T-Scan aids marine corrosion inspection
TWI Bulletin, March/April 1990
Graham Edwards BSC (Hons), DMS, MInstNDT, MITD is a Principal Research Engineer in the NDT Research Department. As well as leading the specialist site services team he also works on NDT of non-metals and in expert systems.
Use of a T-scan ultrasonic system for detection of corrosion in ship hulls ensures that isolated corrosion pits are not missed, and helps in diagnosis of corrosion problems, as Graham Edwards explains.
Recently, corrosion surveys have been undertaken with the T-scan system by TWI's Specialist Site Services team on the hulls of coastal vessels ( Figure 1). The hulls have suffered from corrosion on the inside surface near the keel due to the action of micro organisms in sediments accumulating beneath the engines. Access to view the corroded surfaces from inside the ship is limited, and inspection has to be carried out from the outside while the tug is in dry dock. Inspection takes about two days during which several metres of plate can be scanned.
Fig. 1. Vessel examined for internal corrosion beneath the engines by using the T-scan system on the outside of the hull
Corrosion
Corrosion of metal components costs industry billions of pounds annually, despite the precautions taken to select corrosion-resistant materials and to control the environment in which metals are used. Since it is often impossible to avoid corrosion, efforts must be made to monitor its progress so that corroded parts can be replaced before they fail.
Unfortunately, corrosion often occurs on surfaces which are not accessible for visual inspection and it becomes necessary to employ non-destructive testing (NDT) methods. All five principal NDT methods - ultrasonics, radiography, magnetics, eddy currents and penetrants - can be used in detecting corrosion. As part of its Specialist Site Services, TWI is able to offer advanced NDT techniques for detecting and monitoring corrosion.
NDT methods are most frequently used in the power-generating and petrochemical industries. Detection of corrosion in condenser tubes has led to development of the most advanced eddy-current testing systems. Ultrasonic thickness surveys over the surfaces of oil storage tanks and fractionating columns are a major business for NDT service companies. Recently, magnetic-stray-flux methods have been introduced for rapidly scanning the floors of storage tanks. Even gamma-radiography has been used in procedures for detecting corrosion on the inside surfaces of pipe risers in oil platforms.
Use of NDT in the detection of corrosion on aircraft is often vital. The effects of corrosion beneath wing skins and at fasteners can be catastrophic; as well as regular visual inspections, specialised radiographic, eddy-current and ultrasonic techniques have been developed to detect corrosion in inaccessible areas.
Detection of corrosion in the marine environment has received a great deal of attention since the installation of steel oil platforms in the North Sea. Here, there were two problems which made the application of conventional NDT techniques unsatisfactory.
The first was the need to carry out inspections underwater, which led to development of simple-to-use underwater digital ultrasonic thickness gauges. The second was the detection of deep, isolated corrosion pits in large surface areas of the structures, feasible only by computerised ultrasonic systems such as the T(thickness)-scan described in this article. T-scan is able to image the corroded surface and thus detect isolated corrosion pits that could all too easily be missed when taking individual thickness measurements with a digital ultrasonic thickness gauge.
Ultrasonic detection
The principle behind the use of ultrasonics for detection of corrosion on the inside surface of a pipe, for example, is as follows. Short pulses (3-5 cycles long) of high-frequency sound (between 2 and 5MHz) are fired by a piezoelectric transducer along a perpendicular beam from the outer to the inner surface of the testpiece ( Figure 2). The reflected pulse from the backwall is received, reconverted to an electrical pulse, and the time of flight between transmission and reception measured by the ultrasonic test instrument. With a knowledge of the velocity of ultrasound through the material (5950 m/s for low carbon steel), the distance between the outer and inner surfaces can be calculated. The ultrasonic pulse-echo technique therefore provides a method of wall-thickness measurement when only one surface is accessible. It is perhaps the earliest of all applications for ultrasonic NDT.
Fig. 2. Pulse-echo ultrasonics
In earlier ultrasonic instruments, the electrical pulses were displayed in an A-scan presentation. The time base of the A-scan would be calibrated by obtaining echoes through sections of known wall thickness of the test material; the operator would then read off the thickness of the testpiece by noting the position of its backwall echoes on the calibrated time base.
Digital electronics made reading the thickness much easier. In a digital ultrasonic thickness gauge, a quartz oscillator provides high-frequency oscillations which are counted between the transmission of the pulse and the reception of the echo. The time of flight can be displayed directly as a thickness value, provided that the correct velocity constant for the test material has been entered into the instrument.
The drawbacks
Despite their ease of operation and high resolution (0.lmm under ideal test conditions), digital ultrasonic thickness gauges suffer from a number of drawbacks.
