Introduction
Research Aims and Objectives
The aim of this project is to directly compare the California Bearing Ratio (CBR) swell test as set out in BS 1924-2: 1990 with the European accelerated swelling test (BS EN 13286-49:
This research examines the 2004 guidelines for fine-grained stabilized soils and evaluates existing test procedures, including the loss of strength on immersion test as outlined in the Manual of Contract Documents for Highways Works, Volume 1, Series 800, Clause 880.4.
The objectives of this project are:
• to review literature and case studies, and liaise with industry to establish current practice;
• to examine a range of materials with differing mineralogies and total potential sulfate contents;
• to compare the CBR swell test with the European accelerated swell test and the loss of strength on immersion test;
• to relate this testing to the current stabilisation process considering design versus screening and control, and;
• to disseminate the findings by inclusion within Highways Agency (HA) documents as appropriate and the production of industry guidance through Britpave and by academic publications.
Research Methodology
The research methodology comprises three phases as detailed below:
Classification testing, mineralogical analysis and analysis of the physical and chemical properties of selected materials:
• Glacial Till low plasticity material (LPM);
• Weathered London Clay high plasticity material (HPM);
• Oxford Clay with high total potential sulfate (H-TPS);
• Oxford Clay with low total potential sulfate (L-TPS)
Manufacture of trial mixtures with high and low plasticity and comparison of design and workmanship (pulverisation and compaction) using swell testing procedures:
• European accelerated swelling test (BS EN 13286-49:2004)
Information gathered is used to influence the selection of material
Desk study and literature review
Manufacture of trial mixtures using Oxford Clay with high and low total potential sulfate and comparison using swell test procedures:
• European accelerated swelling test (BS EN 13286-49:2004)
• Loss of strength on immersion test (MCHW1 series 800, clause 880:4)
Dissemination of findings via technical reports and academic publications
Review
The Properties of Soil
Soil with the exception of peat is formed by the breakdown of rock masses, either by weathering or erosion
The soils may accumulate in place, or undergo a certain amount of transport
The soils characteristics may also have been effected by its geological past, i.e being covered by ice, excessive heat, wind & rain, etc.
Problematic Soils
Different types of soils can lead to various issues, including shrinkage, expansion, collapse, and insufficient strength or stiffness Each soil type presents unique challenges that must be addressed for effective land use and construction.
Soils are generally made up of a combination of up to four different groups of differing particle size:
Clays themselves can be further divided into hundreds of different clay mineral types, of which three principle types are:
Group Minerals Mean Chief Physical
Kaolin kaolinite ~ 1à non-swelling, low plasticity, halloysite low cohesion
Illite illite and partially ~ 0.1à expansive, medium plasticity, degraded micas low permeability
Smectite montmorillonite, ≤ 0.01à highly expansive, very plastic,
“bentonite” and extremely low permeability mixed layer expansive clays
Expansive Soils
Volume changes in clay soils pose significant challenges for the construction industry, as these fluctuations can lead to considerable shrinkage during drying or swelling due to water absorption Such changes may ultimately result in damage to highways, buildings, and various construction projects.
It was estimated that in Britain in a typical 10year period between 1995-2005 swelling and shrinkage in clay soils had caused over 3 billion pounds worth of damage
Volume change in soils is generally a function of its moisture content, density, void ratio, stresses applied or released, along with the internal soil structure and mineralogy
The principle cause of expansion in natural soils is the presence of swelling clay minerals such as montmorillonite
Grim (1962), set out two types of swelling characteristics in clay soils, namely intercrystalline and intracrystalline swelling
Intercrystalline is when the uptake of water is restricted to the external crystal surfaces and the void spaces between the crystals
Intracrystalline is when water enters not only between the crystals but also between the unit layers which comprise the crystals
The capacity of clay minerals to adsorb water is influenced by their distinct types, which are determined by the arrangement of their tetrahedral (silicon) and octahedral (aluminum) sheets Despite having similar structural components, the specific configuration of these sheets defines the unique characteristics of each clay mineral.
Kaolinite: No Inter layer (intracrystalline) swelling:
Surface area is typically 10-20m 2 /gm
The layers are held together by Van der Waals forces and strong hydrogen bonds, preventing inter-layer swelling.
Illite is characterized by the presence of a potassium atom that perfectly fits into the hexagonal holes of the tetrahedral sheet, resulting in robust interlayer bonding While illites may contain water lenses within their interlayer spaces, they do not possess complete water layers and do not exhibit intracrystalline swelling.
