Search Results (196)

This study addresses the integration of sustainability-oriented analytics and decision support in CNC milling through a hybrid information system combining structured data management, sustainability KPIs, rule-based expert reasoning, and machine learning models. The proposed framework is evaluated using the publicly available NASA Ames Milling Tool Wear Dataset. Sustainability indicators related to operational energy demand, tool degradation, and vibration/acoustic-emission response are computed from machining parameters and sensor-derived features. Random Forest and Support Vector Machine models are used for tool wear classification. The expert system applies deterministic rules to identify operational risks, which are combined with machine learning predictions through a hierarchical decision-fusion strategy. Under case-wise cross-validation, the Random Forest achieved a mean classification accuracy of 73.8%, while the Support Vector Machine achieved 68.8%. The expert system most frequently identified elevated vibration-index and acoustic-emission conditions, while critical clean-production risks occurred rarely. Overall, the results demonstrate the feasibility of integrating expert knowledge, sustainability KPIs, and data-driven models into a hybrid information system for decision support in CNC milling environments. The proposed framework provides a foundation for future research on hybrid information systems supporting sustainable manufacturing and intelligent decision making. Full article

Industrial IoT maintenance often stops at health prediction, leaving maintenance, rescheduling, safety, and communication to separate decision processes. This study presents MIRA, a safety-constrained graph-based multi-agent reinforcement learning architecture for joint prescriptive maintenance, production rescheduling, and event-triggered communication. Machine condition was estimated from CNC milling data using temporal convolutional models; because predictive uncertainty failed a predefined validation gate, the controller used deterministic health estimates. Evaluation covered five controllers, six simulated scenarios, and 1800 matched episodes. Relative to Graph-MAPPO, MIRA reduced operational cost by 9.38%, weighted tardiness by 28.10%, unexpected failures by 17.39%, message count by 84.98%, and transmitted data by 83.83%, while increasing on-time completion by 23.55%, without a detectable difference in corrected critical-message recall. Across the three independently trained seeds, failures, safety violations, and message count favored MIRA consistently, whereas cost and tardiness favored MIRA in two seeds. Disabling the execution shield increased safety violations from 0 to 3.56 per episode. Post-training variation in the projected-health safe-start threshold from 0.124 to 0.132 produced no safety violations and only small changes in aggregate operational outcomes. Cross-domain health transfer to PHM 2010 failed without adaptation. The results support simulator-level decision coordination, while broader replication, variable-size deployment, and factory validation remain necessary. Full article

In conventional CNC machining, the workpiece clamping pose is registered with a preset CAD model under multiple geometric constraints to establish the machining reference frame. The tool path, generated from this model, is subsequently used to produce components of identical geometry. However, this paradigm proves inadequate when a final shape must accommodate morphological variations specific to each individual blank. Manual grinding, as an alternative, is not only inefficient and hazardous but also relies heavily on subjective quality assessment. To address these challenges, we propose an adaptive local-region milling strategy tailored for blanks with similar yet non-identical surface morphologies, enabling the finished geometry to adjust dynamically to each workpiece. Under conditions of under-constrained clamping, visual positioning is first employed to automatically locate the target regions. Line laser scanning is then conducted over the planned area to acquire high-density point clouds. Through segmentation, points lying outside the region to be machined are extracted, from which a theoretical post-machining surface is reconstructed. Milling toolpaths are subsequently planned based on this reconstructed model to compensate for surface variations across different blanks. Experimental validation on a three-axis CNC milling machine demonstrates that the proposed adaptive strategy effectively replaces manual grinding by removing the bulk of the machining allowance from locally variant surfaces. With the reconstructed model serving as the reference, 77.1 percent of the machining errors fall below 0.055 mm. These results confirm that the method yields a smooth and level surface finish, thereby meeting the fundamental requirements for such adaptive machining tasks. Full article

Corner cleaning is a critical sub-stage of finishing and is used to remove residual material left by previous machining operations, thereby ensuring the designed dimensional precision and surface quality. Although essential in CNC machining, many current industrial CAM software like UG NX, PowerMill systems were largely developed based on early-stage theoretical frameworks. Meanwhile, the geometric shapes of machined workpieces are becoming increasingly complex, making such software gradually unable to meet the growing requirements for tool path quality. This paper establishes a rigorous mathematical framework to bridge the gap between industrial practice and theoretical modeling. Based on this framework, an enhanced Principal Component Analysis (PCA) method is proposed to generate an optimal drive curve by integrating geometric variance maximization with vector field guided directional optimization. Furthermore, a robust offset strategy is presented, incorporating self-intersection detection, critical-point trimming, and segment connection mechanisms to ensure path continuity and smoothness. Experimental results and real machining cases demonstrate that the proposed method outperforms the widely used commercial software in terms of robustness and path smoothness, validating its effectiveness and practical applicability. Full article

