Electromagnetic acoustic transducer

Abstract

An electromagnetic acoustic transducer (EMAT) for exciting ultrasonic vibrations within a test object 10, the EMAT comprising: at least one magnet 2 configured to generate a magnetic field; an electri

Description

The present technique relates to the field of ultrasonic non-destructive testing using electromagnetic acoustic transducers. Ultrasonic non-destructive testing may be used to detect a property of an object under test. An ultrasonic detector located at the surface of a solid object can detect output vibrations received from within the object in response to transmission of a pulse of input vibrations into the surface of the object, so that properties of the object which may not be visible at the surface can be detected. For example, where the object is a wall of a pipe, vessel, tank or other conduit or container, the ultrasonic non-destructive testing can be used to detect changes in the inner surface of the conduit / container, such as monitoring changes in the thickness or surface roughness of the wall, for example caused by corrosion or erosion. As an example, using monitoring techniques to track the corrosion or erosion of the inner surface of a pipe in a refinery may permit the safe refining of oil which would otherwise be regarded as too difficult due to the way in which it corrodes or erodes the pipes of the refinery. Input vibrations may be excited within the test object by use of electromagnetic acoustic transducers (EMATs). EMATs may also be used to detect output vibrations received from the test object. It would be desirable to increase a signal-to-noise ratio in the vibrations measured by an EMAT, to enable more accurate determination of properties of the test object. At least some examples provide an electromagnetic acoustic transducer for exciting ultrasonic vibrations within a test object, said electromagnetic acoustic transducer comprising: at least one magnet configured to generate a magnetic field; an electrical coil configured to carry current for inducing electrical currents in the test object; and an electrically conductive core disposed between the at least one magnet and the electrical coil, for conveying the magnetic field from the at least one magnet to the electrical coil; wherein the electrically conductive core has a structure configured to reduce intensity of ultrasonic vibrations within the electrically conductive core. At least some examples provide a method for reducing noise in a signal detected using an electromagnetic acoustic transducer for exciting ultrasonic vibrations within a test object, comprising: generating a magnetic field with at least one magnet; generating an electrical current in an electrical coil, to induce electrical currents in the test object; and conveying the magnetic field from the at least one magnet to the electrical coil using an electrically conductive core disposed between the at least one magnet and the electrical coil; wherein the electrically conductive core has a structure configured to reduce intensity of ultrasonic vibrations within the electrically conductive core. Further aspects, features and advantages of the present technique will be apparent from the following description of examples, which is to be read in conjunction with the accompanying drawings, in which: Figure 1 is a schematic diagram illustrating an electromagnetic acoustic transducer (EMAT) in place on the surface of a test object; Figure 2A is a schematic illustration of a laminated core; Figure 2B illustrates a portion of a coil on the surface of a test object; Figure 2C illustrates a laminated core in position between a magnet and an electrical coil in an EMAT placed on a test object; Figures 3A to 3C illustrate a multi-block core; Figure 4 is a graph comparing attenuation of ultrasonic vibrations in two cores having different structures; Figure 5A is a graph showing amplitude over time of a signal measured by an EMAT having a non-conductive soft magnetic core; Figure 5B is a graph showing amplitude over time of a signal measured by an EMAT having a laminated conductive core comprising a single block; Figure 5C is a graph showing amplitude over time of a signal measured by an EMAT having the laminated conductive core of Figure 3C; Figure 6 is a graph illustrating the signal-to-noise ratio (SNR) of an EMAT having a core as shown in Figure 3C over an extended period of time in a test environment; Figure 7 is a graph illustrating the effectiveness of cores having different values of magnetic permeability in a test arrangement representative of an EMAT; Figure 8 illustrates a method of using an EMAT to excite ultrasonic vibrations in a test object and measure a resulting signal from the test object; Figure 9 is a graph illustrating the effect on noise in an EMAT attributable to varying the thickness of the laminated layers in a laminated electrically conductive core; and Figure 10 is a graph illustrating the effect on noise in an EMAT attributable to varying a grain size for an electrically conductive material. An electromagnetic acoustic transducer (EMAT) for exciting ultrasonic vibrations within a test object comprises at least one magnet configured to generate a magnetic field. The magnetic field may for example be a static magnetic field generated by a permanent magnet. In other examples the at least one magnet may comprise a continuous or pulsed electromagnet for generating the magnetic field. The EMAT also comprises an electrical coil configured to carry current for inducing electrical currents in the test object. The electrical coil can carry a time-varying current (e.g., an alternating current) to generate a time-varying magnetic field in the vicinity of the test object. The varying magnetic field induces eddy currents in the surface of the test object when the test object is made from a conductive material. The eddy currents generated in the test object interact with the magnetic field generated by the at least one magnet, and through the Lorentz force generate ultrasonic waves in the test object. The same principles also enable an EMAT to measure ultrasonic vibrations in the test object. Vibrations in the test object cause the displacement of charge carriers in the test object. The movement of charge carriers in the magnetic field generated by the at least one magnet causes those charge carriers to experience a Lorentz force which causes a varying current to be generated in a surface of the test object. The varying surface current generates a varying magnetic field which induces a current in the coil. The current in the coil can be measured to provide a signal indicative of the ultrasonic vibrations in the test object. The measurement can be repeated a number of times and the resulting signals averaged to provide an output value indicative of the ultrasonic vibrations in the test object. Averaging may be performed because, in some examples, the signal due to the ultrasonic vibrations may be lower than the random electrical noise in the signal. Appropriate averaging can however reduce the effect of random noise to isolate the signal caused by ultrasonic vibrations in the test object. The magnitude of vibrations induced in the test object and the magnitude of currents induced in the electrical coil when the EMAT is used as a detector are both proportional to the strength of the magnetic field at the surface of the test object. An EMAT may be configured to separate the at least one magnet from the electrical coil, for example because the same mechanism which generates vibrations in the test object could otherwise also generate vibrations in the magnet, leading to noise in the EMAT. Separating the magnet from the electrical coil reduces the strength of induced currents in the magnet and hence reduces noise. However, in many examples the electrical coil may be arranged to be provided close to the test object, and separating the at least one magnet from the electrical coil can lead to the strength of the magnetic field experienced at the surface of the test object being significantly reduced. For example, the strength of a static magnetic field may have an inverse cube relationship with distance, meaning that providing a gap between the coil (at the test surface) and the magnet can significantly reduce the strength of the magnetic field at the