During its initial use the vehicles go through a period of adjustment, in which the structure and moving parts show their problems of looseness, tightness, stress, excess or lack of torque by emitting characteristic noises of different intensity and frequency. These noises are bothersome for the user. Car manufactures and private companies have specially-trained personnel who are able to identify the source of those noises simply by listening to them. Once the source is identified, the problem is often solved by tightening the moving parts and/or using polymer materials. For the companies that provide the service of elimination of annoying noises inside the cabin, however, this dependence on such specially trained people is a problem.
As an alternative, there is a proposal to develop a system that measures the noise patterns inside the cabin during dynamic tests, subtracts the engine noise, identifies the annoying noise, obtains the frequency spectra of the detected noise by using the Fourier and Hilbert transforms, compares these spectra with the spectra of the most frequently annoying noises found inside cabins of vehicles and, finally, reports the potential sources of the annoying noise detected.
This paper describes the system assembled for this purpose and its performance in the elimination of annoying noises inside vehicle cabins.