Ultrasonic Time-of-Flight Diffraction was invented in the early 1970s and initially developed as a research tool. Its rate of development was dramatically changed by the decision at the beginning of the 1980s to plan for a Pressurised Water Reactor (PWR) in the United Kingdom. Although such reactors were common in other countries, a considerable body of opinion in the UK was sceptical of the safety of PWRs. A thorough safety case was therefore required to present to the public enquiry which was almost inevitable once a site for the power station had been named.
Nuclear reactors of the PWR type have thick steel walls withstanding considerable internal pressure. It is therefore necessary to establish with a very high level of confidence that there are no cracks bigger than the critical size, in the parent metal, or in the welds. At about the time that the decision to build one of these reactors in the UK was taken, results were published which suggested that conventional ultrasonic inspection techniques could not size planar cracks bigger than the critical size as accurately as would be necessary to achieve the confidence level required.
This led the nuclear industry in both the United Kingdom and Europe, to invest heavily in a research and development programmes aimed at improving ultrasonic inspection of thick-section steel. The programme in the UK covered conventional ultrasonic inspection techniques but also devoted considerable effort to ultrasonic Time-of-Flight Diffraction because it had already shown great promise as a tool capable of accurately sizing planar, through-wall cracks -exactly what was required.
The first edition of our book, published in 1989, came at a time when much of the development work had been completed and several test-block trials had also been undertaken. The technique had proved itself and was being adopted as one of the essential tools, alongside enhanced pulse-echo inspection, for nuclear reactor inspection. Out hope then was that the technique would spread into other industrial sectors. In the intervening years, this has taken place and the technique is now a mature one.
As we enter a new millennium, it seems the right time to bring our exposition of the technique up-to-date. To do this we have kept much the same form as the previous edition, starting with the theoretical background. One of the strengths of ultrasonic Time-of-Flight Diffraction is that theoretical understanding was developed at an early stage and this has been used consistently to develop the inspection techniques used in real applications. The technique, if used correctly, is capable of yielding very accurate measurements of crack size but, to achieve this, it is necessary to have a good understanding of potential sources of error. We have therefore considerably extended the section on errors and how to minimise them.
Since the technique now has more data to back it up, both from more complex test-block trials and more realistic field applications, we have extended the sections covering both these aspects. As a mature technique it has begun to be specified in codes and standards and we have described the current status in this area.
No other industry has been pressing for such a thorough understanding as the safety case for a PWR required, so only a small amount of additional development work has been done since the first edition of the book. Somewhat surprisingly, some of the signal processing techniques that were covered in the first edition are still not regularly applied, despite computer processing power having increased a thousandfold since then. There is room for further work in this area to demonstrate what could be achieved with modern technology.
This book aims to provide a thorough background to the theory and practice of the technique and we hope that it will encourage an even wider range of applications and further advances in capability.
J. P. Charlesworth, Dartmouth, Devon
J. A. G. Temple, Upton, Oxfordshire
September 26, 2001
Chapter 1: Introduction
(14 Pages)
Whenever we tum on a domestic appliance, fill the petrol tank of a car, travel by road, rail, sea or air, we rely directly or indirectly on some equipment or structure working reliably under stress. For example: most electricity generation involves high pressure steam boilers heated either by the burning of fossil fuel or by a nuclear reaction; gas is transported from the North Sea to the users by high pressure pipelines; hydrocarbon fuels are produced in refineries containing much high pressure plant; most modern forms of transport rely on the integrity of components subjected to large and rapidly varying stresses.
Components are designed with more than adequate strength to resist the stresses arising in normal service and even to tolerate certain levels of abnormal conditions. When failure occurs, it is often because the component contained a defect, normally of a crack-like nature, sufficiently large to cause a major reduction of strength. Such defects may arise from faulty manufacture or the effects of service in a corrosive environment and may be enlarged by fatigue. To ensure their absence after manufacture or to detect them in service, a variety of non-destructive testing (NDT) techniques may be used. Of these, ultrasonic testing is the most widely applicable, being capable of detecting and sizing cracks in a wide variety of locations and orientations, in many materials used in engineering and even for considerable thickness of material (greater than 300mm in steel, for example). A particular type of ultrasonic testing technique is the subject of this book.
Ultrasonic testing makes use of high frequency, but very low amplitude, sound waves to detect, characterise and size defects in components. The sources and receivers of these ultrasonic waves are transducers, usually, but not always, made from a piezoelectric material which deforms under the application of a voltage. Applying a voltage generates a mechanical distortion which propagates into und travels through the component as a wave. When such a wave arrives at the receiver, the piezoelectric material converts this into a voltage which depends on the orientation and magnitude of the distortion.
