Macro and mesoscopic mechanical behavior of concrete with actual aggregate segmented by hybrid Transformers and convolutional neural networks

The precise establishment of mesoscale models for concrete is crucial for understanding its mechanical behavior through numerical simulations. This study presents a deep learning-based numerical framework aimed at investigating both macro- and mesoscopic mechanical behavior of concrete, focusing on...

Full description

Saved in:
Bibliographic Details
Main Authors: Dong Wang, Junxing Zheng, Jichen Zhong, Lin Gao, Shuling Huang, Jiajia Zheng
Format: Article
Language:English
Published: Elsevier 2025-07-01
Series:Case Studies in Construction Materials
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2214509525002128
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850239083040735232
author Dong Wang
Junxing Zheng
Jichen Zhong
Lin Gao
Shuling Huang
Jiajia Zheng
author_facet Dong Wang
Junxing Zheng
Jichen Zhong
Lin Gao
Shuling Huang
Jiajia Zheng
author_sort Dong Wang
collection DOAJ
description The precise establishment of mesoscale models for concrete is crucial for understanding its mechanical behavior through numerical simulations. This study presents a deep learning-based numerical framework aimed at investigating both macro- and mesoscopic mechanical behavior of concrete, focusing on the actual aggregate morphology and distribution in 2D cross-sections. A dual encoder concrete aggregate segmentation network (DECAS-Net) based on CNN and Transformer is developed to segment aggregates and construct a refined mesoscale model compatible with the discrete element method (DEM). A multi-branch parallel dilated convolutional feature fusion module is designed to effectively integrate multi-scale features from the different encoders. Uniaxial compression tests are conducted to simulate the effects of actual aggregate shape and distribution on the macro- and mesoscopic mechanical behavior of concrete. Comparative analyses are performed utilizing models with circular aggregates and randomly distributed irregular aggregates, providing further insights into the influence of actual aggregate morphology and distribution on concrete mechanical behavior. The results demonstrate that the proposed DECAS-Net method provides accurate foundational data for establishing refined 2D mesoscale models, and highlights the significant impact of aggregate morphology and distribution on the mechanical behavior of concrete.
format Article
id doaj-art-57205f9f05ce4a1c96e204e050ee256c
institution OA Journals
issn 2214-5095
language English
publishDate 2025-07-01
publisher Elsevier
record_format Article
series Case Studies in Construction Materials
spelling doaj-art-57205f9f05ce4a1c96e204e050ee256c2025-08-20T02:01:15ZengElsevierCase Studies in Construction Materials2214-50952025-07-0122e0441410.1016/j.cscm.2025.e04414Macro and mesoscopic mechanical behavior of concrete with actual aggregate segmented by hybrid Transformers and convolutional neural networksDong Wang0Junxing Zheng1Jichen Zhong2Lin Gao3Shuling Huang4Jiajia Zheng5School of Civil and Hydraulic Engineering, Huazhong University of Science and Technology, Wuhan 430074, ChinaSchool of Civil and Hydraulic Engineering, Huazhong University of Science and Technology, Wuhan 430074, China; Corresponding authors.School of Civil and Hydraulic Engineering, Huazhong University of Science and Technology, Wuhan 430074, China; Corresponding authors.School of Civil and Hydraulic Engineering, Huazhong University of Science and Technology, Wuhan 430074, ChinaKey Laboratory of Geotechnical Mechanics and Engineering of the Ministry of Water Resources, Changjiang River Scientific Research Institute, Wuhan 430010, ChinaChina Road and Bridge Engineering Co., Ltd., Beijing 100011, ChinaThe precise establishment of mesoscale models for concrete is crucial for understanding its mechanical behavior through numerical simulations. This study presents a deep learning-based numerical framework aimed at investigating both macro- and mesoscopic mechanical behavior of concrete, focusing on the actual aggregate morphology and distribution in 2D cross-sections. A dual encoder concrete aggregate segmentation network (DECAS-Net) based on CNN and Transformer is developed to segment aggregates and construct a refined mesoscale model compatible with the discrete element method (DEM). A multi-branch parallel dilated convolutional feature fusion module is designed to effectively integrate multi-scale features from the different encoders. Uniaxial compression tests are conducted to simulate the effects of actual aggregate shape and distribution on the macro- and mesoscopic mechanical behavior of concrete. Comparative analyses are performed utilizing models with circular aggregates and randomly distributed irregular aggregates, providing further insights into the influence of actual aggregate morphology and distribution on concrete mechanical behavior. The results demonstrate that the proposed DECAS-Net method provides accurate foundational data for establishing refined 2D mesoscale models, and highlights the significant impact of aggregate morphology and distribution on the mechanical behavior of concrete.http://www.sciencedirect.com/science/article/pii/S2214509525002128Concrete macro and mesoscopic mechanical behaviorActual aggregate morphology and distributionDeep learningAggregate segmentationDiscrete element method
spellingShingle Dong Wang
Junxing Zheng
Jichen Zhong
Lin Gao
Shuling Huang
Jiajia Zheng
Macro and mesoscopic mechanical behavior of concrete with actual aggregate segmented by hybrid Transformers and convolutional neural networks
Case Studies in Construction Materials
Concrete macro and mesoscopic mechanical behavior
Actual aggregate morphology and distribution
Deep learning
Aggregate segmentation
Discrete element method
title Macro and mesoscopic mechanical behavior of concrete with actual aggregate segmented by hybrid Transformers and convolutional neural networks
title_full Macro and mesoscopic mechanical behavior of concrete with actual aggregate segmented by hybrid Transformers and convolutional neural networks
title_fullStr Macro and mesoscopic mechanical behavior of concrete with actual aggregate segmented by hybrid Transformers and convolutional neural networks
title_full_unstemmed Macro and mesoscopic mechanical behavior of concrete with actual aggregate segmented by hybrid Transformers and convolutional neural networks
title_short Macro and mesoscopic mechanical behavior of concrete with actual aggregate segmented by hybrid Transformers and convolutional neural networks
title_sort macro and mesoscopic mechanical behavior of concrete with actual aggregate segmented by hybrid transformers and convolutional neural networks
topic Concrete macro and mesoscopic mechanical behavior
Actual aggregate morphology and distribution
Deep learning
Aggregate segmentation
Discrete element method
url http://www.sciencedirect.com/science/article/pii/S2214509525002128
work_keys_str_mv AT dongwang macroandmesoscopicmechanicalbehaviorofconcretewithactualaggregatesegmentedbyhybridtransformersandconvolutionalneuralnetworks
AT junxingzheng macroandmesoscopicmechanicalbehaviorofconcretewithactualaggregatesegmentedbyhybridtransformersandconvolutionalneuralnetworks
AT jichenzhong macroandmesoscopicmechanicalbehaviorofconcretewithactualaggregatesegmentedbyhybridtransformersandconvolutionalneuralnetworks
AT lingao macroandmesoscopicmechanicalbehaviorofconcretewithactualaggregatesegmentedbyhybridtransformersandconvolutionalneuralnetworks
AT shulinghuang macroandmesoscopicmechanicalbehaviorofconcretewithactualaggregatesegmentedbyhybridtransformersandconvolutionalneuralnetworks
AT jiajiazheng macroandmesoscopicmechanicalbehaviorofconcretewithactualaggregatesegmentedbyhybridtransformersandconvolutionalneuralnetworks