Search results

In the field of steel defect detection, the existing detection algorithms struggle to achieve a satisfactory balance between detection accuracy, computational cost and inference speed due to the interference from complex background information, the variety of defect types and significant variations in defect morphology. To solve this problem, this paper aims to propose an efficient detector based on multi-scale information extraction (MSI-YOLO), which uses YOLOv8s as the baseline model.

First, the authors introduce an efficient multi-scale convolution with different-sized convolution kernels, which enables the feature extraction network to accommodate significant variations in defect morphology. Furthermore, the authors introduce the channel prior convolutional attention mechanism, which allows the network to focus on defect areas and ignore complex background interference. Considering the lightweight design and accuracy improvement, the authors introduce a more lightweight feature fusion network (Slim-neck) to improve the fusion effect of feature maps.

MSI-YOLO achieves 79.9% mean average precision on the public data set Northeastern University (NEU)-DET, with a model size of only 19.0 MB and an frames per second of 62.5. Compared with other state-of-the-art detectors, MSI-YOLO greatly improves the recognition accuracy and has significant advantages in computational cost and inference speed. Additionally, the strong generalization ability of MSI-YOLO is verified on the collected industrial site steel data set.

This paper proposes an efficient steel defect detector with high accuracy, low computational cost, excellent detection speed and strong generalization ability, which is more valuable for practical applications in resource-limited industrial production.

The purpose of this research is to enhance the Laser Powder Bed Fusion (LPBF) additive manufacturing technique by addressing its susceptibility to defects, specifically lack of fusion. The primary goal is to optimize the LPBF process using a digital twin (DT) approach, integrating physics-based modeling and machine learning to predict the lack of fusion.

This research uses finite element modeling to simulate the physics of LPBF for an AISI 316L stainless steel alloy. Various process parameters are systematically varied to generate a comprehensive data set that captures the relationship between factors such as power and scan speed and the quality of fusion. A novel DT architecture is proposed, combining a classification model (recurrent neural network) with reinforcement learning. This DT model leverages real-time sensor data to predict the lack of fusion and adjusts process parameters through the reinforcement learning system, ensuring the system remains within a controllable zone.

This study's findings reveal that the proposed DT approach successfully predicts and mitigates the lack of fusion in the LPBF process. By using a combination of physics-based modeling and machine learning, the research establishes an efficient framework for optimizing fusion in metal LPBF processes. The DT's ability to adapt and control parameters in real time, guided by machine learning predictions, provides a promising solution to the challenges associated with lack of fusion, potentially overcoming the traditional and costly trial-and-error experimental approach.

Originality lies in the development of a novel DT architecture that integrates physics-based modeling with machine learning techniques, specifically a recurrent neural network and reinforcement learning.

A section of in-service PE gas pipeline in Guocun, Beijing, was found to appear gas leaking at the electrofusion (EF) joint. This study is dedicated to reveal the material cause of EF joint failure to help with a more accurate prediction of service life of PE gas pipe and further normalize the construction of PE gas pipeline.

Defect detection was carried out on the leaking EF joint using ultrasonic phased array. The mechanical degradation and structural aging behavior was studied by tension test, FTIR technology, TG test and DSC test. The organic components in the soil surrounding the PE gas pipe failure area were qualitatively identified.

The results showed that the organic surfactants in the soil environment could accelerate the aging behavior of PE material, leading to a deterioration of mechanical properties and a serious reduction in the ability of the PE pipe and EF joint, especially at the welding defect, to resist external force.

A novel study was conducted to investigate the failure cause of the EF joint of in-service PE gas pipe, incorporating the analysis of environmental factors and structural deterioration.

This paper aims to present an integrated system designed for quality control and inspection in additive manufacturing (AM) technologies.

The study undertakes a comprehensive examination of the process in three distinct stages. First, the quality of the feedstock material is inspected during the preprocessing step. Subsequently, the main research topic of the study is directed toward the 3D printing process itself with real-time monitoring procedures using computer vision methods. Finally, an evaluation of the 3D printed parts is conducted, using measuring methods and mechanical experiments.