test surface. To reduce the decrease of the magnetic field strength caused by the gap, the EMAT also comprises a core disposed between the at least one magnet and the electrical coil, for conveying the magnetic field from the at least one magnet to the electrical coil, and hence to the surface of the test object when in use. The core may also be referred to as a pole piece. The magnetic permeability of the core may define the capability of the core to guide magnetic field from the at least one magnet to the electrical coil. In a conventional approach, the core may be made from a non-conductive material. If the core were made from an electrically conductive material, then the same mechanism which induces eddy currents in the test object could also induces eddy currents in the core. Ultrasonic vibrations could therefore be induced in the core by the interaction of the eddy currents and the magnetic field generated by the at least one magnet. Ultrasonic vibrations in the core may act as a source of coherent noise. The noise which may be generated in an electrically conductive core can have a similar waveform to ultrasound generated in the test object, and hence can be difficult to isolate from a signal measured from the test object. For example, averaging used to reduce random electrical noise in the signal may not reduce the impact of coherent noise caused by ultrasonic vibrations in the core. Therefore, ultrasonic vibrations induced in the core could significantly reduce the signal-to-noise ratio of measurements made using an EMAT. Use of a non-conductive material for the core can suppress the generation of eddy currents in the core and hence can reduce the generation of noise in the core. Hence, it would typically be considered counter-intuitive to provide a core made from an electrically conductive material. However, the inventors have recognised that electrically conductive materials typically have properties which are desirable for the core. Electrically conductive materials may for example generally have higher magnetic permeability than non-conductive materials, and may therefore be more effective at conveying magnetic field from the at least one magnet to the electrical coil. By conveying a larger proportion of the magnetic field to the test surface of the EMAT, then either the device can be made more accurate or the size of the at least one magnet can be reduced whilst maintaining accuracy. Hence, an improved EMAT can be provided by using an electrically conductive core. The inventors have also realised that the intensity of ultrasonic vibrations in the core can be reduced even when electrically conductive materials are used for the core by configuring the structure of the electrically conductive core to reduce intensity of ultrasonic vibrations within the electrically conductive core. The structure of the core may be configured in various ways to reduce the intensity of ultrasonic vibrations. These will be discussed in greater detail below. In some examples, the configuration of the core suppresses the generation of ultrasonic noise in the core by reducing the ability of the core to support the eddy currents which are responsible for the generation of ultrasonic noise. For example, the surface of the core may be divided into regions which have a size small enough that generation of eddy currents is suppressed in the surface region of the core. The structure of the core may also or alternatively be configured to cause attenuation of ultrasonic vibrations which are generated in the core, to reduce the intensity of said ultrasonic vibrations. For example, the core may comprise scattering sites or may have a shape which can cause dissipation of energy from ultrasonic waves in the core. In either case, the inventors have surprisingly found that by controlling the structure of the core, an electrically conductive core can be used in an EMAT, and hence measurement performance can be improved, whilst reducing the intensity of noise arising from the core. The core may not be entirely made from an electrically conductive material, but at least comprises portions of electrically conductive material. In some examples, the core may be electrically conductive in at least one direction. In some examples, an electrically conductive material is one in which eddy currents may be induced by an external electric field. For example, this definition may include metals and therefore in some examples the electrically conductive core comprises at least one metallic portion. In some examples, in at least one direction the electrically conductive core may have an electrical conductivity greater than 1x106 Q^nr1. The inventors have determined that at least cores made from a material having a conductivity above this threshold may have a conductivity high enough that eddy currents could form in the core and that, absent the current techniques, such a core could act as a significant source of coherent noise in an EMAT. In at least some examples, the core may be made from NO20 electrical steel, having an electrical conductivity of 1.96 x106 Q'1m'1. The inventors have found that NO20 electrical steel is a material having desirable magnetic permeability and, despite being an electrical conductor, can be used to make a practical EMAT core using the present techniques, allowing high signal strengths to be achieved with reduced noise. In some examples, the electrically conductive core has a magnetic relative permeability greater than or equal to 70, where relative permeability is the ratio of the permeability of the core to the permeability of free space. Materials having a relative permeability of 70 or greater may achieve a magnetic flux density at the surface of the test object of over 95% of the maximum magnetic flux density which could be achieved at the surface. A relative permeability of 70 or greater may be considerably more difficult to obtain in materials which are not electrically conductive. Such materials may for example be expensive or less stable, and hence may not be suitable for environments in which an EMAT may be used. On the other hand, a relative permeability of over 70 can be obtained more easily with electrically conductive materials. For example, various forms of steel may have a relative permeability over 1000. In some examples, the electrically conductive core may comprise two or more electrically conductive portions defining, at their intersection, one or more interfaces within the electrically conductive core. The interface may include an air gap between two portions of the core or, for example, could include a region of adhesive between two portions of the core. Generally a material in the interface may have a different speed of sound than the speed of sound in the electrically conductive material forming the core. The inventors have found that the inclusion of at least one such interface within the core can cause attenuation of ultrasonic vibrations which are induced in the core. Reflection of ultrasonic waves from an interface, or transmission through the interface, can cause energy to be dissipated from the ultrasonic waves. Therefore, configuring the structure of the core to include at least one interface can reduce intensity of ultrasonic vibrations within the electrically conductive core. In some examples, the electrically conductive portions may be sheets of a laminated core, and interfaces may be provided in layers between laminated sheets. Alternatively, or in addition, the electrically conductive portions may be blocks of a multi-block core, with one or more interfaces between blocks. In some examples the electrically conductive core comprises, in at least a test face region of the core, laminated layers of an electrically conductive material separated by layers of an electrically insulating material. The test face region may be a region of the core closest to the electrical coil, which in use may also be the region of the core closest to the test object. Being closest to the electrical coil, the test face region will experience the strongest induction of eddy currents from the electrical coil. Eddy currents induced in the core by the electrical coil will be limited to the surface of the electrically conductive core, as the current density decreases exponentially over the depth of the surface. The skin depth is defined as