Lovejoy, D.J. (2004), "Non-destructive testing of engineering composite materials and structures", Aircraft Engineering and Aerospace Technology, Vol. 76 No. 3. https://doi.org/10.1108/aeat.2004.12776caf.001
Emerald Group Publishing Limited
Copyright © 2004, Emerald Group Publishing Limited
Non-destructive testing of engineering composite materials and structures
Non-destructive testing of engineering composite materials and structures
Keywords: Nondestructive testing, Engineering, Materials
While composite materials may be considered as novel it must be remembered that modern composites have been in use for decades and that the oldest engineering materials of all, namely wood, stone, bone and a variety of vegetable and animal materials, belong to this class of materials. Strictly speaking steels are also composite materials; however, this paper is restricted to the combinations of materials which comprise of honeycomb sandwiches and laminates.
An exciting aspect of the non-destructive testing of these materials and structures made from them is that consideration of how best to test them is focussed very strongly on the artefacts made from them rather than the individual methods which are used to carry out the tests.
Obvious parallels exist between the approach and programmes for testing composites and the more conventional metallic items in engineering. The origins of defects follow the familiar pattern of the intrinsic defects which arise in the preparation of the materials, processing of the materials whereby parts are joined or holes drilled are processing defects, and defects arise during the service of the parts. There is also a further parallel in the varying needs of testing newly manufactured parts and those which have been in service. Despite these parallels many types of defects which are detected in composites are quite different from those which are familiar in metallic components.
The defects which can be detected in composites
Many of the defects which are detected in composite materials and structures are specific to the class of material under investigation. It is well to remember that most metallic materials are fairly isotropic with their properties being largely similar in all directions through the material. The variations in grain direction which are the result of cold working and other processes do affect the proper application of a programme of non-destructive testing, but these are minor considerations when compared with the very marked anisotropy which is the characteristic of most engineering composites. The major defects which need to be detected in composites are:
unbonds, disbonds or voids in the interface between the outer skin and the adhesive, unbonds, disbonds, and voids in the interface between the core material and the adhesive, voids in foam adhesives, unbonds or disbonds between the adhesive and a closure member at the core-closure member, and fluid ingress.
The defects which are frequently detected in laminates are:
porosity in the laminates or bond lines,
incorrect state of the resin cure,
incorrect overall fibre fraction,
misaligned or broken fibres,
non-uniform fibre distribution leading to matrix rich regions,
gaps, overlaps, and other faults in the arrangement of plies,
pores and voids in matrix regions,
unbonded or disbonded interlaminar regions,
resin cracks or transverse ply cracks,
unbonds in thermoplastic composite materials due to the failure of separate flows to re-weld during moulding,
mechanical damage around machined holes, and
local bond failure in adhesively bonded materials 1.
A further consideration which has a parallel with the non-destructive testing of conventional materials is the question “What constitutes a defective state?” The standard definition of “Some aspect of the workpiece which renders it unfit for its intended purpose”. Still serves well.
Non-destructive testing methods
Of the six widely used NDT methods, which are familiar from the testing of metallic items, only magnetic particle testing has no application to the currently available engineering composites. The remaining five, eddy current, penetrant, radiography, ultrasonic testing, and visual inspection are all available to provide data on the state of composites. The specific requirements of the individual materials and structures must be considered to obtain the optimum information to assess the status of the item. Among the newer technologies which are used are acoustic emission, holography and other optical methods, mechanical impedance measurements and vibrational techniques, and thermographic methods. The order in which the various methods are discussed does not reflect any relative significance and they are dealt with alphabetically.
The principle of acoustic is the application of a small stress of a representative pattern to a workpiece and by the use of sensitive transducers to monitor the stress waves which are generated by the energy released by small failure events within the material. When appropriate instrumentation is used to analyse the collected data it is possible to obtain information as to the location, severity, and sometimes, even the nature of the events and the discontinuities associated with them. Even with the sophistication of modem electronics, it can be difficult to interpret all the data from testing essentially isotropic materials and it is no surprise that the problem increases when anisotropic materials such as composites present further problems. The fact that composites do not exhibit the Kaiser effect is familiar to anyone who has a creaking wooden boars in a staircase since it continues to creak after repeated loadings. A new Felicity effect has been introduced to describe the behaviour of wood and other composites.
Much of the application of acoustic emission to testing composites has been in condition monitoring and material evaluation. Laboratory studies continue to find ways of extending the attractive features of this technology.
