Smart thrust bearing with integrated discharge-based energy harvesting and deep-learning fault detection

초록

Triboelectric nanogenerators (TENGs) are promising energy harvesting devices that convert wasted energy into electrical energy from ambient environments, especially in the form of friction. However, they have a few drawbacks, such as low output and severe wear. To address these drawbacks, we adopt a phenomenon known as electric discharge. Rotation is a representative input form for mechanical components, and we are inspired by a thrust bearing structure for generating our TENG as bearings are widely used in many industries and are constantly rotating. In this paper, a thrust ball bearing TENG (TBB-TENG) is proposed. The TBB-TENG exhibits characteristic patterns on the stator by modifying its surface to induce an electric discharge and ultimately enhance the output. Additionally, a mechanical stability analysis is performed theoretically in response to the modification. Furthermore, parametric studies are conducted to optimize the proposed TBB-TENG. Finally, we present the practical use of TBB-TENG as a self-powered defect diagnosis sensor using a one-dimensional convolutional neural network (1D-CNN), demonstrating its high accuracy.

키워드