NDE involves the use of various non-invasive measurement disciplines, such as Radiography (RT), Thermography (IRT), Shearography (OT) and Ultrasonics (UT) to determine the integrity of a component, structure or material without destroying the usefulness of the item. For the oil and gas industry these disciplines have been empirically derived from years of experience in inspecting more traditional homogenous fabrication materials such as steel and its alloys. A search of the available standards revealed that there are no specific standards available that focus on all of the objectives of this project. The results of this project will help to address this issue. BS EN ISO 14692; 2002 is the most up to date and leading standard regarding the use of 'Glass-reinforced plastic piping in the petroleum and natural gas industries'. Of the advanced techniques for development within this project only conventional ultrasonics and radiography are mentioned. Within the standard the preferred method of NDE is a visual inspection (often viewed from outside with a light source inside), obviously this is not an effective method for installed or in-service piping due to the intrusive nature of the routine that would require plant shut down and disassembly. Consequently, the inspection techniques available are not optimised for the inspection of defects typically found in GRP Pipe components. The limited NDE techniques that have been developed to locate cracks, porosity, and volumous disbonds using the above-mentioned NDE disciplines contain a number of major technical limitations that have stopped them from being used on a large commercial basis in industry. The area of most concern is the extremely low probability of detecting the kissing bonds (KB's) case. Currently, there are no NDE techniques that can reliably detect KBs

Technical Limitations of Existing Technologies

Owing to the lack of strategic research dedicated to the inspection of GRP components and manual inspection routines, there is a considerable deficiency in industrial standards and validated techniques issued for NDE of GRP components. Recent studies have concluded that the current level of NDE performance is inadequate, and go on to recommend development of NDE techniques and equipment. Recent PANI trials, carried out to assess the effectiveness of manual inspections have shown that operators detect only 50% of defects. Despite all the development to date, NDE engineers are restricted to manual technology that is unrepeatable and subjective, being prohibited from employing new techniques that are quantifiable, robust, repeatable and reproducible until research provides evidence of solutions, to the problems. Several other industry reports confirm this point of view.

Radiography is more suited to the detection of larger flaws only and relies upon the variation in the level of interaction between the X-rays and material under inspection. For voids with a large dimension in the direction of propagation of the X-ray beam the variation in attenuation between the GRP and air makes an easily detectable signal. Unfortunately, for KBs or narrow voids (unbonds) there is not sufficient difference in opaqueness for the unbond to be reliably detected. This is due to the fact that the absorption coefficients of the glass fibre, resin matrix and the adhesive are very similar. Image contrast is very low unless low energy sources are used Current systems use film-based RT thus requiring increased exposure times for inspection and increasing the risk of operator exposure. Furthermore film based systems have high costs due to the fact that X-ray films have to be developed, washed, fixed, dried and then viewed on an illuminated screen and manually interpreted involving high labour and material costs. The written procedure for the X-ray NDE inspection may be 200 pages long, and 100 or more RT exposures may be required due to the low dynamic range of conventional analogue film techniques. This level of inspection can add more than €5,000 to the cost of a single component inspection. Digital equipment is currently available but has to date failed to meet the resolution levels achievable with film. Moreover, in the case of testing GRP and other plastic materials, the low energy of the radiation used produces digital images with low contrast. Consequently, there are no standards available for the application of such technology.

Shearography is subjective with the stressing or heat transfer levels un-quantified. Portable vacuum units capable of operation by one technician have such a small field of operation as to negate the potential speed of coverage advantage. Additionally portable hoods do not easily conform to small radii surfaces as found in GRP pipe applications. Currently available mature systems are only reliably capable of monitoring strains that are 'out of plane'- that is normal to the surface of the test component. 'In plane' systems have significant limitations including collimation of the beam and the test object must also be flat and viewed directly along the camera axis. Departures from these conditions introduce error sources within the instrument by distorting the fringe patterns. Mathematical analysis tells us that it is the second differential of out of plane displacement (or the first differential of the in-plane displacement), which allows a map of the structure's strain field to be generated. However, any attempt to use these fringes directly to determine the strain is limited by noise and the sensitivity constraint particularly on curved surfaces.



The defect quite clearly (encircled in red) the thinner section cools more quickly than the surround thicker sections, inducing high stress levels (greater concentration of fringes) in the surrounding area.

Thermography is also subjective with the stressing or heat transfer levels un-quantified and the size to depth aspect ratio in the region of unity. This means that the technique is limited to detecting defects of the same diameter as their depth in the material. Also the heat absorptivity, reflectivity and transmissibility coefficients are difficult to assess. The glass fibre material is particularly porous and the air packets enclosed act as an ultrasonic reflector as well as a thermal insulator. In practical terms this means that the introduction of heat into bondline and internal areas, as required by active IRT is particularly difficult. Whereby passive IRT systems are of little or no benefit for the inspection of GRP pipes, as there are no internal heat generation or degradation sources. Lamb waves are ultrasonic waves that can propagate over tens of metres and are therefore recommended for global long range NDE in metallic structures. Lamb waves are commonly generated as multimode waves, which create problems in analysing after acquisition. However, the received data has the advantage of a broad spectrum of information in one scan, due to specific sensitivity of each mode to various discontinuities. Currently this technology has only been proven on metallic structures.









Ultrasonic NDE methods cannot easily detect defects due to the porous nature of GRP fabrications. Ultrasound is reflected at interfaces of high acoustic impedance mismatch, which is the case between GRP and air boundaries. The ultra high level of porosity, typically found in GRP lay ups presents thousands of reflectors that tend to scatter the ultrasound and prevent propagation of the sound wave. The currently undetectable KB's problem that is prevalent in GRP joints is a particular challenge as the whole philosophy of UT inspection is based on the detection of voids or volumous gaps in the material. Additionally, the trade off between minute fault sizes and large area scanning provides a conflict for UT as a large area UT transducer cannot characterise a small defect and a small transducer cannot quickly inspect a large area. The use of liquid as a couplant also adds considerable cost to an inspection especially for porous materials where the couplant can be absorbed and cause internal damage over a period of time. To date the use of probes that do not require coupling is restricted to medium frequencies and/or single element probes.







Other NDE disciplines are only capable of detecting flaws in traditional materials. One other technique that has shown promise is the Acoustic Emission methodology that uses acoustic probes to monitor and analyse the sound emitted when a fault is influenced an external stressing medium of sufficient energy to induce crack or defect growth. This discipline has been excluded from this project as the practice of deteriorating the health of the structure under test goes against the philosophy of Non-Destructive Testing.

Application of the above techniques by manual methods This is one of the most severe limitations of current NDE techniques used for GRP inspection. Manual methods are slow, inaccurate, unrepeatable and open to operator subjectivity thus reducing confidence levels of the inspection. Additionally, most automation currently available tends to be of a Cartesian flat bed type X-Y scanner with only limited motion in the Z-axis. Modern manufacturing processes facilitate design optimisation that generally leads to complex geometry components, making traditional scanners redundant. The need to automate the inspection has long been recognised as an important step on grounds of safety and economics.