Machining Process Monitoring Research Articles

Жизненный цикл изделия: мониторинг процессов механической обработки и фильтрация виброакустических сигналов

Introduction. In modern manufacturing, the product life cycle comprises various stages, from conception to disposal. Among these stages, machining plays a significant role, as it directly influences the durability and functionality of the finished product. With increasing competition and the need to reduce production costs, optimizing machining processes has become a crucial task. Traditionally, conservative technological approaches have been used to ensure processing quality. However, this often leads to decreased productivity and higher costs. Modern monitoring and diagnostic techniques can significantly improve process control, particularly through tool condition monitoring. The subject. This paper discusses the stages of the product life cycle and emphasizes the importance of monitoring machining processes. It explores the potential of using vibroacoustic signals to continuously monitor equipment and product conditions. Special attention is paid to the use of vibroacoustic signals for diagnostics and quality control. Modern approaches to filtering these signals, including the use of the fast Fourier transform and various window functions, are analyzed in order to improve the accuracy of the analysis and identify potential defects. The purpose of this work is to develop an algorithm for an online monitoring system that will monitor the condition of cutting tools based on the creation of a digital shadow using a vibroacoustic complex. The main tasks to be solved are to establish the ranges of applicability of frequency response of acoustic signals and optimal window functions, as well as to establish the relationship between the degree of wear on the cutting tool and the results of vibration diagnostics and surface roughness. The methods and technologies for filtering vibroacoustic signals and their application in real–world production settings are discussed. Special attention is given to the role of digital twins in integrating monitoring and filtering data, allowing for the creation of a virtual model of a product to predict its behavior and optimize processes throughout the life cycle. A comparison of various monitoring methods and technologies is conducted, as well as an analysis of practical examples of digital twin implementation in production processes and its impact on improved control. Results and discussion are presented, identifying current research and practical advancements, while also proposing existing challenges and promising areas for future research in the fields of monitoring, signal filtering, and the use of digital twins in mechanical manufacturing.

Cutting model integrated digital twin–based process monitoring in small-batch machining

Cutting model integrated digital twin–based process monitoring in small-batch machining

Development of smart manufacturing framework for micromilling of thin-walled Ti6Al4V

Identification of chatter is mostly carried out offline after capturing the machining signals or by analyzing the machined surface after completing or interrupting the machining process. The offline methods for analyzing the machined signals are tedious and not very useful for microcomponents attributed to permanent deformation of thin-wall if chatter has already been occurred during process. Real-time information is necessary to monitor and control the micromilling process. Consequently, a smart manufacturing framework has been developed in the present work to monitor the chatter onset and the deviation in geometrical shape of the thin-wall during micromilling. An interactive web interface has been developed to project the real-time information necessary to monitor the micromilling process. The excitation of natural modes of the thin-walled structure of Ti6Al4V has been observed in real-time on the web browser for micromilling at 80 µm radial depth of cut, 50,000 rpm and a feed rate of 3 µm/flute. The maximum permanent deformation of 97.43 µm for thin-wall has been obtained at 130 µm radial depth of cut. The developed layout of web browser can be used for real-time monitoring of any other machining process as well.

Health monitoring of CNC machining processes using machine learning and wavelet packet transform

Monitoring the health of a machining process plays a crucial role in advancing fully automated manufacturing. Faulty machining processes (abnormal process) leads to part rejection and can damage the CNC machines. In this context, data-driven machine learning (ML) models with wavelet packet transform (WPT)-based feature extraction method have gained significant importance in health monitoring. The study of different ML models, mother wavelets, and decomposition levels on real production data is important for effective process health monitoring but are not explored well in the literature. This study analyzes the performance of fifty mother wavelets, two decomposition levels (L2 and L3), and five ML models using actual production facility CNC machine dataset to achieve accurate process health monitoring. Various performance indicators obtained from the confusion matrix have been used to evaluate the ML models. The findings indicate that the Random Forest (RF) and Light-Gradient boosting machine (GBM), using the coif8 and db14 mother wavelet, at decomposition level (L3), deliver the best classification performance. Furthermore, the RF model consistently achieves good precision, sensitivity, accuracy, and F1 scores across most classes, making it a reliable and efficient option for the classification of normal and abnormal machining processes. The SVC, MLP, and CNN models, however, also exhibit competitive performance.

