Computers & Structures最新文献

Deep learning-based post-earthquake structural damage level recognition

IF 4.4 2区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Structures

Pub Date : 2025-04-25 DOI: 10.1016/j.compstruc.2025.107761

Xiaoying Zhuang , Than V. Tran , H. Nguyen-Xuan , Timon Rabczuk

Rapid assessment of building damage levels has become very important and has received considerable attention in structural engineering. Traditional methods for this work involve manual inspection, which is often tedious and time-consuming. Deep learning technology in computer vision has developed rapidly in recent years and has proven its superiority. This paper aims to develop an efficient approach to recognize quick post-earthquake structural damage levels. First, we develop a feature extraction with seven pre-trained CNN models (Xception, InceptionV3, InceptionResNetV2, MobileNet, MobileNetV2, DenseNet121, NASNetMobile) on a small dataset of 2000 images. The CNN models are then trained by five fold cross-validation. The performance of the models is compared on a testing set, the MobileNet model demonstrated the best classifier performance with an accuracy of 90.89 %. Second, the Bayesian optimization method and the fine-tuning strategy are used to find the optimal hyperparameters of the MobileNet model. The results revealed that the performance of the MobileNet model increased significantly with an accuracy of 96.11 %. Third, Gradient-weighted class activation mapping (Grad-CAM) is used to highlight crucial regions on structural damage images for CNN’s prediction. Finally, the generalizability of the MobileNet model is improved by training it on an extended dataset of 3600 images. The proposed approach demonstrates the feasibility and potential uses of deep learning in image-based structural damage level recognition.

{"title":"Deep learning-based post-earthquake structural damage level recognition","authors":"Xiaoying Zhuang , Than V. Tran , H. Nguyen-Xuan , Timon Rabczuk","doi":"10.1016/j.compstruc.2025.107761","DOIUrl":"10.1016/j.compstruc.2025.107761","url":null,"abstract":"<div><div>Rapid assessment of building damage levels has become very important and has received considerable attention in structural engineering. Traditional methods for this work involve manual inspection, which is often tedious and time-consuming. Deep learning technology in computer vision has developed rapidly in recent years and has proven its superiority. This paper aims to develop an efficient approach to recognize quick post-earthquake structural damage levels. First, we develop a feature extraction with seven pre-trained CNN models (Xception, InceptionV3, InceptionResNetV2, MobileNet, MobileNetV2, DenseNet121, NASNetMobile) on a small dataset of 2000 images. The CNN models are then trained by five fold cross-validation. The performance of the models is compared on a testing set, the MobileNet model demonstrated the best classifier performance with an accuracy of 90.89 %. Second, the Bayesian optimization method and the fine-tuning strategy are used to find the optimal hyperparameters of the MobileNet model. The results revealed that the performance of the MobileNet model increased significantly with an accuracy of 96.11 %. Third, Gradient-weighted class activation mapping (Grad-CAM) is used to highlight crucial regions on structural damage images for CNN’s prediction. Finally, the generalizability of the MobileNet model is improved by training it on an extended dataset of 3600 images. The proposed approach demonstrates the feasibility and potential uses of deep learning in image-based structural damage level recognition.</div></div>","PeriodicalId":50626,"journal":{"name":"Computers & Structures","volume":"315 ","pages":"Article 107761"},"PeriodicalIF":4.4,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143868517","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Generalized reconfigurations and growth mechanics of biological structures considering regular and irregular features: A computational study

