Abstract
Powder metallurgy (PM) is a rapidly growing field that offers a cost-effective way to manufacture complex-shaped metal components with superior properties compared to conventional manufacturing methods. However, PM involves the production of green compacts, which are prone to defects such as cracks, voids, and inclusions due to their fragile nature. These defects can compromise the final properties of the sintered components, resulting in high rejection rates and increased costs. Therefore, the need for reliable and efficient non-destructive testing (NDT) methods to detect defects in green compacts has become a focal point of research. Traditional methods for crack detection are based on destructive imaging techniques, where, polished sections of the specimen are observed under an optical microscope or scanning electron microscope. These methods, however, are time-consuming and cannot predict the initiation and likelihood of a crack in green compacts. In contrast, this study reviews different NDT methods proposed by researchers in the field to detect cracks and defects in green compacts. The discussed NDT methods include thermal imaging, X-Ray imaging, acoustic and ultrasound testing (UT), eddy current based testing (ECT), camera-based techniques, and simulations based on the Drucker-Prager Cap (DPC) model. We further review the usage of machine learning (ML) techniques for classification and decision-making. All methods are introduced along with a discussion of their limitations and merits.