Publication Date

5-8-2025

Document Type

Student Project

First Advisor

Pohlman, Nicholas

Degree Name

B.S. (Bachelor of Science)

Department

Department of Mechanical Engineering

Abstract

Over prolonged periods of time, mechanical machinery requires regular maintenance to ensure optimal performance. Conventionally this maintenance is done on a regular schedule with predefined procedures to repair all aspects of the machine. As machine learning models become more applicable in everyday life, these algorithms can be leveraged to predict failures in mechanical devices. This project worked toward creating a predictive maintenance machine learning model for centrifugal pumps that can identify and predict discrete failure modes. As a result of the project, a pump loop setup was constructed. The purpose of this report is to provide a guide for how the setup can be run in its current form. Additionally, information on changes that could be made to the setup to be used as an instructional tool for lab experiments is present. The framework for such a lab experiment that has been drafted is also included, providing a starting point for a future fluid mechanics lab.

Publisher

Northern Illinois University

Download

Share

COinS