Research overview

Spur gear tooth modelling

Spur gear cracked tooth modelling

Time-Varying Mesh Stiffness (TVMS) Model

Developed analytically improved model ofTVMS(IAM-TVMS) of spur gear pair by considering misalignment of gear root circle and base circle with accurate transition curve, non-linear Hertzian contact stiffness, revised fillet-foundation stiffness by incorporating structure coupling effects of nearby loaded tooth.
Developed IAM-TVMS model for various faults on spur gear i.e., cracked tooth, chipped tooth, and missing tooth.
Developed improved cracked model of spur gear by taking parabolic curve as a limiting line with straight line crack propagation path.
Developed IAM-TVMS model of carburized spur gear pair by incorporating coating of thickness ‘t’ on the outer surface of gear.

A unified EM modeling of a single-stage spur gear with cracks at tooth root is developed by applying modified Lagrangian approach on whole EM system.
The EMmodel is lashed with IAM-TVMS model and improved crack model for accurate analysis.
The input torque is calculated by differentiating the kinetic energy of the EM system with respect to motor’s angular displacement.
The effect of various crack depth on EM system are studied by motor current.
Cepstrum analysis is done on residual current and vibration signal to check the effectiveness of the developed model.

First time the TVMS for carburized spur gear pair including tooth root cracks are formulated.
The carburized part, in combination with soft core, is expected to form sandwich structure.
Effect of carburization is considered in the IAM-TVMS model.
Hertz contact for coated gear pair is also considered in the IAM-TVMS model.
Comaparison is done between uncarburized and carburized gear to show the effect of carburization on gear pair.
Carburized gear pair has more stiffness as compared to uncarburized gear pair.
The spectral entropy successfully demarcate the uncarburized and carburized gear pair.

Developed an AI-based non-parametric filter approach for gearbox fault diagnosis which has the ability to give higher classification accuracy on those signals which are acquired at an affordable sampling rate with a comparatively smaller number of samples per fault condition.
Proposed new features based on the Logarithmic amplitude of higher order moment (LHOM) for gearbox fault diagnosis.
Studied the effect of sampling rate on parametric and non-parametric type data pre-processing on gearbox fault diagnosis.
Studied the importance of non-parametric filter on deep learning based gearbox fault diagnosis.

Flow chart of proposed technique for gearbox fault diagnosis