| NDT.net - May 2002, Vol. 7 No.05 |
The roll out of a new phased array imaging system was immediately put to the test toward a nuclear power application during the Spring of 2001. The IntraSpectÔ UT Phased Array imaging system was brought on board late into an inspection qualification program as a backup capability. Its role was to provide an alternative inspection technique for flaw sizing in areas where success with conventional ultrasonics was originally deemed questionable.
The efforts to conduct the phased array imaging portion of the qualification and its first-use readiness are described herein.
The feedwater inlet nozzle on some boiling water reactor (BWR) vessels contain a stainless steel cladding over its carbon steel base material, the combined total being approximately 6" (15 cm) thick. Significant thermal stresses on the inner radius of such nozzles have in some cases caused in-service thermal fatigue cracking of the clad having the potential to propagate into the base material.
This problem was first defined in 1975 by the US Nuclear Regulatory Commission (NRC) and associated BWR suppliers. Two remedies were proposed, one being the removal of the cladding, and the other being a modification to the feedwater nozzle sparger or both. At this time, cracks were identified at a given BWR plant using penetrant and ultrasonic testing (PT & UT) and removed. Grind outs to remove the cracks resulted in divots in the cladding.
Following the repairs, the subject plant conducted manual UT examinations in accordance with NUREG 0619 in the period of time between 1980 and 1995. The plant then implemented automated UT examinations from the nozzle ID in period from 1995 to 2001.
In preparation for the inspection in 2001, efforts were further explored and qualified to detect and size indications through the combination of conventional and phased array automated UT examinations from the vessel OD. This split duty approach was partially driven by a narrow window of time to fully qualify the relatively new phased array technology for all inspection areas.
The use of phased array technology for this application was one of the aspects recognized and developed through the BWR Owners’ Group. Prior to the 2001 outage, the plant issued a letter to the NRC identifying their intent to implement an "Alternative Feedwater Nozzle Inspection" in accordance with the approved guidelines of the BWR Owners’ Group Report GE-NE-523-A71-0594-A, Revision 1.
A mockup fabricated many years ago and owned by the plant, was sent to the Electric Power Research Institute (EPRI) for used in the technique development and qualification effort. EPRI then modified some reflectors with the permission of The plant had most of the reflectors modified, resulting in the full-scale mockup with notches and replicas of grind outs. The overall mockup is shown in Figure 1.
Fig 1:
|
Westinghouse’s NDE Products and WesDyne groups teamed with EPRI and the customer and an industry expert institute to accomplish the qualification efforts at the NDE Center.
The EPRI NDE Center spreadsheet model was used to develop UT techniques to examine the nozzle. The model predicted probe angles and skews for very specific beam angle intercepts in the examination zones.
The basic inspection philosophy was to use a tangential beam for detection, and a 60-degree beam for tip diffraction sizing. One of the major targets was to look for indications originating at the base of the divots. The physical geometry of the vessel and blend OD surfaces (i.e., varying profile due to the nozzle outer radius) required custom contoured wedges to be used with the conventional and phased array probes.
Note that contoured wedges are an area in where phased array in field conditions could result in significant material savings as compared to conventional probes. Conventional probes would require many more wedges to provide the range of transducer angles , frequencies, and contours skews to fully interrogate the inspection zones. In addition, significant time savings would also result from phased array imaging in both the time saved in changing out conventional shoes (especially in a nuclear environment), as well as requiring less physical scanning to cover the same volume.
The required inspection involves 4 areas of the nozzle. The conventional ID radius between the shell and nozzle planes is zone 1. Areas under the nozzle boss are zones 2A and 2B. The remaining barrel section is zone 3. Each zone required detection and sizing techniques. The inspection requirement was to be able to detect and size any flaw greater than 0.250" (0.635cm) with an accuracy within 0.100" (0.254cm) in sizing.
Equipment and personnel were dispatched to the EPRI NDE center to conduct the qualification efforts. The team worked closely together to develop a successful detection and sizing technique, and subsequently applied it in a non-blind test condition to complete the qualification.
