Abstract Press-hardened steel (PHS), characterized by its extremely high strength, has wide applications in vehicle body manufacturing as an innovative lightweight material. Nevertheless, the poor weldability of PHS results in poor weld toughness and a high risk of interfacial fracture (IF), posing challenges to the resistance spot welding (RSW) process. Introducing an external magnetic field in the welding process to perform electromagnetic stirring (EMS), the magnetically assisted RSW (MA-RSW) process has been proven an effective method to improve the weld toughness of high-strength steel, but it may increase the risk of expulsion. In response to these challenges, this study introduces a new process called SPMA-RSW to improve the weldability of PHS by combining MA-RSW and the stepped-current pulses (SP) technique, which can enlarge the weld lobe. Nugget appearance, microstructure, microhardness, and mechanical properties were systematically investigated by comparing traditional RSW, MA-RSW, SP-RSW, and SPMA-RSW. The result showed that the SPMA-RSW process would significantly increase the nugget size, inhibit the shrinkage voids, finer the grain size of PHS welds, and harden the nugget region. This increased the lap-shear strength and changed the fracture mode from brittle IF mode to ductile plug fracture (PF) mode at the same heat input. Specifically, the peak load and energy absorption were increased by 32.3% and 84.2%, respectively. Then, an analytical model was developed to reveal the mechanism of the effect of EMS on the fracture mode transition and was verified by experiment. This work can help improve the weld quality and thermal efficiency of the RSW process for PHS.

Abstract Despite the importance of product repairability, current methods for assessing and grading repairability are limited, which hampers the efforts of designers, remanufacturers, original equipment manufacturers (OEMs), and repair shops. To improve the efficiency of assessing product repairability, this study introduces two artificial intelligence (AI) based approaches. The first approach is a supervised learning framework that utilizes object detection on product teardown images to measure repairability. Transfer learning is employed with machine learning architectures such as ConvNeXt, GoogLeNet, ResNet50, and VGG16 to evaluate repairability scores. The second approach is an unsupervised learning framework that combines feature extraction and cluster learning to identify product design features and group devices with similar designs. It utilizes an oriented FAST and rotated BRIEF feature extractor (ORB) along with k-means clustering to extract features from teardown images and categorize products with similar designs. To demonstrate the application of these assessment approaches, smartphones are used as a case study. The results highlight the potential of artificial intelligence in developing an automated system for assessing and rating product repairability.

Abstract The five-axis ball-end milling dynamics of thin-walled blades is presented. The cutting forces are predicted from the ball-end mill–blade geometry engagement maps along the toolpath. The Frequency Response Function (FRF) of the thin-walled blade is predicted using Finite Element shell elements, and it is updated along the toolpath as the metal is removed. The predicted cutting forces are applied on both the workpiece and tool FRFs to predict the forced vibrations and chatter stability at each tool location. A simplified method to update the cutter-workpiece engagement (CWE) is used to obtain the three-dimensional stability lobe diagram at each desired point on the blade. The integrated model is used to simulate the five-axis machining of thin-walled blades in the digital environment. The proposed digital model is experimentally validated by machining a series of thin-walled rectangular plates and a twisted fan blade.

Abstract Powder bed fusion (PBF) is an additive manufacturing process in which laser heat liquefies blown powder particles on top of a powder bed, and cooling solidifies the melted powder particles. During this process, the laser beam heat interacts with the powder causing thermal emission and affecting the melt pool. This paper aims to predict heat emission in PBF by harnessing the strengths of recurrent neural networks. Long short-term memory (LSTM) networks are developed to learn from sequential data (emission readings), while the learning is guided by process physics including laser power, laser speed, layer number, and scanning patterns. To reduce the computational efforts on model training, the LSTM models are integrated with a new approach for down-sampling the pyrometry raw data and extracting useful statistical features from raw data. The structure and hyperparameters of the LSTM model reflect several iterations of tuning based on the training on the pyrometer readings data. Results reveal useful knowledge on how raw pyrometer data should be processed to work the best with LSTM, how physics features are informative in predicting overheating, and the effectiveness of physics-guided LSTM in emission prediction.

Abstract This paper introduces a new control method for web tension control on a complex roll-to-roll winding machine used in battery production. A traditional web tension control method cannot perform well enough under high winding speed: the parameter tuning process is time-consuming, the disturbance rejection performance is not satisfying, and the control performance is not stable. A hybrid control method is proposed, and it is easy to be implemented on common programming platform for commercial winding machines with an easy tuning process, while providing superior control performance to the traditional control method. The system modeling used in the control method is much simpler than the modeling in most of the tension control researches, providing better feasibility for industrial application.

