Artificial Frozen Soil Shear Testing: Laboratory Methods, Strength Parameters, and Engineering Appli

Artificial frozen soil shear testing is a specialized laboratory procedure used to determine the shear strength parameters of ground that has been artificially frozen for construction projects such as mine shafts, metro tunnels, and cross-passages. By measuring cohesion and internal friction angle under controlled sub-zero temperatures, engineers can verify frozen wall thickness, evaluate short-term and long-term stability, and optimize artificial ground freezing designs. This article explains the sample types, test items, testing methods, applicable standards, and practical applications of shear testing on artificially frozen soils.

Technical Overview

Artificial ground freezing (AGF) is a ground improvement technique in which a coolant, typically calcium chloride brine or liquid nitrogen, is circulated through freeze pipes installed in water-bearing soil or weak rock. Pore water solidifies into ice, cementing soil particles together and producing a strong, impermeable frozen body. Because this frozen body acts as both a temporary structural element and a groundwater cutoff, its mechanical properties, particularly shear strength, directly govern the safety and economy of the excavation it protects.

Artificial frozen soil shear testing measures the shear resistance of frozen specimens under controlled temperature, normal stress, and strain rate conditions. Results are expressed through the Mohr-Coulomb failure criterion, in which shear strength equals apparent cohesion plus the frictional component mobilized by normal stress. Unlike unfrozen soil, frozen soil strength is strongly temperature dependent, time dependent due to creep of the ice matrix, and sensitive to water content, salinity, and loading rate. Dedicated low-temperature laboratory testing is therefore essential for reliable frozen wall design.

• Strength increases significantly as temperature decreases, roughly in proportion to the fraction of unfrozen water that converts to ice.
• Cohesion rises dramatically upon freezing because ice acts as a bonding agent between particles.
• Frozen soil behaves viscoplastically, so measured peak strength depends strongly on the applied strain rate.
• Confining pressure can cause pressure melting at ice-particle contacts, producing non-linear strength envelopes at higher stresses.

Testing Samples

Remolded or Reconstituted Specimens

Most routine programs use remolded specimens prepared at the in-situ dry density and natural water content of the stratum under investigation. Soil is compacted in layers into molds, adjusted to the target moisture condition, and then frozen under controlled conditions. Remolding allows numerous parallel specimens to be produced, which is necessary because a complete strength envelope requires testing at multiple temperatures and stress levels.

Undisturbed Specimens

When natural soil structure, layering, or ice fabric must be preserved, undisturbed samples are recovered using thin-walled samplers from exploration boreholes or by block sampling from trial pits. Trimming and handling are performed in a low-temperature environment to prevent partial thawing and structural disturbance. Undisturbed testing is preferred for stratified, varved, or fissured deposits where fabric strongly influences strength.

Specimen Geometry and Conditioning

Common geometries include cylindrical specimens of 39.1 mm by 80 mm or 61.8 mm by 125 mm for triaxial testing, and cylindrical or cubic specimens of 50 mm to 100 mm for direct shear testing. After trimming, specimens are placed in a cryostat or environmental chamber and conditioned at the target temperature for at least 24 hours, and often 24 to 48 hours depending on specimen size, so that the internal temperature is uniform before loading begins.

Testing Items

A comprehensive artificial frozen soil shear testing program typically includes the following items, each performed at one or more design temperatures such as -5, -10, -15, and -20 degrees Celsius:

• Instantaneous direct shear strength at four or more normal stress levels
• Triaxial shear strength under unconsolidated-undrained or consolidated-undrained conditions at sub-zero temperatures
• Determination of apparent cohesion and internal friction angle at each temperature
• Complete stress-strain curves, failure strain, and secant deformation modulus
• Strain-rate sensitivity testing and long-term or creep-limited strength evaluation
• Influence of dry density, water content, and pore water salinity on shear strength
• Comparison of strength before freezing, after freezing, and after thawing where excavation sequencing requires it

Testing Methods

Direct Shear Test

A temperature-controlled direct shear apparatus houses the frozen specimen inside a cooling chamber or refrigerated shear box. The specimen is first brought to the target temperature, a constant normal stress is applied, and a horizontal shear load is then imposed at a controlled displacement rate, commonly in the range of 0.5 to 2.0 mm per minute for instantaneous strength determination. Peak and residual shear stresses are recorded, and the test is repeated at progressively higher normal stresses. Plotting peak shear stress against normal stress yields a failure envelope from which cohesion and friction angle are derived.

