Login
Register
Submit
Experts Review
Office Work
Author Center
Supervisor: China Iron and Steel Association
Sponsor: Central Research Institute of Building and Construction Co., Ltd., MCC Group, China;China Steel Construction Society，China
Editor-in-Chief: Qingrui Yue
Download:
Call for PapersMore >
Steel Construction Published the List of Highly Influential Articles of 2020
Articles in latest articles have been peer-reviewed and accepted, which are not yet assigned to volumes /issues, but are citable by Digital Object Identifier (DOI).
Display Method:
Display Method:
2026, 41(6): 1-9.
doi: 10.13206/j.gjgS25092901
Abstract
PDF (1971KB)
Abstract:
Sufficiently rich and diverse sample data are essential prerequisites for intelligent structural damage diagnosis. However， the scarcity of damage-state data in practical engineering applications limits the implementation of such methods. Unsupervised deep learning algorithms can partially address this challenge， and an autoencoder-based framework for structural damage early warning combining convolutional neural networks （CNNs） and autoencoders （AEs） was therefore proposed. The proposed framework comprises three modules： data preprocessing， feature extraction， and damage early warning. The feature extraction module uses a CNN to derive discriminative features from healthy-state samples， which serve as input to the damage early warning module. This module adopts an autoencoder model trained exclusively on healthy-state data. Structural damage is detected by monitoring the reconstruction error of input data. Damage is identified when this error exceeds a predefined threshold， which is determined through statistical analysis of the reconstruction error distribution from the healthy-state dataset. The feasibility and reliability of the method were validated using the Qatar Stadium benchmark dataset and controlled indoor damage tests. Experimental results demonstrated that the framework achieved a recognition accuracy exceeding 98% at a 20 dB signal-to-noise ratio， a damage recognition accuracy exceeding 93% even at a 5 dB signal-to-noise ratio， demonstrating its excellent noise immunity.
Sufficiently rich and diverse sample data are essential prerequisites for intelligent structural damage diagnosis. However， the scarcity of damage-state data in practical engineering applications limits the implementation of such methods. Unsupervised deep learning algorithms can partially address this challenge， and an autoencoder-based framework for structural damage early warning combining convolutional neural networks （CNNs） and autoencoders （AEs） was therefore proposed. The proposed framework comprises three modules： data preprocessing， feature extraction， and damage early warning. The feature extraction module uses a CNN to derive discriminative features from healthy-state samples， which serve as input to the damage early warning module. This module adopts an autoencoder model trained exclusively on healthy-state data. Structural damage is detected by monitoring the reconstruction error of input data. Damage is identified when this error exceeds a predefined threshold， which is determined through statistical analysis of the reconstruction error distribution from the healthy-state dataset. The feasibility and reliability of the method were validated using the Qatar Stadium benchmark dataset and controlled indoor damage tests. Experimental results demonstrated that the framework achieved a recognition accuracy exceeding 98% at a 20 dB signal-to-noise ratio， a damage recognition accuracy exceeding 93% even at a 5 dB signal-to-noise ratio， demonstrating its excellent noise immunity.
2026, 41(6): 10-19.
doi: 10.13206/j.gjgS26042702
Abstract:
To address the difficulty in rapidly and accurately evaluating the peak displacement of concrete-filled steel tube （CFST） members under lateral impact， as well as the insufficient interpretation of the internal decision-making mechanism of existing data-driven methods， a peak-displacement database of CFST members was established based on published domestic and international lateral impact test data， and a machine learning model that balances predictive accuracy and interpretability was developed. The database consists of 193 valid test samples， covering geometric parameters， material properties， and impact conditions. Eight parameters， namely hammer mass， impact velocity， section diameter， steel tube thickness， axial compression ratio， steel yield strength， concrete compressive strength， and member length—were selected as input features， and the mid-span peak displacement was taken as the output feature. Based on the experimental database， six ensemble learning algorithms were employed to establish peak-displacement prediction models， and the Optuna framework combined with five-fold cross-validation was used for automatic hyperparameter optimization. Meanwhile， To enhance the transparency of the model results， Shapley Additive Explanations （SHAP） were employed to conduct an interpretability analysis of the CatBoost model， quantifying the contribution degree， influence direction， and potential interactions among variables of each input parameter on the predicted peak displacement. The results indicate that among the six ensemble learning models， the CatBoost model exhibits the best overall performance， effectively characterizing the nonlinear variation pattern of peak displacement in CFST members under multi-parameter coupling. Impact velocity contributes the most to the prediction of peak displacement and serves as the dominant parameter. Hammer mass and section diameter rank next in importance， whereas the influence of the axial compression ratio is relatively weak. Overall， impact velocity， hammer mass， and specimen length tend to amplify the peak displacement， while section diameter， steel tube thickness， and steel yield strength generally mitigate its development. The effects of concrete compressive strength and axial compression ratio remain relatively limited within the current data range. Furthermore， a significant interaction exists between steel yield strength and concrete compressive strength.
