In a previous research project [SPR 2418; "Study of the performance of acoustic barriers for Indiana toll roads,"] the influence of environmental factors and of advanced sound barrier concepts was investigated. The presence of temperature gradients over pavements was found to have a strong influence on sound propagation. Refraction of sound waves emitted by tire-road interactions in the vicinity of the ground also affects sound barrier performance. Modified ray tracing model suggested that prevailing winds have an influence on barrier performance at large distances. Randomized edge configurations were found not to improve barrier performance for traffic noise. Random edges simply scatter sound energy without any net noise reduction. Although the edge can be optimized for specific frequency components and locations, it appears that optimization for broadband noise control is difficult. The study also suggested that adding sound absorptive material along the barrier edge could enhance barrier performance. The present study is the continuation of the latter effort to confirm the findings related to the sound absorptive treatment on the barrier through a more rigorous study and to apply the design concept to a realistic situation. A comparison was made between barriers that incorporated sound absorptive treatments and barriers with T-shaped tops. The results confirmed that a sound absorptive treatment near the barrier edge resulted in a performance improvement over corresponding rigid barriers. A design optimization study of the most effective shape of acoustic treatments concluded that a circular shape works best. The performance of two different acoustical materials was also compared. Use of glassfiber resulted in better performance in the high frequency region, while polyolefin foam with closed cells achieved a relatively large insertion loss at low frequencies. Efforts were made to develop a numerical predictive model. The boundary element method was used to model the infinite size surrounding fluid effectively. The disadvantage of the boundary element model is the calculation load associated with the large number of elements required for high frequency analysis. A mesh optimization procedure was successfully implemented in the boundary element model to reduce the calculation time while satisfying the tolerances for analysis accuracy at each analysis frequency. Octave band averaging was also adapted to facilitate the comparison between the numerical results and experimental data. It was found that the results from the boundary element model agree relatively well with the experimental results up to 6300 Hz at selected locations. The insertion loss distribution proved the numerical model's capability of reproducing the rather complicated interference pattern on the receiver plane correctly at one-third octave band frequencies from 1000 Hz to 6300 Hz. Spatial-averaged insertion losses over different size receiver planes showed that the numerical model was less reliable when the averaging was done close to the ground level. Preliminary measurements for an actual, real-size barrier were performed in South Bend, Indiana, to identify a measurement location that can be used to verify the effectiveness of the proposed add-on device. The add-on device was designed based on laboratory experiments and numerical studies, and was proven to be effective in a realistic highway environment. Federal Highway Administration (FHWA) Traffic Noise Model (TNM) was exercised for the comparison of the measurements. On-site measurements were performed to evaluate the absorptive treatment. Application of the treatment over a limited (6 m) region improved the performance of the barrier by 2 to 5 dB at the frequencies from 2000 Hz to 5000 Hz.

Report Number



noise barriers, traffic noise, sound propagation, sound absorptive material, acoustic treatment, SPR-2593

SPR Number


Project Number


File Number


Performing Organization

Joint Transportation Research Program

Publisher Place

West Lafayette, IN

Date of this Version