Wear Classification Research Articles

Digital Twin-Driven Tool Wear Monitoring and Predicting Method for the Turning Process

Accurate monitoring and prediction of tool wear conditions have an important influence on the cutting performance, thereby improving the machining precision of the workpiece and reducing the production cost. However, traditional methods cannot easily achieve exact supervision in real time because of the complexity and time-varying nature of the cutting process. A method based on Digital Twin (DT), which establish a symmetrical virtual tool system matching exactly the actual tool system, is presented herein to realize high precision in monitoring and predicting tool wear. Firstly, the framework of the cutting tool system DT is designed, and the components and operations rationale of the framework are detailed. Secondly, the key enabling technologies of the framework are elaborated. In terms of the cutting mechanism, a virtual cutting tool model is built to simulate the cutting process. The modifications and data fusion of the model are carried out to keep the symmetry between physical and virtual systems. Tool wear classification and prediction are presented based on the hybrid-driven method. With the technologies, the physical–virtual symmetry of the DT model is achieved to mapping the real-time status of tool wear accurately. Finally, a case study of the turning process is presented to verify the feasibility of the framework.

Stamping Tool Conditions Diagnosis: A Deep Metric Learning Approach

Stamping processes remain crucial in manufacturing processes; therefore, diagnosing the condition of stamping tools is critical. One of the challenges in diagnosing stamping tool conditions is that traditionally, the tools need to be visually checked, and the production processes thus need to be halted. With the development of Industry 4.0, intelligent monitoring systems have been developed by using accelerometers and algorithms to diagnose the wear classification of stamping tools. Although several deep learning models such as the convolutional neural network (CNN), auto encoder (AE), and recurrent neural network (RNN) models have demonstrated promising results for classifying complex signals including accelerometer signals, the practicality of those methods are restricted due to the flexibility of adding new classes and low accuracy when faced to low numbers of samples per class. In this study, we applied deep metric learning (DML) methods to overcome these problems. DML involves extracting meaningful features using feature extraction modules to map inputs into embedding features. We compared the probability method, the contrastive method, and a triplet network to determine which method was most suitable for our case. The experimental results revealed that, compared with other models, a triplet network can be more effectively trained with limited training data. The triplet network demonstrated the best test results of the compared methods in the noised test data. Finally, when tested using unseen class, the triplet network and the probability method demonstrated similar results.

Tool wear classification based on machined surface images using convolution neural networks

Tool wear classification based on machined surface images using convolution neural networks

Automated and Rapid Seal Wear Classification Based on Acoustic Emission and Support Vector Machine

Seal wear in hydraulic cylinders results in fluid leakage, and instability of the piston rod movement. Therefore, regular inspection of seals is required using automated approaches to improve productivity and to reduce unscheduled maintenance. In literature, successful attempts have been made using acoustic emission-based condition monitoring to classify the seal wear. However, limited attempts have been made to implement automated approaches to classify seal wear using acoustic emission features. Therefore, this article presents an automated approach for rapid and computationally economical diagnosis of seal wear using acoustic emission. The experiments were performed at varying pressure conditions on a hydraulic test rig that can simulate fluid leakage conditions similar to that of hydraulic cylinders. From the acoustic emission spectrum, the frequency bands were identified with peak power and by heterodyning the signal. This technique results in 10X downsampling without losing the information of interest. Further, the signal was divided into smaller “snapshots” to facilitate rapid diagnosis. In these tests, the diagnosis was made on short-time windows, as low as 0.3 seconds in length. A set of time and frequency domain features were designed, and the Support Vector Machine (SVM) was trained on 60 % of the test cases generated with seals of three different conditions. The SVM was able to accurately classify the status irrespective of the pressure conditions, with an accuracy of ~99 %. Therefore, the proposed automated seal wear classification technique based on acoustic emission and SVM can be used for real-time monitoring of seal wear in hydraulic cylinders.

Deep Learning Approach for Vibration Signals Applications.

