Research on the Application of Statistical Energy Analysis to the Bridge-borne Noise.

Associate Professor Yasuyuki Miyaki

The noise pollution for automotive and railway transport over steel bridges is an important issue. The steel bridges, rail tracks, or their support are mainly recognized as the noise source for Bridge-borne noise, because the vibrations excite the bridge structure and cause the amplification of noise. This type of noise is dominant in a frequency band from 10Hz to 1kHz. The noise analysis using Finite Element Method (FEM) is restricted to frequencies below 100Hz while noise analysis using Statistical Energy Analysis (SEA) is restricted to high frequencies.

In this research, establishment of a statistical method to analyze and predict the noise at low and high frequencies is focused. A Z-shaped plate structure is used here as a simulation model. First, FEM and SEA is used to distinguish the range of frequency that is valid for both methods. In FEM, modal analysis is done to extract the first 20 modes. Harmonic analysis is then also performed to evaluate displacement as a response to sinusoidal force. Finally, to validate the simulation results described earlier, an experiment to measure the sound pressure level is conducted.

Simulation and experimental results showed that sound pressure level at 63Hz and above can be predicted by statistical energy analysis for the Z-shaped plate model used in this research.