Online Tool Condition Monitoring of CNC Turnings based on Motor Current Signature Analysis

Methodology for Tool Wear Detection in CNC Machines Based on Fusion Flux Current of Motor and Image Workpieces

“…The three most important parameters for machining operations are the depth of cut (d), cutting speed (V) and tool feed rate (f ). The depth of cut is the average between the initial diameter of the workpiece (D o ) and its final diameter (D f ), as shown in (1); the cutting speed is the superficial (tangential) speed of the workpiece, in m/min, as a function of D o and the spindle speed (N) in rpm, as shown in (2); and the tool feed rate is the distance the tool travels in one rotation of the spindle, in mm/rev, and can be defined by the linear feed rate of the tool through the workpiece (v f ) and N, as defined in (3).…”

“…Obtaining information on the condition of the tool through machine elements is ideal due to the relationship between the variables measured on the machine and the evolution of wear; in this sense, the spindle motor current is commonly analyzed in motor current signature analyses (MCSA). Zou et al [3] employed an MCSA to generate an online system from the spindle motor current with a bispectrum signal modulation (BSM) algorithm by identifying the magnitude and phase of the BSM to distinguish between the effects of three levels of wear upon changes in the depth of cut and workpiece diameter of an A3 steel piece, although only one of the obtained features could identify the wear for each depth of cut. On the other hand, Marani et al [4] implemented a feature extraction of current signals where the root mean square (RMS) values of the tests were obtained and a network with a long short-term memory (LSTM) model was used to predict the evolution of tool wear on a coated carbide tool used for the turning of steel alloy, comparing the accuracy of 30 different architectures to obtain the best suited for the proposed experiment.…”

“…Several literature reviews have identified the main techniques, tools and trends for processing [21][22][23][24]. Firstly, signals are processed directly in the time domain [3][4][5][6][7][8][9][10][11][12][13][14][15][16][17][18] or techniques or transforms are used for their analyses in the frequency [3,5,[18][19][20] or time-frequency [19,20] domains. From here, several techniques are used to obtain features that allow the analysis to be carried out in a better way, such as the use of statistical indicators [3][4][5][6]8,14,[18][19][20], time or time-frequency transforms for direct feature extraction [3][4][5]19,20] and, in some cases, methods for the selection of the most appropriate features or dimensionality reduction such as heuristic techniques [5,6,22], linear discriminant analysis (LDA) [19] or principal component analysis (PCA).…”

“…The use of intelligent techniques is of great value when processing large amounts of information generated by fused sensors, allowing a more adequate manipulation and interpretation of the data. Another point to note is the implementation of the methodologies, since they can be performed online, allowing process monitoring [3][4][5][7][8][9][10][11][12][14][15][16][17][18][19], or offline once the machining of a part has been carried out [6,13,20,26]. This is an important aspect, since in methodologies that use vision systems or images, the machining process must be stopped for the acquisition of images of the tool or they may be taken after the process; however, this depends on the needs and planning of the process.…”

Method of Machining Centre Sliding System Fault Detection using Torque Signals and Autoencoder

“…Very often sliding systems are assessed indirectly by evaluating the quality of machining parts [3]. Direct assessment of the components of the sliding system can be performed using both off-line and on-line techniques, based on methods such as vibration analysis [4][5][6][7], laser interferometry [8,9], noise analysis [10,11], visual inspection [12,13], temperature monitoring [14,15], acoustic emission monitoring [16,17], motor current signature analysis [18,19] and thermal imaging [20][21][22].…”