Nowadays, the “non-demountable” trend of IC engine diagnostics is becoming ever more widespread. One of the techniques used in such diagnostics is vibroacoustic monitoring. It is used in safety and ecology, but not in in car workshops. Keeping in mind that this method may be used for all kinds of mechanisms, vibroacoustic diagnostic tool for IC engine has a great potential.
Vibroacoustic technique is based on sound analysis. Traditional approaches to analysis of physically realized signals, patterns and time series generated by complex polymodal dynamic systems are generally based on statistical methods, and more often on spectral correlation methods and their variations. And here, the dynamic nature of the processes that generate signals is made insignificant by the statistical approaches applied to analysis of such signals. Only a dynamic approach to complex system analysis characteristic of the modern nonlinear dynamics permits factoring in signal analysis in the process of identification of dynamic systems according to test data analysis results.
Considering this, author developed a fault detection algorithm. At the first algorithm stage, sound waves are recorded with a vibration sensor, and readings of a cylinder pressure gauge and of the first cylinder spark plug are taken for use in synchronizing vibration sensor signals. Afterwards, the frequency spectrum of the sound signal recorded is divided into intervals. This completed, the next signal interval is selected and expanded in a Fourier series. Proceeding from the readings of a knock sensor one must determine whether the signal is phased or not. If it is, the surge phase is determined, and after that the signal source is forecast and a wavelet transform is applied to the signal. If it is not, then a wavelet transform is first applied to the signal, and then, as per the data obtained, the peak amplitude frequency is determined.
Proposed approach enables the determination of the following parameters: misfires, crankshaft angular position and defects in main IC engine mating pairs. Experiments, carried out using for Lada 2110, 1.5l 16 valve engine were a success. Vibroacoustic diagnostic proved itself worthy for valve train diagnostics, cylinder-piston group diagnostics and diagnostics of roller bearings. Afterwards the parameters obtained can be used for simplified diagnostics of engines with similar geometrics. Thus, the experimental findings were easily applied to a 1.5L Hyundai Accent engine.
A mobile device for vibroacoustic diagnostic is being developed. At the same time author researches possibilities of vibroacoustic diagnostic application to other parts of the car, such as transmission and suspension.