method is takes advantage electrical parameters of a piezoelectric transducer that is either bonded on the structure or embedded inside it. Due to electromechanical coupling of the transducer with the structure, mechanical resonances of structure can be seen in spectra of electrical characteristic of the piezoelectric transducer. Authors investigated the problem of damage assessment using EMI method under the influence of varying temperature. The studies were focused on the composite materials. Authors investigated both the influence of changing temperature and the influence of damage on resistance spectra gathered at the transducer. Results showed that temperature compensation is needed for the purpose of damage detection. Authors proposed temperature compensation algorithm based on signal correlation. Composite materials especially with sandwich structure are characterised by large material damping which could be a problem in the field of damage assessment. Due to the fact of large damping, the damage sensitive area for EMI method could be strongly reduced. This problem was also investigated in the conducted research.
Télécharger le document.
11:50
THU.1.B.5
Vibration health criterion based on the combination of bending and torsional modes to monitor the health status of safety-critical composite structures
Olivier Ponte Felgueiras (Mines ParisTech)
Despite high interest in composite materials (mechanical performance and mass reduction), their application in the transport industry is still limited: suspension, wheels, steering connection elements (axle fronts and postpones, direction and hanging devices,…). In order to ensure the reliability of these safety-critical parts, it is essential to dimension them properly in the design phase, validate correctly the prototypes and, especially, to be able to diagnose their health status throughout their lifetime. Nowadays, multiple methods are available, but measuring equipment which their use is often expensive, with long acquisition and postprocessing times, and often requiring immobilizing the vehicle or extracting parts of its structure. It is therefore desirable to develop a non-intrusive measurement method allowing the assessment of material degradation, from the mesoscopic to the structural scale (cracking, delamination, …), without the need to disassemble the structure. Using the principles of experimental modal analysis, we aim at establishing a clear link between the evolution of certain modal parameters and the damage observed under tensile loading. In order to do this, the load is applied via a piezoelectric actuator, the response of the system is measured by a 3D vibrometer laser and the damage identification is carried out along a section of the specimen using a motorized optical microscope. The tensile modal basis is characterized for several stages of loading and damage. We propose a vibration criterion, using a combination of bending and torsional modes. The idea consists, firstly, of introducing specific damage, controlled by the orientation of composite plies in the specimen, and then allowing it to appear in a natural way (for instance originating at a porosity). This experimental vibration health criterion is then compared to its digital counterpart in order to validate its relevance.
Télécharger le document.
12:10
THU.1.A.6
Damage monitoring framework based on probability modeling under varying structural boundary conditions
Fang Fang (Nanjing University of Aeronautics and Astronautics)
A key challenge for the practical application of aircraft s tructural health monitoring (SHM) is to realize reliable damage monitoring under time - varying conditions. To deal with this challenge, a SHM framework based on probability modeling is proposed to provide a modular architecture with high generalization capability. In this framework, double probability models combining short term and long term SHM information are updated dynamica lly and an improved probability modelling method is proposed to construct models in a stable, adaptive and efficient way. Finally, the probability similarity of models is measured to realize normalized and reliable damage detection under time - varying condi tions. The framework applied with the guided wave - based SHM technology is validated by the hole - edge cracks monitoring of an aircraft wing spar and t he crack s are monitored reliably under varying structural boundary conditions.
Télécharger le document.
THU.1.C
Data processing & Statistical methods I
AUDIT C

Data processing & Statistical methods I

Session chairs :
Room : AUDIT C
10:30
THU.1.C.1
Innovative Learning-by-Examples Approaches for RealTime NDT/NDE of Complex Structures
Marco Salucci (ELEDIA Research Center)
Non-destructive testing and evaluation (NDT&E) is aimed at non-invasively evaluating the structural integrity of complex structures and guarantee human safety in many civil and industrial (e.g., aerospace and nuclear) applications. Within this context, probing the structure under test (SUT) through electromagnetic (EM) fields and processing the resulting "signature" measured over an external observation domain allows one to retrieve preliminary diagnoses before proceeding with ad-hoc (often invasive) repairing/substitution actions. Typical NDT&E analyses can be subdivided into a sequence of inter-connected tasks aimed at progressively infer the SUT status. More in detail, the first task is aimed at detecting the presence of one or multiple defects with respect to a (a-priori known) healthy scenario. Then, a second task is devoted to estimate qualitative information such as the number and the position of the detected targets, while a third step is aimed at retrieving a guess of their shape and size. Finally, depending on the targeted application, a fourth task is the estimation of quantitative information (i.e., material composition) on the imaged defects. Although fast and accurate EM forward solvers have been recently introduced, many state-of-the-art imaging approaches are not suitable for applications where an immediate feedback is required because of their iterative nature and the resulting very high computational load. On the other hand, learning-by-examples (LBE) strategies recently attracted particular attention since they can yield accurate and robust diagnoses with real-time capabilities. LBE methods are based on the generation of a training database of input/output pairs to build a computationally fast surrogate model of the (unknown) inverse operator relating measurements and SUT status.  More in detail, such a process can be seen as a three-step procedure aimed at (a) reducing the dimensionality of the input/data space by identifying a reduced set of features carrying the largest information on the predicted output (i.e., the SUT descriptors), (b) collecting the lowest possible number of I/O training samples able to accurately model the underlying physical phenomena, and (c) train a machine learning (ML) algorithm with good generalization capabilities in order to effectively process previously unseen data during the on-line test phase. Accordingly, this work is aimed at providing an overview of the most recent advances on LBE techniques as applied to NDT&E scenarios, highlighting their current limitations and challenges as well as envisaging future trends including their natural evolution within the ever-growing deep learning framework.
Télécharger le document.
10:50
THU.1.C.2
Options to undertake the AI digital transformation journey for NDT in Aerospace
Peter Chow (Fujitsu)
Artificial intelligence (AI) in machine learning (ML) and deep learning (DL) technologies are making a colossal change in all areas of our lives, societies and businesses. Compared to previous generation of AI the main factor is no coding for every situations, just sufficient data to train the system to the needed accuracy level. This is a huge efficiency in scalability and cost. Also, previous out-of-scope applications such as non-destructive testing and evaluation (NDT&E) like ultrasound and similar methods for automated inspection and structural health monitoring (SHM) are now realisable. Combining with advanced image process technology for defect sizing assessment, the automated quality assurance process provides same high consistency across plants anywhere in the world. This AI digital transformation undertaking is no simple task for any organisations. In the manufacturing NDT&E space some of the technical challenges are: • Lack of expertise and skill personnel • In a good production process the lack of data (e.g. defects) and the fabrication method frequently protected • Further R&D or production insertion suitable On the organisation side some of the reasons for undertaking that need to face are: • Business need e.g. measurable benefits from AI digital transformation and why now • Consequences of not undertaking e.g. delay taking actions • Strategic fit e.g. enabling consistency across both within and across facilities • Fit with programme plan e.g. digitalisation and/or Industry 4.0 roadmap To leverage this technology to support NDT&E and SHM, what are the options today for companies? In this presentation we will present some of the options, “Continue As-Is”, “Develop In-House”, “Strategic University Collaborations”, and “Strategic Alliance with AI service providers” to help organisations thinking or already undertaking the AI journey.
Télécharger le document.
11:10
THU.1.C.3
Microwave NDT/NDE through a Probabilistic Compressive Sensing Method
Marco Salucci (ELEDIA Research Center)
During the last years, microwave imaging techniques have emerged in the framework of non-destructive testing and evaluation (NDT/NDE) as an effective alternative to well-known methods such as X-rays or eddy current testing (ECT). Such methods afford several advantages, among which (i) the use of low-power/non-ionizing electromagnetic waves, (ii) the possibility to avoid a direct contact of the sources with the material under test, and (iii) the strong penetrability of microwaves into dielectric materials. Moreover, they can be exploited in several applicative framework ranging from structural health monitoring to biomedical diagnosis. In this framework, this paper presents a novel method devoted to the retrieval of unknown defects or anomalies embedded within physically inaccessible dielectric media by processing the measured scattered radiation. Towards this goal, the Born approximation is applied to linearize the inverse scattering problem at hand, the total field within the undamaged structure being assumed almost equal to that in presence of the defects. Moreover, in order to counteract the strong ill-posedness and ill-conditioning issues, the proposed method properly combines prior information available on (i) the host media to be investigated and on (ii) the sparsity of the defect/anomaly (namely, physical sparsity or with respect to a suitable set of basis functions), by using a differential compressive sensing (CS) approach enabling an efficient retrieval of a few number of non-zero coefficients modelling the unknown solution. Finally, the inversion is formulated in a multi-task Bayesian compressive sensing (MT-BCS) framework and a fast Relevance Vector Machine (RVM) is used to reconstruct the unknown defects by taking into account the inter-relationship between the linear systems arising from multiple (sequential) illuminations/measurements of the host specimen. Both numerical and experimental results are presented in order to assess the effectiveness of the proposed methodology.
Télécharger le document.
11:30
THU.1.C.4
Efficient TR-MUSIC damage detection in composites with a limited number of sensors
Vasileios Dimopoulos (DMMS Lab)
Time Reversal Multiple Signal Classification (TR-MUSIC) has been shown to successfully localize extended defects in complex media. To this end, an extended degradation is approximated through a set of independent and effective scatterers. These effective contributions dictate the minimum number of sensors and excitations needed for adequate measurements and precise classification. More specifically, in the algorithm’s first step the obtained scattered field is analyzed into a signal and a noise space. In the second step the signal space information is used to highlight steering vectors focusing on the defect’s position. Overall, localization’s quality becomes proportional to the accuracy of the space separation which in turn depends on the number of sensors and scatterers existing in the system. However, in many real-life applications the latter is unknown. Therefore, it becomes problematic to identify and design sensing arrays a priori. In this paper, for an extended degradation, we use a limited number of sensors to identify the scatterer’s position. Although it is common to use large arrays, here we prove through a monitoring network of reduced dimensions, we are able to obtain characteristic information for the defect. Over this process, the ideal signal space is approximated with only a limited number of observations. In addition, these approximations are later on enhanced by increasing data’s dimensionality. In practice, we prove that one may use a single sensor accompanied with multiple excitations and still be able to achieve accuracy analogous to the ideal network. Furthermore, due to reciprocity, this relationship may be inverted. The proposed approach has been evaluated with a sensing array of reduced dimensions performing damage localization on degraded composite structures. To demonstrate the algorithm’s robustness, information from only the vibro-acoustic range is exploited enabling for fast and inexpensive detection for subwavelength defects.
Télécharger le document.
11:50
THU.1C.5
NDT 4.0 : Application of Industry 4.0 in NDT and impact on aerospace companies
Philippe Meynard (Visiconsult)
Industries are living today a change called Industry 4.0. After the version 3.0 based on the electronics, computers and automation, this new evolution is now oriented with cloud, connected devices, Internet of things and networks. Actually, on aerospace companies, the need is to maintain a high-quality production despite higher volume, less time for control. Industry 4.0 could be an opportunity to match this new challenge. In this paper, we present a description about NDT 4.0 and examples of this new concept on aerospace companies. NDT 4.0 is the declination of industry 4.0 dedicated to the NDT world. It gives new tendencies, key points for this market linked to Industry 4.0 with a description of the cycle (Autonomous robot; simulation; system integration; Internet of things; Cybersecurity; Cloud computing; Additive manufacturing; augmented reality and big data) We will give examples of applications and customized systems applied with the NDT 4.0 philosophy. We will mention new challenges and developments using x-ray technologies.
Télécharger le document.
12:10
THU.1.C.6
Deep Learning frameworks for wave propagation-based damage detection in 1D-waveguides
Mahindra Rautela (Department of Aerospace Engineering, Indian Institute of Science, Bangalore, India)
Deep Learning methodologies are said to mimic the brain cognition capabilities and are revolutionizing across various engineering domains. Wave propagation-based damage detection methodology is one of the preferred candidates for online health monitoring systems. In this paper, we have used deep learning models to detect cracks in 1D-waveguides using axial waves. A discontinuity in the waveguide in the form of a crack introduces prior reflections and these signatures are utilized to classify them. In this work, different cracks are introduced with varying crack lengths across different locations. High-frequency tone-burst signals are used to excite the waveguide and their time-domain representations are converted into the frequency domain. Spectral finite element formulations are used to model the cracks in the frequency domain. The solution is converted back into the time-domain and is used to form the feature space (inputs) for deep learning frameworks. We have used classification based supervised deep learning models: Dense Neural Networks (DNNs), 1D-Convolutional Neural Networks (1D-CNNs), Recurrent Neural Networks (RNNs) and Long Short-Term Memory (LSTMs) to detect damage in the waveguide. Alternatively, time-frequency analysis in the form of wavelet transform is also employed to train 2D-CNNs for damage detection. The proposed models are implemented in Python using TensorFlow APIs and the models are trained to learn decision boundary mappings from the feature space to the target space. New signatures are fed into the trained models to detect damages autonomously in real-time without resorting to time-taking pre-processing step and expert’s analysis. Metrics like accuracy, binary crossentropy loss and training time per epoch are used to compare the performance of these frameworks. Their ability to learn and generalize over the phenomenon of damage detection is also discussed.
Télécharger le document.
12:30
Lunch-Buffet

