Table of Contents ECNDT '98 Session: Nuclear Industry

VT Examination of BWR Re-Circulation Pump Internals J. Larsen, L. Vesth, L. Jeppesen - FORCE Inst., Denmark. FORCE Institute, Park Allé 345, DK-2605 Brøndby, Denmark Corresponding Contact: E-mail: lv@force.dk, Internet: http://www.force.dk

Introduction

FORCE Institute was asked to design and build a manipulator for a complete visual examination for cracks in the inner surface of the above mentioned pumps. Mounting of the manipulator on the pump flange should be designed to minimise the total dose taken by the operators of the equipment. For the same reason the number of changes of the camera set-up should be minimised.

Requirements to the manipulator

The camera should cover the inner surface completely keeping an angle of view not more than 15 degrees from a normal to the object in the centre of the image. The radiation level was low as the pump had just been de-contaminated, but the manipulator should be designed for service in a non-decontaminated pump as well. Repositioning accuracy within ± 2 mm. should be possible to enable further examination of indications.

Requirements to the NDT system

Figure 1 Manipulator placed between tower elevator legs (left). Inserted image: The top ring on the elevator pressed against pump flange

Figure 2 Main user interface of the PC-based control system

Figure 3 Calibration of the camera in relation to the manipulator

Figure 4 One of the CAD system generated VT scan patterns

Figure 5 Axis configuration. P axes have prismatic movements and R axes have Rotational movements

The capabilities of the camera system chain should be verified on test blocks with real cracks. Zoom should be remote controlled for closer examination of possible indications. A stop for evaluation of each frame should be possible in order to minimise the negative effects of moving objects in the image. The results should be documented on video tape or in digital form. Position of the image should be stored with the documentation. The light should be adjustable to compensate for differences in light conditions

3D CAD simulation of pump and manipulator

In order to be able to test the coverage of the surface and at the same time test for collision between manipulator and pump internals, it was decided to build a 3D CAD model of the pump. This 3D CAD model was based on as-built drawings of the pump. After modelling of the pump, the first design of the manipulator including the camera system was drawn in 3D and the kinematics of the different axes was defined enabeling a simulation of the movements using RobCad software. Please refer to Figure 1 for an overview of the 3D CAD model.

Manipulator construction - Control System

The motors and master/slave modules from the flexible and modular PS-4 system developed by FORCE Institute was chosen for all the critical axes. The motor for positioning of the manipulator in the pump housing was operated by an on/off switch independent of the manipulator control system for VT. As the UT unit built into the PS-4 system would not be used for this examination, it was decided to write a small PC based application, which would position and stop the manipulator for evaluation of each frame. This application should read an ASCII file defining the different frame positions, and move the manipulator accordingly either with fixed time intervals or at the command of the operator. Furthermore, it should be possible to control the functions of the camera from the same user interface to minimise the workload of the operator. Also storage of digital images on hard-disk should be controlled from this interface. The position of the manipulator and the camera orientation was routed to the PC via an Ethernet connection. The image was sent through an image processing board in the PC, where the position infomation was added to the image in real-time prior to saving on tape or hard-disk. Please refer to Figure 2 where the layout of the control system interface is shown. The camera was controled via a RS232/RS422 communication link. The distance from the control room to the scanner could be up to 50 meters. Please refer to Figure 3 for an image of the camera unit.

Manipulator construction - Generation of VT examination path

The 3D simulation of manipulator/camera movements within the pump was used to generate the necessary data used by the manipulator control system. During this simulation, collision was detected. This then led to adjustments in the design until the coverage was complete and no collision was detected. After generation of the necessary data files for the manipulator control system, a test was carried out using a mock-up of the pump internals and the Steady Stick manipulator. Please refer to Figure 4 for an example of the scanning path pattern covered by one of the data files.

Manipulator construction - mechanical design

Please refer to Figure 1 for an overview of the manipulator To enable easy transportation and placement of the manipulator in the different pumps, a tower elevator placed on wheels was constructed. The tower could be elongated or made shorter by use of pneumatic cylinders. After locking the tower to the pump flange, the manipulator was hoisted up and into the pump by an electrical motor independent of the manipulator control system. Inside the pump, the manipulator had 5 axes for positioning of the camera. The R1 axis rotates all other axes. The P2 axis lifts or lowers the camera. The P3 axis moves the camera further away from or closer to the rotation axis R1. The R4 axis can swing the camera 90 degrees horisontally to either side of the P3 axis (only when P3 is extended). The R5 is the vertical rotation of the camera unit.

Verification of VT system sensitivity

After verification of all scan paths in the mock-up of the pump, a verification of the VT system was carried out. This was done by placing small test blocks in the mock-up of the pump and locating these using the manipulator control system. The results was documented on S-VHS video tape and screen dumps, which could be activated directly from the same control system.

Practical experience

The manipulator and the VT system performed very well during the actual pump examination. However, it was not possible to exploit the tower elevator functions completely. Many parallel activities prohibited pushing it from pump to pump - it had to be dismounted after inspection of one pump and re-assembled at the next pump.

Conclusion

• 3D CAD simulation of the mechanical functionality of a manipulator in the design phase
• prevents in-service problems.
• Off-line generation of manipulator positions as a result of CAD simulation is a cost-effective and safe alternative to on-line monitoring/colision detection.
• The manipulator functioned as specified/expected except for the tranportation.
• The integrated control system covering manipulator movements, camera control and documentation of results including position information was easy to use.