There is a growing demand for compressed air in the industry for various applications. Majority of industrial requirements is in line with screw compressor operating range. Design and construction of screw compressors are demanding tasks that require advanced calculations and theoretical knowledge. Clearances play a major role in the performance and reliability aspects of a screw compressor. Seals are provided in compressors to fit around rotor shafts in order to prevent the leakage of lubricating oil and working medium. However there is a small clearance between the seal and rotor shaft, which can cause potential leakage of the working medium. The performance of the compressor is directly related to the leakage rate through the seals. The labyrinth seal is a special type of seal, used in screw compressors and turbo-machinery for sealing purpose. Labyrinth seal is a non-contacting type seal that uses a tortuous path to minimize the gas leakage. The pressure drop occurs at each labyrinth tooth as the medium is squeezed between the labyrinth tooth and the rotor. The leakage through the seal is directly related to the labyrinth profile and also the clearance between the rotor and the labyrinth tooth. The present work is carried out to reduce the leakage through the labyrinth seal by optimising the tooth profile and operating clearances. Heat transfer analysis is carried out on the housing of the labyrinth seal to find out the boundary temperature of the seal. Also the heat transfer analysis on the labyrinth seal followed by Thermo-structural analysis is carried out to find out the accurate operating clearance of the seal. By using CFD as a tool, the optimisation is carried out on different design configurations of labyrinth seal by comparing the deviation in leakage rates. Effect of rotor speed, width of seal and pressure ratio on air leakage rate is also investigated. A set of labyrinth seals has been designed based on the above optimisation and tested in the compressor. The results have been compared with the CFD prediction.

This content is only available via PDF.
You do not currently have access to this content.