Search Results (351)

Accurate Prognostics and Health Management (PHM) of cutting tools in Computer Numerical Control (CNC) milling machines is essential for minimizing downtime, improving product quality, and reducing maintenance costs. Previous studies have frequently applied deep learning, particularly Long Short-Term Memory (LSTM) neural networks, for tool wear prediction and Remaining Useful Life (RUL) prediction. However, they often rely on simplified datasets or single architectures limiting industrial relevance. This study proposes a novel ensemble-LSTM framework that combines LSTM, BiLSTM, Stacked LSTM, and Stacked BiLSTM architectures using a GRU-based meta-learner to exploit their complementary strengths. The framework is evaluated using the publicly available PHM’2010 milling dataset, a well-established industrial benchmark comprising comprehensive time-series sensor measurements collected under variable loads and realistic machining conditions. Experimental results show that the ensemble-LSTM outperforms individual LSTM models, achieving an RMSE of 2.4018 and an MAE of 1.9969, accurately capturing progressive tool wear trends and adapting to unseen operating conditions. The approach provides a robust, reliable solution for real-time predictive maintenance and demonstrates strong potential for industrial tool condition monitoring. Full article

The development of modern machining and manufacturing industry puts forward higher requirements for the machining accuracy of machine tools. The thermal error of the machine tool spindle directly affects the accuracy of the machined workpiece. To improve the accuracy of thermal error prediction, this paper conducts temperature field analysis for the thermal error of the machine tool spindle and employs the Whale Optimization Algorithm (WOA) to optimize the temperature field parameters, aiming to establish a spindle temperature field model. This approach avoids the problem that traditional measurement methods cannot obtain the temperature of key rotational positions of the spindle and provides a new method for the selection of temperature-sensitive points in the thermal error measurement process. Initially, a spindle Product of Exponentials (POE) error model is constructed to map the five errors of the spindle to three-dimensional vectors in the machine tool space. Subsequently, the Whale Optimization Algorithm (WOA) is used to optimize the physical parameters of the spindle, and the optimal spindle temperature field model is determined. The calculated spindle thermal error data and temperature field model data are input into the OLGWO-SHO-CNN model for training. Finally, a case study is carried out on a machining center, and the trained model is used to perform compensation verification under constant and variable speed conditions, respectively. The experimental results show that under the constant speed condition, the compensation rates of the X-axis, Y-axis, and Z-axis are 77.2%, 73.1%, and 88.7%, respectively; under the variable speed condition, the compensation rates of the X-axis, Y-axis, and Z-axis are 74.7%, 78.2%, and 88.0%, respectively. The compensation results indicate that the established spindle temperature field model and the OLGWO-SHO-CNN model have good robustness and accuracy. Full article

This study addresses the fabrication challenges associated with producing diverse geometries for concrete dry connections, particularly regarding cost, time, and geometric limitations. The research investigates methods for fabricating precise, rebar-free dry connections in concrete, focusing on stamping and green-state computer numerical control (CNC) milling. These methods are evaluated using metrics such as dimensional accuracy, tool abrasion, and energy consumption. In the stamping process, a design of experiments (DOE) approach varied water content, concrete age, stamping load, and operational factors (vibration and formwork) across cone, truncated cone, truncated pyramid, and pyramid geometries. An optimal age range of 90 to 105 min, within a broader operational window of 90 to 120 min, was identified. Geometry-specific exceptions, such as approximately 68 min for the truncated cone and 130 min for the pyramid, were attributed to interactions between shape and age rather than deviations from general guidance. Within the tested parameters, water fraction primarily influenced lateral geometric error (diameter or width), while age most significantly affected vertical error. For green-state milling, both extrusion- and shotcrete-printed stock were machined at 90 min, 1 day, and 1 week. From 90 min to 1 week, the total milling energy increased on average by about 35%, and at one week end-face (head) passes caused substantially higher tool wear, with mean circumference losses of about 3.2 mm for head engagement and about 1.0 mm for side passes. Tool abrasion and energy demand increased with curing time, and extrusion required marginally more energy at equivalent ages. Milling was conducted in two engagement modes: side (flank) and end-face (head), which were evaluated separately. End-face engagement resulted in substantially greater tool abrasion than side passes, providing a clear explanation for tolerance drift in final joint geometries. Additionally, soil-based forming, which involves imprinting the stamp into soft, oil-treated fine sand to create a reversible mold, produced high-fidelity replicas with clean release for intricate patterns. This approach offers a practical alternative where friction and demolding constraints limit the effectiveness of direct stamping. Full article