This study discusses convolutional neural networks (CNNs) for vibration signals analysis, including applications in machining surface roughness estimation, bearing faults diagnosis, and tool wear detection. The one-dimensional CNNs (1DCNN) and two-dimensional CNNs (2DCNN) are applied for regression and classification applications using different types of inputs, e.g., raw signals, and time-frequency spectra images by short time Fourier transform. In the application of regression and the estimation of machining surface roughness, the 1DCNN is utilized and the corresponding CNN structure (hyper parameters) optimization is proposed by using uniform experimental design (UED), neural network, multiple regression, and particle swarm optimization. It demonstrates the effectiveness of the proposed approach to obtain a structure with better performance. In applications of classification, bearing faults and tool wear classification are carried out by vibration signals analysis and CNN. Finally, the experimental results are shown to demonstrate the effectiveness and performance of our approach.

Statistical and Machine Learning Models for Classification of Human Wear and Delivery Days in Accelerometry Data.

Accelerometers are increasingly being used in biomedical research, but the analysis of accelerometry data is often complicated by both the massive size of the datasets and the collection of unwanted data from the process of delivery to study participants. Current methods for removing delivery data involve arduous manual review of dense datasets. We aimed to develop models for the classification of days in accelerometry data as activity from human wear or the delivery process. These models can be used to automate the cleaning of accelerometry datasets that are adulterated with activity from delivery. We developed statistical and machine learning models for the classification of accelerometry data in a supervised learning context using a large human activity and delivery labeled accelerometry dataset. Model performances were assessed and compared using Monte Carlo cross-validation. We found that a hybrid convolutional recurrent neural network performed best in the classification task with an F1 score of 0.960 but simpler models such as logistic regression and random forest also had excellent performance with F1 scores of 0.951 and 0.957, respectively. The best performing models and related data processing techniques are made publicly available in the R package, Physical Activity.

Wear detection of WC-Cu based impregnated diamond bit matrix based on SEM image and deep learning

The working efficiency and lifetime of impregnated diamond tools are closely related to their wear conditions, among which different wear modes of the metal matrix play an essential role. Since traditional qualitative description cannot meet the requirement of mathematical relationship establishment, a deep learning method, Mask R-CNN, was applied for the quantitative determination of the matrix wear based on scanning electron microscope (SEM) images. A series of WC-Cu based metal matrix composite (MMC) samples had been prepared by hot-pressed sintering, followed by a pin-on-disc wear test to obtain the wear surface images, and the datasets were established based on a normal wear classification principle where classification is of four basic types: abrasive wear, adhesive wear, fatigue wear and corrosion wear (corrosion wear is not involved in this study). After training, validation, and test based on the SEM wear image datasets, the wear segmentation results from the trained model indicated that Mask R-CNN could automatically identify the wear of metal matrices efficiently and accurately, which was in good agreement with the results obtained by manual labelling. By modifying the algorithm codes, the masks of abrasive, adhesive, and fatigue wear were extracted and counted for model effectiveness evaluation. Moreover, the wear condition values (i.e., wear region areas) obtained from extracted masks would be easily applied for correlation analysis between cutting tool qualities and drilling efficiencies in future research as well. In comparison with statistic results by artificial cognition, the three types of wear showed an average wear region mask IoU over 70%, and an average wear region area loss of less than 3%. In the process of wear detection on similar wear images in published work, the Mask R-CNN model also presented good performances. All related codes and SEM image datasets are available at https://github.com/sunwucheng/IDB_matrix_wear.

Acoustic emission and machine learning based classification of wear generated using a pin-on-disc tribometer equipped with a digital holographic microscope

