 Table of Contents ECNDT '98 Session: Nuclear Industry | Examination of core shroud weldsJens Larsen*, Hans Kristensen & Leif Jeppesen
*Corresponding Author Contact: | jl@force.dk ; http://www.force.dk |
Late 1995, FORCE Institute was asked to design and build a manipulator for in situ ultrasonic examination of the stainless steel core shroud welds in two different nuclear reactor types in Sweden. The core shroud is approximately 5 meters in diameter and 6 meters in height. The examination should cover as close as possible 100 % of the 8 longitudinal and the 3 circumferential welds. The first, was to be examined summer 1996.
The manipulator should be able to operate within the narrow space between the core shroud and the reactor vessel without collision with several non-removable obstacles. It should be water tight at least for the duration of a whole inspection period at a water depth of at least 25 meters. The materials should withstand radiation levels up to 1200 Sv/h in core zone. No emission of chlorides or halogens into the reactor water from the materials was permitted - i.e. no grease was used. Repositioning accuracy of all axes: within ± 2 mm. The scanning movement should be possible in both longitudinal and transverse direction for longituinal as well as for circumferential welds.
New rules called for qualification test of: the manipulator, the NDT system, NDT procedures for data collection, NDT procedures for sizing, personnel for data collection and/or evaluation before the examination could begin. This resulted in a very tight schedule for design, production and test of the manipulator and the NDT system. This left no margins for mistakes which would require redesign.
Due to the above mentioned requirements, it was decided to establish two 3D CAD models:
To control the movements of the manipulator, the modular PS-4 control system was chosen. In this case it would control 6 axes. The PS-4 control system consists of individual modules (Master-, Link- and Driver Modules) containing control and communication electronics. The communication between the units is performed through a local ring network based on twisted pair wire and RS422 drivers. For communication the HDLC protocol was used. The communication in the local network and the communication between the network and the base station (the Master Controller) were controlled by the Master module. Two driver modules can be connected to a single Master module. For connecting more drivers to the system three extra link modules were used. These contain an extra communication computer and an extra power supply (30V / 5 V). Three different types of driver units are used in the control hardware. A DC-motor driver unit built into the DC-motor housing, an independent DC-motor driver unit for three smaller DC motors and a valve control driver for double action pneumatic valves. The Master Control for the two circuits is connected to the control system computer via two normal RS-232 serial interfaces.
 Fig. 5: The core shroud manipulator - print out from the 3D model |
The mechanical design is minimised to allow parallel activities as fuel in/out loading and central mast activities during inspection with the manipulator. To prevent water from getting into the motors, these were pressurised and tested to 4 bar. Purpose built cables were used to minimise the number of water tight connectors and cables in the basin. In order to ensure sufficient repositioning accuracy, tooth rails and threaded spindles or tooth belts were used. Please refer to Figure 5 for an overview of the main components of the manipulator
Two different positioning possibilities exist: Relative to a fix point on the core shroud or absolute co-ordinates determined by zero switches on each of the position critical axes on the manipulator. During examination any movement is fed back to a work-station with the CAD model of the core shroud and the manipulator. Collision detection helps the operators avoid collision during transport and scanning. This collision detection proved to be essential in areas where the cameras can not reach.
Vital components were tested in Co-60 radiation to determine the component exchange and service intervals. Radiation monitoring was carried out during the examination by placing small dosimeter bricks in water tight boxes at relevant positions (close to the electronic components). To increase the lifetime of the cameras for supervision (Sword docking and manipulator movements) these were placed on the Docking Station approximately 1.5 metre above the fuel elements. Encoders in the lower motors were specified to be exchanged after each examination.
The new PS-4 system developed by FORCE Institute was chosen to enable collection of A-scan data. This would enable further analysis in case indications were found. The speed will depend on number of probes applied and the volume to be covered.
 Fig. 1: Longitudinal weld geoetry  Fig. 2: Circumferential weld geometry at top flange (flange at left)  Fig. 3: Circumferrential weld geometry  Fig. 4: Image form the PS-4 system with A-scan, P-scan and TOFT images displayed |
Please refer to Figure 1, Figure 2 and Figure 3 for examples of weld geometry covered.
Before evaluation, the A-scan data were converted to easily read and evaluated P-scan images. Had this conversion taken place during scanning, the amount of data to be stored would have been reduced significantly, but then re-scan would be necessary, if A-scan data should be required for further analysis.
Please refer to Figure 4 for an image of the screen for evaluation of an indication in a test block
The system has been used for UT, ET and visual testing on 3 different core shrouds. During visual examination, the position of the manipulator can be overlayed in the image. The docking station can be adjusted to suit rings built for different core shroud diameters. The Sword can be built in different lengths (two at present). Examination of the inside surface of the reactor vessel is also possible, if the "hand" holding the probes is turned 180 degrees.
Flexibility
Safe operation due to
Economic
Repeatability
Proven in practice
| NDT.net |