IF 4.4 2区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Structures

Pub Date : 2025-04-24 DOI: 10.1016/j.compstruc.2025.107781

Nasser Firouzi , Krzysztof Kamil Żur , Timon Rabczuk , Xiaoying Zhuang

Many soft biological structures have natural features of viscoelastic and hyperelastic materials. Research focused on the growth biomechanics of these structures is challenging from theoretical and experimental points of view, especially when irregular forms/defects of biological objects should be considered. To this aim, an effort is made in this paper to develop a general nonlinear finite element model for the growth of biological soft structures such as arteries, skin or different tissues. The non-Newtonian fluid is considered for viscoelastic branches. The effect of variation in thickness growth and irregular geometry as well as defects of biostructure is taken into account for the first time. The general nonlinear formulations are obtained for isotropic as well as anisotropic material properties. Furthermore, to resolve evolution equations resulting of internal variables for growth as well as viscoelastic branches, two effective implicit trapezoidal time integration schemes are employed. To study the applicability of the proposed model, the obtained results are compared with results from clinical studies for skin growth, available in the literature. The results demonstrate that the present model enables to capture of the experimental observations with very good accuracy. Additionally, the presented model enables to study of different shapes of biostructure, and variation in thickness growth, including regular and irregular defects, which have never been investigated previously.

{"title":"Generalized reconfigurations and growth mechanics of biological structures considering regular and irregular features: A computational study","authors":"Nasser Firouzi , Krzysztof Kamil Żur , Timon Rabczuk , Xiaoying Zhuang","doi":"10.1016/j.compstruc.2025.107781","DOIUrl":"10.1016/j.compstruc.2025.107781","url":null,"abstract":"<div><div>Many soft biological structures have natural features of viscoelastic and hyperelastic materials. Research focused on the growth biomechanics of these structures is challenging from theoretical and experimental points of view, especially when irregular forms/defects of biological objects should be considered. To this aim, an effort is made in this paper to develop a general nonlinear finite element model for the growth of biological soft structures such as arteries, skin or different tissues. The non-Newtonian fluid is considered for viscoelastic branches. The effect of variation in thickness growth and irregular geometry as well as defects of biostructure is taken into account for the first time. The general nonlinear formulations are obtained for isotropic as well as anisotropic material properties. Furthermore, to resolve evolution equations resulting of internal variables for growth as well as viscoelastic branches, two effective implicit trapezoidal time integration schemes are employed. To study the applicability of the proposed model, the obtained results are compared with results from clinical studies for skin growth, available in the literature. The results demonstrate that the present model enables to capture of the experimental observations with very good accuracy. Additionally, the presented model enables to study of different shapes of biostructure, and variation in thickness growth, including regular and irregular defects, which have never been investigated previously.</div></div>","PeriodicalId":50626,"journal":{"name":"Computers & Structures","volume":"315 ","pages":"Article 107781"},"PeriodicalIF":4.4,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143868514","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Stochastic static finite element model updating using the Bayesian method integrating homotopy surrogate model

IF 4.4 2区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Structures

Pub Date : 2025-04-24 DOI: 10.1016/j.compstruc.2025.107769

Bin Huang , Ming Sun , Hui Chen , Zhifeng Wu

The Bayesian model updating method usually involves tens of thousands of finite element model calculations, which will bring huge computational costs to large structures such as bridges. To reduce the computational costs, this paper develops a highly efficient Bayesian model updating method based on a new static homotopy surrogate model. The new surrogate model is established on the basis of the finite element model using the stochastic homotopy method, which is different from the existing surrogate models that depend on the selected samples. Then by using the hybrid Monte Carlo sampling algorithm integrating the homotopy surrogate model, the static Bayesian model updating of structure is implemented. The numerical example of a plate demonstrates that the established surrogate model has higher accuracy than the polynomial response surface model and Kriging model. Based on the uncertain static test data, the finite element model of a continuous concrete box-girder bridge is efficiently updated using the new method. And the statistics of the displacements in the updated bridge are in good agreement with that of the uncertain measurement data.

