WO2016016663A2

WO2016016663A2 - System for non-destructive detection of internal defects

Info

Publication number: WO2016016663A2
Application number: PCT/GB2015/052231
Authority: WO
Inventors: Geoff DIAMOND
Original assignee: V-Viz Ltd
Priority date: 2014-07-31
Filing date: 2015-07-31
Publication date: 2016-02-04
Also published as: GB201413566D0; WO2016016663A3

Abstract

There is provided a device for non-destructive detection of internal defects comprising two or more imaging systems selected from the group comprising; a millimetre-wave imaging system comprising a millimetre-wave radiation detector configured to detect electromagnetic radiation with a frequency in the millimetre range; a near-infra-red imaging system comprising a near-infra-red laser source and a near-infra-red radiation detector; and a capacitance imaging system comprising two coplanar electrodes separated by an insulating gap, an alternating voltage source and a means for measuring voltage output by one or more electrodes. The device further comprises a processor configured to analyse data from each of the imaging systems, so as to determine the presence of damage under the surface and further configured to combine image data from each of the two or more imaging systems.

Description

System for non-destructive detection of internal defects

Technical field of invention

The present invention relates to the detection of internal defects using non-destructive techniques. In particular, the invention relates to an integrated system for the detection of corrosion, cracks and other damage or defects present on or within surfaces concealed by paint or other coatings.

Background to the invention

Corrosion on civilian and industrial infrastructure is a universal challenge that has an average cost to societies globally of 3-4% of GDP. For example, damage present on bridges and other structures is often concealed under a coating and difficult and costly to reveal, such as corrosion under paint (CUP). Large structures as ship hulls, ballast tanks, offshore platforms, buoys, refineries, pipelines, chemical storage plants and bridges are all constructions and objects where damage may be concealed under a surface coating. Early detection is costly and time consuming and in many circumstances not possible until such time that the problem has become severe enough to be visible to the naked eye. At this time most of the structural damage is already done and structures may have become potentially dangerous and maintenance much more costly than if done at an early stage.

Detection of CUP and other concealed defects in such structures as those mentioned above must be carried out using non-destructive testing (NDT) techniques, in order not to damage the structure. Current NDT techniques used for the detection of CUP and other internal defects include X-ray radiography, thermography, microwave inspection, ultrasonic inspection, techniques using eddy currents, magnetic particle inspection (MPI) and penetrants. These techniques require inspection in a highly localised area around an existing defect or region of CUP in order to detect the problem. Accordingly measurements taken by known techniques typically cover only a small area and hence inspection speeds are very slow. Moreover, they are usually highly labour intensive, require extensive preparation and often involve shutting down operation. For example, for large structures it is often necessary to erect scaffolding in order to allow a user to get inspection equipment close enough to make a usable measurement. As a result, current NDT techniques for the detection of CUP and other defects are highly expensive. MPI does not function well when detecting corrosion, cracks or other defects under paint over 75 microns thick.

The known DT technologies referenced above also suffer from a number of disadvantages. X-ray radiography requires bulky and expensive equipment, presents safety hazards to the users and other personnel, is highly directional and sensitive to flaw orientation. Another disadvantage with radiographic techniques is that they typically require the placement of a radiation source on the opposite side of a surface to the detection equipment. Flash thermography requires the use of large halogen sources that are both bulky and heavy, which often need to be positioned in hard to access locations - the halogen sources must be moved between many locations in order to fully inspect a structure. Microwave inspection is highly susceptible to errors caused by variations in paint thickness and sensor drift when looking for CUP. Ultrasonic inspection requires intimate contact and impedance-matching coupling between an ultrasonic device and the inspected surface which is a very time-consuming and expensive to achieve. Typically such techniques have a coverage rate of less than 0.5 square metre an hour. Eddy current techniques are time consuming and in general have poor sensitivity and spatial resolution.

Due to the growth in the number of steel and metal structures which are susceptible to corrosion damage in critical sectors such as the offshore oil and gas industry, bridges, refineries and chemical plants, there is an urgent need to develop systems and techniques that enable industries to adopt more cost efficient and reliable inspection & monitoring methods.

Summary of invention

In order to mitigate at least some of the problems above, the present invention provides nondestructive testing techniques to identify regions of an object corresponding to the presence of damage on the object, in particular to the presence of corrosion, cracks or other defects concealed by a coating such as paint.

In a preferred embodiment, there is provided a system comprising two or more of a millimetre-wave and sub-millimetre-wave (MMW/sub-MMW) imaging device, a near-infrared (NIR) imaging device and a capacitance imaging device, wherein the system is configured to detect damage to an object. Each of these component devices output complementary data that is combined to provide a more detailed and comprehensive analysis of the object.

Each component device is discussed below in turn.

Millimetre-wave and sub-millimetre-wave (MMW/sub-MMW) imaging device

In the preferred embodiment of the MMW/sub-MMW imaging device, there is included a broadband MMW/sub-MMW radiation detection unit, which allows a wide range of wavelengths of MMW/sub-MMW radiation to be detected, encompassing bands in which radiation is transmitted for many different coatings. Thus the present invention makes detection of corrosion, cracks or other defects or damage concealed under a wide range of paint and other coatings more reliable.

The MMW/sub-MMW imaging device also comprises focussing optics and a processor. The focussing optics are arranged relative to the detection unit such that the reflected MMW/sub- MMW radiation forms an image of the object on the detection unit. The detection unit measures the intensity of MMW/sub-MMW radiation at each part of the image and outputs this data to the processor.

The intensity of a particular part of the image depends on the properties of a corresponding part of the object being analysed. For example a portion of the object on which surface corrosion is present, or which corresponds to a crack or other damage or defect of the object has a different reflectivity to MMW/sub-MMW radiation, hence the intensity of detected radiation corresponding to an area of damage will be different to that corresponding to an undamaged part of the object.

The processor of the MMW/sub-MMW imaging device is configured to perform analysis to identify areas of the image corresponding to damage present on the object.

In a preferred embodiment, the processor is further configured to generate a visual image corresponding to the MMW/sub-MMW image detected by the detection unit. In some embodiments, the processor is further configured to provide a visual or audio indication that damage has been detected, a visual indication identifying a specific region where damage has been detected, or a combination thereof. Preferably, the detection unit comprises a plurality of detectors arranged in an array. Such an array has the advantage that the image acquisition time is greatly decreased.

In some embodiments the focussing optics comprise a polymer lens. Preferably the polymer is a high density polyethylene. Advantageously this permits the fabrication of the focussing optics cheaply using rapid prototyping techniques, thereby minimising production costs.

Preferably, the MMW/sub-MMW imaging device also includes a MMW/sub-MMW radiation source. The source is used to illuminate the object to be analysed, and has particular advantage in situations where it is not possible to utilise background "cold sky" illumination, for example in areas in shadow such as the underside of bridges, or indoors. Preferably, the device also includes a beam splitter that allows the device to both illuminate the object and detect reflected radiation along substantially the same axis. This is of particular advantage in situations where diffuse on axis light (DOAL) is required, such as in machine vision applications.

In the preferred embodiment comprising a MMW/sub-MMW radiation source, preferably the device further comprises a modulator, configured to modulate radiation emitted by the source such that MMW/sub-MMW radiation incident on the object to be analysed has a low degree of, or no coherence. Advantageously, such de-coherence has the dual advantages of minimising specular reflections which could obscure parts of the detected image, as well as reducing undesirable standing wave effects.

In some embodiments the device further comprises communication circuitry, configured to allow the processor to communicate with an external data processing system over wired protocols, wireless protocols, or both. In some embodiments the communication circuitry is used to transmit data to a remote device for further analysis, or to inform a remote user of the detection of an area of damage. In some embodiments the communication circuitry is configured to allow the device to be operated remotely, advantageous in potentially hazardous situations.

Near-infra-red (NIR) imaging device The NIR imaging device comprises a NIR radiation source, an aperture, a detection unit and a processor. Radiation from the source exits the device through the aperture, and is used to illuminate an object of interest. A portion of NIR radiation reflected by the object enters the device through the aperture. The aperture is positioned such that the reflected radiation that enters the device is incident on the detection unit. The detection unit measures the intensity of reflected NIR radiation and outputs this data to the processor.

Preferably the radiation source comprises a first NIR laser that emits NIR radiation of a first wavelength, the beam size of the laser being such that only a small area of the object is illuminated at any one time. Advantageously, the illumination of such a small area of the object improves the level of detail the imaging device can resolve. In another embodiment, a high intensity NIR light emitting diode (LED) is used in place of a first NIR laser.

In a preferred embodiment, the device also comprises means to change the position of the small area of illumination over time. Advantageously, this allows a large area of the object to be scanned by the device whilst only ever illuminating a small area of the object. Preferably neighbouring areas of the object are scanned in sequence, following a raster pattern. In some embodiments, the means comprises a deflection unit that changes the angle at which the beam of laser radiation exits the aperture, for example one or more mirrors connected to galvanometers.

In another embodiment, the radiation source preferably comprises a plurality of NIR lasers and/or NIR LEDs emitting radiation of a first wavelength arranged in an array, each emitting radiation that illuminates a different part of the object, and configured such that only a single source (either laser or LED) emits radiation at any one time, different LEDs emitting radiation at different points in time. In an embodiment including an array of LEDs emitting radiation of a first wavelength, the detection unit preferably comprises an array of NIR detectors, configured such that only a single detector is active at a point in time, the time at which a particular detector is active being the time at which a particular corresponding LED is emitting radiation. This embodiment has the advantage that the components are arranged to make a small, hand-held device suitable to operating at close distances to the object of interest. In alternative embodiments comprising an array of radiation emitting elements, the radiation source comprises an array of NIR lasers. In a preferred embodiment of the present invention, the detection unit outputs data describing the measured intensity of detected reflected NIR radiation at a given point of time to the processor. The processor is preferably configured to associate the intensity of detected radiation at a point of time with the position of the illuminated region at that time, thereby generating an image of the reflected radiation corresponding to the object.

