Challenge details:

Currently, plain bearing (consists of a cylinder with a hole bored through the centre axis) refer Fig. 1, conditions are monitored manually or with basic tools, which often fail to detect early signs of wear, resulting in unexpected failures and costly repairs. In the Rolling Stand Mandrel, the Roll Shaft is placed on a plain bearing and rotates at high speed to resize the hot billet according to the pass design of each SKU (stock keeping unit). During this process, lubrication oil is injected into the gap between the Roll Shaft and the bearing to create a hydrodynamic lubrication state between the surfaces. This state is influenced by factors such as shaft speed, lubrication viscosity, temperature, and contamination levels. However, hydrodynamic lubrication is not always effective. At the start and end of rotation, when the shaft speed is low, boundary lubrication occurs, causing intermittent contact between the Roll Shaft and the plain bearing, leading to wear over time. Traditional monitoring has failed to provide real-time, accurate data for predicting failures. Additionally, there are constraints on installing sensors due to lack of space and the need to operate at high temperatures.

We are looking for a solution that provides real-time condition monitoring of bearing with high accuracy, non-intrusive method that can integrate seamlessly with existing system, and ability to withstand high-temperature (for high speed, 45-80 C) and harsh environments (presence of water and dust).

Currently, vibration and temperature sensors have been installed to measure the health of the bearing.