To address the difficulty in rapidly and accurately evaluating the peak displacement of concrete-filled steel tube （CFST） members under lateral impact， as well as the insufficient interpretation of the internal decision-making mechanism of existing data-driven methods， a peak-displacement database of CFST members was established based on published domestic and international lateral impact test data， and a machine learning model that balances predictive accuracy and interpretability was developed. The database consists of 193 valid test samples， covering geometric parameters， material properties， and impact conditions. Eight parameters， namely hammer mass， impact velocity， section diameter， steel tube thickness， axial compression ratio， steel yield strength， concrete compressive strength， and member length—were selected as input features， and the mid-span peak displacement was taken as the output feature. Based on the experimental database， six ensemble learning algorithms were employed to establish peak-displacement prediction models， and the Optuna framework combined with five-fold cross-validation was used for automatic hyperparameter optimization. Meanwhile， To enhance the transparency of the model results， Shapley Additive Explanations （SHAP） were employed to conduct an interpretability analysis of the CatBoost model， quantifying the contribution degree， influence direction， and potential interactions among variables of each input parameter on the predicted peak displacement. The results indicate that among the six ensemble learning models， the CatBoost model exhibits the best overall performance， effectively characterizing the nonlinear variation pattern of peak displacement in CFST members under multi-parameter coupling. Impact velocity contributes the most to the prediction of peak displacement and serves as the dominant parameter. Hammer mass and section diameter rank next in importance， whereas the influence of the axial compression ratio is relatively weak. Overall， impact velocity， hammer mass， and specimen length tend to amplify the peak displacement， while section diameter， steel tube thickness， and steel yield strength generally mitigate its development. The effects of concrete compressive strength and axial compression ratio remain relatively limited within the current data range. Furthermore， a significant interaction exists between steel yield strength and concrete compressive strength.
2026, 41(6): 20-29.
doi: 10.13206/j.gjgS26041001
Abstract:
As critical special equipment， elevators heavily rely on periodic inspection reports for fault diagnosis and risk assessment. Existing approaches primarily focus on sensor and time-series data while overlooking the rich knowledge embedded in textual inspection reports. To address this limitation， this paper proposes a large language model （LLM）-based method for constructing an elevator inspection and diagnostic knowledge graph. Based on elevator inspection technical specifications， an ontology for elevator inspection and diagnosis is first developed. On this basis， contextual prompt learning is employed to automatically extract entities and relationships from inspection reports. This method addresses challenges such as limited domain-specific semantic representation and reasoning capability， and mitigates the limitations of general-purpose LLMs. Subsequently， entity alignment is performed through predefined rules， and a high-quality building elevator diagnostic knowledge graph is constructed. The proposed method requires neither manual annotation nor model training， and can efficiently process multi-source heterogeneous elevator inspection data. Experimental results demonstrate that LLM-based prompt learning under domain ontology constraints significantly improves knowledge graph coverage and accuracy. Compared with zero-shot learning， few-shot learning improves entity recognition precision， recall， and F1-score by 6.1 percentage point， 27.20 percentage point， and 17.61 percentage point， respectively， thereby providing effective data support for downstream intelligent fault diagnosis.
