International Journal of Advanced Manufacturing Technology最新文献

This research explores the fabrication and characterisation of metamaterial-architecture-inspired 3D-printed polymer substrates with complex geometries, subsequently functionalized with titanium (Ti) and diamond-like carbon (DLC) coatings deposited by direct current magnetron sputtering. In this context, the term metamaterial-architecture-inspired refers exclusively to the engineered surface geometry and does not imply the experimental demonstration of emergent metamaterial properties. Polymeric substrates were fabricated via laser stereolithography using both an industrial (SLA-3500) and a low-cost (Form 1+) printing system, employing photoreactive resins in a layer-by-layer process. Ti and DLC thin films were subsequently deposited, and the resulting surfaces were characterised using reflected light optical microscopy and Raman spectroscopy to assess geometrical fidelity, coating conformity, and chemical-structural stability. Uniform coatings were successfully achieved on complex three-dimensional microtextures using both SLA systems. Substrates printed with the SLA-3500 exhibited well-defined layers and an average increase in valley curvature of approximately 2.8%, whereas Form 1 + printed samples showed a higher deviation of about 17.7% relative to the original design. Raman spectroscopy confirmed the presence of characteristic D and G bands at 1396 cm⁻¹ and 1589 cm⁻¹ in DLC-coated samples on both Accura^®60 and Clear FLGPCL 02 substrates, indicating graphitic carbon domains while preserving the chemical integrity of the underlying polymer. Ti-coated surfaces exhibited increased broadband intensity between 1200 and 1420 cm⁻¹, attributed to resin-metal interactions. Despite minor variations in spectral intensity, no significant shifts in vibrational frequencies were observed, demonstrating comparable molecular stability of both substrate systems following coating deposition. These results establish a reliable framework for the fabrication and surface functionalization of architected polymer substrates, enabling future investigations into structure-property relationships and application-specific functional performance.

Additive Manufacturing is an innovative technology that fabricates parts layer by layer. However, in Laser Powder Bed Fusion (LPBF), printed metal parts often exhibit residual stresses, deformations, and other defects due to non-uniform temperature distribution during the printing process. To mitigate these issues, an optimized scan sequence within each layer can improve thermal uniformity. Traditional optimization methods, which rely on domain knowledge and employ trial-and-error or heuristic approaches, often fail to achieve optimal solutions due to the complex nature of the problem. One major challenge in improving scan strategies lies in the vast search space required to optimize the scan sequence for individual scan tracks within each layer, making it difficult to identify the best solution. To overcome this challenge, this work proposes an innovative scan strategy, Reinforced Scan, that leverages reinforcement learning to intelligently determine the optimal scan sequence. The method introduces a novel reward function that accounts not only for temperature variance but also for the spatial uniformity of the temperature field. By structuring the optimization problem into multiple hierarchical levels, the approach significantly reduces computational demand and enhances the manageability of the optimization process. The effectiveness of the proposed Reinforced Scan is validated through Netfabb™ Local Simulation and real-world laser scanning experiments on a Ti-6Al-4V thin plate. Its performance is compared against conventional heuristic scan sequences. Both simulation and experimental results demonstrate that Reinforced Scan achieves superior outcomes, notably reducing residual stress compared to traditional methods.

Modelling complex manufacturing processes presents significant challenges related to accuracy and explainability. Physics-based models, while interpretable and generalizable, often suffer from reduced accuracy due to simplifications and incomplete system understanding. On the other hand, purely data-driven models are typically more accurate but lack transparency, limiting their trust and adoption in critical manufacturing applications. Existing hybrid approaches attempt to address these issues but often retain black-box AI components that compromise interpretability. In this study, we propose a novel hybrid modelling framework that intrinsically integrates physics-based models with explainable AI, to correct for modelling inaccuracies. This approach offers both high accuracy and transparent, traceable decision-making. Its effectiveness is demonstrated through a case study predicting the real-time position of cutting tools from accelerometer signals during ultra-precision diamond turning.