A Digital Twin Platform Integrating Process Parameter Simulation Solution for Intelligent Manufacturing

The present work aims to develop a digital twin system typical of intelligent manufacturing applications, which has integrated visualization technologies, as well as the process parameter simulation solution. The application under consideration is a typical machining process, with a gantry machine tool controlled by Siemens Programmable Logic Controller(PLC) S7-1200. With the establishment of dual-directional data communication between the physical machine tool and its virtual counterpart based on TCP/IP protocol, real-time visualization, monitoring, and control of the entire working process can be achieved. Furthermore, we integrated with the digital twin system as a solution for real-time process parameter simulation based on finite element modeling (FEM), which enables the real-time monitoring of necessary process parameters, e.g., surface deformation, during the machining process. A preliminary experiment was conducted to validate our proposed digital twin system, and the results demonstrated that our proposed method has satisfactory performance in terms of both control and monitoring of the traditional machining process, and synchronization between the physical and virtual models is also proven to be positive with minimal latency.

In-process monitoring of the ultraprecision machining process with convolution neural networks

ABSTRACT In-process monitoring and quality control are the most critical aspects of the manufacturing industry, especially in ultra-precision machining (UPM) at an industrial scale. However, in-process ensuring product quality has been difficult, as any subtle change in the process influences the UPM process dynamics and the process outcome. In order to meet the increasingly soaring demand for precision components, intelligent monitoring of the machining process is essentially important and much needed. Capturing complex signal patterns through conventional signal processing for the UPM process is often challenging due to the comparably high noise levels in the industrial environment. Signals obtained during UPM are inherent transients and non-stationary, necessitating extensive and accurate features for classification. Accurate detection of anomalies may allow for quick corrective actions, reducing the degree of damage. Earlier research revealed multi-sensor analysis, which yields richer signal feature information, but the unavoidable sensor failure in conjunction with heterogeneous sensing made it challenging. In order to address the challenges, this paper investigates the feasibility of convolution neural network (CNN) for classifying abnormal and normal machining in the UPM process. The vibrational signals obtained from B&J 4533-B accelerometer during diamond turning are transformed into time-frequency-based log-spectrogram images. These images are classified using CNN, and the results show that a proposed convolutional neural network algorithm has demonstrated an accuracy of 85.92% in classifying images and thus the corresponding in-process machining status.

Effect of applied standard wood machining fluid on colour and chemical composition of the machined wood surface

Appropriate monitoring of wood machining processes is a key issue to ensure the expected quality of the processed wood, expected efficiency and minimize energy consumption of production processes. A new trend is the design of environmentally friendly machining fluids. In this paper, as a preliminary study in this field, the effect of applied standard wood machining fluid on changes in the colour and chemical composition of the machined wood surface is presented. Scots pine wood (Pinus sylvestris L.) was used for this research. Colour measurements were carried out based on the three-axis CIELab system test in time intervals and coefficients such as: colour chroma (Cab*), colour saturation (Sab*), colour hue (h°), and total colour changes (ΔE*). Changes in chemical composition were analysed on the Attenuated Total Reflectance Fourier Transform Infrared Spectroscopy (ATR FT-IR). The results confirmed that standard machining fluids cause a significant change in the colour of the treated pine surface, which decreases over time but is still present even after 24 h. For the spectral analysis, no chemical changes were observed between the machining fluid and the wood. However, the fluid particles remained in the wood after 24 h. In order to reduce the effect of the machining fluid on the colour of the wood, its composition should be changed to allow and/or accelerate the evaporation of their components from the treated wood surface.

Non-invasive milling force monitoring through spindle vibration with LSTM and DNN in CNC machine tools

Milling force reflects the complex mechanical interaction between machine tool and workpiece, it is of great significance to acquire milling force accurately to monitor and optimize machining process. In this paper, a convenient method to monitor milling force non-invasively is proposed, which can predict milling force based on spindle vibration in real-time. Long and short-term memory network (LSTM) and deep neural network (DNN) are used to build the mapping between milling force and spindle vibration in data-driven manner. Temporal features of training data are extracted through LSTM, while DNN is used to reinforced key features. Accuracy of the proposed method is demonstrated by experiments. Comparing with existing data-driven prediction methods, prediction accuracy is improved at least by 16.7% using LSTM-DNN. This paper provides an easy implemented and accurate approach to acquire milling force of machine tool and has great value for machining process monitoring.