the distance from the surface of the material to the depth where the induced current density is decreased by a factor of 1 / e (i.e. to approximately 37% of the surface current density). The test face region may therefore be defined as the portion of the face of the core facing the electrical coil in the region from the surface to the skin depth. The test face region is the region in which the eddy currents causing ultrasonic vibrations in the core may be primarily generated. Hence, when providing in a core a structure to reduce the formation of eddy currents, the structure may be provided in at least the test face region. Laminated layers of an electrically conductive material and an electrically insulating material provide a particularly effective structure for reducing the formation of eddy currents. Laminated layers may be provided as parallel planar sheets of a conductive material separated by an electrically insulating material. The insulating material may in some examples be an adhesive joining the conductive sheets. Eddy currents generated in the surface of a conductive material may form as current loops. If the dimensions of the conductive material are limited then the maximum diameter of the current loops is also limited, meaning that the magnitude of induced eddy currents can be reduced. Laminated sheets provide a constraint on the dimensions of the conductive material and hence on the induced current loops because the induced eddy currents cannot cross between laminated sheets due to the insulating layers. A core which is laminated, at least in the test face region, can therefore reduce the intensity of ultrasonic noise in the core by reducing the formation of eddy currents and hence reducing generation of ultrasonic vibrations. The laminated sheets may be provided at least in the test face region, but may also be provided in other parts of the core. In some examples, for example, an entire core may be provided from laminated sheets of electrically conductive material separated by electrically insulating material. Alternatively, at least a block (discussed below) comprising the test face region may be provided from laminated sheets. As discussed above, providing a number of laminated layers also provides a number of interfaces within the core. Therefore, in addition to reducing the generation of ultrasonic vibrations in the core by suppressing formation of eddy currents, providing a laminated core 7 also increases attenuation of generated ultrasonic vibrations by increasing dissipation of ultrasound within the core. The thickness of the laminated layers, their extent in a direction orthogonal to the plane of the laminated layers, may not be particularly limited, and may vary in different implementations of the present techniques. The performance achieved using layers of different thicknesses may vary depending on a frequency of the signal carried by the electrical coil, and hence an acceptable level of performance may be achieved using different thicknesses depending on a particular implementation. However, in general performance may be improved by using layers having lower thicknesses, as this may suppress the ability of the electrically conductive layers to support current loops. In some examples, the thickness of an electrically conductive layer may be less than 1mm. A plurality of the electrically conductive layers, e.g., all of the electrically conductive layers, may have equal thicknesses, and in some examples the plurality of layers may have a thickness of less than 1mm. The inventors have determined that, for a range of useful frequencies for the signal supported by the electrical coil, a thickness of 1mm may be the largest value for which a signal-to-noise ratio is acceptable. For example, for a range of frequencies in the range 2MHz to 5MHz, a thickness of 1mm may enable performance loss caused by the presence of coherent noise generated in the core to remain below 12dB. In some examples, the thickness of an electrically conductive layer may be less than 0.7mm, which may be the largest thickness which can enable a higher level of performance. For example, for a range of frequencies in the range 2MHz to 5MHz, a maximum thickness of 0.7mm may enable performance loss caused by the presence of coherent noise generated in the core to remain below 6dB. In some examples, the thickness may be less than 0.6mm, to enable worst case performance to remain below 3dB performance loss. The orientation of the laminated layers of the core may affect their effectiveness at reducing formation of eddy currents. The electrical coil may be substantially planar, such that the plane of the coil may be parallel with the surface of the test object in use. Providing laminated layers parallel to the plane of the coil can reduce the intensity of ultrasonic vibrations in the core by providing a set of interfaces in the core for attenuating ultrasonic vibrations. In this orientation the eddy currents may be generated in the plane of the layers and hence the layers may not provide much restriction on the extent of the eddy currents, and hence may not substantially limit formation of eddy currents. Therefore, in some examples, a direction orthogonal to a plane of the laminated layers may lie in a plane of the electrical coil. In other words, the plane of the laminated layers may be orthogonal to the plane of the electrical core. In this orientation the eddy currents may not be generated in the plane of the layers and hence the laminations are able to provide a greater limitation on the extent of eddy currents which can be formed in the core. In some examples, the coil is configured to carry current in a substantially linear direction in a region adjacent to the electrically conductive core. For example, the coil may have a racetrack configuration and the core may be provided along a straight portion of the racetrack. In this configuration, the eddy currents may be induced in the core in a direction aligned with the linear direction of the coil. In some examples, the formation of eddy currents can be further reduced when the core is configured such that the direction orthogonal to the plane of the laminated layers is parallel to said linear direction of the coil. In this orientation, the laminated layers cut across the direction in which eddy currents are preferentially generated. Hence, the extent of conductive material is reduced in the direction in which eddy current are preferentially generated. Therefore, the formation of eddy currents can be further suppressed by the use of a core having a laminated test face region. In some examples, the electrically conductive core may be formed from two or more blocks of material. For example, the two or more blocks may include a test face block comprising the test face region, and at least one further block. The two or more blocks may combine to form the electrically conductive core. For example, the blocks may be held together by an interference fit. In other examples, the blocks may be held together by an adhesive. Providing several blocks forming the core can provide several advantages, discussed below. A first advantage is that the use of several blocks provides at least one interface within the core which can act to dissipate ultrasonic vibrations generated in the core. In some examples, the test face block may be entirely laminated. Compared to providing laminations only in a test face region, a fully laminated test face block may be more straightforward to manufacture and may dissipate ultrasound more effectively. In an example of the present technique using a laminated test face block, there may be no requirement for blocks other than the test face block to include laminations, as these blocks may be further from the electrical coil and hence less susceptible to the formation of eddy currents. In some examples, the laminated layers in the test face block may be stacked along a first direction, and the laminated layers may have a common cross section when viewed along the first direction. For example, each of the layers forming the laminated test face block may be identical. This can simplify manufacture of the test face block compared to examples in which the laminated sheets differ. The shape of the core can be an important consideration because cores which fill the gap between the at least