An IntraSpect Ultrasonic Phased Array Imaging system was used both for qualification and subsequent in-service field inspection support. The IntraSpect system is designed and manufactured by Westinghouse’s NDE Products group. Its phased array version used here is the cooperative effort between Krautkrammer Branson (KB USA) and NDE Products. The phased array probe and front-end electronics are KB hardware, and have been fully integrated into an IntraSpect system to result in an advanced phased array imaging system. In particular, the easy to use software features common in IntraSpect systems have been successfully applied to the complexity of phased array imaging. The system consists of the following major subcomponents as shown in Figures 2 and 3:
Fig 2:
|
Fig 3:
|
A 2.25 MHz, 16 element phased array transducer approximately 5/8" (1.6 cm) square was scanned across the inspection zones, stroking in the circumferential direction, and indexing in the radial direction. Data was collected in a pulse-echo mode. Phased array allows the resultant UT beam to be swept through a range of angles into the material, skewed through a range of angles across the material, and focused at a range in the depth of the material. Beam steering and focusing is accomplished through constructive and destructive interference of UT signals from each element as a result of time based triggering. For a simplified example, each element is fired slightly later then its neighbor progressing across the phased array probe to obtain an angle beam wave. Similarly, a focused beam is created by firing the end elements first, and the center element last (i.e., hyperbolic triggering sequence).
The overall inspection required scans to be made on three OD areas to obtain the necessary coverage of zones as follows:
Note that phased array was used in an R&D mode most of the time and then officially qualified for sizing in two of the zones.
Where applied, the basic approach was to use phased array imaging ultrasonics in a forward looking sweeping sector scan to shotgun the inspection volume (Figure 4). The image shown provides the sector scan and its associated A-Scan (RF signal) at the red cursor position.
Once an indication was detected, a B-Scan side view slice was produced at its associated angle and skew to assist in sizing the indication (Figure 5). This image shows a combined C-Scan, B-Scan and A-Scan (top to bottom of image respectively). The B-Scan shows the clad/notch corner trap, the clad to shell interface, and the notch tip at the following metal paths: 10.1, 9.6, and 8.6. Note that this B-Scan was collected in a manner scanning from left to right as seen from the angled cursor.
The same area was also scanned from the opposite direction as shown in Figure 6. The clad roll is more easily seen in this larger area view. The tip signal again shows up sooner in time (and associated metal path).
Fig 4:
|
Fig 5:
|
Fig 6:
|
Once data has been collected, additional IntraSpect analysis tools can be applied. For example, a calibrated depth feature allows an operator to more precisely measure the differences between viewed indications such as the corner trap and crack tip. A delta curser feature also allows the overall size of an indication to be directly measured (i.e., length and width).
The combination of conventional and phased array UT qualification efforts demonstrated sizing techniques for Zones 1, 2A and 2B. The resulting sizing error RMSE was +/- 0.060" (0.152 cm).
Flaw detection was ultimately accomplished in the qualification effort with fixed angles via conventional automated UT. It was determined that sizing could also be accomplished using fixed angles. The intent of the phased array capability was to have it available and ready to apply for certain zones where the flaw sizing success of conventional testing was originally deemed questionable. This objective was satisfied.
The proven phased array imaging technique and hardware were then taken to the field to support flaw sizing (if flaws were detected) for the inspection of the vessel. As no flaws were detected, sizing via phased array imaging was readied, but not applied under field conditions.
As a result of this effort, the potential benefits of phased array imaging toward BWR vessel nozzle inspections were recognized and partially qualified. It is now the intent of the plant to plan sufficient time to qualify the remaining inspection areas to be performed by phased array imaging for future inspections.
For additional information about the IntraSpect UT Phased Array system, please contact NDT.net exhibitor AMDATA NDE Technology LLC in the USA at +1-860-627-8750 or via e-mail at amdatande@aol.com
| © NDT.net - info@ndt.net | |Top| |