{"title":"Stochastic static finite element model updating using the Bayesian method integrating homotopy surrogate model","authors":"Bin Huang , Ming Sun , Hui Chen , Zhifeng Wu","doi":"10.1016/j.compstruc.2025.107769","DOIUrl":"10.1016/j.compstruc.2025.107769","url":null,"abstract":"<div><div>The Bayesian model updating method usually involves tens of thousands of finite element model calculations, which will bring huge computational costs to large structures such as bridges. To reduce the computational costs, this paper develops a highly efficient Bayesian model updating method based on a new static homotopy surrogate model. The new surrogate model is established on the basis of the finite element model using the stochastic homotopy method, which is different from the existing surrogate models that depend on the selected samples. Then by using the hybrid Monte Carlo sampling algorithm integrating the homotopy surrogate model, the static Bayesian model updating of structure is implemented. The numerical example of a plate demonstrates that the established surrogate model has higher accuracy than the polynomial response surface model and Kriging model. Based on the uncertain static test data, the finite element model of a continuous concrete box-girder bridge is efficiently updated using the new method. And the statistics of the displacements in the updated bridge are in good agreement with that of the uncertain measurement data.</div></div>","PeriodicalId":50626,"journal":{"name":"Computers & Structures","volume":"315 ","pages":"Article 107769"},"PeriodicalIF":4.4,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143868516","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A probabilistic semi-explicit model for crack propagation in concrete structures under dynamic loading

IF 4.4 2区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Structures

Pub Date : 2025-04-24 DOI: 10.1016/j.compstruc.2025.107783

Gustavo Luz Xavier da Costa , Pierre Rossi , Mariane Rodrigues Rita , Magno Teixeira Mota , Rodolfo Giacomim Mendes de Andrade , Eduardo de Moraes Rego Fairbairn

In this paper, concrete cracking is investigated in dynamics through finite element modeling. A probabilistic approach is employed to translate the effects of material heterogeneity on tensile strength and fracture energy. Both parameters depend on compressive strength and heterogeneity degree (volumetric ratio between finite element and largest aggregate). Material softening is modeled through damage theory. The actual concrete strengthening effect is modeled by an empirical formulation and similar reasoning is adopted for fracture energy. The apparent strengthening effect is naturally captured when mass and damping are included in the equation of motion. A convergence test is shown, indicating the probabilistic model proposed here is mesh-insensitive, converging in the average sense when both finite element size and load/time increment decrease. Then, experimental data are selected from literature for a wide range of loading rates. The effect of strain rate on the dispersion of crack pattern, load-carrying capacity and load–displacement curve is discussed. The influence of structural damping on the shape of load–displacement curve is also remarked.

{"title":"A probabilistic semi-explicit model for crack propagation in concrete structures under dynamic loading","authors":"Gustavo Luz Xavier da Costa , Pierre Rossi , Mariane Rodrigues Rita , Magno Teixeira Mota , Rodolfo Giacomim Mendes de Andrade , Eduardo de Moraes Rego Fairbairn","doi":"10.1016/j.compstruc.2025.107783","DOIUrl":"10.1016/j.compstruc.2025.107783","url":null,"abstract":"<div><div>In this paper, concrete cracking is investigated in dynamics through finite element modeling. A probabilistic approach is employed to translate the effects of material heterogeneity on tensile strength and fracture energy. Both parameters depend on compressive strength and heterogeneity degree (volumetric ratio between finite element and largest aggregate). Material softening is modeled through damage theory. The actual concrete strengthening effect is modeled by an empirical formulation and similar reasoning is adopted for fracture energy. The apparent strengthening effect is naturally captured when mass and damping are included in the equation of motion. A convergence test is shown, indicating the probabilistic model proposed here is mesh-insensitive, converging in the average sense when both finite element size and load/time increment decrease. Then, experimental data are selected from literature for a wide range of loading rates. The effect of strain rate on the dispersion of crack pattern, load-carrying capacity and load–displacement curve is discussed. The influence of structural damping on the shape of load–displacement curve is also remarked.</div></div>","PeriodicalId":50626,"journal":{"name":"Computers & Structures","volume":"315 ","pages":"Article 107783"},"PeriodicalIF":4.4,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143868515","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Transfer learning-based artificial neural networks for hysteresis response prediction of steel braces