Other methods of creating and detecting ultrasonic waves are possible, such as electromagnetic acoustic transducers (EMATs) which essentially use (electro)magnetostriction as the method of translating a distortion into a voltage and vice versa, or the use of lasers to ablate part of die surface to generate an ultrasonic pulse coupled with an interferometer to read the surface ripples on the component when signals arrive back. While most of what we discuss in this book is independent of the mode of generation or reception of the ultrasonic waves, we usually have in mind ceramic piezoelectric transducers.
The physical method of sending and receiving signals may be unimportant bot the characteristics of the signals generated and received can be important. As we shall see later, the pulse length, the angular spread of the ultrasonic beam, the polarisation of the waves in the signal and their phase are all important.
Pulse-echo ultrasonic inspection techniques rely on the arnplitude and range of a signal returned from the defect to the interrogating equipment in order for the defect to be detected, sized and, possibly, characterised. The process governing the amplitude is usually specular reflection, in which any crack acts like a mirror for die ultrasound. For a given arrangement of ultrasonic transducers on the component undergoing inspection, this process of specular reflection can only occur for a limited range of orientations of the defect. In the absence of a specular reflection, the signals retumed will be those arising from diffuse scattering from the surfaces of the crack and by diffraction from the edges of the crack. These diffracted signals are of particular interest, since, being associated with the extremities of the defect, they may be used to determine the size of the defect accurately and thus assess the integrity of the component. The ultrasonic Time-of-Flight Diffraction technique is based on the exploitation of the exploitation of these signals diffracted from the defect edges.
Chapter 2: Theoretical Basis of
Time-of-Flight Diffraction
Fig1: The two probe basis of the Time-of-Flight Diffraction technique
|
(35 Pages)
In this chapter we consider the technique from a theoretical point of view. We do not present detailed theory but we illustrate conclusions drawn from modelling work and discuss the way in which these conclusions affect the design of TOFD inspection. We give the types of waves which can propagate and examples of their wavespeeds. We explain elementary diffraction with emphasis on the radiation of the diffracted energy into a wide range of angles. This gives the technique one of its advantages over conventional methods of defect detection and sizing -its relative insensitivity to defect orientation. We explain how the TOFD technique is used to measure defect sizes and we discuss the accuracy of such measurements in considerable detail. We also describe some important features of the signals observed when a TOM probe assembly is scanned across the location of a defect. Finally, we very briefly describe ways of using diffracted signals with only a single transducer.
Chapter 3: Signal Amplitudes and
Comparison with other
Techniques
(19 Pages)
As we have already discussed, Time-of-Flight Diffraction makes use of the signals diffracted by the extremities of defects, in particular by crack tips. Although, unlike pulse-echo ultrasonic techniques, signal amplitude is not used as a measure of defect size, signals must be large enough to be observable. It is also useful to understand: how this amplitude might be affected by a variety of factors, such as the angle of the incident ultrasonic beam with respect to the plane of the crack; what happens if the crack is skewed so that the plane of the crack is not normal to the line joining transmitter and receiver transducers; and how large are the signals relative to more familiar signals such as those from a small flat-bottomed hole. We can build this understanding through a mathematical model of the underlying physical processes.
In this chapter we present predictions, based on just such a mathematical model, of the magnitude of Time-of-Flight Diffraction signals, compared with those from a reference reflector. We use a model for two reasons. First, it is much easier to vary the parameters of defects within a model than it is experimentally with test-blocks. Secondly, it is possible to isolate the different factors influencing the outcome in a more straightforward way. The theoretical predictions are compared with experimental data as appropriate.
The results of these calculations illustrate how the signals from a variety of cracklike defects are expected to vary with the shape, size and orientation of the defect and highlight one of the strengths of the technique: its relative insensitivity to the orientation of cracks.
The first results we shall present are for the case in which the centre of a planar crack and the centres of both transmitter and receiver lie in the same plane, normal to the inspection surface. This is not a severe restriction, since in almost all cases the probes will be scanned over the defective region and this configuration will be passed through during the scan. Sometimes the shape of the component may preclude reaching such a position and so we later present results applicable to less restricted geometry.
