DocumentCode :
1336395
Title :
A System for Nuclear Fuel Inspection Based on Ultrasonic Pulse-Echo Technique
Author :
Thomé, Zieli Dutra ; Pereira, Wagner Coelho Albuquerque ; Machado, João Carlos ; Seixas, José Manoel ; Soares-Filho, William
Author_Institution :
COPPE/UFRJ-Fed. Univ. of Rio de Janeiro, Rio de Janeiro, Brazil
Volume :
58
Issue :
5
fYear :
2011
Firstpage :
2452
Lastpage :
2458
Abstract :
Nuclear Pressurized Water Reactor (PWR) technology has been widely used for electric energy generation. The follow-up of the plant operation has pointed out the most important items to optimize the safety and operational conditions. The identification of nuclear fuel failures is in this context. The adoption of this operational policy is due to recognition of the detrimental impact that fuel failures have on operating cost, plant availability, and radiation exposure. In this scenario, the defect detection in rods, before fuel reloading, has become an important issue. This paper describes a prototype of an ultrasonic pulse-echo system designed to inspect failed rods (with water inside) from PWR. This system combines development of hardware (ultrasonic transducer, mechanical scanner and pulser-receiver instrumentation) as well as of software (data acquisition control, signal processing and data classification). The ultrasonic system operates at center frequency of 25 MHz and failed rod detection is based on the envelope amplitude decay of successive echoes reverberating inside the clad wall. The echoes are classified by three different methods. Two of them (Linear Fisher Discriminant and Neural Network) have presented 93% of probability to identify failed rods, which is above the current accepted level of 90%. These results suggest that a combination of a reliable data acquisition system with powerful classification methods can improve the overall performance of the ultrasonic method for failed rod detection.
Keywords :
data acquisition; fission reactor fuel; fission reactor operation; fission reactor safety; light water reactors; neural nets; ultrasonic transducers; PWR technology; data acquisition control; data classification; electric energy generation; frequency 25 MHz; linear Fisher discriminant; mechanical scanner; neural network; nuclear fuel inspection; nuclear pressurized water reactor; operating cost; plant availability; pulser-receiver instrumentation; radiation exposure; signal processing; ultrasonic pulse-echo technique; ultrasonic transducer; Acoustics; Assembly; Electron tubes; Face; Fuels; Inspection; Probes; Neural network; non-destructive testing; nuclear fuel inspection; ultrasound;
fLanguage :
English
Journal_Title :
Nuclear Science, IEEE Transactions on
Publisher :
ieee
ISSN :
0018-9499
Type :
jour
DOI :
10.1109/TNS.2011.2164557
Filename :
6031787
Link To Document :
بازگشت