Tool Wear Condition Research Articles

Prediction of tool wear in CFRP drilling based on neural network with multicharacteristics and multisignal sources

Carbon fiber-reinforced polymer (CFRP) drilling is a typical process in the aircraft industry. Because the components of CFRP are different and uneven, it is difficult to extract tool wear characteristics from the machining signals, which are composed of the processing characteristics of various materials and the tool state characteristics. The aim of this work is to present a new comprehensive approach based on multicharacteristics and multisignal sources to predict the tool wear state during CFRP drilling through a combination of a backpropagation (BP) artificial neural network (ANN) model and an efficient automatic system depending on the sliding window algorithm. It was verified that the peak factor and Kurtosis coefficient of different signals and the energy value of the d5 layer of the thrust force signal and the d3 layer of the vibration signal after wavelet decomposition were related to tool wear. Among them, the energy value of the d3 layer of the vibration signal was selected as the wear indicator and was able to describe the state of the tool during the CFRP drilling process regardless of the drilling conditions and individual tool differences. A confirmatory drilling experiment using 6-mm-diameter polycrystalline diamond twist drilling under different processing parameters was conducted to verify the ANN model based on multicharacteristics and multisignal sources. A lower feed speed and a higher cutting speed were both highly correlated with the VB value of flank wear. Drill wear accelerated because of the occurrence of adhesive wear when the number of drilled holes reached around 90. The accuracy of the neural network model is 80–87% when using the value of only one characteristic but clearly increases based on multicharacteristics and multisignal sources in real time, indicating that the BP ANN model has higher accuracy in predicting the tool state in CFRP drilling through the sensor signal fusion method.

Experimental study on cutting flexible textile materials and the sharpening process of a vibrating blade

Blade wear is an essential factor that influence textile cutting quality and precision and it is important to seek an appropriate method to predict the cutting tool conditions. In this study, the vertical force range of the blade are measured to determine the vibrating blade’s abrasive wear during different cutting conditions of superimposed plies of a flexible material “Denim”. The relationship between cutting force against the abrasive wear is studied and identified through the evolution of the roughness and the cutting edge radius, which provides a technical foundations for online monitoring of the blade’s wear conditions and cutting quality. Moreover, this study analyzes the sharpening conditions of three different HSS blades in order to regenerate the suitable tool’s cutting edge. It is shown that with incorporation of the vertical cutting force factor we can predict conveniently the tool wear condition and start the sharpening procedure at the accurate time.

Tool Wear Status Recognition and Prediction Model of Milling Cutter Based on Deep Learning

In order to ensure the reliability and stability of the manufacturing process, tool wear state should be realized real-time and accurate monitoring. This paper proposes a tool wear state recognize and predictive framework model based on Stacking Sparse De-noising Auto-Encoder (SSDAE), the Particle Swarm Optimization (PSO) and the Least Squares Support Vector Machine (LSSVM). The Stacking Sparse De-noising Auto-Encoder (SSDAE) technique is utilized to realize multi-feature signal dimension reduction with the aim of improving the prediction accuracy, which reduces the dependence on the prior knowledge of feature selection and greatly improves the modeling efficiency. PSO technique is helpful for adaptive optimization of kernel parameters, which greatly improved computing power and LSSVM model prediction accuracy. A dataset from a real machining process is utilized to verify the effectiveness of proposed model in improving the prediction accuracy. The experimental results show that a high correlation coefficient greater than 0.95 is used to extract feature vector from time domain, frequency domain and time-frequency domain three directions, and the proposed SSDAE-PSO-LSSVM model performs better than Partial Least Squares Regression (PLSR), Back Propagation Neural Network (BPNN) and Extreme Learning Machine (ELM) in terms of prediction accuracy.

Research on Tool Wear Based on 3D FEM Simulation for Milling Process

In the process of metal cutting, the anti-wear performance of the tool determines the life of the tool and affects the surface quality of the workpiece. The finite element simulation method can directly show the tool wear state and morphology, but due to the limitations of the simulation time and complex boundary conditions, it has not been commonly used in tool life prediction. Based on this, a tool wear model was established on the platform of a finite element simulation software for the cutting process of titanium alloy TC4 by end milling. The key technique is to embed different types of tool wear models into the finite element model in combination with the consequent development technology. The effectiveness of the tool wear model was validated by comparing the experimental results with the simulation results. At the same time, in order to quickly predict the tool life, an empirical prediction formula of tool wear was established, and the change course of tool wear under time change was obtained.

