Engine Bearing Analysis Under Diverse Conditions via Response Surface Methodology

With the increasing demand for reliable diagnostics in automotive components, accurately assessing engine bearing health under various conditions has become crucial. This study presents a robust methodology for identifying and distinguishing faults in internal combustion engine bearings. Specifically, the focus is on differentiating between healthy and faulty bearings by analyzing key predictive variables—temperature, engine speed, and humidity—and their effects on bearing vibration responses. Experiments were conducted at different levels using a Design of Experiments (DOE) framework, providing valuable insights into the nonlinear impacts of each factor on the vibration response of the engine bearing system. Root Mean Square (RMS) vibration data were analyzed using Analysis of Variance (ANOVA) to assess the significance of the model terms, indicating that engine rotation speed has the most significant effect on bearing vibrations, while environmental humidity exhibited the least significant effect. The rate of increase in RMS vibration was higher at sub-zero temperatures for healthy bearings and greater above 30° C , for faulty bearings. Additionally, interactions among all three predictors were found to be insignificant, demonstrating that each parameter independently influenced the vibration response. This combined approach—integrating DOE and Response Surface Methodology (RSM)—shows promising potential for predicting the dynamic behavior of main journal bearings in internal combustion engines.