The main results of this technical paper are the development and presentation of an integrated solution for quality control and inspection in AM processes.

The proposed solution entails the development of a promising tool for the optimization of the quality in 3D prints based on machine learning algorithms.

The purpose of this study is therefore to detail an additive manufacturing process for printing TiD parts for implant applications. Titanium–diamond (TiD) is a new composite that provides biocompatible three-dimensional multimaterial structures. Thus, the authors report a powder-deposition and print optimization strategy to overcome the dual-functionality gap by printing bulk TiD parts. However, despite favorable customization outcomes, relatively few additive manufacturing (AM) feedstock powders offer the biocompatibility required for medical implant and device technologies.

AM offers a platform to fabricate customized patient-specific parts. Developing feedstock that can be 3D printed into specific 3D structures while providing a favorable interface with the human tissue remains a challenge. Using laser metal deposition, feedstock powder comprising diamond and titanium was co-printed into TiD parts for mechanical testing to determine optimal manufacturing parameters.

TiD parts were fabricated comprising 30% and 50% diamond. The composite powder had a Hausner ratio of 1.13 and 1.21 for 30% and 50% TiD, respectively. The flow analysis (Carney flow) for TiD 30% and 50% was 7.53 and 5.15 g/s. The authors report that the printing-specific conditions significantly affect the integrity of the printed part and thus provide the optimal manufacturing parameters for structural integrity as determined by micro-computed tomography, nanoindentation and biocompatibility of TiD parts. The hardness, ultimate tensile strength and yield strength for TiD are 4–6 GPa (depending on build position), 426 MPa and 375 MPa, respectively. Furthermore, the authors show that increasing diamond composition to 30% results in higher osteoblast viability and lower bacteria count than titanium.

In this study, the authors provide a clear strategy to manufacture TiD parts with high integrity, performance and biocompatibility, expanding the material feedstock library and paving the way to customized diamond implants. Diamond is showing strong potential as a biomedical material; however, upscale is limited by conventional techniques. By optimizing AM as the avenue to make complex shapes, the authors open up the possibility of patient-specific diamond implant solutions.

Extended reality (XR) is an emerging technology, with its popularity rising in different industry sectors, where its application has been recently considered in construction safety. This study aims to investigate the applications of XR technologies in the safety of construction through projects lifecycle perspective.

Scientometric analysis was conducted to discover trends, keywords, contribution of countries and publication outlets in the literature. The content analysis was applied to categorize previous studies into three groups concerning the phase of lifecycle in which they used XR.

Results of the content analysis showed that the application of XR in the construction safety is mostly covered in two areas, namely, safety training and risk management. It was found that virtual reality was the most used XR tool with most of its application dedicated to safety training in the design phase. The amount of research on the application of augmented reality and mixed reality in safety training, and risk management in all phases of lifecycle is still insignificant. Finally, this study proposed three main areas for using the XR technologies regarding the safety issues in future research, namely, control of safety regulations and safety coordination in construction phase, and safety reports in the operation phase.

This paper inspected the utilization of all types of XR for safety in each phase of construction lifecycle and proposed future directions for research by addressing the safety challenges in each phase.

Clause 8 is about time; the time period for the performance of the Works (the beginning and the end), management of the Programme, the Contractor’s right to Extension of Time and the Employer’s right to suspend progress of the Works. Once the Works have been completed, Clause 9 [Tests on Completion] deals with the testing at completion and Clause 10 [Employer’s Taking Over] deals with the mechanics for the Employer’s Taking Over of the Works. Clause 11 [Defects after Taking Over] deals with defects after the Employer’s Taking Over. See also Figure 8.