The processor is configured to perform analysis to identify areas of the image corresponding to damage present on the object. In some embodiments the analysis involves comparing the measured intensity of a first part of the image to one or more other parts, wherein a difference in intensity higher than a predetermined threshold is indicative of the presence of damage.

In the preferred embodiment, the device also comprises a beam splitter, positioned such that a portion of the radiation emitted from the radiation source passes through the beam splitter before exiting the device through the aperture, and a portion of the reflected radiation travelling along a similar optical axis is diverted onto the detection unit after entering the device through the aperture. Advantageously, by detecting reflected radiation along an axis close to, or the same as, that of the radiation incident on the object, parallax errors are reduced, whilst maximising contrast and collection efficiency of reflected NIR radiation.

In some embodiments, the radiation source also includes a second NIR laser having a second wavelength. The processor is configured to construct first and second images of the object corresponding to the intensities of reflected radiation of the first and second wavelengths. The processor is further configured to perform dual-energy absorptiometry analysis, comprising applying a scaling factor to each image dependent on the wavelengths used, and subtract the images to create a third image. Advantageously, this further compensates for the scattering of radiation by the surface coating, providing a more detailed image of corrosion and other damage or defects.

In some embodiments the device includes a narrow band, thin film optical interference filter, positioned so as to filter radiation entering the device before it reaches the detector, thereby reducing noise due to fluorescence effects caused by incident radiation interacting with the object's surface coating. In some embodiments the device further comprises communication circuitry, configured to allow the processor to communicate with an external data processing system over wired protocols, wireless protocols, or both. In some embodiments the communication circuitry is used to transmit data to a remote device for further analysis, or to inform a remote user of the detection of an area of damage. In some embodiments the communication circuitry is configured to allow the device to be operated remotely, advantageous in potentially hazardous situations.

NIR illumination of longer wavelengths (for example 9 to 11 microns) typically also heats an object being analysed. The heated object then emits NIR radiation at a range of wavelengths. In another embodiment the radiation source includes a long wavelength NIR laser, and the device is employed in a different, thermographic mode of operation, wherein the device is configured to detect emitted radiation from the object after heating by long wavelength NIR laser. Advantageously, this mode of operation allows analysis to be conducted at a distance away from the object using a single device, and also analyses a large area of the object without a user having to move the device. Preferably in this embodiment, the device is configured to illuminate a region of the object for a certain period of time, and to only commence detection of radiation after illumination has ceased, therefore only allowing detection of emitted, and not reflected radiation. Preferably for thermographic mode, data analysis is performed by the processor in the manner described above with respect to the detection of reflected radiation.

In a preferred embodiment, the NIR imaging device is capable of alternating between detecting reflected radiation and emitted radiation (thermographic mode), thus allowing a user to gain more detailed data regarding possible corrosion and other damage and defects for a given object. In thermographic mode the device first illuminates the object, then ceases illumination before detecting emitted radiation in order to prevent reflected radiation from the source also being detected.

Capacitance imaging device

In the preferred embodiment there is provided a capacitance imaging device comprising a plurality of coplanar electrodes, a voltage source configured to apply an alternating voltage to the electrodes, a voltage detector configured to measure an induced voltage in each of the electrodes and a processor. The sensor electrode on the PCB comprises one plate of a virtual capacitor and the surface under inspection is the other plate which is grounded with respect to the sensor input.

The device is positioned such that the electrodes are held parallel to, and at a fixed distance from the surface of the object of interest. In some embodiments wheels or rollers are provided to maintain the fixed distance whilst allowing the device to be easily moved over the surface of the object. In a preferred embodiment, the electrodes are imprinted on the surface of a flexible substrate to allow the electrodes to conform to curved surfaces of objects to be tested, for example the exterior of pipes, improving device sensitivity.

The alternating voltage applied to the electrodes generates an electric field with a distribution dependent on the local environment of the source electrode, which in turn induces a voltage in the electrodes. The capacitance imaging device is moved over the surface of the object, and the value of the induced voltage in the electrode is measured by the voltage detector and output to the processor.

The processor then determines whether changes in the measured induced voltage with position are indicative of the presence of damage or defects under the surface coating, through comparison against predetermined values for different object-surface coating interface conditions (for example, equivalent values for a pristine interface, delaminated surface coating, or a crack in the object).

In a preferred embodiment the processor is further configured to record the position of the device on the object of interest over time, and associate each measured induced voltages with the position that it was measured at, thereby creating an image of the object corresponding to the measured induced voltages. Preferably the processor is further configured to generate a visual image based on the induced voltage image that can be output to a display and viewed by a user.

Combined non-destructive imaging apparatus

Preferably there is provided non-destructive imaging apparatus comprising one or more of the imaging devices above, configured to synergistically analyse the data output by each device to give a user a detailed view of corrosion, cracks and other damage or defects present on the device that are concealed from visual inspection from a surface coating such as paint.

When combining MMW/sub-MMW and NIR imaging devices, there is preferably provided a hybrid focussing optics to allow the simultaneous focussing of MMW/sub-MMW and NIR radiation onto respective MMW/sub-MMW and NIR detectors, thereby minimising both space taken up by the components of the apparatus, and the total number of components to save cost.

Other aspects of the invention will become apparent from the appended claim set. Brief description of drawings

Figure 1 is a schematic of a passive millimetre-wave and sub-millimetre-wave imaging device.

Figure 2 is a schematic of an active millimetre-wave and sub-millimetre-wave imaging device.

Figure 3 is a schematic of a near-infra-red (NIR) imaging device.

Figure 4A is a perspective view of a hand-held NIR imaging device.

Figure 4B is a cross sectional view of a hand-held NIR imaging device.

Figure 5 is a schematic of a capacitance imaging device.

Figure 6 shows an arrangement of electrodes for a capacitance imaging device.

Figure 7 is a schematic of a combined near infrared and millimetre-wave/sub-millimetre- wave imaging device

Figure 8A is a cross sectional view of hybrid focussing optics for use in a combined millimetre-wave and sub-millimetre-wave imaging device and a near-infra-red imaging device.

Figure 8B is a top view of hybrid focussing optics for use in a combined millimetre-wave and sub-millimetre-wave imaging device and a near-infra-red imaging device. Figure 9A shows a photograph of a test sample.

Figures 9B and 9C show test results obtained by an active millimetre-wave and sub- millimetre-wave imaging device and a capacitance imaging device respectively.

Figures lOA-lOC show exemplary graphical user interfaces for use with imaging devices.

Description of an embodiment of the invention

In the preferred embodiment, there is provided a system comprising two or more of a millimetre-wave and sub-millimetre-wave (MMW/sub-MMW) imaging device, a near-infrared (NIR) imaging device and a capacitance imaging device, wherein the system is configured to detect damage to an object. Each of these component devices output complementary data that is combined to provide a more detailed and comprehensive analysis of the object.

Each component device is discussed below in turn.

Millimetre-wave and sub-millimetre-wave (MMW/sub-MMW) imaging device

Coatings such as paints and primers typically have transmission characteristics that allow certain wavelength bands of MMW/sub-MMW radiation to pass through them substantially unaffected, hence MMW/sub-MMW imaging techniques are useful for detecting corrosion and other damage to objects of interest (for example part of a pipeline, bridge, ship's hull, etc.) concealed to visual inspection by a surface coating. Transmission characteristics typically differ between different paints and other coatings. Known MMW/sub-MMW imaging devices typically use narrow band detection close to a single chosen wavelength, which might not be transmitted by a particular coating. We note that whilst the following discussion relates specifically to detection of damage/corrosion/defects on objects concealed by coatings, the MMW/sub-MMW devices and techniques may also be applied in other contexts, including industrial inspection, medical imaging, and other fields.

Figure 1 shows a schematic of a passive millimetre-wave and sub-millimetre-wave imaging device, which performs non-destructive testing for damage or defects (such as corrosion and cracks) present on an object covered by a surface coating. The device 100 includes a broadband MMW/sub-MMW detection unit 104, advantageously allowing a wide range of wavelengths of MMW/sub-MMW radiation to be detected, encompassing bands in which radiation is transmitted for many different surface coatings. Preferably the detection unit 104 comprises a plurality of detectors 105-107 to allow detection of a greater amount of MMW/sub-MMW radiation per unit time, and reduce the time taken for image capture. In further embodiments the invention is achieved using a single detector. The device further comprises focussing optics 108 and a processor 110.

In a preferred embodiment, each detector 105-107 comprises a glow discharge detector (GDD). Each GDD comprises a sealed glass envelope filled with a gas, a cathode and an anode disposed such that both the cathode and anode have a part situated within the glass envelope. Each detector further comprises a visible light detection element, preferably a CCD sensor configured to detect light emitted within the glass envelope.

