Modeling experiments are reported which have investigated the frequency dependence of barrier insertion loss for various noise barrier designs. The effect of ground surfaces has been studied, treating both grass-covered ground and asphalt. Reshits show that interference effects are an important feature of observed behavior. Use is made of this knowledge in the design and testing of new barrier designs, which deliberately introduce a beneficial destructive interference phenomenon to increase insertion losses over well-defined ranges of frequency.

A scale model study is presented of the performance of thin, perfectly reflecting semi-infinite barriers in the presence of both asphalt and grass-covered surfaces. The barrier insertion loss is shown to be strongly dependent on the type of ground on either side of the barrier. Observed ß behavior is explained by a study of the interference phenomena occurring in both the barrier's presence and its absence.

No abstract

No abstract

An investigation is presented into the performance of parallel barrier configurations, using acoustical scale modeling. A realistic geometry is investigated, with the source being positioned over a paved roadway and the receiver over grass-covered ground. The grass-covered ground surface was properly modeled in terms of its impedance. Results were obtained for a range of barrier types, and demonstrate that frequency dependent effects are evident in barrier insertion loss data. In most cases, the barrier on the far side of the source did not significantly affect sound levels at the receiver. The most effective barrier design was found to be that of a gradual grass-covered slope up to an upright, thin barrier.