In the current context of the manufacturing industry, optimizing the cutting parameters to achieve a controlled surface roughness involves costly and time-consuming experimental efforts. The present study addresses this challenge by developing robust machine learning-based approximate functions for predicting surface roughness (Ra) resulting from toroidal milling on a five-axis CNC. The research includes an experimental design conducted under real production conditions on C45 steel. The relatively small experimental dataset was augmented, normalized, and then scripts were written for four prediction models: two artificial neural network architectures and two models based on decision trees. Their performance was analyzed based on MSE, RMSE, R², and MRA metrics. The results obtained reveal significant differences between the models, highlighting solutions with high accuracy, excellent robustness, and superior generalization capacity for new data. The study highlights the high potential of prediction models in optimizing machining processes, providing an effective way to reduce costly physical experiments and increase productivity in industrial environments. Full article

This study investigates the critical relationship between machining-induced surface integrity and the effectiveness of subsequent anti-corrosion protection for high-strength Al 7136-T76511 aerospace alloy. Given the alloy’s susceptibility to exfoliation corrosion, ensuring high-quality surface substrates for protective coatings is paramount. The research aims to model the influence of end milling parameters—cutting speed, depth of cut, and feed per tooth— on surface roughness to establish a topographical risk prognosis framework for subsequent coating vulnerability. A comprehensive full-factorial experimental design involving 150 distinct cutting regimes was evaluated on a CNC machining center. Statistical analysis using ANOVA showed that cutting speed is the most significant factor, contributing 83.89% to the variance of longitudinal Ra. A critical resonance zone was identified between 570 and 610 m/min, where the model predicts high instability and surface integrity degradation. The developed mathematical models achieved high precision, with coefficients of determination (R²) ranging between 85% and 88%. The research identifies a critical “danger zone” of dynamic instability between 570 and 610 m/min, where resonance significantly increases data dispersion (standard deviation = 0.112 µm compared to 0.051 µm in stable regimes). Findings demonstrate that even when average R_a values remain within industrial limits, vibration-induced micro-cracks, and severe chatter marks function as geometric precursors that theoretically lower the structural barrier efficiency of subsequent protective films. This study establishes that prioritizing process stability over nominal roughness minimization is essential for the structural integrity of critical aerospace components. Full article

To address the pervasive engineering challenges of rigid interference and subpar machining efficiency encountered during the complex freeform surface polishing of P20 mold steel, this study proposes and fabricates a structurally designed, five-petal composite compliant polishing tool via fused granulation fabrication (FGF). The tool structurally integrates a passive thermoplastic polyurethane (TPU) compliant buffer layer with an active PA66/diamond micro-cutting functional layer, achieving monolithic precision assembly through dual-temperature-zone 3D printing. Tensile mechanical characterization (n = 6) reveals that the composite interface attains an average ultimate tensile strength (UTS) of 59.39 ± 15.41 MPa (with a peak of 78.90 MPa) and an average elongation at break of 27.42 ± 7.41%, demonstrating exceptional structural robustness and fracture toughness under heavy-load abrasive machining conditions. During adaptive polishing validations on complex convex topographies and deep concave mold cavities, the compliant tool effectively compensated for normal vector spatial errors intrinsic to three-axis CNC machining via passive geometric adaptation. Topographical evaluations suggest a ductile-regime, differential asperity planarization material removal paradigm, which is attributed to the macroscopic 3D elastic deformation of the tool synergized with the proposed compliance of the polymer matrix. Following high-intensity sequential polishing regimens, the original macroscopic milling striations were substantially reduced. Quantitative profilometric analysis reveals that the average surface roughness of the convex profiles decreased from an initial 13.33 µm to 7.42 µm, while that of the restrictive deep concave features was reduced from 10.84 µm to 4.11 µm. Ultimately, this technological framework circumvents the traditional reliance on capital-intensive, six-degree-of-freedom robotic platforms, providing a scalable automated polishing protocol compatible with standard CNC systems for the cost-effective surface planarization of precision molds. Full article