Data processing & Statistical methods I

Session chairs :
Room : AUDIT C
10:30
THU.1.A.1
Sizing the width of tilted cracks using laser spot lock-in-thermography
Arantza Mendioroz (Universidad del País Vasco UPV/EHU)
Laser spot thermography has been successfully applied to characterize the width of narrow (down to 0.5 micron) vertical cracks in both, modulated and pulsed regimes. In addition to the experimental thermographic procedure, the accurate quantification of the properties of the crack involves the calculation of the surface temperature distribution. The solution of this heat diffusion problem can only be accomplished analytically in the case of infinite vertical cracks. For more realistic situations, i.e., when the crack is finite, is tilted, or is not flat, etc., numerical approaches are needed. In this work, we address the characterization of infinite tilted cracks using laser spot lock-in thermography. Continuous finite element models (FEM) have been widely implemented to solve the heat diffusion problem in cracked specimens. These models propose separate domains for the sound material and the crack, which is treated as a volume filled with air. As a consequence, the system is discretized by using conforming meshes that match at the interfaces between the bulk and the crack, leading to continuous temperature distributions in both domains. Even if these methods provide appropriate solutions, they present important shortcomings such as the treatment of narrow cracks, which involves extremely fine meshes leading to very poor computational efficiency. To overcome this difficulty, we have developed a new numerical approach based of discontinuous Galerking finite elements modelling (DG-FEM) which is much more efficient to deal with narrow cracks (below 10 micron). The use of DG-FEM enables non-continuous solutions of the thermal diffusion equations over the crack, which is modeled as a single interface characterized by a thermal resistance (Rth) value. The procedure we propose to evaluate this thermal resistance lies in fitting the experimental surface temperature profile perpendicular to the crack, through the center of the laser spot, to the theoretical model. Once determined, this thermal resistance can be related to an air gap width (L) through the thermal conductivity of air (Kair) by Rth = L / Kair. In order to validate the methodology, we have performed laser spot lock-in thermography experiments on stainless steel (AISI-304) samples containing infinite artificial slanted cracks of different widths at pre-selected angles (30, 45 and 60 degrees) with the sample surface. The experimental temperature amplitude profiles were introduced as input data on the developed inversion algorithm. The results confirm the success of the proposed methodology, combining DG-FEM and laser spot lock-in thermography, in characterizing the width of tilted cracks.
Télécharger le document.
10:50
THU.1.A.2
Radome Inspection with Terahertz Waves
Joachim Jonuscheit (Fraunhofer ITWM, Kaisers-lautern, Germany)
We present a terahertz imaging system for aircraft radome inspection based on a two-frequency FMCW radarn working at center frequencies of 100 and 150 GHz and with 40 and60 GHz bandwidth, respectively. The imaging sensor combines two all-electronic multiplier chains at the respective frequencies and is integrated in the production environment of radomes built of glass fiber-reinforced composite structures. While images acquired at the two distinct working frequencies can yield different information of possible structural defects, at the same time, a data fusion algorithm is applied combining the signals of both terahertz sensors for enhanced depth resolution by a total FMCW bandwidth of 100 GHz. I. INTRODUCTION Fiber-reinforced plastic (FRP) composite structures are widely used in mechanical construction where a combination of light weight, high structural integrity, and – in the case of aircraft radomes – transparency for radio and microwave frequencies is required. Such composite structures commonly consist of multiple functional layers of various materials, e.g., aramide and glass fiber material. Since the structures are built layer by layer, an inline production quality control is desirable to detect possible delaminations, structural imperfections and other defects already during manufacturing. Defect detection is furthermore desired in in-field inspection of complete radome structures during maintenance procedures. Frequency-modulated continuous-wave (FMCW) terahertz imaging systems have shown to present an excellent combination of spatial and depth resolution for typical defect sizes while providing large penetration depths in the respective materials [1,2]. In this contribution we demonstrate a two-frequency FMCW imaging system which was integrated in an aircraft radome production environment [3]. We show first measurements of a test radome structure as well as images of flat FRP panels with embedded artificial defects as well as real radomes. REFERENCES [1] I. Amenabar, F. Lopez, and A. Mendikute, Journal of Infrared, Millimeter, and Terahertz Waves, vol. 34, no. 2, p. 152, 2013. [2] E. Cristofani et al., Optical Engineering, vol. 53, no. 3, p. 031211, 2014. [3] F. Friederich et al., “Terahertz Radome Inspection,” Photonics, vol. 5, no. 1, p. 1, 2018.
Télécharger le document.
11:10
THU.1.A.3
Single-shot pulsed terahertz non-destructive evaluation systems for in-line production control and automated inspection
Uli Schmidhammer - Teratonics
The regulated reduction of greenhouse gas emissions is accelerating the adoption of increasingly lightweight innovative materials - such as composites, plastics and coatings - in the transportation industries. New means of Non-Destructive Testing and Evaluation (NDT&E) are needed to meet the quality and safety requirements for these materials during production and maintenance operations. Supported by the "Industry 4.0", a production integrated NDT&E is often needed during the manufacturing process itself (as opposed to a control of the finished product) in order to minimise the waste of time, raw materials and energy. These new controls require stability, operational simplicity, speed, depth of analysis and characterization of internal defects. These needs are not all met by current methods (X-rays, ultrasound, optical methods...). Teratonics satisfies these industrial requirements by providing non-destructive, fast, non-contact and non-ionizing control solutions based on a breakthrough single-shot terahertz (THz) pulse technology invented initially at the University Paris Sud / CNRS, Paris Saclay. By reducing the acquisition time by more than 10,000,000 times compared to the commercial state of the art, our technology opens up a major field of application for THz radiation analysis. Thanks to its unparalleled speed and stability, our solution can be used as a production integrated NDT&E system for in-line inspection. It can also be used as a mobile NDT&E equipment for maintenance operations. It can perform dimensional control (including thicknesses with micrometric precision), and defect sizing and imaging on the surface and inside products. For example, in composite tapes, it is possible to control the fiber / resin ratio and the spatial homogeneity. During the placement of the tapes, the orientation and the contact of the tapes can be checked. When producing composites by RTM or pultrusion, typical defects that can be detected are fiber misalignment/waviness, fiber density variations, bubbles, inclusions, or delamination. In glued products, our technology allows to characterize the distribution and the thickness of the bead of glue and to reveal bubbles or inclusions. It can also be used to detect weld defects, thickness variations or bubbles in welded plastics or composites. Finally, our technology can control thicknesses of single or multi-layer coatings with accuracies of the order of a few microns on both metal and plastic / composite surfaces. Our presentation will present the major results obtained to date and will suggest future directions, both on the NDT system design and on the data analytics.
Télécharger le document.
11:30
THU.1.A.4
Non-destructive inspection on heat damaged composite materials
Andrew Ngo (IMRE, A*STAR)
Over the years, composite has become one of the most widely used materials in the aerospace industry. With higher usage, composite materials will be subjected to damage from external factors, with the primary source being heat. Over exposure to heat or contact to excessive thermal energy will lead to exterior and interior flaws like discoloration, delamination or core damage which all result to degradation in the structural integrity of composite materials. The most common and efficient method to examine the material is by performing non-destructive inspection. Inspections are done to detect and locate damages on the surface, subsurface as well as the internal core without affecting the serviceability of the composite component. However, there is a risk for minor flaw to be left undetected and overtime result in an accumulation of thermal stress. Structural failure of the material could happen. In this study, we evaluate the detection sensitivity of ultrasonic and thermography inspection for composite specimens heated at various temperature. The composite sample consists of 12 layers in orientation sequence of [0/90/+45/-45/90/0]s, and the temperature ranges from 50oC to 300oC in steps of 50oC.
Télécharger le document.
11:50
THU.1.A.5
3D Surface Topography Measurement Using Elastomeric Contact
Thierry Mantel (GelSight Inc.)
GelSight is a 3D surface imaging technology based on an elastomeric sensor that enables high-resolution surface metrology and compliant tactile sensing. Using controlled illumination and advanced computer vision techniques, GelSight can transform multiple 2D images into a detailed 3D surface within seconds. A painted membrane on the gel closely conforms with any surface it touches. Then multiple images of the surface with different lighting conditions are taken in a very short time from a single high-resolution digital camera. A 3D reconstruction algorithm generates a height map allowing precise measurement of surface shapes regardless of the optical properties of the material. Based on these concepts, GelSight, Inc. has developed a portable high-resolution 3D measurement system that can quantify surface geometry down to one micrometer on any rigid material, including metals, composites, plastics, glass, fabric, and leather. The system can be taken to the factory floor or in a maintenance facility for insitu measurements on aerospace components. Working with researchers in the aerospace industry, GelSight, Inc. has developed streamlined algorithms of interest to the NDT community, including automatic measurement of the deepest point of a scratch, hole diameter and circularity, and fastener flushness. The GelSight Mobile system also has basic measurements such as gap widths, offsets, Z height profiles, radius of curvature and roughness. The system provides detailed digital information that can easily be integrated in advanced manufacturing organizations or in advanced on-line maintenance management services. For more advanced applications, the system can export data under formats readable by commercial analysis and metrology software. GelSight systems are currently deployed in various industries such as aerospace, automotive, energy and many others.
Télécharger le document.
12:10
THU.1.A.6
Flying Line Active IR Thermography applied to the inspection of environmental barrier coatings
Ludovic Gaverina (Onera)
In the aerospace industry, Environmental barrier coated (EBC) ceramic matrix composites (CMCs) are considered as an attractive option for use in turbine engines. When the system is subjected to thermal gradients, throughthickness cracks can appear. It is therefore crucial to investigate damage evolutions of the EBC. Twenty years ago, Krapez et al. proposed the “flying spot technique” based on a constant displacement of the laser spot to detect cracks in steel. Later, the Pulsed Flying Spot (P.F.S) was developed to obtain in-plane thermal diffusivity fields on heterogeneous and anisotropic materials. Though the laser spot is a very interesting approach due to its known analytical solution, it has a lower scanning rate than the line scan method. For this reason, the flying line method has been proposed instead to detect cracks in EBCs. Recently, ONERA has developed and improved two laser based setup: (i) one to perform thermal fatigue test and create a cracking network on the EBCs and (ii) a second one dedicated to the non-destructive inspection of the EBCs by the flying spot method. The present paper is based on experimental tests carried out on a coated woven SiC/SiC CMC. First, advanced post-processing methods are developed and applied to the thermal sequences acquired by flying line thermography, in order to improve the detection of the cracks. Then, damage localization characterization approaches are introduced and discussed. Finally, conclusions and perspectives are addressed.
Télécharger le document.
14:00
THU.2.A
NDT Methods- Phased Array I
AUDIT A

NDT Methods- Phased Array I

Session chairs :
Room : AUDIT A
14:00
THU.2.A.1
Application of TFM imaging for enhanced NDT in Aerospace
Hubert Voillaume (Eddyfi Technologies)
Although ultrasonic phased array is now widely used for aerospace NDT, a new imaging technique called Total Focusing Method (TFM) is bringing new performances with high definition, realistic images. TFM imaging is now available with real time capabilities on rugged UT instruments, and the number of successful applications is growing. Several applications have been demonstrated for aerospace NDT showing advantages in terms of easy to setup, quality, analysis and reliability. In this paper, we will compare PAUT and TFM and discuss how the improved images of TFM give practical advantages over other phased array techniques. Application cases of weld testing on Aluminium and Titanium assemblies, commonly used in aerospace industry, are presented. ATFM, an adaptive version of the imaging technique, is applied to inspect irregularly shaped components produced by the weld bead. Other application cases are presented regarding productivity of base steel for aerospace; how to increase coverage and quality of inspection in production stage.
Télécharger le document.
14:20
THU.2.A.2
Multi-element ultrasonic evaluation of scattering solids by reflection matrix analysis
Cécile Brütt (Safran Tech, Institut Langevin)
Multiple scattering due to strong heterogeneities of the microstructure make ultrasonic testing of forged metal alloys often difficult. Indeed, ultrasound imaging is based on the assumption that the RF signals come from single scattered waves only, so that there is a one-to-one correspondence between depth and time. Multiple scattering breaks this equivalence; as a consequence, the echo of a defect is harder to distinguish from structural noise. Thus, an estimation of the proportion of single versus multiple scattering would be a valuable indicator to assess the quality of ultrasonic inspection. In this work, we present several estimators of the single-to-multiple scattering ratio, obtained by analyzing the scattering reflection matrix. Data were either numerically simulated, or acquired by actual ultrasonic measurements on samples of interest (TA6V and Ti17 titanium alloys billets, 3 MHz linear probe of 128 transducers). The dataset consists in a Full Matrix Capture, which contains all combinations of A-scans with one probe element firing an ultrasonic wave and all elements receiving the backscattered echoes. Two estimators are compared. The first one, introduced by Aubry and Derode (PRL, 2009) and rearranged by Baelde & al. (Ultrasonics, 2018) consists in projecting the data matrix onto a theoretical subspace of single scattering. More than just estimating the single scattering proportion by taking the norm of the projection on the subspace basis, this method also enables to isolate single scattering information. The second method, built by Badon & al. (Sci. Adv., 2016), projects the data matrix in a focused basis to extract the local scattering parameters. In this basis, single scattering appears on the main diagonal of the matrix, whereas multiple scattering results in non-diagonal elements. The results show that in the studied titanium alloys, the single scattering proportion can be as low as 20%: multiple scattering is not negligible and strongly affects the backscattered signals. This conclusion is discussed in relation to the scattering parameters such as the diffusion constant and the scattering mean-free path.
Télécharger le document.
14:40
THU.2.A.3
Aircraft Engine Fan Blade Inspection by means of UT Phased Array and Eddy Current Array
Thibault Darmedru (Safran AE)
In order to improve the NDT inspection of civil aircraft engines fan blades, Safran Aircraf Engines has decided to develop with Olympus on field and in shop two complementary methods : Eddy Current Array and Phased Array Ultrasound. These methods are due to be implemented at regular intervals and/or number of cycles. While Phased Array inspection is being done more often, on the wing, the Eddy Current Array shall be performed in the maintenance workshop following blade preparation. This paper describes the method and the tools specifically developed for this application. Data sets will be presented that confirm the well known advantages of multi-element techniques over the single element ones.  Imaging of the results and real-time coupling check reduces the human factor influence, volume and surfacecoverage with the use of multi-group technology improves the inspection, electronic focalization increases the sensitivity while data storage and availability, speed of inspection and repeatability make the described method particularly interesting. The paper finally shows that the use of dedicated Phased Array and EC Array probes and setups can bring clear advantages for the operator in term of time and ease of use.
Télécharger le document.
15:00
THU.2.A.4
Ensuring complete integrity of aeronautic composite parts thanks to a fully automated non-destructive inspection solution and process
Angélique Raude (Testia)
Composite materials have widely been developed and used in the Aeronautic industry over the years, thanks to their unique combination of high strength, light weight and durability ensuring an excellent strength to weight ratio compare to other types of material. They can be manufactured in various shapes, forms and compositions, including glass or carbon fibre, fibre-reinforced matrix systems or any combination of any of these. This naturally explain why almost all airplanes incorporate some kind of composite materials in their design. And in some cases, they can account for up to 50 to 70 percent of a complete airplane structure. It is therefore critical to inspect such materials / component to ensure their complete integrity at a manufacturing stage to look for known flaws such as: porosity, delamination, lack of bonding, foreign objects, etc. The non-destructive inspection of composite components can be challenging due their varying nature and shapes and tedious depending on the chosen technique. The most common technique used is ultrasonic testing. Although this technique is very beneficial, it has some downsides such as: human dependencies, slow manual process, and the area under investigation is limited by the transducer active area. This has now been addressed by the development of a complete automated solution improving both inspection efficiency and reliability. This paper describes the studies carried out to establish dedicated inspection methods based on conventional and phased array ultrasonic testing, the development of a tailored and fully automated acquisition solution and the creation of a masked analysis process, as well as their deployment. Those tools aim to simplify and speed-up the part integrity diagnosis.
Télécharger le document.
15:20
THU.2.A.5
Flexible wedge phased array transducers for inspecting variable-geometry or complex components
Philippe Dumas - Imasonic
The transmission of ultrasound from the transducer into the inspected component is a determining factor in the performance of ultrasound inspections. Various coupling solutions exist to ensure this transmission. The most frequently used are: • Immersion of the component in water tank: This coupling presents the best acoustic performance (low attenuation, coupling homogeneity, no intermediate interface). However, the inspected parts need to be fully immersed and thus complex control systems are required. • Coupling by direct contact with a liquid couplant, or via a rigid wedge or a delay line with liquid couplant at the interfaces: This coupling requires simpler control systems for the inspection, but the homogeneity of the couplant film and attenuation in the wedges deteriorate the signal. The geometry of the inspected part can make the coupling more difficult to setup, particularly if the surface is complex or varies from one point to another. The problem becomes critical when the dimensions of the transducer are large in comparison with the local curvature of the interface. The use of transducers that are flexible, or that are fitted with a flexible wedge, improves the quality of the coupling for components with complex or variable geometry, and in some cases, makes it possible to do certain inspections that currently have no solution. This article presents the recent developments and results obtained in the context of transducers with flexible wedges, in particular: • Design options; • Flexible membranes development; • Mechanical interfaces development for multi-element transducers; • Mechanical supports development for manual or automated use; • Acoustic performance tests; • Wear resistance tests. These studies have demonstrated the contribution of flexible wedge transducers to various applications, with acoustic performances similar to that of immersion and easy implementation comparable to standard contact inspections, while remaining compatible with an industrial use. The detailed results will be presented, as well as the possibilities for the future developments of transducers with flexible wedges.
Télécharger le document.
THU.2.B
NDT & SHM Reliability assessment
AUDIT B

