Preliminary case studies to acoustics-based condition monitoring of industrial fans

Authors

Dominik Tóth  - Budapest University of Technology and Economics, Faculty of Mechanical Engineering, Department of Fluid Mechanics, Hungary
Tamás Divinyi-Zsoldos - Budapest University of Technology and Economics, Faculty of Mechanical Engineering, Department of Fluid Mechanics, Hungary
Lendvai Bálint - Budapest University of Technology and Economics, Faculty of Mechanical Engineering, Department of Fluid Mechanics, Hungary
János Vad - Budapest University of Technology and Economics, Faculty of Mechanical Engineering, Department of Fluid Mechanics, Hungary

Abstract

The application of the Industry 4.0 approach to industrial air technology and ventilation results in the foreseen realization of smart fans operating in smart factories. Smart industrial fans are envisaged to be equipped with smart features; among others, with continuous condition monitoring, based on advanced processing of measurement data acquired on the fan on the site of application. One possibility is acoustics-based condition monitoring, in which patterns of noise radiated by the fan are gathered, processed, and compared to reference data associated with normal fan operation. The discrepancy between the actually measured and reference patterns are considered as potential signatures of extraordinary fan condition, thus aiding the demand-based maintenance of the fan. This paper serves for adding to the experimental background of acoustics-based condition monitoring. The subjects of the presented case studies are representative fans of radial and axial flow type, in installations corresponding to real-case industrial layouts. The fans are either in an isolated layout, connected to inlet and outlet ducts; or are incorporated in a compact equipment of Heating, Ventilation and Air Conditioning (HVAC). The noise data are acquired out of either the inlet or the outlet duct attached to the machinery, as tailored by the acoustic accessibility of the particular industrial fan installation. Noise-pattern modifying effects caused by the ductwork as well as by the surrounding HVAC equipment are intended to be considered in data evaluation. Special duct installations, having future perspectives in acoustics-based condition monitoring as well as in low-noise, high-efficiency air technical systems of smart factories are studied, and are compared to traditional metal ducts from the viewpoint of acoustic monitoring. Such special installations incorporate acoustically transparent duct segments, and textile air duct systems. The aim of the paper is as follows. a) Making a comparison of actually measured and processed sound pressure spectra to reference spectra corresponding to normal fan condition. b) In such comparison, assessing the detectability of acoustic signatures of extraordinary fan operation. In the presented case studies, the extraordinary scenarios represent fan operational points of drastic throttling; or incorporate an improperly fixed, and thus, extremely vibrating machine element inside of the machinery. Practical aspects of microphone arrangement, as well as data acquisition, processing, and evaluation are discussed in detail. By such means, the paper intends to serve with first-hand empirical guidelines in execution of acoustics-based condition monitoring of fans in real industrial installations, and to initiate applied research responding to the challenges and questions arisen in the truly practical, industrial approach applied herein.