one magnet and the electrical coil more completely may be more effective at conveying magnetic field to the electrical coil. However, providing a test face block with laminations having a common cross section may limit the shape of the test face block to a prism, which in some configurations may not be an ideal shape for filling the gap. Providing a core having two or more blocks can overcome this problem because the overall shape of the core can be varied whilst providing a test face block having laminations with a common cross section. Such a core therefore benefits from simplified manufacture whilst providing a more effective shape. In some examples, the at least one further block is configured to have a shape which interlocks with the test face block to restrict relative movement of the laminated layers of the test face block. For example, at least one further block may wrap around the test face block to restrict movement of the laminated layers of the test face block in one or more directions. For example, the relative movement of the laminated layers may be restricted in a direction orthogonal to the plane of the laminated layers. The insulating layers between the conductive laminated sheets may act as an adhesive to hold the core together. However, the inventors have realised that certain environments in which an EMAT may be used may reduce the effectiveness of the adhesive, and therefore there may be a risk of a laminated core becoming separated. For example, an EMAT may be used for testing oil pipes in a high temperature environment which can thermally degrade the adhesive. A multi-block core in which at least one further block is configured to have a shape which interlocks with the test face block to restrict relative movement of the laminated layers of the test face block provides a particularly effective mechanical approach for holding the test face block together even in situations where an adhesive would be less effective. For example, the test face block may comprise a cut-out extending through the test face element in a direction orthogonal to the plane of the laminated layers, and a further block may be configured to occupy the cut-out to prevent relative movement in the plane of the laminated layers, and wrap around the external faces of the test face element either side of the cut-out to prevent relative movement orthogonal to the plane of the laminated layers. A cut-out can be provided in a test face block having laminated layers with a common cross section, and therefore is particularly convenient in a laminated core. In some examples, the test face block and a further block may be configured to intersect at right angles to provide an electrically conductive core having a cruciform cross-section in a plane parallel to a plane of the electrical coil. A cruciform core may provide a particularly effective core shape. A cruciform core can fill the gap between at least one magnet and the electrical coil more effectively than a prism may be able to, whilst allowing the test face block to be constructed from layers having a uniform cross section, and supporting a further block for restricting relative movement of the layers in the test face block. In addition, blocks intersecting at right angles may provide a particularly effective structure in which the further block can act to restrict relative movement of layers within the test face block, and may provide greater resistance to relative movement of the layers than examples in which the further block does not intersect at right angles. As discussed above, in some examples the at least one further block of the core may not be laminated. For example, the at least one further block may be made of a solid piece of an electrically conductive material. However, in other examples the at least one further block has a laminated structure having laminations in a plane which is not parallel with a plane comprising the laminated layers of the test face block. Including laminations in at least one further block can further dissipate ultrasonic vibrations generated in the core. However, the at least one further block may be more distanced from the electrical coil and hence the direction of the laminations may not be as important for reducing eddy currents as for the test face block. Hence, the laminations may not be parallel to the laminations in the test face block. For example, the laminations may be perpendicular to those in the test face block, as this can increase the ability of the further block to hold the laminated layers of the test face block together. In some examples, the electrically conductive core may comprise an electrically conductive material having an average grain size greater than 8pm. Grains may for example represent regions of an electrically conductive material such as a metal. The inventors have identified a correlation between the average grain size in the material used to provide the electrically conductive core, and the presence of noise in the measured signal. In particular, the inventors have noted a reduction in unwanted echoes in a measured signal when using larger grain sizes for the electrically conductive material. Hence, by appropriately selecting the material grain size, the electrically conductive core can be provided with a structure configured to reduce intensity of ultrasonic vibrations within the electrically conductive core. For example, for each of the frequencies in the range 2MHz to 5MHz, an average grain size of greater than 8pm may provide a performance improvement of greater than 1.5dB, where a performance improvement represents the intensity of the signal compared to the level of noise in the signal. In some examples, the electrically conductive material may have an average grain size greater than 12pm. The inventors have determined that an average grain size greater than 12pm can enable a performance improvement of 6dB to be achieved for at least one frequency in the range 2MHz to 5MHz, with other frequencies in that range also experiencing a smaller performance improvement, where a performance improvement of 6dB (representing a signal of double the noise floor) may be desirable to enable effective signal processing of a measured signal. In some examples, the electrically conductive core may be tapered so that an area of a magnet face of the electrically conductive core adjacent to the at least one magnet is greater than an area of a coil face (test face) of the electrically conductive core adjacent to the electrical coil. A tapered core can act to focus magnetic field from the at least one magnet to the test surface, and therefore may increase the sensitivity of the EMAT. In some examples, the EMAT may comprise a protective cover on an opposing side of the electrical coil to the at least one magnet. A protective cover, or wear plate, can protect the electrical coil from damage. The surface of the EMAT for facing the test object may be the surface on the opposing side of the electrical coil to the at least one magnet, and hence the wear plate may be provided on this surface. Examples will now be discussed with reference to the Figures. Figure 1 is a schematic diagram illustrating an electromagnetic acoustic transducer (EMAT) in place on the surface of a test object 10 when the EMAT is in use for ultrasonic nondestructive testing. It will be appreciated that the test object 10 itself is not part of the EMAT. The EMAT comprises at least one magnet 2, and an electrical coil 6. The at least one magnet may be arranged such that the magnetic field is generated in a direction substantially orthogonal to the surface of the test object 10 and a plane of the coil 6. There may be a gap between the magnet 2 and the coil 6 to reduce noise generated within the magnet due to eddy currents generated in the magnet 2 by the coil 6. The EMAT therefore also comprises an electrically conductive core 4 disposed between the magnet and coil, for conveying magnetic field from the magnet to the coil, and to the test object 10. A wear plate 8 is also provided between the coil 6 and the test object 10. The coil 6 may have a generally planar shape to increase the proportion of the coil 6 nearby to the test object 10, and therefore increase sensitivity of the EMAT. The coil 6 may be connected to driving circuitry 12. Driving circuitry 12 can provide a varying