Search

Airborne ultrasound is currently used for a variety of applications, including distance ranging, level monitoring, broken ply detection and for the limited non-destructive testing of low Z materials (such as green-state ceramics, rubber and carbon fiber composite). Ideally we would like to be able to use airborne ultrasound for high quality non-destructive testing of high Z materials, as this would allow testing of parts without contamination by a couplant, and could potentially allow very rapid testing. The testing of most materials using airborne ultrasound is impeded by the high acoustic impedance mismatch between air and the test-piece, giving a very high reflection coefficient. This, (combined with other factors such as the high attenuation of ultrasound in air, and the slow velocity of ultrasound in air) make testing with airborne ultrasound very difficult.

Development of air-coupled capacitance ultrasonic transducers that operate above 100kHz have recently been of interest. The main applications for these devices include the location and imaging of objects, and in the nondestructive evaluation of engineering materials. To control the response of either as a source or a receiver, we needed to investigate the effect of using different backplate materials and manufacturing techniques. Backplates were produced in a range of materials such as silicon, copper and stainless steel. Various techniques of micromachining such as chemical etching, laser machining and ion beam machining were used.

A neural network is an artificial intelligence technique inspired by a simplistic model of biological neurons and their connectivity. A neural network has the ability to learn an input-output function without a priori knowledge of the relationship between them. Typically a neural network consists of layers of neurons, whereby each neuron in a given layer is fully connected to neurons in adjacent layers. Figure 1 shows such an arrangement with three layers, called the input, hidden and output layers. The connection strengths between neurons, often referred to as weights, are modified by a training phase. The training phase used here utilizes an error back propagation algorithm [1]. During training the neural network is presented with input which propagates through the network producing a corresponding output. A comparison of the actual output with the desired or target output generates an error which is used to adjust the neural network’s weights according to an error gradient descent technique [2]. This procedure is repeated for many different input and desired output pairs allowing the neural network to learn the input-output function.

The ultrasonic immersion technique, using a water coupling medium, is now a standard method for the non-destructive inspection of anisotropic composite materials. Using a goniometer to investigate a composite material in any direction of propagation, their anisotropic elastic l,2 or viscoelastic 3,4 properties can be determined. A comparative analysis of the through-transmission ultrasonic bulk wave methods was recently published by Y.C.Chu and S.I.Rokhlin 5.

Silicon backplate capacitance air-coupled transducers have particular advantages over other types of air coupled ultrasonic transducer. Such advantages include the ease of manufacture and the ability to reproduce field characteristics. This is achieved by using a polished silicon backplate and introducing controlled surface roughness by etching holes into the backplate to create a series of air springs. The performance of such devices has been investigated [1], and it was found that the combination of bandwidth and sensitivity leads to a range of NDE applications. The aim of the present paper is to compare the pressure field of these transducers to that predicted by theory for a plane piston transducer in air.

The testing of CANDU Zr-Nb pressure tubes for the presence of hydrides is a problem which is of current interest, as the premature replacement of pressure tubes could occur if the estimation of hydride levels is not made correctly. It has been shown that changes in the ultrasonic and thermal expansion properties of the material occur at a particular temperature, known as the terminal solid solubility (TSS) temperature [1]. Here, stable zirconium hydride platelets present at low temperatures dissociate to form free hydrogen on heating, and also reform from free hydrogen on cooling. Note that the TSS temperature is likely to be different for heating and cooling, due to hysteresis and thermodynamic effects, and will be dependent on the heating or cooling rate. The TSS temperature is known to be a function of the concentration of hydrogen present, and in fact follows a well known phase diagram [2], where the two-phase system becomes a single phase plus free hydrogen on heating (see Figure 1). The region of low atomic % is represented at the extreme left of this phase diagram. At a given value of atomic % of hydrogen (or the equivalent in parts per million (ppm)), a phase change will occur on heating from the two phase system containing the hydride to the single phase plus hydrogen. A transition in the reverse direction will occur on cooling. The temperature at which this happens is the TSS temperature, the value of which depends on the % of hydrogen present. Hence, determination of the TSS temperature leads directly to an estimation of hydride content.

It is often necessary to evaluate materials using non-contact ultrasonic techniques, for example when the test sample is hot, moving, or highly absorbent to conventional fluid couplants. Several non-contact methods are available, such as various optical techniques [1–3], which are generally expensive and require the sample to have optimized optical characteristics. Electro-magnetic acoustic transducers (EMATs) [4,5] and capacitance devices [6] may be used, but require an electrically conductive sample, and a small stand-off distance of a few millimeters or less. There has been much interest recently in the use of air-coupled transducers [7], which may be piezoelectric, using piezopolymers such as PVDF [8], piezocomposites of PZT and epoxy [9,10], or piezoceramics with impedance matching layers on the transducer face [11]. Another type of device is the electrostatic or capacitance transducer [12,13], which consists of a metallized polymer membrane against a backplate electrode to which a bias voltage is applied. Motion of the membrane causes the charge on the backplate to change, which may be detected using a suitable charge amplifier. These devices in general have a wider bandwidth than their piezoelectric counterparts, and improved sensitivity. The backplates are usually mechanically roughened metal, and it is therefore difficult to manufacture two identical devices. However, using a silicon backplate [14–17] and standard etching techniques, the surface of the backplate may be precisely controlled. Such a device is shown schematically in Figure 1. The backplate consists of a silicon wafer into which pits 40μm in diameter and 80μm apart have been anisotropically etched to a depth of approximately 40μm. A gold electrode is then evaporated onto the backplate, and a thin metallized polymer membrane is then placed next to the plate.