IF 4.4 2区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Structures

Pub Date : 2025-04-23 DOI: 10.1016/j.compstruc.2025.107777

Sepehr Pessiyan, Fardad Mokhtari, Ali Imanpour

This paper proposes a novel data-driven surrogate model for predicting the hysteresis response, i.e., axial force – axial deformation, of steel braces in concentrically braced frames under seismic loading using transfer learning-based artificial neural networks. Transfer learning is utilized to leverage pre-trained baseline long short-term memory networks and transfer its knowledge to the new hysteresis surrogate model. The proposed model is validated using four case studies involving various combinations of input data obtained from laboratory tests and data generated using random earthquake-induced vibration, featuring a wide range of frequency contents, amplitudes, and durations. A pseudo-dynamic analysis is then performed on a steel braced frame system to demonstrate the application of the proposed surrogate model in system-level response evaluation while verifying the performance of the model in real-time seismic simulations. The results obtained from the validation study confirm that the proposed brace hysteresis model can properly estimate the underlying physical relationship between the input displacement and output force using the transfer learning approach. The proposed model offers an efficient method to evaluate the dynamic response of steel braced frames.

{"title":"Transfer learning-based artificial neural networks for hysteresis response prediction of steel braces","authors":"Sepehr Pessiyan, Fardad Mokhtari, Ali Imanpour","doi":"10.1016/j.compstruc.2025.107777","DOIUrl":"10.1016/j.compstruc.2025.107777","url":null,"abstract":"<div><div>This paper proposes a novel data-driven surrogate model for predicting the hysteresis response, i.e., axial force – axial deformation, of steel braces in concentrically braced frames under seismic loading using transfer learning-based artificial neural networks. Transfer learning is utilized to leverage pre-trained baseline long short-term memory networks and transfer its knowledge to the new hysteresis surrogate model. The proposed model is validated using four case studies involving various combinations of input data obtained from laboratory tests and data generated using random earthquake-induced vibration, featuring a wide range of frequency contents, amplitudes, and durations. A pseudo-dynamic analysis is then performed on a steel braced frame system to demonstrate the application of the proposed surrogate model in system-level response evaluation while verifying the performance of the model in real-time seismic simulations. The results obtained from the validation study confirm that the proposed brace hysteresis model can properly estimate the underlying physical relationship between the input displacement and output force using the transfer learning approach. The proposed model offers an efficient method to evaluate the dynamic response of steel braced frames.</div></div>","PeriodicalId":50626,"journal":{"name":"Computers & Structures","volume":"315 ","pages":"Article 107777"},"PeriodicalIF":4.4,"publicationDate":"2025-04-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143859630","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Simulation of the TNT-based melt-cast explosive charging process using hot mandrel assisted solidification

IF 4.4 2区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Structures

Pub Date : 2025-04-22 DOI: 10.1016/j.compstruc.2025.107780

Xuezhen Zhai, Yongjia Zhang, Ge Kang, Pengwan Chen

The melt-cast charging process, widely used in warheads for its adaptability, cost efficiency, and automation, requires optimization to minimize defects such as shrinkage cavities and porosity that compromise explosive quality, destructive power, and safety, particularly in large-volume munitions. The hot mandrel technique, by providing localized heating during solidification, helps maintain an open feeding channel, thereby reducing defect formation and improving charge integrity. In this study, the solidification process of a TNT-based melt-cast explosive is investigated using ProCAST combined with an orthogonal test approach, focusing on the hot mandrel charging technique for a warhead. The influence of three primary process parameters—the hot mandrel length, heating time, and temperature—on the solidification process is analyzed. The results demonstrate that, compared to traditional natural solidification, the solidification process with hot mandrel assistance significantly reduces the occurrence of shrinkage cavities and porosity defects, decreases the volume of shrinkage-related flaws, and enhances the overall charge quality. Among the parameters studied, the heating time of the hot mandrel exerts the greatest influence on charge quality, followed by its temperature and length. Prolonging the heating time not only reduces shrinkage defects but also extends the solidification duration. Considering both defect reduction and solidification efficiency, the optimal process conditions within the tested range are as follows: a hot mandrel length of 350 mm, a heating time of 4000 s, and a hot mandrel temperature of 90

^{\circ}

C. This study innovatively develops a numerical simulation approach using ProCAST for hot mandrel-assisted solidification, systematically investigating the effects of three critical parameters on charge quality. The proposed optimization framework balances defect control with production efficiency, providing theoretical guidance for industrial implementation.

