Herein, the quality and accuracy to manufacture delicate parts from NiTi powder using electron beam powder bed fusion (EB-PBF) technology is investigated. Therefore, benchmarks with thin cylinders and thin walls are designed and fabricated using two distinct scan strategies of EB-PBF manufacturing (i.e., continuous melting and spot melting) with different process parameter sets. After these optimizations, four different lattice structures (i.e., octahedron, cell gyroid, sheet gyroid, and channel) are manufactured and characterized. It is shown both continuous melting and spot melting modes are able to manufacture lattices with relative densities over 97%. And as-built lattice structures exhibit an excellent pseudoelasticity up to 8% depending on the design of the structure, e.g., the channel structure shows more deformation recoverability than the cell gyroid. This is attributed to the integrity of geometry as well as compressive mode of the mechanical loading. Of course, the compressive strength and ultimate compressive strength also increase with the increasing volume fraction. Moreover, the spot melting can be used as an engineering tool to customize a delicate beam-shaped structure with a superior pseudoelasticity.

Additive manufacturing (AM) using powder bed fusion is becoming a mature technology that offers great possibilities and design freedom for manufacturing of near net shape components. However, for many gas turbine and aerospace applications, machining is still required, which motivates further research on the machinability and work piece integrity of additive-manufactured superalloys. In this work, turning tests have been performed on components made with both Powder Bed Fusion for Laser Beam (PBF-LB) and Electron Beam (PBF-EB) in as-built and heat-treated conditions. The two AM processes and the respective heat-treatments have generated different microstructural features that have a great impact on both the tool wear and the work piece surface integrity. The results show that the PBF-EB components have relatively lower geometrical accuracy, a rough surface topography, a coarse microstructure with hard precipitates and low residual stresses after printing. Turning of the PBF-EB material results in high cutting tool wear, which induces moderate tensile surface stresses that are balanced by deep compressive stresses and a superficial deformed surface that is greater for the heat-treated material. In comparison, the PBF-LB components have a higher geometrical accuracy, a relatively smooth topography and a fine microstructure, but with high tensile stresses after printing. Machining of PBF-LB material resulted in higher tool wear for the heat-treated material, increase of 49%, and significantly higher tensile surface stresses followed by shallower compressive stresses below the surface compared to the PBF-EB materials, but with no superficially deformed surface. It is further observed an 87% higher tool wear for PBF-EB in as-built condition and 43% in the heat-treated condition compared to the PBF-LB material. These results show that the selection of cutting tools and cutting settings are critical, which requires the development of suitable machining parameters that are designed for the microstructure of the material.

Surface roughness plays an indispensable and fundamental role as a leading indicator of the surface quality of machined parts in the manufacturing process. The precise and effective monitoring and prediction of surface roughness is crucial for surface quality control. In this regard, the development of an in-process surface quality monitoring system is necessary, which has the promising potential to achieve this goal. Such a system typically comprises data-driven models for decision-making and sensing techniques for detecting associated process information. However, some challenges still exist in building such systems. Firstly, the architecture design and deployment of data-driven models, specifically deep learning (DL)-based models, demand adequate domain knowledge. Secondly, most models trained on specific tasks with limited datasets are prone to suppressing their versatility and generalization across different machining conditions. Additionally, in most cases, reliance on handcrafted features to represent dynamic information on various signals during model training necessitates extensive expertise in selecting appropriate feature types. Furthermore, due to the nature of their low dimensionality, handcrafted features have difficulty in capturing of overall process-related underlying patterns from dynamics signatures, which is time-varying and often occurs in transient events. To address these challenges, this paper proposes the regression-based pre-trained convolutional neural network (pre-trained CNN) combined with Mel-spectrogram images based on the transfer learning method for surface roughness prediction. Within the context, the architecture of the transfer model is slightly adapted from already well-trained CNNs. Initial weights in each layer of the CNN model are directly inherited and then fine-tuned through the Bayesian optimization tuning method. Besides, the audible sound signals are captured and subsequently converted into 2D Mel-spectrogram images with variant time lengths, which are separately engaged to retrain and validate four existing pre-trained CNN models (VGG16, VGG19, ResNet50V2 and InceptionResNetV2). Eventually, the effectiveness of proposed models and comparison of their predictive capabilities are further validated through a case study in the turning process. The results demonstrate that each applied pre-trained CNN model is capable of effectively predicting surface quality with satisfactory prediction results. Therefore, the proposed method can facilitate the establishment of a machining monitoring system concerning its accuracy, reliability, and robustness.

