| Design of planetary gearsets is significantly more complicated than their parallel axis counter parts as they have multiple planet branches meshing simultaneously with the ring gear and the sun gear as well as the complicated kinematics that carrier's relative motion introduces. Double-planet planetary gears are further complicated to design but addition of a second planet gear in each planet branch allows for greater flexibility in sizing and synthesizing transmission kinematic chains. The key system-level design factors for these gearsets are transmitted torque, speed ratio, and available space. When the basic design factors and strength criteria are met, little effort is made to further improve these systems as the manual design process of gearsets fails to fully assess all performance related aspects of the system. Moreover, the design process does not assess the dynamic response of the double planet planetary gearsets (or any gearset type) which may be critical depending on the duty cycle they are subject to. A statistical design optimization process is needed to systematically improve gearset designs resulting from manual design process by assessing key design configurations within some logical design space.;Computational methods can be used to simulate gear designs and assess their performance, removing the need to construct test systems to extract performance data. However, most computational models are computationally expensive to run for each potential design change in a limited time. In this study, a system-level design, analysis and optimization approach that is computationally strategic and efficient is proposed. The proposed approach utilizes a combination of gear analysis methods to assess the performance of existing double planet planetary gearsets and determine if improvements through design adjustments can be made. A finite element based computational tool as well as a lumped parameter dynamic model of double planet gears are developed and used to simulate candidate modifications to existing gearset designs. Design of experiments (DOE) is used to reduce the number of analyses run in order to reduce the overall computation time. This methodology is specifically developed but not exclusive to double planet planetary gearsets and can be applied to any gearset design including to single planet planetary gear sets. |