Sustainable development is currently a key priority in improving production systems, requiring an integrated approach that combines economic efficiency, environmental responsibility, and rational energy management. In response to these challenges, this article presents a novel application of computer simulation as a tool for comprehensively assessing the impact of technological improvements in the machining process. The study introduces and compares two models: a baseline model representing the actual state of the machinery fleet with conventional machine tools, and an innovative alternative model incorporating modern numerically controlled (CNC) machines. The results demonstrate, for the first time in this context, that the implementation of CNC technology not only significantly reduces process time and energy demand but also improves resource efficiency, thereby lowering CO₂ emissions and operating costs. This research highlights the innovative use of computer simulation to support decision-making in sustainable manufacturing, offering a practical framework for evaluating technological modernization options and promoting the sustainable development of production enterprises. Full article

To address brittle cracks in ceramic tools, an exogenous precursor ceramic repair agent was developed and applied to Al₂O₃/TiC/NiMo composite ceramic tools, which were treated by a two-step heat treatment process (heating at 3 °C/min to 300 °C for 60 min, heating the sample at 5 °C/min to 500, 600, 700, 800, and 900 °C, holding each for 60 min). The crack healing mechanism and temperature dependency of the repair agent were investigated. Cutting performance, including surface roughness, cutting force, and tool life, was optimized using an L9(34) orthogonal design. The results show that at 900 °C, the repair agent decomposed to form SiOC (Silicon Oxycarbide) amorphous phase and TiB₂ reinforced phase, filling the cracks and achieving atomic-level diffusion bonding. The flexural strength of the repaired sample recovered to 79.9% of the initial value (456.5 MPa), a 196.4% increase compared to the unrepaired sample. Optimal cutting parameters were found to be a cutting speed of 200 m/min, back draft of 0.1 mm, and feed of 0.1 mm/r. Under these conditions, surface roughness was 0.845 μm, cutting temperature was 258 °C, and stable tangential force was 70 N. The effective cutting distance of the repaired tool was increased from 1300 m to 1700 m. Wear was primarily abrasive and adhesive wear, and the SiOC phase formed by the repair agent helped to fill and repair the flank, thus extending tool life. Full article

The electric spindle serves as a critical component in enabling a highly dynamic response, stable torque output, and precise motion control for the main cutting operations of CNC machine tools. The design precision and control performance of its drive motor directly influence the geometric accuracy, surface quality, and overall machining efficiency of the workpiece, thereby determining the comprehensive performance of advanced CNC systems. This paper begins with a systematic review of the global industrial layout of CNC machine tool and electric spindle manufacturers, highlighting regional clustering patterns and technological development trends across key manufacturing regions. Subsequently, it classifies and elaborates on the differentiated technical requirements for the electric spindle motor in terms of wide-speed-range servo capability, high-efficiency operation, adaptability to high-speed and high-power cutting loads, and precision maintenance under high-speed conditions, based on the process characteristics of different types of CNC machine tools. A comprehensive overview of the current state of research is provided with respect to electric spindle motor design and control technologies. Finally, forward-looking perspectives are presented on future development directions, particularly in the areas of multi-physics coupling co-design and the integration of intelligent control algorithms, aiming to offer a solid theoretical foundation and strategic guidance for the advancement and engineering application of high-performance electric spindles. Full article

Thermal error is a significant factor affecting the machining accuracy of machine tools, and error compensation is an economical and effective method to improve machine tool accuracy. However, traditional modeling methods face challenges such as insufficient nonlinear mapping capability and difficulty in parameter optimization when processing time-series data. This paper establishes a thermal error model using a Long Short-Term Memory (LSTM) neural network optimized by the Particle Swarm Optimization (PSO) algorithm (PSO-LSTM). Through thermal characteristic experiments, thermal error data and temperature rise data at various points of the T55II-500 CNC machine tool during actual machining were collected. First, fuzzy clustering and global sensitivity analysis were employed to identify the temperature-sensitive points of the machine tool. Using the temperature rise data of these sensitive points and the thermal errors of machined workpieces as data samples and optimizing the LSTM prediction model with the PSO algorithm, a PSO-LSTM thermal error prediction model was established. To verify its superiority and practicality, this paper conducts a comparative analysis with traditional thermal error prediction models based on Backpropagation (BP) neural network, Long Short-Term Memory (LSTM) network, Multiple Linear Regression (MLR), and Multivariate Nonlinear Regression (MNR). The results show that the PSO-LSTM model outperforms the other models in terms of relative error, average residual, maximum residual, and mean squared error. On this basis, a real-time thermal error compensation system was developed. Under the conditions of near-constant temperature (19.34–20.36 °C), warm natural ventilation (20.63–22.13 °C), and a wider variable temperature range (18.64–28.24 °C), the compensated thermal errors converge from 52 μm, 57 μm, and 67 μm to 4–12 μm, 6–11 μm, and 5–9 μm, respectively, with precision improved by 86%, 88%, and 86%. This effectively reduces the impact of thermal errors and improves the machining accuracy of the machine tool. Full article

