Chemical Stabilization of Subgrades and Bases

The chemical stabilization materials covered are lime, cement, fly ash, and asphalt; proprietary compaction aiders, polymers, and other organic materials are not discussed, due to limited published information and a tendency for many of these materials to be only locally available. Two soil-index-property charts in the source document (Figure 1, “Selection of stabilizers,” and Figure 2, “Suggested stabilizing admixtures suitable for use with soils”) map liquid limit, plasticity index, and gradation combinations to a recommended stabilizer type; both are complex reference charts and are described rather than reproduced here.

In general terms: lime suits plastic, fine-grained clays; cement suits a broad range of granular-to-moderately-plastic soils; fly ash suits soils with little or no plasticity; and asphalt suits granular, silty, and sandy soils with low plasticity.

Determining lime content is the primary objective of mixture design for lime stabilization. The optimum lime content depends on how the stabilized material will be used and on the soil's constituents; design objectives may involve a reduction in plasticity, construction expediency, or permanent strength/stiffness changes. Mixture preparation, specimen preparation, curing conditions, and testing are the four factors considered in a laboratory testing program. Special testing is required for sulfate-bearing clay to prevent deleterious sulfate-induced heave.

Based on use of 4–6 percent lime for clay soils and 2–4 percent for granular and clay-gravel types. Triaxial and cohesiometer values are based on approximately 18 days of laboratory curing, CBR on 4 days curing (soaked), and R-value on about 2 days curing. Stability values of lime-treated specimens increase markedly with longer or accelerated curing.

Because lime stabilization applications are broad, several mix design methods have been developed. According to TRB (1987), these include the California procedure (Terrel et al. 1979), Eades and Grim procedure, Illinois procedure (Terrel et al. 1979), Oklahoma procedure, South Dakota procedure, Texas procedure (AASHTO T-220), Thompson procedure (Thompson 1970), and Virginia procedure (VTM-11).

As an example, the Texas procedure is summarized below:

1. Based on the grain size and PI data, the lime percentage is selected from a stabilizer-selection chart (Figure 3 in the source document, a complex reference chart not reproduced here).
2. Optimum moisture and maximum dry density of the mixture are determined in accordance with AASHTO T-212 or Tex-113-E.
3. Test specimens 6 inches (15.2 cm) in diameter and 8 inches (20.3 cm) in height are compacted at optimum moisture content to maximum dry density.
4. All specimens are placed in a triaxial cell and cured as follows: (a) cool to room temperature; (b) dry at a temperature not exceeding 60°C (140°F) for about 6 hours, until one-third of the molding moisture is removed; (c) cool for at least 8 hours; (d) subject specimens to water exposure via capillary action for 10 days (AASHTO T-212).
5. The cured specimens are tested in unconfined compression per AASHTO T-212 sections 7 and 8, or Tex-117-E.

The design process flow chart (Figure 4, a complex flow diagram not reproduced here) applies two design criteria — pavement structural behavior and durability requirement — and swell must also be reduced to a satisfactory level for lime-modified soil.

Sulfate Risk Guidelines

Although long-term stiffness benefits of chemically stabilized soil are not considered in most current U.S. design approaches, Qubain et al. (2000) found that lime-treated subgrade soil can be successfully incorporated into pavement design with economic benefit, using one of three approaches: an effective resilient modulus for the lime-treated subgrade; a conservative CBR of 15 to account for lime stabilization; or treating the lime-stabilized subgrade as a subbase with its own structural layer coefficient. Little information is available in the literature documenting the long-term performance of stabilized soils used as permanent foundation materials.

For a cement-stabilized system, the general design considerations are compressive strength, durability, and density; if trial mixes must meet a specified durability requirement, a more detailed testing program is needed to determine cement content. According to Christopher et al. (2006), resistance to sulfate attack differs for cement-treated fine-grained versus coarse-grained soil — granular soil-cements are not susceptible to sulfate attack, but cement treatment is not recommended if fine-grained soil has more than 1% sulfate.

Does not include organic or poorly reacting soils; additional cement may be required for severe exposure conditions such as slope protection.

Table 3 in the source document provides a more granular, two-part lookup grid of average cement requirement (percent by weight) for B and C horizon silty and clayey soils, and separately for sandy soils — each cross-referencing an AASHTO group index or sieve-retention percentage against maximum dry density in 5 lb/ft³ bands. It is a large grid (roughly 30 rows across both parts); a representative sample is condensed below rather than the full grid.

For sandy soils, the same source table ranges from about 5% cement (coarse material, low fines, high density) up to about 12% (finer material, more fines passing 0.05 mm, lower density). Consult the source document directly for exact cell-level values before using this table for a specific design.

The short-cut methods developed by the Portland Cement Association are only applicable to sandy soil, in two variants: Method A, for soils not containing material retained on the No. 4 sieve, and Method B, for soils containing material retained on the No. 4 sieve. The design procedure for Method B is summarized below:

1. Perform a sieve analysis and determine the bulk specific gravity of the soil.
2. Conduct a moisture-density test to determine maximum density and optimum moisture content for a mixture of the soil and Portland cement, per ASTM D 558-57 and AASHTO T 134-57. A source chart (Figure 5) estimates maximum density of the soil-cement mixture; that estimated maximum density, the percentage of material smaller than 0.05 mm, and the percentage retained on the No. 4 sieve are then used with a second source chart (Figure 6) to determine cement content by weight.
3. Use the maximum density from step 2 to determine the indicated cement requirement (Figure 6).
4. Mold compressive-strength test specimens in triplicate at maximum density and optimum moisture content, using the indicated cement factor from step 3.
5. Determine the average compressive strength of the specimens after seven days of moist curing.
6. Determine the minimum allowable compressive strength for the soil-cement mixture from a source chart (Figure 7). If the average strength from step 5 equals or exceeds the minimum allowable strength, the indicated cement factor from step 3 is adequate.

Figures 5, 6, and 7 are density/cement-content and minimum-strength nomographs (complex reference charts) and are described here rather than reproduced.

Several approaches control reflective cracking from cement-treated pavement bases, per Adaska and Luhr (2004): increasing density and reducing moisture content during compaction; using a pre-cracking technique (several passes of a vibratory roller over the cement-stabilized base one to two days after compaction, generating hundreds of micro-cracks so a single transverse crack is prevented); or placing a chip-seal, geotextile, or granular layer between the asphalt surface and the stabilized base to relieve stress concentrations from cracks in the treated base.

Source-document inconsistency: the Portland Cement Association reference is cited with four different years across this technique's documents — “PCA (1992)” in this Design Guidance document's own text, “PCA (1978)” and “PCA (1956)” as footnotes on Tables 2 and 3, and “PCA (1980)” in the QC/QA Procedures document — while the bibliography carries only a single Portland Cement Association entry, dated 1995 (Soil-Cement Construction Handbook). This page cites it uniformly as Portland Cement Association (1995), the bibliography's actual entry, rather than silently picking one of the four in-text years.

Class C fly ash is recommended to stabilize fine-grained plastic soils such as clay, as well as coarse-grained soil. Self-cementing ash hydrates much more rapidly than Portland cement; a 2-hour delay in compaction can decrease maximum density by up to 1.6 kN/m³ (10 pcf) or more, and moisture content also strongly influences compressive strength. To address sulfate-attack risk, fly ash with high sulfate concentrations should be avoided. A laboratory study by Ferguson (1993) recommended a fly ash content of 16% for mixing with subgrade materials to obtain a maximum CBR — this Ferguson (1993) reference is cited only here in the source Design Guidance document and does not appear in this technique's bibliography.

No standard test procedure currently exists for designing materials stabilized with self-cementing ash; an effective procedure can instead be adapted from ASTM C593 (Fly Ash and Other Pozzolans for Use with Lime) and ASTM D1633 (Compressive Strength of Molded Soil-Cement Cylinders):

1. Blend soil, fly ash, and water to make a minimum of five test specimens, with moisture contents between 10% below and 6% above the optimum moisture content for maximum density; specimens should have a height-to-diameter ratio of 1.15.
2. Compact specimens over a wide range of moisture contents, using a compaction time delay of less than two hours and a 102-mm (4.0-inch) diameter by 117-mm (4.625-inch) high mold (standard or modified Proctor compactive energy); alternatively, 50.8-mm (2-inch) diameter by 50.8-mm high specimens save material and time and yield results close to standard Proctor specimens (Oflaherty et al. 1963).
3. Cure test specimens for seven days at 38°C (100°F) per ASTM C593.
4. Determine compressive strength of specimens per ASTM D1633.

For higher degrees of stabilization, additional laboratory tests — including freeze-thaw durability via ASTM C593 — may be needed to assess stabilized-material properties for a specific design procedure.

Army and Air Force (1994) provides an equation to estimate the preliminary quantity of cutback asphalt for subgrade stabilization:

The final cutback or emulsified asphalt content must be verified using Marshall Stability Test results. If a bitumen-stabilized soil does not show increased stability at the designed bituminous material content, the soil's gradation should be modified, or a different bituminous material type should be used.

The use of combination stabilizer materials depends on soil type and stabilizer constituents. A general guideline for selecting stabilizer type and combination appears as Figure 8 in the source document (a complex reference chart, not reproduced here); typically, a combination of stabilizers is recommended when soils have more than 25% passing the No. 200 sieve and a plasticity index greater than 10. The first stabilizer pretreats the soil to alter its properties before the dominant (second) stabilizer is applied, and the second stabilizer's quantity exceeds the first's.

Per Winterkorn and Pamukcu (1990), for lime-fly-ash-and-soil mixtures, lime content varies between 2% and 8%, and fly ash between 8% and 36%; the design ratio of lime to fly ash typically varies from 1:3 to 1:4 for reasons of economy and quality. For cement-fly-ash-and-soil mixtures, the cement-to-fly-ash ratio varies between about 3:7 and 1.5:8.5 for sandy soil, typically around 1:3 or 1:4.

Applicable in regions where climatic conditions are a factor in pavement performance.

Reference(s): Army and Air Force (1994); Austroads (1998); Transportation Research Board (1987). Method summary: phenolphthalein is a color-sensitive indicator sprayed on soil to determine the presence of lime or cement in the field or laboratory; a reddish-pink color develops when a stabilizer is present (pH ≥ 10), verifying stabilizer presence and mixing uniformity.

Accuracy and precision: provides only an estimate of lime and cement content for depth of treatment. Adequacy of coverage: used only for checking soil locally. Implementation requirements: visualizing a reddish-pink color is straightforward, requiring minimal training or experience. General comments: useful for distinguishing properly treated from untreated areas, and helps identify where density-test samples should be taken.

Method summary: soil samples are collected to conduct classification, index, strength, and modulus testing. Strength and modulus testing on laboratory-compacted samples may yield different properties than field samples, due to differences in stress conditions and soil structure.

Accuracy and precision: depends on the test method; most are widely used and standardized by ASTM and AASHTO. Adequacy of coverage: measures only the sampled soil's properties; experience is required to interpret layering information from multiple borings and samples. Implementation requirements: a drill rig is needed for undisturbed samples from greater depths; shallow undisturbed samples can use hammer-driven Shelby tubes or test pit excavation.

Reference(s): Austroads (1998); Army and Air Force (1994); Little, D.N. (1995); Portland Cement Association (1995); Transportation Research Board (1987). Method summary: checks the stabilizer application rate. A spot check places a tray or mat of known area on the surface and weighs the stabilizer deposited on it after spreading; an overall check verifies the distance or area over which a truckload of stabilizer of known weight is spread.

Accuracy and precision: not documented in the references. Adequacy of coverage: a spot check typically uses a canvas of about 1 m² (1 yd²|); an overall check can cover the distance or area over which a truckload of cement of known weight is spread. Implementation requirements: the ability to weigh the truck before and after application (overall method), or to weigh and measure a canvas (spot check). General comments: field personnel should know the required stabilizer quantity per unit length or area of pavement; spot checking with a canvas of known area, or an overall check against known tonnage, helps assure the proper quantity is applied.

Reference(s): Austroads (1998); Qubain, Heirendt, and Li (2006). Method summary: evaluates element and mineralogy constituents of stabilized soil and stabilizer using X-ray fluorescence (XRF) and X-ray diffraction (XRD).

Accuracy and precision: not documented in the references. Adequacy of coverage: only representative samples are tested. Implementation requirements: both XRF and XRD testing require experience and knowledge to conduct. General comments: used for lime, cement, and fly ash content verification for uncured mixture, and for binder content measurement.

Reference(s): Austroads (1998); Little, D.N. (1995); Qubain, Heirendt, and Li (2006); Portland Cement Association (1995); Transportation Research Board (1987); White et al. (2005). Method summary: commonly used to determine compacted field density and moisture content of stabilized soil (ASTM D2922, D3017, D6938; AASHTO T238, T239) — a non-destructive test performed in a matter of minutes.

Accuracy and precision: per ASTM D6938, repeatability from a single operator gives a standard deviation of 0.3 to 1.2 lb/ft³ for wet density and 0.3 to 0.5% for moisture content, depending on soil type and test method (direct transmission or back scatter). Adequacy of coverage: commonly taken at several locations at intervals of 150 meters (500 feet) in the field. Implementation requirements: proper operation, calibration, and maintenance are essential; operator training and licensing are required. General comments: used to determine moisture content at the start of and during processing, and to check field density regularly; cannot measure density below 250–300 mm (10–12 inches). Comparing laboratory compaction results with field moisture-density readings helps confirm sufficient compaction.

Reference(s): Army and Air Force (1994); Portland Cement Association (1995). Method summary: trenches may be dug for visual inspection to evaluate depth and uniformity of mixing; a uniform color along the soil profile usually indicates thorough mixing.

Accuracy and precision: cannot be quantified — qualitative and dependent on inspector experience. Adequacy of coverage: limited to the observed area. Implementation requirements: some experience is necessary to recognize inadequate mixing or lack of stabilizer. General comments: adequate mixing shows a uniform soil color; a streaked appearance indicates non-uniform mixing.

Reference(s): Qubain, Heirendt, and Li (2006); Vennapusa and White (2009); White et al. (2005). Method summary: Light Weight Deflectometer (LWD), Falling Weight Deflectometer (FWD), and Clegg hammer tests obtain plate deflections under dynamic impulse loading. LWD applies relatively low contact stress (about 0.2 MPa or less) versus FWD (up to 1.0 MPa); FWD is often performed with an array of deflection sensors to develop deflection-basin data assessing subsurface stiffness/modulus to a depth of about 2 meters. The Clegg hammer measures deceleration (g's) under impulse loading (ASTM D5874).

FWD equipment is trailer-mounted; LWD and Clegg hammer devices are enclosed-box units carried by truck. Different manufacturers use different sensor types (geophones, accelerometers, or sensiometers), and device configuration affects the modulus value; LWDs are generally set up with 200 and 300 mm diameter plates, FWDs with 300 and 450 mm plates.

Accuracy and precision: LWD measurement values may vary by device and sensor type, and repeatability varies with the LWD device used. FWD results are widely used in the U.S. for direct measurement of pavement foundation material properties. For the Clegg hammer, ASTM D587 indicates a coefficient of variation of 2% or better in good laboratory workshop conditions; overall accuracy is not documented for any of the three, as there is no standard comparison method. Adequacy of coverage: the number of tests needed depends on observed variability; tests are relatively fast (about 5 minutes or less). LWD and Clegg hammer have shallower measurement depth (≤0.5 m) than FWD (up to about 2 m). Implementation requirements: experience and special equipment are necessary; LWD and Clegg hammer generally require less training and are less expensive than FWD. General comments: provides a direct measure of dynamic elastic modulus and can serve as a QC/QA tool.

Reference(s): Qubain, Heirendt, and Li (2006). Method summary: DCP tests assess compaction improvement depth by comparing before- and after-stabilization results, generally used in pavement foundation layer QA and able to measure soil properties to about 2 meters. Difficulties include large particles or boulders in the subsurface. Per ASTM D6951, the test drives a cone tip into the soil by lifting an 8 kg sliding hammer to a 575 mm drop height and releasing it; total penetration for a given number of blows is measured and recorded as mm/blow (penetration resistance). ASTM D6951 provides correlations between CBR and DCP results for different soil types; correlations to undrained shear strength and modulus are also documented in the literature.

Accuracy and precision: testing may be difficult in subsurface conditions with large boulders or rocks; ASTM D6951 indicates repeatability standard deviation of less than 2 mm/blow. Adequacy of coverage: sufficient evaluation requires many tests, depending on subsurface variability; penetration tests can assess improvement to greater depths than other surface tests. Implementation requirements: some experience and special equipment are necessary; DCP tests can be performed by one or two people. General comments: empirical correlations relate penetration-test results to soil engineering properties (strength, bearing capacity, stiffness/modulus, liquefaction susceptibility); useful for both QA and for QC assessment of improvement with increasing passes.

Reference(s): Vennapusa and White (2009). Method summary: Static Plate Load Tests (PLT) obtain load-versus-deflection curves to determine modulus of subgrade reaction, soil elastic modulus, or bearing capacity, by increasing a static load on the soil and recording plate deflections. ASTM D1195-93 describes repetitive static PLTs for airport and highway pavement evaluation and design; AASHTO T-222 describes a non-repetitive static PLT method for soils and flexible pavement components.

Accuracy and precision: widely accepted in geotechnical engineering; ASTM D1195-93 notes that precision could not be determined due to soil variability, and accuracy could not be determined for lack of a reference test. Adequacy of coverage: sufficient evaluation requires many tests; one referenced specification recommends one static PLT over a trial compaction section about 4 by 40 meters. Implementation requirements: experience and special equipment are necessary to produce sufficient reaction force (a heavy truck, dozer, or other heavy equipment is typically used); a trained field engineer is required to analyze load-deflection curves against design assumptions. General comments: provides a direct measure of modulus and/or bearing capacity and can serve as a QA tool for verifying design values.

(a) Description. Covers constructing one or more courses of an aggregate-and-cement mixture, or an AFLC mixture, on a prepared roadbed. Treated aggregate courses are designated as cement or AFLC; aggregate grading is designated per Table 703-2.

(b) Material. Aggregate, blotter, chemical admixture (set-retarding), emulsified asphalt, fly ash, hydraulic cement, lime, and water, each conforming to its own subsection of the parent specification.

(c) Construction Requirements. Submit a treated aggregate course mix design 30 days before production. Provide a minimum average compressive strength of 500 psi with no single test below 400 psi (AFLC mixture molded, cured, and tested per ASTM C 593 parts 10 and 11, with the curing period revised from 7 to 28 days at 100±3°F).

Include with the mix design: source of each component material; results of applicable tests; target values for each specified aggregate sieve size; a 200-pound sample of aggregate; a 50-pound sample of fly ash; a 25-pound sample of lime; a 25-pound sample of hydraulic cement; and a 5-pound sample of the retarder or other admixtures. Begin production only after the mix design is approved; furnish a new mix design if a material source changes.

(d) General. Store chemical additives and admixtures in closed, weatherproof containers. Prepare the underlying surface per Section 204 or 303 as applicable. Determine optimum moisture content per AASHTO T 180, Method D.

(e) Mixing. Do not begin mixing when atmospheric temperature is expected to fall below 40°F within 48 hours; do not place a treated aggregate course on a frozen or muddy surface, or while it is raining or snowing. Mix components until uniform, adding sufficient water for optimum moisture content for compaction plus 2 percent. Equip the mixer with batching or metering devices; maintain amounts within tolerance: aggregate ±2.0% by mass, fly ash ±1.5%, lime or cement ±0.5%, retarder or other additive ±2.0%, water ±2.0%. A retarder may slow initial set up to 2 hours (dissolved in water, added uniformly). When a central plant is used, transport the mixture in vehicles that maintain moisture content and prevent segregation and loss of fines.

(f) Placing, Compacting, and Finishing. Place, compact, and finish per Subsections 301.04, 301.05, and 301.06, maintaining moisture content within ±2 percent of optimum. Do not leave uncompacted treated aggregate undisturbed for more than 30 minutes; complete compaction and finishing within 1 hour of adding water (longer with a retarder). The compacted surface must be smooth, dense, and free of compaction planes, ridges, or loose material. Provide a construction joint if the time between placing adjacent partial widths exceeds 30 minutes.

(g) Construction Joints. For lime and fly ash mixtures, tie each day's operation into the previous day's completed work by remixing about 2 feet of the completed course before processing additional sections, adding 50 percent of the original lime or fly ash amount to the remixed material. For cement mixtures, or when a lime or fly ash mixture remains undisturbed more than 24 hours, cut back into the completed work to form an approximately vertical transverse construction joint.

(h) Curing. No traffic on the treated aggregate course. Keep the completed layer continuously moist (fine, uniform spray under pressure) until the next layer is placed, within 7 days of compacting and finishing. Placement may be deferred up to 21 days by sealing the surface with rapid-setting emulsified asphalt (not on intermediate layers): keep the surface continuously moist for at least 7 days, then apply undiluted CRS-2 or RS-2 emulsified asphalt at 0.25 gal/yd² per Section 409, providing a continuous film; if opened to public traffic, apply blotter per Section 411. If the course loses stability, density, or finish before the next course or acceptance, reprocess, recompact, and add additives as necessary, and reapply the asphalt seal where damaged.

(i) Acceptance. Per Table 302-3 for sampling/testing requirements and acceptance quality characteristic category. Fly ash, lime, cement, and chemical admixtures are evaluated under Subsections 106.02 and 106.03; emulsified asphalt under 106.03 and 702.09. Aggregate gradation's upper/lower specification limits equal the calculated mean of test results plus or minus the allowable deviation (Tables 703-2/703-3), with adjustment rules when the mean exceeds the maximum or falls below the minimum gradation value. Construction of treated aggregate courses is evaluated under Subsections 106.02 and 106.04; surface preparation under Section 204 or 303; emulsified asphalt seal under Section 409; blotter under Section 411.

(j) Measurement. Per Subsection 109.02 and as applicable: square yard width measured horizontally to include the top-of-aggregate width including designed widenings; square yard length measured horizontally along the roadway centerline. Emulsified asphalt measured under Section 409; blotter under Section 411.

(k) Payment. Accepted quantities are paid at the contract price per unit of measurement, adjusted per Subsection 106.05, for the Section 302 pay items listed in the bid schedule; payment is full compensation for the work prescribed in this section (see Subsection 109.05).