PurposeOne of the paramount challenges in the field of structural health monitoring (SHM) is the precise assessment of structural damage. Recently, many methods have been proposed to diagnose damage in structures using time series data. Nevertheless, these techniques are restricted in their ability to capture the bidirectional temporal and spatial relationships among the data. To remedy these shortcomings, this study proposes an advanced deep learning (DL) technique that integrates one-dimensional convolutional neural networks (1DCNN) with bidirectional long short-term memory (BiLSTM) networks, termed BiLSTM-1DCNN.Design/methodology/approachThis study proposes an advanced deep learning (DL) technique that integrates one-dimensional convolutional neural networks (1DCNN) with bidirectional long short-term memory (BiLSTM) networks, termed BiLSTM-1DCNN. The novelty of BiLSTM-1DCNN comes in its effective fusion of 1DCNN’s local feature extraction and BiLSTM’s capacity to capture long-term information from both directions (forward and backward).FindingsValidation using a dataset from the Nam O steel truss bridge demonstrates that the proposed method surpasses conventional models, including 1DCNN (87.9%), LSTM (72.5%), BiLSTM (76.1%) and LSTM-1DCNN (91.1%), achieving an accuracy rate of 96.4%. This approach provides a significant advancement in SHM, paving the way for more accurate and reliable damage detection.Research limitations/implicationsThe study’s limitations include potential overfitting due to the model’s complexity and the high computational resources required. The BiLSTM-1DCNN model’s performance is also influenced by the quality and quantity of the training data as well as the specific characteristics of the dataset. Additionally, while the model performs well in detecting damage in the Nam O steel truss bridge, its generalizability to other types of structures and environmental conditions remains untested.Practical implicationsThe BiLSTM-1DCNN model offers significant practical benefits for structural health monitoring by enhancing damage detection accuracy. Its ability to effectively capture both local features and long-term temporal dependencies makes it a valuable tool for real-time monitoring and early detection of structural issues. This can lead to more timely maintenance actions, potentially reducing repair costs and improving safety. The model’s high accuracy in detecting damage can also support infrastructure management decisions, contributing to better resource allocation and enhanced overall safety of critical structures.Social implicationsThe BiLSTM-1DCNN model’s advanced damage detection capabilities have positive social implications, such as improving the safety and reliability of public infrastructure like bridges. Enhanced structural health monitoring can reduce the risk of catastrophic failures, protecting lives and reducing accidents. By providing more accurate and timely assessments, the model can also lead to more efficient use of maintenance and repairs, ensuring that resources are allocated effectively.Originality/valueThe originality of the BiLSTM-1DCNN model lies in its innovative integration of one-dimensional convolutional neural networks (1DCNN) with bidirectional long short-term memory (BiLSTM) networks. This unique combination leverages 1DCNN’s strength in local feature extraction and BiLSTM’s capacity to capture temporal dependencies in both directions, addressing the limitations of existing methods. The model’s high accuracy in detecting structural damage, as demonstrated by its performance on the Nam O steel truss bridge dataset, highlights its value in advancing structural health monitoring technologies and offers a significant improvement over traditional approaches.

Purpose Accurate and premature prediction of corrosion onset in reinforced concrete structures is crucial for designing sustainable and resilient structures, which will have a lesser carbon footprints, reduced diminution of natural resources and effective utilization of waste materials. Design/methodology/approach This study envisages the corrosion resistance and corrosion initiation across seven distinct phases at 15, 30, 45, 60, 75, 90 and 105 days in concrete blended with Portland Slag Cement (PSC) and microsilica. Concrete cube specimens with variable microsilica content (0, 5, 10, 15 and 20%), each containing a centrally sited rebar, were subjected to accelerated corrosion under severe chloride circumstances (5% NaCl) for 105 days, following ASTM C876-91 stipulation. Findings Experimental results and scanning electron microscope investigation revealed that samples with 10% microsilica exhibited the superior corrosion resistance, attributed to the enhanced development of calcium silicate hydrate gel. Additionally, the experimental findings have also been validated using an artificial neural network (ANN). The ANN model predictions closely align with experimental predictions, by achieving higher correlation coefficients (“R” values of 0.99935, 0.99835, 0.99906 and 0.99917 for training, testing, validation and the overall model, respectively). Experimental findings and model predictions suggest that PSC partially replaced with 10% microsilica holds noteworthy potential for developing concrete mixes suited for extreme environments. Originality/value Accurate and timely prediction of corrosion initiation minimizes potential hazards, ensuring structural longevity and reliability. Furthermore, the partial replacement of cement with microsilica enhances corrosion resistance, reducing cement consumption and promoting the effective utilization of waste materials for sustainable construction.