NDT & SHM Reliability assessment

Session chairs :
Room : AUDIT B
14:00
THU.2.B.1
Model Assisted Probability of Detection in Structural Health Monitoring applied to impacted composite structures
Olivier Mesnil (CEA-List)
In Guided Wave Structural Health Monitoring (GW-SHM), reliability and performance demonstration is one of the main challenge to overcome to ensure industry adoption. However, the cost of computing a Probability of Detection (POD) from experimental data is much higher in SHM than in NDE due to the permanent integration of sensors. In addition, SHM presents specificities (such as the dependency on environmental and operational conditions, the durability of sensors overt time, the need for complex algorithms in post-processing) which makes especially relevant the use of simulation in the reliability assessment process. A methodology of performance demonstration based on numerical tools and dedicated to GW-SHM is therefore required. This work proposes the application of the Model Assisted POD (MAPOD) approach on the case of a guided wave imaging technique, which consists of creating an image representing the health of the inspected structure using a sparse grid of ultrasonic transducers. The use of simulation allows studying the response of the system for a large set of possible configurations, accounting for the variability of defect morphology, of transducer performances or environmental conditions. In order to compute the POD for the range of input parameters of interest, a metamodel of the configuration is built from the forward model of the CIVA software. The illustrative use-case is an aerospace composite panel instrumented by piezoelectric transducers and the studied defect is an impact-like delamination of various size, position and morphology. A dedicated detection criterion is developed for the application and the false alarm rate is studied with Receiving Operating Curves.
Télécharger le document.
14:20
THU.2.B.2
Digital Twin approach for critical engine components using PCRT
Thomas Köhler (Vibrant GmbH)
This paper describes ways to apply Process Compensated Resonance Testing (PCRT) to critical engine components to improve quality and to reduce risk using a Digital Twin and Birth Certificate approach. Over the last few years there have been several high-profile engine blade failures, putting a spotlight on risk management for these critical components. Current inspections can cover many of these risks, but some conditions are more readily addressed with traditional methods than others.  Historically, some of these risks have been handled with destructive “sampling” evaluations or other process control approaches, because there is insufficient linkage to the actual fitness of the component at the time of inspection. Where visual inspections are perfectly suited judging the appearance of a component, structural information is required to understand the development of a critical part in its life cycle. Component failures are usually caused by abnormal behaviour of an individual component compared to the predicted model. Digital Twin approaches can be used to manage the lifecycle of critical components by generating models of the component. Birth and Life Cycle Certificates using resonant data provide structural information of the real-world component. PCRT creates birth certificate data using the resonant fingerprint of a component. This fingerprint is unique but follows the variation of the underlying production process. Hence, PCRT fingerprint data is a perfect source to select components based on a Digital Twin model approach. As components age under the stress of the engine operation, the change in the fingerprint data provides a valuable source of identifying components that change in an unusual or unexpected fashion compared to the Digital Twin model. Combining this information with inflight data as well as visual inspection criteria opens new opportunities to reduce risk of critical components and improve safety.
Télécharger le document.
14:40
THU.2.B.3
Effect of Oxide Induced Microstructure Degradation in Aero-Engine Components on POD of NDT Techniques
Vamsi Krishna Rentala (School of Engineering Sciences and Technology, University of Hyderabad, India)
Successful implementation of damage tolerant (DT) methodology for aero-engines requires estimation of reliability of various NDT techniques. Probability of Detection (POD), being a standard metric for measuring the NDT reliability, is usually established with the help of a large number of service expired aero-engine components containing several fatigue cracks. However, EDM notches or starter cracks generated at room temperature conditions are considered for POD studies in the absence of service expired components with defects. This methodology may or may not consider the effect of crack morphology and the service induced microstructural changes present in a service exposed component. Therefore, considering the POD value generated using such EDM based starter defects for estimating NDT reliability can be unrealistic in nature. For example, eddy current technique is dependent on electrical conductivity of material which is in-turn dependent on oxide induced microstructural degradation and crack morphology. Hence, it is essential to understand the relation between POD of NDT technique and oxide induced microstructure degradation. It is also understood that PODs not all NDT techniques are affected by microstructural variations. Therefore, in the current study, an attempt has been made to understand the effect of service induced oxide cracks in aero-engine components on POD studies. Accordingly, samples with representative geometric stress concentration factor of a typical bolt hole in a turbine disc were prepared from an aero-engine turbine disc. These samples were subjected to high temperature (650oC ~ typical engine service temperatures) fatigue testing with a stress ratio (R) of 0.1 and constant number of fatigue cycles using Gleeble test system. Natural fatigue cracks generated in these samples under laboratory conditions were inspected using fluorescent penetrant and eddy current inspection techniques. This paper discusses the variability associated in the POD or a90/95(flaw detection with 90 % probability and 95 % confidence) values with respect to the oxide induced microstructure degradation. A comparison will be drawn between the POD curves generated using high temperature and room temperature tested samples to understand the effect of oxidation.
Télécharger le document.
15:00
THU.2.B.4
Integration of experimental data in Model Assisted Probability Of Detection computations
Christophe Reboud (CEA)
Probability of detection (POD) is a recognized performance indicator of non-destructive methods in various industrial fields, like Aeronautics or Energy. It consists in calculating the probability to detect a flaw in a specimen according to a characteristic parameter, usually linked to its size. To access this curve experimentally, one needs to establish a design of experiment (DoE) covering a certain range of random effects that may affect the specimen. To be pertinent, this DoE typically involves a quite large number of mock-ups and repeated inspections with different operators, which makes the complete process costly and time consuming. In recent years, the use of simulation has emerged to improve the calculation of PoD curves. This is known as Model-assisted Probability of Detection (MaPoD). The advantages of simulation is its competiveness in terms of cost and speed with respect to experimental approach, as well as its ability to investigate wider ranges of parameters (either characteristic or uncertain) and to produce large datasets that are statistically representative. The main problem with MaPoD in general is its arbitrary description of uncertainty, which has of course a great impact on the calculated POD and relies currently mainly on experts judgments. While some laws of these parameters are well characterized and easily accessible (for example in datasheets), others are more difficult to quantify (human factors). In this work, we propose to enhance and evaluate performance of a Bayesian Inversion implementation to infer the laws of uncertain parameters based on integration of information coming from real experiments. This methodology is evaluated on two use cases. The first one is purely numeric and allows us to access the limit of the method in terms of precision and data requirement to converge to the solution. The second one is experimental and consists in an inspection with an ultrasonic sensor of a planar specimen affected by calibrated defects. We monitor in real time the sensor position parameters (that are uncertain parameters in the context of a manual inspection) thanks to an optical tracking tool. With this system, we have access to the true distribution of uncertain parameters and we can validate the approach experimentally.
Télécharger le document.
15:20
THU.2.B.5
Identification of Key Performance Indicators for SHM in Structures of Increasing Complexity Based on Artificial Neural Networks
Aadhik Asokkumar (Dresden International University (DIU), Dresden/Germany)
Lamb wave based technology being one of the advanced approaches in the field of ultrasonics, is being widely considered to perform Structural Health Monitoring (SHM) in damage tolerant aeronautical structures. Damage tolerance is a concept in which damage up to a tolerable size is allowed to exist in structural components before those components will have to be repaired. To get such damages detected reliably, an automated inspection process based on SHM technology is considerable. However, it is important to know at what features such SHM technology needs to look at such that the appropriate key performance indicators (KPI) can be well identified. Such an identification may be best done through numerical simulation which is the purpose of the paper being proposed here. Referring to guided ultrasonic waves and here specifically to Lamb waves different modes being sensitive to the tolerable damages defined have to be identified such as the S0 mode for through thickness cracks and possibly others for further types of damage. To optimize the quality of the Lamb wave modes generated in the first place, two actuators have been placed opposite to each other on the top and the bottom of the plate like structure considered. Changes in the plate geometry such as thickness have been detected through a mode conversion and the dispersive effect of the mode considered i.e. the S0 mode. In this paper the approach on how to determine the input features for an artificial neural network (ANN) will be described on a numerical basis as well as the resulting KPIs obtained through the ANN. At first, the concept is explained using a plate with five holes from which cracks emanate and will be finally explained for a riveted patched repair such as widely used in aeronautical structures.
Télécharger le document.
THU.2.C
Data processing & Computed Tomography
AUDIT C

Data processing & Computed Tomography

Session chairs :
Room : AUDIT C
14:00
THU.2.C.1
Computed Radiography for High Resolution Imaging Applications of Aircraft Structures
Muzibur Khan (National Research Council Canada)
Industrial radiography provides high throughput in non-destructive inspection particularly for aircraft structural components and therefore an important tool for aircraft maintenance. Film-based radiography requires consumables (films, toxic chemicals and proper disposal of chemical waste), darkroom facility and manual processing which is not only time consuming, but also requires more radiation exposure than digital systems. The industrial radiography community is making an intense effort to replace the conventional film technique with digital technologies such as digital radiography and computed radiography (CR). The CR technology uses a reusable phosphor imaging plate instead of a film, and therefore allows faster/easier digital image acquisition. Moreover, CR has advantage over conventional film as wider dynamic range, higher sensitive to radiation, requires lower radiation dose and shorter exposure times. CR imaging performance for general applications such as water ingress, composite honeycomb structures, and foreign object damage in aircraft structures is similar to that of film radiography. However, CR imaging for high resolution requirement applications such detecting fatigue cracks or weld defects is much more challenging; making it difficult to match the detection limit obtain by film. Performance assessment of the CR technology is required to determine if it can effectively provide equal or better performance than the existing film-based technology for high resolution applications, and to provide satisfactory and repeatable inspection results This paper highlights the CR performance metrics, lessons learned and key issues faced while performing experimental validation of CR imaging for detecting fatigue cracks in aircraft structural components. CR system quality parameters such as spatial resolution, contrast sensitivity, contrast-to-noise ratio (CNR), signal-to-noise ratio (SNR), equivalent penetrameter sensitivity (EPS) as well as other system hardware integrity features were evaluated as part of an assessment of CR suitability for aerospace applications.
Télécharger le document.
14:20
THU.2.C.2
A new bi-imaging NDT system for simultaneous recovery of attenuation and electronic density maps
Cécilia Tarpau (Laboratoire de Physique Théorique et Modélisation - LPTM)
Computed Tomography (CT) is a widely used imaging technique for non-destructive testing and evaluation (NDTE) in industry. CT modalities exploit only primary radiation, which is un-deviated but attenuated radiation going through matter. However, in the energy range of X and gamma rays used in NDTE, Compton effect is an important physical phenomenon which should be taken into account. Compton Scattering Tomography (CST) is precisely the imaging technique, which not only accounts for the Compton effect but uses it to image material electronic density. Compton scattering tomography is appropriate for metal objects and yields higher contrast images as compared to computed tomography. This paper proposes the concept of a CST scanner, which is formed by a source and several detectors placed on a circular ring. When the detectors are set to register the energy of primary photons, the system works as Fan-Beam Computed Tomography (FBCT) scanner. But if the detectors are set to register the energy of scattered photons, the system operates as Compton Scattering Tomography scanner. Thus such a device, which provides both attenuation map (by the fan-beam tomography mode) and electronic density (by the Compton scatter tomography mode) of an object, is called a bi-imaging tomographic system. Both information from primary and scattered rays is then wisely exploited. The mathematical modelling of this system makes use of a Radon transform on circular arcs, a generalization of the classical Radon transform (on straight lines). Numerical simulations are carried out in order to show the theoretical feasibility of the proposed system.
Télécharger le document.
14:40
THU.2.C.3
X-ray scatter removal for artifact free CT imaging
Quan Wen (Carl Zeiss IMT GmbH)
In the past decades, industrial computed tomography (CT) has become a powerful tool for the nondestructive, three-dimensional (3D) examination of intricate parts and their inner structures. In particular the possibility to inspect parts for inner defects allows a much better understanding of manufacturing processes and thus paves the way for systematic tuning of process parameters, for example in the field of additive manufacturing of aerospace components. However, the necessary x-ray energies to allow inspection of high density materials required in these applications come with the drawback that the interaction probability of scattering processes become more likely, which poses a limit to the resulting 3D image quality. Scattered radiation leads to a nonlinear image formation resulting in image artifacts like streaking or cupping. In the worst case, these artifacts may limit defect recognition possibilities due to a mix of the actual defect signal and scattered radiation or it may inhibit further evaluation of the measured data as the resulting 3D image quality is not sufficient for, e.g., a high quality surface determination. In this paper we present a method to compensate for scattered radiation artifacts. Using a newly developed workflow to determine the scattered radiation, scatter artifacts can be dramatically reduced to a level, where further evaluation of the data becomes possible with state-of-the-art algorithms. The method intrinsically corrects for all kinds of scatter problems that may occur in a CT system including scattering inside the object. We will demonstrate the approach for the inspection of turbine blades and large aluminum castings. We will show that using this method an image quality can be achieved, which is otherwise only available using fan-beam tomography with line detectors. In comparison to line detector based fan-beam CT we show the great potential of time reduction to allow high quality scans in minutes instead of hours.
Télécharger le document.
15:00
THU.2.C.4
Computed Tomography in NDT and Metrology for Additively Manufactured Aerospace Components
Dominik Martinsen (Yxlon International GmbH)
Additive Manufacturing (AM) is becoming an increasingly important production technique also for safety critical aerospace parts. Often these parts are very complex and require new methods of testing to assure their structural health and integrity. Furthermore it is often important to verify certain wall thicknesses or for example the inner diameter of a turbine blade's cooling channel. The Computed Tomography (CT) allows both detecting the mentioned anomalies and carrying out dimensional measurements. Since the CT data represent the full 3D volume of the parts, the analysis covers not only the outer surface, but the complete part including the inner structures. However, due to the many different scan parameters it is often difficult to define he right inspection parameters. At first a sophisticated method will be presented that helps the operator to determine the optimum parameter set, which suits the part material(s) and geometry, the required test specification as well as the side conditions of the available X-ray system. Secondly, we will describe how to adjust the inspection strategies in order to adapt to the very specific requirements of the aircraft industry. Since DR standards as the ASTM E 2737 are still missing their equivalent for CT inspection, we will present a method that helps to determine both qualitatively and quantitatively the inspection quality of the CT scans.
Télécharger le document.
15:20
THU.2.C.5
Computed Tomography and Digital Radiography for Additive Manufacturing process Quality Assurance and parameters definition
Carlos Galleguillos (FADA-CATEC)
The demand of metallic parts produced by Additive Manufacturing (AM) is continuously growing in aerospace industry. X-Ray Computed Tomography (CT) has been stablished as preferable Non Destructive Testing (NDT) method for validating flight components. This technique shows as the most reliable solution for the inspection of complex geometries (e.g. topology optimized parts, lattice structures, etc.) for both the analysis of dimensional accuracy and verifying the internal condition of the fuse material and its indications/defects (voids, cracks, lack of fusion, inclusions, etc.). On the other hand, validation of flight components by Computed Tomography use to be performed in the final stage of the manufacturing route which involve different sub-processes such as part extraction of the building plate, heat treatments, interface machining, surface post-processing, etc. In this context, Non-destructive testing based on Digital Radiography just after part fabrication could help to detect not acceptable indications in an early stage, avoiding costs related with post processing in a not valid component. This work presents results of detection of non-acceptable indications on an early stage of the manufacturing route using industrial radiography. Tests are performed in both, control samples and components with different geometries. Even more, inspection results of samples with induced defects (typical from metal additive manufacturing processes) are presented for helping in the evaluation of the indications Also, CT results are studied as input of the process regarding raw material inclusions analysis, settings of the manufacturing parameters, development of controlled porosity and dimensional deviation analysis related with thermal shrinking during the manufacturing stage.
Télécharger le document.
15:40
Coffee Break

Data processing & Computed Tomography

Session chairs :
Room : AUDIT C
14:00
THU.2.A.1
Application of TFM imaging for enhanced NDT in Aerospace
Hubert Voillaume (Eddyfi Technologies)
Although ultrasonic phased array is now widely used for aerospace NDT, a new imaging technique called Total Focusing Method (TFM) is bringing new performances with high definition, realistic images. TFM imaging is now available with real time capabilities on rugged UT instruments, and the number of successful applications is growing. Several applications have been demonstrated for aerospace NDT showing advantages in terms of easy to setup, quality, analysis and reliability. In this paper, we will compare PAUT and TFM and discuss how the improved images of TFM give practical advantages over other phased array techniques. Application cases of weld testing on Aluminium and Titanium assemblies, commonly used in aerospace industry, are presented. ATFM, an adaptive version of the imaging technique, is applied to inspect irregularly shaped components produced by the weld bead. Other application cases are presented regarding productivity of base steel for aerospace; how to increase coverage and quality of inspection in production stage.
Télécharger le document.
14:20
THU.2.A.2
Multi-element ultrasonic evaluation of scattering solids by reflection matrix analysis
Cécile Brütt (Safran Tech, Institut Langevin)
Multiple scattering due to strong heterogeneities of the microstructure make ultrasonic testing of forged metal alloys often difficult. Indeed, ultrasound imaging is based on the assumption that the RF signals come from single scattered waves only, so that there is a one-to-one correspondence between depth and time. Multiple scattering breaks this equivalence; as a consequence, the echo of a defect is harder to distinguish from structural noise. Thus, an estimation of the proportion of single versus multiple scattering would be a valuable indicator to assess the quality of ultrasonic inspection. In this work, we present several estimators of the single-to-multiple scattering ratio, obtained by analyzing the scattering reflection matrix. Data were either numerically simulated, or acquired by actual ultrasonic measurements on samples of interest (TA6V and Ti17 titanium alloys billets, 3 MHz linear probe of 128 transducers). The dataset consists in a Full Matrix Capture, which contains all combinations of A-scans with one probe element firing an ultrasonic wave and all elements receiving the backscattered echoes. Two estimators are compared. The first one, introduced by Aubry and Derode (PRL, 2009) and rearranged by Baelde & al. (Ultrasonics, 2018) consists in projecting the data matrix onto a theoretical subspace of single scattering. More than just estimating the single scattering proportion by taking the norm of the projection on the subspace basis, this method also enables to isolate single scattering information. The second method, built by Badon & al. (Sci. Adv., 2016), projects the data matrix in a focused basis to extract the local scattering parameters. In this basis, single scattering appears on the main diagonal of the matrix, whereas multiple scattering results in non-diagonal elements. The results show that in the studied titanium alloys, the single scattering proportion can be as low as 20%: multiple scattering is not negligible and strongly affects the backscattered signals. This conclusion is discussed in relation to the scattering parameters such as the diffusion constant and the scattering mean-free path.
Télécharger le document.
14:40
THU.2.A.3
Aircraft Engine Fan Blade Inspection by means of UT Phased Array and Eddy Current Array
Thibault Darmedru (Safran AE)
In order to improve the NDT inspection of civil aircraft engines fan blades, Safran Aircraf Engines has decided to develop with Olympus on field and in shop two complementary methods : Eddy Current Array and Phased Array Ultrasound. These methods are due to be implemented at regular intervals and/or number of cycles. While Phased Array inspection is being done more often, on the wing, the Eddy Current Array shall be performed in the maintenance workshop following blade preparation. This paper describes the method and the tools specifically developed for this application. Data sets will be presented that confirm the well known advantages of multi-element techniques over the single element ones.  Imaging of the results and real-time coupling check reduces the human factor influence, volume and surfacecoverage with the use of multi-group technology improves the inspection, electronic focalization increases the sensitivity while data storage and availability, speed of inspection and repeatability make the described method particularly interesting. The paper finally shows that the use of dedicated Phased Array and EC Array probes and setups can bring clear advantages for the operator in term of time and ease of use.
Télécharger le document.
15:00
THU.2.A.4
Ensuring complete integrity of aeronautic composite parts thanks to a fully automated non-destructive inspection solution and process
Angélique Raude (Testia)
Composite materials have widely been developed and used in the Aeronautic industry over the years, thanks to their unique combination of high strength, light weight and durability ensuring an excellent strength to weight ratio compare to other types of material. They can be manufactured in various shapes, forms and compositions, including glass or carbon fibre, fibre-reinforced matrix systems or any combination of any of these. This naturally explain why almost all airplanes incorporate some kind of composite materials in their design. And in some cases, they can account for up to 50 to 70 percent of a complete airplane structure. It is therefore critical to inspect such materials / component to ensure their complete integrity at a manufacturing stage to look for known flaws such as: porosity, delamination, lack of bonding, foreign objects, etc. The non-destructive inspection of composite components can be challenging due their varying nature and shapes and tedious depending on the chosen technique. The most common technique used is ultrasonic testing. Although this technique is very beneficial, it has some downsides such as: human dependencies, slow manual process, and the area under investigation is limited by the transducer active area. This has now been addressed by the development of a complete automated solution improving both inspection efficiency and reliability. This paper describes the studies carried out to establish dedicated inspection methods based on conventional and phased array ultrasonic testing, the development of a tailored and fully automated acquisition solution and the creation of a masked analysis process, as well as their deployment. Those tools aim to simplify and speed-up the part integrity diagnosis.
Télécharger le document.
15:20
THU.2.A.5
Flexible wedge phased array transducers for inspecting variable-geometry or complex components
Philippe Dumas - Imasonic
The transmission of ultrasound from the transducer into the inspected component is a determining factor in the performance of ultrasound inspections. Various coupling solutions exist to ensure this transmission. The most frequently used are: • Immersion of the component in water tank: This coupling presents the best acoustic performance (low attenuation, coupling homogeneity, no intermediate interface). However, the inspected parts need to be fully immersed and thus complex control systems are required. • Coupling by direct contact with a liquid couplant, or via a rigid wedge or a delay line with liquid couplant at the interfaces: This coupling requires simpler control systems for the inspection, but the homogeneity of the couplant film and attenuation in the wedges deteriorate the signal. The geometry of the inspected part can make the coupling more difficult to setup, particularly if the surface is complex or varies from one point to another. The problem becomes critical when the dimensions of the transducer are large in comparison with the local curvature of the interface. The use of transducers that are flexible, or that are fitted with a flexible wedge, improves the quality of the coupling for components with complex or variable geometry, and in some cases, makes it possible to do certain inspections that currently have no solution. This article presents the recent developments and results obtained in the context of transducers with flexible wedges, in particular: • Design options; • Flexible membranes development; • Mechanical interfaces development for multi-element transducers; • Mechanical supports development for manual or automated use; • Acoustic performance tests; • Wear resistance tests. These studies have demonstrated the contribution of flexible wedge transducers to various applications, with acoustic performances similar to that of immersion and easy implementation comparable to standard contact inspections, while remaining compatible with an industrial use. The detailed results will be presented, as well as the possibilities for the future developments of transducers with flexible wedges.
Télécharger le document.
16:10
THU.3.A
NDT Methods- Advanced Ultrasonics
AUDIT A