The efficiency of processes involving frictional contacts between surfaces is often characterized by wear rates or friction coefficients. However, the classification and forecasting of wear rates in friction related processes is a real industrial challenge that is unsolved today. Hence, an on-line monitoring system able to classify wear rate can be crucial for many industries as it could help in preventing catastrophic failures. Applications include lifetime assessment of various industrial components where a range of wear failures occur such as scuffing (a typical sudden failure mechanism). These tribological processes can now be sensorized, and the corresponding sensor signatures can be modelled and monitored using state-of-the-art Machine learning (ML) algorithms. In this study, we use an Acoustic Emission (AE) sensor and ML frameworks to classify different wear categories simulated with a customized pin-on-disc tribometer. A real-time investigation of the wear track is necessary to find out the origins of the wear scar visible at the surface. To achieve this objective, the experiments were conducted on a pin-on-disc tribometer equipped with a Digital Holographic Microscope (DHM). Experiments were carried out using alumina and steel balls against steel discs at room temperature. Real-time DHM images of the wear track surface were recorded for each lap at the same position. An acoustic emission sensor recorded the AE signals during the complete duration of experiments. The AE signatures, in combination with the real-time DHM images, were correlated as input and ground truth labels for the ML algorithm. Several ML frameworks were compared; they are support vector machine, logistic regression, XGBoost, random forest, neural networks, k-Nearest Neighbor, quadratic discriminant analysis and Naive Bayes. The classifier was trained to differentiate the acoustic emission features of the different wear rates. Most ML algorithms had an average classification accuracy above 80%, and the highest was obtained with support vector machine (84.7%). The classification accuracy can be improved by combining two neighboring categories with limited differences in terms of wear rate. Hence, the proposed method has a significant industrial potential for in-situ and real-time quality monitoring of wear processes since it requires minimum modifications of commercially available industrial machines.

Classification of tooth wear and indexes used in diagnosis

Classification of tooth wear and indexes used in diagnosis

Deep learning methods for drill wear classification based on images of holes drilled in melamine faced chipboard

In this paper, a set of improvements made in drill wear recognition algorithm obtained during previous work is presented. Images of the drilled holes made on melamine faced particleboard were used as its input values. During the presented experiments, three classes were recognized: green, yellow and red, which directly correspond to a tool that is in good shape, shape that needs to be confirmed by an operator, and which should be immediately replaced, since its further use in production process can result in losses due to low product quality. During the experiments, and as a direct result of a dialog with a manufacturer it was noted that while overall accuracy is important, it is far more crucial that the used algorithm can properly distinguish red and green classes and make no (or as little as possible) misclassifications between them. The proposed algorithm is based on an ensemble of possibly diverse models, which performed best under the above conditions. The model has relatively high overall accuracy, with close to none misclassifications between indicated classes. Final classification accuracy reached 80.49% for biggest used window, while making only 7 critical errors (misclassifications between red and green classes).

Time-efficient approach to drill condition monitoring based on images of holes drilled in melamine faced chipboard

This paper presents a time-efficient approach to the drill wear classification problem that achieves a similar accuracy rate compared to more complex and time-consuming solutions. A total of three classes representing drill state are recognized: red for poor state, yellow for elements requiring additional evaluation, and green for good state. Images of holes drilled in melamine faced chipboard were used as input data, focusing on evaluating differences in image color values to determine the overall drill state. It is especially important that there are as few mistakes as possible between the red and green class, as these generate the highest loss for the manufacturer. In green samples presented in gray-scale, most pixels were either black (representing the hole) or white (representing the chipboard), with very few values in between. The current method was based on the assumption that the number of pixels with intermediate values, instead of extreme ones, would be significantly higher for the red class. The presented initial approach was easy to implement, generated results quickly, and achieved a similar accuracy compared to more complex solutions based on convolutional neural networks.