PurposeIn this article, the applications of the improved residual power series method (IRPSM) and its enhanced version incorporating Adomian polynomials (AP) to solve boundary value problems (BVPs) of higher orders are presented.Design/methodology/approachTo show the efficacy and computational efficiency of both approaches, five case studies including those from the fourth, fifth, sixth, tenth and thirteenth orders, are provided.FindingsThe findings obtained indicate that the IRPSM with AP is a more efficient option than the normal IRPSM since it greatly decreases CPU time while retaining greater accuracy. The introduction of Adomian polynomials in a novel approach applied for the first time in this case study proves to be a valuable innovation for solving complex nonlinear BVPs. This work highlights the potential of IRPSM with AP in numerical analysis, offering a faster and more reliable solution method for a wide range of problems.Research limitations/implicationsTo the best of our knowledge, IRPSM is applicable only to finite boundary value problems (BVPs). The same limitation applies to IRPSM with Adomian polynomials, as Adomian polynomials primarily serve to enhance convergence and reduce computational costs rather than extending the method’s applicability to infinite or unbounded domains.Originality/valueThe novelty of this article lies in its innovative approach to solving higher-order BVPs efficiently. By incorporating the Adomian polynomials (AP), this method enhances the IRPSM, significantly reducing the computational time while maintaining higher accuracy. This new integration is a significant addition to the area of numerical analysis as it offers a more dependable and efficient way to solve complicated nonlinear BVPs. For the first time, Adomian polynomials are used in this way, presenting a potentially useful tool for solving higher-order and nonlinear problems that are difficult for conventional approaches to solve effectively. Research objectives: The objectives of this research are to: Investigate the effectiveness of IRPSM when combined with AP for solving higher-order BVPs? Compare and evaluate the computational efficiency and accuracy of IRPSM with AP against traditional IRPSM. Apply IRPSM with AP to test problems of 4th, 5th, 6th, 10th and 13th orders and demonstrate its ability to reduce CPU time while preserving accuracy.Key features of the proposed approach(1)The IRPSM with Adomian polynomials enhances the accuracy of solutions for higher-order BVPs.(2)IRPSM with Adomian polynomials is a more efficient approach compared to IRPSM.(3)The IRPSM with Adomian polynomials improves convergence speed and stability, ensuring faster and more reliable solutions.

PurposeThis study introduces a novel hybrid algorithm, equilibrium optimizer-singular value decomposition (EO-SVD), for structural damage detection in various bridge models. By combining the optimization capabilities of EO with the dimensionality reduction of SVD, the proposed approach achieves superior accuracy and computational efficiency compared to traditional algorithms.Design/methodology/approachBy combining the optimization capabilities of EO with the dimensionality reduction of SVD, the proposed approach achieves superior accuracy and computational efficiency than traditional algorithms. Comprehensive evaluations were performed on multiple bridge models, including a reinforced concrete bridge, a steel truss bridge and a suspension bridge, under scenarios with noise.FindingsThe proposed algorithm achieved the highest accuracy in the three mentioned structures, with values of 2.98E−05, 2.97E−05 and 4.32E−04, demonstrating high accuracy as the highest standard deviation was only 7.97E−04. EO-SVD consistently delivered the lowest errors and fastest convergence. Utilizing finite element models and dynamic features, the algorithm effectively pinpointed damage locations and severities, even under complex scenarios.Research limitations/implicationsAlthough its performance may slightly decrease with highly intricate structures, EO-SVD remains highly robust and reliable, offering a solid foundation for the future development of structural health monitoring systems.Originality/valueUtilizing finite element models and dynamic features, the algorithm effectively pinpointed damage locations and severities, even under complex scenarios.