Eddy current testing
It is self evident that eddy current testing can only be applied to electrically conductive materials. This implies that its use on non-metallic composite workpieces has been on carbon fibre reinforced plastics which have a volume fraction of carbon fibre of at least 40 per cent. There are however, a number of electrically conductive organic polymers including doped N-polyacetylene, polyphenelenevinylene, and polyparaphenylene which have dielectric conductivities comparable with those of copper and silver at 300 K (23°C). The principle of eddy current testing is well established: when a time varying electromagnetic field approaches an electrically conductive material eddy currents are induced in the material. The field induced in the material under test interacts with the primary exciting field and this interaction is analysed and detected by changes in the impedance of a detecting coil. Signals generated indicate the state of the material. The form of the signals which are obtained provides a great deal of information. When eddy current testing is applied to metallic materials the frequency of the applied field dictates the depth of penetration of the field and limits the region of inspection. The resistivity of carbon reinforced non-metallic materials is such that this phenomenon is much reduced. Eddy current testing is, however, very sensitive to the volume fraction of the carbon fibres and this fact must be considered when the method is used. Eddy current testing has been applied successfully to the inspection of carbon fibre reinforced composite workpieces for the location and quantification of inhomogeneities such as defects, the evaluation of volume fraction in unidirectional laminates, the lay up order of cross plied laminates, and the detection and measurement of broken fibres. When metallic fibres are used in organic polymeric matrices the eddy current method of testing has many attractions.
Mechanical impedance and vibrational methods
These methods use the stiffness of the materials and are conveniently divided into two groups. Global techniques whereby the entire workpiece is excited either by a pulse or in a continuous manner. The natural frequencies and damping are measured and correlated to the state of the material. Local methods use excitation of a small area of the workpiece and measure its mechanical response. These methods are normally applied from one surface and have advantage when access to both surfaces is impractical. Unlike ultrasonic methods, they do not require a couplant between the probe and the test surface. In principle, they rely on the differential response of properly bonded and unbonded or disbonded material and other material variations to vibrational energy. They have proven to be very successful for detecting a wide range of defective conditions including unbond or disbond between thin composite skins, foams, and honeycomb structures. They also allow detection of broken or corroded core. In practice, vibrational methods are applied by through transmission, pulse echo, ringing, or damping methods. This group of test methods is currently very widely used in the non- destructive testing of honeycomb structures.
It is reasonable to believe that the original method of inspection by human eye is as old as mankind. It is clear that earlier humans travelled considerable distances from their homes to obtain specific materials for making tools and weapons; it is unreasonable to think that they did not look at the raw material and the finished artefact. Similarly, with the newest classes of material we still find optical methods of inspection extremely helpful. The original method of visual inspection has been developed with magnification and other optical phenomena including Moire fringe methods, holographic interferometry speckle methods, and shearography have been added to the range of these methods.
The long established value of visual inspection has been with the human race for as long as artefacts have been made. In the manufacturing of all engineering items there remains very a good reason to inspect visually with the unaided human eye and with magnification, with endoscopes, and more recently with imaging systems. In the application of this method to engineering composites this activity can detect a wide range of surface flaws and give some indication that further investigations should be carried out when, for example, areas of porosity or resin starvation are seen at the surface. Rapid visual inspection of the edges of laminates can indicate the existence of edge delaminations. Visual inspection has been extended to laser scanning where an intense monochromatic beam of light scans the surface and the intensity of the reflected light is changed where surface breaking defects exist.
Moire patterns can be created on a surface in two ways. They can be created by the use of gratings or by the projection of alternate light and dark in space above the surface. In appearance, they form parallel and crossed lines. Two methods are followed for using the phenomenon. Shadow Moire involves the projection of the pattern on to a surface and the fringes used to analyse the surface form. Moire fringes produced by the use of gratings are applied in the study of surface strain. The application of this effect in the NDT of composite materials has been more in the development of materials and the design of workpieces, so far this phenomenon has not been widely applied in production.
In principle, holographic interferometry offers an extremely attractive inspection tool; by shining a light on to the surface of an object it is possible to characterise the total form; and by implication, any changes in it including discontinuities. Theoretically, it is possible to detect surface displacement, strain, stress, to observe modes of vibration, detect and quantify cracks, and to diagnose subsurface phenomena through their effects at the surface. In practice, the method has been found to be successful in the semi- quantitative assessment of cracks, unbonds, disbonds, and other flaws. The application to the quantitative strain and stress measurement has been less successful up to date. The principle of the method is the comparison of two holographic images of the surface of the workpiece, one taken at rest and the other under a mild strain. Superimposition of the two images gives rise to fringes and defects stand out as bullseye type irregularities in the interference pattern.
