Design rules to enhance HUMS sensitivity to spur gear faults

This dissertation describes the investigation of spur gear design rules that may enhance the sensitivity of conventional fault metrics to typical gear tooth damage. Spur gears represent a fundamental gear geometry that is commonly used in complex transmission system. This thesis explores the influence of simple spur gear design parameters such as diametral pitch, tooth number and pressure angle on gear tooth fault sensitivity. Using static and dynamic analysis, coupled with vibration based fault metrics (FM0, FM4, DI), this work attempts to determine the influence of RPM level, and gear geometry on vibration signatures.; Specifically, three types of gear damage are modeled. These include pitting, crack and wear damage. Using a Cornell's gear tooth deflection method, a quasi-static nonlinear mesh gear model is developed that includes effects associated with gear tooth damage. This quasi-static model is integrated into a rigid body model of spur gear mesh dynamics under an applied load. This model is used to simulate the gear mesh dynamics under various RPM levels and applied torque loads. Additionally, the filtering effect of bearing is investigated based on a bearing model. Simulated vibration signatures of bearing case with and without gear tooth damage are used as input into conventional vibration based fault metrics. Assuming that for each spur gear design the gear tooth critical tensile or contact stress is held constant, this dynamic model enables one to vary gear design parameters to evaluate their effect on the vibration signature with and without damage.; In this dissertation, the sensitivity of gear design parameters to gear tooth damage is investigated under quasi-static and dynamic loading conditions. Experimental validation is carried out for spur gear designs with various diametral pitch values and for spur gears with increasing crack damage. In the static analysis, enhanced sensitivity is measured with respect to the transmission error. In the dynamic analysis and subsequent experimental validation, conventional fault metrics are used to help assess damage sensitivity to spur gear design parameters. Both simulation and experimental testing confirm that incipient crack damage using conventional fault metrics are insensitive to cracks less than the 50% of the tooth depth. Spur gear designs along with RPM variations have a significant effect on the ability of conventional fault metrics to detect gear tooth damage. Enhanced sensitivity can be achieved for most cases by increasing the diametral pitch of the gear design. However, pitting tends to be insensitive to any gear design changes.