In this thesis, a prediction model based on Finite Element Method has been presented for impact noise in different floors of a building. Expensive experiments will be gradually replaced by prediction models which consist of three parts, excitation, system and response. Source of excitation is standard ISO tapping machine. In this research, different methods for calculating the applied forced by ISO tapping machine to a building floor have been investigated and an appropriate one has been selected for calculating the applied force to the light weight building floor. In the selected method, force is obtained using the mobility of excitation point. The obtained force in time domain can be transformed to the frequency domain by means of Fourier transform. In this prediction model, system consists of the building floor, the air filling the rooms and rooms boundaries. The model is simulated using structural and acoustical elements of ANSYS software. In modeling the system, it has been tried to use material properties and geometrical dimensions of an actual test sample. After defining the system and the excitation force, the third part of a prediction model i.e. system response to the excitation source should be determined. The response in this research is the average acoustic pressure level in the test room after an impact is applied on the building floor. To obtain this response, harmonic and transient analyses are used. Frequency and time domain values of the force calculated from previous step are used as excitation source in harmonic and transient analysis respectively. Because of the high computational cost in harmonic analysis, an optimized algorithm for analysis of the problem has been recommended. For reaching the desired response different methods have been used for calculating the average acoustic pressure level of the room in each of the harmonic and transient analyses. In the first method, radiated power to the noise receiving room and in the second method time and spatial average of pressure in different points of the room are used to evaluate the desired responses. Results obtained by these methods have been compared. For verifying the simulation, its results should be compared with actual test results. Results of a test carried out in a laboratory in Sweden have been used for this purpose. Comparison of simulation and test results shows that in a big portion of the frequency range, there is a good agreement between these results and error is less than three percents. In frequency ranges where simulation and test results are different, reasons of this disparity are discussed. The problem was also solved by transient analysis and results were compared with harmonic analysis. Key words: Acoustics, Impact Noise, Harmonic Analysis, Transient Analysis, Finite Element Method