The device 100 is configured to apply a voltage across the cathode and anode of sufficient magnitude to ionise the gas in the glass envelope, thereby causing a DC bias current to flow through the gas and causing visible radiation to be emitted by the gas, at a localised portion of the GDD. In operation, MMW/sub-MMW radiation is focussed onto one or more of the GDDs, and interacts with the ionised gas, increasing the excitation, ionisation and collision rates of gas molecules, and thus increases the amount of visible light emitted by the gas. The visible light detection element, is configured to detect the changes in visible light emitted by the GDD and thereby detect MMW/sub-MMW radiation. Since the visible light is emitted from a localised portion of the GDD corresponding to the location at which MMW/sub- MMW radiation was incident on the gas, the emitted light corresponds to an image of the object 102. Preferably the gas in the glass envelope is neon. In further embodiments other gases that exhibits scintillation, whereby visible radiation is emitted when MMW/sub-MMW radiation is incident upon the gas, are used.

Preferably the discharge bias current applied across the cathode and anode of the GDD may be varied, thereby varying the frequency of electron collisions in the ionised gas.

Beneficially, a detector of this form allows inexpensive visible light detection electronics to be used. Such a detector has the additional advantage that the image of reflected radiation is output in a visual format without the need for further processing. Advantageously, varying the discharge bias current across the cathode and anode allows the frequency of electron collisions to be matched to the frequency of the MMW/sub-MMW being detected. Under these conditions the detection efficiency is enhanced for that particular frequency of MMW/sub-MMW radiation. Beneficially, by allowing the applied discharge bias current to be varied, the device is highly tuneable, and detection can be optimised for a wide range of MMW/sub-MMW radiation.

Although preferably glow discharge detectors, in further embodiments the detectors 105-107 are other detectors capable of detecting a broad band of MMW/sub-MMW wavelengths, including pyroelectric detectors, Schottky diode detectors, and Golay cells.

In operation, an object to be analysed 102 is illuminated by natural background "cold sky" MMW/sub-MMW radiation, though in further embodiments the object 102 can also be provided with additional illumination from external MMW/sub-MMW radiation sources. MMW/sub-MMW radiation is reflected from the object 102, with the amount of radiation reflected in a given direction from any given point on the object being dependent on whether damage or other defects, including corrosion and cracks are present at that point.

The focussing optics 108 are disposed relative to the detection unit 104 such that radiation reflected by the object 1 12 is focussed by the focussing optics 108 so as to form an image of the object 102 on the detection unit 104. The detection unit 104 measures the intensity of the reflected MMW/sub-MMW radiation 112 for each part of the image. The detection unit 104 outputs measured intensity data to the processor 110 via a communication connection (not numbered). The processor 110 then analyses the detected MMW/sub-MMW image of object 102.

In a preferred embodiment, the processor 110 is configured to perform analysis on the image to identify areas of the image corresponding to damage present on the object. Preferably the processor 110 first performs image enhancement including contrast enhancement and de- blurring using known image processing techniques. The processor 110 compares the measured intensity of the reflected MMW/sub-MMW radiation 112 in different parts of the image. In an undamaged material/sample the image generated by the MMW/sun-MMW radiation would be approximately constant as the material being inspected will be substantially uniform. By comparing two or more regions of a sample the presence of damage can be detected by the comparison. If the difference between a first region and a second region exceeds a certain predetermined threshold, the processor indicates that corrosion, a crack, or other damage or defect is present. Such a threshold in an embodiment is empirically determined for the material being tested. Preferably the processor 110 is further configured to use known techniques to plot the boundary between regions in which damage is present and regions where no damage is present based on the relative intensities of detected radiation, and segment the image according to the boundaries. The processor is configured to provide a user with an indication of the regions in which damage has been detected in the form of a visual output, an audible output, or more preferably both.

Preferably the device 100 also comprises communication circuitry (not shown) configured to communicate with an external data processing system using either wired protocols, wireless protocols or both. This advantageously allows the device 100 to be operated remotely. The device 100 is also configured to generate a visual image and/or audio signal indicating the presence of damage which is then transmitted to an external display via communication circuitry.

Preferably the focussing optics 108 comprise a polymer lens fabricated from high density polyethylene, advantageously allowing the use of rapid prototyping techniques to minimise fabrication expense.

Thus the invention allows for fast and effective detection and identification of corrosion, cracks and other damage or defects present in an object concealed from visual inspection by a surface coating, utilising inexpensive components and low power apparatus.

Figure 2 shows a schematic of an active millimetre-wave and sub-millimetre-wave imaging device 200. Such an active device provides a source of MMW/sub-MMW illumination, and may be used in place of the passive device 100 of figure 1 in situations where there is insufficient ambient background MMW/sub-MMW radiation for the detection of damage and defects.

The active device 200 has a broad band MMW/sub-MMW detection unit 204 comprising a plurality of detectors 205-207, focussing optics 208, a processor 210, a MMW/sub-MMW radiation source 214 and an optional beam splitter 216. The active device 200 operates in much the same manner as the passive device 100 (described above), with additional features for the incorporation of a MMW/sub-MMW radiation source 214. In particular, it is noted that preferably the detection unit 204 is the same as the detection unit 104 and that the processor 210 performs the same analysis as the processor 110 as described for the preferred embodiment of the passive MMW/sub-MMW imaging device 100 above. The provision of a source of MMW/sub-MMW illumination has particular advantage in situations where it is not possible to utilise background "cold sky" illumination, for example in areas in shadow such as the underside of bridges, inside ship hulls, indoors etc.

The MMW/sub-MMW radiation source 214, beam splitter 216, detection unit 204 and focussing optics are arranged relative to each other so as to allow the following operation. In operation the source 214 emits incident MMW/sub-MMW radiation 218, a portion of which passes through the optional beam splitter 216 and the focussing optics 208 before being incident on the object to be analysed 202. A portion of the MMW/sub-MMW radiation reflected by the object 212 entering the device is directed onto the detection unit 204 by the beam splitter 216. The focussing optics 208 are positioned such that the reflected MMW/sub- MMW radiation 212 is focussed to form an image of the object 202 on the detection unit 204.

Advantageously, this arrangement allows the device to both illuminate the object and detect reflected radiation along substantially the same axis. This is of particular advantage in situations where diffuse on axis light (DOAL) is required, such as in machine vision applications.

If provided, the beam splitter 216 preferably comprises a layer of electrically conductive film disposed on a polymer substrate. Preferably the beam splitter 216 comprises a sheet of Mylar with a thickness of 10 microns, coated in a layer of sputtered aluminium. Such an arrangement makes use of the "skin-effect", and has the advantage that it is far less expensive than commonly known beam splitters fabricated from silicon wafers. Additionally, such a beam splitter does not suffer from drawbacks associated with traditional beam splitters for example Fabry-Perot interference devices.

Alternatively, the optional beam splitter 216 is not provided. In this case, the radiation source 214 and the detection unit 204 are preferably located adjacent to one another in close proximity, such that emitted MMW/sub-MMW radiation 218 incident on the object to be analysed 202 and the reflected radiation 212 incident on the detection unit 204 travel along substantially the same axis. This particular arrangement has the advantage that the device 200 is able to both illuminate the object and detect reflected radiation along substantially the same axis (again, being useful in DOAL applications). The following discussion in relation to figure 2 is equally applicable to devices 200 that have an optional beam splitter 216 and devices 200 that do not.

Preferably the radiation source 214 comprises a coherent MMW/sub-MMW source, for example a Gunn diode. Active coherent MMW/sub-MMW sources such as Gunn diodes are typically cheaper than active incoherent MMW/sub-MMW sources, which require complicated electronics such as diode multiplier chains to introduce resonant harmonics to MMW/sub-MMW radiation to provide incoherence. In addition, coherent MMW/sub-MMW sources such as Gunn diodes typically provide much more intense illumination than active incoherent sources. Using a coherent as opposed to incoherent source therefore results in a much stronger signal 212 that can be detected by the detection unit 204. Beneficially use of an active coherent source results in a higher signal to noise ratio, and lower device cost than could be achieved using an active incoherent source.

It is also noted that active incoherent sources require expensive receiver technology to detect the radiation, and typically have a narrow bandwidth of output radiation, whereas coherent sources such as Gunn diodes generate radiation across a broad range of frequencies which may be detected using relatively inexpensive detectors. Advantageously, providing a broad range of incident frequencies is beneficial when looking for corrosion or other defects under coatings such as paint, since different types of corrosion, materials or defects may be easier to detect (i.e. may have a higher/lower reflectivity relative to its surroundings) at different frequencies of MMW/sub-MMW radiation. Active coherent sources may also achieve high frequencies - this allows for reduced diffraction effects for smaller illumination beam spot sizes (i.e. smaller illumination beam cross sections), and thus allow for finer detail to be resolved by the device 200.

In addition, it is noted that an active coherent source may provide a stronger signal and broader bandwidth of incident radiation than passive "cold sky" incoherent illumination, which has relatively weak intensity and is subject to filtering at certain wavelengths by the atmosphere.

Preferably the radiation source 200 also comprises a modulator (not shown in figure 2), configured to modulate radiation emitted by the source such that MMW/sub-MMW radiation incident on the object to be analysed has a low degree of coherence, or more preferably no coherence. Advantageously, such de-coherence has the dual advantages of minimising undesirable specular reflections which could obscure parts of the detected image (otherwise known as "hot spots" or "glare"), as well as reducing undesirable standing wave effects (which introduce noise to measurements of detected MMW/sub-MMW radiation).

