Airbag deployments are associated with loud noise of short duration, called impulse noise. Research performed in the late 1960's and early 1970's established several criteria for assessment of the risk of impulse noise-induced hearing loss for military weapons and general exposures. These criteria were modified for airbag noise in the early 1970's, but field accident statistics and experimental results with human volunteers exposed to airbags do not seem to agree with the criteria.

More recent research on impulse noise from weapons firing, in particular that of Price & Kalb of the US Army Research Laboratory, has led to development of a mathematical model of the ear. This model incorporates transfer functions which alter the incident sound pressure through various parts of the ear. It also calculates a function, called the "hazard", that is a measure of mechanical fatigue of the hair cells in the inner ear. The repeatability of the model was examined in the present study by comparing its predictive behavior for airbag noise impulses generated by nominally identical airbag systems. Calculations of potential "hazard" made by the model were also examined for reasonableness based on mechanical and biomechanical considerations. A large number of airbag noise pulses were examined using the model.

The results provide some counter-intuitive insights into the mechanism of noise-induced hearing loss from deployment of airbag systems. They also indicate that, based upon testing of feline subjects (which are believed to be a good indicator of the risk to the more susceptible segment of the human population), there could be a risk of temporary and possibly permanent threshold shifts in approximately sixty seven percent of the 1990-1995 model year vehicles from 19 manufacturers which were tested and assessed using the human ear model. A statistical estimate of the risk for the human population has yet to be quantified, but work is in progress to do so.

Work is also underway to develop a pre-production airbag component test for deployment noise that component suppliers can incorporate into the design and development process. Once injury risk curves and test protocols are established, it is recommended that the ARL Human Ear Model be utilized by the automotive community as the assessment method of choice.