Wide-bandwidth cutting force monitoring via motor current and accelerometer signals

Cutting forces are valuable for monitoring machining processes, as they allow the evaluation of tool wear, process stability, and workpiece quality. This paper presents a new non-intrusive method for monitoring cutting forces using a Kalman filter (KF) and Rauch-Tung-Striebel (RTS) smoothing. The method incorporates structural acceleration measurements as well as feed drive motor currents, thereby achieving wide estimation bandwidth and steady-state (DC) gain preservation. The KF + RTS smoothing are parameterized using spindle speed-sweeping cutting tests, which significantly reduces the experimental effort for model identification. The approach was validated in different machining conditions, which demonstrated good agreement with dynamometer measurements.

Generic Cutting Force Modeling with Comprehensively Considering Tool Edge Radius, Tool Flank Wear and Tool Runout in Micro-End Milling.

Accurate cutting force prediction is crucial in improving machining precision and surface quality in the micro-milling process, in which tool wear and runout are essential factors. A generic analytic cutting force model considering the effect of tool edge radius on tool flank wear and tool runout in the micro-end milling process is proposed. Based on the analytic modeling of the cutting part of the cutting edge in the end face of the micro-end mill bottom, the actual radius model of the worn tool is established, considering the tool edge radius and tool flank wear. The tool edge radius, tool wear, tool runout, trochoidal trajectories of the current cutting edge, and all cutting edges in the previous cycle are comprehensively considered in the instantaneous uncut chip thickness calculation and the cutter-workpiece engagement determination. The cutting force coefficient model including tool wear is established. A series of milling experiments are performed to verify the accuracy and effectiveness of the proposed cutting force model. The results show that the predicted cutting forces are in good agreement with the experimental cutting forces, and it is necessary to consider tool wear in the micro-milling force modeling. The results indicate that tool wear has a significant influence on the cutting forces and cutting force coefficients in the three directions, and the influences of tool wear on the axial cutting force and axial force coefficient are the largest, respectively. The proposed cutting force model can contribute to real-time machining process monitoring, cutting parameters optimization and ensuring machining quality.

Assessment of a three-axis on-rotor sensing performance for machining process monitoring: a case study

Online monitoring of cutting conditions is essential in intelligent manufacturing, and vibrations are one of the most effective signals in monitoring machining conditions. Generally, traditional wired accelerometers should be installed on a motionless or stable platform, such as a tool holder or lathe bed, to sense vibrations. Such installation methods would cause the signals to suffer more serious noise interferences and a low signal-to-noise ratio, resulting in less sensitivity to valuable information. Therefore, this study developed a novel three-axis wireless on-rotor sensing (ORS) system for monitoring the turning process. The Micro Electromechanical System (MEMS) accelerometer sensor node can be mounted on a rotating workpiece or spindle rotor and is more sensitive in detecting the vibrations of the entire rotor system without any modification of the lathe system and interference in the cutting procedure. The processor, data acquisition, and Bluetooth Low Energy (BLE) 5.0+ modules were developed and debugged to cooperate with a piezoelectric triaxial accelerometer, with a vibration amplitude not larger than ± 16 g. A series of turning tests were conducted and the results were compared with those from the commercial wired accelerometers, which proved that the ORS system can measure the vibration signal of the rotor system more effectively and sensitively than wired accelerometers, thus demonstrating the accurate monitoring of machining parameters.

Machined Surface Image Analysis for Process Instability Identification in Micromilling of Ti6Al4V

The low flexural rigidity of micro-end mill used in micromilling process undergoes catastrophic failure at low amplitude chatter and hence it is critical to identify the low amplitude chatter to produce the chatter free machined surfaces. An unfixing of the workpiece to identify the chatter by analyzing the power spectral density of machined surface topography alters the process set-up and even a minute change in workpiece surface flatness during re-fixing the workpiece lead to variation in depth of cut, which changes the precision of machined surface in micromilling process. Consequently, in the present work, an image analysis method has been proposed for machined surface analysis in micromilling of Ti6Al4V. Different statistical parameters like arithmetic gray level average value of the distribution of pixel intensities and optical parameter have been analyzed to detect the process instability. The estimated statistical parameters of the machined surface images for different process parameters has been compared with the measured surface roughness of the machined surfaces. The frequency spectrum analysis of the machined surface image analysis has also been carried out, which shows the distribution of pixel intensities at different frequencies. Arithmetic gray level average of pixel intensity of the machined surface image has been found to be decreasing with an increases of the roughness of the machined surface. The optical parameter of the machined surface increases with the increases of the roughness. The critical value of arithmetic gray level average of pixel intensity and optical parameter was found to be 148 and 0.36, respectively for chatter onset. The distribution of pixel intensity at different frequencies shows the diffused pattern of brightness of the pixel away from the center location in chatter dominant machining while the distribution of brightness of pixel was found to be circular in nature around the central spot in stable machining. The chatter and stable machining condition has also been verified with the vibration spectrum of the workpiece, where the vibration of the workpiece has been captured using accelerometer. The proposed image analysis method can further be used for online monitoring of machining process.

