Turning Dataset for Chatter Diagnosis Using Machine Learning

Published: 29 May 2019| Version 1 | DOI: 10.17632/hvm4wh3jzx.1
Contributors:
,
, Melih Yesilli

Description

This dataset corresponds to sensor signals from several cutting tests including two perpendicular single axis accelerometers, a tri-axial accelerometer, a microphone, and a laser tachometer. We post both the raw data from all the sensors as well as the conditioned and tagged data from one channel of the tri-axial accelerometer. The tags that we use are: no-chatter, intermediate chatter, chatter, and unknown. The cutting test is performed by turning an Aluminum 6061 workpiece on a Clasuing-Gamet 33 cm (13 inch) engine lathe using a 0.04 cm (0.015 inch) radius Titanium nitride coated insert attached to an S10R-SCLCR3S boring bar. We provide a picture of the experimental setup in the included documents. Data from four different cutting con figurations were collected where each cutting con figuration depends on the stickout distance, which is the distance from the heel of the boring rod to the back face of the tool holder. Four stickout distances were considered: 5.08 cm (2 inch), 6.35 cm (2.5inch), 8.89 (3.5inch), and 11.43 cm (4.5 inch). For each stickout distance, we collect data for several combinations of the rotational speed and depth of cut. We include a table showing the parameter combinations and the number of the tagged time series corresponding to each stickout length in the included documents. We also briefly describe the algorithm we used to tag the data.

Files

Institutions

Michigan State University, Technische Universitat Chemnitz

Categories

Machine Learning, Turning, Machining, Time Series

Licence