Smart manufacturing requires real-time monitoring of multidimensional forces at the interface between the tool and workpiece in computer numerical control (CNC) machining. In this study, an innovative iKIT wireless extrasensory toolholder is introduced that is capable of high-fidelity, in situ, high-frequency sensing and monitoring of the cutting force, torque, and two-way bending moments. The hardware design of the system is outlined, highlighting a high-bandwidth miniature wireless transmission method and noncontact power supply and energy storage solution suitable for rotating machining environments. To assess the system performance, comprehensive milling tests were performed on aluminum alloy materials, and the relationship between the process parameters and changes in multidimensional mechanical loads was thoroughly examined. The experimental findings demonstrate that the smart toolholder detects precisely how parameter variations affect the loads. Multidimensional mechanical signals (torque and two-way bending moments) show a strong positive correlation with the feed rate and axial depth of cut, confirming the impact of the material removal rate on the system loads. Conversely, these signals are negatively correlated with spindle speed, accurately reflecting the effects of thermal softening and a reduced friction coefficient in aluminum alloys during high-speed cutting. This study not only offers a dependable hardware framework for integrating miniaturized sensors into toolholders, but also delivers accurate data to support digital twin models and adaptive control in machining processes. Full article

Five-axis CNC machining involves strong coupling among machining quality, material removal efficiency, and operational safety, making it difficult to obtain adaptive and feasible process parameters using conventional scalar-objective optimization methods. To address this problem, this study proposes a physics-constrained multi-objective multi-agent deep deterministic policy gradient framework, termed MOMADDPG, for Pareto-oriented optimization of five-axis machining parameters. A data-calibrated digital twin simulation environment is constructed to model five-axis kinematics, tool-workpiece engagement, cutting force, chatter tendency, spindle power, tool wear, actuator bounds, and collision risk. The PHM Society 2010 milling dataset is used to calibrate the cutting force and tool wear sub-models, while five-axis motion, tool orientation variation, and engagement conditions are generated within the digital twin environment. In the proposed framework, three heterogeneous agents are assigned to quality preservation, efficiency improvement, and safety assurance, respectively. A hierarchical attention Actor is designed to enhance feature extraction under partially observable machining conditions, while vector-valued dual Critics preserve objective-specific value information. Physical constraints are handled using adaptive Lagrangian multipliers, and a Pareto archive-guided preference curriculum is introduced to improve the diversity of feasible non-dominated solutions. Simulation results show that MOMADDPG achieves a task success rate of 98% and a hypervolume value of 0.674 after training. Compared with representative baselines, including DQN, MADDPG, MAAC, and MAPPO, the proposed method provides better Pareto-front approximation, higher task feasibility, and stronger robustness under process perturbations in the data-calibrated five-axis simulation environment. The results demonstrate the potential of combining digital twins and multi-objective multi-agent reinforcement learning for safe and adaptive parameter optimization in five-axis machining simulations. Further validation on physical five-axis CNC systems is still required before industrial deployment. Full article

Controlling macro-geometrical errors in the dry drilling of ductile cast iron remains a critical challenge for sustainable and cost-efficient automotive component manufacturing. This paper investigates the influence of cutting speed (v_c) and feed per revolution (f_n) on the dimensional and shape accuracy of holes drilled in EN-GJS-500-7 ductile cast iron using an HSS DIN 338 helical drill (Ø 11.8 mm, Ceratizit) on an AVIA VMC800 CNC milling centre. A one-factor-at-a-time (OFAT) experimental design was applied: the feed effect was evaluated at v_c = 10 m/min with f_n ∈ {0.10, 0.15, 0.20} mm/rev, while the speed effect was evaluated at f_n = 0.20 mm/rev with v_c ∈ {10, 25, 30} m/min. Cutting forces, torques, and vibration accelerations were recorded using an HBM MSC 10 transducer and a PCB 356A01 tri-axial accelerometer. Hole geometry was assessed on a Zeiss Contura G2 coordinate-measuring machine (CMM), and surface texture was evaluated with a TOPO 01P contact profilometer. The expanded measurement uncertainty (k = 2) was estimated based on duplicate test specimens. All drilled holes fell within the IT12 dimensional tolerance (PN-EN 22768-1:1999 grade c), with diameter oversizes ranging from +0.26 mm to +0.46 mm relative to the nominal bore. Cutting speed was identified as the dominant factor affecting both diameter oversize and cylindricity, which increased by 60% (from 0.10 to 0.16 mm) as v_c rose from 10 to 30 m/min. Vibration accelerations increased nonlinearly between v_c = 25 and 30 m/min (by a factor of 2.5×), indicating an approach to a structural resonance condition. The lowest surface roughness (R_a = 6.6 µm) was obtained at v_c = 25 m/min. These findings establish clear physical baselines for tool deflection limits, demonstrating that managing dynamic process stability is vital for optimising macro-geometrical accuracy in the dry machining of cast iron alloys. Full article

