Advances in Industrial and Manufacturing Engineering最新文献

Factories are complex systems, which are characterized by interlinked and overlapping life cycles of the constituent factory elements. Within this context, the heterogeneity of these life cycles results in life cycle complexity and corresponding conflicts and trade-offs that need to be addressed in decision situations during the planning and operation of factory systems. Also with respect to the transformation need towards environmental sustainability, there is a need for methods and tools for life cycle oriented factory planning and operation. This paper systematically reviews existing life cycle concepts of factory systems as well as frameworks, models and methods for the evaluation and engineering of factory life cycles. In order to respond to the above challenges, a general understanding about the factory life cycle, e.g. life cycle stages, related activities and interdependencies, is developed and action areas of life cycle engineering are discussed that could supplement factory planning. Following that, the paper presents an integrated, model-based evaluation and engineering framework of factory life cycles.

Development and modernization in joining and welding techniques of plastic materials have attracted much attention in the field of scientific research and industry. Friction stir welding and friction stir spot welding, nowadays, have gained a significant advantage over other joining techniques due to the ease of operation, with no need for any adhesives or external heaters, and the use of non-consumable tools. In this research, the friction stir spot welding process of similar and dissimilar polymeric materials (polypropylene and high density polyethylene) was investigated experimentally. The effect of tool rotational speed and dwell time on the lap shear strength was studied, while the tool plunging rate, plunging depth, and tool geometry were kept constant during all tests.

The welding of dissimilar polypropylene to high density polyethylene was successfully performed within the range of 1400–3500 rpm rotational speeds; however, it was unsuccessful with 800 rpm.

Moreover, the friction stir spot welding of similar polyethylene and polypropylene was studied, and the optimum welding conditions were 3500 rpm with a dwell time of 40 s, and 800 rpm with a dwell time of 120 s for polypropylene and polyethylene respectively.

Machine tools are increasingly being equipped with edge computing solutions to record internal drive signals with high frequency. A large amount of available data may be used to develop new data-driven approaches to process optimization and quality monitoring. This paper presents a new approach to predict the quality of finished workpieces for three-axis milling processes with end mills. For this purpose, internal machine tool data provided by an edge computing solution was recorded and used to develop a Machine Learning based method for quality prediction. For the preparation of the data, an introduced domain knowledge-based slicing algorithm is applied, which allows the recorded data to be automatically and precisely assigned to the corresponding geometric elements on the workpiece. During data-driven modeling, 9 Machine Learning algorithms are compared to 4 Deep Learning architectures for multivariate time series classification. The results show that ensemble methods like Random Forest and Extra Trees as well as the Deep Learning algorithms InceptionTime and ResNet reach the best performances for the use case of data-based quality prediction.

Battery cells are central components of electric vehicles determining their operational characteristics, such as driving range, power output, and safety. Automotive OEMs undertake the necessary efforts to ensure the integration of safe and high-performance battery cells in their electrified fleets. In addition, an increased sustainable awareness of their customers and governmental policies force them to not only focus on operational goals, but rather on environmental aspects as well. Especially, battery cell manufacturing is associated with various negative environmental impacts (e.g. carbon dioxide emission). Therefore, this study develops a concept facilitating the development of novel continuous processes in battery cell manufacturing by enabling virtual experiments and an automatic optimization of economic and ecologic targets. Virtual experiments are enabled by training data-driven models that transfer the gained knowledge from development to large-scale production. The concept includes an inline-capable controller adjusting set points of process parameters with respect to a cost model quantifying product quality and environmental aspects. The validity of the proposed concept is demonstrated with data acquired from real battery cell production chain covering a continuous mixing process.

Faced with globalization, high competition, and the demands of a market in constant dynamism, companies strive to adopt measures for increasing their productivity, among which Lean Manufacturing stands out. Although this set of strategies allows optimizing the production by reducing waste, the literature review showed that, in several organizations, the implementation of Lean does not reflect positive impacts on productivity. It is frequently related to the superficial nature of the approach: the tools and methods are applied, but the repercussions on the workers are commonly neglected. In response, companies seek to implement Risk Management policies to assess injury risk factors for operators during task execution. This study highlights the importance of integrating Lean Manufacturing and Ergonomics principles into organizations to increase productivity and improve working conditions simultaneously. Therefore, by identifying improvement opportunities using the VSM tool, this work aims to implement an innovative and systematic intervention model, which enables the integrated application of Single-Minute Exchange of Dies (SMED) and ergonomic analysis in a metallurgical factory. To this end, the innovative Ergonomic SMED (ESMED) Model is proposed, comprising six steps, which, in this study, focus on the setup operations of a molding machine and by including Rapid Upper Limb Assessment (RULA), Rapid Entire Body Assessment (REBA), Job Strain Index (JSI), Key Indicator Methods (KIM), and Shoaf's Model methods. Based on the results obtained, it is possible to evidence the usefulness and effectiveness of the proposed model in this scenario, emphasizing the 55% reduction in setup time and the extreme attenuation of the level of Work-Related Musculoskeletal Disorders (WMSDs) risk in workers.

