Abstract
Frost heave occurs when ground temperature decreases under $0^{\circ}C$ and soil volume expands, which causes roadway and buried pipe line failure due to differential heaving. There are several models to predict volumetric strain caused by frost heave, but they often require expertise in complex numerical analysis. This study presents an evaluation method to predict volumetric strain caused by frost heave using fundamental physical properties of soils. The evaluation method is assessed and validated through comparison with experimental results.
Introduction
As infrastructure projects extend into extreme environments like Siberia and the Antarctic, engineers must contend with the destructive force of frost heave. When soil temperatures drop below freezing, pore water turns to ice, expanding in volume. This expansion leads to "differential heaving," which can buckle roadways and snap buried pipelines. While sophisticated numerical models exist, they are often "black boxes" requiring advanced geotechnical expertise. This paper proposes a simplified calculation method based on standard soil properties—such as water content and porosity—to bridge the gap between high-level research and practical field application.
Theoretical Framework: Multi-Phase Soil Modeling
The proposed method models soil as a multi-phase system to track volume changes accurately. In its initial unsaturated state, the soil is treated as a four-phase system consisting of air, water, soil grains, and any added anti-frost materials.
Upon freezing, the system transitions into a five-phase model. This model explicitly accounts for the fact that not all water freezes simultaneously; it distinguishes between frozen water (ice) and unfrozen water ($w_u$), which remains liquid even below $0^{\circ}C$.
The calculation framework establishes two predictive boundaries:
- Lower Bound Expansion: Assumes that as ice forms, it first fills existing air voids. Expansion only occurs once the ice volume exceeds the available air space.
- Upper Bound Expansion: Assumes air is incompressible or that ice formation immediately contributes to total volume increase, providing a conservative estimate for engineering safety.
Experimental Validation
The model was validated using a Silty-Sand (SM) soil in a "closed-system" frost heave cell. Samples were compacted to 90% and frozen to -12.5°C in a specialized chamber. Key findings include:
- Unfrozen Water Sensitivity: The unfrozen water content ($w_u$) stabilized at approximately 11% at -6°C for the tested soil. The researchers found that ignoring this value leads to a significant overestimation of frost heave.
- Moisture Impact: Frost heave is highly sensitive to initial moisture. For the SM soil tested, a 3% increase in moisture content above the optimum led to a 1.3% increase in the frost heave ratio.
- Predictive Accuracy: Measured volumetric strains (ranging from 0.4% to 2.8%) consistently fell within the bounds predicted by the model, particularly when using the measured $w_u$ value to bracket the results.
Conclusion
This study provides a practical tool for estimating frost heave using basic laboratory data like wet unit weight, porosity, and specific gravity. By utilizing the "Upper Bound" calculation with known unfrozen water content, engineers can obtain reliable, conservative estimates for infrastructure design in freezing environments without requiring specialized numerical simulation software. Future work will focus on expanding the model to include fine-grained clays and "open-system" conditions where groundwater recharge can lead to even greater expansion.
Summary updated