AUTOMATION OF THE DETECTION OF DEFECTS IN THE HEAT-EXCHANGERTUBING OF NUCLEAR STEAM GENERATORS

THE DETECTION OF DEFECTS IN THE HEAT EXCHANGER TUBING OF NUCLEAR STEAM GENERATORS IS CURRENTLY PERFORMED BY AN INSPECTION TEAM OF EXPERIENCED ANALYSTS USING A NUMBER OF VISUAL AND MANUAL OPERATIONS. THE PRESENT PROCEDURE REQUIRES CONSIDERABLE TIME AND EFFORT, AND IS SUBJECT TO THE UNCERTAINTIES OF HUMAN INTERPRETATION AND JUDGEMENT. THE OBJECTIVES OF THIS PROPOSED RESEARCH PROJECT ARE TO EVALUATE ALTERNATIVE DEFECT DETECTION SCHEMES, SELECT AN OPTIMUM DETECTION ALGORITHM, TEST THIS OPTIMUM PROCEDURE, AND EVALUATE THE FEASIBILITY OF INCORPORATING THE ALGORITHM IN AN INSTRUMENTATION PACKAGE THAT WOULD ENABLE AUTOMATIC DEFECT DETECTION TO BE ACHIEVED. A COMPREHENSIVE DATA BASE OF EDDY CURRENT SIGNALS COLLECTED UNDER A VARIETY OF CONDITIONS WILL BE USED IN THE PHASE I RESEARCH. THESE SIGNALS WILL BE ANALYZED USING A VARIETY OF SIGNAL PROCESSING AND STATISTICAL TECHNIQUES TO DETERMINE AN OPTIMUM DEFECT DETECTION STRATEGY.