Traditional methods of reducing specular reflection include using a diffuse source, positioning a source at a distance from an object that is greater than its coherence length, or using near field imaging techniques. However, such techniques (for example broad beam, near field illumination methods) using with coherent sources can still give rise to image artefacts due to standing wave effects between an inspected object and detector. This is because of the coherence/in phase nature of the source - the strength of a reflected signal from the surface depends not only on the surface itself but also on the actual physical distance between the detector and inspected surface. This is because standing waves occur and if the surface is placed in the minimum-intensity part of the wave (the "trough") the signal will tend to be weak - conversely, if the detector happens to be aligned with the maximum intensity part of the wave (the "peak"), the signal will be strong. For example, for a full wavelength of 3.9 mm, there are 2 minima and 2 maxima, with a half-wavelength distance of 1.95mm.

Advantageously to mitigate such problems the modulator is configured to continuously vary the phase of the incident radiation, the reflected radiation, or both, between 0 and 360 degrees. The reflected radiation information detected by the detection unit 204 is averaged over a predetermined time period. Preferably the time period is the time taken to vary the phase of the radiation between 0 and 360 degrees. Advantageously, this results in the radiation incident on the detection unit 204 having both low coherence resulting in low specularity hot spots, and additionally avoids standing wave effects by averaging over different phases.

Preferably the modulator comprises means for mechanically moving either the MMW/sub- MMW source 214 along the direction of the emitted radiation, and/or moving the detection unit 204 along the direction of the received radiation. Said means for mechanically moving the source 214 and/or the detection unit 204 preferably comprise one or more reciprocating mechanical worm drives affixed to the source 214 and/or detection unit, and configured to displace the source 214 and/or detection unit 204 such that the phase of the radiation is varied between 0 to 360. Alternatively, other mechanical displacement means known to the skilled person may be utilised. In operation, the modulator is configured to change the distance between the source 214 and the object to be analysed 202 by an amount equal to one wavelength of the radiation, over a time period equal to that of the time period of the radiation. Alternatively (or in addition), the modulator is configured to change the distance between the detection unit 204 and the object to be analysed 202 by an amount equal to one wavelength of the radiation, over a time period equal to that of the time period of the radiation. Such mechanical modulation results in the phase of the incident beam the radiation incident on the detection unit 204 having low or no coherence, and minimises the effect of specularity and standing wave noise. Advantageously, this mechanical method involves minimal loss in signal intensity, and therefore avoids the need for additional amplifier elements to boost the signal received by the detection unit 204.

In an alternative embodiment, the modulator comprises a cyclical mechanically-controlled analogue phase shift of the source via a mechanically lengthened transmission line (often called a "trombone line"). As with the above embodiment, this results in low coherence and helps mitigate issues with specularity and standing wave noise. Again, this mechanical method involves minimal loss in signal intensity, and therefore avoids the need for additional amplifier elements to boost the signal received by the detection unit 204.

In a further alternative embodiment, the modulator comprises a box, having at least one dimension longer than the wavelength of the radiation, a plurality of internal surfaces on which radiation from the source 214 is reflected, and one or more apertures through which radiation exits the modulator before being incident on the object 204. Radiation leaving the one or more apertures will have a different relative phase depending on which internal surfaces have reflected it, resulting in low coherence incident radiation on the object to be analysed 202. This results in a similar effect as the time-averaged phase modulation discussed above (for example using the mechanical means described above), in that the radiation incident on the object 202 have low coherence and avoids both specularity and standing wave issues.

In a further embodiment, the modulator comprises means to electronically introduce incoherence to radiation. Such means perform electronic phase- shifting of the source radiation via PIN diodes, varactor diodes and/or GaAs Field Effect Transistors. Advantageously electronic means for introducing phase modulation modulate phase at relatively high speed, thus reducing the amount of time taken to vary the phase over a full cycle of 0 to 360 degrees. This has the benefit that the time required to analyse an object 202 can be reduced.

It is noted that specularity issues are particularly pronounced for on-axis coherent MMW/sub- MMW radiation illumination and detection. Thus the provision of a modulator facilitates more effective examination of objects to be analysed in situations where on-axis illumination/measurement is required.

Accordingly by providing an active MMW/sub-MMW coherent source in combination with a modulator, the benefits of low specularity and standing wave noise and high illumination intensity are achieved, thus further enhancing signal to noise ratios.

Optionally a MMW/sub-MMW coherent active source, and a MMW/sub-MMW detector may be provided on a single chip.

Additionally the device 200 is also able to be operated in a passive mode, wherein no radiation is emitted by the radiation source 214, and "cold sky" ambient MMW/sub-MMW radiation is used as an illumination source instead, in the same manner as described above with respect to the passive device 100 of figure 1. Advantageously this reduces the power consumed by the device in locations where ambient MMW/sub-MMW radiation is available.

Optionally, the source 214 comprises additional optics (not shown) configured to optimise the illumination of the object 202.

The processor 210 is configured to analyse the detected MMW/sub-MMW radiation in order to determine the presence of defects, or the like, in the illuminated sample. The processing occurs as described above with respect to Figure 1.

Preferably the active device 200 also comprises communication circuitry (not shown) configured to communicate with an external data processing system using either wired protocols, wireless protocols or both. This advantageously allows the device 200 to be operated remotely. In some embodiments the device 200 also generates a visual image, or audio signal, indicating the presence of damage which is then transmitted to an external display via communication circuitry.

In operation, the active device 200 may also be moved relative to the object being analysed. This may be done manually, or via mechanical means. Preferably the device also includes a positional encoding unit (not shown). The positional encoding unit associates reflected MMW/sub-MMW radiation data with the position on the object to be analysed 202 from which the radiation was reflected. This permits a user to identify specific parts of the object when analysing data from the device. For example, when a region corresponding to corrosion or other damage or defects is detected, the location of the region can be stored, advantageously allowing a user to scan a large area without pausing, and review any detected regions of interest after the scanning is complete. The positional encoding unit may be any such unit as known in the art. Similarly, a positional encoding unit is also preferably provided for a passive device 100.

Therefore the device provides fast and effective detection and identification of corrosion, cracks and other damage or defects present in an object concealed from visual inspection by a surface coating, utilising inexpensive components. It allows detection of damage and defects in situations in which ambient MMW/sub-MMW radiation is unavailable in sufficient quantities for passive detection, and conserves power in situations where it is available.

As well as detecting corrosion under paint in the MMW and sub MMW the invention provides the means for detecting corrosion under paint at a plurality of different wavelengths. Beneficially the ability to detect at different wavelengths allows for effective detection for corrosion and other damage or defects under paint and other surface coatings having different transmission properties over a range of wavelengths of MMW/sub-MMW radiation using a single device. Furthermore, the invention allows for the same target area (simultaneously or sequentially) to be scanned at two or more different wavelengths. As some specific features, are known to be more visible in the infrared than MMW, and vice versa, then the use of the two different methodologies beneficially allows for the improved detection of features which would otherwise not be possible using a single wavelength.

In an example of the invention there is also provided computer software for analysing data generated by a passive or active device 100 200. Preferably such software is configured to automatically identify regions of interest on the object being analysed, for example by identifying data corresponding to corrosion, defects, or damage. Indications of the regions of interest may then be output on a display for viewing by a user, thus enabling the user to easily discern parts of the object where damage/defects or corrosion might need repairing and/or monitoring.

Near-infra-red (NIR) imaging device

Near infra-red is another part of the electromagnetic spectrum in which many paints and other surface coatings have high transmission. This makes an effective wavelength for detecting corrosion and other damage to a particular object concealed to visual inspection by a surface coating.

Figure 3 shows a schematic of a near infra-red imaging device 300 comprising a detection unit 304, an aperture 307, a NIR radiation unit 308 comprising a first laser source 310, a beam deflecting unit 314 in electrical communication with a controller 318, a beam splitter 316 and a processor 319.

The components are arranged such that in operation, the first laser source 310 emits an incident NIR laser radiation beam 320 having a first wavelength onto the beam deflection unit 314. At a first time, the controller 318 configures the deflection unit 314 to direct the incident beam 320 such that it passes through the beam splitter 316 and the aperture 307 onto a first area of an object to be analysed 302. Reflected NIR radiation 322 that has been reflected by the object 302 substantially along the axis defined by the incident radiation 320 enters the device through the aperture 307. The reflected radiation 322 is then diverted by the beam splitter 316 onto the detection unit 304. The intensity of the reflected radiation 322 is measured by the detection unit 304 and output to the processor 319. At a second time, the controller 318 configures the deflection unit 314 to direct the incident beam 320 onto a second area of an object to be analysed 302, and the intensity of radiation reflected by the second area is measured by the detection unit 304 and output to the processor 319.

The device is configured to illuminate further areas of the object 302 and record the intensity of the reflected radiation until a predefined region of the object 302 has been scanned. Preferably the controller 318 is configured to cause the beam deflection unit 314 to illuminate neighbouring areas of the object 302 in turn, the sequence of illuminated areas prescribing a raster pattern. Such scanning allows detection to be performed over a large area of the object, whilst minimising scattering effects that are detrimental to detection efficiency. Advantageously such scanning may be performed at a fast rate. Typically an area of 10 square metres may be scanned in an hour when the device is positioned between 5 and 10 metres from the object 302.

In a preferred embodiment, the aperture 307 is covered by a material transparent to NIR wavelengths, in order to prevent unwanted material, such as dust and debris, entering the device 300.

The detection unit 304 comprises a detector 306, which is a known commercially available photo-resistive detector. Such detectors are preferred as they are inexpensive. In further embodiments other types of detector suitable for detecting NIR radiation are used such as a thermal camera, or a short wave infra-red detector.