{"title":"Simulation of the TNT-based melt-cast explosive charging process using hot mandrel assisted solidification","authors":"Xuezhen Zhai, Yongjia Zhang, Ge Kang, Pengwan Chen","doi":"10.1016/j.compstruc.2025.107780","DOIUrl":"10.1016/j.compstruc.2025.107780","url":null,"abstract":"<div><div>The melt-cast charging process, widely used in warheads for its adaptability, cost efficiency, and automation, requires optimization to minimize defects such as shrinkage cavities and porosity that compromise explosive quality, destructive power, and safety, particularly in large-volume munitions. The hot mandrel technique, by providing localized heating during solidification, helps maintain an open feeding channel, thereby reducing defect formation and improving charge integrity. In this study, the solidification process of a TNT-based melt-cast explosive is investigated using ProCAST combined with an orthogonal test approach, focusing on the hot mandrel charging technique for a warhead. The influence of three primary process parameters—the hot mandrel length, heating time, and temperature—on the solidification process is analyzed. The results demonstrate that, compared to traditional natural solidification, the solidification process with hot mandrel assistance significantly reduces the occurrence of shrinkage cavities and porosity defects, decreases the volume of shrinkage-related flaws, and enhances the overall charge quality. Among the parameters studied, the heating time of the hot mandrel exerts the greatest influence on charge quality, followed by its temperature and length. Prolonging the heating time not only reduces shrinkage defects but also extends the solidification duration. Considering both defect reduction and solidification efficiency, the optimal process conditions within the tested range are as follows: a hot mandrel length of 350 mm, a heating time of 4000 s, and a hot mandrel temperature of 90 <span><math><msup><mspace></mspace><mrow><mo>∘</mo></mrow></msup></math></span>C. This study innovatively develops a numerical simulation approach using ProCAST for hot mandrel-assisted solidification, systematically investigating the effects of three critical parameters on charge quality. The proposed optimization framework balances defect control with production efficiency, providing theoretical guidance for industrial implementation.</div></div>","PeriodicalId":50626,"journal":{"name":"Computers & Structures","volume":"315 ","pages":"Article 107780"},"PeriodicalIF":4.4,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143859631","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A new three-dimensional model of train-track-bridge coupled system based on meshless method and its graph neural network-based surrogate model

IF 4.4 2区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Structures

Pub Date : 2025-04-22 DOI: 10.1016/j.compstruc.2025.107786

Zhanjun Shao , Peng Zhang , Xiaonan Xie , Zihe Wang , Xuan Peng , Zefeng Liu , Yufei Chen , Ping Xiang

A model of train–track–bridge coupled system is proposed to study the interactions between structures in greater detail. The new model employs a meshless method to numerically simulate the box girder bridge and track slab. In the dynamic analysis, the system at each time step is abstracted into a graph structure and trained using a graph neural network to develop a surrogate prediction model. The graph neural network node connections in the bridge top plate are determined by the meshless method. Multiple numerical examples demonstrate the differences in structural response between the proposed model and the conventional model and evaluate the performance and self-evolutionary capabilities of the surrogate model. The results indicate that, compared to the proposed model, the conventional model underestimates vertical responses by approximately 17 %–69 % and lateral responses by one to two orders of magnitude. The surrogate model demonstrates good displacement prediction capabilities for the bridge on the training dataset, achieving an

R^{2}

value as high as 0.99. Furthermore, it exhibits robust prediction and self-evolutionary capabilities on the test dataset under topological changes, with prediction accuracy decreasing by only about 2 %. However, the prediction performance for rail responses is relatively poor, with an

R^{2}

value as low as 0.29.