First, the ultrasonic thickness gauge measures time of flight of the ultrasonic pulse to the backwall and not physical thickness. The velocity of ultrasound through the material must therefore be accurately known. Ideally, the gauge should be calibrated on a step wedge made of exactly the same material as the testpiece.
Second, on normal test surfaces, a twin transducer probe has to be used to improve coupling. Twin probes have an inherent inaccuracy, because the pulses of ultrasound are not propagating along a beam perpendicular to the surface but at a slight angle of pitch between transmitter and receiver. The error becomes significant on thin sections.
Third, and most important in the context of marine corrosion, the ultrasonic thickness gauge may fail to reveal the full extent of a corrosion pit, because the tip of the pit fails to reflect enough ultrasound to register on the gauge ( Figure 3). The consequence is that corrosion may penetrate the full skin thickness without being detected.
Fig. 3. Use of ultrasonic thickness gauge over a corrosion pit
Ultrasonic imaging
The T-scan computerised ultrasonic data recording and acquisition system has been developed to help overcome these problems. The instrument itself is fully computerised and includes menu-driven software to calibrate and control the test, a liquid crystal display (LCD) to show the T-scan images and A-scans in real time, and twin diskettes for storing test data. It is uncommon among computerised instruments in being highly portable and capable of being used in the most arduous test conditions.
The ultrasonic probe is held by a magnetic-wheeled scanner, which guides the probe in a raster fashion across the test surface. The scanner can be either tracked along a prescribed route following a magnetically attached strip or controlled remotely. The magnetic wheels are capable of holding the scanner underneath horizontal steel surfaces.
The site equipment ( Figure 4) comprises the instrument, scanner, scanner control box and water irrigation supply for acoustic coupling. Although the T-scan images can be viewed during the test on the instrument's LCD, the analysis is performed by a personal computer. The T-scans are displayed in colour graphics (reproduced here in two colours, Figure 5) and various operations can be carried out on them using the system's software.
Fig. 4. T-Scan instrumentation on site
Fig. 5. Result of a T-scan (originally in full colour)
T-scan images
Unlike the conventional C-scans used to display information in an ultrasonic immersion, the T-scan image is made up of information taken from the time of flight only and not from the amplitude of the echo reflected from the backwall. It thus provides an image of the metal loss from the area scanned. Each pixel in the T-scan image contains information from one or more data points in the scan. Each data point can be regarded as a position encoded by the scanner, where a pulse has been fired by the ultrasonic probe and its time of flight recorded; at a repetition rate of several hundred pulses per second the process is continuous.
The colours in the T-scan image are graded according to percentage wall thickness; the step size of the colour levels can be made as small as 0.5% of wall thickness. Such high resolution is generally unattainable to a satisfactory accuracy with digital ultrasonic thickness gauges because they can only take discrete readings - which are subject to errors owing to probe coupling and test surface conditions. The T-scan on the other hand plots variations in thickness, and although they too are subject to probe-coupling effects, these can usually be distinguished in the image.
The T-scan images provide a great deal more information about a corroded internal surface than discrete ultrasonic thickness measurements taken on a grid whose lines, for reasons of efficiency on large surfaces, are usually set at intervals of not less than 25mm. Isolated pits may be missed altogether without continuous scanning. Moreover, pits are more readily distinguished from general wall thinning by reason of their concentric coloured contours in the T-scan image.
Another important advantage of the system is that up to four T-scan images can be obtained from one scan. These may be from different probes, or from different gated positions of the A-scan (important in immersion testing), or by incorporating more subtle changes in the test parameters. For example, the thickness measurements may be taken from either the peak or the leading edge of the reflected pulse ( Figure 6). The position along the pulse at which the measurements are taken is obviously very important. Half a cycle at 5MHz is equivalent to 0.59mm in steel. In some applications it may be necessary to measure a floating level beneath the peak but beyond the leading edge ( Figure 6). In a test where the specimen is immersed in water and the probe is set off from the surface by a few centimetres, the thickness measurements need to be taken between the surface echo and the backwall echo. The water path is variable and this must be accommodated in the T-scan images.
Fig. 6. Measurement positions on the ultrasonic pulse
Application
The T-scan provides an image of the corroded surface that can be compared directly with what is observed visually. In the survey mentioned at the start of this article, the spatial distribution of corrosion pits in the hulls of coastal vessels and their relationship to features such as bulkheads can be observed which helps in diagnosis of the corrosion problem. Moreover, scans can be repeated at intervals and the images compared to monitor the progress of the corrosion.
The results of the inspection have enabled the vessel owners and operators to quantify the extent of corrosion in the susceptible regions of the hulls, providing confidence in the safety of the vessels to continue in service, or allowing repairs to be scheduled.
Thanks are due to Allan Clarke at Ocean Fleets Ltd for permission to publish this article.