The purpose of this study was to evaluate the effect of build direction on the mechanical properties of AISI 316L stainless steel using the Laser Directed Energy Deposition (L-DED) process, in the as-built and heat-treated conditions. The test specimens were manufactured in vertical and horizontal directions for tensile and impact tests. In addition, analysis test specimens cube-shaped were manufactured for microstructural and microhardness evaluation. The microstructure in the as-built condition showed a combination of cellular, equiaxial dendritic, cellular dendritic and columnar dendritic, while the microstructure in the heat-treated condition showed a homogeneous characteristic, composed by differently sized grains. The microhardness evaluation in the heat-treated condition presented a reduction of 26.52% compared to the as-build condition. The ultimate tensile strength of horizontal test specimens in the as-built condition was 4.86% higher than the heat-treated condition, whereas the ultimate tensile strength of vertical test specimens in the as-built condition was 5.55% higher than the heat-treated condition. The impact resistance of horizontal test specimens in the heat-treated condition was 12.36% higher than the as-built condition, whereas the impact resistance of vertical test specimens in the heat-treated condition was 18.92% higher than the as-built condition. Briefly, the build direction directly affects the mechanical properties of components manufactured through the L-DED process, and it is possible to improve certain mechanical properties, such as ductility and toughness, through heat treatment.

The results of theoretical and practical research on the synthesis of sodium phosphates from its inorganic salts and the use of these phosphates as binders for the manufacture of molds and cores are presented.

In order to find variants of sodium phosphate synthesis, the processes of interaction of orthophosphoric acid H₃PO₄ with sodium salts of different types (Na₂CO₃ carbonate (salt of chemically weak acid), NaCl chloride (salt of chemically strong acid) and tripolyphosphate Na₅P₃O₁₀ (polyphosphoric salt)) were analyzed. The regularities of the formation of sodium phosphates in all three systems and the conversion of these phosphates when heated in the range from 20 to 1000 ^оС have been researched.

For the first time, thermodynamic parameters were established and the process of obtaining sodium phosphate through the chemical interaction of orthophosphoric acid with sodium chloride was implemented in the laboratory.

It has also been shown that the chemical interaction of sodium tripolyphosphate with orthophosphoric acid forms the strongest binder, which is a disodium pyrophosphate Na₂Н₂P₂O₇.

Synthesized sodium phosphates have an optimal set of functional properties for using in foundry technologies. They provide high strength in compositions with refractory quartz filler and have sufficient thermal stability. Experimentally established, foundry cores based on synthesized binders provide high quality cast surfaces and are easily removed from the internal cavities of cast parts.

Climate change and normative requirements are forcing car manufacturers to make their products ever lighter despite the highest safety requirements. For this case study a high-strength EN AW-7075 alloy is used as an example of how hot formed aluminium components for vehicle bodies can be produced on existing steel hot forming lines. Reusing production plants not only shortens the start-up time for mass production of hot formed and high-strength aluminium components, but also makes a significant contribution to the sustainable use of existing plant technologies and thus to improving the carbon footprint of manufacturing companies.

In addition to the presentation of the process control and the necessary conversion measures for the modification of a hot forming process from steel to aluminium, the most significant technological properties and the production costs of hot formed components made from both materials are compared. To be able to show the economic viability of a reallocation of existing and depreciated manufacturing plants, investigations of possible manufacturing parameters are carried out and presented.

The case study shows that by converting the furnace technology to Jet-Heating, significantly higher heating rates can be realised when compared with using radiation furnaces. As the furnace holding time was found to have no influence on the final strength in T4 and T6 state of EN AW-7075 specimens, the overall furnace cycle time is significantly lower than for press-hardened steels. Although the modifications necessary for the forming of aluminium alloys result in changed requirements for existing production plants, the additional costs per component for the plant conversion are marginal in relation to the assumed delivery volumes. It is demonstrated that for components with low material input, the use of 7000 series aluminium alloy components can become economically viable on a larger scale for OEMs.

Additive manufacturing (AM) is a promising digital manufacturing approach that has seen recent rapid growth. Despite the fast-growing nature of the technology, AM has been slowed by the qualification and certification due to various defects observed in printed parts. On the other hand, Convolutional Neural Networks (CNN), as a deep learning method, have received a great deal of attention over the last decade and demonstrated excellent performance in dealing with image data. Deep learning is a subset of machine learning and refers to any Artificial Neural Network with more than two hidden layers. This article provides a comprehensive overview of CNN’s application to several aspects of the AM process since the emergence of this field. This review also highlights current challenges and possible solutions to provide a horizon for future studies.