Skip to main content

Random Field Modeling and Geotechnical Reliability

Scientific Question: How to adequately consider spatial variability of granular materials, accurately characterize and model their physical and mechanical properties, and elucidate their influence on various engineering problems?

Research Background​

The mechanical behavior of granular materials exhibits typical multiscale correlation characteristics: microscopic particle fabric features (such as size, morphology, porosity, etc.) directly affect complex mechanical responses of macroscopic accumulations, causing their macroscopic physical and mechanical parameters to show significant spatial variability.

Addressing the industry pain points of "difficult parameter variation quantification, difficult disaster mechanism prediction, and lack of engineering control basis" in macroscopic mechanical behavior of granular materials, this direction establishes a multifaceted framework of "investigation optimization — random modeling — reliability analysis" to破解 the key constraints of spatial variability on safety assessment of major geotechnical engineering projects.

Random Field Modeling and Geotechnical Reliability
Random Field Modeling and Geotechnical Reliability

Conceptual schematic of random field modeling and geotechnical reliability (ClaudeBot generated)

Core Research Contents​

1. Investigation Optimization Based on Spatial Variability Characteristics​

Scientific Question: Traditional investigation schemes struggle to balance exploration cost and parameter characterization accuracy, with poor adaptability to complex site morphology.

Innovation Breakthroughs:

  • Random Field Theory Evaluation Method: Establishes investigation scheme reliability evaluation methods based on random field theory, achieving probabilistic fusion of heterogeneous investigation data including geology, boreholes, and CPT
  • Bayesian-Markov Chain Monte Carlo Joint Inversion: Constructs site characterization parameter determination framework, breaking through limitations of traditional methods
  • Centroidal Voronoi Optimization Algorithm: Innovatively proposes Centroidal Voronoi Tessellation (CVT)-driven exploration point optimization algorithm, achieving 20% reliability improvement under equivalent investigation cost compared to traditional grid layout schemes

Application Scope: This method can be extended to site characterization, investigation, and optimization of geotechnical parameters such as site liquefaction potential, compaction degree, and pollutant concentration.

2. Rotated Anisotropic Random Field Modeling​

Scientific Question: Engineering disaster mechanisms under coupling effects of spatial variability and formation anisotropy are unclear, and existing codes have not yet considered parameter spatial distribution characteristics.

Innovation Breakthroughs:

  • Non-stationary Random Field Framework: Proposes random field modeling methods considering spatial non-stationary characteristics and rotated anisotropy
  • Slope Disaster Energy Evolution: Takes slope engineering as research object, studies energy evolution patterns during slope disaster processes, and proposes a new non-homogeneous geotechnical slope failure criterion based on energy evolution
  • Seabed-Pipeline System Analysis: Takes seabed-submarine pipeline system as research object, analyzes wave dynamic load effects on seabed liquefaction potential and pipeline mechanical behavior, revealing the relatively unfavorable 45° rotated anisotropy influence规律 on seabed-submarine pipeline systems

3. Coastal Soft Ground Risk Management Engineering Applications​

Scientific Question: Coastal mixed soft foundations exhibit both significant spatial variability and rotated anisotropy, with traditional investigation layout and risk assessment accuracy limited.

Innovation Breakthroughs:

  • Systematically integrates centroidal Voronoi diagram-driven investigation optimization theory with rotated anisotropy-coupled reliability analysis methods
  • Constructs "intelligent deployment — random field modeling — energy evolution criterion" technical chain
  • Combines deep learning methods to achieve probabilistic fusion and adaptive deployment of multi-source investigation data (boreholes, CPT, geophysical exploration)

Engineering Applications: This method has been applied in projects such as Fangjiuni Bridge, Shenzhen Mawan Project, and Nansha-Nanzhuzhong Intercity, providing geological risk identification and parameter optimization schemes for coastal cross-sea infrastructure, effectively avoiding potential instability risks.

Research Significance​

Provides theoretical framework for engineering characterization and disaster risk assessment of granular material spatial variability, significantly improving reliability of site parameter characterization, serving coastal infrastructure safety design and construction decision-making.

Representative Publications​

Random Field Modeling and Investigation Optimization​

  1. Huang, L., Huang, S., & Lai, Z.# (2020). On the optimization of site investigation programs using centroidal Voronoi tessellation and random field theory. Computers and Geotechnics, 118, 103331. DOI | PDF

  2. Zhu, B., Liu, J., Lai, Z., & Qian, T. (2023). Sampling Gaussian stationary random fields: A stochastic realization approach. ISA Transactions, 142, 386-398. DOI | PDF

Seabed-Pipeline Systems and Reliability Analysis​

  1. Lai, Z., Chen, Q., & Huang, L. (2021). Evaluating the hydromechanical responses of seabed–pipelines with rotated anisotropic heterogeneous seabed properties. Ocean Engineering, 234, 109226. DOI | PDF

  2. Lai, Z., Chen, Q., Wang, C., & Zhou, X. (2019). Modeling dynamic responses of heterogeneous seabed with embedded pipeline through multiresolution random field and coupled hydromechanical simulations. Ocean Engineering, 173, 556-570. DOI | PDF

Slope Stability and Energy Criteria​

  1. Huang, L., Huang, S., & Lai, Z.# (2019). On an energy-based criterion for defining slope failure considering spatially varying soil properties. Engineering Geology, 264, 105323. DOI | PDF

Multiscale Geotechnical Material Modeling​

  1. Lin, Y., Lai, Z., Ma, J., & Huang, L. (2024). A combined weighted Voronoi tessellation and random field approach for modeling heterogeneous rocks with correlated grain structure. Construction and Building Materials, 416, 135228. DOI | PDF

  2. Lin, Y., Ma, J., Lai, Z., Huang, L., & Lei, M. (2023). A FDEM approach to study mechanical and fracturing responses of geo-materials with high inclusion contents using a novel reconstruction strategy. Engineering Fracture Mechanics, 282, 109171. DOI | PDF