Chapter 4: Design of Time-of-Flight
Diffraction Equipment for
Simple Geometries
(13 Pages)
During the early stages of the development of the Time-of-Flight Diffraction technique on samples less than 90mm thick, little theoretical modelling had been done and the design of TOFD inspection systems was largely empirical. When the Defect Detection Trials (DDT) were planned, in which plates up to 250mm thick were to be inspected, the scale of the task, the short timescale and the weight which would be attached to the results made it imperative to use a more systematic approach to achieving optimum experimental design. This process has been described in detail for the two girth weld simulation plates (Plates I and 2) by Curtis and Hawker [ 1983] and for the near surface defect plate by Charlesworth and Hawker [ 1984]. That part of the work related to design of probe arrays and scanning processes is described in this chapter, while information on data display and analysis techniques appears in Chapter 5. The results of the Defect Detection Trials are discussed in Chapter 8. The additional design problems associated with more complex geometries are discussed in Chapter 6. Although the DDT work was done a long time ago, it is still of historic importance because exercises on this scale have not been repeated in recent times. There have however been some advances in inspection design and instrumentation which will be presented to bring the picture more up to date.
Chapter 5: Processing, Display and
Analysis of Time-of-Flight Data
(21 Pages)
In Chapter 4, we discussed the design of signal acquisition equipment for the Timeof-Flight Diffraction technique. In this chapter we shall describe the remainder of the inspection system, devoted to processing the signals and extracting information from them. The use of novel techniques of display and analysis has been one of the marked features of the development of Time-of-Flight Diffraction but many of the techniques have subsequently proved applicable to pulse-echo data.
In this chapter we shall confine ourselves to analysis techniques which are generally applicable and sufficient for a fall analysis in simple geometries like butt welds in flat plates or girth welds in cylindrical vessels. When the geometry is more complicated, the analyst needs some geometrical assistance from the system to help locate the sources of defect signals and this will be described in Chapter 6.
At the time of the Defect Detection Trials, it was the common practice to collect TOM data on systems which had few or no facilities for data analysis, the data being transferred, for analysis, to other computer systems containing what were then very expensive image display systems. As the cost and size of computers and image display equipment decreased, the analysis functions for TOM tended to be more and more integrated into the data collection system, so that, now, it is usual for the whole process of collection and analysis to be carried out on one portable instrument.
Chapter 6: Complex Geometries
(19 Pages)
Much large industrial plant contains pipes and nozzles welded to thick cylindrical components, often working under high pressure. The welds tend to be in regions of high stress and therefore require inspection for weld integrity during manufacture and possibly for crack initiation and propagation during service. As we have already seen, ultrasonics is the most useful nondestructive test, since it can give information on which fracture mechanics assessments of component integrity may be based.
Where two cylinders intersect, for instance, two cylindrical components of an offshore structure or a nozzle attached to a pressure vessel, the weld forms a three dimensional saddle shape. Probes with a given, fixed, beam angle placed on any one of the surfaces cannot always cover the entire weld volume which needs to be inspected. Thus, design of scanners for such geometries necessitates even more care than is taken with scanners for the simpler geometries of flat plates.
The Time-of-Flight Diffraction technique opens up new alternatives for inspection of complex geometries compared with pulse-echo techniques because of its insensitivity to the relative orientation of probes and defect. With pulse-echo techniques relying on specular reflection it often proves very difficult to arrange for probe beams to illuminate areas of concern, such as welds, at near normal incidence. Tandem techniques also prove difficult because the back wall of the specimen is very often not parallel to the inspection surface in nozzle to shell welds or offshore nodes, for example.
Chapter 7: Additional Complexities
(19 Pages)
In this chapter, we discuss some further aspects of the application of the Time-of-Flight Diffraction technique which require consideration in certain circumstances. We cover anisotropic materials, the effects of compressive stress on signals from cracks, and some subtle effects of component curvature.
Most metallic crystals show anisotropic elastic behaviour but, in fine-grained bulk samples with no preferred grain orientation, the macroscopic properties are isotropic. If, however, the grains approach in size the wavelength of the ultrasound, or are preferentially aligned, the resultant anisotropy and scattering affects ultrasonic inspection. This problem is particularly relevant to austenitic steels, both in bulk and in the form of cladding layers on ferritic steel. Section 7.1 discusses the problem of applying the Time-of-Flight Diffraction technique to such material.
Another problem arises from the differences in temperature and stress levels during service and those occurring when inspection is carried out. Examples of this are: aircraft, where in flight at 33,000ft the temperature is -25°C and the pressure is ~ 0.3 bar, which contrasts with typical inspection conditions of a temperature of 20°C and a pressure of ~ 1 bar; nuclear reactor coolant circuits, where inspection is almost always carried out at temperatures and pressures well below their normal operating point; offshore structures, when inspection is carried out in calm weather when the wind and sea loadings are very different from those during severe weather. As a result of these changes in ambient conditions between normal service operation and those during inspection, cracks which were under tensile stress sufficient to cause growth during some conditions of service could be under compressive stress when inspected. The effect of compressive stress on the amplitude of Time-of-Flight Diffraction signals is discussed in Section 7.2.
