Tool Condition Monitoring Using Artificial Neural Network Models

Tool wear is a major factor that affects the productivity of any machining operation and needs to be controlled for achieving automation. It affects the surface finish, tolerances, dimensions of the workpiece, increases machine down time, and sometimes performance of machine tool and personnel are affected. This chapter deals with the application of artificial neural network (ANN) models for tool condition monitoring (TCM) in milling operations. The data required for training and testing the models studied and developed are from live experiments conducted in a machine shop on a widely used steel, medium carbon steel (En 8) using uncoated carbide inserts. Acoustic emission data and surface roughness data has been used in model development. The goal is for developing an optimal ANN model, in terms of compact architecture, least training time, and its ability to generalize well on unseen (test) data. Growing cell structures (GCS) network has been found to achieve these requirements.

Development of multi sensor fusion based DAQ for in-process TCMS: Experimental and empirical analysis

In order to manufacture low-cost, high-quality goods, in-process tool condition monitoring is an essential responsibility in the manufacturing industry. In this study, a multisensor fusion technique was used to build and execute an effective and reliable TCM system in turning operations with coated and uncoated tungsten carbide inserts. In dry turning, an attempt has been made to optimize the turning process parameter and monitor the tool's condition. Acoustic optic emission based sensor i.e., Laser Doppler Vibrometer and FLIR E60 infrared thermal camera are strategically placed near the machining zone. Thermal images and vibration signals are recorded using an appropriate charge amplifier. To extract characteristics from numerous sensor data, a National Instruments data acquisition (NI-DAQ) system is constructed utilizing LabVIEW software. Thermal images are used to gather temperatures from tool-work piece locations. Vibration signals are translated into vibration parameters. These characteristics serve as the foundation for establishing in-process TCMS. Tool wear, vibrational displacements (Disp), and cutting temperature are investigated as a result of varied tool insert materials and process conditions (CT). Utilizing ISO 10816-3, ISO 3685, and ISO-18434-2008 standards, the cutting tool condition was assessed using extracted features from multi sensor fusion techniques. For Ti-6Al-4 V, the displacement of uncoated and coated tools increased by 65.28% and 44.71%, respectively. For AISI 316L flank wear, the uncoated insert effected 41% and the coated insert impacted 24.14%, respectively. While machining Al7075, the relationship of depth of cut and feed rate on flank wear maintains an identical trend. It is discovered that both temperature and displacement have a significant role in the evolution of flank wear, which is examined in depth. This paper recognizes the use of multi-sensor data in tool condition monitoring when rotating with various cutting tool inserts.

Transfer of Process References between Machine Tools for Online Tool Condition Monitoring

Process and tool condition monitoring systems are a prerequisite for autonomous production. One approach to monitoring individual parts without complex cutting simulations is the transfer of knowledge among similar monitoring scenarios. This paper introduces a novel monitoring method which transfers monitoring limits for process signals between different machine tools. The method calculates monitoring limits statistically from cutting processes carried out on one or more similar machines. The monitoring algorithm aims to detect general process anomalies online. Experiments comprise face-turning operations at five different lathes, four of which were of the same model. Results include the successful transfer of monitoring limits between machines of the same model for the detection of material anomalies. In comparison to an approach based on dynamic time warping (DTW) and density-based spatial clustering of applications with noise (DBSCAN), the new method showed fewer false alarms and higher detection rates. However, for the transfer between different models of machines, the successful application of the new method is limited. This is predominantly due to limitations of the employed process component isolation and differences between machine models in terms of signal properties as well as execution speed.

State Monitoring Method for Tool Wear in Aerospace Manufacturing Processes Based on a Convolutional Neural Network (CNN)

In the aerospace manufacturing field, tool conditions are essential to ensure the production quality for aerospace parts and reduce processing failures. Therefore, it is extremely necessary to develop a suitable tool condition monitoring method. Thus, we propose a tool wear process state monitoring method for aerospace manufacturing processes based on convolutional neural networks to recognize intermediate abnormal states in multi-stage processes. There are two innovations and advantages of the proposed approach: one is that the criteria for judging abnormal conditions are extended, which is more useful for practical application. The other is that the proposed approach solved the influence of feature-to-recognition stability. Firstly, the tool wear level was divided into different state modes according to the probability density interval based on the kernel density estimation (KDE), and the corresponding state modes were connected to obtain the point-to-point control limit. Then, the state recognition model based on a convolutional neural network (CNN) was developed, and the sensitivity of the monitoring window was considered in the model. Finally, open-source datasets were used to verify the feasibility of the proposed method, and the results demonstrated the applicability of the proposed method in practice for tool condition monitoring.