NDT Methods- Advanced Ultrasonics

Session chairs :
Room : AUDIT A
16:10
THU.3.A.1
Imaging of impact damage in stitched carbon-fiber reinforced plastic composites using laser ultrasonic technique
J. Mrowka (AGH University of Science and Technology)
Composite materials become increasingly common in a wide range aerospace structures. In these applications carbon-fiber renforced plastics (CFRP) offer numerous advantages over metals due to their flexibility in design and superior stiffness to mass ratio. The main disadvantage of these materials, however, is their sensitivity to incur impact damage, which may be difficult to detect. Their resistance to damage can be, however, improved by stitching the fabric in the manufacturing process with high-strength fibrous yarns such as Kevlar, polyethylene or glass fibers. The stitches increase interlaminar shear strength and reduce delamination progression. The stitching causes, however, additional issues for non-destructive evaluation, particularly for ultrasonic methods. Because of the difference in wave velocity between the stitching fibers and the surrounding material, the phase-distorted wave front is obtained. The reflected wave is integrated within the transducer aperture and reduces signal amplitude. Laser ultrasound (LUS)offers ultra-wideband of generated signals and allows for sub-ply imaging resolution. Moreover, it features very small footprint area of both the ransmitter and the receiver not achievable for other ultrasonic techniqes. In some cases LUS can nearly reach the imaging quality of the X-ray computed tomography. In this paper, we utilize the superior properties of LUS for imaging of Kevlar stitched CFRPs using laser ultrasound technique. In the experiment the diameter of the pulsed laser spot was approx. 1 mm and point-like receiver of diameter approx. 10 um was used. Both laser beams were aligned to perform a non-contact pulse-echo scanning of the samples. Two sets of samples impacted with various energy levels were investigated and damage severity was compared. For comparison, the X-ray images were used as a reference data.
Télécharger le document.
16:30
THU.3.A.2
Evaluation of metallic bonded plates with nonlinear ultrasound and comparison with destructive testing
Ribay Guillemette (CEA List)
In the last decades, the use of structural adhesion has increased. It is indeed able to replace traditional bonded techniques such as rivets or bolts, and reduces the global weight of the structure, which is of great interest in the aeronautic industry. In addition, it allows the assembly of mixed or composite materials with better stress repartition. However, to be used for structural joining and critical application, reliable non-destructive testing techniques are compulsory for evident safety reasons, be it after fabrication or during the whole life of the structure. While linear ultrasound is known to detect easily decohesion or voids in a structure, its use for bond strength inspection is less straightforward. Constraining hypothesis have to be made to retrieve the bond strength [1]. However, in some cases, the effect of such bond degradation on ultrasound is the same as the effect of geometrical fluctuations of the structure (eg thickness of various layers)[2]. Another approach relies on the nonlinear signature of a bond defect inspected by high amplitude ultrasound [3], generated with a chaotic cavity transducer [4]. In the present paper, this method is applied to the inspection of various metallic bonded plates (titanium or aluminum) with several bond defects. The defects were introduced by degradation of the surface preparation process (introduction of PTFE spray, or finger prints). Combined with the pulse inversion technique, high amplitude plane waves were sent in the structure, leading to harmonic components observed in the defect region. Mechanical destructive tests were performed and display a good agreement with nondestructive tests. References: 1. E. Siryabe, M.Rénier, A. Meziane, J.Galy & M. Castaings. (2017). Apparent anisotropy of adhesive bonds with weak adhesion and non-destructive evaluation of interfacial properties. Ultrasonics, 79, 34-51. 2. P. Zabbal, these de doctorat, November 2018 3. P. Zabbal, Nonlinear ultrasound for nondestructive evaluation of adhesive joints, QNDE 2018 4. O. Bou Matar, O., Li, Y. F., & K. Van Den Abeele, (2009). On the use of a chaotic cavity transducer in nonlinear elastic imaging. Applied physics letters, 95(14), 141913.
Télécharger le document.
16:50
THU.3.A.3
Non-contact detection of disbonds in adhesively bonded metallic plates using laser-excited zero group velocity Lamb modes
Jakub Spytek (AGH University of Science and Technology in Kraków)
The multi-layered composite components are broadly used in advanced engineering structures, especially in the aerospace industry. Adhesively bonded plates are one of the more interesting types of structures due to their high strength to weight ratio and inherent vibroacoustic damping properties. Unfortunately, they are also prone to delamination defects occurring during the manufacturing process or during the service period. One of the more effective ultrasonic solutions developed for investigating the quality of bonding in such structures is the Zero Group Velocity (ZGV) spectroscopy. The ZGV modes are the specific form of Lamb waves, which become trapped near the excitation region and do not propagate, which results in a local resonance at the frequency characteristic to a given plate. It has been demonstrated, that the frequency of the ZGV mode is very sensitive to the changes of the plate thickness and material properties. This phenomenon is also observed in multi-layered plates, in which even the slight deterioration of the bonding layer may lead to noticeable shift in the local ZGV frequency. Here, we present a non-contact scanning system for non-destructive inspection of multi-layered aluminium plates based on the ZGV frequency measurement. The setup comprises two laser sources – the excitation diode pulse laser and the measuring laser interferometer, which are spatially superimposed on the same spot. The resulting broadband excitation coupled with high-precision measurement allows the detection of the ZGV frequencies in the megahertz range. The investigated sample is placed on the automated XY translation stage to enable fast scanning of the inspected surface. Analysis of the local frequency spectrum for each measurement point results in a two dimensional image revealing defects. Detailed theoretical analysis of the ZGV frequencies allows for precise identification of the delaminated interfaces.
Télécharger le document.
17:10
THU.3.A.4
Noncontact laser-based ultrasonic detection of dynamic stress modulated cracks
Peilong Yuan (Katholieke Universiteit Leuven)
Laser ultrasonic generation and detection of guided acoustic waves is an attractive non-contact technique for non-destructive testing, which exploits the sensitivity of the wave propagation to the presence of defects that act as scattering heterogeneities. Remote inspection by laser ultrasonic is an important asset, especially when the object of interest is in rotational or translational motion. This study presented exploits the variability of the interaction of a crack defect with probing waves, which results from dynamical stress opening and closing the crack, and can be detected by the accompanying modulation of the detected reflected, transmitted and mode converted waves. The proposed NDT concept is illustrated on a component undergoing cyclic stress during a fatigue treatment. Laser pulses are used to photoacoustically generate wave packets at a high repetition rate. Detection is done by laser Doppler vibrometry. The correlation between the envelope, arrival time, spectrum and other detected wave characteristics at different instants of the dynamic load cycles on one hand, and the state of the growing crack, on the other hand, are analyzed. The results demonstrate the potential of laser-based ultrasonic inspection for simple and rapid detection of fatigue cracks open and closed under dynamic stress.
Télécharger le document.
17:30
THU.3.A.5
Innovative Ultrasonic Nondestructive Testing Technique for Bonding Quality Evaluation
Bengisu Yilmaz (Ultrasound Research Institute, Kaunas University of Technology)
In the aerospace industry, high-performance engineering materials such as aluminium and composites are widely used. In order to maintain the advantages of these materials and to preserve the flexibility in design and maintenance, the choice of joining technology plays an important role. Among different joining techniques, adhesive bonding is one of the most suitable candidates with homogenous load distribution, the ability to join dissimilar materials and high performance to weight ratio. Due to lack of knowledge in the quality control of adhesive bonding with non-destructive testing, they are only allowed to be used in secondary load carrier structures. The state of the art non-destructive testing techniques such as ultrasonic inspection is able to identify debonding at the interface, delamination, and some inclusions. Even though defect identification in adhesive joints has been studied broadly, the reliability in the NDT of adhesive bonding remains questionable due to variations during manufacturing and consolidation. Such contaminations may lead to a dramatic decrease in quality, and in some cases to weak bond. This study offers a new methodology with a conventional immersion ultrasonic testing system in order to evaluate bonding quality. In this study two different substrate containing weak bonds have manufactured with release agent contamination: aluminium – epoxy – aluminium and carbon fibre reinforced epoxy -epoxy-carbon fibre reinforced epoxy. The study investigates the shear wave propagation through weak and perfect adhesive bonds and compares them quantitatively according to received ultrasonic signal shape and amplitude. The study emphasizes new parameters related to behind weak bonds with innovative NDT methodology in immersion testing. During tests, the transmitter sensor stays stable while the sample is rotated and the receiver transducer moves along the direction of propagation respectively. On contrary to classical pulse-echo ultrasonic investigations where only interface condition is evaluated, the received signals contain the information from the whole bonding zone. The quantitative analysis of the weak bond for both aluminium and composite adhesive joints have been reported.
Télécharger le document.
THU.3.B
Components
AUDIT B

Components

Session chairs :
Room : AUDIT B
16:10
THU.3.B.1
X-ray digital radiology to infer welding defects depth
Emmanuel Siryabe (Safran Helicopter Engines)
X-ray digital radiology is increasingly used in aerospace industries and medical applications because of their numerous advantages such as the non-use of consumables and chemicals products, the capability to storing information via digital media, the reduction of time of control, etc. Today, the use of X-ray digital radiology give news needs and challenges for industrials. In this paper, X- ray digital radiology is used to infer welding defects such as depth of collapse or undercut, in welded assemblies, by measuring the grey level difference. Two batches of representatives samples are tested according to different composition: the first one is HA188 and the second HA230. Comparing depths of collapse measured by X-ray technique with those obtained by laser profilometer and micrographics cuts, we observe a good agreement, in the range of 30 - 250 μm. This results demonstrate the capability of X-ray non-destructive technique to estimate the “depth of collapse” in representative samples. Further investigations will be applied on real parts and results will be presented in a new paper.
Télécharger le document.
16:30
THU.3.B.2
Electromagnetic pulse-induced acoustic testing and its application to the non-destructive evaluation of adhesive bonding between carbon fiber composite and metal
Toshiyuki Takagi (Tohoku University)
Composite materials are used in aerospace industries to replace metals and to reduce the structural weight. Some new aircraft, such as Boeing 787 and Airbus A350, use composite materials in more than half of their structures. However, it is evident that composite materials cannot wholly replace metals under the current condition. Therefore, the adhesive bonding between metal and composite material is adopted as one of the methods that satisfy the above demands. However, with the increase in usage time, debonding phenomena may occur at the bonding part. Non-destructive testing for the inspection of metal/composite adhesive bonding is required to ensure the safety of the structure. The authors propose the electromagnetic pulse-induced acoustic testing (EPAT) method for non-destructive testing of adhesive bonding between metals and composite materials. In this paper, we introduce the principle of EPAT. A pulsed magnetic field (from the driving coil) excites metal material and generates an acoustic sound, and defects in adhesive joints are inspected by measuring and analyzing the acoustic signal obtained from composite materials side. Acoustic emission (AE) sensors measure the acoustic sound. Because the debonding brings changes in vibration propagation time and frequency, the presence and location of the debonding can be determined by analyzing the received acoustic signals at various locations. Two adhesive bonding specimens of metal and carbon fiber reinforced plastics (one with debonding and one without debonding) are used in EPAT experiments. Two sets of received signals are compared. Also, a finite element simulation is used to elucidate the mechanism of the method.
Télécharger le document.
16:50
THU.3.B.3
Surface characterization using multi-orientation light technique
Clément Skopinski (Framatome Intercontrôle)
Framatome Intercontrôle has developed a solution that improves visual inspections, in terms of inspection duration and/or indications characterization. This presentation will focus on the Helios technology, which uses multiple light orientations to determine the reflectance equation at any point of the surface. The analysis of the reflectance coefficients allows to apply quantitative detection criteria and then objectively discriminate surface indications as scratches, cracks, porosities, rugosity,... A 3D mapping of the surface indications is also possible with a data processing of the reflectance map of the surface that allows to size the indications in the 3 x-Y-Z directions. This technology is scalable and could be apply to small scale and large scale parts. In this presentation, the technology will be presented illustrated with different experimental results on different applications.
Télécharger le document.
17:10
THU.3.B.4
Laser-based Free-form Surface Modelling for NDT and SHM
Eddy-Current Pulsed Thermography for the Detection of Impact Damage on CFRP
17:30
THU.3.B.5
A study on factors influencing the accuracy evaluation of dimensional X-ray computed tomography with multi-sphere standards
Shixiang Su (Nanjing University of Aeronautics and Astronautics, PR China)
Industrial X-ray Computed Tomography (XCT) is one of the available choices for measuring parts’ internal and external geometries. However, there are still several challenges that limit the current XCT technologies from being widely spread as a dimensional metrology tool. In the study presented in this paper, multi-sphere standards were developed and used for the evaluation of the accuracy of dimensional measurements performed with an open-ended 450 kV XCT system of 0.4 mm focal spot. Normally, the multi-sphere standard refers to a combination of several ball rods of various heights, where a ‘ball rod’ is understood as a ruby sphere mounted on a rod that is usually made of carbon fiber or ceramic. The main purpose was to determine if the rod’s material, the spatial distribution of the ruby spheres, and/or the selection of test lengths (i.e., the center-to-center distance between two ruby spheres) influence the determination of the maximum permissible sphere distance error (MPESD) attributable to XCT data. The experimental data suggested ceramic is more suitable as a rod material than carbon fiber, and also, the spatial distribution of the ruby spheres has no obvious influence on the accuracy evaluation of dimensional measurements. However, the selection of test lengths will significantly affect the accuracy evaluation results. In particular, the results of dimensional measurements showed that the evaluation of the MPESD with multi-sphere standards for the XCT system used in this study can be expressed as MPESD = (159 + L/800) μm
Télécharger le document.
THU.3.C
Data processing & Statistical methods II
AUDIT C