Surface area is typically 65-100m 2 /gm
Montmorillonite particles possess a double layer of charge, as described by the Gouy and Chapman theory, which consists of water and ions surrounding the particle This results in a weak bond between the layers, leading to forces of attraction and repulsion that create a large surface area for cation exchange Consequently, this facilitates easy electron exchange and interlayer swelling, enhancing the material's properties.
Surface area is typically 50-120m 2 /gm (external surface)
Surface area is typically 700-840m 2 /gm (including the inter layer surface)
Bell & Cullshaw (2001), stated that generally kaolinite has the smallest swelling capacity, and nearly all of its swelling is of the intercrystalline type
Intermixed illite and montmorillonite may swell by 60 – 100%
Calcium (Ca) montmorillonite ranging between 50 – 100%
Sodium (Na) montmorillonite can amount to 2000% of the original volume, the clay then having formed a gel
Incorporating lime into harmful clays like Na-montmorillonite can effectively diminish their swelling potential, as calcium ions (Ca²⁺) replace sodium ions (Na⁺) This process not only reduces swelling but also enhances the clay's strength through dehydration and cementation.
The structure of a soil is taken to mean both the geometric arrangement of the particles or mineral grains, as well as the intracrystalline forces which may act between them
Soil fabric however, normally refers only to the geometric arrangement of the particles
Inter particle arrangements of the fabric of the soil can be described in many different ways with particles having various arrangements as follows:
Face to Face: Edge to Face:
Shifted Face to Face: Edge to Edge:
Soils consist of diverse layers of clay minerals, typically featuring both expanded water-bearing and non-water-bearing layers These layers are formed through a process of interstratification and weathering over time.
The capacity of a material to take on water and soften and/or swell, can also be down to the elementary particle arrangements and void spaces
Individual silt or sand particles:
Clothed silt or sand arrangements:
The swelling behavior of coarse-grained soils is primarily influenced by their particle size distribution In contrast, the swelling behavior of fine-grained soils is more significantly affected by their geological history and structural characteristics than by particle size distribution.
In less dense soils, initial expansion occurs in three dimensions within areas of looser soil, or voids, leading to a volumetric expansion that can be perceived as a "softening" of the soil mass.
In densely packed soil with limited void space, any volume changes necessitate immediate swelling of the soil mass When confined, this swelling often occurs in a one-dimensional direction, typically upward, which can lead to significant structural damage.
Highly expansive clays typically exhibit very low permeabilities, resulting in slow moisture movement In contrast, moderately expansive clays, which have a reduced swelling potential but higher permeabilities, may experience greater swelling during a single wet season compared to their more expansive counterparts.
Therefore undisturbed expansive clay soils often have a high resistance to deformation and may be able to absorb significant amounts of swelling pressure, without swelling themselves
When the microstructural arrangement of expansive clay is disturbed through processes like re-mixing or re-moulding during construction or testing, it can lead to increased swelling compared to undisturbed clay samples.
Several factors can lead to changes in soil volume, and this discussion excludes any chemical influences for now.
Understanding the factors that influence the swelling behavior of materials is crucial for comprehending their engineering properties This knowledge is essential for making informed decisions when selecting materials for construction projects.
The ingredients generally necessary for potentially damaging swelling to occur in natural soils are:
1 The presence of montmorillinite in the soil
2 There must be a source of water for the potentially swelling clay.
Volumetric Expansion (Formation of the Mineral Ettringite)
Soil swelling can occur due to the presence of specific chemical elements that, under certain conditions, form crystalline structures As these crystals grow, they exert pressure on the soil matrix, causing it to expand.
A prevalent issue in UK soils is the presence of sulfate, which can react with aluminas (a key component of clay), calcium (from lime or cement), and water to form a highly expansive crystalline mineral known as ettringite (calcium-aluminate-sulfate-hydrate) Specific conditions must typically be satisfied for this reaction to take place.
1 a source of sulfates (including those derived from oxidation of sulfides);
2 the presence of mobile groundwater, and;
3 calcium hydroxide and calcium aluminium hydrate in the cementitious matrix
Ettringite is a hydrous calcium alumino-sulfate mineral that forms in high pH environments with ample sulfate concentration, particularly during the early hydration of calcium aluminate This mineral is notable for its significant expansion potential, which can reach up to 250%.
2005) Thaumasite attack is a secondary concern for stabilised soils and generally occurs as the pH decreases and at low temperatures
Sulfur is widely distributed in the form of sulfides and sulfates
Sulfides primarily oxidize into sulfate, creating acidic ground conditions that can aggressively attack construction materials like steel and concrete The introduction of calcium oxide-based binders, such as lime and cement, can lead to the formation of secondary sulfates like ettringite or thaumasite, resulting in ground heave.