When coherent light is shone on to a diffusely reflective surface it appears to be covered with myriads of bright and dark points giving the appearance of a roughened bright surface. Two distinct causes, both of which involve interference, contribute to this effect. The effect is real and is not dependent on the observer and can be captured photographically. Light captured by a lens is restricted by the aperture and a random speckle pattern appears which defends on the lens. If the light is collimated the speckle pattern ceases to be random and a regular pattern depends on the aperture of the lens. The speckle pattern is a function of both illuminated surface and lens system in use. In applications for NDT, two speckle patterns are compared and the results are analysed.
Shearography involves a form of speckle shearing interferometry, two coherent images of a surface are formed simultaneously with one shifted with respect to the other and they are recorded. A second similar recording is made while the item is under a slight strain. The two sets of images are superimposed and fringes appear. Since the fringes show the contours of the derivative of the displacement rather than the displacement itself the method is much less sensitive to external vibration than is conventional holographic interferometry since the speckle forming rays for any locality in the image are arriving from two different localities on the object. Applications of shearography to the inspection of composite workpieces are essentially similar to those of holographic interferometry with the advantage that external vibration has much less influence on the measurements.
The use of conventional liquid penetrant testing of composite materials is very restricted. The process where a liquid containing conventionally coloured or fluorescent dyes are applied to the surface, the surface excess removed, and some form of developer applied followed by visual inspection must first be controlled to ensure that the penetrant chemicals are compatible. Even modern penetrant materials which are not based on hydrocarbon distillated must be checked in this way before being applied to composites. Where the compatibility is established the method can be applied to the edges of laminates to detect edge delamination however, sometimes staining is persistent and unacceptable. Paniculate fluorescent penetrants have been developed which overcome both staining and compatibility problems and their use may become established. One application where the application of fluorescent penetrant testing is established as a method of choice for a composite is the inspection of dense ceramic items for surface breaking defects.
Thermographic methods of NDT may be passive whereby an external source of heating or cooling is applied or they may be active in which case events within the material provide the energy needed for thermal changes. In either case, the surface temperature of the item under test is monitored for the evidence of defective states. The active mode of thermography has been used in special investigations such as fatigue studies of composites where the energy of fretting provides an internal heat source for the analysis of the differential rates of cooling between defective and sound material.
Many different heat sources have been used in the passive mode depending on the size and form of the area under investigation and the material. This method has been used successfully for the routine checking of helicopter blades for bond integrity between the skin and spar and between the skin and the honeycomb. It is also a method of choice for detecting fluid ingress in honeycomb structures.
Ultrasonic inspection is particularly well suited to the detection of defects which are normal to the direction of the interrogating beam. Ultrasonic techniques are very widely used in the non- destructive testing of composite materials. Many techniques have been developed using frequencies in the range from 1 to SO MHz. These have proven to be very successful in detecting unbonds and disbonds, for determining void content, and for detailed characterisation of defects. The applications include manual use of ultrasonic thickness gauges through to various levels of sophistication to fully computer controlled multi- axis systems.
All applications of ultrasonic testing require some form of couplant which is designed to maximise the transfer of the ultrasound energy between the transducer and the material under test. The efficiency of the couplant is a function of the acoustic properties of the materials involved. A water/carbon reinforced composite interface transmits between 70 and 75 per cent of the incident energy. It is therefore, convenient to immerse such a material and the transducer in water. When an inspection requires manual scanning a gel form couplant is used in order to achieve a similar energy transfer. The carbon reinforced composite/air interface transmits only 0.03 per cent of the incident energy and reflects the rest. Air filled delaminations and any other air filled discontinuities reflect most of the incident energy and produce strong signals.
While manual inspections are sometimes unavoidable immersion systems are preferred for a number of reasons. The couplant is essentially constant and the transducer beams can be collimated by seating the transducers on manipulators, some distance from the surface of the workpiece. When the transducers are mounted on manipulators, they can scan in a prescribed plane relative to the surface of the workpiece. Information can be obtained about the item under test and presented as A, B, or C scan which displays point, cross section, and area information, respectively.
