Search Results

Structure borne sound is one of the most important reasons of noise pollution in the automobiles and aircraft's. Noise is mostly generated by the vibrating panels excited by either a mechanical or an acoustical excitation. Examples of the typical vibrating structures in automobiles are engine cylinder, gearbox cover, transmission system covers, panels of the body etc. Sound radiation characteristics are also important in the phenomenon of resonant sound transmission through a panel. Resonant sound transmission occurs because of resonant modes of the panel within the frequency bandwidth of interest. Typical example of resonant sound transmission is the transmission through a firewall of an automobile, which forms the partition between the engine compartment and the cabin interior. Radiation characteristics can be typically defined by radiated sound power, radiation efficiency and space average mean square velocity of the panel.

This paper reports on an investigation into the effects of varying amounts of background noise on sound power measurements using the two microphone sound intensity approach. Two sets of measurements were taken on machine sources; one set at a distance of approximately 0.5m and the other very close to the sources. Results show that the sound intensity method is a viable technique for measuring the sound power of a source and that indicators can be obtained which show when sound power results are unreliable.

In order to identify noise sources in a diesel engine, specifically exciting forces such as combustion and piston slap, the so-called coherence method which utilizes relationships between the auto and cross spectra of cylinder pressure, cylinder liner acceleration and engine noise has been examined. Also, as an alternative, a multivariable regression analysis in one-third octave band auto spectra of each signal mentioned above has been made. It was shown that the simple coherence model studied does not seem useful for this particular type of noise source identification problem. However, the multivariable regression approach has yielded fairly reasonable results, though some problems have been found in accuracy. From this research, it was found that combustion noise is predominant for heavier engine load conditions, though at lower load and high speed conditions, piston slap noise becomes appreciable.

This paper reviews some previous findings about the combustion noise of diesel engines and compares them with the authors' findings. The reason why the noise of most turbocharged diesel engines is more sensitive to speed than load is explained. The findings of this research suggest that the frequency content of the combustion gas pressure up to about 300 Hz is related to the maximum cylinder pressure, Ṗmax. Between about 300 Hz and 2 kHz it is related to the maximum rate of cylinder pressure rise Ṗmax and above about 2 kHz it is related to both the magnitude and the duration of the second derivative of the cylinder pressure P̈. Pmax is more sensitive to load than speed. No correlation was found between Ṗmax and engine speed or load. The variation in the magnitude of P̈ is thought to be a random process, but the frequencies of the cylinder pressure fluctuations are closely related to the cavity resonance frequencies.

Existing techniques for identifying noise sources are reviewed. One such technique (the lead-wrapping approach) was used to source-identify a diesel engine and measure the sound power radiated from the major surfaces. Two new advanced techniques (surface intensity and acoustic intensity) were developed and used to measure the sound power radiated from the same major surfaces. The conventional lead-wrapping and new intensity results were compared and agreement was good. The advantages of the new intensity techniques are described and suggestions made for reducing the noise of this and other similar diesel engines.

THE MAIN AIMS of the experiments reported in this paper are to obtain a quantitative comparison between the noise ranking of engine components using intensity and lead wrapping techniques, and to analyze the causes of the sudden increase in noise level observed during the initial part of the acceleration process. The results indicate that the surface intensity technique agrees well with the lead wrapping sound power results and, although the surface intensity technique in its present state of development, is still time consuming, it does also produce information about the vibration levels of the surface which is essential when the component has to be redesigned to reduce noise. The acceleration test results indicate that the increased, sound pressure level is directly related to changes in the combustion process while turbocharger lag appears to have negligible effect.

A diesel engine can be modeled as a multiple-input, single output system, sometimes known as a coherence model. Theoretical models for multi-input, single output systems are quite well developed, but only recently have attempts been made to apply these models to practical noise cases such as diesel engines. The results of these models can be used to predict changes in engine noise as cylinder pressure - time history is varied or to attempt to identify sources of noise in the engine. Practical difficulties and simplifying assumptions which must be made are discussed.

This paper reviews general methods of identifying noise sources in machinery and in particular it describes in detail two methods which have been studied by the authors on a truck: near field measurements and the coherence technique. The results of these two methods are compared with results of a third method in which the same truck with different sources wrapped and then sequentially exposed was driven past a microphone using the SAE J366b test. The paper also reviews two different stationary indoor standard tests which have been investigated: the reduced drive-by simulation method, and the sound power method. These tests could be used to supplement (or replace) the standard SAE J366b drive-by test in some cases. Comparisons are made of data from the two tests developed and the SAE J366b drive-by data for the same truck.