The processor 319 is in electrical communication with the controller 318 and the detection unit 304. The controller 319 is configured to output data to the processor indicating the deflection of the incident radiation 320 at a given point in time, and the detection unit 304 to output data indicating the measured intensity of the detected reflected radiation 322 at that point in time. The processor 319 is configured to infer the position of the area of the object 302 being illuminated at a given point in time from the data provided by the controller 318, and associate intensity data from the detection unit 304 at that point in time with the area being illuminated, thereby creating a reflected NIR radiation intensity image corresponding to a region of the object 302.

The intensity of a particular part of the image depends on the properties of a corresponding part of the object being analysed. For example a portion of the object on which surface corrosion is present, or which corresponds to a crack or other damage or defect of the object has a different reflectivity to NIR radiation, hence the intensity of detected radiation corresponding to an area of damage will be different to that corresponding to an undamaged part of the object.

In the preferred embodiment the area of the object 302 that is illuminated by the device 300 at any one time is small, between 0.5 and 2 mm. Advantageously by using a laser source, small areas can be illuminated from a distance of several metres from the object 302. Preferably the beam diameter of radiation emitted by the first laser source 310 is less than 2 mm. Advantageously, the illumination of such a small area of the object improves the level of detail the imaging device can resolve. NIR wavelengths are typically highly scattered by surface coating materials, for example by the constituent pigments, solvents, binders, etc. found in paint. This leads to highly diffuse reflection of NIR radiation - when large areas of an object covered in a surface coating are illuminated, reflected NIR radiation from each part of the area would be incident on a typical NIR imaging device, leading to a poor level of detail being resolved. Thus the device 300 reduces this problem.

In a preferred embodiment, the processor 319 is configured to perform analysis on the image to identify areas of the image corresponding to damage present on the object. Preferably the processor 319 first performs image enhancement including contrast enhancement and de- blurring using known image processing techniques. The processor 319 calculates the difference between the measured intensity of the reflected NIR radiation 322 at different parts of the image. The processor 319 then compares the difference to a predetermined value. If the magnitude of the calculated difference between a first region and a second region exceeds the predetermined value, the processor determines that corrosion, a crack, or other damage or defect is present in either the first or second region. The processor determines which region the damage is present in based on whether the calculated difference is positive or negative. Preferably the processor 319 is further configured to plot the boundary between regions. Such plotting occurs using known techniques. Preferably the processor is further configured to segment the image according to the boundaries between regions, thereby isolating areas of interest in the image. The processor is configured to provide a user with an indication of the regions in which damage has been detected in the form of a visual output, an audible output, or more preferably both.

In a further embodiment, the NIR radiation source 308 comprises both a first laser source 310 and a second laser source 312, configured to emit first and second wavelengths of NIR radiation respectively, the first and second wavelengths being different. In this embodiment the process above is carried out twice; the first time using radiation from the first laser source 310, and the second time using radiation from the second source 312. Accordingly the processor is configured to produce a first and second image of a region of the object 302 corresponding to detected intensity of reflected radiation of the first and second wavelengths respectively. The scattering effect due to a surface coating is typically dependent on wavelength. Therefore a given surface coating has a different attenuation factor for the first and second wavelengths. The processor is further configured to perform dual-energy absorptiometry analysis on the two images, by scaling the first image by an amount dependent on the attenuation factor of the surface coating at the first wavelength, scaling the second image by an amount dependent on the attenuation factor of the surface coating at the second wavelength, and subtracting the two images from each other using known techniques to create a third image. In a preferred embodiment the attenuation factors are determined in advance. The composition of paint used to coat the surface being detected may be known (for example, from information provided by the manufacturer) or determined. The attenuation factor for the particular composition is then determined empirically, or in further embodiments is modelled, and such factors are used. In further embodiments a database of attenuation factors for different surface coatings (e.g. different paint compositions, polymers etc.) is used and the most appropriate coating type selected and the corresponding attenuation factors used. Advantageously this analysis further reduces the effects of scattering, reducing noise and producing a clearer image of any damage or defects present on the object 302 under the surface coating.

The processor 319 is then configured to perform the analysis as described above on the third image. Advantageously, this image subtraction further reduces undesirable effects due to scattering. Though described using only two laser sources, in further embodiments this technique is implemented using three or more different wavelengths of NIR radiation.

Optionally, the NIR imaging device 300 also includes a narrow band thin film optical interference filter (not shown) disposed between the beam splitter 316 and the detection unit 304, which only allows wavelengths corresponding to the reflected radiation 322 to pass through to the detector. Such a filter advantageously reduces noise by reducing or eliminating wavelengths of NIR corresponding to fluorescence caused by the incident NIR radiation 320 interacting with substances present in the surface coating of the object 302.

In addition to the foregoing, the NIR imaging device 300 is also able to be operated in an alternative, thermographic mode. Typically, when an object to be analysed is illuminated with NIR radiation with a radiation between 9 and 11 microns, portions of the radiation will be reflected and absorbed by the object. Through absorbing a portion of the radiation, the temperature of the object will increase, and it will emit radiation at a range of different wavelengths. The intensity of the emitted radiation from any given point on the object is dependent on local properties of the object, and is different for pristine regions as compared with regions in which corrosion, cracks or other damage or defects are present. The radiation source comprises a further long wavelength laser (not shown), preferably a carbon dioxide laser. In thermographic mode long wavelength incident radiation from the radiation source 308 is deflected onto consecutive areas of the object 302 by the deflection unit 314 as above, thus heating the object 302. The detection unit 304 is not activated until after the radiation source 308 has been deactivated, thereby preventing the detector from detecting reflected radiation originating from the radiation source. The object 302 subsequently emits NIR radiation (not shown) which is directed onto the detection unit by the beam splitter 316. Analysis of the intensity of the detected emitted radiation by the processor 319 proceeds as described above in relation to reflected radiation.

Figure 4A shows a perspective view, and figure 4B a schematic of a hand-held NIR imaging device 400, comprising a plurality of NIR detection units 404-414 and a corresponding plurality of radiation sources 416-426, disposed on a support member 401 and arranged in detection unit-radiation source pairs. Each detection unit-radiation source pair is positioned such that incident radiation 428 emitted from a radiation source 416 is reflected by an object to be analysed 402, and reflected radiation 430 falls onto a corresponding detection unit 404 when the device 400 is held a certain distance from the object 402. In operation, neighbouring pairs are activated in sequence, with only a single pair is active at any one time. For example: at a first time, a radiation source 416 emits incident radiation and a detection unit 404 detects reflected radiation; at a second time a second radiation source 418 emits incident radiation and a second detection unit 406 detects reflected radiation; at a third time a third radiation source 420 emits incident radiation and a third detection unit 408 detects reflected radiation, and so on. Thus the device sequentially scans an area of an object in an analogous manner to device 300 above. As with device 300, the radiation provided by each radiation source 416-426 of the device 400 only illuminates a small area of the object 402, typically no greater than 2mm in diameter. Accordingly the device 400 minimises the effects of scattering of NIR radiation by the surface coating of the object 402. Advantageously, the format of NIR imaging device 400 has the advantage that it is small enough to be handheld. Preferably the radiation sources 416-426 each comprise at least one high intensity NIR LED configured to emit radiation at a first, known, wavelength. Advantageously, the use of such LEDs reduces the space required to house each radiation source and therefore reduce the overall size of the handheld device. Preferably each LED is comprises a lens configured to cause a small area of the object 402 to be illuminated. Beneficially, high intensity NIR LEDs provide narrow bandwidth radiation, typically 60nm at full width half maximum.

The hand-held device 400 further comprises a processor (not shown) in electrical communication with the detection units 404-414. The detection units 404-414 are configured to output data describing the measured intensity of the reflected radiation to the processor, which creates an image of the intensities of reflected radiation corresponding to a region of the object 402, and subsequently performs analysis as described above in relation to the NIR imaging device 300.

In a further embodiment the radiation sources 416-426 each further comprise a second NIR LED that emits a second wavelength of radiation. In another embodiment there is provided a second support member (not shown) in addition to the original support member 401, having a plurality of radiation sources and detection units arranged in arrays as described above in relation to figures 4A and 4B. In this embodiment the radiation sources 416-426 disposed on the original support member 401 emit NIR radiation having a first wavelength, and radiation sources disposed on the second support member emit radiation of a second wavelength. In these embodiments, the processor is configured to create two images for each of the wavelengths of radiation, each corresponding to the same region of the object 402. In these embodiments the processor is further configured to perform dual energy absorptiometry analysis, on which further analysis is performed as described above in relation to the device 300.

In further embodiments, the device 400 also comprises an accelerometer, which outputs spatial data to the processor. The processor is configured to use this data to track the position of the device over time. Beneficially, this position tracking can be used to map the tomography of the object being analysed. Additionally, when a region corresponding to corrosion or other damage or defects is detected, the location of the region can be stored, advantageously allowing a user to scan a large area without pausing, and review any detected regions of interest after the scanning is complete. In additional embodiments, data analysis software may be provided as discussed above in relation to the MMW/sub-MMW devices 100 200.

Capacitance imaging device

Figure 5 shows a schematic of a capacitance imaging device 500, comprising an alternating voltage source 502, a plurality of co-planar electrodes 504-514, a voltage detector 518 and a processor 520. Coplanar electrodes 504-514 are split into two groups, source electrodes 504- 508 electrically connected to voltage source 502, and sense electrodes 510-514, electrically connected to the voltage detector 518. The source electrodes 508-508 are separated from the sense electrodes 510-514 by an insulating gap 516.