The prediction of remaining tool life is critical in practical manufacturing, as it enables early identification of tool wear conditions, preventing downtime and quality issues caused by tool failure. This, in turn, helps reduce costs and improve production efficiency. To address the limitations of traditional models in modeling complex temporal dependencies and capturing critical moment information, a hybrid deep learning model based on a TCN-BiLSTM-Attention architecture was proposed. The model first enhanced low-level features from raw multivariate signals using an initial feature extraction module. Local temporal patterns were then extracted by an improved Temporal Convolutional Network (TCN), followed by a Bidirectional Long Short-Term Memory (BiLSTM) network to capture global sequential dependencies. An attention mechanism was further introduced to strengthen the model’s ability to focus on and select critical features over time, enabling deeper fusion and modeling of complex temporal dynamics. Finally, a multi-layer regression network was applied to predict the remaining tool life. Experimental validation on the PHM 2010 dataset demonstrates that the proposed model outperforms other comparative models. Among the three evaluation metrics (RMSE, MAE, and R²), the optimal results are 0.0184, 0.0127, and 0.9959 respectively, which fully proves that the model exhibits excellent prediction performance. Ablation studies further confirm the contribution of each module to the overall performance improvement. Full article

Timely and effective identification of the tool wear condition is crucial for ensuring the machining quality of CNC machine tools. In most industrial scenarios, the cost of sample collection is high, so only a small number of samples are available for model training, making it difficult for the existing tool wear condition monitoring (TCM) methods based on deep learning to achieve high performance. To address this problem, this paper proposes a TCM method based on the improved symmetric dot pattern (SDP) enhanced ResNet18. Firstly, the time series sample data is converted into grayscale matrices through SDP, the correlation coefficient between the grayscale matrices is calculated, and the optimal parameter combination of SDP is determined according to the objective of minimizing the correlation coefficient. Then, the cutting force signal is converted into a lobe diagram of the optimized SDP to enrich the sample feature information. Next, the SDP lobe diagram is input into ResNet18 for few-shot learning. The results of a series of TCM experiments demonstrate that the proposed method is significantly superior to the STFT and GAF based methods. Full article

Against the backdrop of the rapid development of the manufacturing industry, online monitoring of tool wear status is of great significance for enhancing the reliability and intelligence of CNC machine tools. This paper presents an intelligent tool wear condition monitoring model (CPSO-CNN-BiLSTM-AM) that integrates the improved Chaotic Particle Swarm Optimization (CPSO) algorithm with the CNN-BiLSTM network incorporating an attention mechanism. The aim is to extract the global features of long-sequence monitoring data and the local features of multi-spatial data. Chaos theory and the mutation mechanism are introduced into the CPSO algorithm, which enhances the algorithm’s global search ability and its capacity to escape local optimal solutions, enabling more efficient optimization of the hyperparameters of the CNN-BiLSTM network. The CNN-BiLSTM network with the introduced attention mechanism can more accurately extract the spatial features of wear signals and the dependencies of time-series signals, and focus on the key features in wear signals. The study utilized the IEEE PHM2010 Challenge dataset, extracted wear features through time-domain, frequency-domain, and time-frequency domain methods, and divided the training set and validation set using cross-validation. The results show that in the public PHM2010 dataset, the average MAE of the model for tools C1, C4, and C6 is 0.83 μm, 1.01 μm, and 1.34 μm, respectively; the RMSE is 0.99 μm, 1.79 μm, and 0.88 μm, respectively; and the MAPE is 0.95%, 1.41%, and 1.01%, respectively. In the self-built dataset, the average MAE for tools A1, A2, and A3 is 1.35 μm, 1.19 μm, and 1.83 μm, respectively; the RMSE is 1.41 μm, 1.98 μm, and 1.90 μm, respectively; and the MAPE is 1.67%, 1.55%, and 1.81%, respectively. All indicators are superior to those of comparative models such as LSTM and PSO-CNN. The proposed model can effectively capture changes in different stages of tool wear, providing a more accurate solution for tool wear condition monitoring. Full article

In this paper, an integrated geometric and thermal-induced errors prediction approach for active error compensation in machine tools is proposed and evaluated. The proposed approach is based on a hybrid of physical and neural network predictive modeling to drive an adaptive position controller for real-time error compensation including geometric and thermal-induced errors. Error components are formulated as a three-dimensional error field in the time-space domain. This approach involves four key steps for its development and implementation: (i) simplified experimental procedure combining a multicomponent laser interferometer measurement system and sixteen thermal sensors for error components measurement, (ii) artificial neural network-based predictive modeling of both position-dependent and position-independent error components, (iii) tridimensional volumetric error mapping using rigid body kinematics, and finally (iv) implementation of the real-time error compensation. Assessed on a turning center, the proposed approach conducts a significant improvement of the machine accuracy. The maximum error is reduced from 30 µm to less than 3 µm under thermally varying conditions. Full article

