硅酸盐学报

基于生成对抗网络的硬化浆体孔隙结构三维重构

PDF下载

韩雨宸 ¹ , 李雨辰 ¹ , 张增起 ³ , 韩方晖 ¹^,²

硅酸盐学报 | Research Articles 2026,54(10): 3727-3739

收起

硅酸盐学报 |Research Articles 2026 , 54 (10) : 3727 -3739

基于生成对抗网络的硬化浆体孔隙结构三维重构

全屏

韩雨宸¹, 李雨辰¹, 张增起³, 韩方晖¹^,²

作者信息

^1.北京科技大学未来城市学院，北京　100083

^2.北京科技大学城市安全群体智能北京市重点实验室，北京　100083

^3.北京科技大学冶金与生态工程学院，北京　100083

通讯作者:

韩方晖（1988-），女，博士，教授。

作者简介:

韩雨宸（2001-），女，硕士研究生。

Three-Dimensional Reconstruction of Pore Structure in Hardened Paste via Generative Adversarial Networks

Yuchen HAN¹, Yuchen LI¹, Zengqi ZHANG³, Fanghui HAN¹^,²

Affiliations

^1.School of Future Cities, University of Science and Technology Beijing, Beijing 100083, China

^2.Beijing Key Laboratory for Urban Safety and Collective Intelligence, University of Science and Technology Beijing, Beijing 100083, China

^3.School of Metallurgical and Ecological Engineering, University of Science and Technology Beijing, Beijing 100083, China

Corresponding Author:

HAN Fanghui (1988-), female, Ph.D., Professor. E-mail: hanyang-1120@163.com

HAN Yuchen (2001-), female, Master candidate. E-mail: M202320159@xs.ustb.edu.cn

出版时间: 2026-05-22 doi: 10.14062/j.issn.0454-5648.20250791

文章导航

摘要

收起

基于X射线CT图像的水泥净浆微观结构的精确三维重构对水泥基材料的性能预测至关重要，但传统方法难以捕捉复杂的多尺度孔隙结构或计算成本高昂。为解决这些挑战，本工作提出了一种深度卷积生成对抗网络（DCGAN）架构，将自动编码器集成用于多尺度特征学习，并在带梯度惩罚的Wasserstein生成对抗网络（WGAN-GP）框架下实现，以增强训练稳定性。引入孔隙率约束损失函数以确保生成过程中的全局孔隙一致性。该模型基于石灰石煅烧黏土水泥浆体[55%（质量分数）普通硅酸盐水泥，30%煅烧黏土，15%石灰石；水胶比=0.42]在养护龄期为3、28、90 d和180 d时的微CT图像进行训练。利用闵可夫斯基泛函（孔隙率、表面积、欧拉特征]和空间相关函数[两点相关函数（S2）、线路径函数、两点聚类函数和表面相关函数]对重建的三维结构与实验真实微观结构进行验证。结果表明，所提出的模型在保持与原始微结构孔隙特征一致的同时，具有较高的结构保真度，并显著提高了计算效率（提高约150%）。该方法能够实现高精度的微结构重构，进而实现更可靠的基于微观结构的水泥基材料性能模拟。

关键词

硬化浆体 / 神经网络 / CT图像 / 孔隙结构 / 三维重构

Abstract

收起

Introduction

In recent years, the continuous development of 3D reconstruction methods for porous media has promoted the integrated evolution of image modeling, performance prediction, and multiscale microstructure analysis. Although traditional reconstruction approaches achieve satisfactory matches for low-order statistical descriptors, they often struggle to accurately reproduce realistic geometric morphology and topological characteristics across multiple scales. Meanwhile, traditional generative adversarial network (GAN) methods exhibit remarkable advantages in handling complex porous structures (e.g., cement-based materials), yet are prone to issues such as mode collapse and gradient saturation. To address the aforementioned challenges, a high-fidelity 3D reconstruction method based on deep convolutional generative adversarial networks (DCGANs) is proposed in this paper, incorporating an autoencoder module within the generator to enhance multiscale feature representation. The model introduces a porosity constraint term in the adversarial learning framework and integrates Wasserstein generative adversarial networks with gradient penalty (WGAN-GP), significantly improving training stability and convergence performance. Trained on CT images of new silicoaluminate cement paste at different curing ages, the model demonstrates good performance in both geometric morphology and statistical feature consistency of the reconstructed results.