As critical special equipment， elevators heavily rely on periodic inspection reports for fault diagnosis and risk assessment. Existing approaches primarily focus on sensor and time-series data while overlooking the rich knowledge embedded in textual inspection reports. To address this limitation， this paper proposes a large language model （LLM）-based method for constructing an elevator inspection and diagnostic knowledge graph. Based on elevator inspection technical specifications， an ontology for elevator inspection and diagnosis is first developed. On this basis， contextual prompt learning is employed to automatically extract entities and relationships from inspection reports. This method addresses challenges such as limited domain-specific semantic representation and reasoning capability， and mitigates the limitations of general-purpose LLMs. Subsequently， entity alignment is performed through predefined rules， and a high-quality building elevator diagnostic knowledge graph is constructed. The proposed method requires neither manual annotation nor model training， and can efficiently process multi-source heterogeneous elevator inspection data. Experimental results demonstrate that LLM-based prompt learning under domain ontology constraints significantly improves knowledge graph coverage and accuracy. Compared with zero-shot learning， few-shot learning improves entity recognition precision， recall， and F1-score by 6.1 percentage point， 27.20 percentage point， and 17.61 percentage point， respectively， thereby providing effective data support for downstream intelligent fault diagnosis.
2026, 41(6): 30-35.
doi: 10.13206/j.gjgS26051701
Abstract:
Influenced by multiple factors such as construction technology， material shrinkage and creep， and service environment， concrete-filled steel tubes are prone to void defects， which lead to a decrease in bearing capacity and pose a threat to long-term service safety. Traditional manual percussion detection is convenient and fast. However， this method relies on the subjective experience of inspectors and is prone to misjudgment and missed detection. Therefore， a method for identifying void defects in concrete-filled steel tubes using percussion sound signals and a one-dimensional convolutional neural network is proposed. A one-dimensional convolutional neural network is used to construct a model for identifying void defects in concrete-filled steel tubes. With the percussion sound pressure signal as input， defect features in the sound signal are automatically extracted through multi-layer convolution operations for defect identification. A dataset covering the influence of factors such as void position and void depth was constructed using experimental and finite-element data， and was used to verify the feasibility and reliability of the proposed method. The results showed that the average identification accuracy of the constructed defect identification model on the test set was 97.44%. Furthermore， the model offered the advantages of short training time and fast convergence.
Influenced by multiple factors such as construction technology， material shrinkage and creep， and service environment， concrete-filled steel tubes are prone to void defects， which lead to a decrease in bearing capacity and pose a threat to long-term service safety. Traditional manual percussion detection is convenient and fast. However， this method relies on the subjective experience of inspectors and is prone to misjudgment and missed detection. Therefore， a method for identifying void defects in concrete-filled steel tubes using percussion sound signals and a one-dimensional convolutional neural network is proposed. A one-dimensional convolutional neural network is used to construct a model for identifying void defects in concrete-filled steel tubes. With the percussion sound pressure signal as input， defect features in the sound signal are automatically extracted through multi-layer convolution operations for defect identification. A dataset covering the influence of factors such as void position and void depth was constructed using experimental and finite-element data， and was used to verify the feasibility and reliability of the proposed method. The results showed that the average identification accuracy of the constructed defect identification model on the test set was 97.44%. Furthermore， the model offered the advantages of short training time and fast convergence.
2026, 41(6): 36-43.
doi: 10.13206/j.gjgS25121701
Abstract:
To enhance the quality and digitalization in the construction of complex bridge towers， a virtual trial assembly （VTA） method for complex bridge towers based on point cloud data was proposed. The method addressed the issue of describing the complex geometric shapes of bridge towers by introducing a feature-plane-based detection section segmentation method and a prior-knowledge-based feature point extraction method， enabling digital dimensional measurement for complex bridge towers. Based on the fundamental information of different engineering projects， VTA technical routes were constructed from both the design model perspective and the geometric feature perspective. Based on the VTA results， finite element simulation of the flame straightening process was conducted to analyze deformation correction for joint misalignments. This provided technical reference and guidance for subsequent repair and adjustment during construction. The results showed that the deviation between the point-cloud-based dimensional inspection and total station measurements was controlled within ±2 mm. The VTA method effectively identified joint misalignments， and by integrating finite element simulation， deformation correction parameters were obtained in advance to provide guidance for corrective actions during construction. The proposed method integrates digital dimensional measurement， VTA， and correction process simulation， significantly enhancing the digital construction capability of complex bridge towers and providing a valuable reference for similar engineering projects.
