| NDT.net - October 2002, Vol. 7 No.10 |
Phased array technology for ultrasonic NDE was, for many years, restricted to highly specialised applications such as those in the nuclear field. The equipment was bulky and expensive, and the resolution left room for improvement, which limited application to relatively thick walled components.
Nevertheless, the flexibility inherent to the phased array technology is also very attractive for mainstream applications, for example vessels, piping and tanks in the (petro) chemical industry, girth and longitudinal weld inspection in transport pipelines and civil structures.
Recent advances in electronics, ultrasonic probe development and software have made this possible. Very compact solutions are now available, taking phased array into the realm of high volume, rapid and economically viable NDE.
This paper describes the equipment for automated girth weld inspection RTD has developed together with its partners. The design philosophy and the way in which the aspects technology, flexibility and reliability fit into it are explained. Finally, some field applications will be pictured.
Apart from medical applications (ultrasound-scan) phased array technology was (and is) used for specialised applications in ultrasonic NDE. Wider use was limited by several factors:
The reason why phased array technology now offers many more possibilities and flexibility is explained below . Some known applications in various fields are pictured. Also, it is discussed why phased array technology can be advantageous for ultrasonic NDE. A short overview of the principle of phased array is given, based on linear ultrasonic phased array technology.
RTD has more than 40 years of experience in pipeline girth weld inspection. It was used to develop and design a new system for automatic girth weld inspection. The new system, the Phased Array Rotoscan, is based on phased array technology and on the conventional Rotoscan system, which has proven its worth and reliability.
Experiences and test results, both in the laboratory and in the field, are discussed. Possibilities for other mainstream applications can be based on these experiences and test results.
3.1 Possibilities of phased arrays
Phased array technology was used earlier for specialised applications. The principles of
phased array apply to all wave lengths, acoustic waves, radio waves, ultrasonic waves etc.
Some known applications are:
Particularly, in the field of ultrasonic waves, the possibilities of phased array technology offer great potential. For example: in conventional automated ultrasonic NDE of pipeline girth welds, a new transducer set has to be composed in hardware for each new application (i.e. combination of wall thickness and weld geometry). In case of single element transducers, the purpose of this is optimisation of the probe system's main beam characteristics:
When phased array technology is used rather than fixed transducers, the beam characteristics can be controlled and thus optimised with one system and performed by one linear phased array transducer, without changing the hardware. When the linear phased array contains enough elements, configurations for multiple element transducers such as tandem transducers can be formed in this way. The flexibility offers operational benefits and reduction in preparation and inspection time.
3.2 The Principle of linear phased arrays
Basically, the principle of linear phased arrays can be explained by the Huygens principle.
According to this principle, each wave front can be formed by the addition of an infinite
number of individual point sources, shifted in space and/or phase. The elements of a linear
phased array can be seen as line sources This is a two dimensional variant of the Huygens
principle.
The positions of the line sources (elements) of linear arrays are fixed, but their phases can be controlled. The phase shift between the individual elements can be interpreted as a shift in agitation time. Relative to the first agitated element, the other sources are agitated with specified time delays.
The beam characteristics (or directivity field) such as angle and focal spot depend on time delays. The relationship between time delays, beam characteristics and array parameters (which are constants) can be called the ‘Focal Law’ (this word is often incorrectly used for the set of time delays corresponding to a particular directivity field). The simplest form of the focal law is for beam steering in one medium (see [1]). The focal law can be extended for beam focussing in one medium (see [1]) and beam focussing in a second medium taken Snell’s law into account (for example when a wedge is used).
The effect of the different array parameters (size, pitch, frequency, number of elements etc.) on the directivity field is discussed in more detail elsewhere [2]. The best results are obtained if the distance between the elements approaches zero and the number of elements is infinite. However, little improvement will be obtained when the number of elements is larger than 32 (see [2]).
An example of such an application is automated girth weld inspection. Pipeline girth welds are inspected with zone discrimination, see [3]. Hereby, the weld is divided into small zones of typically 2-3 mm in vertical height. Each zone is inspected with one optimised ultrasonic transducer or pair of transducers. The amplitude in a channel corresponding with a particular zone, indicates a possible defect in that zone. The whole weld bevel can be inspected on both sides of the weld, using two identical probe sets opposite each other.
For each zone on both sides of the weld a different transducer is needed in different configurations such as pulse echo, tandem and TOFD. The number of transducers required for the inspection of a weld can easily be more than twelve.
The inspection system is often calibrated on flat bottom holes, whereby each zone has a corresponding flat bottom hole. The echo caused by a flat bottom hole is set at a pre- determined value. Calibrating such a system is time consuming, because each transducer must be optimised, both its exact location with respect to the weld and its sensitivity.
Using linear phased arrays, the number of transducers can be drastically reduced. With one pair of linear phased array transducers, all configurations required can be made to inspect all zones from both sides. Instead of a separate transducer, a different set of delay times is used for each configuration. A normal pulse echo configuration will be obtained when a group of active elements is used for both transmitting and receiving. An active group of elements (typically up to 32) can be multiplexed along an entire array (with typically up to 128 elements) enabling the index point of the beam to be shifted. A tandem configuration is formed when the reflected beam from one active group is received by another active group elsewhere on the array. The receiving group may also partially overlap the sending group, thus eliminating limitations caused by physical crystal dimensions in conventional transducers.