electrical current in the coil 6. The current in the coil 6 generates a varying magnetic field, which induces closed loop eddy currents in a surface region of the test object 10. The eddy currents flow in the conductive test object parallel to the coil 6. Moving charge carriers in the eddy currents in the test object 10 move in the magnetic field generated by the magnet 2, and hence experience a Lorentz force. The Lorentz force causes the charge carriers to move, and thereby cause ultrasonic vibrations to be generated in the test object 10. Similarly, vibrations in the test object (for example reflections of ultrasound waves generated by the EMAT) can also be measured by the EMAT. Vibrations in the test object 10 cause charge carriers in the test object (e.g., free electrons) to move in the magnetic field generated by the magnet 2. These charge carriers experience a Lorentz force, and hence form surface currents in the test object 10. The surface currents formed in the test object generate a varying magnetic field which induces currents in the coil 6. The currents in the coil may be measured by receiving circuitry 14. An EMAT may comprise both driving circuitry 12 and receiving circuitry 14, although in certain configurations EMATs may act to either generate a signal or measure a signal, and hence may provide only one of driving circuitry 12 or receiving circuitry 14. The strength of vibrations induced in the test object, and the sensitivity of the EMAT to vibrations in the test object are both related to the strength of the magnetic field at the surface of the test object, because the Lorentz force experienced by charge carriers is linearly proportional to the external magnetic field strength. Providing a core can enable the magnetic field strength at the test object 10 to be increased because it may be made of a material which guides the magnetic field from the magnet 2 more effectively than the air gap which would be present otherwise. The performance of the core 4 is related to its permeability, with cores having a higher permeability being more effective. The core is typically made from a non-conductive material, to avoid the formation of eddy currents in the core. Eddy currents in the core induced by the coil 6 would cause ultrasonic vibrations to be induced in the core 4 as well as the test object 10. The ultrasonic vibrations in the core 4 would act as a source of noise. The noise generated in the core by the coil 6 is coherent noise because it is generated in the same way as the signals which are measured from the test object, and therefore can be very difficult to isolate from the desired signal. However, in examples of the present technique, an electrically conductive core 4 is provided. The use of an electrically conductive core would typically be counter-intuitive for the reasons given above, but the inventors have determined that by appropriately controlling the structure of the core, a conductive material can be used whilst reducing the drawbacks of coherent noise which would typically be associated with a conductive core. Figure 2A is a schematic illustration of a laminated core, which may be an example of an electrically conductive core having a structure for reducing intensity of ultrasonic vibrations. The laminated core comprises a number of layers of an electrically conductive material separated and joined together by an electrically insulating adhesive. In this way, the laminated sheets of the core are electrically isolated from each other. Eddy currents cannot therefore form across layers and are limited to individual layers of the laminated structure, where the structure of the individual thin layers may not support eddy currents. The laminated structure may therefore not support significant eddy currents in the core of the EMAT, which can enable an EMAT to use electrically conductive materials whilst suppressing the generation of coherent noise. Figure 2A includes an arrow illustrating a direction orthogonal to the plane of the laminated layers of the core. When viewed along this direction, the laminated layers may have a common cross section (square in the example of Figure 2A, but irregular in other examples as shown in Figure 3A). This common cross section can simplify manufacture of laminated cores. Figure 2B illustrates a portion of a coil 6 shown on the surface of a test object 10. The direction of the induced currents in the test object may generally align with the coil tracks, which may have a generally linear direction in a test region, for example as a straight portion of a racetrack coil. An arrow is shown overlaid on the coil tracks to illustrate a direction parallel to the plane of the sheets of a laminated core, as shown in Figure 2C. Figure 2C illustrates a laminated core in position between a magnet 2 and an electrical coil 6 in an EMAT placed on a test object 10. As shown in Figure 2C, the direction orthogonal to the plane of the sheets (the arrow shown in Figure 2A) lies in the plane of the coil, or in other words the laminated sheets are stood on their ends on the coil. This means that it is only the edges of the laminated layers which lie on the coil, reducing the distances of electrically conductive material in the region in which eddy currents could be generated (compared to if the layers lay flat on the coil). In addition, the vector orthogonal to the plane of the sheets (the arrow of Figure 2A) is parallel to the direction in which current is carried in the coil, meaning that the laminated layers cut across the coil track. Whilst an orientation in which the sheets are parallel with the coil track has been found to reduce noise in the core, the orientation in which the sheets cut across the coil track further reduces the generation of eddy currents in the core 4. Figures 3A to 3C illustrate a multi-block core. Figure 3A illustrates a test face block 20, which is the part of the multi-block core comprising the test face region 24 which is the region closest to the electrical coil 6, and hence the region in which the majority of eddy currents will be generated. The test face region, from the surface of the test face block to the skin depth, may be entirely within the test face block. The test face block 20 may be entirely laminated. By providing a laminated block, formation of eddy currents in the test face region 24 is suppressed by limiting the size of electrically conductive regions in the test face region. In addition, the interfaces between laminated layers also act to attenuate any ultrasonic vibrations which are generated. As discussed above, manufacturing limitations may mean that it is desirable to provide a test face block having laminations with a common cross section, as shown in Figure 3A. However, the requirement for a constant cross section could diminish the benefit of using a conductive core, due to poor magnetic optimization. A single part core may for example not provide a shape which effectively guides the magnetic field to the test object. However, this problem can be overcome by using a multi-part core as shown in Figures 3A to 3C. As illustrated, the test face block 20 may be combined with at least one further block 22 (Figure 3B) to form the core 4 (Figure 3C). The overall core does not have to have a common cross section, and hence can take a shape which provides more effective guidance of the magnetic field from the magnet 2 to the test object 10, even when including a test face block 20 made from laminated sheets. The multi-part core provides further benefits. An interface is formed between the test face block 20 and the further block 22. This interface can act to further scatter ultrasonic vibrations induced in the core, in a similar way to the interfaces between laminated layers, for example if the interface includes an air gap or a material having a different speed of sound to the test face and further blocks. Furthermore, a multi-part laminated core 4 can have improved mechanical properties when compared to a single part laminated core. A laminated core may encounter problems when used in certain environments. For example, high temperature may decrease the effectiveness of the adhesive provided between the conductive layers in the laminated core. In a single part core, this could lead to the core losing integrity and falling apart. This can be a significant