Numerical and experimental analysis of the compressor shell

On the world market, refrigerator manufacturers are very actively competing with each other in terms of quality, new technology, long-term storage of food, interior and exterior design of refrigerator, energy and sound power levels. The competitions between the companies lead to continuous research activities to meet the consumer expectations while keeping the costs of the products as low as possible. Controlling the noise and vibration levels of a refrigerator compressors is essential in order to control the noise emission and to comply with the noise regulations and consumer demands. Vibrations in the compressor is one of the most important causes of noise in the refrigerator. Especially the vibrations generated by the mechanical movement in the compressor and the noise generated as a result of this vibration transmitted to the compressor shell in various ways affect the overall noise level quite significantly. The excessive vibrations of the refrigerators not only increase the unwanted noise which adversely affects the comfort of the customer, but also increase the failure rate of the refrigerators. In this thesis, fundamental information about refrigerators is gathered, dominant sources of noise in the refrigerators is determined, and a general literature search is conducted on noise sources in refrigerators. It is seen that noise and vibration problems in refrigerators leads researchers to develop new ways of designing quieter systems with less vibrations. Although there are various methods of reducing the noise and vibration problems in refrigerators, the reduction of vibration amplitudes by using friction dampers for the compressor shells has not found sufficient attention among researches. Therefore, in this thesis, it is aimed to reduce compressor vibration and noise levels by using friction dampers. For this purpose, single and multi-degree of freedom systems are examined and information about basic sound and vibration issues are examined. First of all, the current situation needs to be established in order to create a reference case and make the level of improvements via planned studies in this thesis more meaningful. . Therefore, numerical models of the compressor are created by Finite Element Method. The upper and lower parts of a compressor shell, forming the outer casing of the compressor, are modelled and analyzed individually first and then the whole compressor shell, i.e. the assembly of the upper and lower parts, are modelled and analyzed. Natural frequencies, mode shapes and frequency response functions of the compressor shell models (upper and lower part as well as the compressor shell assembly) are obtained. The vibration and noise characteristics of the compressor is identified experimentally. First, experimental modal analyses of the upper and lower parts of the compressor are carried out via hammer testing. The measured FRFs are processed to identify the natural frequencies and mode shapes using ICATS software. After this, to determine the current noise level of the compressor, the solo sound power levels of the two different types of compressors are measured. Then, this compressor is positioned in its place at the back of a refrigerator and the total sound power level of the refrigerator is measured. These metrics constituted the reference state for determining the level of improvements by various methods investigated within the scope of this thesis. The last experimental method utilized here for determining the current state of the compressor was the acoustic camera measurements which is performed to determine the noise source around the compressor shell in the refrigerator. This measurement leads to the identification of the noise sources due to the compressor and frequency bands where the noise levels were high. Some of the predicted FRFs for the compressor structures modelled in this thesis are compared with their experimental counterparts in order to determine the level of consistency of the results obtained from the analysis. Results confirmed acceptable level of correlation between the predicted and measured behavior. Then, the natural frequencies and mode shapes obtained for the upper and lower shells of the compressor is compared. Even though there were some differences in the predicted and experimentally identified natural frequencies, especially for the lower shell, both the natural frequencies and the mode shapes were found to be compatible with each other. The vibration behavior of the compressor under operating conditions is also investigated experimentally. Experimental results is obtained by using the so-called Operational Deflection Shapes (ODS) method. The most important advantage of this method is that no artificial excitation needs to be applied to the structure or force signals need to be measured. All the estimations in ODS method are based on the response signals only. It is found that the ODS results obtained by this method using the measurements over the upper half of the whole compressor are similar to some of the mode shapes obtained from numerical models. Considering all the correlation studies, it is concluded that, for the purpose of this thesis, numerical and experimental models is correlated with other adequately. In order to provide some additional damping to the system, it is aimed to introduce some additional frictional contacts to the compressor at critical regions suggested by numerical and experimental analyzes of the compressor shell. In this context, first of all, a friction damping device is fixed at a predetermined location on the compressor shell and both the solo compressor sound power level and the refrigerator total sound power level were measured. With this method, there was a reduction of 1.6 dBA in the solo sound power level of the compressor and 0.3 dBA in the total sound power level of the refrigerator. Fairly low level of reduction in the total sound power level of the refrigerator is attributed to the higher level of contribution of the noise at relatively low frequency bands. After this, an alternative solution is proposed and the damper force is designed to be adjustable in order to determine whether this friction damper can be used more effectively. The use of this adjustable friction damper at two different positions (i.e., using two friction dampers) resulted in a reduction of 2.1 dBA in the solo compressor sound power level and reducing the vibration amplitude of the shell. In order to be able to provide additional vibration damping to the compressor shell, another method of providing additional frictional contacts by means of a belt-shaped strip metal across the compressor circumference is utilized. In this application, the strip metal is used in the connection region of the lower and upper shells to create dry friction contact area. This method is applied to two different types of compressors and the total solo sound power levels of a compressors are reduced by 1.5 dBA and 1.8 dBA. As expected, providing additional damping also reduced the vibration levels of the compressor shell. In addition to the alternative solutions explained above, the usage of pressure rings over the compressor shell surface is recommended for providing additional damping in various regions of the compressor shell surfaces. By applying this method to the compressor shell, higher levels of damping hence vibration reduction is achieved compared to the other methods investigated in this thesis, Another method suggested in this thesis for controlling compressor shell vibrations in certain frequency bands is via the use of so-called multilayered compressor body (shell). In layered compressor shell designs, it is especially desirable to reduce noise and vibration by frictional contacts between multilayered shell surfaces. In the scope of the thesis, the upper shell of the prototype compressor was produced in two layers. Even with only the upper shell of the compressor being manufactured in two layers, a reduction of 1.8 dBA in compressor noise power level is achieved.