Finally, in cylindrical geometries, the speed of the lateral wave, which is used as a timing reference, is found to vary from its value on a flat plate. It is necessary to know what this variation is, if Time-of-Flight Diffraction is to be applied confidently to curved geometries and this problem is discussed in Section 7.3.
Chapter 8: Experimental Demonstrations of Capability
(33 Pages)
Non-destructive testing techniques, including ultrasonic inspection, are often used in safety critical applications such as the nuclear industry, submarines, offshore oil platforms, chemical plants, aircraft and gas pipelines. A characteristic of such applications is that there is often a reluctance to adopt new techniques until there has been a build up of experimental evidence of capability. The evidence which leads to successful adoption tends to start with idealised trials in the laboratory followed by an accumulation of satisfactory field experience.
Both the laboratory tests and the field experience can be embodied in a standard procedure for applying the inspection technique once it has reached a certain maturity. The Time-of-Flight Diffraction technique is now at this stage, having been encapsulated in both British and European Standards. This follows several demonstrations of capability in a wide range of large-scale test-block exercises which are reviewed in this chapter. Results from several significant test-block trials are presented in some detail to highlight the capability of Time-of-Flight Diffraction for accurate determination of the through-wall extent of cracks. The test-block exercises test technique capability, rather than reliability or repeatability in practice.
In this chapter, we confine our discussion to the principal test block exercises. A number of other, generally smaller and more specialised, exercises are covered in Chapter 9.
Capability is not sufficient in itself., reliability in practice is also required. The level of reliability required of an inspection is that which, when combined with a knowledge of the severity of defects, will lead to the desired level of structural integrity under normal operation or possible accident loading conditions. Even good techniques applied reliably will exhibit some spread of errors which have implications for the structural integrity of the component under test. This concept is reviewed with special reference to the pressure vessel of a pressurised water reactor; the approach is, however, universally applicable.
Before embarking on a discussion of the results obtained in test-block trials, we bring out some of the limitations of such tests.
Chapter 9: Applications of Time-of-Flight Diffraction
(9 Pages)
As we have demonstrated in earlier chapters, the Time-of-Flight Diffraction technique is a powerful and accurate tool; it is now routinely used to size defects in a large range of components both as a complement to detection by mote conventional methods and in stand-alone applications where both detection and sizing are carried out by Time-of-Flight Diffraction.
It is in the nature of routine applications that they rarely give rise to published papers. Much of the work cited in this chapter on applications could more accurately be described as studies of capability, undertaken before embarking on routine deployment of the technique. Some of the applications have been referred to in earlier chapters but are presented here again to make this survey as comprehensive as possible. The content of the cited papers is described only briefly and the reader is referred to the original references for greater detail.
Wedgwood [19951 reviews the advantages and disadvantages of TOFD vis-a-vis other ultrasonic inspection methods, citing better reliability as one of the major reasons for choosing TOFD. He describes some applications of TOFD to offshore and nuclear plant but does not attempt a comprehensive list of all applications to date. In the following sections we present applications grouped into specific areas.
Chapter 10: Application of Codes and
Standards to TOFD Inspection
(7 Pages)
The aim of any ultrasonic inspection must be to determine whether the inspected component is fit for its purpose. A component that is fit for its purpose will be one which has a very high probability of continuing to perform its function for its designed life, i.e. a very low probability of failure in service. However, the precise probability required will depend on the consequences of failure. Clearly, if the failure of a component could lead to many deaths, it must be many orders of magnitude less likely to fail than one which would merely cause a short maintenance outage.
The ultrasonic inspection is designed to detect, size and possibly characterise defects in the component. The next stage of assessment is to classify those defects as acceptable or unacceptable. Whether a defect is acceptable or not depends on the component in which it is found and the stresses to which it will be subjected. Defects which are sufficient to cause the component to fail under applied loads which might occur in practice will be classed as unacceptable. In order to ensure that the assessment process is properly carried out, procedures are laid down in codes and standards.
Appendix
Bibliography
Index
Distribution
NORTH AMERICA
Taylor & Francis Inc.
International Thompson Publishing, Discovery Distribution Center, Receiving Dept.,
2360 Progress Drive Hebron, Ky. 41048
ASIA-PACIFIC
Hemisphere Publication Services
Golden Wheel Building At 04-03, 41 Kallang Pudding Road, Singapore 349316
UK& EUROPE
ATP Lid,
27/29 Knowl Piece, Wilbury Way, Hitchin, Hertfordshire, England, SG4 OSX
On-line Bookstore: www.ameritech.co.uk
List Price: 55.00 BP
Fig1: