数字孪生边坡有限元稳定分析
Finite Element Stability Analysis of Digital Twin Slopes
摘要: 针对高边坡施工阶段几何边界持续变化、有限元建模前处理重复性强、多开挖阶段和多扰动工况难以快速检算的问题,提出一种基于数字孪生几何底座的边坡自动化有限元建模与稳定性分析方法。根据数字孪生成熟度层级,本文工作定位为边坡稳定分析数字孪生体系中的“数字模型–数字影子”阶段,即以三维几何模型和有限元仿真流程为核心,实现工程几何信息到计算模型的自动映射,但尚未包含现场监测反演和实时闭环控制。该方法以等高线、钻孔分层、开挖台阶和支护设计参数为基础,利用OpenCascade几何内核构建ZK16高边坡三维几何模型,并通过开挖包络体与坡体实体的布尔差集运算表达施工阶段坡面演化。在给定控制剖面后,系统可随施工步自动完成剖面几何更新与数据提取,生成Abaqus可读取的二维剖面数据,并进一步完成材料赋值、边界约束、等效支护、网格划分和强度折减计算。以ZK16典型剖面为对象开展数值分析,结果表明:第1至第4施工步安全系数均达到搜索上限Fs ≥ 2.00,第5、第6施工步分别降至1.88和1.56,说明后期开挖阶段为稳定性控制阶段;强度软化20%和30%时安全系数分别降至1.22和1.10,表明岩土体抗剪强度衰减对边坡稳定性影响显著。在不计Abaqus求解器计算时间的前提下,本文流程可显著减少多施工步、多工况分析中的重复前处理交互时间。进一步基于400组有限元仿真样本开展1D-CNN代理模型初步验证,验证集决定系数R2 = 0.8071,说明自动化有限元样本库具有辅助快速评估的潜力。
Abstract: To address the problems of continuous geometric boundary changes during the staged construction of high slopes, highly repetitive preprocessing in finite element modeling, and the difficulty of rapid stability calculation under multiple excavation stages and disturbance conditions, an automatic finite element modeling and stability analysis method for slopes based on a digital twin geometric framework is proposed. According to the maturity levels of digital twins, this work is positioned at the “digital model-digital shadow” stage within the slope stability analysis digital twin system. It focuses on the three-dimensional geometric model and the finite element simulation workflow to achieve automatic mapping from engineering geometric information to computational models, but does not yet include field monitoring inversion and real-time closed-loop control. Based on contour lines, borehole stratification, excavation benches, and support design parameters, this method utilizes the OpenCascade geometric kernel to construct a three-dimensional geometric model of the ZK16 high slope. The slope surface evolution during the construction stage is represented by Boolean subtraction operations between the excavation envelope bodies and the slope entity. Once a control section is specified, the system automatically updates the section geometry and extracts data according to the construction steps, generating two-dimensional section data readable by Abaqus. It further completes material assignment, boundary constraints, equivalent support modeling, mesh generation, and strength reduction calculations. Numerical analysis is conducted on a typical section of the ZK16 high slope, and the results show that the safety factors of the 1st to 4th construction steps all reach the search upper limit of Fs ≥ 2.00, while those of the 5th and 6th construction steps drop to 1.88 and 1.56, respectively, indicating that the later excavation stages are the stability-controlling stages. Under conditions where the strength softens by 20% and 30%, the safety factors decrease to 1.22 and 1.10, respectively, demonstrating that the degradation of rock and soil shear strength has a significant impact on slope stability. Excluding the calculation time of the Abaqus solver, the proposed workflow can significantly reduce the repetitive preprocessing interaction time in multi-stage and multi-condition analysis. Furthermore, based on 400 sets of finite element simulation samples, a preliminary validation of a 1D-CNN surrogate model is conducted. The coefficient of determination on the validation set is R2 = 0.8071, indicating that the automated finite element sample database has the potential to assist in rapid assessment.
文章引用:明世雄, 何翔, 赵开明, 吴亦强, 黄浪, 龚志勇. 数字孪生边坡有限元稳定分析[J]. 土木工程, 2026, 15(6): 34-43. https://doi.org/10.12677/hjce.2026.156152

参考文献

[1] 郑颖人, 赵尚毅, 张鲁渝. 用有限元强度折减法进行边坡稳定分析[J]. 中国工程科学, 2002, 4(10): 57-61.
[2] 郑颖人, 赵尚毅. 有限元强度折减法在土坡与岩坡中的应用[J]. 岩石力学与工程学报, 2004, 23(19): 3381-3388.
[3] 荣光旭, 彭艳, 田凯. 基于Python的ABAQUS有限元强度折减法程序在边坡稳定性分析中的应用[J]. 中北大学学报(自然科学版), 2021, 42(4): 332-339.
[4] 李宁, 许建聪. 基于ABAQUS的三维边坡降雨入渗模块的开发及其应用[J]. 岩土工程学报, 2015, 37(4): 667-674.
[5] 杨雪, 苏谦, 牛云彬, 等. 基于BIM + 数字孪生的高边坡施工检算动态反馈方法[J]. 科学技术与工程, 2023, 23(20): 8812-8819.
[6] 阿迪力·如苏力, 李晓军, 芮易, 等. 基于数字孪生的边坡工程及时施工安全控制策略[J]. 同济大学学报(自然科学版), 2025, 53(9): 1361-1371.
[7] 陈超, 阿迪力·如苏力, 李晓军, 等. 基于无人机和数字孪生的边坡施工安全动态评估与反馈方法[J]. 测绘通报, 2025(5): 131-137.
[8] 杨进昆. 数字孪生驱动的公路边坡状态预测研究[D]: [硕士学位论文]. 兰州: 兰州交通大学, 2024.
[9] 邓炜. 边坡稳定性的人工智能模型及大数据样本的构造方法[D]: [硕士学位论文]. 广州: 广东工业大学, 2023.
[10] 武梦婷, 陈秋松, 齐冲冲. 基于机器学习的边坡安全稳定性评价及防护措施[J]. 工程科学学报, 2022, 44(2): 180-188.