Non-Lubricated Bevel Gear Train Optimization using Heuristic Genetic Approach

This research investigated the problem of optimizing bevel gear design in a non-lubricated condition to transmit power between the perpendicular shafts at high speed. The bevel gear design problems can be viewed as a kind of NP hard problem, with standardized parameter values and several additional constraints. In this research, heuristic genetic evolutionary approach has been adopted in two phases for solving this NP-hard design problem. The first phase consists in building good initial solutions followed by genetic operations. The second phase improves the obtained genetic solution by checking with the constraints and the standard values as per the design standards. Results obtained on real data sets are compared with the traditional method and found the results are at par. The special criteria taken for consideration are that there is no lubricating oil and noise should be least possible.