为更详细地研究结构间的相互作用，提出了火车-轨道-桥梁耦合系统模型。新模型采用无网格法对箱梁桥和轨道板进行数值模拟。在动态分析中，将每个时间步的系统抽象为图结构，并使用图神经网络进行训练，以建立代用预测模型。桥梁顶板的图神经网络节点连接由无网格法确定。多个数值实例证明了所提出的模型与传统模型在结构响应方面的差异，并评估了代用模型的性能和自我进化能力。结果表明，与提出的模型相比，传统模型低估了约 17%-69% 的垂直响应和一到两个数量级的横向响应。代用模型在训练数据集上表现出良好的桥梁位移预测能力，R2 值高达 0.99。此外，在拓扑变化条件下，代用模型在测试数据集上表现出稳健的预测和自我进化能力，预测精度仅下降约 2%。然而，铁路响应的预测性能相对较差，R2 值低至 0.29。

{"title":"A new three-dimensional model of train-track-bridge coupled system based on meshless method and its graph neural network-based surrogate model","authors":"Zhanjun Shao , Peng Zhang , Xiaonan Xie , Zihe Wang , Xuan Peng , Zefeng Liu , Yufei Chen , Ping Xiang","doi":"10.1016/j.compstruc.2025.107786","DOIUrl":"10.1016/j.compstruc.2025.107786","url":null,"abstract":"<div><div>A model of train–track–bridge coupled system is proposed to study the interactions between structures in greater detail. The new model employs a meshless method to numerically simulate the box girder bridge and track slab. In the dynamic analysis, the system at each time step is abstracted into a graph structure and trained using a graph neural network to develop a surrogate prediction model. The graph neural network node connections in the bridge top plate are determined by the meshless method. Multiple numerical examples demonstrate the differences in structural response between the proposed model and the conventional model and evaluate the performance and self-evolutionary capabilities of the surrogate model. The results indicate that, compared to the proposed model, the conventional model underestimates vertical responses by approximately 17 %–69 % and lateral responses by one to two orders of magnitude. The surrogate model demonstrates good displacement prediction capabilities for the bridge on the training dataset, achieving an <span><math><msup><mi>R</mi><mn>2</mn></msup></math></span> value as high as 0.99. Furthermore, it exhibits robust prediction and self-evolutionary capabilities on the test dataset under topological changes, with prediction accuracy decreasing by only about 2 %. However, the prediction performance for rail responses is relatively poor, with an <span><math><msup><mi>R</mi><mn>2</mn></msup></math></span> value as low as 0.29.</div></div>","PeriodicalId":50626,"journal":{"name":"Computers & Structures","volume":"315 ","pages":"Article 107786"},"PeriodicalIF":4.4,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143859632","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Computing the dynamic response of periodic waveguides with nonlinear boundaries using the wave finite element method

IF 4.4 2区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Structures

Pub Date : 2025-04-22 DOI: 10.1016/j.compstruc.2025.107778

Vincent Mahé , Adrien Mélot , Benjamin Chouvion , Christophe Droz

A new method to compute the dynamic response of periodic waveguides with localised nonlinearities is introduced and used to investigate the nonlinear shift of a band-edge mode in the bandgap of a locally resonant phononic structure. This nonlinear extension of the Wave Finite Element Method (WFEM) uses a finite-element discretisation of arbitrarily complex unit-cells, and leverages Floquet–Bloch theory to reduce the analysis of the entire waveguide to a state-vector of Bloch waves’ amplitude. Higher harmonics generated by nonlinear effects are addressed using the Harmonic Balance Method and the nonlinear forces are evaluated via an alternating frequency-time procedure. The periodic response of the system is computed through a continuation scheme, taking the Bloch waves’ amplitude as unknowns. The accuracy of the nonlinear WFEM is validated against standard FEM with Craig–Bampton reduction, demonstrating an 83 % speedup in resolution time. Applying the method to a locally resonant metamaterial demonstrates that nonlinear effects can shift resonances from outside to inside bandgaps, resulting in high-amplitude, spatially localised vibrations where small amplitudes are expected from linear theory. The versatility and computational efficiency of this nonlinear dynamic simulation method should facilitate the study of complex metamaterials and civil engineering structures coupled with nonlinear interfaces or singularities.