The calibration of a phased array system can be done with software settings. The linear phased array is placed on a fixed distance from the centre line of the weld. The signals can be optimised without physically moving the transducer, which is the case in a conventional system. Preparing and calibrating the phased array system for the weld inspection, therefore, consumes less time than in the case of a conventional system.
5.1 Design philosophy
Fig 1: RTD conventional Rotoscan system. From left to right, scanner, computer with connector panel,
monitor and keyboard.
|
RTD has many years of field experience in automated ultrasonic girth weld inspection. The high demands on reliability and flexibility of the equipment are well known. The system must be robust and easy manageable for operators.
A very important aspect of the design philosophy is the software. The software must be compatible with the conventional system, including all the proven options for signal interpretation etc., and, in addition, must contain all the control functions for the phased array equipment. The software must also be easy to use.
5.2 The Phased Array Rotoscan system
A system for automatic girth weld inspection was developed by RTD and Technology Design,
based on the above mentioned design philosophy and phased array technology. The system
consists of the following parts (see figure 2):
Fig 2: Scheme of the RTD Phased Array Rotoscan system.
|
In the probe frame, two phased array and two TOFD probes are mounted. The scanner is a standard CRC scanner that can be mounted on a guiding band to keep the distance from the probes to the centre of the weld fixed.
One of the main inconveniences of the phased array is the large number of coaxial cables. Since each of the 128 elements must be individually connected to the electronics, an umbilical cable with the same number of coaxial leads is necessary. In the field, the umbilical cable must be at least 25m long. This makes the umbilical cable fragile and sensitive to noise. To avoid these problems, it was decided to mount the ultrasonic hardware on the scanner. The total length of the coaxial cables is thus reduced to 60 cm. This design required of course a lot of effort, because the electronics needed to be miniaturised, and heat and sealing problems had to be overcome. Nevertheless, the gain in reliability and noise / interference immunity proved that this was the correct approach.
The ultrasonic hardware on the scanner is connected to the remote computer with a standard network link. The ultrasonic hardware can be controlled by the remote computer and the relevant data can be displayed (see figure 3). The network cable is flexible, long, reliable and not noise sensitive.
The ultrasonic hardware on the scanner is placed in a specially designed waterproof housing. The entire box is about the size of a small shoebox. The phased array part of the software can address 2 x 64 channels, with 2 x 16 active elements. There are also temperature monitors and eight conventional channels for separate TOFD transducers or transducers to detect transverse weld defects. The weight of the scanner with the electronics box is the same as for the conventional scanner. Although of course the box increases the weight of the scanner, it is compensated by the reduced number of transducers in the frame.
Fig 3: Phased Array Rotoscan system. From left to right, scanner with TD focus scan, computer with
connector panel, monitor and keyboard.
|
Fig 4: (left) Control interface for the phased array. (right) Data display.
|
The software of the RTD Phased Array Rotoscan system is based on the software for the conventional Rotoscan system. The software is compatible with the conventional system. An optional button provides each channel with the possibility of phased array. The software automatically calculates default delay settings for the given weld preparation and number of zones. The signal caused by a flat bottom hole corresponding with a zone, during calibration, can be optimised by changing the index point, the angle and, where necessary, the focal point of the beam. A graphic presentation visualises the sound path from the elements to the focal point in a zone corresponding with the chosen channel (see figure 4 left). The data is displayed in the same way as in a conventional inspection ( figure 4 right).
5.3 The field trials
The equipment developed was put through extensive tests, designed to establish the capability
of the Phased Array method to achieve the quality standards required. Tests were also done to
find out if the system would be able to cope with the harsh environment in which the current
conventional equipment is used.
For the lab tests, a number of welds were scanned with conventional and with Phased Array equipment and the results were compared. The results clearly showed:
Fig 5: Field trial of the Phased Array Rotoscan system. 1) TD focus scan box 2) probe frame 3) phased
array probe on wedge.
|
The first field tests were done on a 42" pipeline (see figure 5). RTD Rotoscan crews had been active over the summer on this project, and the opportunity was given to re-inspect some of the welds with the new equipment. Already on the first day of field testing an inspection speed and quality level sufficient for normal operation was achieved. Furthermore, the operators were very pleased with the reliability of the equipment, as it proved stable, shock, mud and waterproof. Finally, the manageability of the equipment was very good, as it is more compact and easier to handle than the current Rotoscan and does not have a heavy umbilical cable.
When hardware, software and phased array probes become available off the shelf, Phased array technology becomes attractive for the inspection of parts with a simple geometry such as tank floors and welds. The flexibility of phased array technology can also be used for applications where complex geometries are inspected. These applications are characterised by the need of multiple specialised transducers. Examples of such applications are:
It is also expected that in the future computers will become even faster and more data can be stored. This offers more possibilities to combine and display the data (real time), such as:
The experience in the development and use of the Phased Array Rotoscan system illustrate that the full benefit of phased array technology can now be exploited for mainstream applications.
Comparable quality against comparable cost is of importance. Recent advances in hardware and software support this last aspect. As was the case with the Phased Array Rotoscan system, it is expected that experience will prove comparable quality for other mainstream applications as well. In addition to the benefit of flexibility, the potential benefits of data display and characterisation will also become more attractive.
| © NDT.net - info@ndt.net | |Top| |