Data processing & Statistical methods II

Session chairs :
Room : AUDIT C
16:10
THU.3.C.1
Automatic analysis of ultrasonic data for large and complex CFRP aircraft components
Julien Walter (Centre technologique en aérospatial, Canada)
16:30
THU.3.C.2
Development of dedicated classification tools for the automated diagnostic from non-destructive testing data
Damien Segur (CEA List)
An important problematic in the field of non-destructive testing is the management of large datasets of inspection data and their efficient exploitation to monitor the life of various structures and components. A collaboration between CEA LIST and Airbus has led to the development of dedicated classification algorithms. The application case proposed by Airbus was addressed in the context of the CORAC [1] project INNOFAB. Airbus group collects, with its software NDTKIT [2], data coming from inspections of airfoil stiffeners made of composite material. These experimental databases contain quantitative indicators extracted from ultrasonic testing signals and a set of tags affected to the measurement by technical experts, in particular the category of damage. The volume of inspection of those structures is typically 60 per year, each one being equipped with 18 stiffeners. An unsupervised classifier has been designed and trained to attribute to input data a class of damage among a set of four classes (layer porosity, volume porosity, gap between tows and irrelevant indications). Specific treatments applied to the database, like dimensionality reduction and handling of over/under-representation of classes willbe described, as well as results obtained in terms of performance on experimental datasets. In terms of perspectives, this work demonstrates the potential of learning by example techniques to assist the experts and provide automated diagnostic.
Télécharger le document.
16:50
THU.3.C.3
Defect sizing in guided wave imaging using a convolutional neural network
Roberto Miorelli (CEA-List)
In Guided Wave (GW) Structural Health Monitoring (SHM), a sparse array of ultrasonic sensors, usually piezoelectric transducers, are permanently integrated on the inspected structure to periodically investigate its health. Among the set of post-processing techniques to analyze such measurements, Guided Wave Imaging (GWI) is a promising approach as it allows immediate defect detection and localization. With a large number of sensors and tomographic algorithms, GWI can provide quantitative information on the defect size. However, with a limited number of sensors, inversion of the defect size from the imaging algorithm is a non-linear problem depending on many parameters, including the defect size and position, the frequency of inspection and the positions of the sensors. In this work, a deep learning strategy is implemented to invert the size of defects based on GWI with a sparse array of transducers. A database of guided wave images is first built with the simulation software CIVA for various defect positions and sizes. The database is then used to train a Convolutional Neural Network (CNN) to learn an inversion model. A CNN is a type of neural network relying on successive convolution-like operations applied to images, and is usually used in classification problems. The constructed model is then used to invert the size of defects on experimental images. This methodology is applied on an aluminum panel with a circular through-hole of various positions and diameters, due to the simplicity of creating such experimental samples. Successful inversion of experimental defects is conducted with the numerically trained model. The defect size is inverted with an accuracy of the order of the millimeter for defect in the range of 5 to 15 mm with an inspection frequency at 40 kHz.
Télécharger le document.
17:10
THU.3.C.4
NDT of Welds Using High Speed Data Collection and an Artificial Neural Network Real-Time Processing System During the Welding Operation
Jerald E. Jones (EnergynTech, Inc.)
A New Artificial Neural Network System has been developed and then embedded into a system called "ArcSentry". The system collects real-time data during a robotic or automated welding process, including wire feed speed, travel speed, voltage, current, welding torch travel angle, and welding torch work angle. This data is fed into a neural network model, which instantaneously calculates "slices" of the weld cross-section size and shape. The cross-section can be instantaneously compared to the weld specification. Post welding, the "slices" can then be used to create a 3D image of the weld. The parts which were being welded can then be turned transparent and a 3D image of the weld deposit can be viewed. It can be rotated to any viewing angle. Also, the weld deposit can be color coded: green = meets all specification requirements; yellow = one or more of the weld cross-section measurements at that location in the weld are nearly outside of the specification requirements; and red = one or more of the weld cross-section measurements at that location in the weld are outside of the allowable specification requirements. The data can be transmitted to a central database so that all welds monitored by ArcSentry can have a full NDT Record stored electronically and viewable by Quality Personnel. It is also possible for the ArcSentry System to provide real-time feedback to the welding equipment, to modify the welding procedure and correct the welding operation, in order to assure that the weld meets the specification requirements.
Télécharger le document.
17:30
THU.3.C.5
On The Improvement of Height Indication Measurement in Welds
Abd Ennour Bouzenad (Institut de Soudure / Laboratoire d’Acoustique de l’Université du Mans)
Phased Array Ultrasonic Testing (PAUT) is one of the most mature Non-Destructive Testing (NDT) imaging techniques in industrial activities. As all ultrasonic-based techniques, one of its main advantages over other NDT imaging techniques lies on its detection capability deep within materials. However, PAUT is subjected to beam spreading phenomenon within materials under test, which introduces an overestimation of the height and length of an indication during the sizing operation. In some applications, such as fatigue tests, this overestimation can severely impact the conclusion of a study. In this contribution, a two-step algorithm is proposed in order to automatically measure the indication height in weld-beads. This is performed by considering sectorial scanning images (S-Scan) and applying a correction method to get closer to the existing height. The first step includes the measurement of beam spread in PAUT data collected on a calibration block in the same environmental conditions and using the same probe, wedge and acquisition chain. The second step aims at detecting an indication and measuring its height. The indication’s height is then corrected by considering the beam spread of the different focal laws and the indication depth as well. To validate the present algorithm, two PAUT databases collected on carbon steel welded samples containing several types of real and artificial welding defects were used to quantify the sizing error in correlation with the type of defects. In addition, the comparison is also performed between the proposed algorithm height sizing and the human interpretation of conventional phased array ultrasonic images.
Télécharger le document.
18:00
NDT in Aerospace Center : Refreshment & Exhibition

Data processing & Statistical methods II

Session chairs :
Room : AUDIT C
16:10
THU.3.A.1
Imaging of impact damage in stitched carbon-fiber reinforced plastic composites using laser ultrasonic technique
J. Mrowka (AGH University of Science and Technology)
Composite materials become increasingly common in a wide range aerospace structures. In these applications carbon-fiber renforced plastics (CFRP) offer numerous advantages over metals due to their flexibility in design and superior stiffness to mass ratio. The main disadvantage of these materials, however, is their sensitivity to incur impact damage, which may be difficult to detect. Their resistance to damage can be, however, improved by stitching the fabric in the manufacturing process with high-strength fibrous yarns such as Kevlar, polyethylene or glass fibers. The stitches increase interlaminar shear strength and reduce delamination progression. The stitching causes, however, additional issues for non-destructive evaluation, particularly for ultrasonic methods. Because of the difference in wave velocity between the stitching fibers and the surrounding material, the phase-distorted wave front is obtained. The reflected wave is integrated within the transducer aperture and reduces signal amplitude. Laser ultrasound (LUS)offers ultra-wideband of generated signals and allows for sub-ply imaging resolution. Moreover, it features very small footprint area of both the ransmitter and the receiver not achievable for other ultrasonic techniqes. In some cases LUS can nearly reach the imaging quality of the X-ray computed tomography. In this paper, we utilize the superior properties of LUS for imaging of Kevlar stitched CFRPs using laser ultrasound technique. In the experiment the diameter of the pulsed laser spot was approx. 1 mm and point-like receiver of diameter approx. 10 um was used. Both laser beams were aligned to perform a non-contact pulse-echo scanning of the samples. Two sets of samples impacted with various energy levels were investigated and damage severity was compared. For comparison, the X-ray images were used as a reference data.
Télécharger le document.
16:30
THU.3.A.2
Evaluation of metallic bonded plates with nonlinear ultrasound and comparison with destructive testing
Ribay Guillemette (CEA, List)
In the last decades, the use of structural adhesion has increased. It is indeed able to replace traditional bonded techniques such as rivets or bolts, and reduces the global weight of the structure, which is of great interest in the aeronautic industry. In addition, it allows the assembly of mixed or composite materials with better stress repartition. However, to be used for structural joining and critical application, reliable non-destructive testing techniques are compulsory for evident safety reasons, be it after fabrication or during the whole life of the structure. While linear ultrasound is known to detect easily decohesion or voids in a structure, its use for bond strength inspection is less straightforward. Constraining hypothesis have to be made to retrieve the bond strength [1]. However, in some cases, the effect of such bond degradation on ultrasound is the same as the effect of geometrical fluctuations of the structure (eg thickness of various layers)[2]. Another approach relies on the nonlinear signature of a bond defect inspected by high amplitude ultrasound [3], generated with a chaotic cavity transducer [4]. In the present paper, this method is applied to the inspection of various metallic bonded plates (titanium or aluminum) with several bond defects. The defects were introduced by degradation of the surface preparation process (introduction of PTFE spray, or finger prints). Combined with the pulse inversion technique, high amplitude plane waves were sent in the structure, leading to harmonic components observed in the defect region. Mechanical destructive tests were performed and display a good agreement with nondestructive tests. References: 1. E. Siryabe, M.Rénier, A. Meziane, J.Galy & M. Castaings. (2017). Apparent anisotropy of adhesive bonds with weak adhesion and non-destructive evaluation of interfacial properties. Ultrasonics, 79, 34-51. 2. P. Zabbal, these de doctorat, November 2018 3. P. Zabbal, Nonlinear ultrasound for nondestructive evaluation of adhesive joints, QNDE 2018 4. O. Bou Matar, O., Li, Y. F., & K. Van Den Abeele, (2009). On the use of a chaotic cavity transducer in nonlinear elastic imaging. Applied physics letters, 95(14), 141913.
Télécharger le document.
16:50
THU.3.A.3
Non-contact detection of disbonds in adhesively bonded metallic plates using laser-excited zero group velocity Lamb modes
Jakub Spytek (AGH University of Science and Technology in Kraków)
The multi-layered composite components are broadly used in advanced engineering structures, especially in the aerospace industry. Adhesively bonded plates are one of the more interesting types of structures due to their high strength to weight ratio and inherent vibroacoustic damping properties. Unfortunately, they are also prone to delamination defects occurring during the manufacturing process or during the service period. One of the more effective ultrasonic solutions developed for investigating the quality of bonding in such structures is the Zero Group Velocity (ZGV) spectroscopy. The ZGV modes are the specific form of Lamb waves, which become trapped near the excitation region and do not propagate, which results in a local resonance at the frequency characteristic to a given plate. It has been demonstrated, that the frequency of the ZGV mode is very sensitive to the changes of the plate thickness and material properties. This phenomenon is also observed in multi-layered plates, in which even the slight deterioration of the bonding layer may lead to noticeable shift in the local ZGV frequency. Here, we present a non-contact scanning system for non-destructive inspection of multi-layered aluminium plates based on the ZGV frequency measurement. The setup comprises two laser sources – the excitation diode pulse laser and the measuring laser interferometer, which are spatially superimposed on the same spot. The resulting broadband excitation coupled with high-precision measurement allows the detection of the ZGV frequencies in the megahertz range. The investigated sample is placed on the automated XY translation stage to enable fast scanning of the inspected surface. Analysis of the local frequency spectrum for each measurement point results in a two dimensional image revealing defects. Detailed theoretical analysis of the ZGV frequencies allows for precise identification of the delaminated interfaces.
Télécharger le document.
17:10
THU.3.A.4
Noncontact laser-based ultrasonic detection of dynamic stress modulated cracks
Peilong Yuan (Katholieke Universiteit Leuven)
Laser ultrasonic generation and detection of guided acoustic waves is an attractive non-contact technique for non-destructive testing, which exploits the sensitivity of the wave propagation to the presence of defects that act as scattering heterogeneities. Remote inspection by laser ultrasonic is an important asset, especially when the object of interest is in rotational or translational motion. This study presented exploits the variability of the interaction of a crack defect with probing waves, which results from dynamical stress opening and closing the crack, and can be detected by the accompanying modulation of the detected reflected, transmitted and mode converted waves. The proposed NDT concept is illustrated on a component undergoing cyclic stress during a fatigue treatment. Laser pulses are used to photoacoustically generate wave packets at a high repetition rate. Detection is done by laser Doppler vibrometry. The correlation between the envelope, arrival time, spectrum and other detected wave characteristics at different instants of the dynamic load cycles on one hand, and the state of the growing crack, on the other hand, are analyzed. The results demonstrate the potential of laser-based ultrasonic inspection for simple and rapid detection of fatigue cracks open and closed under dynamic stress.
Télécharger le document.
17:30
THU.3.A.5
Innovative Ultrasonic Nondestructive Testing Technique for Bonding Quality Evaluation
Bengisu Yilmaz (Ultrasound Research Institute, Kaunas University of Technology)
In the aerospace industry, high-performance engineering materials such as aluminium and composites are widely used. In order to maintain the advantages of these materials and to preserve the flexibility in design and maintenance, the choice of joining technology plays an important role. Among different joining techniques, adhesive bonding is one of the most suitable candidates with homogenous load distribution, the ability to join dissimilar materials and high performance to weight ratio. Due to lack of knowledge in the quality control of adhesive bonding with non-destructive testing, they are only allowed to be used in secondary load carrier structures. The state of the art non-destructive testing techniques such as ultrasonic inspection is able to identify debonding at the interface, delamination, and some inclusions. Even though defect identification in adhesive joints has been studied broadly, the reliability in the NDT of adhesive bonding remains questionable due to variations during manufacturing and consolidation. Such contaminations may lead to a dramatic decrease in quality, and in some cases to weak bond. This study offers a new methodology with a conventional immersion ultrasonic testing system in order to evaluate bonding quality. In this study two different substrate containing weak bonds have manufactured with release agent contamination: aluminium – epoxy – aluminium and carbon fibre reinforced epoxy -epoxy-carbon fibre reinforced epoxy. The study investigates the shear wave propagation through weak and perfect adhesive bonds and compares them quantitatively according to received ultrasonic signal shape and amplitude. The study emphasizes new parameters related to behind weak bonds with innovative NDT methodology in immersion testing. During tests, the transmitter sensor stays stable while the sample is rotated and the receiver transducer moves along the direction of propagation respectively. On contrary to classical pulse-echo ultrasonic investigations where only interface condition is evaluated, the received signals contain the information from the whole bonding zone. The quantitative analysis of the weak bond for both aluminium and composite adhesive joints have been reported.
Télécharger le document.
18:30
Shuttles transfers to Vaux le Cernay

Data processing & Statistical methods II

Session chairs :
Room : AUDIT C
19:15
Gala Dinner at the Abbaye of Vaux le Cernay

Data processing & Statistical methods II

Session chairs :
Room : AUDIT C
22:30
Shuttles transfers back to the Symposium at Campanile, Novotel Saclay, & Mercure Massy TGV

Data processing & Statistical methods II

Session chairs :
Room : AUDIT C
08:00
Departure of the Shuttles from Novotel Saclay & Mercure Massy TGV to Saclay, NDT Aerospace 2019 Symposium

Data processing & Statistical methods II

Session chairs :
Room : AUDIT C
08:15
Opening of the Symposium, Welcome Coffee and Exhibition

Data processing & Statistical methods II

Session chairs :
Room : AUDIT C
08:30
Plenary Session

Plenary Session

Session chairs :
Room :
08:30
Applications of Modelling in Aerospace NDT : Past, Present and Future
John Aldrin - Computational Tools
The main objective of the talk is to highlight the value of NDT modeling, through past and recent aerospace applications.  As well, ‘The Ten Laws of NDT Modeling’ will be introduced, encouraging certain best-practices and emphasizing the experiences of the author over the past two decades. Modeling and simulation tools have been shown to support the NDT engineer to better understand the physics of complex inspection problems and provide significant benefits at several stages of NDT technique development.  An overview of past and current NDT modeling and simulation tools is summarized with an emphasis on ultrasonic and eddy current testing.  Several case studies are presented demonstrating the application of simulations to improve NDE design.  As model accuracy has improved over the years, it is now feasible to integrate numerical models as part of an iterative inverse method scheme for sizing cracks dimensions using eddy current techniques.  Recent results on eddy current characterization of discontinuities in propulsion components using model-based inversion are introduced.  Best-practices on the topics of model-assisted probability of detection (MAPOD) evaluation and training machine learning (ML) models with limited experimental data are discussed.   Lastly, models can provide additional value during the transition of an NDT technique and be especially helpful in the training of inspectors.  Throughout the talk, opportunities for the broader use of models in the future will be emphasized.
Télécharger le document.
09:15
Coffee Break

