For a large structure (a bridge, offshore platform, pipeline or storage tank) the structural integrity and safety of operation are paramount. However, structures are always subjected to some dynamic loading (such as heavy traffic, wind or water waves) that may cause fatigue and cut short their technological life span. It is therefore important to establish the integrity during service, and preferably without interrupting its serviceability.
The research and technology dedicated to this task belong to the field of Non-Destructive Testing (NDT), which some prefer to call Non-Destructive Evaluation (NDE).
The inspection focuses on vital points in the structure where fatigue is most likely to occur. These points may be determined through a visual inspection or on the basis of the experience of a structural engineer. The vital points are usually welded joints where steel structural components, such as plates or pipes, connect. There, the dynamic loading stresses may cause fatigue of the metal which can result in a slowly growing crack. It is of vital importance to detect and characterize cracks in an early stage of growth, so they can be fixed before causing serious problems. An example of the possible consequences of fatigue is shown in the figure below, where the dynamic loading was imposed in a laboratory.
Our inspection technique is based on scattering of waves. The basic idea is that information on the state of a structural component can be obtained by analysing its wave scattering properties. To this end the component is irradiated, after which the scattered wave field is measured. Retrieving the desired geometrical crack information from the experimental data requires that wave generation, wave propagation and the wave scattering process at a crack be modelled in accordance with the experimental conditions.
In principle, a wave field may be mechanical (waves in fluids and solids), electromagnetic (electrical currents, radar, x-rays) or quantum mechanical (particle beams). Each type of wave field has its advantages for inspection and a choice should be made on the basis of the required precautions for the inspection technique, its sensitivity for the relevant material and/or geometrical parameters, and on the ease of operation. We investigate a technique that is based on elastic wave scattering and uses waves with high frequencies, for which reason it is commonly referred to as Ultrasonic Inspection(UI). This inspection technique is safe, both for the person who operates it and for structures involving inflammable materials, such as storage tanks or oil/gas platforms.
Elastic waves are quite sensitive both to the geometry and to the material of the test object they propagate in. Fortunately, the wave speeds vary little between different types of steel, but the wave damping properties may vary substantially and is dependent on the composition and microstructure of the steel. Uncontrollable are physical processes such as temperature dependence and wear, which may influence the geometry and the surface conditions of the test object. Especially the surface conditions are important, since the wave transducer must maintain a good acoustical contact with the object, which is usually achieved by employing a liquid couplant between transducer and test object. In addition, noise and measurement errors are unavoidable facts of experimental reality. All these complications necessitate research to develop our knowledge of waves and to improve our understanding of the influence of the parameters involved. This research is intended to innovate our technology to manipulate wave fields so that new, more robust, and more widely applicable inspection techniques become feasible.
Development of MTOFT: an improved long-range imaging technique suitable for single-sided inspection
At the section of Structural Mechanics of TU Delft we have developed the two-dimensional (2D) "multi-path time of flight technique" (MTOFT) for the characterization of both surface breaking and buried cracks in steel plates and welded connections of steel plates. The data may be acquired in a single-sided scan using either pulse-echo or pitch-catch measurements. The data may be acquired from distances of up to 1 meter from the defect. Considering the wave lengths used (~1 mm at 3 MHz ), this may be called a long-range application. Although the use of very small wavelengths (thus high frequencies) would be preferable from the viewpoint of sensitivity and resolution, practically the grain scattering and dissipation of steel limit the highest usable frequencies to ~5 MHz.
Since MTOFT utilizes only the arrival times of the waves it is called a "time of flight" technique. For this technique it is required that the plates are sufficiently thick, so the received data will consist of individual pulses from which the arrival times can be easily determined. From a practical point of view, the plate should be thicker than 5 times the maximum wavelength, which in our case amounts to a minimum thickness of ~10 mm. How thin the plate may actually be taken in practice before one enters the region of plate waves (or modes of propagation instead of bulk waves) is left to be determined experimentally. The wavefield and arrival times in a thick plate may be effectively modelled by a ray-tracing approach. Various tests have been performed on straight plates and a T-connection of straight plates. These tests have demonstrated the excellent performance of MTOFT (resolution in the order of millimetres) and its potential for practical long-range inspection. More detailed results of this research may be found in the thesis "Non-destructive inspection techniques based on elastodynamics: development and validation" by M.C.M. Bakker (TU Delft,11 April 2000).
Comparison of MTOFT imaging with TOFD and SAFT
Standard TOFD (Time Of Flight Diffraction) techniques utilize the directly diffracted waves to locate the tip of a crack. The multiple wave paths are therefore neglected. This necessitates that the transmitter and receiver are as close to the test zone as practicable, which makes TOFD a local technique. Since MTOFT utilizes all occurring wave paths (also the multiples) it may be applied at larger distances from the defect. Another difference is that MTOFT may employ all types of waves (including mode-converted ones) for a characterization. Transmitting and receiving with TOFD is commonly done in a double-sided inspection, i.e. the receiver and transmitter are located at different sides of the test zone. Unfortunately, in many practical situations the test zone can be inspected only from one side. For this task both MTOFT and SAFT (Synthetic Aperture Focusing Technique) are better suited.
For SAFT one first chooses a small spatial region where the crack is thought to be located to keep the computational effort to an acceptable level. For MTOFT this region may be much larger since the computational effort is much smaller. SAFT requires a large amount of data as its principle relies on statistical averaging. This principle makes SAFT very robust. However, data acquisition and processing can be very time consuming. In contrary, MTOFT requires only a limited amount of data for a characterization. Typically, SAFT requires two measurements per wavelength, while MTOFT has shown to yield good results with only one measurement every 5 wavelengths. For an optimal performance of the SAFT technique one first has to eliminate propagation effects from the data, such as geometrical attenuation and influences of reflections. This is not required for MTOFT that relies solely on arrival times. This makes MTOFT more easily adaptable for inspection of different objects with a complex geometry.