In operation, the device 500 is held at a fixed distance from the surface of an object to be analysed (not shown), orientated such that the plane containing the coplanar electrodes 504- 514 is substantially parallel to the surface of the object. In a preferred embodiment the device 500 further comprises a plurality of supports, for example wheels or rollers, configured to support the device such that the electrodes are held at a fixed distance from the surface of the object being tested.

In a preferred embodiment the electrodes 504-514 are disposed on a flexible substrate. The sensitivity of the capacitance imaging device 500 decreases as the distance between the device and the surface of the object increases. Advantageously use of a flexible substrate allows the electrodes to conform to curved surfaces of objects to be tested, for example the exterior of pipes, improving device sensitivity and speed of scanning.

In a preferred embodiment, electrodes 504-514 are connected to both the voltage source 502 and the voltage detector 518. In this embodiment the electrodes together act as a first plate of a capacitor. In the case of a metal or other conductive object (for example with a surface coating), the object itself acts as a second plate of a capacitor. In this way the device 500 in and the object in combination, act as a virtual capacitor. In operation, the alternating voltage source 502 applies an alternating voltage to the co-planar electrodes 504-514. The internal electronics present at the device 500 are held at a virtual ground. Since the object will effectively be at ground with respect to the voltage provided to the electrodes 504-514, a potential difference is induced between the electrodes 504-514 and the object when the voltage is applied, thereby creating a time varying capacitance between the electrodes 504- 514 and the object. The capacitance will vary depending on whether a defect (such as corrosion or a crack) is present, since such defects will alter the electrical properties of the object in the region of the defect. Such changes affect the actual potential of the electrodes for a given applied voltage applied by the voltage source 502. These changes are detected by voltage detector 518. Accordingly by measuring the changes in voltages of the electrodes 504-514 caused by changes in the capacitance between the electrodes 504-514 and the object, the device 500 can determine regions in which the capacitance between the electrodes 504- 514 and the object corresponds to the presence of a defect.

Advantageously this virtual capacitor arrangement removes the need to provide a "second" capacitor electrode - in other words, there is no need for another electrode or set of co-planar electrodes to be placed on the opposite side of the object to the electrodes 504-514 of the present device 500. Providing testing equipment on both sides of an object may be difficult, if not impossible in certain situations, for example when testing oil pipelines or ballast tanks. Accordingly this embodiment allows for detection of corrosion under paint in a wider range of circumstances than techniques involving two capacitor plates placed either side of the object.

In addition, this arrangement leads to very high spatial resolution when testing an object. This virtual capacitor arrangement provides a highly uniform electric field between the source electrodes 504-508 and the object itself, meaning that the areas of the object being measured by the device 500 corresponds to the areas of the coplanar electrodes 504-514 themselves, which may be made small, for example having dimensions of 3mm or less. Preferably, the electrodes 504-514 are either circular or annular in shape, which advantageously further ensures both high electric field uniformity and high spatial resolution. Figure 6 shows a photograph of an array of annular electrodes 600 mounted on a substrate 602.

In an alternative embodiment, in operation, the alternating voltage source 502 applies an alternating source voltage to the plurality of source electrodes 504-508, thereby applying a time varying electric field to the object. The electric field induces alternating voltages in each of the plurality of sense electrodes 510-514. The induced voltage in each of the sense electrodes 510-514 is measured by the voltage detector 518, and the measured values output to the processor 520. The device 500 is moved over the surface of the object. In some embodiments the device is moved manually by a user, alternatively the device is moved by mechanical means to achieve the same effect. For example, in one embodiment the device is incorporated in a remotely operated vehicle, allowing remote data collection in potentially hazardous or difficult to access areas. The voltage induced in each of the sense electrodes 510-514 is dependent on the electric field strength at the particular sense electrode. The electric field strength is determined by the properties at the object- surface coating interface, for example the field strength and hence induced voltage is different for a pristine interface to that of a portion of the interface where corrosion, cracks or other damage or defects are present.

The processor 520 is configured to record the position of the device at any given time. The position of the detector is determined using known means such as by measuring the displacement of the detector from a known, fixed, position and the processor 520 associates the measured induced voltage data from the voltage detector 518 at a certain time with a location on the surface of the object, thereby constructing an image of a region of the object corresponding to measured induced voltage data (i.e. the device performs positional encoding). Alternatively, or in addition, other positional encoding means may be provided, wherein such other positional encoding means are known in the art. In a preferred embodiment, the processor 520 is further configured to generate a visual image based on the induced voltage image that can be output to a display and viewed by a user. For example, each voltage value is converted into a hue, saturation, or brightness value or combination thereof. Thus regions of the image having particular visual properties (for example colour or brightness) provide an indication that corrosion, cracks or other damage or defects are present at a location of the object corresponding to those parts of the visual image. In some embodiments, the device 500 includes a display suitable for displaying the visual image, communication circuitry for transmitting the visual image to an external display device, or both.

In a preferred embodiment the processor 520 is configured to compare the measured induced voltage from the voltage detector 518 corresponding to a first location on the object being analysed to a value measured at a second location on the object. The processor 520 determines whether the difference in the measured value of induced voltage at each location is greater than a predetermined value, and if so determine that corrosion, a crack, or other damage or defect is present, providing an indication that this is the case to a user. In some embodiments the indication is visual, audible, or both. Advantageously, the arrangement of co-planar electrodes (504-514) supported a certain distance over the surface of an object to be analysed, in combination with either of the voltage sensing methods outlined above, allow for highly sensitive detection of corrosion, cracks and other defects using a single device operated from one side of the object to be analysed. This has advantages over known tomographic techniques which require sensors to be placed on either side on an object to be tested, and typically have lower sensitivity to surface defects. As noted above providing testing equipment on both sides of an object may be difficult, if not impossible in certain situations, for example when testing pipelines or ballast tanks.

In addition, the structure described above allows for measurements to be made even when the surface of the object (for example its coating) is conductive, which would not be possible using traditional tomographic techniques.

In a preferred embodiment, the processor 520 is configured to perform analysis on the image to identify areas of the image corresponding to damage present on the object. Preferably the processor 520 first enhances the image using known techniques, preferably including contrast enhancement and de-blurring processes. The processor 520 compares the measured voltage at different parts of the image by calculating the difference between the measured voltage at a first and second region of the image. Preferably a threshold value is chosen in advance through either theoretical or empirical means, corresponding to a voltage difference indicative of the presence of damage on the object. If the measured voltage difference is equal to or greater than the threshold value, the processor indicates that corrosion, a crack, or other damage or defect is present. Preferably the processor 520 is further configured to use known techniques to plot the boundary between regions in which damage is present and regions where no damage is present based on the relative measured voltages. Preferably the processor 520 further segments the image along the boundaries thereby forming discrete sections of the image. The processor is configured to provide a user with an indication of the sections corresponding to regions in which damage has been detected in the form of a visual output, an audible output, or more preferably both. In further embodiments the device 500 comprises means to store data describing the sections corresponding to detected damage, advantageously allowing retrospective data review by a user. If distinct source and sense electrodes are provided, preferably each source electrode 504-508 is paired with a corresponding sense electrode 510-514. The measured induced voltage of the sense electrode of each pairs at each position of the device 500 represents a pixel of the image generated by the processor 520.

Preferably the device 500 is further configured to measure the thickness of an object comprising an electrically conducting material, through contactless impedance measurements. Advantageously this allows the wall thickness of structures formed from conducting materials such as metals or conductive carbon fibre to be measured without damaging a surface coating. Such measurements are made in addition to the imaging functionality described above. Where the device 500 is configured to measure the thickness of the object, the device 500 further comprises compensation circuitry configured to balance the capacitive component of the impedance, such that the measured impedance comprises only resistive and inductive components. Preferably the compensation circuitry is a Schering bridge.

In order to measure the thickness of the object, an alternating voltage is applied to an electrode 504-514, the induced voltage in the electrode 504-514 is measured, from which the impedance of the part of the object corresponding to the area of the electrode is calculated. Once the capacitive component of the impedance has been balanced, the resistive component may be calculated. The measurement is then repeated for the same part of the object, using an applied voltage having a lower frequency than that used for the first measurement, beneficially increasing the depth to which the induced electric field penetrates the object being analysed. The measurement is repeated using successively lower frequencies of applied alternating voltage at the same part of the object. The measurements are preferably fitted to an n-degree polynomial which is then differentiated - beneficially this fitting process reduces noise. An area possessing a higher resistive component of the complex AC impedance is indicative of an area in which the object is thinner. Advantageously, this technique allows for the identification of regions of the object that have become thinner through corrosion on either the front or rear surface of the object. This allows for identification of corrosion in situations in which it is difficult or even impossible to place detection instrumentation on one side of a surface, for example, the interior surfaces of a pipeline or a ship's ballast tank. Advantageously the device 500 allows for the contactless measurement of the wall thickness of an object being analysed - known techniques for measuring wall thickness require that there is physical contact between the surface of the object itself and a measurement device and thus the surface coating to be damaged.

The above thickness measurement technique may also be performed in embodiments having distinct source and sense electrodes. In this case the voltage is applied to the source electrode 504-508 and the induced voltage in sense electrode 510-514 is measured. The method then proceeds as above.

In further embodiments, the device 500 also comprises an accelerometer, which outputs spatial data to the processor. The processor is configured to use this data to track the position of the device over time. Beneficially, this position tracking can be used to map the tomography of the object being analysed (in other words, the accelerometer and processor act as a positional encoding unit). Additionally, when a region corresponding to corrosion or other damage or defects is detected, the location of the region can be stored, advantageously allowing a user to scan a large area without pausing, and review any detected regions of interest after the scanning is complete. Alternatively, or in addition, other types of positional encoding units may be provided, wherein such other positional encoding units are known in the art.