Plenary Session

Session chairs :
Room :
09:30
FRI.1.A
NDT Methods - Mobile NDT
AUDIT A

NDT Methods - Mobile NDT

Session chairs :
Room : AUDIT A
09:30
FRI.1.A.1
Automatic Inspection of Aeronautical Mechanical Assemblies using 2D and 3D computer vision
Hamdi Ben Abdallah (IMT Mines Albi/Institut Clément Ader (ICA)
Quality control is of key importance in the aerospace industry. This paper deals with the automatic inspection of mechanical aeronautical assemblies. For that purpose, we have developed a computer-vision-based system made of a robot equipped with several 2D cameras and a 3D scanner. The 3D CAD model of the mechanical assembly is available. It is used as a reference and it describes the assembly as it should be. The objective is to verify that the mechanical assembly conforms with the CAD model. Several types of inspection are required. For instance, we must check that the needed elements of the assembly are present, and that they have been mounted in the correct position. For this kind of inspection we use the 2D cameras and we have developed inspection solutions based on 2D image analysis. We have found that some types of inspection cannot be performed by using only 2D image analysis. A typical example of such types is the interference between elements. It requires to check if two flexible elements (e.g. cables, harnesses) or a flexible and a rigid element (e.g. pipe, support) are at a safe distance from each other. For this type of situations, we use the 3D data provided by the 3D scanner and we have developed inspection solutions based on 3D point cloud analysis. We have also developed a method to compute the best viewpoints for the sensor held by the robot, in order that an optimal view of each component to be inspected can be obtained. The view-point selection is performed offline (before the on-line inspection) and it exploits the CAD model of the mechanical assembly. The proposed automatic robotized computer-vision-based inspection system has been validated in the context of industrial applications.
Télécharger le document.
09:50
FRI.1.A.2
Accurate & Flexible Solution for Non Destructive Testing in Aerospace Industry Using Robotised System
Mr Thomas Gramberger (Fill Gesellschaft m.b.H)
An ever increasing demand for reducing cost of air travel, rising cost of fuel & environmental constraints has driven the need to further reduce weight of aircrafts & improve performance. Carbon Fibre Composite offers an effective high strength low weight solution. The challenge in using carbon fibre composite for aircraft structures is to ensure that there are no manufacturing defects in the material. Unlike aluminium, composite parts do not “dint” or bend when damaged & detecting defects, especially internal to surfaces become difficult. NDT methods of which several are available become an essential approach. Each method has specific benefits and disadvantages. Today Ultrasonic testing (UT) methods are becoming more popular since they are flexible and relatively easy to implement in production environments. Any selected inspection system needs to be accurate and fast thereby ruling out the use of manually operated scanning devices for large parts. Traditionally, in aerospace industry use of linear kinematics machines was a standard approach for UT systems for Through Transmission or Pulse Echo inspection due to achievable accuracy of positioning required for true inspection paths. Particularly for Through Transmission (TTU) using a Squirter high positional accuracy is necessary to be compliant with stringent inspection specifications. Robot based systems are an attractive alternative as accuracy enhancements in robot kinematics allow them to have the same or even better path accuracy. The challenge however is that although repeatability of standard robotics solutions is excellent the dynamic positional accuracy along a path is not perfect & is only achievable with reduced inspection speed & low data acquisition rates based on the position feedback. Additionally, standard robots need to be protected against water splashes particularly in Squirter Through Transmission technique. Also, programming the robots and ensuring collision free motion through the large number of points is an impossible task using the conventional teach pendant programming methods. An automated approach for programming the system for inspection of all complex aircraft parts including double curvature surfaces e.g. Nose cones and nacelles are essential for successful implementation of a robotised solution. This paper will present a comprehensive and operational NDT solution for aerospace industry with an integral offline programming system which is based on standard Stäubli TX200L robots controlled by industry standard Numerical Controller SIEMENS Sinumerik 840D and secondary encoders to achieve high precision on complex scan paths while ensuring a high inspection speed.
Télécharger le document.
10:10
FRI.1.A.3
Ultrasound Non Destructive Testing of Stage 8-10 Spools On Wing
Brian Pradalet (Air France Industries KLM Engineering and Maintenance)
For 2 years now, Air France Industries KLM Engineering & Maintenance has been carrying out ultrasound nondestructive testing on Stage 8-10 Spools. This is a core part of GE90-9 engines and is a component of the High- Pressure Compressor module. The ultrasound inspections carried out in our hangars or on site for our customers means we can identify any signs of cracks that can lead to the loss of the engine in the event of part failure. The operation is extremely delicate and one of the most difficult to perform.The On-Wing method consists in using an ultrasound probe directly on the wing-mounted engine and significantly cuts TAT. To carry out an ultrasound inspection, the HPC module must be accessible. The access space is very limited and calls for precise movements. This operation requires a degree of dexterity: the technician must be able to place the probe in direct contact with the Stage 8-10 Spool through a borescope hole using a facing camera. In addition, once the probe has been positioned, the inspectors must make sure that the camera probe does not interfere with the Stage 8 compressor blades during the engine core rotation. By manually moving the accessory case, the engine can be rotated to carry out this type of inspection. The probe is very sensitive and fragile and must be handled with extreme care during insertion. Technician are like men with the gold fingers, their expertise ensures high-quality procedures. Inspection of the Stage 8 Web guarantees the detection of any signs likely to emerge in this critical area of the part. Any results outside the OEM’s acceptance criteria will require engine removal for shop inspection and possible repairs.
Télécharger le document.
FRI.1.B
Components - Composites
AUDIT B

Components - Composites

Session chairs :
Room : AUDIT B
09:30
FRI.1.B.1
Advanced thermographic NDT applied on composite aircraft components
Daniel Chauveau (Institut de Soudure Industrie)
An advanced robotised thermographic NDT solution, usable as an alternative to ultrasonic squirter systems, is described. The main innovations are based on: • the analysis and processing of the thermal signal from each pixel of a thermal image. As a matter of fact, the thermal signal resulting of the thermal wave diffusion through the component is mathematically processed and treated using multiple derivatives. The final signal is then compared to the original signal obtained on a reference part in order to detect defects such as delamination, inclusion, void, porosity. • the automation and integration in a fully robotic cell allowing industrial production control with consistency and reliable NDT analysis. The system is designed to be able to perform daily inspection of flying aircraft parts with higher rate and lower costs compared to UT squirter devices. So the inspection rate is now suitable with the production rate. One of the main advantage of thermography is no coupling medium is necessary. There is no longer need of water in the shop and then to dry the parts. Benefits, performance and limitations of the robotised technique are presented and discussed
Télécharger le document.
09:50
FRI.1.B.2
Comparative study of NDT methods for stamping-overmolding process
Xavier Tardif (IRT Jules Verne)
Today, the aeronautics industry faces a major challenge: it must combine a high production rate with a reduction of cost production while maintaining a high level of product technicality (mechanical performance and lightness). It requires process innovations that must lead to a low scrap rate and an ability to identify them early in the process. Stamping of thermoplastic composites coupled with overmolding is one of the best processes to combine speed and reliability of the product and is then subject of interest of such industry. Indeed, the stamping step is very fast while the overmolding step allows the production of much more complex part by adding new functions without adding assembly operations. Meanwhile, as this process is composed of successive steps, each of them generates specific defects. Firstly, during stamping operation the typical defects occurring are mainly either porosities or wrinkles. Porosities are involved by a lack of pressure during the stamping step leading to a bad impregnation of the fibers by the resin. The wrinkles can be induced by mistakes in the positioning of the material in the mold or by the process itself. Secondly, during the overmolding step, the polymer is expected to adhere to the stamped part. However, depending on the process conditions (low injection temperature or low pressure) adhesion may not occurs. In the same time, porosities in the overmolded part could also appear depending of the injection conditions. Today, ultrasonic methods is considered as the reference NDT method. nevertheless, online ultrasonic measurement remains complicated. They require a long implementation time due to the sweeping allowing covering the entire surface to be investigated. In parallel, active thermography which is a full field could significantly reduce measurement times but its performance has to be demonstrated. from an industrial point of view , the performance criteria are not only the probability of detection (POD) but also the duration of a measurement and its ability to be introduced into the industrial tool. Hence, the aim of this work is to investigate the capacity of thermography, ultrasonic phased array and Laser ultrasound NDT method to identify and differentiate the different defects occurring in an overmoulding process. The ability to these NDT method to be introduced in an industrial tool will also be discussed.
Télécharger le document.
10:10
FRI.1.B.3
Comparison of Advanced Post-Processing Techniques for Flash Thermography for Rapid NDT of CFRP Aircraft Component: A Case Study
Gaétan Poelman (Mechanics of Materials and Structures - UGent-MMS)
Over the last few decades, composite materials such as carbon fiber reinforced polymers (CFRP) have become widely used due to their high specific stiffness and strength in comparison to traditional metals. At present, they are being implemented for critical components in multiple industrial sectors (e.g. automotive, aerospace, …). However, due to their layered structure, they are prone to internal damage features introduced during manufacturing or during operation, which may deteriorate their mechanical performance. Hence, in order to ensure their structural integrity, these materials need to be inspected non-destructively. Flash thermography, which is a specific type of optical infrared thermography, is a nondestructive testing method that can be adopted to evaluate a material’s internal structure. In this technique, a component’s surface temperature is rapidly elevated by means of an intense optical flash (5 ms at 6 kJ). Subsequently, a high-end infrared camera is used to record the cooling down regime of the sample’s surface. Internal defects cause heat to be built up above them, making them detectable as localized anomalies in the surface’s temperature profile. Considering that CFRPs have an anisotropic thermal diffusivity and the fact that thermal waves are strongly damped, appropriate post-processing techniques are indispensable for the detection of deeper defects. In this study, flash thermography is applied to a stiffened CFRP aircraft panel (vertical tail of an Airbus A320) with several real production defects. The performance and computational effort of various classical post-processing approaches (e.g. Pulsed phase thermography, Principal component thermography) as well as several more recent post-processing techniques (e.g. partial least-squares thermography, gapped smoothing thermography) is investigated and evaluated with respect to improving the defects’ detectability. Acknowledgments The authors acknowledge both the Research Foundation Flanders (FWO grants 12T5418N and 1148018N) and the SBO project DETECT-IV (Grant no. 160455), which fits in the SIM research program MacroModelMat (M3) coordinated by Siemens (Siemens PLM software, Belgium) and funded by SIM (Strategic Initiative Materials in Flanders) and VLAIO (Flemish government agency Flanders Innovation & Entrepreneurship). The authors express their gratitude towards SABCA for supplying material for this research.
Télécharger le document.
FRI.1.C
Materials Characterisation
AUDIT C

Materials Characterisation

Session chairs :
Room : AUDIT C
09:30
FRI.1.C.1
In Situ Thermographic Inspection of Automated Fiber Placement at NASA Langley Research Center
Elizabeth D. Gregory, Ph.D. NASA Langley Research Center
Automated fiber placement (AFP) can be used to fabricate complex structure with highly tailorable properties but it can also introduce critical manufacturing flaws that are difficult to detect. Inspection of parts comprises a significant amount of the overall manufacturing time for AFP and is often done manually, by human visual inspection. The In Situ Thermographic Inspection System (ISTIS) developed at NASA Langley Research Center aims to reduce the time needed to inspect these parts. ISTIS uses an algorithmic, data-rich process for detection and can assess the quality of cohesion between plies beyond the ability of human visual inspection. ISTIS consists of a thermal camera and a customizable mini PC mounted on the AFP head with image acquisition triggered by the AFP system at known, discrete spatial increments. The ISTIS system has been used to create three-dimensional data sets of the thermal signal of each ply. ISTIS has been used by multiple projects at NASA Langley to observe flaws that would not have been detectable by visual inspection. ISTIS has proven to be a useful tool for process development in addition to part inspection by providing unique information about substrate heating. ISTIS has been able to detect ply debonding resulting from the steering geometry of subsequent plies. ISTIS has been able to illustrate heating variability due to AFP head position during fabrication of a complex geometry part. The NASA Langley team is also planning to expand the amount of data that will be collected and the type of automatic analysis that will occur at the time of collection incorporating machine learning for flaw classification. This presentation will also discuss upcoming improvements to the system and data analysis algorithm.
Télécharger le document.
09:50
FRI.1.C.2
Multi-functional phased array ultrasonic imaging: 4 modes for better analysis
Choon-Su Park, Korea Research Institute of Standards and Science, Republic of Korea
Phased array ultrasound testing (PAUT) is able to visualize defects inside structures. As industry develops, various structural components are made of many kinds of materials, and they need appropriate frequencies corresponding materials for non-destructive PAUT. Conventional PAUT, however, generally needs an array transducer that has a single resonant frequency. In addition, nonlinear PAUT with high resolution [C. Park et al., NDT&E Int., 2016] has been newly proposed to look for hidden defects from linear PAUT. These recent trends lead us to develop multi-functional phased array ultrasonic testing (M-PAUT), which provides two linear images and two nonlinear images. The M-PAUT needs an inter-leaved array transducer composed of two different resonant frequencies intervening each other. Two linear phased array images are able to be from each resonant frequency array, and two nonlinear phased array images are higher-harmonic image and sub-harmonic image by corresponding transmitting-receiving combination. The 4 images with an array transducer must be able to delineate more information on the status of structures. The multi-functional phased array imaging scheme is introduced, and computer simulations have assured that the proposed method successfully shows two linear phased array images compatible with conventional PAUT image. In the end, some experiments clearly demonstrate that the system for M-PAUT works well for linear phased array images as well as non-linear phased array images.
Télécharger le document.
10:10
FRI.1.C.3
Use of microfocus X-ray system and pulse thermography to measure porosity content into aerospace composite parts
Pierre Servais (NDT service composite testing, Belgium)
A major problem for aerospace composite parts is to size and measure the porosity content. Ultrasonic is being used since many years but relies on the backwall echo amplitude to determine the porosity content of the defect. As production parts have now more and more complex geometries with internal ribs and locally high curvature, it is here proposed to use the latest state-of-the-art X-ray system using high resolution Flat Panel combined with a microfocus X-ray tube. This new digital technique is more and more portable and wireless and can be combined with robots to be able to X-ray zones in less than 20 seconds for each area (25x30 cm). Future artificial intelligence will recombine the different Xray images to provide a full mapping of the part and assess directly the porosity percentage. New design of Inspection Reference Panel for porosities are now used to build up a first correlation between the ADU graylevel of different porous zones using CNR as given by ASTM. Porosity is usually very difficult to measure for chamfer zone or joggle or curved areas for the historic Ultrasonic Testing method but digital Xray is a very good way to visualize and analyze the porous areas based on quantitative measurement of radiographic attenuation which was not possible for the old Silver Film technique. This paper will study different artificial and natural porous zones inspected by Ultrasonics and compared to digital X-ray (RT) and pulse thermography (IRT). A combination of both digital methods (RT and IRT) will be presented as a future alternative to Ultrasonics as X-ray is sensitive to any defect perpendicular to plies orientation and Thermography is sensitive to any air gap parallel to the plies. Combining both images (RT and IRT) will be a very clever way to inspect composite parts, it is very fast and digital and without any water consumption like ultrasonics and it permits to push further the NDT Data Fusion algorithms used in Artificial Intelligence.
Télécharger le document.
10:30
Coffee Break

Materials Characterisation

Session chairs :
Room : AUDIT C
10:45
FRI.2.A
NDT Methods- Phased Array II
AUDIT A

