California Bearing Ratio (CBR): Pavement Subgrade Assessment

A typical flexible pavement structure consists of four layers: the subgrade (the prepared natural soil or fill material), the subbase course (typically crushed stone or recycled asphalt), the base course (higher quality crushed stone or stabilized material), and the asphalt surface course. Each layer plays a distinct role in distributing traffic loads downward through the pavement structure. When a truck wheel applies a concentrated load to the pavement surface—say 80 kN from a single tire—that load must be distributed through progressively larger areas as it moves down through the layers. The asphalt surface layer is designed to resist rutting and cracking. The base course distributes loads more widely and provides structural support. The subbase course further distributes loads to larger areas. The subgrade is the foundation—it must support all the load transferred to it without excessive settlement or deformation.

The subgrade strength fundamentally determines how much material must be placed above it to adequately support traffic. Imagine two identical road projects in different locations. Project A has a subgrade with a CBR of 40% (strong, well-drained granular soil). Project B has a subgrade with a CBR of 3% (weak clay soil with poor drainage). Using standard AASHTO design procedures with identical traffic loads and design standards, Project A might need 250 mm of asphalt and base course combined. Project B might need 450 mm of the same materials—80% more thickness at significant cost. The difference comes entirely from subgrade strength.

Weak subgrades create multiple problems. They settle more under load, causing differential settlement that leads to pavement cracking and rutting. They are susceptible to moisture changes—wet clay expands; drying clay shrinks. This cyclic movement breaks up the pavement above. Weak subgrades are particularly problematic in areas with seasonal water table fluctuations. In wet seasons, the water table rises and saturates the clay subgrade, dramatically reducing its strength. In dry seasons, the soil dries and shrinks, creating voids and settlement. This yearly cycle of saturation and drying can degrade a poorly designed pavement in just 5-10 years. Meanwhile, a properly designed pavement over a strong subgrade might last 20+ years with minimal distress. The engineer's job is to measure subgrade strength accurately, use that data in rigorous design calculations, and specify the correct pavement thickness. CBR testing provides this critical measurement.

The California Bearing Ratio test was developed in the 1930s by the California Department of Transportation (Caltrans) as a method to correlate soil strength with pavement performance. At the time, pavement design was largely empirical—engineers built roads, observed what happened, and adjusted designs based on performance. There was no standardized way to quickly measure soil bearing capacity in the laboratory. Caltrans engineers needed a simple, repeatable test that could be performed at many project sites to guide pavement design. The CBR test was born from practical necessity.

The test is elegantly simple in concept. A standard penetrometer (a rod with a 50 mm diameter flat tip) is pushed into a soil sample at a constant rate. The force required to push the penetrometer into the soil is measured at specific depths. The higher the force, the stronger the soil. By comparing this force to the force required to push the penetrometer into a standard material (a well-graded crushed stone aggregate), a ratio is calculated. This ratio, expressed as a percentage, is the CBR value.

What made CBR successful is its strong correlation with pavement performance. After decades of use, enormous amounts of field data have been collected showing how pavements with different designs perform over time. This data has been analyzed to develop design curves that relate CBR value, traffic load, and pavement thickness to expected service life. These relationships are now encoded in design standards used worldwide (AASHTO, British Standards, Austroads, and others). The test has remained largely unchanged since the 1930s because it works so well. While more sophisticated tests exist (California Bearing Ratio Extended, Triaxial tests, Resilient Modulus tests), CBR remains the standard for routine pavement design work because it is inexpensive, quick, and provides reliable results for practical engineering decisions.

The CBR test begins with soil sampling and preparation. Soil is obtained from the project site or proposed borrow source using standard boring and sampling techniques. The sample is typically taken at the depth where the subgrade will be formed—usually the top 300-500 mm of the natural soil after removal of unsuitable surface material. The sample must be representative of the actual soil that will be in place. If a project spans an area with varying soil types, multiple samples are taken at different locations. Samples are also taken at different depths to identify weak layers that might control design.

Once the field sample is obtained, it is brought to the laboratory where moisture content and grain size analysis are performed. The engineer needs to understand the soil's composition—what percentage is gravel, sand, silt, and clay. This helps identify whether the soil is a clayey sand, silty gravel, clay, or other classification. The soil is then tested using the Proctor compaction test to determine the relationship between moisture content and compaction density. The standard Proctor test (lighter compaction effort, simulating light construction equipment) or modified Proctor test (heavier effort, simulating modern heavy compaction equipment) is selected based on the project's specifications. The Proctor test produces a curve showing that at very low moisture content, the soil doesn't compact well (dry soil is friable and won't bond together). As moisture is added, the soil compacts better because moisture acts as a lubricant, allowing soil particles to move closer together. Eventually, at optimum moisture content, the soil reaches maximum density. Adding more moisture actually reduces density because the water occupies space that soil particles could occupy. The Proctor test identifies this optimum moisture content.

For the CBR test, a soil sample is prepared at the optimum moisture content (or sometimes at a slightly wetter condition to represent field conditions after rainfall). The sample is compacted into a standard mold—a cylinder 152.4 mm in diameter and 177.8 mm tall, with a capacity of about 2100 cm³. The compaction is performed using the same effort as the Proctor test that was selected (standard or modified). For highway subgrades, modified Proctor compaction is typical, representing field compaction by heavy equipment. The sample is compacted in three equal layers, with each layer receiving the specified number of blows from a standard 4.54 kg hammer dropped from a specific height. This systematic compaction ensures a reproducible, consistent sample that represents how the soil will be compacted in the field.

After compaction, the sample in the mold is carefully removed from the compaction apparatus and weighed. The sample density is calculated and recorded. A surcharge weight (a disk weight sitting on top of the sample) may be applied to simulate the weight of pavement and traffic loads pressing down on the subgrade in service. This surcharge is typically 50 mm of steel weights, equivalent to about 2.3 kPa. The purpose of the surcharge is to create a stress condition in the sample during soaking that is similar to field conditions.

After initial compaction and measurement, the CBR sample is placed in a soaking apparatus—a basin with water maintained at a relatively constant level. The sample soaks for 96 hours (four days) while fully submerged. This extended soaking simulates moisture conditions after the pavement has been in service for several years and the subgrade has become fully saturated through infiltration of rainwater and groundwater rise. The 96-hour soaking period is based on extensive empirical data showing that this duration reliably saturates most soil samples to a depth that is representative of field conditions.

During the soaking period, the sample is monitored for swell—vertical expansion as water is absorbed into the soil. Swell is particularly important for clay soils. A dense, well-graded sandy gravel might show zero swell—water fills pores but doesn't cause volume change. A clay soil, especially one with high plasticity (clay minerals with strong affinity for water), can expand significantly as water is absorbed. Fine-grained clay can expand 2-5% or more as it hydrates. This swell data is critical for identifying problem soils. A high swell value (greater than 3-4%) indicates a soil prone to expansion and contraction with moisture changes. Such soils are risky for long-term pavement performance because the pavement above will be subjected to differential settlement as the clay expands when wet and contracts when dry. This cyclic movement induces cracking in the pavement.

The surcharge weight applied to the top of the sample during soaking restrains some of this swell. The constraint approximates the restrained conditions in the field where pavement and overlying soil restrict vertical movement. However, the swell measurement still reflects the soil's inherent propensity to absorb water and expand. The combination of high swell potential and compressibility creates particularly problematic soils for pavement. If a subgrade is identified as having high swell potential (>3-5%), special treatments might be needed: the soil might be treated with lime or cement to reduce its plasticity and swell potential, the drainage system might be improved to minimize water infiltration, or the pavement thickness might be increased to distribute surface loads more broadly and reduce differential settlements.

After the 96-hour soaking period, the CBR sample is removed from the soaking basin and allowed to drain for a standardized time (typically 15 minutes). The sample is then placed in a load frame—a mechanical testing apparatus equipped with a hydraulic jack or mechanical screw that applies force and a load cell that measures the applied force with high precision. A penetrometer (a steel rod with a flat, circular tip 50 mm in diameter) is positioned against the top surface of the sample. The load frame is adjusted so that the penetrometer just makes contact with the sample surface under negligible load.

The load frame then advances the penetrometer into the sample at a constant rate of 1.27 mm per minute (specifically 0.05 inches per minute in the original standard). As the penetrometer advances, it encounters resistance from the soil. Stronger, more dense soil offers more resistance; weaker soil offers less. The load cell continuously measures the force required to maintain the constant advance rate. This force is recorded as a function of penetration depth. Typically, forces are recorded at penetration depths of 0.5 mm, 1.0 mm, 1.5 mm, 2.0 mm, 2.5 mm, and then continuing at 0.5 mm intervals up to 7.5 mm or more.

The test is continued until either a total penetration of 7.5 mm is reached or the force reaches a maximum value and begins to stabilize. At standardized penetration depths of 2.5 mm and 5.0 mm, the measured force is compared to standard reference forces for penetration of a high-quality crushed stone aggregate. These reference forces were established by testing a standard, well-graded gravel and are specified in the standards. The reference force for 2.5 mm penetration is 132.4 kN, and the reference force for 5.0 mm penetration is 264.8 kN. The CBR value is calculated as: CBR (%) = (Measured Force / Reference Force) × 100.

For example, if penetration of the soil sample at 2.5 mm depth requires 39.7 kN of force, the CBR value would be (39.7 / 132.4) × 100 = 30%. This means the soil is 30% as strong as the standard reference aggregate at 2.5 mm penetration. Typically, CBR at 2.5 mm penetration is reported as the primary result because it often reflects conditions of maximum strain difference between soil and the penetrometer (at 5 mm penetration, the soil around the penetrometer has deformed and rearranged, making the measurement less representative of actual bearing capacity). However, both values are recorded and reported.

CBR values range from less than 1% for very soft clays to over 100% for well-cemented gravels and crushed rock. Most soils encountered in routine pavement work fall in a much narrower range. Very poor subgrades (plastic clays in poor drainage conditions, saturated silts) have CBR values of 1-5%. Poor subgrades (low-plasticity clays, fine sands) have CBR values of 5-10%. Fair subgrades (clayey sands, sandy clays) have CBR values of 10-20%. Good subgrades (well-graded sands and gravels) have CBR values of 20-40%. Excellent subgrades (dense well-graded gravels, crushed stone) have CBR values of 40-80%. Very strong materials like quarry rock or well-compacted crushed limestone can exceed 80% CBR.

For typical highway projects, a representative highway subgrade has a CBR in the range of 3-15%. A well-designed project will have subgrades compacted to a specified density (often 95% of maximum density from modified Proctor) and tested to verify that the specified CBR is achieved. If field testing during construction shows that the subgrade CBR is significantly lower than predicted from laboratory testing, design adjustments might be needed.

The CBR value is strongly influenced by several factors. Moisture content is critical—the same soil tested at optimum moisture might have CBR of 20%, but at saturation (after 96-hour soaking), the same soil might have CBR of only 8%. This is why the saturation condition in the test is so important—it represents worst-case field conditions. Compaction density also strongly influences CBR. The same soil compacted to 90% of maximum density might have CBR of 6%, but at 98% density, the same soil might have CBR of 15%. This is why field compaction specification and verification are crucial. Soil type and gradation matter enormously. A well-graded gravel (containing a range of particle sizes that fit together efficiently) has much higher CBR than a uniform sand (containing particles of similar size that don't pack as densely). Plasticity matters too—the more plastic clay present, generally the lower the CBR. Fine-grained soils (clays and silts) almost always have lower CBR than coarse-grained soils (sands and gravels).

Pavement design standards translate CBR values into pavement thickness recommendations. The most widely used standard is AASHTO (American Association of State Highway and Transportation Officials) Mechanistic-Empirical Pavement Design Guide, though traditional AASHTO Empirical Design, TRB (Transportation Research Board) methods, and Austroads in Australia are also common. These design methods typically use the CBR value as an input along with several other parameters: design traffic (expressed as Equivalent Single Axle Loads or ESALs over the design period, typically 10-20 years), environmental factors (freeze-thaw potential, moisture conditions, temperature), design reliability (the confidence level that the pavement will perform adequately—higher reliability requires thicker pavements), and performance criteria (acceptable levels of rutting and cracking).

