Power Loss in Planetary Gearboxes Including the Influence of Gear Elastic and Dynamic Effects

In recent years, environmental and sustainability issues have acquired major importance, propelling the investments in renewable energy sources. Worldwide e orts have been made to increase the energetic and operational e ciency of equipments. Many machines have a motor or some other kind of propelling mechanism that have an input that rotates at xed speed or in a range of speeds that is not the desired one. In many situations in order to overcome these issues gearboxes are used. Designing a transmission with the care of maximizing e ciency while keeping proper operating safety factors has become a necessity. The main purpose of this work was to study the in uence of operating conditions and gear oil formulation in the e ciency of meshing gears (applied to a planetary gearbox) including elastic and dynamic e ects. Four wind turbine gear oils were selected, characterized and gearbox e ciency tests were performed in a gearbox test rig with recirculating power. A no-load power loss gearbox test rig was also developed, so that both load and no-load power loss measurements were performed. A numerical classical gearbox power loss model was developed aiming to understand the in uence of each component in the gearbox power loss. At nominal load the most important power loss sources are the gears. An accurate gear load loss prediction is quite dependent of the gear tooth load sharing model that is considered. Three quasi-static gear load sharing models were developed for both internal or external spur and helical gears. Aiming to understand the in uence of gear dynamics in gear power loss, a four degree of freedom lumped mass gear dynamics model accounting, for time varying mesh sti ness, friction and damping, was also developed. The ndings and the models developed for a simple gear pair were applied to study the power loss behaviour of a planetary gearbox. The experimental tests in the planetary gearbox showed that di erent lubricants promote very di erent power loss behaviours. The more re ned solutions for the load distribution model proved to be fundamental in obtaining accurate power loss predictions. The dynamic power loss model results have showed that average power loss can be a ected by gear dynamics.