NDT Methods- Phased Array II

Session chairs :
Room : AUDIT A
10:45
FRI.2.A.1
High-Performance Ultrasonic Transducers based on PMN-PT Single Crystals for NDT of aerospace materials
Thomas Herzog (Fraunhofer Institute for Ceramic Technologies and Systems IKTS)
Ultrasonic transducers for NDE applications are commonly based on Lead Zirconate Titanate or PZT, an inorganic compound and ceramic perovskite material. Until now the advantages of PMN-PT are used in medical applications, but are not implemented in NDE. For applications with low signal amplitudes, high electronic noise and small transducer elements, the perfor-mance of ultrasonic probes can be significantly enhanced by using Lead Magnesium Niobate-Lead Titanate (PMN-PT) instead of PZT. This single-crystal material offers significantly better piezo parameters and leads to a higher sensitivity and larger bandwidth. There is a better depth resolution possible with this transducers as practical tests have shown. Similar to PZT it can also be fabricated in 1-3 piezocomposite technology. In a cooperation between Fraunho-fer IKTS and Ibule Photonics, PMN-PT ultrasonic transducers are developed and optimized. The performance of phased array probes and single element transducers was measured by a so called PCUS® pro electronic front end from Fraunhofer and compared with equivalent PZT-based probes. As a result various single element and phased array transducers with im-proved performances are available for NDT of typical aerospace materials and applications.
Télécharger le document.
11:05
FRI.2.A.2
Advanced DDF for inspection of thick Aluminium and Titanium materials
Xavier Harrich (Socomate International)
FAAST is a UT Phased Array system especially designed to meet with the high productivity inspection needs, thanks to its outstanding capability at replacing multiple conventional Phased Array systems working in parallel. The FAAST technology has the capability to transmit multiple sound beams, multi-oriented and/or multi-focused in one single shot using standard 1D or 2D phased array probes. More features are integrated into the instrument, such the possibility to shot at different frequencies within the same shot. Furthermore, it allows to work in full parallel or by using several active apertures on the same probe. Hereafter is a list of applications where the FAAST technology brings added value within the Aircraft industry: Using FAAST for special alloy and titanium turbine disc inspection complying with Multi-zone testing procedures aiming at the detection of down to Ø0.4mm FBH at 2.5mm up to 140mm depth from the surface. Multi-zone testing procedure requests Multi-focused and/or Multi-oriented beams generation within a single spray throughout a 2D matrix probe, reducing thus considerably the inspection time per turbine disc. Using FAAST in Aircraft Industry for aluminium plate has been performed using the multiple focused aiming at the detection of Ø0.8mm FBH from 2mm up to 160mm depth from the surface using a single 1D linear 128 elements phased array probe with a width of 120mm. As the FAAST allows to generate all focusing delay laws in one single shot, the scanning speed can reach up to 700mm/s while the performances answer the Aircraft standards in terms of SNR. Using FAAST in Bars application has been performed using the multiple angles to detect Ø0.8mm FBH and longitudinal notch. By using a curved phased array probe, the FAAST generates the 0° and ±45° angles in one single shot allowing thus an increase of speed of 60.
Télécharger le document.
11:25
FRI.2.A.3
Dynamic tracking of phased-array probe and 3D C-scan reconstruction for in-service inspection of CFRP aircrafts components
Alexandre Beausoleil (Centre technologique en aérospatiale)
Ultrasonic testing is the preferred method to perform in-service inspection of CFRP aerospace parts. Obtaining Cscan images is highly desirable in order to increase the probability of detection, to improve the results traceability or to provide guidance for repairs. Different phased-array solutions exist to obtain C-scans during in-service manual inspections, but they require mechanical encoders or scanners which bring limitations: limited inspection surface, restrictions to the operator movement, problems adapting the scanner to curves and complex geometries that can be found on an aircraft. An improved inspection technique is proposed using a portable phased-array UT instrument coupled to a dynamic tracking system. Using markers, a stereoscopic camera collects the real-time 3D position of a phased-array probe. A software application was developed to perform real-time transformation of the 3D coordinates measured by the tracking system into 2D “scan-index” coordinates required by the UT instrument. C-scans can thus be generated as the probe moves along the inspected surface. As the link between the probe’s path and the C-scan axes is determined dynamically, the skew angle of the probe can also be calculated. If the probe moves in a direction normal to the surface, the acquisition is interrupted until the inspector puts the probe back on the surface. This allows to resume an inspection and to scan discontinuous surfaces. Moreover, a reference tool can be used to allow a displacement of the camera. As the inspection resumes, no discontinuity is observed on the C-scan. In addition, for every pixel of the C-scan, the probe’s 3D position is recorded to allow for a post-treatment 3D reconstruction of the inspected part. The first step is to create a mesh of the surface based on valid probe positions. Then, for each valid position, a C-scan pixel cell is associated to the face of a 3D surface voxel so that the C-scan is locally stretched or compressed to fit the part’s surface. The technique was tested on reference CFRP specimens including artificial defects as well as real aerospace components. The accuracy of the defects positioning, and sizing and the overall performance of the technique are discussed.
Télécharger le document.
11:45
FRI.2.A.4
Inspection of Composite Laminate Material using Adaptive Focusing Technology
Etienne Grondin (Olympus)
For the past several years, the aviation industry has seen important growth due, in part, to an increased presence of low-cost carriers and regional airlines which has changed the business dynamics by introducing new low-cost business models. As a result of this growth, production rates for new airplanes have increased and new aircraft programs are being launched. Consequently, aviation component manufacturers are facing new challenges including a rise in production rates, a high probability of detection (POD) due to the critical nature of the parts being manufactured, lack of skilled operators, and parts with increasingly complex geometry. Ultrasonic phased array (PA) instruments have evolved, enabling the implementation of advanced acquisition strategies. The introduction of these acquisition strategies helps manufacturers address the inspection challenges they are facing. The evolution of electronic components enables advanced acquisition strategies, such as adaptive focusing, to be implemented. Adaptive focusing simplifies the inspection of variable radiuses, variable opening angle and twisted components and will also compensate for probe misalignment by using innovative signalprocessing algorithms. This paper presents an overview of the adaptive focusing technology with the goal of helping NDT integrators and composite material manufacturers to address the challenges they are facing in terms of system performance, production output, and quality control.
Télécharger le document.
12:05
FRI.2.A.5
Modeling and Inspection Challenges for Evaluation of Composite Materials and Structures
Deborah Hopkins, Bercli
The challenges in inspecting and modeling composites are significant, and it is often the case that both experiments and modeling are necessary to design and optimize inspections. Whereas laboratory and field experiments have long been the standard for validating models, the complexity and heterogeneity of compositematerials makes it extremely difficult to replicate test specimens and experimental conditions as is necessary to achieve comparable results and statistically valid results. Similarly, it is extremely difficult to capture the complexity and material variability of composites in numerical models. Previous results, including successful implementation of composite-inspection systems, have demonstrated the ability to meet aerospace requirements for both in-line and off-line applications [1-3]. Ongoing work and research by others [4] promise to further advance ultrasonic-inspection methodologies. In spite of these advances and the success of existing industrial implementations, significant challenges remain, particularly for polymer-matrix composites with multiple layers and complex geometry. The approach taken for the work reported here is to use modeling (CIVA 16 with and without Surface Adaptive Ultrasound [SAUL]) in conjunction with laboratory experiments to identify and characterize the variability for specific cases. The laboratory experiments were performed using both flat and curved linear arrays, as well as a matrix phased array, for a variety of composite aerospace parts and a specially designed test specimen. Among the specimen variables studied were thickness, shape, and for curved specimens, the effect of changes in the radius of curvature across the part. For the SAUL measurements, variables included the number of active elements used for each shot. All of these variables are shown to be important for the design of probes and optimization of inspection strategies for both in-line and off-line inspection of composite structures. 1. D. Hopkins, M. Datuin, M. Brassard, “Challenges and solutions for ultrasonic phased-array inspection of polymer-matrix composites at production rates,” Proc. 45th Annual Rev. of Prog. in QNDE, (2018). 2. M. Brassard, D. Hopkins, J. N. Noiret, “Integration of robotics and surface-adaptive phased-array ultrasound for fully automated inspection of complex composite parts,” JEC Composites Magazine, 83, (2013). 3. S. Mahaut, O. Roy, S. Chatillon, P. Calmon, “Modeling and application of phased array techniques dedicated to complex geometry inspection,” Proc. 39th Annual Rev. of Prog. in QNDE, (2012). 4. E. Carcreff, G. Dao and D. Braconnier, “Fast total focusing method for ultrasonic imaging,” QNDE, (2016).
Télécharger le document.
FRI.2.B
Components - Composites
AUDIT B

Components - Composites

Session chairs :
Room : AUDIT B
10:45
FRI.2.B.1
Electromagnetic nondestructive inspection of fiber misalignment in CFRP with directional eddy current
Hiroyuki Kosukegawa (Institute of Fluid Science, Tohoku University)
Carbon fiber reinforced plastics (CFRPs) have been increasingly employed for structural materials in aircraft these days. However, the development of nondestructive testing (NDT) technique for CFRPs is still not sufficient. In particular, “fiber waviness” and “misorientation” are representative defects which largely affect mechanical properties of the product and cannot be inspected by conventional NDT technique. We introduce electromagnetic nondestructive technique with directional eddy current to inspect and evaluate these fiber misalignment defects. Because directional eddy current is highly sensitive to the direction of carbon fiber, eddy current testing (ECT) using directional eddy current can efficiently evaluate fiber misalignment by comparing with conventional isotropic eddy current. In this paper, we adopt three types of eddy current probes: (i) circular driving (CirD), (ii) symmetrical driving (SymD), (iii) uniform driving (UniD). CirD probe is a conventional type and generates isotropic magnetic flux. SymD probe generates directional and sharp eddy current near the center of the probe. UniD generates directional but widely uniform magnetic flux. The detectability of fiber waviness and misorientation are evaluated with three probes above. CirD probe shows fairly good performance on both misalignment defects. SymD probe exhibits high performance on the evaluation of fiber waviness but poor ability on misorientation. UniD probe shows the highest performance on both fiber waviness and misorientation. UniD probe can detect fiber waviness with a waviness angle of more than 2º. The lowest misorientation angle which UniD probe can detect is less than 3º. These capacities of UniD probe is effective to assess the quality in mechanical properties of the CFRP products. We conclude that ECT with directional eddy current can be a powerful technique to evaluate fiber misalignment and useful for quality assurance of CFRP products.
Télécharger le document.
11:05
FRI.2.B.2
Ultrasound defect detection in thick composites with out-of-plane waveness
Cui Fangsen (Institute of High Performance Computing, A*Star, Singapore)
Thick composites (usually thickness > 10 mm) are increasingly applied in aerospace engineering and prone to outof- plane waviness in the manufacturing process. Besides the drastic decrease of compressive strength due to outof- plane waviness, local resin rich regions may also exist in the composite, both perplexing the ultrasound propagation therein. In this study, a multi-frequency ultrasound-based method is proposed to characterize both waviness and delamination. Numerical modeling with frequency sweeping is conducted first, to quantitatively analyze influence of waviness and local resin rich region on ultrasound propagation in pulse-echo mode. It is found that at the resonance frequency corresponding to single composite ply, ultrasound B-scan built from signal phase can well characterize the waviness. And ultrasound with a lower frequency, compared with the ply resonance frequency, enables more energy to penetrate the local resin rich region and propagates with minor wave vector deviation through the wavy region. Then, a two-step composite characterization method is proposed. In the first step, ultrasound with frequency corresponding to the ply resonance is used to characterize waviness, and in the second step, ultrasound with a lower frequency is generated to detect delamination within wavy composite based on reflected wave signal. With this method, defects can be detected in the coexistence of waviness and local resin rich region. Finally, experimental validation work is performed, showing the method is promising in detecting defects.
Télécharger le document.
11:25
FRI.2.B.3
Ad-hoc solutions for ultrasonic inspection of radii in closed composite structures
Sergio González (CATEC-Advanced Center for Aerospace Technologies)
Nowadays, the ultrasonic inspection (UT) is the most common Non Destructive method (NDT) largely applied in the aerospace industry due to its accuracy, reliability and the degree of industrialization. Although UT is widely developed, new manufacturing processes and part concepts are continuously pushing the technology for new improvements and applications. This is the case of the geometries resulting from novel composites manufacturing methods, such as Resin Transfer Moulding (RTM) or Liquid Resin Infusion (LRI). The last allows the fabrication of complex parts with internal stiffeners and closed geometries in one shoot, reducing assembly time but also limiting the accessibility for NDT inspections. In this sense, one of the critical issues in the aeronautic sector is the improvement of radii inspection capabilities. The requirement of the aeronautic industry of guaranteeing the quality of the primary composite structures implies the inspection of the whole component through certified technologies. This constraint, together with the lack of technological solutions for the inspection of regions with limited accessibility is blocking the industrial implementation of optimized manufacturing processes. The present work describes the design and development of adaptive hardware solutions for the inspection of internal radii in composite structures. The final positioning tool has been obtained as result of an iterative design process and manufactured through additive manufacturing technology (Powder Bed Laser Fusion, formerly SLSSelective Laser Sintering) using polyamide. This device was validated for inspection of inner radii of five outboard boxes, belonging to the winglet structure of a Green Regional Aircraft prototype (Clean Sky 2 programme). The inspections under real operational environment demonstrated the validity of the devices, enabling a certified inspection in the whole radii length and reducing the lead time of the process.
Télécharger le document.
11:45
FRI.2.B.4
Full wave field signal processing techniques for NDT of CFRP aircraft panel: A case study
Joost Segers (Mechanics of Materials and Structures, Ghent University)
Composite materials (e.g. carbon fiber reinforced polymer CFRP) are increasingly used for critical components in several industrial sectors (e.g. aerospace, automotive,…). A major challenge is the detection of internal damages in these composites which may have occurred during manufacturing or during operational life. One possibility for damage detection in thin-walled composite structures is to analyze the elastic wave characteristics measured on the surface. Many different wave actuation and sensing configurations are possible combined with advanced data processing methods. In this study, a single piezoelectric actuator is used for excitation combined with a 3D scanning laser Doppler vibrometer to measure the resulting full wave field. Depending on the frequency of excitation, different approaches are used. At relatively low frequencies, i.e. 5 to 100 kHz, damages can be localized by searching for local defect resonances (LDR). In this case, the full field steady-state vibration is analyzed in the frequency domain. At higher frequencies, i.e. 100 to 300 kHz, wave propagation analysis is conducted in time domain to search for defect-wave interactions e.g. wave amplification, wave scattering, change in local wave number, … In this study, the applicability of full wave field signal processing techniques for the detection of production defects in a CFRP aircraft panel with stringers (A320 tail spar) is investigated. Multiple data processing strategies (e.g. automated LDR detection and acoustic wave number spectroscopy) are used and the damage identification and localization performance is critically compared and evaluated. Acknowledgments: Joost Segers is a PhD fellow of the Research Foundation-Flanders (FWO, Grant no. 1148018N). The authors further acknowledge FWO (through grants 12T5418N) and the SBO project DETECT-IV (Grant no. 160455), which fits in the SIM research program MacroModelMat (M3) coordinated by Siemens (Siemens PLM software, Belgium) and funded by SIM (Strategic Initiative Materials in Flanders) and VLAIO (Flemish government agency Flanders Innovation & Entrepreneurship).
Télécharger le document.
12:05
FRI.2.B.5
Laser-generated shock waves for NDE : application to structural bonding and delamination of composite materials
Mathieu Ducousso (Safran Group)
We propose in this study to use intense acoustic shock waves to evaluate mechanical strength of structural bondings or to create delamination in monolithic composite materials. Such last application is dedicated to the creation of delaminated reference samples for NDE procedure and to allow validation of SHM procedure. The shock waves were generated using ns-scale intense laser pulses in the GW/cm2 intensity range. Such illumination creates a plasma expansion from the sample surface and, thanks to reaction principles, an intense stress, in the GPa order, propagates into the bulk of the material. The sample surface is protected from laser and plasma radiation using an aluminum tape. After back face reflection, phase change occurs, and the compressive shock wave becomes a release shock wave. Such release wave can be used to create a local tension, high enough to evaluate mechanical strength of a structural bonding[1] or to delaminate a composite material[2]. A symmetrical laser shock approach on the structure of interest with controlled time delay between laser pulses can be used to localize the maximal tensile strength anywhere in the structure. First numerical and experimental works are presented here. Bondings are structural epoxy bonds of different mechanical strength between a TA6V4 titanium alloy and a 3D woven carbon/epoxy composite material. composite materials are CFRP of thicknesses from 1.8 to 6 mm. Results demonstrate the high potential of laser-generated shock waves for NDE. References [1] M. Ducousso, S. Bardy, Y. Rouchausse, T. Bergara, F. Jenson, L. Berthe, L. Videau, N. Cuvillier, Quantitative evaluation of the mechanical strength of titanium/composite bonding using laser-generated shock waves, Appl. Phys. Lett., 112, 2018, 111904 [2] M. Ghrib, L. Berthe, N. Mechbal, M. Rébillat, M. Guskov, R. Ecault, N. Bedreddine, Generation of controlled delaminations in composites using symmetrical laser shock configuration, Composite Structures, 171, 2017, 286
Télécharger le document.
FRI.2.C
Materials Characterisation
AUDIT C