Ettringite, a compound represented by the formula [Ca6Al2(OH)12(SO4)3·27H2O], is created through the reaction of soluble sulfates, such as those from gypsum, with dissolved alumina, which is generated by the high pH levels resulting from lime addition The crystallization process of ettringite is expansive and undergoes significant volume changes as it absorbs water Additionally, thaumasite (Ca3Si(CO3)(SO4)(OH)6) can form from ettringite without any alteration in volume.
P Sherwood (1993) explains the importance in understanding the chemical factors which are likely to affect Stabilised Soil
“The reactions between sulfates and the hydrated silicates and alluminates lead to products that occupy a greater volume than the combined volume of the reacting constituents”
Sulfate attack requires water for the reaction to occur, and typically, there is not enough water during mixing to dissolve significant amounts of sulfate Therefore, unless additional water can penetrate the material, substantial sulfate attack is unlikely, even in the presence of high sulfate concentrations.
Soil may contain sulfides, such as iron pyrites, which can oxidize to form jarosite or gypsum when combined with calcium carbonate Both jarosite and gypsum are known for their expansive properties, and jarosite can form independently of calcium minerals, even in unstabilized materials.
To accurately assess soil materials, it is essential to test for both sulfates and sulfides This is typically done by measuring the total sulfur content and multiplying the result by three to estimate the Total Potential Sulfate (TPS) content, assuming complete oxidation of sulfides To determine the sulfide content, subtract the total sulfate content from the total sulfur content Comprehensive guidelines for testing soils for sulfate and sulfide, along with acceptable limits, are detailed in TRL447 and HA74/07.
The Principles of Ground Improvement
Ground improvement principles focus on enhancing soil properties to provide various benefits, including financial savings, extended design life, improved construction ease, and increased durability.
Two forms of ground improvement technique are “Modification” and “Stabilisation”
The Modification process can generally be described as an improvement of either the moisture content and/or the workability of the material
Adding quick lime (CaO) to a material initiates a drying effect through an exothermic reaction as the lime hydrates with the free water present in the surrounding material, effectively reducing moisture levels It is crucial to ensure that the lime is fully hydrated and that the material has the appropriate moisture content for optimal compaction, which is typically assessed through laboratory tests that evaluate the relationship between dry density and moisture content of the soil.
Cation exchange plays a crucial role in enhancing the workability of materials, particularly when calcium ions from lime interact with metal ions in the clay structure This process can lead to an increase in the plastic limit (PL), which subsequently reduces the plasticity index (PI), resulting in a more friable and workable material.
This plasticity and moisture change can typically change a materials workability in typical cases by 3% moisture for every 1% CaO added
M o is tu re c o nt en t ( p er c ent ) x y x = 10% above PL y = 5% below PL
Initially, the soil had a moisture content of 35%, which was 10% wetter than its natural plastic limit (PL) Following the addition of 2% lime, and assuming no moisture loss from exothermic reactions, the clay's moisture content became 5% drier than its treated PL of 40% This adjustment results in a broader workable moisture content range for the soil.
The modification process utilizes only the minimum necessary amount of binder to transform materials previously considered unsuitable, such as overly wet substances, into suitable ones by bringing them within an acceptable moisture range While the process does not require a permanent increase in strength beyond that of the natural material, some short-term stabilization or strengthening may occur However, this initial strength gain may be reversible once the material achieves its long-term equilibrium moisture content.
Stabilisation is achieved by adding sufficient CaO to elevate the pH to around 12.4, creating an environment that facilitates the dissolution of silica and alumina from clay particles This process allows the silica and alumina to react with calcium ions from lime, resulting in the formation of calcium silicate hydrates (CSH) and calcium aluminate hydrates (CAH) Over time, these hydrates crystallise into compounds similar to those found in cement.
The pozzolanic reaction, named after the siliceous volcanic ash found in Pozzuoli, Italy, occurs slowly from a cementing perspective and involves the use of pozzolanic materials like pulverised fuel ash (PFA), which are rich in silica but contain little to no lime While the chemical effects of lime stabilization can resemble those of cement, the presence of unrefined silica and alumina in the soil may result in lower strength at 28 days However, the long-term pozzolanic reaction can persist for several years, potentially achieving strength levels that equal or surpass those of cement-bound materials.