Tool wear in CNC milling increases friction and torque demand at the tool-workpiece interface, which is reflected in spindle motor current. This study develops a non-intrusive tool wear condition classification method using spindle motor current monitoring during practical CNC milling of commercial medium-carbon steel workpieces (JIS S50C/AISI SAE 1050-equivalent; as-received and non-heat-treated; nominal laboratory hardness approximately 4.3 HRC). Experiments were performed on a Tongtai MDV-508 vertical machining center at fixed cutting conditions (3000 rpm spindle speed, 2 mm axial depth of cut, 5 mm cutting width, and 300 mm/min feed rate) using eight TiAlN-coated fine-grain WC–Co solid carbide end mills (10 mm diameter, four flutes; nominal Co binder approximately 10 wt%). An oil-based HS Highstart/HS-SSHS-BH10 cutting fluid was applied through the machine external coolant nozzle in flood mode at an estimated nominal flow rate of approximately 3 L/min and near-room coolant temperature (25 ± 2 °C), and was used as supplied without dilution. A clamp-type AC current sensor was installed on one phase line supplying the spindle motor, and current was acquired using an NI-9221 module at 20 kHz. Cutting intervals were isolated by envelope-based segmentation, concatenated, and divided into 1 s windows (0.5 s overlap) for feature extraction. Three feature sets were evaluated: time-domain statistics, frequency-domain statistics, and an FFT→PCA hybrid representation. Tool states (New, Mid-life, Old) were labeled using post-process surface roughness Ra thresholds supported by microscope observation. The PCA transformation was fitted only on training data and then applied to the held-out test data. A logistic regression classifier achieved 97.44% test accuracy (152/156 windows; 95% Wilson CI: 93.59–99.00%) with the PCA-hybrid features, outperforming time-domain (89.74%) and frequency-domain (94.87%) models. The results support spindle current monitoring as a low-cost approach for quality-aligned tool condition monitoring, while the external validity remains limited to the tested machine, material, tool, coolant, and cutting-parameter combination. Full article

In the measurement of volumetric errors in CNC machine tools using multi-station laser tracking technology, the coordinate calibration accuracy of external measurement base stations is a key factor determining the system’s final accuracy. Traditional calibration approaches typically use the commanded positions of the machine tool directly to inversely determine base station coordinates, which results in strong coupling between inherent geometric errors and base station parameters. Consequently, the measurement accuracy cannot be properly evaluated, and metrological traceability of the results remains difficult to achieve. To address this issue, this paper proposes a novel calibration principle based on an independent external physical standard and develops a base station calibrator independently. This device employs a precision turntable, G5-grade precision spheres, and electromagnet groups to construct an equivalent target with four feature points at the spindle end. Verified by a high-precision coordinate measuring machine (CMM), the maximum difference in repeated calibrations of the device is 1.4 µm, indicating its excellent positioning repeatability. The calibrator was further applied to measure the positioning errors of a CNC milling machine, and comparative experiments were performed with a Renishaw XL-80 laser interferometer. The results indicate that the error variation trends obtained from the two measurement principles are highly consistent. In both the X-axis and Y-axis directions, the maximum deviations of linear errors are controlled within 3.7 µm, while the maximum deviations of angular errors remain within 4.6 µrad. Furthermore, the reliability of the system data was confirmed through an uncertainty analysis. The external physical standard developed in this study ensures that base station calibration accuracy is not affected by the inherent errors of the machine tool, providing a novel and reliable scheme for high-precision calibration and metrological traceability of machine tool spatial errors. Full article