To enhance the quality and digitalization in the construction of complex bridge towers， a virtual trial assembly （VTA） method for complex bridge towers based on point cloud data was proposed. The method addressed the issue of describing the complex geometric shapes of bridge towers by introducing a feature-plane-based detection section segmentation method and a prior-knowledge-based feature point extraction method， enabling digital dimensional measurement for complex bridge towers. Based on the fundamental information of different engineering projects， VTA technical routes were constructed from both the design model perspective and the geometric feature perspective. Based on the VTA results， finite element simulation of the flame straightening process was conducted to analyze deformation correction for joint misalignments. This provided technical reference and guidance for subsequent repair and adjustment during construction. The results showed that the deviation between the point-cloud-based dimensional inspection and total station measurements was controlled within ±2 mm. The VTA method effectively identified joint misalignments， and by integrating finite element simulation， deformation correction parameters were obtained in advance to provide guidance for corrective actions during construction. The proposed method integrates digital dimensional measurement， VTA， and correction process simulation， significantly enhancing the digital construction capability of complex bridge towers and providing a valuable reference for similar engineering projects.
2026, 41(6): 44-54.
doi: 10.13206/j.gjgS25090103
Abstract:
To reasonably assess the structural reliability of disk-lock scaffold systems， this paper proposes a modeling and calculation method based on multi-level plastic hinges and joint stiffness. Taking a typical high formwork support project of Kunming Changshui T2 Terminal as a case study， this study established a statistical model for the uncertainty of key parameters. The reliability indexes of the scaffold system were quantitatively evaluated via a surrogate model approach， and a method for determining reliability warning thresholds was developed. The results indicated that， compared with corresponding experimental data， the proposed calculation method based on multi-level plastic hinge theory and joint stiffness achieved an error within 5%， which met engineering requirements. Furthermore， based on the existing scaffold response dataset， the response surface-Monte Carlo surrogate modeling method realized rapid reliability assessment for high formwork support systems. In addition， the quantitative determination method of warning thresholds based on exceedance probability supported the real-time adjustment of thresholds， thereby improving the early warning accuracy for high formwork structures.
To reasonably assess the structural reliability of disk-lock scaffold systems， this paper proposes a modeling and calculation method based on multi-level plastic hinges and joint stiffness. Taking a typical high formwork support project of Kunming Changshui T2 Terminal as a case study， this study established a statistical model for the uncertainty of key parameters. The reliability indexes of the scaffold system were quantitatively evaluated via a surrogate model approach， and a method for determining reliability warning thresholds was developed. The results indicated that， compared with corresponding experimental data， the proposed calculation method based on multi-level plastic hinge theory and joint stiffness achieved an error within 5%， which met engineering requirements. Furthermore， based on the existing scaffold response dataset， the response surface-Monte Carlo surrogate modeling method realized rapid reliability assessment for high formwork support systems. In addition， the quantitative determination method of warning thresholds based on exceedance probability supported the real-time adjustment of thresholds， thereby improving the early warning accuracy for high formwork structures.
2026, 41(6): 55-61.
doi: 10.13206/j.gjgS26042301
Abstract:
Based on the Xiamen Bailu West Tower project， the whole process of high-altitude dismantling and modification of the integral steel platform was designed for the core tube of the super high-rise tower with an inward-inclined wall. Finite element simulation and dynamic monitoring were also conducted. In this paper， MIDAS Gen was used to carry out the scheme design and finite element simulation for the modification of the integral steel platform passing through the inclined wall， The stress values and levelness of key load-bearing components were dynamically and intelligently monitored throughout the renovation process. According to the oblique retraction of the tower shear wall， a staged dismantling and reinforcement scheme for the integral steel platform formwork system was designed. Finite element calculations were carried out for five key dismantling and modification construction states， with main component parameter values including stress ratio， displacement， and support reaction. The stress distribution and mechanism of the structure under key construction conditions such as formwork lifting， reinforcement binding， and steel platform lifting were thoroughly analyzed. The results showed that the addition of temporary vertical bracing effectively shared the overall load of the structure and reduced the maximum stress ratio. The safety redundancy of the structure after dismantling and modification was higher than that before dismantling and modification， indicating a safer structure. The monitoring revealed that the internal force distribution of the cylinder frame columns， platform beams， and corbels fluctuated within a range lower than the simulation results， and the levelness remained within the early warning value range during climbing. These measures comprehensively ensured construction safety throughout the entire process of integral steel platform modification. This study can provide a reference for the construction and process monitoring of similar super high-rise inclined walls.
