NDT 2007 Abstracts: 3C-4C
Wednesday 19 September

Session 3C – Data Processing

Long term data storage management – the options
Robin Evans
Key Account Manager, GE Inspection Technologies. T: +44 7967 388976; E: robin.evans@ge.com

The increasing use of ever more powerful software and analysis tools for varying NDT methods and the need to store larger and larger data files is creating a corresponding storage management problem. Files for digital and computed radiography can regularly be 120 MB per image and the desire to store complete waveforms from ultrasonic scans is generating masses of data.

These files have to be managed for access and also for long term storage. This may be for a few months, a few years or even the complete life of the product plus, which can be twenty years or more.

This presentation will look at some of the options available and compare the advantages and disadvantages of these systems to try and find the optimum solution to short term problem with long term consequences.

Data fusion in automated inspection systems
Markus Friedrich, Walter Galbraith, Gordon Hayward
Centre for Ultrasonic Engineering, Department of Electronic and Electrical Engineering, University of Strathclyde, 204 George Street, Glasgow G1 1XW, UK. E: markus.friedrich@eee.strath.ac.uk

Teams of small modular inspection vehicles for automated inspection tasks offer the possibility of employing a variety of different NDE inspection methods simultaneously. By synergistically utilising information derived from multiple sources, individual deficiencies and limitations can be compensated for, leading to a more accurate and precise evaluation of the condition of the engineering structure under test. This paper presents approaches based on fusion of NDE data that have been obtained by a heterogeneous team of small inspection robots which are equipped with payloads for magnetic flux leakage, eddy current and ultrasonic inspection. Any potential uncertainties in individual measurements regarding the location of defects constitute the basis for fusion methods based on statistical and probabilistic algorithms. Images of a two-dimensional test structure have been constructed from data derived from different scans, indicating the positions of detected artificial defects. Applying Dempster-Shafer Theory of Evidence and Bayesian Analysis, the confidence level in the accuracy of these images is increased and the uncertainty reduced. The fused data results show particular improvement in image quality, when the underlying inspection technique is characterised by poor reliability.

Potential drop data inversion for crack depth profiling
Giuseppe Sposito1, Peter Cawley1 and Peter B Nagy1,2
1RCNDE, Department of Mechanical Engineering, Imperial College London, London SW7 2AZ, UK
2Department of Aerospace Engineering and Engineering Mechanics, University of Cincinnati, Cincinnati, OH 45221

Estimating the shape and size of a defect is a problem of major interest in many industrial applications, since the depth of a crack is often a key parameter in calculations of structural integrity. Previous studies have shown that the potential field created on the surface of a test-piece by the injection of direct or alternating currents for PD measurements can be calculated with a simple three-dimensional Finite Element (FE) model; this model is therefore able to give an accurate solution to the direct problem of predicting the response of a probe to a surface-breaking defect of known geometry. Most previous work on the inverse problem of using values of transfer resistance measured at a number of different locations to calculate the depth profile of an unknown defect have assumed a priori knowledge of the defect shape; crack gauges are commercially available that assume the defect has a semi-circular form. However, this assumption is not always correct. The aim of the present study is to develop an inversion technique of more general validity. The results of both FE analyses and experimental tests on specimens with EDM notches of various shapes and sizes were used to develop an inversion algorithm, which was subsequently applied to experimental data obtained with an array probe on samples with laboratory-grown stress corrosion cracks (SCC), in order to reconstruct the depth profile of the cracks.

Monte Carlo solution of the ultrasonic defect characterisation inverse problem
Stephen Mein and Bogdan Matuszewski
Department of Technology, University of Central Lancashire, Preston, PR1 2HE. E: sjmein@uclan.ac.uk

An important problem in the non-destructive testing of materials is the accurate determination of the location, size, shape and orientation of any detected flaws; this is essential for assessing integrity of inspected objects. This paper presents an algorithm wherein recovery of the geometric parameters of flaw captured in pulse-echo ultrasonic inspection data for immersion testing at low angles of incidence is treated as an inverse problem.  The forward problem of modelling the ultrasonic echo signals is given by the Impulse Response Method, a frequency independent numerical solution to the acoustic wave equation. To solve the inverse problem, Markov Chain Monte Carlo methodology is employed as an iterative optimisation strategy, minimising a cost function between observations and a projected model. Modelled flaw parameters are updated stochastically such that they converge toward the true values over successive iterations.  The advantages of this approach over standard deterministic algorithms are an ability to circumvent local minima in the cost function and a means to incorporate a priori knowledge into the solution space. Results are shown for simulated and real test data, demonstrating converge properties of algorithm and contrasted with those for standard deterministic optimisation strategies thus illustrating the potential of the method.

