Abstract:
Noise is a ubiquitous problem in the United States, particularly in the construction industry. Because over-exposure to noise does not usually have immediate adverse consequences, there has been little focus on exposure assessment and control of construction noise. Noise-related studies have been done in the construction industry, but the date are not comprehensive. While sound pressure levels associated with certain tasks have been fairly well characterized, frequency spectra have not.
Different frequencies of sound have varying physiological impacts on the ear. Therefore, the risk of developing hearing loss depends not only on the intensity of the noise that one is exposed to, but also the frequency characteristics of that noise. In addition to the effect that different frequencies have on hearing loss, they interact differently with physical objects as well. This means that characteristics of noise such as frequency distributions and perhaps the patterns of the exposures need to be accounted for in exposure assessment and control implementation.
Basic noise measurement instrumentation is generally designed for measurement of discrete samples of continuous noise and is not capable of frequency analysis as the complete continuous noise signal is not captured during the measurement. Recording the continuous signal using digital recording technology and then performing a subsequent spectral analysis is an approach which allows for the capture and derivation of true frequency distributions.
This study involved detailed analysis of noise exposures in the construction industry. Task-and tool-based measurements were made via personal digital sampling at various types of construction sites. Samples were collected at 441.1 kHz and then downloaded into a computer with Cool Edit software and subsequent t1/3 octave band analyses were performed with LabView software. Time-domain results were obtained by downloading the sound level meter data with QuestSuite Software.
Results indicated that there was a difference between both time-and frequency-domain characteristics for different tasks. Further, tasks (and tools) were typically characterized by a few 1/3 octave bands in which the majority of energy was contained. The spectral data for a tool was found to be representative of its corresponding task spectra. Time-domain analysis illustrated that all of the tasks which were included in this study were characterized by high sound pressure levels (>89 dB). Although ambient area noise levels were high as well, for the most part they contributed little to the task results.