Laser generated ultrasound is being investigated [1,2] for testing structures made of both conventional metals and carbon fibre reinforced polymer (CFRP). Laser interferometers are widely used in such work to detect the normal surface motion caused by ultrasonic pulses. Interferometers offer non-contact, remote and high-fidelity detection, together with a potential to cover large areas rapidly by optical scanning. However their cost is high and only in testing large and/or expensive structures may the cost be justified. A lower cost alternative, but with some compromise on the virtues of an interferometer, would be to use an air transducer as a receiver.

The use of laser generated ultrasound for the pulse-echo testing of carbon fibre reinforced polymer (CFRP) aerospace components is being established [1,2]. Our particular interest is in the possibility of using ultrasound propagating from the laser source along the component, as bulk or plate waves, as a means of testing a larger area for a single step in a scan. As part of an investigation into how laser generated pulses propagate in CFRP we have examined their propagation in a 40 ply, 5.5mm thick (300x300mm area) unidirectional plate. The plate was made from carbon fibre reinforced epoxy resin pre-preg (Enka Tenax HTA fibre; ICI 7716H epoxy resin). Plies were cut out by hand, laid up and then cured in a compression mould.

Tomographic imaging is a technique to determine values of a spatially varying parameter on a cross-sectional slice through an object. In the ultrasonic case of tomography, waves are propagated through the sample between a series of source and receiver locations placed in a plane around the object. These locations are chosen such that the rays pass through as large a fraction of the object plane as possible, and to conform to any requirements for regular positioning in the reconstruction procedure. If the propagation delay is measured for each raypath between source and receiver, the acoustic slowness can be determined along the cross-sectional plane (slowness is the reciprocal of velocity). There are in existence a variety of algorithms that can construct tomographic images, such as filtered backprojection and algebraic reconstruction tomography (ART) [1,2].

There has been a lot of recent interest in the use of pulsed lasers for the generation of ultrasound in a range of media [1], due to advantages inherent in the generation process. These include the non-contact nature of the method, and the generation of a wide bandwidth over a small source area. The size and shape of the source can also be changed using suitable optics, as can the generation mechanism itself.

Tomographic imaging involves the collection of data from a series of projections through an object, and the reconstruction of variations in a chosen parameter across a two-dimensional section. It is now an established method for the reconstruction of images from ultrasonic data, and has been used widely in such areas as medical ultrasound and non-destructive evaluation [1–3].

Illumination of a specimen by a pulsed focused laser is known to induce elastic waves, and models are available for both cases of thermoelastic generation and liquid evaporation. In this article, the inverse problem of recovering the elastic constants of a specimen from experimental waveforms recorded on‐ and off‐epicenter is considered. An algorithm, based on nonlinear least‐squares fitting, is presented and has been used on experimental displacement signals recorded on an aluminum plate with a Michelson interferometer. The choice of the receiver position is discussed and experimental results are given.

Focusing transducers such as spherical bowls, annular arrays and planar transducers fitted with lenses have many uses in underwater acoustics, and various other areas involving acoustical imaging. In this paper, the characteristics of the first two types of transducer will be compared theoretically, with a comparison to experiment for the case of the spherical bowl transducer.

A series of theoretical programs have been developed for the prediction of the radiated pressure fields of transient radiators in the form of discs and annuli. Such transducers have been constructed using polyvinylidene difluoride (PVDF) polymer film, and their pressure fields sampled in two-dimensional planes following transient excitation. It is shown that good agreement is possible between theory and experiment for such transducers.

The present paper is concerned with the ultrasonic inspection of composite panels, fabricated by attaching a thin (

The metals manufacturing industry makes routine use of ultrasonic nondestructive testing techniques for the detection of defects such as cracking. It is becoming increasingly clear, however, that more advanced non-contact techniques are required for the testing of materials which are either in motion or at an elevated temperature. An example is in the continuous casting of aluminum, where differential cooling across the material can lead to stresses which, in turn, can result in porosity or cracks. In this case, an on-line defect detection system is desirable, to detect the onset of manufacturing problems. Hence, the aim is not to detect all defects, but to give an indication of problems in the solidification process.

Standard ultrasonic inspection schemes utilize piezoelectric ultrasonic transducers, which are applied either directly to the surface of the sample under test or used as part of an immersion technique. The latter may be inconvenient, whereas the former is dependent on a satisfactory coupling of the transducer with the specimen. These approaches are also inapplicable to inspection at elevated temperatures or in hostile environments, and to samples having surfaces which are rough or which have extensive coatings or oxide layers. The situation is further aggravated when reproducible, wide-band measurements are necessary, especially when a consistent transducer/specimen bond cannot be guaranteed.

The radiated fields from ultrasonic transducers have been investigated theoretically using many approaches, and both nearfields and farfields have been examined. Some treat the continuous-wave case1-3 , whereas others have examined pulsed behaviour4. The references cited investigate principally the uniformly vibrating disc, where the front face of the transducer is assumed to vibrate in phase, with a constant amplitude. The transducer is usually assumed to be mounted in an infinite baffle, a perfectly-reflecting barrier preventing modification of the field by waves radiated from the back surface. Various other theoretical approaches have been presented, which have treated variations in baffle configuration5,6.