In the turning process, the surface roughness of the machined part is considered a critical indicator of quality control. Provided the conventional offline quality measurement and control is time-consuming, with slow feedback and an intensive workforce, this paper presents an online monitoring and prediction system for the effective and precise prediction of surface roughness of the machined parts during the machining process. In this system, the audible sound signal captured through the microphone is employed to extract the features related to surface roughness prediction. However, owing to the nonlinear phenomena and complex mechanism causing surface quality in the whole process, the selection of statistical features of the sound signal in both the time and frequency domains varies from one case to another. This variation may lead to false prediction results as sufficient domain knowledge is required. Therefore, the versatile and knowledge-independent features extraction method is proposed, which exploits deep transfer learning to automatically extract sound signal features in the time-frequency domain through pre-trained convolution neural networks (pre-trained CNN). The performance of prediction models based on two feature extraction methods – statistical feature extraction and automatic feature extraction was further tested and validated in the case study. The results demonstrate that the performances of the prediction model built on the automatically extracted features outperformed that developed with the statistical feature method concerning the accuracy and generalization of the prediction model. In addition, this study also provides solid theoretical and experimental support for developing a more precise and robust online surface quality monitoring system.

This chapter presents a guiding framework for circular economy implementation in supply chains. Closing the loop for resource efficiency is a well-known practice in the industry. To concretize the circular economy implementation strategies, closed-loop thinking requires innovation and adaptation. Circular supply chains (CSCs) are one of the key enablers in closing the loop by design or intention for value recovery and profit maximization. CSC is an emerging area, and the view of CSC where forward and reverse supply chain is seamlessly integrated with the overall aim to achieve system-wide circularity is missing in the academic debate. By offering a cross-functional and systemic perspective of circular supply chains, we present a guiding framework to structure and understand the underlying complexities and highlight the crucial elements of circular supply chain implementation. The framework categorizes the circular supply chain into four building blocks: systemic approach, main drivers, levels of decision making, and mechanisms to manage the full loop closure and minimize the inherent uncertainties of a complex system. We conclude the chapter by illustrating the applicability of the circular supply chain framework using two industrial cases that are transitioning toward the circular economy.

The transition to circular manufacturing systems (CMS) is crucial for achieving sustainable growth, addressing the environmental concerns and resource scarcity challenges. Shifting towards CMS requires a systemic approach that integrates value proposition models, product design, and supply chains (SCs). Circular supply chains (CSCs) emerge as a core pillar of CMS, incorporating value delivery, use, recovery, and reuse. CSCs are inherently more complex and dynamic than linear SCs requiring a holistic analysis approach to capture their complex and dynamic attributes. This research proposes an integrated analysis framework combining qualitative and quantitative approaches to explore the complexities and dynamics of CSCs and assess their economic, environmental, and technical performance. Through the lens of two different CMS implementation case studies, one in automotive parts remanufacturing and one in white goods manufacturing, this research illustrates the framework's applicability. In the automotive case, centralizing core management activities was found to improve economic performance by 50-54 %. However, the introduction of regional logistics hubs, while economically efficient, led to a 20 % increase in CO2-equivalent emissions. On the other hand, the white goods case study highlighted the trade-offs in centralizing end-of-life recovery facilities, where financial savings of up to 60 % were offset by increased transportation costs and increased CO2 emissions. The analysis of CSCs in these two distinct manufacturing sectors underscores the relevance and flexibility of the proposed framework, providing decision-makers with a tool to examine how different CSCs configurations and strategies impact overall performance. This guidance is crucial for developing optimal CSCs design and implementation strategies.