In the context of Industry 4.0 and smart manufacturing, predicting cutting tool remaining useful life (RUL) is crucial for enabling and enhancing the reliability and efficiency of CNC machining. This paper presents an innovative predictive model based on the data fusion architecture of Graph Neural Networks (GNNs) and Transformers to address the complexity of shallow multimodal data fusion, insufficient relational modeling, and single-task limitations simultaneously. The model harnesses time-series data, geometric information, operational parameters, and phase contexts through dedicated encoders, employs graph attention networks (GATs) to infer complex structural dependencies, and utilizes a cross-modal Transformer decoder to generate fused features. A dual-head output enables collaborative RUL regression and health state classification of cutting tools. Experiments are conducted on a multimodal dataset of 824 entries derived from multi-sensor data, constructing a systematic framework centered on tool flank wear width (VB), which includes correlation analysis, trend modeling, and risk assessment. Results demonstrate that the proposed model outperforms baseline models, with MSE reduced by 26–41%, MAE by 33–43%, R² improved by 6–12%, accuracy by 6–12%, and F1-Score by 7–14%. Full article

This study compares one traditional prediction model—the conventional single-variable interpolation method—with two newly developed models: an improved multi-variable interpolation model extended from the single-variable formulation, and a more advanced NURBS-based multi-variable interpolation model. All models are integrated into a real-time volumetric error compensation framework embedded within a CNC controller, enabling in-kernel correction without external hardware. Accuracy verification is carried out using planar body diagonal measurements obtained from a dense on-machine PDGE data grid across the coupling plane. Quantitatively, the improved multi-variable interpolation model reduces diagonal errors by 71%–74%, while the NURBS-based model achieves 82% (T1) and 84% (T2) reductions, delivering an additional 18%–19% improvement relative to the single-variable baseline. The in-kernel evaluation satisfies 2–4 ms interpolation cycles, confirming real-time feasibility. The proposed framework provides a compact, data-driven solution for predicting and compensating cross-talk-induced PDGEs in precision machine tools. Full article

This paper investigates the influence of substrate grain size on the behavior of a multilayer CrN/CrAlN coating, with the bilayer thickness varying across the cross-section in the range of 200–1000 nm. The substrate tools were made of WC-Co sintered carbide with three different grain sizes. The coatings were subjected to mechanical and tribological tests to assess their performance, including nanohardness, scratch resistance, and tribological testing. The coating’s roughness was measured using a 2D profilometer. Additionally, the chemical composition and surface morphology were analyzed using Scanning Electron Microscopy (SEM) and Energy Dispersive X-Ray Spectroscopy (EDX). The durability tests were performed on an industrial CNC machine tool on the particleboard. The results revealed that tools with ultra-fine nano-grain (S) and micro-grain (T) WC-Co substrates exhibited a significant increase in tool durability by 28% and 44%, respectively. Significant differences in the microgeometry of the substrate U, especially in relation to the tool based on substrate S, explain the lack of improvement in its durability despite the use of a multilayer coating. Full article

Indexable face milling tools are often used to machine workpieces with large axial and radial depth of cuts, and thus, the inserts quickly wear out in machining. A kernel technique of smart machining is tool wear compensation, which is to regularly and automatically measure the insert radius/length with a laser tool setter on the machine table during machining, and compensate them in the subsequent machining. Another technique is tool condition monitoring, which is to calculate the insert flank wear width for tool condition and compare with its threshold. When it is less than but close to its threshold of invalid inserts, the cutting tool is automatically changed right before it becomes invalid. On-machine laser tool setters have been equipped in CNC machine tools for several years; however, they cannot conduct cutting tool condition monitoring. The main reason is that the insert flank wear width cannot be measured on the on-machine laser tool setter, and the status quo is that the cutting tool is replaced either too early or too late. To address this problem, a novel tool condition monitoring technique of determining the insert flank wear width of indexable face milling tool using on-machine laser tool setters is proposed. According to the insert geometry, the worn cutting edge and a new workpiece milling mechanism proposed in this work, the insert flank wear width can be calculated. In machining, the insert radius wear is measured on the on-machine laser tool setter, and the insert flank wear width is calculated to evaluate whether it is invalid soon. The results indicate that the optimal height for radius measurement is located near the intersection of the corner and side edges point MR₃, and close to the cutting depth point MR₅. A wear land width threshold of 0.10 mm is established to define tool failure. The proposed calculation method achieves high accuracy, maintaining calculation errors within 14.00%. The inserts can be used in good condition with the maximum lifespan. This method has been verified in machining applications and can be directly applied in industry. Full article