INFORMATION TECHNOLOGY FOR AUTOMATED ASSESSMENT OF THE ARTILLERY BARRELS WEAR BASED ON SVM CLASSIFIER

An information technology for the automated assessment of the has been developed wear level has been developed. Information technology is based on the analysis of acoustic fields accompanying a shot. The acoustic field of the shot consists of a ballistic wave accompanying a projectile flying out at a supersonic speed, and a muzzle wave generated when propellant gases are ejected from the barrel. The parameters of the ballistic and muzzle waves depend significantly on the level of barrel wear. This makes it possible to construct an automatic classifier of the barrel wear level based on the analysis of informative features of acoustic signals recorded by microphones near the weapon's firing position. The information technology is based on a binary SVM classifier. A set of records of acoustic fields of shots was synthesized on the basis of real signals recorded when firing a 155 mm howitzer. From the set of records, a training and test set of information features were formed for training the classifier and assessing its quality. Methods of preliminary data normalization of training and test samples are investigated. A technique for optimizing the classifier hyperparameters with instance cross-validation has been developed. The technique is a two-stage method for finding the optimal values of hyperparameters. In the first stage, the search is performed on an exponential decimal grid. At the second stage, the optimal values of hyperparameters are refined on a linear grid. A method for the binary classification of artillery barrels according to the wear level has been formulated. Checking the classifier on a consistent test sample showed that it provides the correct classification of barrel wear with a probability of 0.94. An information technology has been developed for classifying artillery barrels by wear level based on the analysis of acoustic fields of shots. Information technology consists of three stages: data preparation, construction, training an optimization of the binary SVM classifier and the operation of the binary SVM classifier. A field experiment was carried out, which confirmed the correctness of the basic scientific and technical solutions. An automated system has been developed for classifying wellbores by wear level.

Indirect cutting tool wear classification using deep learning and chip colour analysis

In the growing Industry 4.0 market, there is strong need to implement automatic inspection methods to support manufacturing processes. Tool wear in turning is one of the biggest concerns that most expert operators are able to indirectly infer through the analysis of the removed chips. Automatising this operation would enable developing more efficient cutting processes that turns in easier process planning management toward the Zero Defect Manufacturing paradigm. This paper presents a deep learning approach, based on image processing applied to turning chips for indirectly identifying tool wear levels. The procedure extracts different indicators from the RGB and HSV image channels and instructs a neural network for classifying the chips, based on tool state conditions. Images were collected with a high-resolution digital camera during an experimental cutting campaign involving tool wear analysis with direct microscope imaging. The sensitivity analysis confirmed that the most sensible image channels are the hue value H that were used to teach the network, leading to performances in the range of 95 of proper classification. The feasibility of the deep learning approach for indirectly understanding the tool wear from the chip colour characterisation is confirmed. However, due to the big effects on chip colours of variables as the workpiece material and cutting process parameters, the applicability is limited to stable production flows. An industrial implementation can be foreseen by populating proper large databases and by implementing real-time chip segmentation analysis.

Bread and porridge at Early Neolithic Göbekli Tepe: A new method to recognize products of cereal processing using quantitative functional analyses on grinding stones

The Early Neolithic site Göbekli Tepe which has delivered the largest corpus of grinding stones known by now in Northern Mesopotamia is used as a case study to test the possibility of wear classification and the application of quantification methods in use-wear caused by the processing of cereals. The research design is based on combined optical and tactile analyses as well as geometric analyses of shape deformation and 3D-modelling of surface roughness on original grinding stones and experimentally used replicas. This approach can be used to differentiate between fine and coarse flour as products of cereal processing as they produce different wear markers. Also, it can be applied to quantify use-lives of handstones. At Göbekli Tepe, handstones with markers for fine flour production, which can be possibly linked to bread-like products, are in the minority while handstones with markers for coarse flour clearly predominate. The social implications of these results have to be discussed with a view towards Göbekli Tepe’s interpretation as a ritual site. We propose fine flour as a special product within feasting, in accordance with previous interpretations of the site.

CNN based tool monitoring system to predict life of cutting tool

In this study, we present tool wear prediction system to monitor the flank wear of a cutting tool by Machine Learning technique namely, Convolutional Neural Network (CNN). Experimentations were performed on mild steel components under dry cutting condition by carbide inserts as cutting tool. Images of cutting tool and turned component were taken at regular interval using an inverted microscope to measure the progression of flank wear and the corresponding image of component was noted. These images were used as an input to the CNN model that extract the features and classify cutting tool in one of the three wear class namely, low, medium and high. The result of the CNN training set was used to monitor the life of cutting tool and predict its remaining useful life. In this work which is first of its kind, the CNN model gives an accuracy of 87.26% to predict the remaining useful life of a cutting tool. In particular, the study exhibits that CNN method gives good response to the data in the form of images, when used as an indicator of tool wear classification in different classes.

