Fault detection and diagnosis of historical vehicle engines using acoustic emission techniques

Abstract

<span lang="EN-GB">The reactivation of artefact mechanisms is always a challenge for conservators. Non-invasive diagnostic techniques, applicable directly on the artifacts, allows for performing early-stage diagnostics and avoiding damage. The Acoustic Emission Monitoring of Historical Vehicles (ACUME_HV) project represents the first use of acoustic emission (AE) as a non-invasive technique for the diagnostics of historical vehicles. The aim of this project is to develop an objective, human-independent method. This will help museum personnel to make decisions regarding the reactivation of historical vehicle engines using measurements and data analysis rather than merely personal experience. Herein, we present the results of the first phase of the ACUME_HV project, which was focused on the development of a protocol for the use of AE during cold tests.</span>

Data analysis techniques for the visualization and classification of historical vehicle engines’ health status using data-driven solutions

“…In this paper, we investigate the use of data analysis techniques for the visualization and classification of the engine health status using data-driven solutions such as supervised machine learning. The paper focuses on the analysis of vibro-acoustic signals for the characterization of engines in different conditions (data presented in previous publications [17][18][19][20][21]). Research questions studied in this article focus on the use of AE signals and machine learning, and can be summarized as: 1) identify the engine that produces the signal; 2) determine if an engine has an anomaly; 3) determine which engine a record belongs to and if it has an anomaly.…”

“…The developed methodology and the results of the ACUME_HV project have been previously published in different journals and conference proceedings [17][18][19][20][21]. The ACUME_HV project showed that acoustic emission can be used to establish the state of preservation of historical engines before their reactivation and to detect the most common faults [17][18][19][20][21].…”

“…The developed methodology and the results of the ACUME_HV project have been previously published in different journals and conference proceedings [17][18][19][20][21]. The ACUME_HV project showed that acoustic emission can be used to establish the state of preservation of historical engines before their reactivation and to detect the most common faults [17][18][19][20][21]. Obtained AE data can be post-processed and interpreted in different ways, based on the extraction of specific features, such as peak amplitude, duration, rise time, frequency, and energy [14].…”

“…Obtained AE data can be post-processed and interpreted in different ways, based on the extraction of specific features, such as peak amplitude, duration, rise time, frequency, and energy [14]. In the first phase of the project, the data analysis was limited to the detection and evolution over time of AE events' amplitude (also called "hits"), as a function of the rotational speed of the engine [17][18][19][20][21]. Those events are produced by the movement of the different contact pairs during the functioning of the engine and are directly related to the thermodynamic cycle of the engine itself (fuel injection, ignition, expansion, and exhaust) [17][18][19][20][21].…”

“…In the first phase of the project, the data analysis was limited to the detection and evolution over time of AE events' amplitude (also called "hits"), as a function of the rotational speed of the engine [17][18][19][20][21]. Those events are produced by the movement of the different contact pairs during the functioning of the engine and are directly related to the thermodynamic cycle of the engine itself (fuel injection, ignition, expansion, and exhaust) [17][18][19][20][21]. Interpretation of the data in this phase was limited to combining the obtained measurements with a good knowledge of the system and with the expertise of conservation professionals.…”