A Multi-Index Generative Adversarial Network for Tool Wear Detection with Imbalanced Data

The scarcity of abnormal data leads to imbalanced data in the field of monitoring tool wear conditions. In this paper, a novel multi-index generative adversarial network (MI-GAN) is proposed to detect the tool wear conditions subject to imbalanced signal data. First, the generator in the MI-GAN is trained to produce fake normal signals, and the discriminator computes scores of testing signals and generated signals. Next, the generator detects abnormal signals based on the performance of imitating testing signals, and the discriminator will compute the scores of testing signals and generated signals. Subsequently, two indexes, i.e., L 2 -norm and temporal correlation coefficient (CORT), are put forward to measure the similarity between generated signals and testing signals. Finally, our decision-making function further combines L 2 -norm and CORT with two discriminator scores to determine the tool conditions. Experimental results show that our method obtains 97% accuracy in tool wear detection based on imbalanced data without manual feature extraction, which outperforms traditional machine learning methods.

A New Multitask Learning Method for Tool Wear Condition and Part Surface Quality Prediction

Deep learning has been gradually used in the field of machining condition monitoring. However, at present only single-task prediction can be performed, which results in increased experimental costs, wasted datasets, and repetitive work. In this article, a new multitask learning method based on a deep belief network (DBN) is proposed, which can be used to predict the tool wear condition and part surface quality. The single-task data transmission of the last few hidden layers of the DBN network is improved to multitask parallel data transmission so that the improved DBN can realize multitask learning. The loss function of the multitask learning model is defined as the weighted sum of all single-task loss functions. According to the loss of different tasks in the iteration process, the weight of corresponding tasks can be adjusted automatically. Furthermore, the multitask deep learning method can realize information sharing, suppress overfitting, improve prediction accuracy, and require less computing time. Combined with the abovementioned improvements, a multitask model for tool wear and part surface quality was developed. Experimental verification was performed on a KVC850M three-axis vertical machining center. The results show that the accuracy of the proposed multitask prediction model is 99% for the tool wear prediction and 92.86% for part surface quality prediction.

An Improved Deep Learning Model for Online Tool Condition Monitoring Using Output Power Signals

Something like normal functionality of tools in a manufacturing process is typically designed to ensure reliability, where fast and accurate identification of tool abnormal operation plays a vital role in intelligent manufacturing. In this study, a novel method is proposed to assess the cutting tool condition, which consists of a convolutional neural network with wider first-layer kernels (W-CONV), and long short-term memory (LSTM). The analysis benefits from the use of output power signals from the cutting tool, since they can be obtained easily and efficiently, enabling the proposed method to be applicable in practical operation for online condition monitoring. Moreover, effectiveness of the proposed method is investigated, using test data from cutting tools at various tool wear conditions. Results demonstrate that with the proposed method, tool wear condition can be identified accurately and efficiently. Furthermore, with test data collected at cutting tools with different sizes, the robustness of the proposed method can be further clarified.

Use of image processing to monitor tool wear in micro milling

Micro milling is a versatile machining process owing to its capability to machine, by material removal, micro-sized components with complex geometrical features. However, micro milling tools wear quickly, being a key issue to determine tool wear condition in order to prevent excessive tool wear or a sudden tool breakage while machining, which would waste the workpiece. Due to the small size of micro milling tools, direct measurement of the worn tool is not possible. In order to overcome this drawback, this paper presents a new method based on digital image processing where image captures of the micro tool and subsequent analysis provides a valuable information to determine the progression of tool wear. Tool wear is measured in flank wear, crater wear and tool radius reduction. Different approaches are compared so as to determine the best option for every set of images. These methods are based on the use of morphological operations, k-means clustering and Otsu Multilevel algorithm. Results show a good performance with differences of 5% between predicted and actual worn area, which satisfies the industrial requirements. This procedure can be transferred to industrial environments and implemented in collaborative robots, increasing the level of automation and facilitating the decision-making process.

Micro-milling tool wear monitoring under variable cutting parameters and runout using fast cutting force coefficient identification method

Extracting discriminative tool wear features is of great importance for tool wear monitoring in micro-milling. However, due to the dependency on tool runout and cutting parameters, the traditional tool wear features are incompetent to monitor the tool wear condition in micro-milling with significant tool runout and varied cutting parameter interactions. In this study, micro-milling cutting force is represented by a parametric model including variable cutting parameters, tool runout, and tool wear. The cutting force coefficient in the model, which is not only discriminative to the tool wear condition but also independent to the tool runout and cutting parameters, is extracted as the micro-milling tool wear feature. To reduce the computation cost, a fast neural network–based method is proposed to identify the tool runout and the cutting force coefficient from the cutting force signal. Experimental results show that the proposed cutting force coefficient–based approach is efficient to monitor the micro-milling tool wear under varied cutting parameters and tool runout.