Industry-oriented method for dynamic force identification in peripheral milling based on FSC-LSQR using acceleration signals

Online measurement of milling force is very important for machining process monitoring and control. In practice, it is difficult to measure the milling force directly during the milling process. This paper develops a method for milling force identification called least square QR-factorization with the fast stopping criterion (FSC-LSQR) method, and the queue buffer structure (QBS) is employed for the online identification of milling force using acceleration signals. The convolution integral of milling force and acceleration signals is discretized, which turns the problem of milling force identification into a linear discrete ill-posed problem. The FSC-LSQR algorithm is adopted for milling force identification because of its high efficiency and accuracy, which can effectively handle the linear discrete ill-posed problem. The online identification of milling force can be realized using the acceleration signal enqueue and the milling force dequeue operations of the QBS. Finally, the effectiveness of the method is verified by milling tests under different milling parameters.

A new BRTCN model for predicting discharge status of WEDM based on acoustic emission

Machine status detection is very critical to ensure the smoothness of processing workpieces in terms of efficiency and quality. In light of the significance of acoustic emission (AE) signals for the monitoring of wire electrical discharge machining (WEDM) processes and the strong feature extraction ability of deep learning, this paper proposes a physical propagation mechanism of AE and develops an effective deep learning dual-input model called batch relevance temporal convolution neural network (BRTCN) with a new labeling method after analyzing the collected signals. BRTCN, mainly composed of noise extractor, encoder, batch relevance and decoder, is applied to build the AE model that can accurately predict discharge status. Through comparative experiments, the encoder of BRTCN is capable of extracting the AE local sequence features, grabbing long dependencies and reducing computational costs. It is found that the noise extractor in BRTCN is essential for the AE model to converge stably. This paper innovatively detects the discharge status of WEDM based on dual channel AE signals and the proposed BRTCN model examines the relationship between AE and pulse time series with low computation and high accuracy.

Intelligent support system for monitoring machining processes

The intelligent support system for monitoring machining processes was objective of our research. The paper presents process of development of this system which can used in real enterprises. Methodology: to create of the system we use methods of artificial intelligence (neural networks, decision trees and rules, fuzzy logic), using AI methods we created procedures support monitoring machining processes: diagnosis of machining process disturbances using decision rules, control of stability of the machining process based on control card analysis using neural networks, control of surface roughness parameter using neural networks and decision trees and monitoring machine tool through control and compensation of thermal deformations of ball screws of CNC machine tools using neural networks. To optimize of AI models we compared many the models with different structures, chose the best models and used them in intelligent system.

Bio-Inspired Mechano-Sensor Based on the Deformation of Slit Wake

Internal mechano-sensors, as an indispensable part of the proprioceptive system of intelligent equipment, have attracted enormous research interest because of their extremely crucial role in monitoring machining processes, real-time diagnosis of equipment faults, adaptive motor control and so on. The mechano-sensory structure with signal-transduction function is an important factor in determining the sensing performance of a mechano-sensor. However, contrary to the wide application of the cantilever beam as the sensory structure of external mechano-sensors in order to guarantee their exteroceptive ability, there is still a lack of an effective and widely used sensory structure to significantly improve the sensing performance of internal mechano-sensors. Here, inspired by the scorpion using the specialized slit as the sensory structure of internal mechano-sensilla, the slit is ingeniously used in the design of the engineered internal mechano-sensor. In order to improve the deformability of the slit wake, the hollowed-out design around the slit tail of biological mechano-sensilla is researched. Meanwhile, to mimic the easily deformed flexible cuticular membrane covering the slit, the ultrathin, flexible, crack-based strain sensor is used as the sensing element to cover the controllable slit wake. Based on the coupling deformation of the slit wake, as well as the flexible strain sensor, the slit-based mechano-sensor shows excellent sensing performance to various mechanical signals such as displacement and vibration signals.