Circular business models are gaining traction in academia and industry as an instrument to deliver environmental benefits while being economically profitable. Despite this increased interest, studies that quantitatively assess the impact of circular business models are limited. This study provides researchers and practitioners with a quantitative analysis tool to assess the dynamics of circular business models from both an economic and environmental perspective. Using the case study of white goods-as-a-service, this study employs multi-method simulation modelling in combination with statistical design and analysis of simulation experiments to investigate the effect of different factors including payment schemes (i.e., fixed fee, pay-per-use, and hybrid) and subscription contract duration (i.e., long-term, mid-term, and short-term) on the economic and environmental performance of access-based models. In addition, as the adoption of access-based models is economically challenging for manufacturers due to a discrepancy between costs and revenue streams, this study analyses the effect of different levers to improve the liquidity performance of access-based business models including deposit schemes, cancellation and collection fees, as well as partnering with financial institutions to cover the initial revenue gap. 

This paper develops a hybrid experimental/simulation method for the first time to assess the thermal stresses generated during electron beam melting (EBM) at high temperatures. The bending and rupture of trusses supporting Inconel 625 alloy panels at similar to 1050 degrees C are experimentally measured for various scanning strategies. The generated thermal stresses and strains are thereafter simulated using the Finite-Element Method (FEM). It is shown that the thermal stresses on the trusses may reach the material UTS without causing failure. Failure is only reached after the part experiences a certain magnitude of plastic strain (similar to 0.33 +/- 0.01 here). As the most influential factor, the plastic strain increases with the scanning length. In addition, it is shown that continuous scanning is necessary since the interrupted chessboard strategy induces cracking at the overlapping regions. Therefore, the associated thermal deformation is to be minimized using a proper layer rotation according to the part length. Although this is similar to the literature reported for selective laser melting (SLM), the effect of scanning pattern is found to differ, as no significant difference in thermal stresses/strains is observed between bidirectional and unidirectional patterns from EBM.

The electron beam powder bed fusion (PBF-EB) is limitedly used to manufacture complex structures such as delicate lattices. Nickel titanium (NiTi) has been chosen for fabricating the lattice structure due to its widely utilization in the biomedical sector. However, issues may arise when manufacturing angled trusses while the dimensional inaccuracy increased with the increasing of the angle between the truss member and the vertical build direction. Therefore, two different scan strategies: spot melting and linear melting were used to manufacture the lattice structures respectively to compare the dimensional accuracy of different structures. This investigation highlights that linear melting is prone to maintain the geometrical accuracy of line-based structure with a limited influence from the scan speed while the spot melting is more capable of manufacturing the point-based structure with a higher geometrical resolution.  

Electron beam powder bed fusion (PBF-EB) is used to manufacture dense nickel titanium parts using various parameter sets, including the beam current, scan speed and post cooling condition. The density of manufactured NiTi parts are investigated with relation to the linear energy input. The results implies the part density increases with increasing linear energy density to over 98% of the bulk density. With a constant energy input, a combination of low power and low scan speed leads to denser parts. This is attributed to lower electrostatic repulsive forces from lower number density of the impacting electrons. After manufacturing, densest parts with distinct parameter sets are categorized into three groups: i) high power with high scan speed and vacuum slow cooling, ii) low power with low scan speed and vacuum slow cooling and iii) low power with low scan speed and medium cooling rate in helium gas. Among these, a faster cooling rate suppresses phase transformation temperatures, while vacuum cooling combinations do not affect the phase transformation temperatures significantly. All the printed parts in this study exhibit almost 8% pseudoelasticity regardless of the process parameters, while the parts cooled in helium have a higher energy dissipation efficiency ( ), which implies faster damping of oscillations.