problem in an EMAT, which may be commonly used in high temperature environments, for example when used for testing of an oil pipe. However, the inventors have realised that providing an interlocking multi-part core can address this drawback associated with laminated cores, because the interlocking shape can mechanically hold the core together, such that the core retains integrity even if the adhesive loses effectiveness. Figures 3A to 3C illustrate an example interlocking design, in which the test face block 20 and further block 22 each contain a cut-out into which the other block fits. Each block 20, 22 prevents lateral movement of the layers of the other block by occupying the cut-out. In addition, the further block 22 reaches around the layers of the laminated test face block 20 and hence also prevents the layers from separating along the direction orthogonal to the plane of the laminated layers. In this way, the further block mechanically holds the laminated layers of the test face block 20 together, even if they are not held by an adhesive. The further block 22 may not be laminated, and in some examples may be made from a solid piece of electrically conductive material for example. However, in some examples, the further block 22 may also be laminated. Whilst these laminations may not reduce the formation of eddy currents, because the further block 22 may not be provided in a region susceptible to eddy currents, the laminations in the further block 22 may further attenuate ultrasonic vibrations in the core by providing a series of interfaces. The direction of the laminations of the further block 22 is not as important as the direction of the laminations in the test face block 20 because, as mentioned, the laminations may not significantly reduce formation of eddy currents. The laminations may therefore be non-parallel with the laminations in the test face block 20., The further block 20 may, as with the test face block 22, have laminations having a common crosssection when viewed along the direction orthogonal to the plane of the laminations as shown in Figure 3B. This can enable simplified manufacture. The multi-part block can have a structure which also mechanically holds the further block 22 together when the further block 22 is laminated. The core 4 shown in Figure 3C also has a tapered shape, in which a face closest to the magnet 2 (the top face as shown in Figure 3C) has a larger area than a face closest to the coil 6 (the bottom face as shown in Figure 3C). This tapered shape acts to focus the magnetic field from the magnet onto a test region of the test object 10. The tapered shape also contributes to geometric attenuation of ultrasonic vibrations in the core. Figure 4 is a graph comparing attenuation of ultrasonic vibrations in a 2D simulation of two cores having different structures. A solid core is compared against an interlocking laminated core. The cross section of both cores has the same shape, with the solid core being made of a solid piece of electrically conductive material while the laminated core (shown on the bottom row of Figure 4) comprises laminated layers. The laminated core can be seen as a crosssection of the core shown in Figures 3A to 3C. As can be seen in Figure 4, energy is dissipated more effectively in the interlocking laminated core compared to a core having a solid structure, despite both cores being made from an electrically conductive material. Hence, controlling the structure of the core can significantly reduce the intensity of ultrasonic vibrations in the core, and therefore can enable an electrically conductive core to be used whilst suppressing the generation of coherent noise in an EMAT. Figures 5A to 5C show a series of graphs showing amplitude over time of a signal measured by EMATs having different cores. Figure 5A shows a signal measured by an EMAT having a non-conductive soft magnetic core formed as a 10x10x10mm cube of Mix63 material. Figure 5B shows a signal measured by an EMAT having a laminated conductive core formed from a single 10x10x10mm cube of laminated sheets of electrically conductive material separated by insulating material. Figure 5C shows a signal measured by an EMAT having the laminated conductive core of Figure 3C, in which the width and depth are 10x10mm at the top of the core, and 5x6mm at the bottom. In all examples the magnet was a 10x10x10mm cube. All graphs show a pulse excited in the test object 10 at time 0, and the resulting echo measured at around 25 us, representing the response of the test object to the initial pulse. As can be seen by comparing Figures 5A and 5B, providing an electrically conductive core increases the amplitude of the measured signal (shown at around 25us). In the example of Figure 5, the peak amplitude of the signal is increased by around 2 times (6 dB). This increase in signal strength is due to the increased magnetic permeability of the electrically conductive core compared to the non-conductive core used previously. Comparing Figures 5A and 5B also shows a decrease in noise in the laminated conductive core. In an ideal case, there should be no signal between around 10 and 20 us, and signal shown in this region can be attributed to noise. As can be seen, providing a laminated core reduces noise compared even to a non-conductive core. As can be seen by comparing Figures 5B and 5C, providing a multi-part core having an optimized shape further increases the amplitude of the measured signal. This can be attributed to the increased guidance of magnetic field onto the test object by the core having an improved shape, which therefore increases magnetic field strength at the test object. In addition, a comparison of Figures 5B and 5C shows a further decrease in noise. This further decrease in noise can be attributed to the geometric attenuation provided by the tapered shape, and to the attenuation provided by the interface between blocks in the multi-block core. Figure 6 is a graph illustrating the signal-to-noise ratio (SNR) of an EMAT having a core as shown in Figure 3C over an extended period of time in a test environment. The environment provides a temperature of 100C above the maximum temperature of the adhesive as provided by the adhesive datasheet, which may be representative of some of the environments in which an EMAT may find use. Figure 6 illustrates that initially the SNR begins to decrease. The decrease in SNR can be attributed to degradation of the adhesive between layers leading to the layers beginning to separate. The separation of layers provides gaps in the core, reducing the ability of the core to effectively convey magnetic field from the magnet to the test object, and hence decreases magnetic field strength at the test object. The reduced magnetic field strength at the test object reduces the magnitude of induced vibrations and reduces the sensitivity of the EMAT to vibrations in the test object, and hence reduces a signal strength detected by the EMAT. The reduction in signal strength leads to a decrease in SNR. Figure 6 shows that the SNR degrades until a certain level, at which point it flattens out and thereafter remains stable. The SNR stops decreasing at the point where the interlocking blocks of the core prevent the layers from separating any further. Figure 6 therefore illustrates the mechanical stability provided to the core by the interlocking design. Figure 7 is a graph illustrating the effectiveness of cores having different values of magnetic permeability in a test arrangement representative of an EMAT. The horizontal axis of Figure 7 shows different relative permeability values, and the vertical axis of Figure 7 shows the amount magnetic flux density conveyed to the test object by a core having that permeability, quantified as the surface flux density normalized against the highest permeability assessed, and also quantified as the expected difference in performance compared against the highest permeability assessed. Figure 7 shows that performance increases rapidly between relative permeability of around 0 and 100, whilst performance improvements taper off at higher permeability. Figure 7 illustrates that a core having a relative permeability of 70 is expected to provide 95% of the magnetic flux density of a core having a relative permeability of 1000. Replacing a non-conductive core with a conductive core may