The asphalt freeze-thaw splitting test is a laboratory method used to evaluate the resistance of compacted asphalt mixtures to moisture-induced damage under freeze-thaw conditioning, and it is widely applied in mix design verification and quality control for highway and airport pavements. By comparing the indirect tensile strength of unconditioned specimens with specimens subjected to vacuum saturation, freezing, and warm-water immersion, the test yields the Tensile Strength Ratio (TSR), a key indicator of water stability. This article explains the test principle, applicable standards such as JTG E20 (T 0729), AASHTO T 283, and ASTM D4867, specimen requirements, step-by-step procedure, typical acceptance criteria, and practical engineering applications.
Technical Overview
The asphalt freeze-thaw splitting test, also known as the freeze-thaw indirect tensile test or the modified Lottman test, is a standardized method for assessing the moisture susceptibility of compacted bituminous mixtures. The test accelerates moisture damage in the laboratory by combining vacuum water saturation, a low-temperature freezing cycle, and a warm-water soaking cycle, after which the indirect tensile (splitting) strength of the conditioned specimens is compared with that of dry, unconditioned specimens. The result is expressed as the Tensile Strength Ratio (TSR), which quantifies how much strength the mixture retains after severe moisture and temperature exposure.
The test is standardized internationally by AASHTO T 283 (Resistance of Compacted Asphalt Mixtures to Moisture-Induced Damage) and ASTM D4867 (Effect of Moisture on Asphalt Concrete Paving Mixtures). In China, the equivalent procedure is specified in the highway engineering test standard JTG E20-2011 as T 0729-2011 (Freeze-Thaw Splitting Test for Bituminous Mixtures), and the acceptance thresholds are defined in the construction specification JTG F40-2004.
Moisture damage in asphalt pavements occurs through two principal mechanisms: loss of adhesion at the aggregate-binder interface (stripping) and loss of cohesion within the binder film. Freeze-thaw action intensifies both mechanisms because water trapped in the air voids expands upon freezing, generating hydraulic pressure that physically disrupts the binder-aggregate bond. Mixtures with poor water stability typically exhibit raveling, potholes, and crumbling of the surface layer, particularly in seasonally frozen or high-rainfall regions.
Testing Samples
Marshall Compacted Specimens
The standard specimen for the Chinese procedure (JTG E20 T 0729) is a Marshall cylinder with a diameter of 101.6 mm and a height of 63.5 ± 1.3 mm, compacted with a standard Marshall hammer, typically 50 blows per face. The target air void content is 7.0 ± 1.0%, which represents a realistic in-service void level at which moisture can penetrate the mixture. Specimens must be compacted, cored, or sawn so that the two end faces are parallel and perpendicular to the axis.
Gyratory Compacted Specimens
Under AASHTO T 283 and ASTM D4867, specimens may be prepared with a Superpave gyratory compactor (100 mm or 150 mm diameter) or by coring from laboratory-molded slabs, with air voids controlled at 7.0 ± 0.5%. Specimens outside the tolerance band must be discarded, because void content strongly influences the degree of saturation and therefore the test outcome.
Specimen Grouping
A minimum of three specimens (preferably four) is required for each of the two groups, randomly assigned from the same batch:
- Group A (unconditioned reference group): stored dry at room temperature and tested to establish the baseline indirect tensile strength.
- Group B (conditioned group): subjected to vacuum saturation, freezing, and warm-water immersion before strength testing.
Before testing, all specimens are measured for diameter, height, and bulk density, and the air void content is calculated for each specimen. Specimens with abnormal voids, visible defects, or density outliers are excluded so that the two groups are statistically comparable.
Testing Items
The freeze-thaw splitting test determines the following parameters:
- Unconditioned indirect tensile strength RT1 (MPa): the average splitting strength of the dry reference group tested at 25 °C.
- Conditioned indirect tensile strength RT2 (MPa): the average splitting strength of the freeze-thaw conditioned group tested at 25 °C.
- Tensile Strength Ratio (TSR): the ratio of conditioned to unconditioned strength, calculated as TSR = RT2 / RT1 × 100% and reported to the nearest 0.1%.
- Degree of saturation of Group B specimens: verified to fall within the specified range of 70–80% after vacuum saturation.
- Peak load, specimen dimensions, and individual strength values for each specimen, reported together with the group averages.
Common acceptance criteria require a TSR of no less than 75–80% depending on climate zone, traffic level, and binder type; many specifications for polymer-modified asphalt mixtures and heavily loaded pavements specify a minimum of 80%. Mixtures failing the threshold must be redesigned or treated with moisture-damage mitigation measures.
Testing Methods
Conditioning Procedure
The standard freeze-thaw conditioning sequence per JTG E20 T 0729 consists of the following steps:
- Vacuum saturation: Group B specimens are submerged in water and subjected to a partial vacuum so that the degree of saturation reaches 70–80%. The saturation degree is verified by weighing each specimen before and after saturation.