PurposeWe provide a dataset for advancing Deep Generative Models (DGMs) in engineering design and propose methods to automate data labeling by utilizing large-scale foundation models.Design/methodology/approachGeoBiked is curated to contain 4,355 bicycle images, annotated with structural and technical features and is used to investigate two automated labeling techniques: The utilization of consolidated latent features (Hyperfeatures) from image-generation models to detect geometric correspondences (e.g. the position of the wheel center) in structural images. For the generation of diverse text descriptions for structural images, GPT-4o, a vision-language model (VLM), is instructed to analyze images and produce diverse descriptions aligned with the system-prompt.FindingsBy representing technical images as Diffusion-Hyperfeatures, drawing geometric correspondences between them is possible. The detection accuracy of geometric points in unseen samples is improved by presenting multiple annotated source images. GPT-4o has sufficient capabilities to generate accurate descriptions of technical images. Grounding the generation only on images leads to diverse descriptions but causes hallucinations, while grounding it on categorical labels restricts the diversity. Using both as input balances creativity and accuracy.Research limitations/implicationsSuccessfully using hyperfeatures for geometric correspondence suggests that this approach can be used for general point-detection and annotation tasks in technical images. Labeling such images with text descriptions using VLMs is possible, but dependent on the models detection capabilities, careful prompt-engineering and the selection of input information.Originality/valueApplying foundation models in engineering design is largely unexplored. We aim to bridge this gap with a dataset to explore training, finetuning and conditioning DGMs in this field and suggesting approaches to bootstrap foundation models to process technical images.

PurposeThis paper aims to develop a probabilistic framework which combines uncoupled cofferdam stability analysis, random forest and Monte Carlo simulation for cofferdam reliability analysis.Design/methodology/approachThe finite element method and limit equilibrium method are used to calculate the seepage field and stability of cofferdam, respectively. Sufficient training and validating random samples are generated to obtain a random forest surrogate model with acceptable accuracy. The calibrated random forest model combined with MCS is used to conduct cofferdam reliability analysis. The proposed methodology is illustrated using a typical cofferdam model.FindingsThe numerical simulation results demonstrate that a larger pore water pressure leads to a lower stability of the cofferdam and vice versa. The increase in the slope angle significantly reduces the stability of cofferdam on the corresponding side, while the stability of cofferdam on the other side is mainly affected by the internal pore water pressure. The increase in the width and height of the reverse pressure platform significantly enhances the stability of cofferdam, and the changes in the angle of the reverse pressure platform affect the stability of cofferdam to some extent. The probability of failure (Pf) of cofferdam increases gradually with increasing vertical and horizontal scales of fluctuation, coefficient of variation, and cross-correlation coefficient when the degradation degree of soil properties is low. It is worth noting that the effect of vertical and horizontal scales of fluctuation, coefficient of variation, and cross-correlation coefficient on the Pf of cofferdam changes significantly when degradation coefficient decreases to a critical value.Practical implicationsA geotechnical engineer could use the proposed method to perform cofferdam reliability analysis.Originality/valueThe reliability of cofferdam can be efficiently and accurately studied using the proposed framework.

PurposeThe paper attempts to establish a failure prediction method for machine tools.Design/methodology/approachWith respect to these problems of failure prediction for machine tools, by analyzing the dimensional accuracy of the machined parts of the main shaft bearing to predict the bearing replacement period, from the perspective of an inverse problem, considering the influence of bearing type and other factors, based on data mining technology and prediction methods, we propose a grey periodic extension model combined with multi-scale graph correlation (MGC) to screen out the dimensional accuracy types closely related to the bearing failure and predict and analyze the change trend of the part accuracy.FindingsThe results show that the proposed model can help predict the change in dimensional accuracy of parts by building a combinatorial model and the bearing replacement time and has stronger universality compared with judging the condition of bearings based on their own data changes.Originality/valueFrom the perspective of inverse problems, based on data changes, the proposed approach can well deal with failure prediction for machine tools.