increase permeability from below 70 (e.g., below 50) to above 70 (e.g., above 100) and therefore due to the steep curve shown in Figure 7 can be associated with significant improvements in performance. Figure 8 illustrates a method of using an EMAT to excite ultrasonic vibrations in a test object and measure a resulting signal from the test object. At step 1000 the driving circuitry 12 generates a driving signal comprising a time-varying electrical current. At step 1002 the driving circuitry provides the time-varying electrical current to the electrical coil 6, causing a magnetic field to be generated in the vicinity of the coil and therefore an electrical current to be induced in the surface of the test object 10. An external magnetic field is generated by a magnet 2 and conveyed to the surface of the test object 10 via an electrically conductive core 4. The magnetic field at the surface of the test object 10 interacts with the moving charge carriers in the electric currents induced in the surface of the test object 10, and causes those charge carriers to move via the Lorentz force, inducing vibrations in the test object. The vibrations induced in the test object 10 travel away from the surface and reflect off structures within the test object, such as an opposite edge of the test object or imperfections within the test object. The reflected ultrasonic vibrations return to the surface of the test object and, via their interaction with the external magnetic field created by the magnet 2, induce current in the coil 6. At step 1004 the current induced in the coil by ultrasonic vibrations in the test object are measured by receiving circuitry 14. The measured signal can be processed to determine features of the test object, such as a distance between opposing walls indicating a thickness of the test object (e.g., the wall thickness of a tested pipe) or the presence of imperfections in or on the test object (e.g., via dispersion of the induced ultrasonic vibrations). Figure 9 is a graph illustrating the effect on noise in an EMAT attributable to varying the thickness of the laminated layers in a laminated electrically conductive core, such as the core shown in Figure 2A. The performance loss may for example represent a level of noise in the signal compared to a baseline. Figure 9 illustrates performance loss for a range of laminate thicknesses for a selection of test frequencies (frequency of the signal in the electrical coil for inducing vibrations in the test object, and frequency of vibrations in the test object) in the range 2MHz to 5MHz, which may be representative of frequencies most likely to be used in practical implementations of the EMAT. Figure 9 illustrates the general trend that decreasing laminate thickness reduces performance loss, and hence illustrates that it may be desirable for laminate thickness to be decreased in general. This trend is due to the reduced amount of eddy currents which may be supported for thinner laminated layers at a given frequency. Figure 9 also illustrates that performance loss is generally higher for higher frequencies at a given laminate thickness. As the frequency increases, the size of eddy current loops induced in the electrically conductive core decreases, and hence for a given thickness more current loops may be supported for higher frequencies. At a certain minimum thickness for each frequency, the performance loss reaches a minimum value when the thickness of the laminate layers becomes too small to support eddy current loops at that frequency. A maximum acceptable performance loss due to noise, to enable effective signal processing to extract information about the test object, may in some examples be 12dB. As can be seen in Figure 9, for a maximum performance loss of 12dB the largest laminate thickness in the range 2MHz to 5MHz is 1mm, and hence 1mm may represent the largest viable laminate thickness in some examples. Smaller laminate thicknesses improve performance for all frequencies. In some examples, to enable all frequencies in the range 2MHz to 5MHz to have a maximum performance loss less than 12dB, the maximum laminate thickness may be 0.4mm. In some examples, to enable performance loss to be less than 6dB for at least one frequency the maximum laminate thickness may be 0.7mm, and for performance loss to be less than 6dB for all frequencies in the range the maximum laminate thickness may be 0.3mm. Figure 10 is a graph illustrating the effect on noise in an EMAT attributable to varying a grain size for an electrically conductive material, based on measurements of a test core. The performance improvement represents a signal intensity compared to an intensity of the noise in the signal, and hence a larger performance improvement is preferable to a lower performance improvement. The grain size was measured using a procedure which provides a benchmark method which takes account of aspects of statistical variability and microstructural variability. A cast iron electrically conductive material was used to provide data for Figure 10, in which a grain size was measured according to the following procedure: 1. Prepare specimen and etch with Nital to give appropriate contrast with reflected light microscopy. 2. Select a magnification that (a) clearly resolves the microstructure, and (b) enables the capture of a micrograph that will contain at least 200 grains. 3. Capture an image and use a series of horizontal and vertical lines to count boundaries. No grain, or feature (other than graphite flakes), should be sampled more than once with a vertical line or more than once with a horizontal line. 4. Move to a ‘random’ location 1 mm away and then repeat step 3. Carry out steps 3 and 4 four times, or for large grains, until more than 200 intersections have been counted in each direction (i.e. horizontal and vertical). 5. The grain size is given by the total line length divided by the number of intersections (i.e. the mean linear intercept value). 6. A ‘grain boundary’ for the purposes of this specific procedure includes boundaries between ferrite, graphite flakes, phosphide eutectic constituents, pearlite colonies and prior austenite grains. It will be appreciated that different approaches could be adopted for measuring grain sizes. However, the overall trends illustrated by Figure 10 are present regardless of the technique used to quantify grain sizes. As shown in Figure 10, performance improvement is larger for larger grain sizes across all frequencies in the range of 2MHz to 5MHz. Figure 10 also shows that performance improvement is larger for higher frequencies at a given grain size. For all frequencies in the range 2MHz to 5MHz, a minimum performance improvement of 1,5dB can be achieved for average grain sizes larger than 8pm. In some examples, a minimum performance improvement of 6dB may be desirable to enable effective signal processing. A 6dB improvement may be achieved for at least one frequency in the 2MHz to 5MHz range for a 12pm average grain size, and hence in some examples a minimum average grain size of 12pm may be desired. In the present application, the words “configured to...” are used to mean that an element of an apparatus has a configuration able to carry out the defined operation. In this context, a “configuration” means an arrangement or manner of interconnection. “Configured to” does not imply that the apparatus element needs to be changed in any way in order to provide the defined operation. In the present application, lists of features preceded with the phrase “at least one of” mean that any one or more of those features can be provided either individually or in combination. For example, “at least one of: [A], [B] and [C]” encompasses any of the following options: A alone (without B or C), B alone (without A or C), C alone (without A or B), A and B in combination (without C), A and C in combination (without B), B and C in combination (without A), or A, B and C in combination. Although illustrative embodiments of the invention have been described in detail herein with reference to the accompanying drawings, it is to be understood that the invention is not 5 limited to those precise embodiments, and that various changes and modifications can be effected therein by one skilled in the art without departing from the scope of the invention as defined by the appended claims.

