Detection of Damage in Composite Materials Using Classification and Novelty Detection Methods

Abstract

AbstractThe increased use of composite materials in engineering applications, and their susceptibility to damage means that it’s imperative that robust testing technique are developed to help in the detection of damage. Many of the detection techniques currently available are highly complex, difficult to conduct and rely on human interpretation of data. Simple testing methods are available but are too unreliable to be used effectively. This investigation explores the development of simple testing methods which use classification and novelty detection methods to detect the presence of damage in composite materials, making the process of damage detection much quicker, simpler and more versatile.KeywordsCompositeNDTNDIDamage DetectionDelaminationFourier TransformDigital Signal ProcessingNeural NetworkNovelty Detection