Transfer Path Analysis of Multi-Structure Acoustic Systems Using a Simplified Measurement Object Samira Mohamady, Markus Müller-Trapet, Michael Vorländer Institute of Technical Acoustics, RWTH Aachen University, Germany, Email :{smo, mmt, mvo}@akustik.rwth-aachen.de Abstract Sound energy produced by heavy machinery surrounds our everyday life. Vibro-acoustic analysis of such a systems leads to identify the problem to eliminate unnecessary noise generation. Most of the sound generators can be entitled to multiple structure acoustics system due the coupling of structure and fluid in most of the machines. The flow of vibro-acoustic energy can be obtained using Transfer path analysis (TPA). The aim of the present study is to predict and evaluate the sound energy propagate through multiple structure acoustics system using the TPA method. The system is a rectangular enclosure with 2 MDF plates placed perpendicular in the center of the arrangement. The results are also estimated using the finite element method (FEM). In this study COMSOL Multiphysics © is used as FE software. Introduction Transfer path analysis (TPA) is a well know method to trace the flow of vibro-acoustic energy from the excitation sources to the given receiver locations [1-4]. Since the method requires numerous inputs and multiply force analysis, heavy structures are not easy to study. Therefore simulations of such systems are proposed to identify the amplitude of every transfer paths in early implementation stages. The simulation should be accurate enough to advance the analysis process without need of disassembly and assembly of the system. Simplification of the study object could lead to better understanding of the simulation accuracy and relatively easy understanding of the physics behind it. This paper provides the TPA analysis of a simplified multi- structure acoustic system. The analysis is performed experimentally and analytically to analyze the accuracy of the simulated model. A Comparison of the simulation and experiment is also carried out, which can be extended to the complex structure. Experimental Arrangement A rectangular enclosure with dimensions of       according to [1] is used as simple structure with verity arrangement as follow: Arrangement1: Box with 6 medium-density fiberboard (MDF) sides. Arrangement 2: Box with 5 MDF sides and an aluminum plate on top of it. Arrangement 3: Arrangement.2 with two MDF boards placed in the middle of it. The mechanical properties of the MDF and the aluminum are given in the Table_1. Following modifications of the test object are performed to simplify the measurement process; the necessary cables of sensors and actuators are passed inside the box using designed cable-guide [1], to prevent any changes of setup during measurement a small window of size     is embedded in a large side of the box. The aluminum plate is also connected to the MDF wood using modeling clay which prevents air leakage in the junctions. Material Mechanic properties Thickness  Mass density (    Young’s modulus  Poisson ratio MDF Wood 0.022 618 3.2 0.33 Aluminum 0.001 2700 70 0.33 Table 1: Mechanical properties of materials under study In this experiment a small dodecahedron sound source with diameter of 0.01 m is used to generate an omni-directional sound field. The measurement of air-borne excitation is performed using Sennheiser KE4 electret microphones. Figure 1 shows all experiment arrangements and measurement devices. Figure 1. Experimental setup: a) MDF box, b) box with the aluminum plate on top of it, c) box with the aluminum plate and 2 MDF plates inside it. d) dodecahedron sound source e) Sennheiser KE4 electret microphone Simulation Criteria The COMSOL ® Multiphysics software is used as a finite element tool to model all arrangements introduced in the experimental setup. The geometry of the arrangements are modeled using graphical user interface of COMSOL ® software. Modeling of the interior acoustic field inside the box and vibration of the aluminum plate are performed using acoustic shell interaction module. The model is excited using primary sound flow point source and sound field is calculated in a specific corner of the box. The model of geometry in all here arrangements and source and receiver placements are illustrated in Figure 2. As it is shown the placement of the sound source and receiver follows exactly the same as experiment. a b c d e