Testing of the suitability of a number of hearing protection products with respect to the audibility of signals at railway workplaces by determining of the masked thresholds of signals in background noise

Status:

completed 12/2014

Aims:

In two IFA projects (4152 "Protection against noise for engine drivers and engine driver/shunters" and 4154 "Protection against noise for engine drivers and engine driver/shunters"), now completed, the noise exposure at workplaces in rail operations (those of engine drivers and engine driver/shunters) was determined and a selection method for hearing protection developed that assures adequate audibility of signals. The selection of the hearing protection was divided into two parts. The first was a calculation-based preliminary selection at the IFA based upon a psychoacoustic model according to Zwicker, which took account of the spectra of typical warning signals and background noises at the workplaces; the second was a personal hearing test on site at which a work situation was simulated. This method was accepted by the railway inspectorate and laid down in BG Information BGI/GUV-I 5147 by the responsible accident insurance institutions (EUK and VBG).

The calculation-based method according to Zwicker was used to study the effect of a hearing protector upon the perceptibility of a signal in the presence of background noise. Spectra of twelve background noises and 18 warning signals were considered for this purpose. Criteria for a suitable hearing protector can be determined with reference to different parameters for the perceptibility. For rail operations, the same criteria have been used to date as in permanent way construction (code S in the IFA list of approved products). These requirements are satisfied by only 13% of the products on the list of approved products.

In the present project, hearing tests on test subjects were to be used to determine whether the criteria of the calculation-based method could be relaxed without the audibility of signals against background noise being excessively impaired.

Activities/Methods:

Masked threshold measurements were performed in the laboratory (in the IFA's semi-soundproofed chamber). For this purpose, the masked threshold of a warning signal against the background noise was determined for a test subject. The masked threshold is the level at which a signal at a given background noise is perceived in 50% of all cases.

The tests were performed with approximately five different hearing protectors that the calculations had shown to be reasonably suitable for the hearing of signals. Preference was given to standard products on the market. From the signals and background noises used in the calculation-based method according to Zwicker, three or four of each had to be selected for which the results of the calculations suggested prospects for good differentiation between the hearing protectors studied.

Two measurements were required for each hearing protector/signal/background noise scenario: one performed with and one without hearing protection. The variable describing the suitability of a hearing protector for the signal audibility was the masked threshold shift, as with the calculation-based method according to Zwicker.

Where necessitated by the results, the list of suitable hearing protectors was to be extended.

Results:

The test arrangement was designed and tested and is available.

During supervision of a project commissioned externally in which a new method was developed for the hearing test with hearing protection for rail operations, new observations were made regarding the criteria for selection of the hearing protection. The present project, which was intended to test these criteria, was therefore postponed.

Plans are for the masked threshold measurements planned for this project to be integrated into a further project in which the program used to date for computing the signal audibility with hearing protection is to be reprogrammed. The new program is to be validated by suitable measurements in the course of this project.

Last Update: