Helical gears pair optimal design using Genetic Algorithms

Nowadays environmental concerns and the constant demands of obtaining efficient designs when tackling complex structural problems forced mechanical engineers to use the optimal design methodology. In this paper, a Genetic Algorithm is used to minimize the mass of a helical gear pair mechanism - a key component of an Electric Continuously Variable Transmission - eCVT used as automatic power transmission of various hybrid cars. Designing such mechanisms is an iterative process in which it is necessary to make some tentative choices (for example, the impact of a certain choice of gear width and center distance may yield a minimum mass gearing, but the selection of these values for that parameters may cascade through subsequent steps of the gear design process to ultimately lead to a helical gearing that does not meet the specific strength or manufacturing conditions). The objective function considered here i.e. the mass of the helical gears pairs is computed exactly through a unique method developed by the authors. The optimal design problem is described by a set of 5 design variables (i.e. integer, discrete, and real), and the objective function is subjected to a set o 17 highly non-linear constraints. The results obtained by using GA show significant improvement over the results obtained by traditional design (a 24.21% mass reduction).