Abstract: This article presents a multi-phase-field poromechanics model that simulates the growth and thaw of ice lenses and the resultant frozen heave and thaw settlement in multi-constituent frozen soils. In this model, the growth of segregated ice inside the freezing-induced fracture is implicitly represented by the evolution of two phase fields that indicate the locations of segregated ice and the damaged zone, respectively. The evolution of two phase fields are driven by the driving forces that capture the physical mechanisms of ice and crack growths respectively, while the phase field governing equations are coupled with the balance laws such that the coupling among heat transfer, solid deformation, fluid diffusion, crack growth, and phase transition can be observed numerically. Unlike phenomenological approaches that indirectly captures the freezing influence on the shear strength, the multi-phase-field model introduces an immersed approach where both the homogeneous freezing and the ice lens growth are distinctively captured by the freezing characteristic function and the driving force accordingly. Verification and validation examples are provided to demonstrate the capacities of the proposed models.
Introduction
When soil freezes, it does not simply transform into a uniform solid block. Instead, water often migrates through the pores to specific locations, forming thick layers of pure ice known as ice lenses. As these lenses grow, they exert massive pressure, lifting the ground in a process known as frost heave, which can crack roads and foundations. Conversely, when these lenses melt, the soil loses its structural integrity, leading to thaw settlement and potentially catastrophic landslides.
Existing models for these processes are often limited to one dimension or rely on "phenomenological" rules—essentially empirical observations rather than fundamental physics. These older models struggle to explain the initiation of new ice lenses or the mechanical failure of the soil skeleton required to accommodate ice growth. This research addresses these gaps by treating ice lens formation as a physics-driven fracturing process.
Methodology: The Multi-Phase-Field Approach
The core innovation of this research is the multi-phase-field microporomechanics model. Developed by Hyoung Suk Suh and WaiChing Sun, the model utilizes two distinct "phase fields" to track the state of the material without needing to explicitly define every complex boundary:
- The Fluid Phase Field ($c$): This tracks whether the water in the pores is liquid ($c=1$) or frozen ($c=0$). It is governed by the Allen-Cahn equation, a standard tool for modeling phase transitions.
- The Damage Phase Field ($d$): This tracks the "cracking" or damage of the soil skeleton. It treats the formation of an ice lens as a hydraulic fracture where the growing ice ($d=1$) literally pries the soil apart.
By coupling these phase fields with the laws of thermodynamics, fluid diffusion (Darcy’s law), and solid mechanics, the simulation captures the "cryo-suction" effect. In this phenomenon, freezing ice acts as a pump, pulling liquid water toward the freezing front to feed the growing lens. The system was implemented using the FEniCS finite element library and solved using a staggered approach for numerical stability.
Verification and Validation
The model was validated against established physical benchmarks and experimental data rather than traditional machine-learning datasets:
- Terzaghi’s 1D Consolidation: Used to verify the fundamental hydro-mechanical coupling of the soil skeleton.
- Experimental Freezing Data: The model was validated against physical experiments involving water-saturated copper foams to observe the movement of the freezing front.
- Heterogeneous Soil Profiles: Synthetic profiles with random porosity were generated to test how the model handles the non-uniform reality of natural ground.
Key Insights and Results
Ice Lenses as Fractures
By treating the ice lens as a segregated ice fracture rather than just a frozen pore, the model captures the loss of soil shear strength and the abrupt expansion that occurs when a lens initiates. The simulations showed that ice lenses are not random; they tend to form in regions with higher initial porosity and lower fracture resistance.
Coupled Physics and Latent Heat
The model successfully replicated the "temperature plateau" observed in nature, where soil temperature remains constant at the freezing point while water undergoes phase transition. In simulations of clayey soil, the model produced a distinct "sawtooth" pattern in ground displacement, where each jump corresponds to the initiation of a new, discrete ice lens deeper in the soil.
Thaw Settlement
The model also successfully simulated the abrupt settlement that occurs during thawing. It demonstrated how the melting of segregated ice leads to a rapid loss of volume and soil stability, providing a mechanistic explanation for thaw-induced infrastructure failure.
Conclusion and Future Work
While the multi-phase-field model provides a robust physical framework for understanding frost heave, it is computationally intensive. Current simulations are primarily limited to 1D and 2D due to the fine mesh required to capture crack gradients.
Furthermore, the current model assumes brittle soil behavior. In reality, frozen soil and ice can exhibit ductility, flowing and deforming slowly over time (creep). Future work will focus on incorporating plasticity and creep into the solid skeleton and the ice itself to improve long-term predictions for permafrost engineering in a changing climate.
Summary updated