Measurement xxx (xxxx) xxx Please cite this article as: Rajiv Tiwari, Measurement, https://doi.org/10.1016/j.measurement.2020.108676 Available online 2 November 2020 0263-2241/© 2020 Elsevier Ltd. All rights reserved. Blockage and cavitation detection in centrifugal pumps from dynamic pressure signal using deep learning algorithm Rajiv Tiwari * , D.J. Bordoloi , Aakash Dewangan Department of Mechanical Engineering, Indian Institute of Technology Guwahati, Guwahati 781039, India A R T I C L E INFO Keywords: Centrifugal pump Blockage Cavitation Pressure signature Modulating valve Deep learning Binary data classifcation ABSTRACT Centrifugal pumps play an important role in many industrial applications even in harsh environment for prolong duration. High effciency with very low power consumption makes them very popular in industry. However, during their operation, they may fail due to some operationally developed faults, which may subsequently lead to the interruption in the continuous operation of pumps. Therefore, monitoring the health status of the centrifugal pumps is essential to prevent unwanted stoppage, which may further lead to the breakdown of the whole system. The main focus of this study is to propose a methodology to identify the presence and severity of blockages, and cavitation in the centrifugal pump using fuid pressure, which is very vital for fuid related faults. To simulate the blockage in the pump, the fow area of the suction pipe is restricted by dividing into six equal intervals (i.e., 0%, 16.7%, 33.3%, 50%, 66.6% and 83.33%) using a mechanical modulating valve. Due to blockage and cavitation, the main parameter which directly gets affected is the fuid dynamic pressure. Hence, in the present study, pressure signatures were captured at different blockage levels and at different running speeds with the help of a pressure transducer, which was mounted on the circumference of the centrifugal pump casing. Deep learning based binary data classifcation methodology is used to classify the data acquired from the pressure transducer. To get better performance of the data classifer, statistical features are extracted from time domain pressure signals. In order to identify the severity of the faults, binary classifcation of the data is performed at different blockage levels and running speeds. Finally, based on the results obtained from the classifer, existence of the faults (i.e., blockage and the cavitation), their severity levels are presented. 1. Introduction Pumps are very popular hydraulic machinery in the industrial and domestic applications. Most of industrial pumps are centrifugal pumps, when it comes to high volume of the fuid to be handled. They are crucial part of many industries, to regulate the fow of processes [1]. Pumps are used in the power plants, oil–refneries, building services, industry and water engineering, chemical and process industries, and biomedical applications [1]. Continuous use of centrifugal pumps at high speed without any maintenance work results in failure of its mechanical components and sometimes causes serious damage to the whole assembly. Faults can be mechanically induced, system faults, operationally developed or any combination of these. Mechanically induced faults occur due to faulty parts of the pump, viz. the bent rotor, misalignments, and bearing faults. Faulty installation and leakage type faults are system faults. Opera- tionally developed faults are the ones, which develop while the pump is running, viz. blockage, cavitation and fow related problems [2]. Typical faults that frequently develop in the pump are cavitation (due to pres- sure of the liquid below vapour pressure), casing and impeller damage (due to pressure pulsations associated with internal recirculation), blockage of the suction and discharge pipe (due to presence of the solid particles and impurities in the pumping liquid), installation faults, dry run, entrained vapour (due to vortex in the tank, insuffcient water source or incorrect net positive suction head conditions) [2]. These faults can develop at any stage regardless the condition of the pump. Therefore, in the present work, blockage fault in the suction pipe and cavitation in the centrifugal pump are studied. In centrifugal pumps, radially outward movement of the fuid is due to rotating blades of the impeller. The movement of the fuid results in fall of the pressure in the suction pipe of the pump. When pressure falls below the vapour pressure of the fuid, it causes bubble formation. This phenomenon is called the cavitation [3]. When fuid comes in the high- pressure zone, bubbles collapse and generate shock waves. Cavitation * Corresponding author. E-mail address: rtiwari@iitg.ac.in (R. Tiwari). Contents lists available at ScienceDirect Measurement journal homepage: www.elsevier.com/locate/measurement https://doi.org/10.1016/j.measurement.2020.108676 Received 4 July 2020; Received in revised form 7 October 2020; Accepted 26 October 2020