Materials engineering

Using deep learning to classify steel materials objectively

Research News /

Rolling bearings are installed wherever something is in rotation. The wide range of applications extends from large wind turbines to small electric toothbrushes. These bearings, which consist of steel components, must be carefully selected and tested with regard to their quality and the application in question. The grain size has a crucial effect on the mechanical properties of the steel. Up to now, the size of the microscopic crystallites has been assessed by metallographers by way of visual inspection — a subjective and error-prone method. Researchers at the Fraunhofer Institute for Mechanics of Materials IWM, in collaboration with Schaeffler Technologies AG & Co. KG, have developed a deep learning model that enables objective and automated assessment and determination of the grain size.