Tool condition monitoring framework for predictive maintenance: a case study on milling process

ABSTRACT In metal cutting processes, tool condition monitoring has a great importance to prevent surface damage and maintaining the quality of surface finishing. With the development of digitalisation and connection of industrial machines, it has become possible to collect real-time data from various types of sensors (e.g. vibration, acoustic or emission) during the process execution. However, information fusion from multiple sensor signals and tool health prediction still present a big challenge. The aim of this paper is to present a data-driven framework to estimate the tool wear status and predict its remaining useful life by using machine learning techniques. The first part of the framework is dedicated to sensor data preprocessing and feature engineering, while the second part deals with the development of prediction models. Different types of machine learning algorithms are used and compared to find the best result. A case study in a milling process is presented to illustrate the potentialities of the proposed framework for tool condition monitoring.

Discrimination and Prediction of Tool Wear State Based on Gray Theory

Abstract In allusion to the problems that random noise causes that interfere with the extraction of wear characteristics from cutting vibration signals and cause low accuracy of lathe tool wear state discrimination and prediction, a new method was proposed in this study. The proposed method extracted the wear time-domain characteristics of tool vibration signals through wavelet packet transform and the correlation coefficient method. Next, noises in wear time-domain characteristics were reduced by singular-value decomposition. The gray proximity correlation between the characteristic data series corresponding to the initial cutting and current cutting was calculated and used to represent the characteristic of tool wearing and discriminate the wear state. The gray decision-making model of metabolism GM(1,1) was established based on the wear characteristic series and was used to predict the variation trend of tool wear state, thus deciding the necessity of cutting tool changing. Cutting wear experiments and wear state predictions were carried out on the ZCK20 numerical control lathe (Tuoman, Zhejiang, China) by using three pieces of WNMG080408-TM T9125 lathe tools (Tungaloy Corporation, Iwaki, Japan). Experimental results demonstrated that the proposed method could eliminate noise effectively, acquire the optimal wear characteristics of tools, and discriminate and predict the wear state of tools accurately.

Study on tool wear mechanism and cutting performance in helical milling of CFRP with stepped bi-directional milling cutters

Helical milling, a new machining technology, has obvious advantages in hole making of carbon fiber reinforced plastic (CFRP), and the technology, nonetheless, shows some defects such as hole diameter tolerance and surface damage aggravation caused by fast tool wear. Aiming to improve the hole making quality of CFRP, this paper proposes the bi-directional helical milling strategy, carries out the experimental research on milling CFRP with stepped bi-directional milling cutters, analyzes the wear mechanism of forward and backward cutting edges, and thus obtains the change laws of cutting forces and hole-making quality in different tool wear conditions. The experimental results show that in the initial milling stage, strip-shaped and spoon-shaped wear bands are formed by abrasive wear mechanism on both forward and backward cutting edges; with aggravation of the tool wear, the wear mechanism of forward cutting edge is gradually dominated by the fatigue wear, and subsequently, a large-area coating peels off the tool surface, causing the rapid wear of substrate. Compared with the forward cutting edge, the backward cutting edge produces less and more stable axial cutting force, and causes all the cutting forces to rise more slowly. In comparison with traditional drilling and milling of CFRP, bi-directional hole-making can effectively suppress some defects such as hole diameter shrinking, burrs and tearing, and thus it shows better machining quality.

Tool Wear Condition Monitoring by Combining Variational Mode Decomposition and Ensemble Learning.

Most online tool condition monitoring (TCM) methods easily cause machining interference. To solve this problem, we propose a method based on the analysis of the spindle motor current signal of a machine tool. Firstly, cutting experiments under multi-conditions were carried out at a Fanuc vertical machining center, using the Fanuc Servo Guide software to obtain the spindle motor current data of the built-in current sensor of the machine tool, which can not only apply to the actual processing conditions but, also, save costs. Secondly, we propose the variational mode decomposition (VMD) algorithm for feature extraction, which can describe the tool conditions under different cutting conditions due to its excellent performance in processing the nonstationary current signal. In contrast with the popular wavelet packet decomposition (WPD) method, the VMD method was verified as a more effective signal-processing technique according to the experimental results. Thirdly, the most indicative features that relate to the tool condition were fed into the ensemble learning (EL) classifier to establish a nonlinear mapping relationship between the features and the tool wear level. Compared with existing TCM methods based on current sensor signals, the operation process and experimental results show that using the proposed method for the monitoring signal acquisition is suitable for the actual processing conditions, and the established tool wear prediction model has better performance in both accuracy and robustness due to its good generalization capability.