The stabilisation process can therefore continue for years as long as there is enough lime still available to react with the clay minerals
Stabilizing silts typically involves the use of cement, as lime alone is insufficient; it necessitates the incorporation of pozzolanic materials like clay, PFA, or ground granulated blast furnace slag (GGBS) to achieve the desired cementing effect.
Stabilisation as with modification can change the workability of a material, but unlike modification it is primarily used to permanently enhance the strength/stiffness of the material as well
When evaluating ground improvement techniques, it's essential to consider their effectiveness in meeting project requirements Techniques such as stabilization and modification involve earth moving, which may not be suitable for densely populated areas In such cases, alternatives like piling with concrete or lime/cement recycled materials, vibro stone columns, and dynamic compaction may be more appropriate solutions.
In urban areas with limited space, utilizing piling, vibro stone columns, or dynamic compaction techniques may be more advantageous than traditional methods like stabilization or modification, which involve digging, treating, and relaying.
The materials present in a contract may not be suitable for vibro or dynamic techniques, particularly when dealing with made ground or silts that can be sensitive and prone to instability under vibration In such cases, these materials are better suited for stabilization or piling methods instead.
These other forms of ground improvement techniques, are outside the scope of this paper.
Durability
Durability refers to the ability of materials to withstand various environmental factors, including weathering, erosion from traffic and construction activities, and resistance to soil chemistry.
As can be seen from the previous sections on expansive soils, there are two very different reasons for swelling in clays, and they are:
1 The natural soils ability to take on (adsorb) water and therefore change its volume due to its minerality, particle size distribution, voids ratio etc, and these have been classed under the design durability section of this paper
2 The materials chemical durability /resistance to chemical reactions such as ettringite formation, which may result in a volume change and a decrease/increase in performance compared to its design durability/performance
As with all construction materials there is generally a desire to upgrade a materials strength/stiffness and in turn hopefully increase its durability
Durability testing can vary, and there are numerous methods which can be adopted to establish whether or not a material can be defined as being durable
Understanding the two distinct types of durability issues—Chemical and Design durability— is crucial when stabilizing soil These issues can manifest simultaneously within the same soil, affecting its stabilization in various ways.
Chemical durability refers to a material's ability to withstand chemical reactions caused by specific elements, such as sulfates, which can diminish its performance relative to its intended durability.
Organic matter content significantly influences soil durability by consuming some of the lime added to the soil This neutralization process reduces the amount of available lime, which is essential for reacting with soil elements, potentially leading to a weaker and less durable material.
Chemical durability tests have been well documented and publicised in the format of DMRB Volume 4, section 1, Part 6, HA74/07 and TRL447:
Design durability can be defined as the effects on a material due to workmanship or design elements (such as inadequate compaction, frost, poor choice of binder, etc)
Various durability tests for stabilized soils are conducted to evaluate specific aspects of their engineering behavior, aiding in the assessment of their end-use characteristics Similar to ground improvement techniques, each test serves a distinct purpose, requiring engineers to carefully weigh the unique advantages and disadvantages of each option before selecting the appropriate tests to employ.
Simple classification tests, including moisture content, particle size distribution, and Atterberg limits, play a crucial role in assessing the long-term durability of materials Although they may not initially appear to be design durability tests, each provides valuable insights into material stability For instance, the plastic limit test reveals a material's workability across various moisture contents and indicates its equilibrium state, whether it is dry or wet relative to its optimal long-term moisture content, thus demonstrating its stability and tendency to absorb or release moisture.
Testing various parameters is essential for assessing a material's durability Materials with high clay content are often prone to frost heave under specific weather conditions Additionally, silt content can render a material unsuitable for certain ground improvement techniques, like dynamic compaction, due to its instability and sensitivity to vibrations.
Engineers rely on a multitude of tests to guide their material selection and design choices effectively Numerous studies illustrate how materials perform in comparison to one another and under various conditions Consequently, it is essential for engineers to trust the design tools they use and to ensure that the parameters for each test and process are accurately defined.
The CBR swell test serves as an essential design tool for engineers evaluating the suitability of stabilized materials in construction projects Recognized by the TRL for its effectiveness in determining chemical durability and design strength, this test has been chosen for comparison with the European Accelerated Swelling test, the only other swell test considered by the Highways Agency for assessing stabilized soil durability.
There are many durability tests which could have been considered for use by the Highways Agency and some of them are described in brief as follows:
1.8.2.1 European Accelerated Swelling Test: BS EN 13286-49:2004
The standard specifies the production of three specimens, each measuring 50 mm in diameter and 50 mm in height, to achieve a wet density of 96+0.5% of the "Normal Proctor" using axial compression These specimens must be created from materials that pass through a 6.3 mm sieve and stored at 20+2 °C with over 90% humidity for 1.5 to 2 times the material's workability period Subsequently, they should be fully immersed in water at 40+2 °C for 168+4 hours before testing The workability period is defined in the producer's technical sheet or determined per BS EN 13286-45: 2003 According to MCHW Series 800, cement has an approximate workability of 2 hours at 20 °C, equating to 35 degree hours from the time of cement addition, calculated by summing the mean ambient air temperature above 0 °C over each hour.
The European accelerated swelling test is an unconfined method that permits expansion in all directions, as illustrated in Figure 1 Given that the manufactured specimen depicted in Figure 2 is smaller, it is expected to achieve saturation in a shorter time frame compared to a CBR specimen.
Figure 1 Test BS EN 13286-49:2004 allows expansion in all directions
Figure 2 European accelerated swelling test compaction cylinders (above) specimen being weighed (below)
1.8.2.2 Laboratory Determination of the California Bearing Ratio: BS 1924-2:1990
A sample of stabilised material is compacted into a CBR mould (152 mm diameter × 127 mm high) in approximately three equal layers, using a 2.5 kg rammer (62 blows/layer), then sealed at 20±2 o C for three days
The procedure entails soaking specimens to determine a soaked California Bearing Ratio (CBR) value This is achieved by placing the specimens in a water bath maintained at 20±2 °C A collar is positioned on top of each specimen, and a perforated base plate is affixed to the bottom to facilitate water ingress, ensuring the water level remains just below the collar's top.
Figure 3 CBR Swell Test (BS 1924-2: 1990)
During the immersion process, water enters the sample through capillary action If there is minimal or no water at the top of the specimen after the initial three days, additional water should be added for the final 24 hours before strength testing The standard curing procedure for the California Bearing Ratio (CBR) test involves a total of seven days, consisting of three days of air curing followed by four days of soaking Similarly, the 28-day swell test follows the same curing method but requires soaking for 28 days instead of four.
In HA74/07, the requirements for testing specimens have been updated from HA74/00, now mandating that soaked CBR specimens be evaluated for both strength and swell The material suitability criteria detailed in HA74/07 indicate that a specimen is deemed durable if the average CBR value exceeds 15%, with no individual result falling below 8% Additionally, the average heave must be 5 mm or less, with no individual measurement exceeding 10 mm.
Historical Literature Review
The literature review aimed to address practitioners' concerns regarding the use and confidence in stabilized materials within the construction industry The Highways Agency (HA) is recognized as a leader in the UK industry, yet recent failures in several HA contracts have raised questions about the effectiveness of current specifications and testing methods in identifying potential risks before contract initiation.
The Government, through the Secretary of State for Transport, utilizes the Highways Agency to manage and maintain England's strategic road network In collaboration with the Department of Transport, the Highways Agency reviews and establishes specifications and guidelines for the development and construction of road schemes.
The Highways Agency collaborates with academic research organizations like the Transport Research Laboratory (TRL) to explore technical challenges and enhance specifications and guidance Currently, three key documents are utilized by the Highways Agency and other UK professionals to assist in the application of stabilization techniques.
1 Design Manual for Roads and Bridges (DMRB), Volume 4; section 1; Highways Advice note; HA 74/00: Treatment of Fill and Capping materials using either Lime or Cement or both
2 Manual of Contract Documents for Highway Works (MCHW); Volume 1; Specification for Highway Works(SHW); Series 600(Earthworks) and Series 800 (Road Pavements)
3 Transport for Research Laboratory (TRL) Report 505; Swell test requirements for Lime stabilised materials
The documents in question reference several British Standards, including BS1924:1990, which outlines testing methods for cement-stabilised and lime-stabilised materials These standards have been utilized or cited in relation to various Highway contracts that are currently under scrutiny.
Since the 1950s thousands of miles of state highway and major airports have been stabilised in America, South Africa, and some European countries
In the early 1980s, lime and cement stabilization gained traction in the UK when the British Airports Authority (BAA) implemented this method for enhancing their car parks, taxiways, and runways The process was first trialed at Gatwick Airport, where a lime stabilization treatment was also used to modify a lake area designated for fire training activities.
-1992 The application of lime and cement soil stabilization at BAA airports Proceedings of the Institution of Civil Engineers (Transport), 95, February, 11-49)
The effects of sulfates on stabilised materials had previously been investigated
( P T Sherwood 1962: the effects of sulphates on Cement and Lime stabilisation)
This study examines how sulfates and mineral composition influence lime or cement-stabilized soils It highlights the relationship between soil immersion, clay fraction percentage, and sulfate content The findings indicate that lime or cement-treated soils are prone to disintegration when they have a high clay fraction in sulfate solutions under specific conditions.
Removing the clay fraction eliminates the risk of disintegration in cylindrical specimens The study observed that while some specimens disintegrated when fully immersed in an unconfined state, others remained intact Additionally, all specimens behaved as anticipated when exposed to increasing concentrations of sulfate solution.
However, interestingly, when the sulfate levels increased above 1% with some specimens, the amount of strength loss levelled off
The study found that certain lime clay specimens disintegrated within days when immersed in sulfate solution, while cement-treated sand specimens with identical sulfate quantities remained unaffected even after one year of immersion.
The extent of soil disintegration correlates with the clay content, as a reaction between clay and sulfate ions occurs when lime and excess water are present.
The paper suggests that it is the soil grain size that affects the durability and not the sulfate content The larger the grain size, the better the durability
M J Dumbleton (1962) conducted investigations in the UK to evaluate the potential of lime for soil stabilization, revealing that the impact of lime on plasticity is influenced by the type of clay, the proportion of lime added, and the duration of treatment.
Research presented in Rogers & Glendinning's paper "Modification of Clay Soils using Lime" at the Lime Stabilisation seminar in Loughborough University in 1996 highlights both immediate and significant long-term changes in the plasticity of clay materials The study suggests that the extent of these changes may be influenced by the degree of pulverisation during on-site mixing, potentially causing delays in the observed effects.
In 1975, the Specification for Roads & Bridges (DoT SRBW:1975) did not provide guidelines for lime stabilization, but it did include specifications for cement stabilization, which set a maximum limit of 0.25% total sulfate.
In 1986, the Department of Transport's Specification for Highway Works (Sixth Edition) replaced sulfate content as a measure for controlling cement-stabilized materials with an immersion test This immersion test, detailed in BS1924:1975, remains the preferred suitability and durability assessment for hydraulically bound materials, as outlined in the MCHW 2005 series 800.
The basic principle of the test is as follows:
In this study, a total of 10 specimens, either cubic or cylindrical, are created from the same laboratory mix Five specimens are air-cured for 14 days before testing their unconfined compressive strength (UCS), while the remaining five are air-cured for 7 days and then immersed in water for an additional 7 days To be considered durable, the soaked specimens must achieve over 80% of the UCS of the un-soaked specimens without any visible signs of deterioration All curing processes are conducted at a consistent temperature of 20°C.
These specimens were deemed to be strong enough to handle after 7 days, and large enough to be representative of the mixed material
The total sulfate test was considered too variable because obtaining a small, representative sample of just 10 grams from a larger material area may lead to potential sampling issues.
Please note that the MCHW 2005; series 800 differs only from B.S.1924 (1975), in the respect that the curing periods have now been extended to 28days (14air cure + 14days in soak)
Conclusions of Literature Review
The CBR test methodology, combined with past experiences, indicates that achieving swell due to oxidation is challenging under the existing testing procedures and curing conditions.
The Thomas, Kettle & Morton report indicates that oxidation of pyrite occurred in both fully saturated (regime 2) and partially saturated specimens (regime 3), with minimal differences observed between the two This trial was performed on stabilized colliery shale, which exhibited relatively high strengths It raises the question of whether different curing regimes would yield varying results if applied to a more cohesive material.
There is a pressing need to reassess current durability testing procedures to effectively predict in-situ performance, especially since the MCHW does not set upper limits for sulfates or sulfides and relies solely on laboratory tests for suitability This necessity arises from failures in several stabilization contracts, including the A10 Wadesmill Bypass, Saxmundham Bypass, and M40 motorway projects Additionally, the introduction of IAN73 (HD25) Design Guidance for Road Pavement Foundations and the review of HA74/00 (now replaced by HA74/07) regarding the treatment of fill and capping materials with lime or cement further emphasize the importance of this review.
Concerns from industry specialists suggest that the California Bearing Ratio (CBR) swell test (BS 1924-2:1990) may not accurately reflect material behavior under actual site conditions, potentially leading to construction failures despite passing the test In contrast, the European accelerated swelling test (BS EN 13286-49: 2004) remains underutilized and poorly understood within UK practices, highlighting the need for better assessment methods in material evaluation.
Britpave (2005) provided guidance for the stabilization of sulfate-bearing soils to aid practitioners, highlighting risk mitigation strategies like the incorporation of ground granulated blast furnace slag to reduce expansive sulfate reactions However, it is crucial that stabilization specifications are not solely based on limiting values for chemical constituents; a dependable laboratory testing procedure to assess volumetric stability is vital for effective soil management.
Several variables identified in the literature review can influence the formation of ettringite and thaumasite, either individually or collectively Additionally, these factors may impact the softening and swelling properties of stabilized fine-grained soils A comprehensive list of these factors is provided in Table 3.
Table 3 Variables which may have an effect on the swelling characteristics of a soil
Chemistry Classification Workability (site and laboratory)
Plasticity index Activity index Particle size distribution Moisture content (and or MCV)
Binder content Mellowing Degree of pulverisation Degree of compaction Density gradient Curing time
Confining conditions Permeability Strength / stiffness
Industry practitioners have indicated that the current laboratory CBR swell test (BS1924:1990 method) fails to replicate the oxidation conditions of pyrite and the swell development observed in the field Consequently, laboratory swell measurements are smaller than those found in situ, primarily due to the inhibited formation of expansive minerals like ettringite This discrepancy has been identified as a critical area for further research.
The current BS 1924-2 1990 CBR swell test and its associated swell limits may not effectively predict in situ behavior, as evidenced by reported failures of stabilized soils, particularly on the M40 contract, where significant heave occurred in materials deemed suitable by the CBR swell test Furthermore, it is important to distinguish between "Softening" and "Swell," as both phenomena can arise during the immersion of specimens; for this work, specific definitions will be provided for clarity.
• Softening refers to the process by which a material loses strength/stiffness and softens, possibly due to design or workmanship limitations , and;
• Swell or swelling refers to the process by which a material expands, possibly due to insufficient chemical durability.
Phase 2 - Classification testing
Four sources of material were selected for inclusion within this laboratory study They were selected to represent materials that are common candidates for soil stabilisation within the
The first two sources were selected to allow an initial comparison between low and high plasticity clay:
• Glacial Till: Brown Silty SAND with some Clay, Chalk & Gravel Report reference: LPM
• Weathered London Clay: Orange-Brown mottled grey silty CLAY Report reference: HPM
The second two sources were selected to allow a comparison between relative levels of TPS:
• Oxford Clay: High Total Potential Sulfate/Sulfur Material Report reference: H-TPS
• Oxford Clay: Low Total Potential Sulfate/Sulfhur Material Report reference: L-TPS
Classification testing was performed to assess the physical and chemical properties of the material, as detailed in Table 4 The chemical analysis of sulfide and sulfate content followed the guidelines of TRL report 447, while other material property tests adhered to relevant British Standards The activity index, representing the ratio of the plasticity index to the mass percentage of soil particles smaller than 2 microns, was calculated Additionally, the Initial Lime Consumption Value, Plasticity Index, and pH Value were determined according to BS 1924-2: 1990, BS 1377-2: 1990, and BS 1377-3: 1990, respectively.
Table 4 Classification test and chemical analysis
The report details various soil properties, including the Plastic Limit, Liquid Limit, and Plasticity Index, which assess the soil's plasticity and consistency It also highlights the Clay Fraction and Activity Index, indicating the soil's clay content and its reactivity Additionally, the report examines Organic Matter Content, Sulfate Content, and Water Soluble Sulfate levels, which are essential for understanding soil chemistry and potential environmental impacts Total Sulfur percentage and pH Value are also included, alongside Initial Lime Consumption Value, which is crucial for determining soil amendment needs.
Oxford Clays in southern and eastern England typically contain 5-15% sulfides, which are a reduced form of sulfur However, chemical tests reveal that most sulfur exists as sulfate rather than sulfides The concentration of sulfate in clay largely depends on local ground conditions rather than the clay type itself High sulfate levels can accumulate at the base of the weathering zone, leading to significantly lower concentrations in the upper layers due to the influence of mobile groundwater.
The water content and moisture condition value (MCV) were determined in accordance with
BS EN 1097-5: 1999 and BS EN 13286-46: 2003, respectively The results are presented in table 5
Table 5: Moisture content and MCV value
Phase 3 – Experimental Laboratory Testing Stage 1
3.1 Test Procedure - Trial Mixture Design
Phase 3 Stage 1 trial mixtures were developed to assess how design and workmanship influence the durability of stabilised fine-grained soils The study utilized Glacial Till and London Clay with low sulfate content to reduce the impact of sulfate reactions on durability testing outcomes The focus was on evaluating how binder content, degree of pulverisation, and compactive effort affect specimen durability, as measured by two key test procedures identified in the literature review.
1.8.2.1: European Accelerated swelling test: BS EN13286-49: 2004
1.8.2.2: Soaked California Bearing Ratio test: BS 1924: 1990: Part 2
Mix 1: Low Plasticity Material (LPM), with a low Total Potential Sulfate content (L-TPS)
The material was intentionally blended with a low binder content and insufficient pulverization, combined with minimal compactive effort, to produce a lime-treated substance that is likely to fail the CBR swell test.
Mix 2: Low Plasticity Material (LPM), with a low Total Potential Sulfate content (L-TPS)
To achieve stabilization, the material was blended with an adequate amount of lime at its Initial Lime Consumption Value, ensuring thorough pulverization and effective compaction This process aimed to produce a lime-treated material capable of passing the CBR swell test.
Mix 3: Low Plasticity Material (LPM), with a low Total Potential Sulfate content (L-TPS)
The material was blended with an adequate amount of lime and cement to facilitate stabilization, ensuring thorough pulverization and effective compaction This process aimed to produce a mixture capable of passing the California Bearing Ratio (CBR) swell test.
Mix 4: High Plasticity Material (HPM), with a low Total Potential Sulfate content (L-TPS)
The material was intentionally blended with a low binder content and subjected to minimal pulverization and compaction to produce a lime-treated substance that is likely to fail the CBR swell test.
Mix 5: High Plasticity Material (HPM), with a low Total Potential Sulfate content (L-TPS)
The material was combined with an adequate amount of lime to achieve complete stabilization at its Initial Lime Consumption Value This process involved thorough pulverization and effective compaction to produce a lime-treated material capable of passing the California Bearing Ratio (CBR) swell test.
Mix 6: High Plasticity Material (HPM), with a low Total Potential Sulfate content (L-TPS)
The material was blended with an adequate amount of lime and cement to ensure complete stabilization, achieving effective pulverization and compaction This process aimed to produce a lime and cement-treated material capable of passing the CBR swell test.
Materials were sourced and classified following the guidelines outlined in HA74/07 Specimens for each mix were prepared with varying water contents, specifically within a moisture content variation (MCV) range of 8 to 12.
The BS EN 13286-49: 2004 European accelerated swelling test, essential for compliance with the Series 800 (MCHW1, 2007) specification for soil lime/cement treated materials, mandates a curing period of 1.5 to 2 times the cement's workability period before immersion However, concerns have emerged regarding the adequacy of this short curing time for cement-treated materials To address these concerns, an additional curing regimen was implemented, resulting in a separate set of specimens being cured for an extended duration of 72 hours.
20 o C) prior to being totally immersed and tested This curing regime is herein referred to as
‘non standard’ in relation to the European accelerated swelling test
3.2 Results of Phase 3, Stage 1 Laboratory Testing
The results of the Phase 3 Stage 1 laboratory testing are presented in Table 6 to table 8
Table 6 Preliminary testing of Phase 3, Stage 1 mixes
Table 6, 7, and 8 present key data, including dry density (DD) measured in Mg/m³ and water content (WC) expressed as a percentage The lime and cement contents are calculated as a percentage of the dry mass of soil Additionally, water content, the percentage of material passing the 5mm BS test sieve, and the moisture control value (MCV) were recorded at the time of compaction.
= standard test method, ns = non-standard test method All of the above CBR specimens were prepared, made, cured and tested in accordance with BS 1924-2:1990 method 4.5 (using a 2.5kg rammer)
Table 7: 7 day and 28 day CBR results for Phase 3, Stage 1 mixes
Average 7 day CBR 28 day CBR (%)
DD (Mg/m 3 ) Top Bottom Average
Table 8: Swell and expansion measurements for Phase 3, Stage 1 trial mixes
3.3 Discussion of Phase 3, Stage 1 Laboratory Testing
Mix 1 (LPM 1 – L-TPS) results show that the 7 day CBR swell test has passed the CBR requirement of >15% (achieving 16-22%); however after 28 days soaking the CBR decreased (achieving 10-12%), but was still greater than the minimum of 8% More importantly, this implies that stabilisation had not occurred The 28 day CBR swell results were all 15% (achieving 17-24%) After 28 days soaking, the CBR values increased further (achieving 26-37%), and recorded 15% (achieving 35-70%) After 28 days soaking, the CBR values increased further (achieving 45-90%) and recorded