Methods

The samples were prepared using the new silicoaluminate (limestone and calcined clay cement) system with 55% (mass fraction) Portland cement, 30% calcined clay, 15% limestone powder, and a water-to-binder ratio of 0.42. High-resolution micro-CT images were obtained at curing ages of 3, 28, 90 d, and 180 d. The images were preprocessed using Gaussian filtering and threshold segmentation to obtain binarized pore structures. A DCGAN architecture was developed by integrating 3D generative adversarial networks (GANs) with an autoencoder module. The framework incorporated gradient penalty to establish a WGAN-GP based on the Wasserstein distance, thereby constraining the Lipschitz continuity of the discriminator. A porosity constraint term was introduced into the adversarial learning framework to ensure global statistical consistency of the reconstructed microstructures. The reconstructed 3D volumes were validated against experimental microstructures using Minkowski functionals (porosity, surface area, Euler characteristic) and spatial correlation descriptors including the two-point correlation function (S2), lineal-path function, two-point cluster function, and surface correlation function.

Results and discussion

The results demonstrate that the proposed model effectively reconstructs pore morphology and distribution characteristics across different curing ages, with the porosity of reconstructed volumes showing good consistency with measured samples. The introduction of a gradient penalty term in the loss function enhanced the training stability and convergence efficiency of the model. By incorporating porosity as a regularization term in the adversarial learning framework, the model was effectively guided to learn the porosity evolution trends across different curing ages, thereby improving the consistency between generated structures and experimental data. The generator employed a combined ReLU and GELU activation mechanism, which contributed to stable generator training. The integration of Instance Normalization in the discriminator further improved the quality of generated results and the generalization capability of the model. Comparative analyses based on Minkowski functionals and statistical descriptors revealed that the generated structures exhibit high agreement with the original CT images in terms of geometric morphology, connectivity, and spatial correlation. Compared to conventional reconstruction methods, the proposed approach improves computational efficiency by approximately 150% while maintaining structural fidelity.

Conclusions

This study constructed an efficient 3D generative adversarial network (GAN) model based on CT images of cement paste, achieving high-precision 3D reconstruction of pore structures at different curing ages. The proposed model demonstrates good stability and robustness during training and achieves high-quality reconstruction of pore structures in cement-based materials while effectively preserving structural diversity and morphological realism.

Key words

hardened paste / neural network / CT images / pore structure / three-dimensional reconstruction

引用本文

韩雨宸, 李雨辰, 张增起, 韩方晖. 基于生成对抗网络的硬化浆体孔隙结构三维重构. 硅酸盐学报, 2026 , 54 (10) : 3727 -3739 . DOI: 10.14062/j.issn.0454-5648.20250791

Yuchen HAN, Yuchen LI, Zengqi ZHANG, Fanghui HAN. Three-Dimensional Reconstruction of Pore Structure in Hardened Paste via Generative Adversarial Networks[J]. Journal of the Chinese Ceramic Society, 2026 , 54 (10) : 3727 -3739 . DOI: 10.14062/j.issn.0454-5648.20250791

参考文献

收起

文献

收起

参考文献引证文献

排序方式：

[1]

ZHANG T, JI X, LU F F. 3D reconstruction of porous media by combining scaling transformation and multi-scale discrimination using generative adversarial networks[J]. J Petrol Sci Eng, 2022, 209: 109815.

[2]

HONG S W, KIM S Y, PARK K, et al. Mechanical property evaluation of 3D multi-phase cement paste microstructures reconstructed using generative adversarial networks[J]. Cem Concr Compos, 2024, 152: 105646.

[3]

TONG Z, WANG Z J, WANG X F, et al. Characterization of hydration and dry shrinkage behavior of cement emulsified asphalt composites using deep learning[J]. Constr Build Mater, 2021, 274: 121898.

[4]

SUNDARARAGHAVAN V, ZABARAS N. Classification and reconstruction of three-dimensional microstructures using support vector machines[J]. Comput Mater Sci, 2005, 32(2): 223-239.

[5]

SHAMS R, MASIHI M, BOOZARJOMEHRY R B, et al. Coupled generative adversarial and auto-encoder neural networks to reconstruct three-dimensional multi-scale porous media[J]. J Petrol Sci Eng, 2020, 186: 106794.

[6]

XU H Y, DIKIN D A, BURKHART C, et al. Descriptor-based methodology for statistical characterization and 3D reconstruction of microstructural materials[J]. Comput Mater Sci, 2014, 85: 206-216.

[7]

HENKES A, WESSELS H. Three-dimensional microstructure generation using generative adversarial neural networks in the context of continuum micromechanics[J]. Comput Meth Appl Mech Eng, 2022, 400: 115497.

[8]

KIM J S, KIM J H, HAN T S. Microstructure characterization of cement paste from micro-CT and correlations with mechanical properties evaluated from virtual and real experiments[J]. Mater Charact, 2019, 155: 109807.