A U-Net-Based Approach for Tool Wear Area Detection and Identification

The tool wear condition monitoring is key to ensuring product quality. This article develops a direct technique dealing with cutting tool images to automate the tool wear detection and identification. The constructed U-Net-based network can realize an effective and reliable extraction of the tool wear area. The introduction of deep supervision with a Matthews correlation coefficient (MCC)-based surrogate loss function helps to address the few-shot and data imbalance issues. Experiments on the images with wear on the flank face of cutting tools from a computer numerical control (CNC) turning machine show the effectiveness, competitiveness, and reliability of the proposed method under different types of loss functions.

Tool wear condition monitoring based on a two-layer angle kernel extreme learning machine using sound sensor for milling process

Tool condition monitoring (TCM) in numerical control machines plays an essential role in ensuring high manufacturing quality. The TCM process is conducted according to the data obtained from one or more of a variety of sensors, among which acoustic sensors offer numerous practical advantages. However, acoustic sensor data suffer from strong noise, which can severely limit the accuracy of predictions regarding tool condition. The present work addresses this issue by proposing a novel TCM method that employs only a few appropriate feature parameters of acoustic sensor signals in conjunction with a two-layer angle kernel extreme learning machine. The two-layer network structure is applied to enhance the learning of features associated with complex nonlinear data, and two angle kernel functions without hyperparameters are employed to avoid the complications associated with the use of preset hyperparameters in conventional kernel functions. The proposed TCM method is experimentally demonstrated to achieve superior TCM performance relative to other state-of-the-art methods based on sound sensor data.

Prediction of residual stress with multi-physics model for orthogonal cutting Ti-6Al-4V under various tool wear morphologies

• The multi-physics modeling was proposed to analyze the residual stress distribution. • The additional stresses due to tool wear morphologies were integrated. • The relationship between residual stress index and tool wear was developed. The machining-induced residual stress has a pivotal impact on the corrosion resistance, fatigue, and functionality of manufactured components. Tool wear process induces the tool geometrical alterations, thereby affects the thermo-mechanical loads and residual stress distribution in the manufactured components. This work proposed a multi-physics model with the considerations of wear induced tool geometrical alterations to predict the residual stresses distribution for orthogonal turning Ti-6Al-4 V. These tool geometrical alterations included the flank wear, cutting edge radius, and rake angle. The proposed model framework involved five steps: (i) cutting force, (ii) temperature distribution, (iii) mechanical stress, (iv) thermal stress, and (v) calculation residual stresses by relaxation procedure. The experimental results verified that this model could effectively evaluate the residual stress distribution characteristics under tool wear conditions. Based on the prediction results, the response relationship analysis between the residual stresses and tool wear conditions was developed. This work could provide a novel method to control the residual stress distribution through monitoring of the tool wear morphology.

Tool wear state recognition based on gradient boosting decision tree and hybrid classification RBM

Machined surface quality and dimensional accuracy are significantly affected by tool wear in machining process. Tool wear state (TWS) recognition is highly desirable to realize automated machining process. In order to improve the accuracy of TWS recognition, this research develops a TWS recognition scheme using an indirect measurement method which selects signal features that are strongly correlated with tool wear to recognize TWS. Firstly, three time domain features are proposed, including dynamic time warping feature and two entropy features. The time, frequency, and time-frequency domain features of the vibration and force signals are extracted to form a feature set. Secondly, gradient boosting decision tree (GBDT) is adopted to select the optimal feature subset. Lastly, contrastive divergence (CD) and RMSspectral are used to train hybrid classification RBM (H-ClassRBM). The trained H-ClassRBM is used for TWS recognition. The PHM challenge 2010 data set is used to validate the proposed scheme. Experimental results show that the proposed features have better monotonicity and correlation than the classical features. Compared with CD and Adadelta, CD and Adagrad, and CD and stochastic gradient descent with momentum, the H-ClassRBM trained by CD and RMSspectral improves recognition accuracy by 1%, 2%, and 2%, respectively. Compared with feedforward neural network, probabilistic neural network, Gaussian kernel support vector machine, and H-ClassRBM, the proposed TWS recognition scheme improves recognition accuracy by 37%, 51%, 9%, and 8%, respectively. Therefore, the proposed TWS recognition scheme is beneficial in improving the recognition accuracy of TWS, and provides an effective guide for decision-making in the machining process.

