·Home ·Table of Contents ·General

Application of Filmless radiography during the production of large diameter pipes Th. Kersting, L. Oesterlein, Europipe Deutschland, Mülheim an der Ruhr (Germany); N. Schönartz, Mannesmann Forschungsinstitut, Duisburg (Germany) Contact

Summary

At the beginning of 1998, the first equipment for a filmless radiographic inspection of welds at linepipes with wall thicknesses up to 45 mm was installed and taken into operation in the Mülheim works of Europipe. The fully automatic facility operates under the control of the production data recording system (PRODIS) and replaces two conventional X-ray chambers thanks to its considerably reduced exposure times. Radiographic images are evaluated off-line at a high-resolution monitor with additional monitors also being used at the repair station and in the office of the responsible NDT-engineer. The images are archived on CD-R.

This equipment represents the first-time use of the filmless radiography with a sensitivity equivalent to that of a film in a fixed-cycle production line for welds at heavy-walled pipes governed by the most exacting safety requirements. It is planned to change the mill's entire radiographic testing to this technology within the next few years.

1.  Purpose

In the Mülheim mill, EUROPIPE produces longitudinally welded large-diameter pipes up to a length of 18.3 m, with diameters ranging from 24^" to 64^" and wall thicknesses between 8 mm and 45 mm. The required inspection procedures result in a number of up to 1000 radiographs per day. Each of these X-ray images has to be evaluated and recorded in corresponding lists; a large part of the films is presented to customers' representatives in connection with acceptance testing. A film preservation period of twelve years as required by the relevant guarantee regulations causes a considerable logistic expenditure for the procurement and timely availability of the films as well as high storage costs. From the ecological point of view, a problematic factor are the large quantities of chemicals needed for the film processing and their rather costly disposal.

The a.m. situation is simplified by the use of the filmless radiography. Images are automatically stored and managed on CD-R, customers' representatives may readily view the images on their own monitor, and both chemicals and film costs are completely eliminated.

So far, the mill had operated 5 X-ray chambers with 4 tubes each in order to handle the required amount of films. Through the use of filmless radiography in the internal test cycle, two of these chambers could already be replaced by one chamber with only one tube without causing any bottlenecks. The aim is to operate a total of 3 to 4 chambers for the volume of inspections needed.

Studies have proved that even experienced evaluators only detect approx. 70 % to 80 % of all defects. The use of a fully automatic image evaluation is intended to increase this percentage, so that a human interpretation should only be required in cases of doubt.

On a medium-term basis, EUROPIPE intends to completely replace the X-ray film technology by the filmless radiography.

2. Process of filmless radiography

Within the framework of a project promoted by the ECSC (contract no. 7215-EB/101) a pilot facility was used to demonstrate that the filmless radiography can replace the film without impairing the testing quality. The production facility described in the following represents another development of this pilot plant with a further improvement of test features. It is fully integrated into the production process.

Only the use of a high-resolution radioscopy system with 1024 x 1024 pixels allows a restriction of enlargement to values below 1.5, which is required to use an X-ray tube of 0.4 mm or 0.6 mm nominal focal spot size with a sufficient power to penetrate the entire production range. At the same time, the covered weld area is still large enough for inspections to be carried out within the individual cycle times.

In view of the given dimensions it is not possible to move the image converter within the pipe. Radiographic testing therefore has to be performed from the inside to the outside. Consequently, the image converter is suspended above the pipe from a vertically adjustable trolley travelling in longitudinal direction of the pipe. The X-ray tube moves to the individual test positions (fig. 1) in a cantilevered spar within the pipe, which is open on the upper side but supported on both sides during the test.

Fig 1: Filmless radiography at EUROPIPE in Mülheim

Fig 2: Computer configuration of filmless radiography

In the absence of a technically satisfactory solution for the problem of operating a conventional X-ray tube with high-voltage cable and coolant lines within a pipe of more than 18 m length, an air-cooled 250 kV-X-ray tube with integrated medium-frequency high-voltage generator and 0.6 mm focal spot was purpose-designed for this facility. The supply and control cables required for the operation of the tube are guided to the tube within the spar via a customary trailing cable device. The video signal supplied by the image converter is digitised by a plug-in frame grabber card and is integrated by the computer in real time by means of the corresponding software. The computer system is designed as redundant cluster of two high-speed Alpha workstations interfaced via SCSI-bus and Fast Ethernet. Archiving of images is done on CR-R by a special computer connected via Ethernet. This configuration guarantees both max. data safety and max. availability of the system (fig.2). Major technical data of the facility are listed in table 1.

X-ray tube
max. tube voltage	250 kV
max. power dissipation	600 W
focal spot size (nominal)	0.6 mm
Image converter
input screen diameter	220 mm
usable input window	95 mm x 170 mm
optics	high-resolution, high-intensity tandem optics
camera	HDTV-CCD-camera
Image processing
digitalisation	1024 x 1024 pixels
integration	by software, 25 images/s
functions	filter, measuring, histogram etc.
storage	redundant on 2 separate hard disks
Computer system
SCSI-cluster	2 x Digital AlphaStation 500/400
operating system	OpenVMS
user environment	OSF/MOTIF
Archiving
medium	CD-R
data system	ISO 9660
format of image files	TIFF (non-compressed)
capacity	> 500 images
Table 1: Technical data of filmless radiography

The equipment is located within the so-called U-cycle, i.e. before the mechanical expander. A certain amount of weld repairs is permissible in this area. After the expansion of the pipe, repairs are generally not allowed. Under the economic point of view, the reliability of NDT in the U-cycle is thus of considerable importance for the production process (fig. 3). The processes used have to anticipate the results of final inspection and testing. In order to be able to observe the required cycle times, the entire process has to be automated. The operator merely enters the production number of the pipes and monitors the smooth course of inspection.

