No abstract

Experimental and computational results are given for active control systems minimizing sound radiation from steel bridges for trains. The performance of the active systems with respect to noise reduction, convergence, stability, and actuator requirements is described. At frequencies above approximately 300 Hz passive means can be used to reduce the radiation of noise from railway track. At lower frequencies, however, passive means to reduce acoustic radiation are not possible due to various reasons. Additionally, the main source of acoustic radiation at low frequencies is the steel bridge instead of the railway track. Therefore, the active noise control system for the steel bridge is based on the active isolation of track vibration by minimizing the mean-squared force between the steel bridge and the support for the rails. This is accomplished by controlling the distance between the rail supports and steel bridge with piezoelectric stack actuators. Transfer functions, load impedances, and actuator characteristics obtained from finite-element analysis and measurements lead to noise level reductions of 10 to 20 dB, depending on the control strategy and the available computational resources of the real-time implementation.