Identification of cutting tool wear condition in turning using self-organizing map trained with imbalanced data

One of the most important parameters in machining process is tool wear. Thus, monitoring the wear of cutting tools is essential to ensure product quality, increase productivity, reduce environmental impact and avoid catastrophic damages. As wear is related to the vibrations of the process, the vibration signal is commonly used to monitor the process non-intrusively. Traditional wear monitoring techniques present a number of problems such as: the difficulty of identifying vibration features sensitive to wear evolution, the specialist requirement for supervising the model training and an endless series of tests to work with balanced data. To overcome these difficulties, this paper aims to propose a new approach in the application of unsupervised artificial intelligence technique with imbalanced data to identify the cutting tool wear condition during the turning process. The methodology will allow industrial applications since no supervision is required in the model training when machining condition is changed. From vibration signals collected during each tool pass, a self-organizing map model was used to identify the ideal moment of tool change. The classifier used was compared to benchmark supervised methods (weighted k-nearest neighbor and support vector machine). Imbalanced data sets were used to simulate the industrial reality. Tool tests were performed under different wear conditions and changing the cutting parameters. The results showed that it is possible to predict the cutting tool wear condition with a self-organizing map neural for imbalanced data, using only the vibration signal with up to 92% accuracy.

Tool wear induced modifications of plastic flow and deformed material depth in new generated surfaces during turning Ti-6Al-4V

The thermo-mechanical loads induced by tribological interfaces complicate the descriptions of the local plastic deformation behaviours in high-speed machining. This study utilized finite element simulation technique and microstructural observation to investigate the local plastic deformation behaviours of new generated chips and machined surface in turning Ti-6Al-4 V. The plastic deformation behaviours modifications of plastic flow and deformed material depth under different tool wear states were discussed. On one hand, the Arbitrary-Lagrangian-Eulerian (ALE) modelling based on hybrid sticking-sliding friction model was proposed with dependence of tool wear geometrical changes. The multi-physics variables distributions such as cutting temperature, strain, and strain rate were acquired to provide the basic analysis of thermo-mechanical loads induced by tribological effects. On other hand, the experimental evidences of local plastic deformation behaviours including intensive microstructural changes in primary/secondary/tertiary deformation zones were analysed in details. Both simulated and experimental results indicated that thermo-mechanical loads due to tool wear effects were the critical driving factor for local plastic behaviours evolution. The deformation strength of primary/secondary/tertiary zones increased because of the gradual increasing contact areas and heat generation. This study can help to control the microstructural evolutions by limiting tool wear and enhance the surface quality.

Simple machine learning allied with data-driven methods for monitoring tool wear in machining processes

The aim of this work was to identify the occurrence of machine tool wear in carbide inserts applied in a machine turning center with two steel materials. Through the data collected with an open-source communication protocol during machining, eighty trials of twenty runs each were performed using central composite design experiments, resulting in a data set of eighty lines for each tested material. The data set consisted of forty lines with the tool wear condition and forty lines without. Machining parameters were set to be in the range of the usual industrial values. The cutting parameters in the machining process were cutting speed, feed rate, cutting depth, and cutting fluid applied in the abundance condition and without cutting fluid (dry machining). The collected data were the spindle motor load, X-axis motor load, and Z-axis motor load in terms of the percentage used. AISI P20 and AISI 1045 steels workpieces were tested with both new and worn inserts, and a flank tool wear of 0.3 mm was artificially induced by machining with the same material before the data collecting experiment. Two approaches were used in order to analyze the data and create the machine learning process (MLP), in a prior analysis. The collected data set was tested without any previous treatment, with an optimal linear associative memory (OLAM) neural network, and the results showed 65% correct answers in predicting tool wear, considering 3/4 of the data set for training and 1/4 for validating. For the second approach, statistical data mining methods (DMM) and data-driven methods (DDM), known as a self-organizing deep learning method, were employed in order to increase the success ratio of the model. Both DMM and DDM applied along with the MLP OLAM neural network showed an increase in hitting the right answers to 93.8%. This model can be useful in machine monitoring using Industry 4.0 concepts, where one of the key challenges in machining components is finding the appropriate moment for a tool change.