{"title":"Computing the dynamic response of periodic waveguides with nonlinear boundaries using the wave finite element method","authors":"Vincent Mahé , Adrien Mélot , Benjamin Chouvion , Christophe Droz","doi":"10.1016/j.compstruc.2025.107778","DOIUrl":"10.1016/j.compstruc.2025.107778","url":null,"abstract":"<div><div>A new method to compute the dynamic response of periodic waveguides with localised nonlinearities is introduced and used to investigate the nonlinear shift of a band-edge mode in the bandgap of a locally resonant phononic structure. This nonlinear extension of the Wave Finite Element Method (WFEM) uses a finite-element discretisation of arbitrarily complex unit-cells, and leverages Floquet–Bloch theory to reduce the analysis of the entire waveguide to a state-vector of Bloch waves’ amplitude. Higher harmonics generated by nonlinear effects are addressed using the Harmonic Balance Method and the nonlinear forces are evaluated via an alternating frequency-time procedure. The periodic response of the system is computed through a continuation scheme, taking the Bloch waves’ amplitude as unknowns. The accuracy of the nonlinear WFEM is validated against standard FEM with Craig–Bampton reduction, demonstrating an 83 % speedup in resolution time. Applying the method to a locally resonant metamaterial demonstrates that nonlinear effects can shift resonances from outside to inside bandgaps, resulting in high-amplitude, spatially localised vibrations where small amplitudes are expected from linear theory. The versatility and computational efficiency of this nonlinear dynamic simulation method should facilitate the study of complex metamaterials and civil engineering structures coupled with nonlinear interfaces or singularities.</div></div>","PeriodicalId":50626,"journal":{"name":"Computers & Structures","volume":"315 ","pages":"Article 107778"},"PeriodicalIF":4.4,"publicationDate":"2025-04-22","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143855019","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Structure mode shapes classification using graph convolutional networks in automotive application

IF 4.4 2区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Structures

Pub Date : 2025-04-19 DOI: 10.1016/j.compstruc.2025.107767

Sitthichart Tohmuang , Mohammad Fard , Pier Marzocca , James L. Swayze , John E. Huber , Haytham M. Fayek

Classifying vibration mode shapes of a structure in an engineering design cycle can be a labor intensive and repetitive task. Although several methods have been proposed to automatically classify mode shapes, most existing models cannot fully represent mode shapes using both structural and modal information, limiting their application to specific structures. In this paper, we propose a graph convolutional network (GCN) model, which learns to classify mode shapes from a graph perspective. The mode shape graphs were generated using both geometric and modal information derived from the Finite Element Method (FEM). In order to validate the model’s performance, Finite Element (FE) models of Sport Utility Vehicle (SUV) types were developed as representatives of the real-world complex structures. Both quantitative and qualitative assessments are performed to emphasise the advantages of representing mode shapes as graph data. Within the developed dataset, the classification results show that GCN models achieve 100 % precision across diverse geometric configurations and varying input conditions, outperforming existing methods such as Modal Assurance Criteria (MAC) and traditional Machine Learning (ML) techniques.

