Sound Datasets of a Rolling Element Bearing under Various Operating Conditions

Welcome to the FSTF Mechanical Laboratory at Sidi Mohamed Ben Abdellah University.

Within this repository, we present a comprehensive collection of ball-bearing test data encompassing both normal and faulty bearings across diffrent operating speeds. The experimental procedures were conducted utilizing a specialized test rig rotor system, with sound data meticulously measured at proximal and distal locations relative to the housing bearings. These web pages serve as a unique resource, meticulously documenting the precise test conditions of the motor and detailing the bearing fault status for each conducted experiment.

The datasets are bifurcated into two distinct categories: the first category entails recordings obtained using a stethoscope, denoted as datasets 1, while the second category comprises recordings conducted without a stethoscope under diffrent operating conditions, referred to as datasets 2. In both cases, bearings—specifically SKF model 6004, grooved ball bearings—were installed in a rotor system specified in Appendix A, as provided by the Gunt Company [1]. The defective bearings featured artificial damage in the inner race, outer race, and ball element, representing localized defects, as well as a bearing afflicted with a looseness defect due to prolonged operation.

Data acquisition procedures involved the recording of sound signals emanating from the bearings under diverse conditions, employing a mobile application named VibroTeak, developed by our team. This application facilitates the recording of sound signals in (.wav) format at a sampling rate of 44.1 kHz. Subsequently, the sound data was imported into Matlab using the "audioread" function for further processing. Figure 1 (Appendix A) illustrates the acquisition of acoustic signals utilizing a stethoscope connected to a smartphone via the jack input.

The datasets presented herein offer a novel avenue for researchers to delve into, assess, and propose innovative diagnostic algorithms tailored to this specific type of data.

References:
[1] Gunt, “PT 500.12 Roller bearing faults kit.” https://www.gunt.de/en/products/mechatronics/machinery-diagnosis/roller-bearing-faults-kit/052.50012/pt500-12/glct-1:pa-148:ca-77:pr-1030.

搜集汇总

背景与挑战

背景概述

该数据集由FSTF机械实验室提供，是一个关于滚动轴承在不同运行条件下的声音数据集合，包含正常和故障轴承（型号SKF 6004）的声音记录，故障类型涵盖内圈、外圈、球体局部缺陷以及松动缺陷。数据分为两类：使用听诊器采集的数据集1和不使用听诊器在不同运行条件下采集的数据集2，采样率为44.1 kHz，格式为.wav，旨在支持研究人员开发和评估针对此类数据的诊断算法。