Accurate tracking of modal characteristics is a valuable diagnostic tool for condition monitoring of machine tool spindle units. While experimental modal analysis (EMA) is the conventional method used for machine tool modal identification, it is often impractical to implement in production settings due to the invasive and manual nature of the impact hammer test. In this study, a new technique for operational modal analysis (OMA) based on output-only vibration measurements obtained during a milling operation with variable spindle speed is proposed. Modal identification is performed using two OMA standard methods, namely stochastic subspace identification (SSI) and frequency domain decomposition (FDD). The modal characteristics are compared to values obtained from conventional EMA from impulse hammer testing on the static spindle, and from the operational spindle during cutting using force measurements collected by a table dynamometer. The percentage difference between the natural frequencies identified by the proposed OMA method and frequencies identified by conventional impulse hammer testing was less than 10%, and for the operational spindle during cutting tests, the difference was less than 3%. These results demonstrate the validity of a new modal identification method that can be practically implemented in production.

We present a scalable and environmentally friendly method for fabricating mechanically robust superamphiphobic coatings on aluminum alloy and zinc substrates using a dual-step process combining sandblasting (SB) and steam treatment (ST), followed by surface energy reduction with fluorinated molecules. This approach creates hierarchical micro/nano structures essential for omniphobic performance. On Al-alloy SB + ST surfaces we measured static contact angles of 162.0° (water), 156.1° (ethylene glycol), and 154.4° (peanut oil), while the corresponding Zn surfaces reached 160.1°, 156.0°, and 152.8°, respectively, with sliding angles below 5° across all tested liquids. The coatings retained high repellency after 50 tape-peeling cycles and 100 cm of sandpaper abrasion under a 500 g load (e.g., ethylene glycol > 140° and peanut oil ≈ 120°). They also showed resistance to water jet impact, excellent self-cleaning, and anti-fogging performance. Compared to conventional hot water treatment or chemical etching, this ST-based method enables faster, cleaner fabrication and significantly enhances mechanical durability making it a promising candidate for large-scale applications in anti-fouling, anti-corrosion, and protective surface technologies.

In heavy industries like oil and gas, and shipbuilding, maintaining process quality is challenging. These sectors face inconsistent manual procedures and a shortage of skilled operators regarding thermal cutting and bevelling for welding preparation tasks. Manual fitting and repetitive quality control modifications, especially during thermal cutting, significantly increase time consumption and hinder productivity. Traditional thermal cutting methods are prone to human error, resulting in inconsistent cut quality, and demand high expertise leading to variability in cut precision, increased rework, and material wastage. The objective of this work is to address these challenges by introducing real-time ultrasonic sensing into a robotic plasma cutting control system to automate the steel plate bevelling process. The ultrasonic sensor enables the system to dynamically adapt to variations in steel plate thickness before cutting, ensuring precise and consistent results. The solution begins by presenting an automated method for measuring thickness and computing bevel distance per sample. Secondly, it proposes adaptive adjustments to cutting parameters per sample, leveraging the ultrasonic sensor data to enhance accuracy and reduce the need for manual intervention. Finally, the approach introduces adaptive robotic path generation for cutting and utilizing real-time ultrasonic sensor data to optimize cutting paths. The outcome of this study is the successful development and validation of an adaptive robotic plasma cutting system for steel plate bevel applications, which leverages real-time ultrasonic sensor data to automate the parameter input process and robotic motion planning, demonstrating improved accuracy and efficiency compared to traditional approaches. The results demonstrate that ultrasonic-driven robotic cutting significantly reduces the average error cut percentage to 4.47% with deviations ranging from 0.13 to 0.23° for the bevel angle and 14.27% with deviations between 0.02 and 0.05 mm for root face deviation, compared to the standard cutting approach which has an average error of 18% with deviations ranging from 0.10 to 0.38 mm and 77.1% with deviation between 0.48 to 0.90°, respectively. This paper highlights the benefits of using advanced sensing technology, particularly ultrasonic sensors, to automate plasma bevel cutting for metal plates in the steel fabrication and welding sectors.