The design process involves selecting a trial pavement structure (thickness and material type of each layer) and calculating whether that structure will provide adequate performance for the design traffic and subgrade conditions. If the calculated performance is inadequate, the designer increases thickness or improves material quality and recalculates. Iterations continue until an adequate design is found. CBR of the subgrade is one of the first inputs entered into this process. A higher CBR subgrade requires less total thickness because the subgrade provides more inherent strength. A lower CBR subgrade requires greater thickness.

As a practical example, consider a regional highway expected to carry 1 million ESALs over a 20-year design life. The designer obtains a CBR value of 5% from subgrade testing. Using AASHTO design, the calculation shows that 200 mm of asphalt concrete over 300 mm of crushed stone base course (500 mm total) is required. This design is expected to keep rutting below 12 mm and cracking below 20% over the 20-year period. Now consider the same project with a different location where the subgrade CBR is 25%. Using the same AASHTO method with identical traffic and design life, the calculation shows that only 120 mm of asphalt over 150 mm of base course (270 mm total) is adequate. The thinner design for the stronger subgrade saves significant money per kilometer of roadway.

Designers must also consider whether the CBR value represents the most critical condition. If multiple samples are tested at different depths and locations, the lowest CBR value typically controls the design because it represents the weakest layer in the subgrade profile. If a boring shows a weak clay layer at 1.5 m depth beneath an otherwise adequate granular subgrade, the weak layer might not control design if it's deep enough that it receives reduced stress from surface loading. However, a weak layer within the top 1-1.5 m depth (within the zone of significant stress influence from surface traffic) must be considered in design.

CBR testing sometimes reveals problem soils that cannot be adequately supported by pavement thickness alone. The primary concern is high swell potential. A soil with swell greater than 5% is considered problematic. Such soils have very high clay content and clay minerals (particularly montmorillonite) with strong affinity for water. They expand dramatically when wetted and shrink when dried. This cyclic volume change induces stress in the pavement above, leading to cracking and accelerated deterioration. For projects on such soils, several treatment options exist.

Lime stabilization is a common approach. Hydrated lime (calcium hydroxide) is mixed into the clay soil, typically at rates of 3-8% by weight. The lime reacts with the clay minerals in a process called pozzolanic reaction, converting the clay structure and reducing both plasticity and swell potential. Lime-treated clay becomes less expansive and more granular in behavior. The stabilized material often develops sufficient strength to function as a base course rather than just subgrade, adding structural benefit beyond just reducing swell. A subgrade treated with 5% lime might reduce swell from 8% to 2% and increase CBR from 3% to 8%—a dramatic improvement. However, lime stabilization is not permanent. In very wet climates or with poor drainage, lime-stabilized clay can eventually reabsorb water and regain some of its swelling tendency. Therefore, lime stabilization is most effective when combined with improved drainage design.

Cement stabilization is an alternative that produces more durable results. Portland cement mixed into soil at 3-6% by weight causes the soil to harden and develop higher strength. Cement-stabilized clay becomes cemented soil, almost like a weak concrete. CBR values can increase dramatically—from 3% to 20% or more—and swell becomes negligible. However, cement stabilization is more expensive than lime stabilization and requires more careful quality control during construction. The stabilized layer must be properly cured (kept moist for several days to allow cement hydration) and compacted to achieve design properties.

For very problematic soils, excavation and replacement might be the most cost-effective solution. If a weak, high-swell clay layer is less than 0.5-1.0 m thick, it might be economical to excavate and remove the poor soil and replace it with good-quality granular material imported from a borrow source. The replacement material is selected to have high CBR (typically 40% or greater) and minimal swell. The additional cost of excavation, hauling, and compaction must be evaluated against the cost of stabilization or additional pavement thickness to determine the most economical approach for each project.

Improved drainage is often essential regardless of which treatment is selected. If poor drainage allows water to accumulate at the subgrade level, even good-quality soil can experience reduced strength and accelerated deterioration. Proper drainage design—including permeable base courses, subdrains, and sloped gravel shoulders—ensures that water does not accumulate in the subgrade.