Analysis of rolling contact fatigue cracks on rails using computed tomography
M Ph Papaelias1, J Garnham1, C Forrest2, S Kenny2, C L Davis1
1Rail Research UK Centre, Gisbert Kapp Building, University of Birmingham, Birmingham, B15 2TT
2TWI Technology Centre (Wales), ECM2, Margam, Port Talbot, SA13 2EZ, UK

Rolling contact fatigue (RCF) cracks develop in the railhead due to the severe stresses experienced during vehicle passage. The cracks initiate at shallow angles to the rail surface and, after propagation to depths of around 5 mm, can turn up, leading to spalling, or turn down into the railhead potentially leading to rail breaks. The cracks are also known to have multiple branches resulting in complex three-dimensional shapes. Crack shape is currently determined using sectioning of the railhead and microscopy or by breaking open the rail and observing the fatigue and fracture surfaces. In the former case only a two-dimensional view is obtained and in the latter only the dominant crack is observed.

This paper discusses the initial results obtained for a number of rail head samples with moderate to severe RCF cracks using computed tomography to reveal the three-dimensional nature of the cracks. ACPD measurements and metallographic analysis of the samples examined are also presented as a means of comparison to the radiographic results. The results obtained up-to-date demonstrate the possibility of using computed tomography as a non-destructive investigative tool for the examination and evaluation of RCF cracking.

SRAS: imaging grains and microstructure using non-contact ultrasonics
Steve D Sharples, Matt Clark, Mike G Somekh
Applied Optics Group, School of Electrical & Electronic Engineering, University of Nottingham, University Park, Nottingham NG7 2RD. T: 0115 9515220; F: 0115 9515616; E: steve.sharples@nottingham.ac.uk

The mechanical performance of many materials is heavily influenced by their microstructure. In terms of metals, this is related to the size, degree of randomness, and orientation of the grains that make up the material. The ability therefore to image the microstructure and to get an indication of grain orientation is of profound importance to those in industries where materials are working at their mechanical limits. It may also be of interest from a process control perspective for those developing and refining the production of certain high performance alloys.

SRAS (spatially resolved acoustic spectroscopy) is a relatively new technique that can rapidly image material microstructure, currently with lateral resolutions of a few hundred microns. The technique uses lasers to excite and detect high frequency surface acoustic waves, and accurately measure their velocity to provide contrast between grains, and give an indication of orientation. Despite still being a lab-based instrument, the technique is fast, non-contact, completely non-destructive, can potentially be used on rough surfaces and image very large samples. As well as describing the technique and instrumentation in detail, striking images are presented to illustrate the capabilities of the technique.

Quantitative analysis of flash thermography data
Steven M Shepard
Thermal Wave Imaging, Inc., Ferndale, MI, USA
 
In the past decade, flash thermography has evolved from its origins as a qualitative adjunct to traditional NDT methods to become a widely accepted first-line inspection methodology. Although a limited amount of information about the subsurface condition of a sample can be gained by viewing the infrared image sequence of the sample surface after flash heating, modern thermographic NDT systems treat each pixel as an independent temperature-time history of a point on the sample. The Thermographic Signal Reconstruction (TSR) technique has been remarkably effective in facilitating analysis of single-pixel data. In addition to significant temporal noise reduction, the TSR time derivatives allow unambiguous measurement of depth, thickness and porosity, as well as characterisation of interfaces and signatures associated with particular sample types.  Unlike most other flash thermography analysis techniques, TSR is based on absolute, invariant characteristics of the signal, and does not depend on the presence of contrast in the image, so that even a single pixel contains useful information about the host sample. The net result is detection of features far beyond the sensitivity achievable by direct viewing of the image sequence, or by 2-dimensional image processing of the constituent images.

Session 4C – Conformable/Flexible Arrays

Advanced matrix phased arrays for inspection
Philippe Benoist1, Philippe Dubois1, Sylvain Chatillon1, Laurent Le Ber2 and Olivier Roy2
1CEA/LIST, Gif-sur-Yvette, F-91191, France
2M2M, Les Ulis, 91940, France

Phased arrays are now widely used in ultrasonic non-destructive testing, as they provide improved versatility with comparison to conventional monolithic probes. Although these techniques are quite well known in the NDT field, most applications have been carried out with linear phased arrays (1D phased arrays). More complex applications now require now advanced phased array techniques based on matrix array design, which imply increased number of elements and improved phased arrays driving software and systems. Developments carried out at CEA and M2M aim at providing user-friendly tools to conceive and to fully exploit electronic commutation and scanning features over matrix arrays (2D phased arrays), for which one can easily define groups of transmitters and receivers, draw their electronic scanning trajectories in the full array pattern, and modify their active aperture as well as applied delay laws at transmission and/or reception. The collected data may therefore be displayed in the real component frame thanks to 3D CAD features based upon paths associated to each ultrasonic shots carried out during the acquisitions. Several inspection configurations performed over complex components (nozzles, CAD specimen…) are presented, both using simulation and experimental data, which illustrates some potential applications of matrix array settings.