Combined non-destructive imaging apparatus

In the preferred embodiment, there is provided a non-destructive imaging apparatus comprising at least two of the MMW/sub-MMW imaging device, NIR imaging device, and capacitance imaging device as described above. Each imaging device provides complementary data for each region of an object being analysed, which are synergistically combined, advantageously providing a user with an enhanced, more detailed indication of areas of the object in which corrosion, cracks or other damage or defects are present.

In a first embodiment, a MMW/sub-MMW imaging device (as detailed above with reference to either figure 1 or figure 2) and an NIR imaging device 300 are combined into a single nondestructive imaging system. Preferably the apparatus is configured to detect reflected MMW/sub-MMW radiation and reflected NIR radiation that has been reflected from the same area of an object being analysed simultaneously, advantageously reducing the time required to scan the object.

To ensure that the detected MMW/sub-MMW and NIR radiation was reflected by the same area of the object, it is necessary to detect MMW/sub-MMW and NIR radiation travelling along substantially the same axis. Preferably the first embodiment also includes a single hybrid MMW/sub-MMW/NIR focussing component that allows reflected MMW/sub-MMW and NIR radiation entering the device along substantially the same axis to be separated, and focusses reflected MMW/sub-MMW radiation onto a MMW/sub-MMW detection unit and focusses reflected NIR radiation onto a NIR detection unit. Advantageously, providing a single hybrid component capable of focussing both MMW/sub-MMW and NIR radiation minimises the amount of space needed to house the apparatus, and the number of optical components required hence minimising weight. Thus the portability of the apparatus is improved. It is also noted that by housing MMW/sub-MMW and NIR imaging devices within a single unit and sharing optical components, the cost of the apparatus is reduced in comparison with separate MMW/sub-MMW and NIR imaging devices in isolation.

Figure 6 shows a schematic of a combined MMW/sub-MMW and NIR imaging device 700. In operation both MMW/sub-MMW 710 and NIR radiation 712 that has been reflected or emitted by an object being analysed enter the device, and are directed onto hybrid focussing optics 704 by a beam deflection unit 702. The hybrid focussing optics 702 are configured to simultaneously reflect MMW/sub-MMW radiation 710 onto a MMW/sub-MMW detection unit 708, and NIR radiation 712 onto an NIR detection unit 706.

Preferably the focussing optics 704 are parabolic in cross section. Preferably the device 700 further comprises a MMW/sub-MMW radiation source and an NIR radiation source (not shown) which operate as described above, and a processor (not shown) configured to perform the analysis as described above in relation to the MMW/sub-MMW and NIR devices of figures 1-3.

Figure 8A shows a cross sectional view, and figure 8B a top view of hybrid focussing optics 800, comprising a first portion 802 and a second portion 804. Together, the first and second portions 802 804 define a parabolic surface of a mirror. Both the first portion 802 and the second portion focus MMW/sub-MMW radiation onto the MMW/sub-MMW detection unit. The second portion 804 focusses NIR radiation onto the NIR detection unit. As the angular resolution of the device is directly proportional to the wavelength of radiation being used the mirror can be used to detect at both wavelengths simultaneously. Advantageously by having a large focusing area (the combined first and second portions) for MMW/sub-MMW radiation, as well as the focussing area of the second portion for NIR radiation, a high level of detail is resolved for both wavelength regimes. To allow optimised image formation, the surface of the focussing optics is smooth to within a quarter of the wavelength being focussed. Therefore, for a mirror to effectively focus NIR radiation having a short wavelength, it must have a far higher degree of smoothness than would be necessary to effectively focus MMW/sub-MMW radiation having a much longer wavelength. Advantageously, the configuration of the hybrid focussing optics 800 means that only the second portion 804, which has a smaller focussing area than the first portion 802, needs to conform to the high smoothness condition required for NIR radiation, reducing the costs associated with precision manufacture. In a preferred embodiment the first portion is fabricated from inexpensive spun aluminium.

In a second embodiment, an NIR imaging device 400 as shown in figures 4A and 4B and a capacitance imaging device 500 are combined into a single unit. Advantageously the small lightweight configuration of the handheld NIR imaging device 400 makes it ideal for incorporation with a capacitance imaging device 500, which must be physically moved over the surface of an object to be analysed. Additionally, the configuration of NIR radiation sources and NIR detection units as shown in figures 4A and 4B can be easily optimised for the detection of reflected NIR radiation from a surface close to the radiation sources.

In other embodiments, other combinations of MMW/sub-MMW imaging device, NIR imaging device, and capacitance imaging device are used.

Regardless of the combination of imaging devices used, the non-destructive imaging apparatus is preferably configured to synergistically analyse data output from each imaging device. Preferably the non-destructive imaging system comprises a data processing system, configured to overlay image data from each of the individual devices (produced as described above) to create a further image containing a greater degree of detail. Preferably this image is presented in visual form either at a display forming part of the system, or at an external display in communication with communication circuitry included in the system. Advantageously a user of the system is presented with a more complete picture allowing fast and intuitive identification of corrosion, cracks or other damage or defects that may be present on an object concealed from visual inspection under a surface coating. By using a combination of techniques it is also possible to locate damage or defects that are more effectively identified using one of the imaging techniques than another - this is particularly the case in situation where the object being analysed has multiple different surface coating materials in multiple different areas, in which case corrosion might be more readily detected using MMW/sub-MMW techniques in one area, NIR techniques in another area, and capacitance techniques in a other areas.

The combination of one or more of the imaging devices above in a single apparatus has additional benefits over using each of the imaging devices in isolation. By combining the devices into a single apparatus the time taken by a user to complete analysis of a given area of an object is greatly reduced. This is particularly important when testing an object in an industrial environment that requires normal operations to be stopped whilst testing is in progress.

Test results

To demonstrate the effectiveness of certain of the above embodiments, the following tests have been carried out.

Figure 9A shows a photograph of a sample 900 made of steel. The sample 900 includes areas of corrosion 902. The sample 900 also includes areas 904 in which material has been removed from the surface of the sample 900, and the resulting defect has been filled in with material.

After taking the visible photograph shown in figure 9A, the sample 900 was coated with a corrosion resistant zinc oxide primer and then covered with several layers of micaceous metal oxide based paint. Such coatings are typically used on industrial structures such as pipelines. The surface defects 902 904 were concealed by the coating such that they were no longer visible to the naked eye.

Figure 9B shows imaging results 906 obtained by an active MMW/sub-MMW imaging device (in accordance with the preferred embodiment of device 200 described above in relation to figure 2) when imaging the coated sample 900. The results 906 are shown overlaid on a photograph of sample 900 taken before the sample 900 was coated. The results 906 clearly show the presence of the surface defects 902 904.

Figure 9C shows imaging results 908 obtained by a capacitance imaging device (in accordance with the preferred embodiment of device 500 described above in relation to figure 5) when imaging the same coated sample 900 as shown in figure 9A and as tested by the capacitance device 500 in figure 9B. Again, the results 908 clearly show the presence of the surface defects 902 904.

Accordingly the preferred embodiments of the active MMW/sub-MMW imaging device 200 and the capacitance imaging device 500 provide clear and accurate indications of surface defects on a metal sample 900 concealed by layers of zinc oxide primer and micaceous metal oxide based paint.

Figures 1 OA- IOC show exemplary graphical user interfaces (GUI) 1002 1006 1010 for use with software for analysing data obtained from one or more imaging devices 100 200 300 400 500. Figure 10A shows a GUI 1002 displaying capacitance imaging results 1004 after scanning a sample 900. Figure 10B shows a GUI 1006 displaying MMW/sub-MMW imaging results 1008 after scanning a sample 900.

In operation, analysis software identifies parts of the scanned sample that correspond to damage or corrosion etc. for both the capacitance results 1004 and MMW/sub-MMW results (other combinations of imaging results may also be used). The software then identifies parts of the sample where defects appear for both the capacitance and MMW/sub-MMW results, in combination with data output by the positional encoding means provided at least device, and provides an indication of these parts to a user (again, other combinations of imaging results may also be used). Figure IOC shows a GUI 1010 in which regions 1012 corresponding to possible defects identified in results from both a capacitance imaging device 500 and a MMW/sub-MMW device 200 are displayed to a user. As can be seen in figures 10A and 10B, the areas indicating defects 1004 1008 overlap. Advantageously, by identifying areas of an object that correspond to damage identified by two or more of the imaging techniques described above, certain damage can be located with greater accuracy, and properties of the damage can be more accurately inferred. For example, certain types of damage may be seen to a lesser or greater extent using certain techniques, or using certain wavelengths - therefore by identifying regions corresponding to defects for each imaging method or wavelength, properties of the defect may be inferred.

Claims

1. A capacitance imaging device which performs non-destructive detection of damage present under a surface coating on a conductive object comprising:

one or more coplanar electrodes;

a voltage source; and

a processor;

wherein each of the one or more coplanar electrodes define one or more first capacitor plates;

wherein the device is configured such that in operation, when held on or near the surface of the object being imaged, the surface of the object acts as a second capacitor plate wherein the voltage source is configured to apply an alternating voltage to at least one electrode, and the processor is configured to measure an output voltage of the least one electrode;

wherein the processor is further configured to identify areas of the object

corresponding to detected changes in the output voltage corresponding to changes in a capacitance between the one or more first capacitor plates and the second capacitor plate, wherein the change in capacitance is indicative of the presence of damage under the surface coating, when the one or more coplanar electrodes are moved parallel to the surface of the object; and

wherein the processor is further configured to provide an output indicating the areas of the object identified as corresponding to the presence of damage under the surface coating.

