Prediction of Exhaust System Vibration Through Harmonic Analysis & High Cycle Fatigue Life Evaluation

Abstract

Exhaust system typically experiences vibration during engine operating conditions due to periodic disturbing forces (firing force and inertia force) which are generated from the engine. Natural frequency of the exhaust system gets excited due to the periodic forces causing resonance which often leads to high cycle fatigue (HCF) failure. Turbocharger is a part of exhaust system and it is mounted on the exhaust manifold. The periodic forces are transferred from engine base (Cylinder head and Block) and these forces gets amplified to overhanging components like exhaust system turbocharger. It is an industrywide practice to perform modal analysis to determine the natural frequencies of the system. However, modal analysis cannot predict the intensity with which the system would vibrate. Thus, we need to make some assumptions about the system vibration ‘g’ levels. Based on accuracy of this assumption, we may end up under-designing or over-designing the system. Harmonic analysis enables us to accurately predict the ‘g’ level at turbocharger using experimental cylinder head base excitations. After recording the correlation with experimental data in many cases it was found that this approach further aided in establishing damping constant factor of the exhaust manifold at elevated temperature. This analysis process has been validated with multiple cases as it has turned out to be a potential approach while doing design risk assessments and optimizing the engine vibration validation efforts. The benefit of prediction of exhaust system vibration level allows us to avoid iterative design process in the early stage of product development thus optimizing the design by taking advantage of shifting the natural frequency of exhaust system to lower source excitation (cylinder head). This saves vast amount of simulation lead time. Another benefit of this process is that the prediction of resonance condition of exhaust system through simulation helps us to estimate the fatigue life against the predicted ‘g’ level.