Laboratory CBR testing predicts what a soil will do under specified conditions, but field conditions during construction rarely match laboratory conditions exactly. Therefore, systematic field verification testing is essential. As the subgrade is being prepared and compacted during construction, samples of the compacted material are tested to verify that the specified density and CBR are being achieved. Multiple testing methods are used in the field.

Nuclear density gauges use gamma ray technology to measure soil density in place without removing samples. A probe is inserted into a borehole drilled to the desired depth, and gamma rays are used to measure the density of soil surrounding the probe. This gives a quick, non-destructive measurement of whether the soil is compacted to the specified density. Multiple readings are taken across the project to verify uniform compaction. The results are compared to the maximum density determined in the Proctor test. Field compaction is typically required to be 95-98% of maximum density, depending on project specifications. If field density is below the minimum specified, additional compaction equipment is deployed and the area is re-worked until the specification is met.

Sand cone testing is a more traditional method where a hole is dug in the compacted subgrade, the soil from the hole is carefully collected, and its volume and moisture content are measured in the laboratory. The field density is calculated from the mass and volume. Sand cone testing is more labor-intensive than nuclear testing and provides results for one point rather than the average over a larger area, but it is still widely used, particularly for verification and dispute resolution when nuclear gauge results are questioned.

Infrared moisture meters or laboratory oven-drying can verify that soil moisture content in the field matches the specification. If the subgrade is compacted too dry, it might not achieve design density. If too wet, compaction equipment can displace the soil rather than properly consolidating it. Moisture verification is particularly important in arid regions where evaporation during construction can cause moisture content to drop below optimum.

During construction, if field verification testing shows that either density or moisture is not meeting specifications, corrective action must be taken. This typically means adding water (if dry) or allowing additional time for evaporation (if wet), followed by re-working with compaction equipment and re-testing. A few days of schedule delay during construction is far preferable to having a poorly compacted subgrade that fails prematurely after the project is open to traffic.

After the base course and surface course are placed, the project is monitored for performance during the initial years. Pavement condition surveys document any cracking, rutting, or other distress. If distress appears earlier than predicted by design calculations, it might indicate that field subgrade conditions were worse than laboratory CBR testing suggested, or that traffic loads exceeded design assumptions. Performance monitoring data feeds back into design refinement for future projects on similar soils. Over decades of operation, performance data from thousands of projects has validated and refined the relationships between CBR and pavement longevity.

While CBR testing is the industry standard, it has recognized limitations. The test measures bearing capacity at a relatively slow penetration rate under static load. In reality, pavements experience dynamic, moving loads from vehicle wheels. Modern testing methods like Resilient Modulus testing measure the elastic behavior of soil under cyclic loading, which better represents actual field conditions. However, resilient modulus testing requires expensive equipment and specialized expertise, so it is not standard for routine design work.

CBR results can be affected by sample preparation and handling. Clay soils are particularly sensitive—the same soil can show different CBR values depending on compaction technique, moisture content variations, and how long the sample was stored before testing. This is why quality control in the laboratory is essential and why field verification is critical. Professional laboratories participating in proficiency testing programs and following strict quality protocols produce more reliable results than casual testing.

The saturation soaking condition (96 hours full submersion) might not represent all field conditions. In arid regions where groundwater is deep and rainfall is minimal, the subgrade might never reach full saturation. In such cases, using a saturated CBR value in design is overly conservative—it results in thicker pavements than necessary. Conversely, in coastal areas with high water tables or in freeze-thaw regions where moisture accumulates during spring thaw, saturation conditions might be encountered regularly, making the standard test appropriate or even unconservative.

CBR does not account for the stress history of the soil. The test applies increasing stress in one direction (vertical penetration) but does not account for how three-dimensional stresses develop under actual traffic loading. More sophisticated models using finite element analysis and multi-dimensional material properties provide greater insight but are computationally intensive and require more expertise. For routine design, simpler relationships based on decades of CBR-based design experience remain practical and adequate.

Regional variations in soil and climate mean that while CBR-based design works well on average, individual projects might deviate from the average. An exceptionally well-drained sandy gravel might perform better than predicted. A poorly drained clay with high swell might perform worse. A pragmatic approach combines CBR testing with site-specific knowledge, field observation during construction, and monitoring of performance.