Page 45 - Winter 2020

The Goley et al. (2011) method provides an adjustment to the noise exposure that is independent of the sample duration. Assuming that the permissible exposure is an 8-hour A-weighted sound pressure level of 85 dB(A), the kurtosis correction would adjust the level and the potential contribution to the dose accordingly
D = 100*(C1 ⁄T1 +C2 ⁄T2 ...+CN ⁄TN) (4) where the Ci values are the times at a given exposure level
and the Ti are the permissible times for that same level
Ti = (5)
If a worker were exposed to 91 dB with a kurtosis level of β = 3, the allowable time would be 2 hours (120 minutes) before a 100% dose would be reached (NIOSH, 1998). If the kurtosis for that same noise exposure level were 15, the kurtosis-adjusted noise level would be about 3 dB higher [4.02*log10(15/3) = 2.81], resulting in a reduction of exposure time by approximately 50%.
The Kurtosis Metric in Other Applications
Use of Kurtosis to Evaluate the Combined Effects of Noise and Solvent Exposures Fuente et al. (2018) conducted a pilot study to test the possible interactive effects of noise and a mixture of chemical solvents that can cause hearing loss. They selected 20 workers exposed to noise plus a mixture of solvents (toluene,xylene,ethylbenzene,andstyrene)and20exposed only to noise. Each worker was assessed for solvent exposure and for noise using the CNE and the kurtosis-adjusted CNE. Interactions between the noise exposure (unadjusted) and solvent exposure were not significant for the hearing thresholds 1,000 Hz through 8,000 Hz. There was, however, a significant effect at 6,000 Hz when the cumulative noise exposure was adjusted for kurtosis. This pilot study provides evidence that when noise is combined with certain solvents, the use of kurtosis may help identify a harmful effect.
The Effect of Hearing Protection Devices
on Kurtosis
Currently, hearing protection devices are commonly used to protect workers from high-level noise exposures. The effects of hearing protection on kurtotic noise exposures is largely unknown. Murphy (2019) investigated the effect of hearing protection devices on the kurtosis of noise from a jackhammer (a pneumatic or electric chisel used
to break up concrete) using an acoustic test fixture placed approximately two to three meters away from the operator. He evaluated hearing protector conditions, alternating between the ears-covered and ears-uncovered conditions as the jackhammer chiseled through a 6-inch-thick slab of 5,000 psi concrete. The ears-uncovered kurtosis of the jackhammer noise was 15, whereas the ears-covered conditions varied between 2.6 and 12.1 depending on the hearing protector. The earmuff condition reduced more of the high-frequency components of the noise than the low- and midfrequencies and yielded a kurtosis of about 2.6. One of the earplugs with a nonlinear filter yielded a kurtosis of 12.1 in the open-filter condition. The kurtosis adjustment to the noise level varied between −0.3 and +2.3 dB. The A-weighted attenuation of the protectors, however, varied between 21 and 41 dB. For this exposure, a properly fitted hearing protector had a greater effect on the allowable exposure time than the reduction in exposure time due to a high kurtosis level. Future efforts could include a better understanding of the interaction between kurtosis and the frequency characteristics of both the noise environment and the hearing protector attenuation.
Kurtosis in the Future
Kurtosis in Practice
Further research is needed to apply kurtosis to programs for the prevention of noise-induced hearing loss.
Establishment of Protocol for Kurtosis Application
The window length necessary for computing kurtosis and the sampling rate for noise recordings directly affects the kurtosis value of a noise. Therefore, in specific environments such as industrial settings, noise exposures can be clearly and effectively characterized by kurtosis only when the window length and the noise sampling rate are fixed or standardized. The choice of window duration should consider computation efficiency and accuracy to predict hearing damage, along with the choice of noise sampling rate. The effective range of human hearing is 20 to 20,000 kHz. A sampling rate of 48 kHz can cover the audible frequency range for humans and should be suitable for industrial noise sampling. However, military impulse-noise bursts are usually very short, and sampling rates of 200 kHz or above are often used for detailed analysis of these impulse noises. Therefore, it is necessary to develop a kurtosis analysis protocol according to the application situation, the type of noise (impulse or impact noise), effects of sampling rates, and filtering.
Winter 2020 • Acoustics Today 45