A Methodology for Tribo-Mechanical Characterization of Metallic Alloys under Extreme Loading and Temperature Conditions Typical of Metal Cutting Processes

The present paper proposes a combined tribo-mechanical methodology for assessing friction under conditions representative of metal cutting, without resorting to machining process monitoring. The purpose is to withdraw the size effect’s contribution due to tool edge radius to the well-known overestimation of the friction coefficient. Comparative numerical analysis of several tribological tests led us to conclude that the ring compression test is one of the most suitable for reproducing the frictional conditions at the chip–tool interface. Two distinct metallic alloys were selected to demonstrate the application of the proposed methodology (UNS L51120 lead alloy and 18Ni300 maraging steel in conventional and additively manufactured conditions). The results help to better explain the influences of process parameters on the friction coefficient value under high temperature and high strain rate conditions. Results showed a typical increase in the coefficient of friction of up to 20% due to both temperature and strain rate parameters for 18Ni300. The results are of interest because they allow considering potential sources of error in the numerical simulation of metal cutting when the same friction coefficient value is considered for a wide range of cutting parameters.

Three-axis displacement sensor-integrated spindle system for carbon fiber-reinforced plastic (CFRP) machining process monitoring

We present a carbon fiber-reinforced plastic (CFRP) milling process monitoring method based on the multi-axis displacement sensor integrated with the precision spindle system. Unlike conventional machining process monitoring methods by using accelerometers or dynamometers, in this study, 2-axis curved-edge sensor (CES) and single-axis capacitive sensor (CS) were integrated with the spindle and measured 3-axis spindle shaft motions along the radial and axial directions, respectively, while machining. Also, the CES showed a status of the spindle balancing while correcting the unbalanced mass on the spindle. The 3 × 3 stiffness matrix of the spindle shaft was multiplied by the 3-axis spindle shaft motions, and simultaneously estimates the cutting forces while cutting the unidirectional CFRP samples with 0o, 45o, and 90o fiber orientation. The conventional 3-axis force sensor installed under the workpiece was used for a baseline comparison with the cutting force results measured by the proposed method. To monitor and characterize the defects of CFRP, the time-frequency domain spectrogram images analyzed from the cutting force data were used. Unique features were extracted according to the fact of existence and nonexistence of the defects.

Special Issue on Self-Optimizing Machining Systems

The concept of Self-Optimizing Machining Systems (SOMS) has been proposed against the background of Industry 4.0 and the Digital Twin concept, based on cyber-physical systems. In order to improve manufacturing productivity, quality, and efficiency, each component technology related to the machining process, such as CAD/CAM, process modeling/simulation, process monitoring/control, and workpiece assessment, as well as the machine tools themselves, has been developed independently to date. However, series of processes, including the interactions among these component technologies, have finally determined the machining performance and the quality of the products. SOMS deals with the information links among these components comprehensively and plays the important role of combining these links and functionalities to optimize the overall machining system. Nevertheless, an intensive implementation and combination of these technologies has yet to become state-of-the-art in industry, while further research and development for SOMS is required for Industry 4.0 and Digital Twin. This special issue focuses on the research trends of SOMS, especially the interaction links among machine tools, process monitoring, and work assessment. From researchers who are active on the front lines of manufacturing engineering, the latest achievements related to the development of SOMS are presented in 6 papers. On one hand, the development of sensor-integrated components is indispensable for SOMS to monitor the status of a process and feed it back to a related component in order to control the machining process and its environment. On the other hand, it can be said that visual simulation, virtual metrology, and other epoch-making, on-machine technologies for evaluating machined surfaces, as well as process optimization based on machined surface information, are strongly required. We hope this special issue will contribute to future research and development for researchers and engineers in the field of manufacturing and machining systems.

Multivariate statistical process monitoring and control of machining process using principal component-based Hotelling T<SUP align="right">2</SUP> charts: a machine vision approach

Multivariate statistical process monitoring and control of machining process using principal component-based Hotelling T<SUP align="right">2</SUP> charts: a machine vision approach