Flexible arrays for ultrasonic testing of solid high attenuative cased material
Andre Schilde1, Katherine Kirk1, Alan Fenwick2 and Lyn Jones2
1Microscale Sensors, School of Engineering and Science, University of Paisley, Paisley, Scotland, PA1 2BE
2NDE group, QinetiQ plc, Farnborough, England, GU14 0LX

Flexible arrays are new developing ultrasonic transducer technology to insonify complex structures[1,2]. In this project we aim to detect cracks and flaws in solid rocket motor fuel which is encased in cylindrical tubes of varying diameters. Previous work on flaw detection in rocket motors has been reported in [3] and we will expand this to use a flexible phased array.

The flexible array will be based on a large flexible PZT piezocomposite, with a continuous bottom electrode and discrete linear elements for top electrodes. The array will be operated with linear frequency modulated (LFM) signals, e.g. a chirp, for better penetration and to increase the signal to noise ratio.

Measurements have been taken on solid cylinders and tubes by wrapping a contact probe in a sector around the curved surface. The piezocomposite probe was made from PZT 5R surrounded by the flexible epoxy CY221. Results were compared with measurements from commercial probes in immersion experiments using the insertion loss method. We found out the behaviour and appropriate frequencies of this device and investigated the measurement configuration using ray tracing software.

[1] D J Powell, G Hayward, IEEE Trans. UFFC, 43(3), 393-402, 1996.
[2] O Casula, C Poidevin, G Cattiaux, Ph Dumas, Ultrasonics 44, 647-651, 2006.
[3] T H Gan, D A Hutchins, D R Billson, F C Wong, IEEE Ultrasonics Symposium, 823-826, 2000.

Further development of a conformable phased array device for the inspection of irregular surfaces
R Long and P Cawley
RCNDE, Imperial College

A conformable phased array device that allows reliable ultrasonic inspection of components with irregular surfaces such as weld caps has been developed. The device uses a standard linear phased array transducer which is coupled to the surface under test by a water path encapsulated by a low loss synthetic rubber membrane. The reservoir of water encapsulated in the device can be subjected to either constant pressure, by use of a header tank, or constant volume. The profile of the surface under test is measured utilising the phased array incorporated into the device. The computation of updated delay laws, required to recover the inspection performance when inspecting over an irregular surface, has been improved by employing the Fermat minimum time principle. The research incorporated beam modelling using the CEA CIVA software and comparisons with experimental measurements. It is shown that the conformable membrane device using a standard array transducer provides a solution for ultrasonic inspection around the weld region of welded pipes.

Transducer array for integration with miniature electronics for NDT
J F Saillant1, S Triger2, J Wallace2, S Cochran1 and D Cumming2
1University of Paisley, Paisley, Scotland, UK
2University of Glasgow, Glasgow, Scotland, UK

Ultrasonic inspection in industry has for decades been performed using mono-element transducers. This technology, whilst being very useful, is highly subjective and operators can be led to misinterpretation. The concept of 3D imaging systems shows great potential for reducing those risks and improving the quality of the inspections. However, the technology involved in such systems is several degrees of complexity higher than that of monoelement transducer ones. This presents issues in the system’s main technical areas: design and fabrication of the electronics, image processing, and the design and fabrication of the transducer array.

This paper will focus on the particular issue of the design and fabrication of the 2D arrays for an ultrasonic mosaic transducer system[1]. This system can be seen as a mosaic in which each tile, in the present case, includes a 16-element piezocomposite transducer array integrated with the electronics necessary for full transmit-receive capability on all 16 channels. The beam forming capability of this mosaic can then be varied according to the number of tiles directly juxtaposed next to one another laterally and the software used to define the system’s operation.

One particularity of these ultrasonic tiles is that they can fit different transducer designs, as long as the array dimension matches the footprint of the electronics and that it has 16 elements or less. Monolayer and multilayer configurations are considered in this paper and the benefits of multilayer technology are explained. The different stages of the design of the fabrication of the array are described, including the design of the piezocomposite, matching layer and backing materials, array fabrication and finally transducer characterisation.

[1] S Triger, J Wallace, L Wang, S Cochran, J Saillant, F Afroukh and D R S Cumming, 'A modular FPGA-based ultrasonic array system for applications including non-destructive testing', these proceedings.