The wear classification of MoDTC-derived particles on silicon and hydrogenated diamond-like carbon at room temperature

Molybdenum dithiocarbamate (MoDTC) is well-known as a friction-modifier in engine oil but has wear accelerating effects on DLC/steel sliding contact. Since MoDTC was reported degradable to other materials, the effect of each material on wear acceleration has not been clarified yet. Therefore, it is necessary to divide which Mo-derived compound has a major role to enhance wear. Thus, powder-type MoDTC, MoO3, Mo2C, MoS2, and Mo were dispersed into Poly-alpha-olefin (PAO), and tested on a-C:H and Si-DLC against steel ball at room temperature. For Mo2C, MoDTC, and MoO3, the specific wear rate of a-C:H was approximately 3.4 × 10−5, 3.2 × 10−6 and 1.5 × 10−6 mm3/Nm, respectively. MoDTC indicated higher affection than MoO3 regardless of the particles’ hardness. Si-DLC showed better wear prevention compared to a-C:H.

Convolutional neural network-based tool condition monitoring in vertical milling operations using acoustic signals

Sonic monitoring presents itself as one of the least invasive but easiest to implement methods of machine condition characterization. This work investigates the viability of categorically classifying cutting tool wear using only sonic output from a vertical milling center and proposes a statistical model of milling acoustic signals as well as a novel machine learning-integrated method of acoustic signal differentiation. To this end, a deep convolutional neural network is used for data classification. Experimental results support the proposed sonic model and demonstrate that tool wear classification accuracy as high as 99.5% is possible using a two-dimensional deep convolutional neural network.

Classifiers ensemble of transfer learning for improved drill wear classification using convolutional neural network

In this paper we introduce the enhanced drill wear recognition method, based on classifiers ensemble, obtained using transfer learning and data augmentation methods. Red, green and yellow classes are used to describe the current drill state. The first one corresponds to the case when drill should be immediately replaced. The second one denotes a tool that is still in a good condition. The final class refers to the case when a drill is suspected of being worn out, and a human expert evaluation would be required. The proposed algorithm uses three different, pretrained network models and adjusts them to the drill wear classification problem. To ensure satisfactory results, each of the methods used was required to achieve accuracy above 90% for the given classification task. Final evaluation is achieved by voting of all three classifiers. Since the initial data set was small (242 instances), the data augmentation method was used to artificially increase the total number of drill hole images. The experiments performed confirmed that the presented approach can achieve high accuracy, even with such a limited set of training data.

Data augmentation techniques for transfer learning improvement in drill wear classification using convolutional neural network

This paper presents an improved method for recognizing the drill state on the basis of hole images drilled in a laminated chipboard, using convolutional neural network (CNN) and data augmentation techniques. Three classes were used to describe the drill state: red - for drill that is worn out and should be replaced, yellow - for state in which the system should send a warning to the operator, indicating that this element should be checked manually, and green - denoting the drill that is still in good condition, which allows for further use in the production process. The presented method combines the advantages of transfer learning and data augmentation methods to improve the accuracy of the received evaluations. In contrast to the classical deep learning methods, transfer learning requires much smaller training data sets to achieve acceptable results. At the same time, data augmentation customized for drill wear recognition makes it possible to expand the original dataset and to improve the overall accuracy. The experiments performed have confirmed the suitability of the presented approach to accurate class recognition in the given problem, even while using a small original dataset.

Tool wear classification using time series imaging and deep learning

Tool condition monitoring (TCM) has become essential to achieve high-quality machining as well as cost-effective production. Identification of the cutting tool state during machining before it reaches its failure stage is critical. This paper presents a novel big data approach for tool wear classification based on signal imaging and deep learning. By combining these two techniques, the approach is able to work with the raw data directly, avoiding the use of statistical pre-processing or filter methods. This aspect is fundamental when dealing with large amounts of data that hold complex evolving features. The imaging process serves as an encoding procedure of the sensor data, meaning that the original time series can be re-created from the image without loss of information. By using an off-the-shelf deep learning implementation, the manual selection of features is avoided, thus making this novel approach more general and suitable when dealing with large datasets. The experimental results have revealed that deep learning is able to identify intrinsic features of sensory raw data, achieving in some cases a classification accuracy above 90%.