Abstract

The tooth contact pattern is a crucial indicator of overall performance for many gear geometries, especially spiral bevel and hypoid gears. It is directly associated to the loaded function of transmission errors, a primary contributor to gear noise and vibration. This is especially relevant in electric vehicles, where gear noise is more pronounced due to the absence of masking noise from internal combustion engines. Consequently, the inspection and control of the contact pattern are essential aspects of gear design and manufacturing. Traditionally, contact patterns are evaluated using a manual process known as the tooth contact roll test. This test is time-consuming, labor-intensive, and relies upon visual inspection by skilled operators. For these reasons, developing an efficient and effective automated inspection process has been a recurring focus of interest for several decades, as it can simultaneously speed up the process and make it more robust. Despite the many efforts and innovative approaches, none have succeeded in gaining popularity. This dissertation revisits the problem with a novel approach centered around leveraging information about the gear and the fact that visible contact patterns are influenced by subtle geometric deviations. The goal is to develop an automated framework capable of analyzing tooth contact patterns with the accuracy and insight of a human inspector. Designed for real-world factory conditions, the framework aims to offer reliable, repeatable performance under varying environmental factors. Experiments have been conducted using RGB images and RGB-D point clouds of gears showing actual contact patterns. These tests were performed on the production floor under uncontrolled conditions. The methods involve using information about the model, the data, or both to transfer a representation of the contact patterns back to the digital twin or provide a standardized output for evaluation. Results demonstrate the practical use of the proposed methodologies, showing that this is a challenging but solvable engineering problem.

Publication Date

5-14-2026

Document Type

Dissertation

Student Type

Graduate

Degree Name

Mechanical and Industrial Engineering (Ph.D)

Department, Program, or Center

Mechanical Engineering

College

Kate Gleason College of Engineering

Advisor

Alfonso Fuentes-Aznar

Advisor/Committee Member

Hermann J. Stadtfeld

Comments

This thesis has been embargoed. The full-text will be available on or around 6/26/2027.

Campus

RIT – Main Campus

Download

Share

COinS