PurposeThis study implements and evaluates the effect of incorporating a truncated weighing function on a nonlocal plane stress projected algorithm with Tikhonov regularization for plane stress conditions.Design/methodology/approachApplying the continuum model to materials undergoing softening after yielding often results in mesh-dependent solutions. To address the mesh dependency problem, several integral-type nonlocal plasticity models have been proposed. However, the shortcoming of this approach is that only the points near the strain-softening point significantly impact its stresses and strains. Therefore, a truncated Gaussian distribution weighing function was implemented on a recently developed plane stress projected nonlocal plasticity model. The efficacy of this nonlocal formulation was measured by modifying the truncated Gaussian distribution limits as a function of the number of standard deviations and assessing the effect of this truncation on the computational time and the distribution of equivalent plastic strains. The study also analyzed the impact of the length scale and regularization parameters on the finite element solutions when the truncated weighing function is incorporated into the nonlocal plane stress projected formulation.FindingsThe output parameters, such as plastic strains, band width and force-deformation curve, showed negligible differences when the Gaussian distribution was truncated to three standard deviations, whereas the computational time decreased by around 5%. However, truncating the weighing function at two standard deviations or less increased the equivalent plastic strain. The output parameters, such as the equivalent plastic strain, are sensitive to length scale modifications and tend to converge as the regularization parameter increases.Originality/valueFor the first time, the present study uses the truncated Gaussian distribution as a weighing function in a 2D plane stress projected nonlocal plasticity model and studies the effect of truncation of the above function on the numerical accuracy and computational efficiency of the model. The effect of the regularization parameter will also be studied, especially for the length scale factor, since the attenuation of the Gaussian function is controlled by the standard deviation depending on the values of the length scale.

PurposeMulti-unmanned aerial vehicle (UAV) missions aim to optimize the execution of multiple missions using limited resources, making it possible to balance the objectives of each mission while minimizing the time to completion.Design/methodology/approachAn algorithm combining cluster analysis and differential evolution particle swarm optimization (DE-PSO) is proposed to solve this problem.FindingsThe investigative study is based on the homogenization of multi-UAV missions in multi-objective task distribution to reduce the total elapsed time.Practical implicationsThis method effectively reduces task time and provides a solution for multi-UAV operations in transmission line cooperation.Originality/valueA novel heuristic algorithm is proposed, and the algorithm fully considers the clustering characteristics under multi-region and the positional relationship characteristics of scene target distribution. It also fully considers the physical characteristics of airport location and UAV power to uniformly optimize the time.

PurposeAs widely used by the design and decision-making of resource allocation and communications in manufacturing, monitoring of mobile robot network and other engineering fields, minimum initial marking (MIM) estimation is to determine the initial marking(s) with a minimum total number of tokens in labeled Petri net models of real-world application systems in engineering. This paper addresses the MIM estimation under the observation of a label sequence. We aim to find all possible solutions while reducing the computational cost in solving this NP-hard problem by a hybrid evolutionary heuristic.Design/methodology/approachFirst, a set of transition firing sequences (TFSs) consistent with the observed label sequences is found to generate the initial population of the genetic algorithm (GA). Second, we calculate the initial markings that correspond to each TFS and are minimal, and the quality of individuals is assessed. Third, simulated annealing (SA) is used to improve the individuals in the population, and then GA is used for a global search. Finally, the above operations are repeated until the termination condition is satisfied.FindingsIn comparison with the existing methods for finding MIMs with heuristic algorithms, our algorithm improves 51.5–316.7% in the number of obtained MIMs. Moreover, the computation time is reduced by 90.1% compared with a classical algorithm.Originality/valueIn this paper, an effective evolutionary algorithm by hybridizing GA and SA is developed to estimate MIMs. The proposed approach can obtain the complete MIM set at a lower computational cost.