[9]

YANG Y F, LIU F G, YAO J, et al. Multi-scale reconstruction of porous media from low-resolution core images using conditional generative adversarial networks[J]. J Nat Gas Sci Eng, 2022, 99: 104411.

[10]

MOSSER L, DUBRULE O, BLUNT M J. Reconstruction of three-dimensional porous media using generative adversarial neural networks[J]. Phys Rev E, 2017, 96(4): 043309.

[11]

SAHIMI M, TAHMASEBI P. Reconstruction, optimization, and design of heterogeneous materials and media: Basic principles, computational algorithms, and applications[J]. Phys Rep, 2021, 939: 1-82.

[12]

CORRALES M, IZZATULLAH M, HOTEIT H, et al. A Wasserstein GAN with gradient penalty for 3D porous media generation[C]//Second EAGE Subsurface Intelligence Workshop. Manama, Bahrain. European Association of Geoscientists & Engineers, 2022: 1-5.

[13]

ZHANG T, NI M L, GUAN Q J, et al. Reconstruction of three-dimensional porous media using multi-scale generative adversarial networks[J]. J Appl Geophys, 2023, 213: 105042.

[14]

WU Z T, HOU B, JIAO L C. Multiscale CNN with autoencoder regularization joint contextual attention network for SAR image classification[J]. IEEE Trans Geosci Remote Sens, 2021, 59(2): 1200-1213.

[15]

周嵘, 吴朝东, 张亚楠. 基于生成对抗网络的砂岩薄片图像视野外重建[J]. 北京大学学报(自然科学版), 2023, 59(2): 231-241.

ZHOU Rong, WU Chaodong, ZHANG Yanan. Acta Sci Nat Univ Pekin, 2023, 59(2): 231-241.

[16]

MASON D R, LONDON A J. Morphological analysis of 3d atom probe data using Minkowski functionals[J]. Ultramicroscopy, 2020, 211: 112940.

[17]

FENG Y B, LI Y Y, XU L. Existence of solutions to the Gaussian dual Minkowski problem[J]. J Differ Equ, 2025, 416: 268-298.

[18]

VERONESI N, MARULLI F, VEROPALUMBO A, et al. Artificial neural networks for galaxy clustering: Learning from the two-point correlation function of BOSS galaxies[J]. Astron Comput, 2023, 42: 100692.

[19]

DEROSSI A, SEVERINI C, DE PILLI T. Reconstruction of food microstructure via statistical correlation functions. The use of lineal-path distribution functions[J]. J Food Eng, 2014, 142: 9-16.

[20]

LI Y, LUO Y Z, LIAN D H, et al. Outdoor LiDAR point cloud classification algorithm based on multilevel point cluster feature fusion[J]. Heliyon, 2025, 11(4): e42623.

[21]

POSTNICOV V, KARSANINA M V, KHLYUPIN A, et al. The 2- and 3-point surface correlation functions calculations: From novel exact continuous approach to improving methodology for discrete images[J]. Phys A Stat Mech Appl, 2023, 628: 129137.

2026年第54卷第10期

PDF下载

引用本文

BibTeX

文章信息

doi: 10.14062/j.issn.0454-5648.20250791

接收时间：2025-10-30
首发时间：2026-10-22
出版时间：2026-05-22

补充材料

相关文章

文章信息

作者

出版历史

收稿日期：2025-10-30
修回日期：2026-01-22

基金

作者信息

^1.北京科技大学未来城市学院，北京　100083

^2.北京科技大学城市安全群体智能北京市重点实验室，北京　100083

^3.北京科技大学冶金与生态工程学院，北京　100083

通讯作者:

韩方晖（1988-），女，博士，教授。

参考文献

分享链接

https://castjournals.cast.org.cn/joweb/gsyxb/CN/10.14062/j.issn.0454-5648.20250791

分享至

全文二维码

扫描看全文

引用本文

BibTeX

本文的引用情况

2种不同金属材料的力学参数

科 Family	属数 Number of genus	种数 Number of species	占总种数比例 Percentage of total species (%)	属 Genus	种数 Number of species	占总种数比例 Percentage of total species (%)
鹅膏菌科Amanitaceae	2	11	5.26	鹅膏菌属 Amanita	10	4.78
小菇科 Mycenaceae	2	12	5.74	丝盖伞属 Inocybe	5	2.39
多孔菌科 Polyporaceae	8	14	6.70	蜡蘑属 Laccaria	5	2.39
红菇科 Russulaceae	3	23	11.00	小皮伞属 Marasmius	6	2.87
				小菇属 Mycena	11	5.26
				光柄菇属 Pluteus	5	2.39
				红菇属 Russula	17	8.13
				栓菌属 Trametes	5	2.39

关闭全屏

BibTeX
EndNote
RefWorks
TxT