This study investigates ultrasonic vibration-assisted (UV) CNC milling of hardened 90CrSi steel cylindrical surfaces, with emphasis on ultrasonic horn design, experimental analysis, and multi-objective optimization of machining parameters, addressing the need for an integrated framework combining system design, experimental validation, and multi-objective optimization. A quarter-wavelength ultrasonic horn was designed and experimentally validated to operate at a frequency of 20 kHz. By adjusting the horn–workpiece system, stable vibration amplitudes were achieved to enable effective ultrasonic-assisted milling of cylindrical surfaces. Milling experiments based on a Box–Behnken design were conducted to examine the effects of vibration amplitude, cutting speed, feed rate, and radial depth of cut on material removal rate (MRR) and surface roughness (Ra). Surrogate models using response surface methodology (RSM) and Gaussian process regression (GPR) were developed to predict machining performance. A GPR-assisted NSGA-II algorithm was then applied to simultaneously maximize MRR and minimize Ra, resulting in a well-defined Pareto front that reveals the trade-off between machining productivity and surface quality. Furthermore, an AHP-based decision-making approach was employed to select preferred machining conditions from the Pareto-optimal solutions. The GPR models demonstrated high predictive accuracy (R² > 0.98), and validation experiments confirmed the reliability of the predicted optimal results, with deviations below 5%. In addition, a comparative analysis between ultrasonic-assisted and conventional milling showed that MRR increased by 10.81–40.17%, Ra decreased by 27.11–44.44%, and cutting force was reduced by 14.2–42.65%, providing direct experimental evidence of improved machinability. The results demonstrate that the proposed integrated framework provides an effective strategy for optimizing ultrasonic vibration-assisted milling processes and improving the machinability of hardened 90CrSi cylindrical surfaces. Overall, the proposed framework provides a practical and cost-effective strategy for enhancing machining performance and offers a robust approach for multi-objective optimization of ultrasonic vibration-assisted milling processes. Full article

Hybrid CNC and additive manufacturing platforms often rely on host-assisted or otherwise overdimensioned control architectures to achieve deterministic multi-axis motion, increasing system cost and complexity. This paper presents a fully microcontroller-based timer–DMA motion execution architecture that eliminates the need for external processors or FPGA-based execution, enabling deterministic multi-axis synchronization under the tested conditions in a simpler, more cost-effective way. The proposed framework integrates motion planning, precise step-time computation, and hardware-assisted pulse generation within a unified embedded control architecture. The main novelty lies in the systematic use of timer and DMA peripherals to offload time-critical pulse execution from the microcontroller core, allowing it to focus on motion planning and precise step-time computation. Unlike segmentation-based approaches, the duration of each individual step is calculated directly without fixed-interval segmentation, enabling high motion resolution while avoiding per-step interrupts that introduce jitter at high motion speeds. The architecture was validated on a hybrid platform capable of both milling and material extrusion. Experimental results confirmed real-time feasibility within practical on-chip memory limits and demonstrated very small interpolation errors caused mainly by timer quantization, comparable to those observed in host-processor-based motion systems. Machining and additive-manufacturing experiments further confirmed stable execution and accurate trajectory tracking under real operating conditions. Full article

Surface quality of machined wood-based panels plays a key role in subsequent processing and product performance; however, its formation during CNC edge milling remains insufficiently understood, particularly for materials with different structural characteristics, including recycled content. This study investigates the influence of feed per tooth, milling strategy, and material structure on surface quality during CNC edge milling of particleboards manufactured from alternative lignocellulosic resources. Six board variants were experimentally produced and machined on a five-axis CNC machining center Morbidelli m100 using a single-edge milling cutter, with feed per tooth varied at three levels and both climb and conventional milling strategies applied. Surface quality was evaluated using a non-contact 3D optical profilometer Keyence VR-6000, and roughness (R_a) and waviness (W_z) parameters were analyzed. The results showed that surface roughness increased with increasing feed per tooth for all materials, with an increase of approximately 30%–70%. Statistical analysis confirmed a significant effect of feed per tooth and material type, while milling strategy and its interaction with material were not statistically significant. Materials with higher surface heterogeneity (CVR_a) showed increased roughness and greater sensitivity to feed. A statistically significant positive relationship was found between surface heterogeneity (CVR_a) and roughness sensitivity (ΔR_a), indicating that materials with higher surface heterogeneity (CVR_a), which likely reflects variability in their internal structure, are more sensitive to changes in feed per tooth. Full article