Fig 3: Schematic presentation of integration into the U-cycle

3. Integration of the facility into the production process

If the automatic ultrasonic test produces any indications, these positions are directly transferred to the internal production data recording system, and the pipe moves on to the X-ray station. As soon as the pipe has reached its position in front of the chamber and the pipe number has been entered, PRODIS transmits all required information, in particular the X-ray positions, to the computer of the filmless radiography, which in turn transmits same to the PLC. If all data are in compliance with the specifications, the pipe is moved into the chamber. During the entry, the tube and the image converter already travel to the first position to be radiographed with a positioning accuracy of ± 2 mm each. The weld is centred below the image converter by means of a weld sensor, in order to ensure a vertical penetration.

As soon as image converter and tube have reached their desired positions, the X-ray inspection is started. The corresponding parameters have already been determined experimentally and have been stored in a parameter data file. The system imports these data and takes the exposures. Generally, energies between 100 keV and 220 keV as well as integration times of 5 s to 10 s (25 images/s) are used. The integrated images are stored on hard disk.

As soon as all exposures of a pipe have been taken, they are made available to the evaluator. After the entry of the pipe number in the PRODIS system, the evaluator receives all X-ray positions. He analyses the images successively by marking possible defects with a frame and entering the result in the PRODIS system. After the evaluation, the images are stored along with a parameter data file, which e.g. includes the enlargement factor and the location of the marking frame.

Depending on the results obtained, the images are made available to the personnel at the repair station, where they can be called up from a workplace terminal. The repair is performed and confirmed in the PRODIS system. The pipe is then automatically returned to the X-ray station to control the effectiveness of the repair. Not until the pipe has left the U-cycle are all images released for archiving, which is done on a routine basis once per shift. A special software compiles the images, converts the data into a generally readable format (TIFF) and burns a CD, which can be read by any customary computer with CD-drive. After a control of the proper storage, the images are erased from the hard disk.

4. Results

The facility was set up in January 1998 and has been used for the entire amount of X-ray inspections required in the U-cycle since the summer of 1998. As an internal test it is not subject to customer's specifications. Standard EN 13068 part 2 and the draft of standard prEN 13068, part 3, are taken as a reference basis for the minimum requirements applicable to the image quality. According to this standard, the system complies with the provisions of system class SC 1 (internal unsharpness U £ 0.4 mm) and the test class SB. The required sensitivity amounts to approx. 1 %, the spatial resolution to wire pair number 9 according to EN 462, part 5 (= 0.13 mm). In order to achieve these values, a geometric enlargement of V »  1.4 is required. Wire pair number 9 corresponds to the resolution of the film (Agfa D5, wall thickness b = 20 mm, FFD = 800 mm, t = 4 mm, fig. 4). Customers' specifications call for contrast sensitivities of ³ 1.5 %, so that the system meets all requirements. The observance of image sharpness and sensitivity is controlled and documented once per shift.

Fig 4: Spatial resolution of film (above) and filmless radiography (below)

Fig 5: Sample with slag inclusion. X-ray film (top), filmless radiography (bottom)

Fig 6: Sample with pores. X-ray film (top), filmless radiography (bottom)

The following radiographs (fig. 5 and 6) show some examples from production runs. The films are original exposures from conventional X-ray chambers. The films were digitised with a spatial resolution of 50 µm. Defect positions were subsequently subjected to another radiographic inspection with the image converter system. Apart from an adjustment of image size and contrast as well as the inversion of grey tone representation for the image converter exposures, the radiographs were not altered in any way.

Once the film evaluators have got accustomed to the new system, image processing does not pose any problems. The various possibilities such as contrast adjustment or negative/positive representation are used systematically. At present it can be stated that the introduction of the filmless radiography has not resulted in an increase of rejected pipes during final inspection, i.e. all defects which can be detected with customer's specified inspection process are already located before by the filmless radiography and can be repaired. Thus the system meets the high quality standards applied by EUROPIPE.

5. Outlook

Following the successful introduction of the filmless radiography in the U-cycle, EUROPIPE intends to abandon the future use of film radiography also during acceptance testing. The change has been planned for 2001. As yet, the technique has not been fully included in standards and specifications for large-diameter pipes. Reservations about the new system still expressed by some customers may, however, be invalidated by corresponding information and a proof of the system's effectiveness. Possible alternatives to the high-resolution image converter now being used are flat-panel detectors, which promise a more comfortable handling and larger inspection windows. Initial tests will still be carried out in the course of this year.

As with any human-based evaluation, the analysis of the X-ray images depends on the knowledge, experience and concentration of the evaluator. It is therefore difficult to maintain a constant level of quality. As mentioned above, one of the goals followed with the set-up is the automatic evaluation of the X-ray images which overcomes these limitations. A prototype version of a software package called AXION (Automatic X-ray Image EvaluatiON) has recently been installed in the mill. In the final version, this software shall analyse all possible defects within the image. According to a given level of confidence the image is either classified as defect-free or suspect. Only in case of doubt, i. e. if the required level of confidence is not reached, the final decision is left to the human evaluator. During the test period, the evaluators will check every decision made by the software. In the long run, however, this approach relieves the operator of routine work and enhances the probability of detection and the total quality of inspection.

© AIPnD , created by NDT.net

|Home|    |Top|