2. The device of claim 1 wherein the processor is further configured to generate an image of the object using the output voltage as the coplanar electrodes are moved over an area of the surface of the object, and is further configured to compare the output voltage at different parts of the image.

3. The device of claim 2 wherein the processor determines whether damage under the surface coating is present for a given area based on the comparison between the output voltage at different parts of the image.

4. The device of any of claims 1 to 3 wherein the processor is further configured to provide an indication to a user that a defect is present.

5. The device of any preceding claim further comprising a support positioned such that the one or more coplanar electrodes are held at a fixed distance from the surface of the object during use.

6. The device of claim 5 wherein the support is a rotating member.

7. The device of any preceding claim, wherein the device is further configured to determine the relative thickness of the object upon which the coating is placed,

wherein the processor is configured to measure at a plurality of locations the resistive component of an AC impedance of the object based on the detected output voltage for each of said locations;

and for each of a plurality of locations in the object the processor is configured to compare the plurality of measured resistive components of AC impedance at the plurality of locations, and to determine whether comparison of the resistive components of AC impedance for the object at a certain position has a value that falls outside a predetermined range of tolerance, and in the event that the value falls outside the predetermined range of tolerance to provide an indication that the object has a reduced thickness at that position.

8. The device of claim 7 wherein the predetermined range of tolerance is based on an average of the plurality of measured resistive components.

9. The device of claim 7 or 8 wherein the resistive component of an AC impedance of the object based on the detected output voltage measured at a plurality of AC frequencies.

10. A millimetre-wave and sub-millimetre-wave imaging device which performs nondestructive detection of damage present on an object covered by a surface coating

comprising:

a radiation source for providing at least one of coherent millimetre-wave radiation and coherent sub-millimetre-wave radiation;

a broadband detection unit configured to detect millimetre-wave and sub- millimetre-wave radiation;

a modulator configured to modulate the radiation so as to produce radiation that is substantially incoherent; and a processor; wherein

the processor is configured to identify detected reflected radiation having an intensity indicative of the presence of damage, and determine a first area of the object from which the radiation indicative of the presence of damage was reflected; and wherein

the processor is further configured to provide an output indicating that the first area.

11. The device of claim 10, wherein the modulator is configured to continuously vary the phase of the radiation provided by the radiation source, between 0 and 360 degrees.

12. The device of claim 10 or claim 11, wherein the modulator comprises means for at least one of:

mechanically moving the radiation source along a direction of radiation emitted by the radiation source; and

mechanically moving the detection unit along a direction of radiation received by the detection unit;

wherein, in operation, at least one of the radiation source and detection unit is moved over a first distance in a first time period.

13. The device of claim 12 wherein the first distance equals at least one wavelength of the radiation provided by the radiation source, and the time period is substantially one time period of the radiation provided by the radiation source.

14. The device of any of claims 10 to 13 further comprising focussing optics, wherein the focussing optics are arranged to focus millimetre-wave and sub-millimetre-wave radiation reflected by the object onto the detection unit.

15. The device of any of claims 10 to 14 wherein the processor is further configured to generate an image of the object using the detected radiation, measure a difference in the intensity of detected radiation between a first and a second region of the image, and determine that damage is present on the object at a location corresponding to the first part of the image based on a comparison of the difference with a predetermined value.

16. The device of any of claims 10 to 15 wherein the detection unit comprises one or more detectors.

17. The device of any of claims 10 to 16 wherein each of the one or more detectors is selected from the group comprising; a glow discharge detector, a pyroelectric detector.

18. The device of claim 17 wherein the detector is a glow discharge detector comprising an anode and a cathode, wherein a voltage is applied across the anode and the cathode, and wherein the magnitude of the voltage is chosen based on the frequency of the radiation.

19. The device of any of claims 10 to 18 further comprising a beam splitter, wherein the beam splitter is positioned such that a portion of incident radiation passes through the beam splitter, and a portion of reflected radiation is diverted onto the detector.

20. The device of claim 19 wherein the beam splitter comprises a metallised polymer sheet.

21. A near-infra-red imaging device which performs non-destructive detection of damage present under a surface coating on an object comprising:

a near-infra-red laser radiation unit comprising at least one laser source; a detection unit comprising at least one detector configured to detect near- infra-red radiation; and

a processor;

wherein the device is configured to project incident near-infra-red radiation from the laser radiation unit onto a first area of the object at a first time;

wherein the device is configured to project incident near-infra-red radiation from the laser radiation unit onto a second area of the object at a second time;

wherein the detector detects near-infra-red radiation reflected by the object; wherein the processor is configured to identify areas of the object

corresponding to detected reflected radiation having an intensity indicative of the presence of damage under the surface coating; and wherein the processor is further configured to provide an output indicating the areas of the object identified as corresponding to the presence of damage under the surface coating.

22. The device of claim 21 wherein the processor is further configured to generate a first image of the object using the detected radiation, and is further configured to measure a difference in the intensity of detected radiation between a first and a second region of the image, and determine that damage is present on the object at a location corresponding to the first part of the image based on a comparison of the difference with a predetermined value.

23. The device of any of claims 21 or 22 further comprising a beam splitter, wherein the beam splitter is configured to direct a portion of near-infra-red radiation that is reflected by the object along the same optical axis as the projected incident near-infra-red radiation onto the detector.

24. The device of any of claims 21 to 23 wherein the laser radiation unit further comprises at least two laser sources, and is further configured to emit near infra-red radiation having at least two wavelengths.

25. The device of any of claims 21 to 24 wherein the device is configured to generate a plurality of images at different wavelengths and further configured to generate a differenced image.

26. The device of any of claims 21 to 25 further comprising a narrow band optical interference filter.

27. The device of any of claims 21 to 26 wherein each of the at least one detectors is selected from the group of a thermal camera, a short wave infra-red detector, a photo-resistive detector.

28. The device of any of claims 21 to 27 wherein the changing of the position of the projected dot over time is performed by a beam deflection unit, the beam deflection unit configured to deflect radiation emitted by at least one laser source by varying angles over time.

29. The device of claim 28 wherein the beam deflection unit comprises: a galvanometer comprising a deflecting member; and

a mirror connected to the deflecting member.

30. The device of any of claims 21 to 29 further comprising a plurality of laser sources, wherein the changing of the position of the projected dot over time is performed by emitting near infra-red radiation from each of the plurality of laser sources in turn.

31. A near-infra-red imaging device for non-destructive detection of internal defects comprising:

a laser source;

a detector configured to detect near-infra-red radiation;

a processor;

wherein the laser projects incident radiation onto an object to be analysed, thereby locally heating a portion of the object;

wherein the detector detects emitted near-infra-red radiation emitted by the locally heated portion; and

wherein the processor is configured to identify areas of the object

corresponding to detected emitted radiation having an intensity indicative of the presence of damage under the surface coating; and

32. The device of claim 31 wherein the processor is further configured to generate a first image of the object using the detected radiation, and is further configured to measure a difference in the intensity of detected radiation between a first and a second region of the image, and determine that damage is present on the object at a location corresponding to the first part of the image based on a comparison of the difference with a predetermined value.

33. The device of any of claims 31 or 32 further comprising a beam splitter, wherein the beam splitter is configured to direct a portion of near-infra-red radiation that is reflected by the object along the same optical axis as the projected incident near-infra-red radiation onto the detector.

34. The device of any of claims 31 to 33 wherein the device is configured to generate a plurality of images at different wavelengths and further configured to generate a differenced image.

35. The device of any of claims 31 to 34 further comprising a narrow band optical interference filter.

36. The device of any of claims 31 to 35 wherein each of the at least one detectors is selected from the group of a thermal camera, a short wave infra-red detector, a photo-resistive detector.

37. The device of any of claims 31 to 36 wherein the location of the locally heated portion is changed over time using a beam deflection unit, the beam deflection unit configured to deflect radiation emitted by at least one laser source by varying angles over time.

38. The device of claim 37 wherein the beam deflection unit comprises:

a galvanometer comprising a deflecting member; and

a mirror connected to the deflecting member.

39. A device for non-destructive detection of internal defects comprising:

two or more imaging systems selected from the group comprising;

a millimetre-wave imaging system comprising a millimetre-wave radiation detector configured to detect electromagnetic radiation with a frequency in the millimetre range;

a near-infra-red imaging system comprising a near-infra-red laser source and a near-infra-red radiation detector; and

a capacitance imaging system comprising two coplanar electrodes separated by an insulating gap, an alternating voltage source and a means for measuring voltage output by one or more electrodes;

the device further comprising a processor; the processor configured to analyse data from each of the imaging systems, so as to determine the presence of damage under the surface and further configured to combine image data from each of the two or more imaging systems.

40. The device of claim 39 wherein the two or more imaging systems include at least a millimetre-wave imaging system and a near-infra-red imaging system; the device further comprising a hybrid focussing optic comprising a substantially parabolic first surface positioned within a substantially parabolic second surface.

41. The device of claim 40 wherein the hybrid focussing optic is positioned within the device such that millimetre-wave radiation reflected by an object being analysed is focussed onto the millimetre-wave radiation detector by the first and second surfaces, and such that near-infra-red radiation reflected by the object is focussed onto the near-infra-red radiation detector by the second surface.

42. The device of either of claims 40 or 41 wherein the first surface is spun aluminium.