Study and development of denoising and contrast enhancement technique for effective flaw characterization by mathematical modeling from ut signals

Abstract

newline Weld quality assessment is an integral area of research as newlineAmerican Society of Mechanical Engineers imposes strict guidelines on newlinethe acceptance of weld pieces. Quality Assessment involves acquiring newlinethe images/signals of the weld pieces, interpreting the data and decision newlinemaking. In computer aided interpretation software has to be developed newlinefor the interpretation and decision making. newlineUltrasonic Testing is the widely accepted Non-Destructive newlineTesting (NDT) technique that uses Ultrasonic signals for the acquisition newlineof signals/images from thick walled weld pieces. In case of Pulse-Echo newlineand phased array, the amplitude of the reflected ultrasonic signal carries newlinethe information whereas in Time of Flight Diffraction, time taken by the newlinediffracted wave provides the desired information. newlineThese signals/images are interpreted and the decision about the newlineacceptance or rejection of the weld pieces is made. In recent years, newlinesoftware is developed to replace human interpreter. In order to develop newlinesoftware, it is necessary to acquire the signals depicting flaws, newlineidentify/develop suitable denoising techniques, identify/develop feature newlineextraction and classification techniques. newlineThe proposed research work aims at acquiring ultrasonic newlinesignals by deliberately introducing the flaws, developing denoising newlinetechniques for reducing noise, feature extraction and classification newlinetechniques. In this research work both planar defects (Lack of Fusion, newlineCrack, Lack of penetration) and volumnar defects (porosity, slag) are newlineintroduced and the ultrasonic signals are acquired using Pulsar Receiver.