A Study of Fault Diagnosis and Prognosis Based on Condition Monitoring for Centrifugal Pump

碩士 === 國立成功大學 === 機械工程學系 === 107 === Centrifugal pumps are used for many different purposes in a wide variety of industries, such as power generation plants, chemical processing plants, water supply, heating and cooling distribution systems. Centrifugal pumps often play an important role in the enti...

Full description

Bibliographic Details
Main Authors: Huan-ChunChen, 陳奐均
Other Authors: Tian-Shiang Yang
Format: Others
Language:en_US
Published: 2019
Online Access:http://ndltd.ncl.edu.tw/handle/pkdvnn
Description
Summary:碩士 === 國立成功大學 === 機械工程學系 === 107 === Centrifugal pumps are used for many different purposes in a wide variety of industries, such as power generation plants, chemical processing plants, water supply, heating and cooling distribution systems. Centrifugal pumps often play an important role in the entire production or process chain, so timely detection of faults is crucial to enhance the production level as well as reducing unexpected cost. In this study, a monitoring system is constructed to timely keep track on the status of the pump, including the flow rate, pressure, temperature, current, and vibration. In addition, clogging experiments and impeller experiments are designed based on the failure causes and effects. There are three kinds of experiments conducted in this study. The first experiment is to mimic the clogging status by shrinking the cross-sectional area of outlet, which can equivalent to the rises in flow resistance. The second experiment is to discuss the degrees of wear impeller blades. The final experiment is to consider the unbalance effects by grinding different number of blades. By collecting the data from sensors and combining with the pump domain knowledge, fault diagnosis and prognosis can be made to deal with the faults. Furthermore, the criterion of maintenance or replacement strategies can be formulated for users to follow, thus helping the centrifugal pump work properly to reduce accidental costs.