Detecting Single Faults in Cam-Follower Systems Using Bond Graph Methodology

Abstract

This study explores using bond graph methodology to detect single faults in cam-follower systems. Cam-follower mechanisms are widely used in various machines, and their proper functioning is crucial for overall system performance. Faults in these systems can lead to abnormal vibrations, noise, and potential performance degradation. The bond graph approach offers a systematic method to model the dynamic behavior of a cam-follower system. By representing the system components (cam, follower mass, spring, and friction) as elements and their interactions through bonds, we can derive a system equation that relates the follower's acceleration to the cam profile, spring constant, and friction coefficient. This study focuses on analyzing the system equation to identify potential single faults that could cause higher follower acceleration than expected. By examining the terms in the equation, we explore how a decrease in spring constant (broken spring) or friction coefficient (worn-out surfaces) can lead to a reduction in the opposing forces acting on the follower, resulting in potentially higher acceleration. The ability to detect these single faults using the bond graph model and system equation provides a valuable tool for diagnosing and maintaining cam-follower systems. By monitoring the system's behavior and comparing it to the expected performance predicted by the model, potential faults can be pinpointed before they cause significant damage or performance issues. This study highlights the effectiveness of bond graph methodology as a non-invasive approach for fault detection in cam-follower systems. The derived system equation provides a foundation for further research on incorporating sensor data and signal processing techniques to develop real-time fault detection and monitoring systems.