This paper aims to investigate and explain the dual fracture behaviour of PA12 specimens sintered by selective laser sintering (SLS) as a function of wall thickness and build direction with a powder mixture 30:70. To achieve this objective, research related to chemical, thermal and structural behaviours as a function of the input variables was carried out to describe and explain why ductile-fragile behaviour occurs during fractures under uniaxial tension manufactured via a methodology of material analysis and manufacturing processes.

The factorial design 32 relates the fracture of PA12 tensile specimens to the horizontal, transverse and vertical build directions at 2.0, 2.5 and 3.0 mm thicknesses, respectively. Fractographic images revealed the fracture surfaces and their dual ductile-fragile behaviour related to the specimens’ measured crystalline, thermal, surface and chemical properties.

The study showed that thermal property variables differ depending on the input variables. The wall thickness variable affected this morphology the most, showing the highest percentage of the ductile area, followed by the transverse and vertical directions. It was determined that the failure in the vertical direction is due to crystalline gradients associated with the layer-by-layer construction process. The pore density may be closely related to generating ductile and brittle areas.

In this paper, fracture characterisation is performed based on the mechanical, chemical, structural, thermal and morphological properties of PA12 manufactured by SLS. In addition, a heatmap of porosities in cross-sections is constructed using a machine learning model (k-means) related to dual fracture behaviour. This research revealed significant differences in the fracture type according to the build direction. In addition, thin-section fractography provides a more detailed explanation of the fragile behaviour of the vertical direction associated with crystalline changes due to the direction of the sintering layers.

This study aims to develop a novel thermal modeling strategy to simulate electron beam powder bed fusion at part scale with machine-varying process parameters strategy. Single-bead and part-scale experiments and modeling were studied. Scanning strategies were described by the process controlling functions that enabled modeling.

The finite element analysis thermal model was used along with the powder bed fusion with electron beam experiments. The proposed strategy involves dividing a part into smaller sections and creating meso-scale models for each subsection. These meso-scale models take into consideration the variable process parameters, including power and velocity of the moving heat source, during part building. Subsequently, these models are integrated to perform partscale simulations, enabling more realistic predictions of thermal accumulation and resulting distortions. The model was built and validated with single-bead experiments and bulky parts with different features.

Single-bead experiments demonstrated an average error rate of 6%–24% for melt pool dimension prediction using the proposed meso-scale models with different scanning control functions. Part-scale simulations for three different geometries (cantilever beams with supports, bulk artifact and topology-optimized transfer arm) showed good agreement between modeled temperature changes and experimental deformation values.

This study presents a novel approach for electron beam powder bed fusion modeling that leverages meso-scale models to capture the influence of variable process parameters on part quality. This strategy offers improved accuracy for predicting part geometry and identifying potential defects, leading to a more efficient additive manufacturing process.

This paper aims to investigate the theoretical and practical links between digital twin (DT) application in heritage facilities management (HFM) from a life cycle management perspective and to signpost the future development directions of DT in HFM.

This state-of-the-art review was conducted using a systematic literature review method. Inclusive and exclusive criteria were identified and used to retrieve relevant literature from renowned literature databases. Shortlisted publications were analysed using the VOSviewer software and then critically reviewed to reveal the status quo of research in the subject area.

The review results show that DT has been mainly adopted to support decision-making on conservation approach and method selection, performance monitoring and prediction, maintenance strategies design and development, and energy evaluation and management. Although many researchers attempted to develop DT models for part of a heritage building at component or system level and test the models using real-life cases, their works were constrained by availability of empirical data. Furthermore, data capture approaches, data acquisition methods and modelling with multi-source data are found to be the existing challenges of DT application in HFM.

In a broader sense, this study contributes to the field of engineering, construction and architectural management by providing an overview of how DT has been applied to support management activities throughout the building life cycle. For the HFM practice, a DT-cum-heritage building information modelling (HBIM) framework was developed to illustrate how DT can be integrated with HBIM to facilitate future DT application in HFM. The overall implication of this study is that it reveals the potential of heritage DT in facilitating HFM in the urban development context.