Claims

1. An electromagnetic acoustic transducer for exciting ultrasonic vibrations within a test object, said electromagnetic acoustic transducer comprising:at least one magnet configured to generate a magnetic field;an electrical coil configured to carry current for inducing electrical currents in the test object; andan electrically conductive core disposed between the at least one magnet and the electrical coil, for conveying the magnetic field from the at least one magnet to the electrical coil;wherein the electrically conductive core has a structure configured to reduce intensity of ultrasonic vibrations within the electrically conductive core.

2. The electromagnetic acoustic transducer according to claim 1, wherein in at least one direction the electrically conductive core has an electrical conductivity greater than 1x106 Q'1m'1.

3. The electromagnetic acoustic transducer according to any preceding claim, wherein the electrically conductive core has a relative permeability greater than or equal to 70.

4. The electromagnetic acoustic transducer according to any preceding claim, wherein the electrically conductive core comprises two or more electrically conductive portions defining, at their intersection, one or more interfaces within the electrically conductive core.

5. The electromagnetic acoustic transducer according to any preceding claim, wherein the electrically conductive core comprises a test face region closest to the electrical coil, the test face region comprising laminated layers of an electrically conductive material and an electrically insulating material.

6. The electromagnetic acoustic transducer according to claim 5, wherein a thickness of an electrically conductive layer is less than 1mm.

7. The electromagnetic acoustic transducer according to any of claims 5 to 6, wherein the electrical coil is substantially planar, and a direction orthogonal to a plane of the laminated layers lies in a plane of the electrical coil.

8. The electromagnetic acoustic transducer according to claim 7, wherein the electrical coil is configured to carry current in a substantially linear direction in a region of the electrical coil adjacent to the electrically conductive core, and the direction orthogonal to the plane of the laminated layers is parallel to said linear direction.

9. The electromagnetic acoustic transducer according to any of claims 5 to 8, wherein the electrically conductive core comprises a test face block forming the test face region, and at least one further block which combines with the test face block to form the electrically conductive core.

10. The electromagnetic acoustic transducer according to claim 9, wherein the laminated layers in the test face block are stacked along a first direction, and the laminated layers have a common cross section when viewed along the first direction.

11. The electromagnetic acoustic transducer according to any of claims 9 and 10, wherein the at least one further block is configured to have a shape which interlocks with the test face block to restrict relative movement of the laminated layers of the test face block.

12. The electromagnetic acoustic transducer according to any of claims 9 to 11, wherein the test face block and a further block are configured to intersect at right angles to provide an electrically conductive core having a cruciform cross-section in a plane parallel to a plane of the electrical coil.

13. The electromagnetic acoustic transducer according to any of claims 9 to 12 wherein the at least one further block has a laminated structure having laminations in a plane which is not parallel with a plane comprising the laminated layers of the test face block.

14. The electromagnetic acoustic transducer according to any preceding claim, wherein the electrically conductive core comprises an electrically conductive material having an average grain size greater than 8pm.

15. The electromagnetic acoustic transducer according to any preceding claim, wherein the electrically conductive core is tapered to have an area of a magnet face of the electrically conductive core adjacent to the at least one magnet which is greater than an area of a coil face of the electrically conductive core adjacent to the electrical coil.

16. The electromagnetic acoustic transducer according to any preceding claim, comprising a protective cover on an opposing side of the electrical coil to the at least one magnet.

17. The electromagnetic acoustic transducer according to any preceding claim, wherein a structure of the electrically conductive core is configured to reduce presence of coherent noise in a signal detected based on the ultrasonic vibrations excited within the test object.

18. A method for reducing noise in a signal detected using an electromagnetic acoustic transducer for exciting ultrasonic vibrations within a test object, comprising:generating a magnetic field with at least one magnet; andgenerating an electrical current in an electrical coil, to induce electrical currents in the 5 test object;wherein the magnetic field is conveyed from the at least one magnet to the electrical coil using an electrically conductive core disposed between the at least one magnet and the electrical coil; andthe electrically conductive core has a structure configured to reduce intensity of10 ultrasonic vibrations within the electrically conductive core.23

Images