Abstract Although the production of polymer/carbon nanotube (CNT) nanocomposites has grown exponentially over the last years for a variety of applications, the availability of polymer/CNT filaments for use in commercial 3D printing systems is very limited, and, currently, little is known about the printability of recycled polymer/CNT nanocomposites. In this respect, the fused filament fabrication (FFF) of recycled thermoplastic polyurethane (TPU)/CNT nanocomposites was investigated with a special focus on the piezoresistive behavior. Mechanically recycled and virgin TPU/CNT nanocomposites with different CNT contents (0.5, 1, 3, and 5 wt% by weight) were subjected to filament extrusion and FFF, and the changes induced by mechanical recycling, CNT contents, and infill orientation were monitored by melt flow index, thermal, mechanical, electrical, and piezoresistive properties. It was found that the recycled TPU nanocomposites exhibit very good printability with mechanical and electrical properties that are generally comparable with those for the virgin nanocomposites, the decrease of the elongation at break at 5 wt% CNTs being the primary challenge for the mechanical recycling of TPU/CNT nanocomposites. The 3D-printed recycled TPU/CNT nanocomposites with 3 wt% and 5 wt% CNTs have very good strain sensing behavior with tunable sensitivity by varying the printing conditions. When targeting strains that suit the human motion (0–68%), the recycled TPU/CNT film sensors printed with 0-infill orientation show higher sensitivity (gauge factor up to 115 and 20 for 3 wt% and 5 wt%, respectively) compared with the virgin TPU/CNT film sensors (gauge factor up to 50 and 15 for 3 wt% and 5 wt%, respectively). The findings of this work provide guidance for assessing the potential of using recycled TPU/CNT nanocomposites for 3D printing strain sensors for a wide range of human motions.

Abstract High-strength and corrosion-resistant materials, such as the nickel-based superalloy Inconel 718, are widely used in the energy and aerospace industries. However, machining these materials results in high process forces and significant tool wear. This tool wear negatively influences the resulting surface topography. Nevertheless, the accuracy requirements for functional surfaces are extremely high. Simulation systems can be used to design these processes. However, time-consuming and cost-intensive experiments often have to be conducted to develop and parameterize the required models. To overcome this problem, an analogy test setup for in-process measurements of wear-dependent properties was developed, which allows a multi-level evaluation of the process. By combining different measurement techniques, wear-dependent process characteristics can be determined and analyzed, which is usually only possible for initial and final conditions requiring a significant measurement effort.

Abstract Wire arc additive manufacturing (WAAM) enables 3D printing of large high-value metal components. However, integrating WAAM into production lines requires a critical understanding of the influence of process parameters on the resulting material characteristics. As such, this research investigates the relationship between WAAM wire feed speed (WFS) and torch speed (TS) on the resulting mechanical characteristics of 316LSi thick parts (2.5 cm (0.98 in.)). The experimental procedure is informed by a training matrix that allows parametric analysis of WFS and TS on the ultimate tensile strength (σult), yield strength (σy), elastic modulus (E), failure strain (εf), hardness (HV0.5), and dimensional accuracy (Da) of the printed samples. The research found that WAAM-processed 316LSi parts feature isotropic material properties despite variations in WFS and TS. The surrogate model developed in this study offers five significant polynomial models capable of accurately predicting the influence of WAAM process parameters on σult, σy, εf, E, and Da. The research found TS to be the most significant WAAM process parameter in comparison to WFS for σult and εf. On the contrary, σy, E, and Da were found to be primarily driven by WFS as opposed to TS. Overall, the paper for the first time presents an accurate surrogate model to predict the mechanical characteristics of WAAM 316LSi thick parts informed by wire feed speed and torch speed. The study demonstrates that the mechanical properties of WAAM-processed steel are primarily influenced by the underlying process parameters offering significant potential for tunable performance.

Abstract Laser powder bed fusion (PBF/L) is a popular metal additive manufacturing (AM) process used to manufacture complex metallic 3D components. Maraging steel is one of the metals used in AM and it belongs to the class of ultra-high-strength steels used in aerospace and tooling industries. In the PBF/L process, a laser beam is used to melt and fuse the metal powder particles. This creates a high thermal gradient and rapid cooling of the melt pool results in columnar grains. The microstructure of AM part is entirely different from the conventionally manufactured case and this necessitates post-AM heat treatments. The current paper reviews the effects of printing parameters and heat treatment on microstructure and mechanical properties of PBF/L produced maraging steel 300 alloy. Tensile, impact, fracture, and fatigue properties of as-built and heat-treated PBF/L parts are discussed in detail.

Abstract The increasing usage of cobot applications reshapes work environments and working conditions, requiring specific advancements in organizational practices for health and safety. Enterprises should shift from a technocentric risk management approach to considering cobot application as socio-technical systems, for which a resilience engineering approach is beneficial. This study presents an instantiation of the resilience analysis grid in cobot applications with the aim of measuring resilience potentials in terms of the four cornerstones of resilience engineering (respond, learn, monitor, and anticipate). The assessment has been provided via a questionnaire to 15 companies making use of cobot applications. Results revealed that companies mainly focus on the risk assessment of cobot applications with a traditional view of machine-centric safety, paying less attention to assessing contexts and process variables. This observation seems to arise mainly due to the lack of formally available safety methods or limited guidance from technical standards. Additionally, traditional industrial approaches to risk management lack monitoring of several risks that are essential for managing resilience, defined as the adaptive capacity of people, organizations, and human–machine systems. In addition, companies strongly rely on data from the cobot manufacturer for their safety assessment. The resilience analysis grid was confirmed as a valuable assessment tool for the participating companies to identify improvement areas and assess health and safety from a resilience engineering perspective.