Triaxial Shear Test

Cryogenic triaxial cells permit testing under cell pressures representative of the depth and frozen wall geometry, typically ranging from 0.5 MPa to several MPa for deep mine shafts. Specimens are sealed in membranes, brought to test temperature, consolidated if the test protocol requires it, and sheared under strain control at rates selected to match either short-term loading or creep-dominated long-term conditions. Triaxial testing captures the confining pressure dependence of frozen strength and provides more realistic parameters for three-dimensional stress states than direct shear.

Temperature Control and Instrumentation

Accurate results depend on strict thermal control. Chamber and specimen temperatures are monitored with thermocouples or PT100 sensors embedded in dummy specimens, with fluctuations typically held within plus or minus 0.5 degrees Celsius or tighter. Load cells, pressure transducers, and displacement sensors must be calibrated for operation at low temperature, and a computerized data acquisition system records load, deformation, temperature, and pore or cell pressure throughout each test.

Applicable Standards

In China, testing is generally performed with reference to the MT/T 593 series, the coal industry standards covering physical and mechanical testing of artificially frozen soils, together with project-specific technical specifications. Internationally, ASTM D3080 for direct shear and ASTM D4767 for consolidated-undrained triaxial testing are commonly adapted to sub-zero conditions. Where freezing is used in transportation or hydraulic projects, national soil testing standards are supplemented by project design criteria for frozen ground structures.

Applications

Shear parameters obtained from artificial frozen soil testing are applied throughout the design and construction verification of frozen ground structures:

• Vertical mine shaft sinking through water-bearing alluvium, where frozen wall thickness and safety factors are checked against measured cohesion and friction angle
• Metro tunnel cross-passage and bracket
• Ground stabilization for shield tunnel launch and reception shafts in soft, saturated ground
• Temporary frozen curtains supporting deep excavations adjacent to sensitive structures and utilities
• Back analysis of frozen wall performance during excavation and prediction of long-term strength under sustained load
• Verification that ground freezing will achieve design strength in saline, organic, or unusually wet soils where freezing effectiveness is uncertain

FAQ

What test temperatures should be selected?

Test temperatures should bracket the expected average temperature of the frozen wall, typically one or two degrees colder than the design frost front temperature. Common programs test at -5, -10, -15, and -20 degrees Celsius so that strength can be interpolated across the thermal gradient that actually exists between the freeze pipes and the excavation face.

How long must specimens be conditioned before testing?

Specimens are conditioned until their core temperature stabilizes at the target value, which normally requires at least 24 hours and frequently 24 to 48 hours for larger specimens. Testing before full thermal equilibrium is reached leads to unconservative strength values.

How much does freezing increase shear strength?

The increase is substantial. Apparent cohesion of a soft saturated clay may rise from tens of kilopascals to several megapascals when frozen, while the friction angle changes comparatively little. The exact gain depends on temperature, ice content, and soil type, which is why laboratory measurement cannot be replaced by empirical assumptions.

Why does strain rate matter so much in frozen soil testing?

Ice and frozen soils exhibit pronounced viscoplastic behavior. Faster loading produces higher measured peak strength, whereas sustained loads cause creep and strength reduction over time. Design of frozen walls that must stand for weeks or months therefore relies on rate-reduced or creep-limited strength rather than the instantaneous peak values.

How do remolded and undisturbed samples compare?

Remolded specimens provide reproducible data suitable for parametric studies and are acceptable when soil is relatively homogeneous. Undisturbed specimens preserve natural fabric and should be used whenever layering, fissures, or previous freeze-thaw history could materially affect shear behavior.

Can the same apparatus test unfrozen and thawed soil?

Many low-temperature shear and triaxial systems operate across a range from below -30 to above ambient temperature, allowing the same equipment to characterize the unfrozen in-situ soil and the post-thaw residual strength, both of which are needed for construction staging and permanent works design.