Based on the Xiamen Bailu West Tower project， the whole process of high-altitude dismantling and modification of the integral steel platform was designed for the core tube of the super high-rise tower with an inward-inclined wall. Finite element simulation and dynamic monitoring were also conducted. In this paper， MIDAS Gen was used to carry out the scheme design and finite element simulation for the modification of the integral steel platform passing through the inclined wall， The stress values and levelness of key load-bearing components were dynamically and intelligently monitored throughout the renovation process. According to the oblique retraction of the tower shear wall， a staged dismantling and reinforcement scheme for the integral steel platform formwork system was designed. Finite element calculations were carried out for five key dismantling and modification construction states， with main component parameter values including stress ratio， displacement， and support reaction. The stress distribution and mechanism of the structure under key construction conditions such as formwork lifting， reinforcement binding， and steel platform lifting were thoroughly analyzed. The results showed that the addition of temporary vertical bracing effectively shared the overall load of the structure and reduced the maximum stress ratio. The safety redundancy of the structure after dismantling and modification was higher than that before dismantling and modification， indicating a safer structure. The monitoring revealed that the internal force distribution of the cylinder frame columns， platform beams， and corbels fluctuated within a range lower than the simulation results， and the levelness remained within the early warning value range during climbing. These measures comprehensively ensured construction safety throughout the entire process of integral steel platform modification. This study can provide a reference for the construction and process monitoring of similar super high-rise inclined walls.
2026, 41(6): 62-68.
doi: 10.13206/j.gjgS26053135
Abstract:
In performance-based seismic design， the maximum strain developed in energy-dissipating components must be calculated and checked against specified limits. To investigate the relationship between the maximum strain in steel beams and the story drift angle of steel frames， a stress-strain curve including the strain-hardening stage was adopted， and the shear deformation of steel beams was taken into account. Given the specified maximum strain at the beam-end section， the displacement ductility factors of the beam end were calculated for four types of steel. It was pointed out that the strength ratio computed from the nominal minimum strength specified in design codes led to inaccurate ductility factors of steel beams. After incorporating the deformation of columns， the story ductility factor of the frame was evaluated. Curves relating the story drift angle to the maximum strain at the beam end were presented， and an approximate formula was proposed. Furthermore， for shear-type energy-dissipating coupling beams， the maximum strain corresponding to a story drift angle of 1/50 was given when such beams were treated as flexural members.
In performance-based seismic design， the maximum strain developed in energy-dissipating components must be calculated and checked against specified limits. To investigate the relationship between the maximum strain in steel beams and the story drift angle of steel frames， a stress-strain curve including the strain-hardening stage was adopted， and the shear deformation of steel beams was taken into account. Given the specified maximum strain at the beam-end section， the displacement ductility factors of the beam end were calculated for four types of steel. It was pointed out that the strength ratio computed from the nominal minimum strength specified in design codes led to inaccurate ductility factors of steel beams. After incorporating the deformation of columns， the story ductility factor of the frame was evaluated. Curves relating the story drift angle to the maximum strain at the beam end were presented， and an approximate formula was proposed. Furthermore， for shear-type energy-dissipating coupling beams， the maximum strain corresponding to a story drift angle of 1/50 was given when such beams were treated as flexural members.
2026, 41(6): 69-71.
doi: 10.13206/j.gjgS26052135
Abstract:
Within the inelastic range， Euler’s formula becomes inapplicable. Engesser successively proposed the tangent modulus theory and the dual modulus theory to address the buckling behavior of axially compressed members in the inelastic region. Subsequent validation confirmed the validity of the tangent modulus theory. Accordingly， the critical stress for axially compressed members in the elastic region can be expressed by Euler’s formula， and that for the inelastic region by the Engesser formula. By introducing a safety factor， the column curve can be obtained. As China’s first steel structure design code， the Design Code for Steel Structures （TJ 17-74） adopts this methodology to derive its column curve. However， the relevant parameters are not theoretical steel strength values but are determined through experimental testing.
Within the inelastic range， Euler’s formula becomes inapplicable. Engesser successively proposed the tangent modulus theory and the dual modulus theory to address the buckling behavior of axially compressed members in the inelastic region. Subsequent validation confirmed the validity of the tangent modulus theory. Accordingly， the critical stress for axially compressed members in the elastic region can be expressed by Euler’s formula， and that for the inelastic region by the Engesser formula. By introducing a safety factor， the column curve can be obtained. As China’s first steel structure design code， the Design Code for Steel Structures （TJ 17-74） adopts this methodology to derive its column curve. However， the relevant parameters are not theoretical steel strength values but are determined through experimental testing.