- Freezing: Each saturated specimen is wrapped in a plastic film bag together with approximately 10 mL of added water and placed in a chamber at -18 ± 2 °C for 16 ± 1 hours.
- Thawing and warm-water immersion: The specimen, still in its sealed bag, is transferred to a water bath at 60 ± 0.5 °C for 24 hours.
- Temperature equilibration: Prior to strength testing, both Group A and Group B specimens are immersed in a water bath at 25 ± 0.5 °C for 2 hours to ensure a uniform test temperature.
Splitting Strength Test
Each specimen is placed between two loading strips in a compression testing machine or Marshall stability press fitted with a splitting fixture, with the loading axis aligned along the specimen diameter. The load is applied at a constant rate of 50 mm/min at 25 °C until failure, and the peak load PT is recorded. The indirect tensile strength is calculated as follows:
RT = 0.006287 × PT / h, where RT is the indirect tensile strength in MPa, PT is the peak load in N, and h is the specimen height in mm. This coefficient is derived from the general expression RT = 2PT / (π × d × h) with a standard diameter d of 101.6 mm.
Data Processing and Quality Control
The individual strength values within each group are averaged, and any specimen deviating excessively from the group mean is investigated or discarded according to the standard's outlier rules. The TSR is then computed as the percentage ratio of the conditioned average to the unconditioned average. Critical quality-control points include strict temperature tolerances during freezing and water bathing, accurate control of the saturation degree, verification of specimen air voids before testing, and regular calibration of the load cell, loading strips, and environmental chambers.
Applications
- Mix design verification: confirming that a proposed job mix formula meets the water-stability (TSR) requirement before field production, as part of the Marshall or Superpave design process.
- Construction quality control and acceptance: checking plant-produced mixtures during paving projects, particularly in wet, seasonally frozen, or heavy-rainfall climate zones.
- Evaluation of anti-stripping measures: comparing the TSR of mixtures with and without liquid anti-stripping agents, hydrated lime, polymer-modified binders, or fibers to quantify their effectiveness.
- Aggregate and binder selection: screening moisture-sensitive aggregates (for example, acidic siliceous aggregates with poor binder affinity) and selecting compatible binder grades.
- Performance evaluation for frost regions: assessing durability of surface, intermediate, and base asphalt layers in cold regions where repeated freeze-thaw cycles accelerate moisture damage.
- Research and failure analysis: supporting studies on aging, recycled asphalt pavement (RAP) content, warm-mix additives, and forensic investigation of raveling or stripping distress in existing pavements.
FAQ
What is the difference between the freeze-thaw splitting test and the immersed Marshall test?
Both tests evaluate water stability, but the freeze-thaw splitting test is far more severe. It combines vacuum saturation, freezing at -18 °C for 16 hours, and 60 °C water immersion for 24 hours before measuring splitting strength, whereas the immersed Marshall test only soaks specimens in 60 °C water for 48 hours and measures stability. The freeze-thaw test therefore better simulates moisture damage in cold and wet climates, and it is generally the controlling requirement in such regions.
What TSR value is considered acceptable?
Acceptance thresholds depend on the governing specification, climate zone, and binder type. Typical requirements range from 75% to 80%, with many highway and modified-asphalt applications specifying TSR ≥ 80%. The exact limit should always be taken from the applicable project specification.
Why must the air void content be controlled at approximately 7%?
The 7 ± 1% void level allows water to enter the specimen at a realistic in-service rate during vacuum saturation, ensuring the conditioning is neither too mild (insufficient moisture ingress at very low voids) nor too destructive (uncontrolled damage at high voids). Specimens outside the tolerance produce non-representative TSR values and must be discarded.
How many specimens are required, and how are results reported?
At least three parallel specimens per group are required, with four preferred for better statistical reliability. The report includes individual and average indirect tensile strengths for both groups, the degree of saturation, specimen dimensions and air voids, conditioning details, and the final TSR to the nearest 0.1%.
What are the main causes of a low TSR, and how can it be improved?
Low TSR values usually result from poor adhesion between the binder and moisture-sensitive aggregates, excessive air voids, inadequate compaction, insufficient binder content, or the use of unmodified binders in severe climates. Improvement measures include adding liquid anti-stripping agents or hydrated lime, using polymer-modified or high-grade binders, selecting more alkaline aggregates, optimizing gradation and binder content, ensuring thorough coating during mixing, and improving field compaction to reduce in-place voids.
Which standards govern the test?
The primary references are JTG E20-2011 (T 0729-2011) for Chinese highway practice, AASHTO T 283 for the modified Lottman procedure, and ASTM D4867 for the effect of moisture on asphalt concrete paving mixtures. Accredited third-party laboratories routinely perform the test in accordance with these standards for mix design approval and construction acceptance.