Materials Characterisation

Session chairs :
Room : AUDIT C
10:45
FRI.2.C.1
Contribution of ultrasonic NDT to the assessment of the effect of out-of-plane waviness defects on the mechanical strength of thermoplastic composite materials
Camille Trottier (Onera)
In the aerospace industry, mass reduction has been an important economic and environmental issue for many years. The replacement of metallic materials by composite materials for mechanically stressed structural parts is a real technological challenge. In the aeronautic industry, the need for increased production rates and reduced costs led to the introduction of thermoplastic resins (TP) which can be used to develop complex-shaped structures. However, the shaping of TP prepregs can create severe damage such as fiber breakage, disorientation of the layer, in plane or out-of-plane waviness. The consolidation step of the manufacturing process can also generate defects such as dry or resin-rich areas and high porosity rate in intra- or inter-layer areas. The optimal use of such complex composite structures requires both non-destructive and mechanical testing to detect and characterize these defects. The aim of the present paper is to closely link ultrasonic NDT measurements to the effect of defects on the structural mechanical integrity. In this study, various defect manufacturing techniques are used to generate calibrated out-of-plane waviness defects in carbon-fiber reinforced TP composite plates. All manufactured plates are inspected and characterized by NDT: first, ultrasonic testing to precisely locate and size the defects; then, laser vibrometry to measure the mechanical modulus. The accuracy of those NDT measurements is then validated by comparisons with optical microscopy observations. Finally, compression tests are carried out on several specimens with and without defects so that the non-destructive characterizations can be linked to the loss of mechanical properties induced by out-of-plane waviness.
Télécharger le document.
11:05
FRI.2.C.2
Non-Destructive Magnetic Testing of the Physical and Mechanical Properties of Mission-Critical Mounting Components in Aviation Technology
Sergey G. Sandomirski (Joint Institute of Mechanical Engineering of the NAS of Belarus, Minsk, Belarus)
Mounting products with mission-critical functions (bolts, studs, rods) used in aviation technology are subject to increased strength, hardness, and wear resistance requirements. They are made of medium carbon steels alloyed with chromium, manganese, or silicium (steel 37Cr4, 41Cr4, 46Cr2, 42CrMo4, and others). The balance between the strength and plastic properties of the products is achieved by the choice of the mode of their heat treatment - quenching and subsequent medium or high temperature tempering. Possible deviations in the chemical composition of the material of the products and the modes of their heat treatment from the prescribed ones lead to unacceptable deviations of the properties of the products. This requires monitoring all products. The physical basis of magnetic structural analysis lies in the fact that mechanical and magnetic properties of steels are sensitive to the structural transformations occurring in them during thermal treatments. It has been proven that the coercive force Нc and the remanent magnetization Мr of many steels are related to their structure. When controlling mass batches of products, the best results in reliability and productivity are achieved by magnetization of products when they fall through a coil with direct current and measurement of the remanent magnetic flux Fd in the product. In this case, Fd in products is not proportional to Mr, but to Hc of the material of the products. A characteristic feature of alloyed steels with a carbon content greater than 0.3%, including steels 37Cr4, 41Cr4, 46Cr2, 42CrMo4 and others, is a non-monotonic change in Hc with an increase in the tempering temperature Tt of products. Therefore, magnetic quality control of products from such steels requires a specific process. To test the tempering mode of such products, we have proposed to expose the magnetized products to a graded influence of a demagnetizing field with intensity Нp, before measuring Fd. The developed method allows us to choose the value of Hp, at which the optimum sensitivity to changes in Tt is achieved while monitoring products of specific sizes. In this report we show that such testing is based on the sensitivity of Fd in the product after its reverse magnetization in the field Hp to the remanent magnetization of the material Mr. We also report the technical means ensuring application of the developed methodology in industrial settings with a control output of up to 2 products per second. Examples of application of the method for controlling the hardness of small bolts made of steel 42CrMo4, intended for use in aircraft construction, are given.
Télécharger le document.
11:25
FRI.2.C.3
Materials Fatigue Characterization through Feature Extraction from Thermographic Images
Alexandr Lozak (Saarland University, Saarbrücken/Germany)
Developments in digital image recording and processing as well as the increased sensitivity of sensing systems has opened a variety of opportunities within the scope of NDT and structural assessment. One of those opportunities is the recording of thermographic images during fatigue tests. When done appropriately thermographic topographies can be obtained that can provide a variety of relationships between, temperature and stresses applied, plastic deformation zones developing, damage to accumulate, cracks to emerge and possibly much more. Traditionally a single fatigue test is considered as a single data point result. However, this may not be required for a thermographic image since it shows the temperature generated within the specimen as a result of fatigue loading assuming all other external thermal influences to be excluded. If a specimen’s loading condition is known then the thermal information recorded could be well correlated. This virtually allows this thermal information to be segmented leading to more than a single data point to be retrieved from the experiment. In the paper to be presented thermal images of “most indicative” zones will be shown first in terms of their topography and its respective change. Such changes can be clear damage indicators i.e. cracks. However, are there possibly more damage related indicators to retrieve including possibly a statistic of some material’s intrinsic properties? How does such a statistic change in case smaller segments on the specimen surface are considered for evaluation? Some thought will be also spent on how to retrieve the information from a single un-notched specimen first before moving to a special five segment mainly un-notched specimen in a second step. This comparison will be based on different feature extraction, signal and image processing, and classification techniques to be applied It will be shown how much the thermographic image can be sectioned such that the information retrieved is still representative when compared to analyzing the complete un-notched specimen. Feature extraction and image processing techniques in terms of smoothening, noise elimination and outlier determination will be addressed as well as related statistics, feature maps, dynamic representation and behavior of features all in relation to material properties. Machine learning algorithms, deep learning or neural networks are further prospective aspects addressed.
Télécharger le document.
11:45
FRI.2.C.4
Micro-magnetic non-destructive testing: aerospace bearing application
Benjamin Ducharne (LGEF)
The quality standards improvement in the aerospace field have revealed a real industrial need for precise and systematic control of the produced and manufactured steels. Different techniques for the control of the manufactured products quality already exist, the so-called “non-destructive testing and evaluation (NDT&E) techniques”. The objective is to evaluate both internal and external states of the tested samples without affecting their integrities. The need for NDT is becoming more prominent. Its evolution is linked to the increasing complexity of the new industrial products geometries and to the instrumentation and computer science improvements. This evolution is also related to the evolution of society preoccupation. In the aerospace bearing industry, the nondestructive testing techniques based on the “local magnetic signature” constitutes a real alternative to the classic NDT methods like the traditional chemical baths which reveal defects in the form of nuances of gray spots (precipitated carbon) and which are generally quite difficult to interpret. The micro-magnetic methods are nonpolluting methods, they consist in controlling the structural modifications of a test sample under high magnetic field excitation [1]-[4]. Using local dedicated magnetic sensors, we can establish a precise cartography of the controlled structure and the treated parts. In SKF Company, France, Barkhausen noise is by instance used since many years now. It is particularly efficient for the control of contact zones between the rolling and the set elements of the bearings. The current industrial uses of the Barkhausen’s noise in particular and of the micro-magnetic NDT&E methods in a general way are very empirical; operators set thresholds of rejection using experimental results but no accurate models are available to refine these thresholds and to improve the phenomena understanding. The purpose of our work is to develop local accurate models of the coupled mechanical/magnetic and of the microstructure to improve the understanding of the material behavior [5]. This will help to refine and generalize the use of the micromagnetic signature control as a method for nondestructive testing. [1] B. Gupta & al., J. of Mag. and Mag. Mat., vol. 486, pp. 165250, 2019. [2] B. Gupta & al., NDT & E Int., vol. 104, pp. 42-50, 2019. [3] B. Ducharne & al., IEEE Trans. on. Mag, vol. 99, pp. 1-6, 2018. [4] B. Ducharne & al., J. of Mag. And Mag. Mat., pp. 231-238, 2017. [5] T. Matsumoto & al., AIP Advance, vol. 9, Iss. 3, 2019.
Télécharger le document.
12:05
FRI.2.C.5
Ultrasonic guided waves as a tool for the evaluation of effective elastic moduli of laminate composite materials
Artem Eremin (Institute for Mathematics, Mechanics and Informatics, Kuban State University, Krasnodar, Russia)
Non-destructive assessment of the effective elastic properties of laminate composite structures is motivated by various practical needs such as the demand for reliable input data for numerical simulation and design, control of material degradation during structure operation, etc. While propagating over a thin-walled structure, elastic guided waves (GWs) carry information about its material properties, i.e., elastic moduli, density, thickness, and lamination scheme. This provides a basis for the development of appropriate assessment strategies for ultrasonic NDT and SHM systems, especially in view of the achievements in non-contact and in-situ approaches for the GW excitation and sensing using laser and piezoelectric emitters and receivers. This approach has been implemented for the evaluation of effective parameters of laminate carbon fiberreinforced plastic (CFRP) composites fabricated from identical unidirectional plies. The developed method is based on the use of the Green matrix of the multilayered anisotropic structure under study in the Fourier transform domain. The effective elastic moduli and, if required, the thickness and density of each sublayer are obtained via the minimization of certain goal functions that specify the discrepancies between the measured and calculated GW characteristics and/or the polar sets of Green’s functions. Previously, we investigated the sensitivity of dispersion properties of fundamental and high-order GWs for various propagation directions to the variation of certain elastic constants in the sublayer stiffness matrix for different typical lamination schemes. As a goal function, we tested two approaches: the sum of deviations between the calculated and measured group velocities, and the inverted Green’s matrix elements, which must be zero at the poles yielding GWs. The practical implementation of the developed method is discussed and illustrated by examples of effective material property restoration for several unidirectional, cross-ply and quasi-isotropic CFRP samples. In the experiments, GWs are excited by surface-attached piezoelectric wafer active sensors (PWAS) and acquired with a non-contact scanning laser vibrometer or other PWASs arranged in a distributed sensor network. The results obtained are validated against destructive tensile tests as well as by comparing the simulated and measured transient signals carried by excited GWs.
Télécharger le document.
12:30
Shuttles Departure for Industrial Visits

Materials Characterisation

Session chairs :
Room : AUDIT C
12:30
SAFRAN Aircraft Engines
12:30
AIR FRANCE Engine Part Repair Center
The Visit Schedules :
  • 12h30 : Shuttle departure with Lunch on Board
  • 13h30-15h30 : Visit of the site
  • 17h30 : Approx. Arrival at orly Airport
This will be a unique occasion to discover some of our synchrotron techniques and to dialog with experts from a state-of-the-art light-source for materials study in Europe. Visitors will have the occasion to discover the following laboratories:
  • Fluorescent Penetrant Testing using Sensitivity Level 2 or 4 Post emulsifiable, apply by immersion
  • Fluorescent Penetrant Testing using Sensitivity Level 2, 3 or 4, Post emulsifiable, apply by electrostatic
  • Radiography inspection (Film Radiography and Digital Radiography)
  • Magnetic Particle Inspection
  • Ultrasonic Inspection
12:30
ARIANE GROUP
12:30
Synchrotron Soleil
The Visit Schedules :
  • 12h30: Lunch-snack at CentraleSupélec
  • 13h15: Shuttle departure
  • 13h30-15h30: Visits of the site
  • 17h00: Approx. Arrival at Orly Airport

The tour will start with a plenary presentation of SOLEIL introducing the platform through  its history and its overall activity. Afterwards, the visit will focus on some of our 29 beamlines and laboratories with high interest for Aerospace testing and measures. This will be a unique occasion to discover some of our synchrotron techniques and to dialog with experts from a state-of-the-art light-source for materials study in Europe. Visitors will have the occasion to discover the following laboratories:

  • HERMES, X-ray nano-microscopy and nano-spectroscopy beamline used for bulk and surface chemical analysis,
  • DIFFABS, X-ray beamline using a coupled approaches in diffraction and fluorescence spectroscopy,
  • ANATOMIX, X-ray nano-micro-tomography beamline with absorption and phase contrasts for advanced non-destructive observations,
  • Surface Laboratory, with STM microscope, vapor deposition in UHV chamber, Auger spectrometer, Argon ion gun, LEED …
12:30
Industrial Visits (detail of the tours)

Industrial Visits (detail of the tours)

Session chairs :
Room :
12:30
SAFRAN Aircraft Engines
The Visit Schedules:
  • 12h30: Shuttle departure with Lunch on Board
  • 13h30-15h30/16h: Visit of the site
  • 17h30/18h: Approx. Arrival at orly Airport
12:30
AIR FRANCE Engine Part Repair Center
The Visit Schedules:
  • 12h30: Shuttle departure with Lunch on Board
  • 13h30-15h30/16h: Visit of the site
  • 17h30/18h: Approx. Arrival at orly Airport
CRMA is a wholly-owned subsidiary of the AIR FRANCE KLM Group specializing in the repair of CFM56, GE90 and GP7200 components and engine modules and the overhaul of a wide range of aircraft systems. With its experienced personnel, innovative engineering division and cutting-edge facilities, CRMA brings airline industry professionals flexible repair and logistics support solutions tailored to their unique requirements. By delivering top-quality services and focusing on adaptability and competitive pricing we guarantee the best possible performance for our customers. In maintenance operations performed on the engines, the NDT applied are:
  • Fluorescent Penetrant Testing using Sensitivity Level 2 or 4 Post emulsifiable, apply by immersion
  • Fluorescent Penetrant Testing using Sensitivity Level 2, 3 or 4, Post emulsifiable, apply by electrostatic
  • Radiography inspection (Film Radiography and Digital Radiography)
  • Magnetic Particle Inspection
  • Ultrasonic Inspection
Télécharger le document.
12:30
ARIANE GROUP
The Visit Schedules:
  • 12h30: Shuttle departure with Lunch on Board
  • 14h-16h/16h30: Visit of the site
  • 18h30/19h: Approx. Arrival at orly Airport
12:30
Synchrotron Soleil
The Visit Schedules:
  • 12h15: Lunch-snack at CentraleSupélec
  • 13h15: Shuttle departure
  • 13h30-15h30/16h: Visits of the site
  • 17h/17h30: Approx. Arrival at Orly Airport

The tour will start with a plenary presentation of SOLEIL introducing the platform through  its history and its overall activity. Afterwards, the visit will focus on some of our 29 beamlines and laboratories with high interest for Aerospace testing and measures. This will be a unique occasion to discover some of our synchrotron techniques and to dialog with experts from a state-of-the-art light-source for materials study in Europe. Visitors will have the occasion to discover the following laboratories:

  • HERMES, X-ray nano-microscopy and nano-spectroscopy beamline used for bulk and surface chemical analysis,
  • DIFFABS, X-ray beamline using a coupled approaches in diffraction and fluorescence spectroscopy,
  • ANATOMIX, X-ray nano-micro-tomography beamline with absorption and phase contrasts for advanced non-destructive observations,
  • Surface Laboratory, with STM microscope, vapor deposition in UHV chamber, Auger spectrometer, Argon ion gun, LEED …
Télécharger le document.

Programme

Final programme is now online,

DOWNLOAD THE PROGRAM

 

INDUSTRIAL SITES VISITS:

On November 13th, participants will be welcomed in CEA List for an exclusive visit of the labs, followed by a cocktail dinner :

DOWNLOAD THE PROGRAM

On Novemver 15th, participants will have the choice between 4 industrial sites to visit. Each visits will last between 2hOO & 2h30. The detailed information will be published soon, in between find below the bus shuttles timetables.


REGISTRATION TO THE VISITS

TO NOTE : Identity cards or passport will have to be presented at the entrance of the industrial sites.

NDT AEROSPACE SHUTTLES

A shuttle system will be set up from 13 to 15 November 2019. It will provide the following transfers:

  • Round Trip Novotel Paris Saclay to NDT Aerospace
  • Round trip transfers for conference dinner on Nov 14, back to Saclay and Novotel
  • On the 15th, transfers to various industrial sites and back Saclay and Orly airport (connection with Air France shuttles to Roissy Charles de Gaulle airport or public transport: Orly Val, RER)

A more detailed timetable schedule will come later

CONFERENCE DINNER

Welcome in the Vaux de Cernay Abbey.

Located in the Chevreuse valley in the Rambouillet state-owned forest, is an enchanted place. Founded in 1118 by a group of monks from Savigny Abbey (in Manche) the Vaux-de-Cernay Abbey enjoys more than eight centuries of history.

Shuttles are planned to the Conference dinner from and to CentraleSupelec and to Novotel in Saclay.

 

Scientific Committee

  • Christian Boller Saarland University, Germany
  • Matthias Buderath Airbus, Germany
  • Pierre Calmon CEA List, France
  • Fu-Kuo Chang Stanford University, USA
  • Younho Cho Pusan National University, South Korea
  • Esmeralda Cuevas Tecnatom, Spain
  • Nicolas Dominguez Airbus, France
  • Zheng Fan Nanyang Technological University, Singapore
  • Gary Georgeson Boeing, USA
  • Christian Große TU München, Germany
  • Srinivasan Gopalakrishnan Indian Institute of Science, India
  • Alfredo Güemes Polytechnic University of Madrid, Spain
  • Henning Heuer Fraunhofer IKTS, Germany
  • Frédéric Jenson Safran Tech, France
  • Johann Kastner FH Oberösterreich, Austria
  • Wieslaw Ostachowicz Polish Academy of Sciences, Poland
  • William Prosser NASA, USA
  • Matthias Purschke DGZfP, Germany
  • Xinlin (Peter) Qing Xiamen University, China
  • Lei Qiu Nanjing University of Aeronautics and Astronautics, China
  • Robert Smith University of Bristol, UK
  • Zhonquing Su Hong Kong Polytechnic University, Hong Kong
  • GuiYun Tian Newcastle University, UK
  • Shenfang Yuan Nanjing University of Aeronautics and Astronautics, China
  • Zhenggan Zhou Beihang University, China

Organising Committee

  • Pierre Calmon CEA List, France, (Chairman)
  • Christian Boller Saarland University, Saarbrücken, Germany, (Co-chair)
  • David Barnoncel Ariane Group, France
  • Bastien Chapuis CEA List, France
  • Nicolas Dominguez Airbus, France
  • Florence Giraud COFREND, France
  • Frédéric Jenson Safran Tech, France
  • Jean Marie Morvan Dassault Aviation, France
  • Christophe Parsis, Afenda FrANDTB
  • Michael Stamm Brussels Airlines/KU Leuven, Belgium (Student Challenge)