Coating sprays play a crucial role in extending the capabilities of optical measuring systems, especially when dealing with reflective surfaces, where excessive reflections, caused by incident light hitting the object surface, lead to increased noise and missing data points in the measurement results. This work focuses on metal additively manufactured parts, and explores how the application of a sublimating matting spray on the measured surfaces can improve measurement performance. The use of sublimating matting sprays is a recent development for achieving temporary coatings that are useful for measurement, but then disappear in the final product. A series of experiments was performed involving measurement by fringe projection on a selected test part pre- and post-application of a sublimating coating layer. A comparison of measurement performance across the experiments was run by computing a selected set of custom-developed point cloud quality indicators: rate of surface coverage, level of sampling density, local point dispersion, variation of selected linear dimensions computed from the point clouds. In addition, measurements were performed using an optical profilometer on the coated and uncoated surfaces to determine both thickness of the coating layer and changes of surface texture (matte effect) due to the presence of the coating layer.

The deep drawing process in the automotive industry generates up to 45% material waste. To address this issue, the folding-shearing process was developed as a drop-in solution, enabling the formation of parts in pure shear with minimal thickness variation. This process involves folding a blank while collecting the excess material in a region called the 'beak', which is subsequently sheared in-plane using a single set of tools moving in one forming direction. This paper investigates the extent to which the curvature of the geometry of the beak influences the resulting thickness distribution. A combination of physical and numerical trials demonstrates that a beak design with a negative Gaussian curvature reduces the maximum thickening by 65%. This reduction in thickening helps minimise the forming loads and tool wear, thereby improving the overall robustness of the process. An analytical model is proposed to predict the resulting thickness distribution and demonstrates accuracy within a 12.5% deviation from experimental results. Finally, a design map is proposed to instantly identify the optimal beak design parameters without the need for extensive numerical or physical validations while ensuring a minimal thickness change.

Electric-field assisted (EA) manufacturing is a promising hybrid manufacturing technique, offering significant advantages over conventional manufacturing methods. Extensive experimental and numerical studies have demonstrated that the application of electric current reduces flow stress in metals and alloys, thereby improving their manufacturability. This enhancement is attributed to the synergistic effects of electroplasticity and Joule heating, both induced by the applied current during processing. Several key manufacturing processes have garnered substantial interest from the research community for their potential enhancement through electric fields. Here, we present a comprehensive review of recent developments in EA manufacturing over the last decade. The findings of various researchers investigating different EA manufacturing processes are discussed, accompanied by detailed tables summarizing the materials and electric current parameters employed in each process.

The growing complexity of manufacturing processes and the increasing diversity of decision-making tools present challenges in selecting effective approaches for process optimisation. Many existing tools are either too narrowly focused or inconsistently applied across sectors, limiting their broader impact. Additionally, the lack of clear integration strategies often hinders their full implementation in industrial settings. This systematic review examines decision-making tools that enable comparative assessments applied at the unit process level in manufacturing, covering both the selection between competing manufacturing routes and the optimisation of specific processes. A total of 37 journal articles were selected through a structured database search and evaluation process. The review analyses commonly used tools such as Multi-Criteria Decision Analysis (MCDA), Life Cycle Assessment (LCA), and Direct Comparison, highlighting their applications, benefits and limitations. Findings show that MCDA offers robust, multi-dimensional evaluations but is often constrained by complexity and data demands. In contrast, simpler methods like Direct Comparison provide more accessible insights but with a limited scope. Advanced tools such as Deep Learning and Computational Simulations hold promise but face challenges in scaling beyond the process level. Notably, there is limited integration of sustainability metrics within process-level decision-making. To address this, the study proposes a structured framework to guide future research and implementation, focusing on data management, AI integration and tool scalability. The results highlight the need for hybrid approaches that combine different tools to balance trade-offs and support long-term sustainability and operational efficiency in manufacturing systems.