{"title":"Structure mode shapes classification using graph convolutional networks in automotive application","authors":"Sitthichart Tohmuang , Mohammad Fard , Pier Marzocca , James L. Swayze , John E. Huber , Haytham M. Fayek","doi":"10.1016/j.compstruc.2025.107767","DOIUrl":"10.1016/j.compstruc.2025.107767","url":null,"abstract":"<div><div>Classifying vibration mode shapes of a structure in an engineering design cycle can be a labor intensive and repetitive task. Although several methods have been proposed to automatically classify mode shapes, most existing models cannot fully represent mode shapes using both structural and modal information, limiting their application to specific structures. In this paper, we propose a graph convolutional network (GCN) model, which learns to classify mode shapes from a graph perspective. The mode shape graphs were generated using both geometric and modal information derived from the Finite Element Method (FEM). In order to validate the model’s performance, Finite Element (FE) models of Sport Utility Vehicle (SUV) types were developed as representatives of the real-world complex structures. Both quantitative and qualitative assessments are performed to emphasise the advantages of representing mode shapes as graph data. Within the developed dataset, the classification results show that GCN models achieve 100 % precision across diverse geometric configurations and varying input conditions, outperforming existing methods such as Modal Assurance Criteria (MAC) and traditional Machine Learning (ML) techniques.</div></div>","PeriodicalId":50626,"journal":{"name":"Computers & Structures","volume":"314 ","pages":"Article 107767"},"PeriodicalIF":4.4,"publicationDate":"2025-04-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143848662","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Bayesian model condensation and selection of master degrees of freedom

IF 4.4 2区工程技术 Q1 COMPUTER SCIENCE, INTERDISCIPLINARY APPLICATIONS

Computers & Structures

Pub Date : 2025-04-17 DOI: 10.1016/j.compstruc.2025.107768

Ce Huang , Ting Liu , Li Wang

Condensation of large-scale finite element models while maintaining high prediction accuracy is crucial for efficient structural analysis and design. To this end, a novel Bayesian framework for model condensation and selection of master degrees of freedom (DOFs) is developed in this paper. The main idea behind it is to recast model condensation into the Bayesian full-field reconstruction problem. In doing so, the key lies in the definition of the response covariance matrix so that the transformation matrix for model condensation is obtained through the conditional Gaussian distribution analysis. It is also shown that this approach can coincide with the conventional static/dynamic condensation or Schur complement schemes after choosing proper response covariance matrices. Moreover, since the response covariance depends on the load, the developed approach can streamline model condensation by leveraging prior information on load locations and spectral properties, ultimately reducing the computational overhead while preserving accuracy. On the other hand, within the Bayesian framework, the master DOFs are efficiently selected to iteratively encompass the slave ones with maximum posterior covariance, and this leads to minimizing the prediction covariance as well as motivating the heuristic automatic determination of the number of master DOFs. Numerical examples on static/dynamic cases and with comparisons to several existing methods are investigated to highlight the performance of the Bayesian model condensation and master DOFs selection approach.

{"title":"Bayesian model condensation and selection of master degrees of freedom","authors":"Ce Huang , Ting Liu , Li Wang","doi":"10.1016/j.compstruc.2025.107768","DOIUrl":"10.1016/j.compstruc.2025.107768","url":null,"abstract":"<div><div>Condensation of large-scale finite element models while maintaining high prediction accuracy is crucial for efficient structural analysis and design. To this end, a novel Bayesian framework for model condensation and selection of master degrees of freedom (DOFs) is developed in this paper. The main idea behind it is to recast model condensation into the Bayesian full-field reconstruction problem. In doing so, the key lies in the definition of the response covariance matrix so that the transformation matrix for model condensation is obtained through the conditional Gaussian distribution analysis. It is also shown that this approach can coincide with the conventional static/dynamic condensation or Schur complement schemes after choosing proper response covariance matrices. Moreover, since the response covariance depends on the load, the developed approach can streamline model condensation by leveraging prior information on load locations and spectral properties, ultimately reducing the computational overhead while preserving accuracy. On the other hand, within the Bayesian framework, the master DOFs are efficiently selected to iteratively encompass the slave ones with maximum posterior covariance, and this leads to minimizing the prediction covariance as well as motivating the heuristic automatic determination of the number of master DOFs. Numerical examples on static/dynamic cases and with comparisons to several existing methods are investigated to highlight the performance of the Bayesian model condensation and master DOFs selection approach.</div></div>","PeriodicalId":50626,"journal":{"name":"Computers & Structures","volume":"314 ","pages":"Article 107768"},"PeriodicalIF":4.4,"publicationDate":"2025-04-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143838949","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":2,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0