Computer Aided Optimal Design of Helical Gears

A random search method for optimum design of a pair of helical gears has been developed. The sequence of optimization consists of two principal components. The first is the selection phase, where the output is the starting solution for the design variables - module, facewidth, helix angle and number of teeth on the pinion. The input for the selection phase includes the application environment, approximate center distance, minimum helix angle, desired values of gear ratio, pinion speed and the power to be transmitted. The limits on each of the design variables and the constraints are imposed interactively during the first phase. A standard tooth form is assumed for the design. Standards published by the American Gear Manufacturers Association are employed for the design process. The second phase consists of the implementation of the optimization procedure to find the minimum weight. The method employs a random number as the search direction with the step size being altered based on the value of the constraints. A number of random directions are generated and a minimum in each of those directions are determined to form a set of feasible solutions. The optimum solution is then determined from the set of feasible solution. Graphs are presented during optimization to create a user interactive environment. The program generates a complete set of manufacturing data for the designed gear