Combining classical and component-based TPA for equivalent load identification

Abstract

The transfer path analysis (TPA) has become a rather standard tool for solving noise and vibration problems, as it helps understanding the mechanisms responsible for the generation and transmission of those quantities. By better understanding the intricate role of multiple sources and propagation paths, it is possible to diagnose and propose effective modifications that would addresses such issues at specific target locations. Although originally an experimental approach, hybrid methods that include modeled sub-systems have been proposed, which allow the assessment of key system features even at stages prior to the construction of full physical prototypes. However, in classical TPA, the operational forces are characteristics of the complete system, which implies that, with each modification in one subsystem, it is necessary to redo all the tests for the correct determination of target functions. For this reason, in recent years, interest has been renewed in the development of faster and simpler techniques for analyzing energy transfer paths, which offer a compromise between workload and accuracy. More specifically, a set of methods called component-based TPA is highlighted, which consists of characterizing the excitation of a source through a set of equivalent forces (or even interface velocities) inherent only to the active subsystem. In this way, the responses at target locations on the passive subsystem could be calculated using these equivalent loads and the dynamics of the complete system, obtained numerically, experimentally or under a hybrid framework. This work presents a critical analysis of the component-based TPA methods and proposes the combined use of these methods with a classical TPA approach in the process of determining equivalent forces of the active subsystem. This set of equivalent forces, combined with the passive subsystem dynamics, allows the prediction of the vibrational behavior of the full assembly at targeted locations, without the need for a full experimental analysis of the assembled system. As the case study presented here consists of a modular academic setup, it allows the qualitative assessment of the method in distinct assembly boundary conditions, in which the subsystems are connected via rigid or flexible joints.