Fault detection and diagnosis of diesel engine valve trains. (May 2016)

• Record Type: Journal Article
• Authors: Flett, Justin; Bone, Gary M.
• Is Part Of: Mechanical systems and signal processing. Volume 72/73(2016)
• Page Start: 316
• Page End: 327
• Publication Date: 2016-05
• Subjects: Diagnostics of mechanical systems -- Engine fault diagnosis -- Signal processing -- Condition monitoring -- Automotive research; Structural dynamics -- Periodicals; Vibration -- Periodicals; Constructions -- Dynamique -- Périodiques; Vibration -- Périodiques; Structural dynamics; Vibration; Periodicals; 621
• DOI: 10.1016/j.ymssp.2015.10.024
• Languages: English
• ISSNs: 0888-3270

Abstract

Abstract: This paper presents the development of a fault detection and diagnosis (FDD) system for use with a diesel internal combustion engine (ICE) valve train. A novel feature is generated for each of the valve closing and combustion impacts. Deformed valve spring faults and abnormal valve clearance faults were seeded on a diesel engine instrumented with one accelerometer. Five classification methods were implemented experimentally and compared. The FDD system using the Naïve-Bayes classification method produced the best overall performance, with a lowest detection accuracy (DA) of 99.95% and a lowest classification accuracy (CA) of 99.95% for the spring faults occurring on individual valves. The lowest DA and CA values for multiple faults occurring simultaneously were 99.95% and 92.45%, respectively. The DA and CA results demonstrate the accuracy of our FDD system for diesel ICE valve train fault scenarios not previously addressed in the literature. Highlights: Accurate results are produced for fault scenarios not previously studied. Naïve-Bayes classification method produced the best overall performance. We introduce a novel feature for quantifying the valve and combustion impacts. Combinations of valve spring and clearance faults are investigated experimentally. We develop a fault detection and diagnosis system for diesel engine valve trains.