Jet or squirter probes offer considerable flexibility when compared with immersion systems. These use water as the couplant, but instead of immersing the workpiece the coupling water is sprayed on as a jet which also acts as a waveguide for the ultrasound. Structures which have complex shape can be inspected in this way with the help of shaped guide tubes. This technique is often applied in a through transmission mode at low frequencies (frequently below 10 MHz). The jet probe method, which is widely used in the inspection of airframe components, is identical to the immersion technique in principle.
After the discovery of X-rays by Roentgen and of gamma rays by Becquerel at the end of the 19th century, these phenomena became applied to industrial inspection very rapidly, indeed. Industrial radiography was well established during the 1914-1918 war when it was applied to the detection of cracks in one of the oldest engineering composites during the inspection of wooden aircraft propellers.
Despite the relatively poor absorption of X-radiation by non-metallic composite materials the method is very useful when low voltages are used, typically SO kV and below. Excellent results can be obtained when the voltages are as low as 12 kV. During such low voltages excellent pictures of the arrangement of fibres within the material can be obtained. Similarly, fluid ingress into composites can be shown quite clearly.
The fact that radiography is very sensitive to the orientation of the radiation beam with respect to that of the discontinuities sought is well known. This characteristic leads to a great difficulty in detecting defects such as delamination or disbond. This difficulty can be resolved by the use of an X-ray opaque liquid penetrant before exposing the item to the radiation beam. This technique is referred to as penetrant enhanced X-radiography (PEXR). While this approach has been very successful in indicating the extent and nature of impact damage on airframe composite panels it is not as widely used as earlier. There are a number of reasons for this. Many of the penetrant liquids used which offer very attractive technical characteristics have unattractive properties from the points of view of health and safety and the environment. Most are volatile organic halides and need to be handled with great care to protect operators and some are listed as ozone depleting chemicals. Fortunately, solutions of inorganic halides such as zinc iodide in an aqueous solution of isopropanol with a little wetting agent have been found to be effective. These however, lack the attraction of ready volatility. A further disadvantage of using radio opaque penetrants is that they are very invasive and can extend well beyond the area of the impact damage thus giving a false impression of the extent of the defective area. X-radiography of carbon fibre reinforced composites is used to determine the volume fraction of the fibres and fir checking fibre alignment and for defects in the fibres.
Stereo radiography involves taking two radiographs: one normal to the surface of the workpiece and the other at an angle of around 15°, which produces an apparent three-dimensional view of the item. This technique is reported to provide a detailed data of the through thickness of composites and localising indications of damage. By this method, it is possible to locate a single ply within a laminate.
The X-ray attenuation coefficients of the different materials in bonded composite to metal joints or adhesively bonded honeycomb structure vary in such a manner that very little information concerning the bond can be obtained by the application of conventional X-ray techniques. Neutron radiography techniques have been found to provide very clear pictures of both bond and composite in the presence of metals.
Computer aided tomography
A normal radiograph is a two-dimensional representation of a three-dimensional object. The density of the image at each point is governed by the attenuation experienced by that part of the X-ray beam incident on it. This provides information about the contribution, which each element of the object makes to the total attenuation. Computer aided tomography (CAT) measures the intensity of the X-ray transmission along many paths in a single plane and, by use of a computer, calculates the attenuation contributed by each element in the plane. In this way, a picture of the material can be built-up. This technique has been used on most engineering materials including composites to gain a full picture of their status.
Compton fluorescence tomography
The phenomenon of Compton scattering has been used to assess the status of carbon fibre reinforced composite materials with a reported spatial resolution of 2-3 mm. At a depth of 50 mm, energy from a radio isotope of less than 200 keV is collimated and a number of detectors are used to acquire scattered photo energy data which are presented in discrete slices. Since the scanning plane during data acquisition is in the x-y plane rather than the x-0 plane single side inspection is possible.
In this brief review of the methods of NDT which are available for the inspection of engineering composites, it has not been possible to discuss any one of them in great detail. Some interesting points do arise however, a number of the methods discussed remain largely or completely restricted to laboratory investigations while others, notable vibrational, sonic, and ultrasonic methods, holography, and thermography are in widespread use. Others such as the Russell effect and corona discharge methods have not been mentioned here as it seems unlikely that they will be applied outside laboratories for some time to come.
Clearly our profession's previous experiences are of both interest and importance in two areas. First, the familiar needs for an understanding of the nature of the material and how they are manufactured is apparent and secondly, as with the non-destructive inspection of metallic workpieces it is unlikely that any single NDT method will provide enough data for informed and useful decisions to be reached concerning fitness for the purpose of critical components.
D.J. LovejoyThe South West School of NDT, Cardiff, South Wales, UK