Soil Improvement

A new method is presented for calculating the net vertical loads applied by embankments and surcharge pressures to geosynthetic reinforcement in column-supported embankments. The new method satisfies force equilibrium and displacement compatibility of the system components, including the embankment, geosynthetic reinforcement, columns, and foundation soil between columns. Nonlinear response of the soil components is incorporated in the analysis. The new method is in good agreement with the results of instrumented case histories, pilot-scale tests, and numerical analyses.

This paper presents the current U.S. state-of-practice for the design of the load transfer platform used in column supported embankments. Two design methods are in use in the U.S. today, the catenary and beam method. Both methods are presented, including the fundamental design assumptions that differentiate the two approaches. The paper also provides guidance on the determination of the geosynthetic properties for the load transfer platform based on the method used to design the platform.

The design of piled embankments is extremely complex. Several methods exist for estimating the magnitude of arching occurring in the embankment fill. However these methods involve detailed and often complex calculations, which are considered tedious for preliminary design. Experimental and numerical analysis of piled embankments have demonstrated the existence of a plane of equal settlement, at the critical height, in the arched embankment fill. A new method, based on the critical height of the embankment, is proposed for calculating the magnitude of arching. The critical height is calculated assuming a log spiral shaped shear plane in the embankment fill, above the pile caps. For the case histories examined, the new method compared favorably with existing more complex methods for calculating the magnitude of arching in piled embankments.

Cemented mixtures composed of rigid sand particles and soft rubber particles are tested to investigate their small strain behavior. Mixtures are prepared with different volumetric sand fractions (sf) to identify cementation and decementation responses using shear wave propagation at the small strain, and k_o-loading. The shear wave velocity increases with increase in the vertical effective stress for uncemented soils. The higher the sand fraction, the higher the shear wave velocity. As the sand fraction increases, the response of the specimens shows transition from rubber-like to sand-like behavior. During cementation hardening, the shear wave velocity dramatically increases even at the fixed stress. While the cementation hinders the apparent behavior during loading stage, transition behavior appears during reloading stage due to the breakage of the cementation hardening.

Approximately 290 million tires are disposed of annually in the United States. Although markets now exist for about 80% of the US waste tires, it is estimated that about 27 million waste tires still need to be disposed of in landfills. Waste tire rubber has been used in several applications such as highway backfill, subgrade and embankments; as drainage, landscaping, and sorptive materials; as well as in alternative methods for energy generation. While waste tire rubber has been used to modify the mechanical properties of coarse-grained soils, just a limited number of studies have addressed the stabilization of fine-grained soils with rubber. None of these studies have tested mixtures of rubber with highly expansive clays, such as the ones widely encountered in Colorado and other western US states. In this study, the effect of adding small particles of waste tire rubber on the swelling potential of an expansive soil from Colorado was evaluated. The index properties and compaction parameters of the rubber, expansive soil, and expansive soil-rubber (ESR) mixture tested were determined. One-dimensional swell-consolidation tests were performed to assess the feasibility of using small particles of waste tire rubber as a mechanical additive to mitigate the swelling potential of the expansive soil. While the ESR mixture is more compressible than the untreated soil, both the swell percent and the swelling pressure are significantly reduced by the addition of rubber to the expansive soil.

A laboratory study was undertaken to evaluate the effectiveness of different percentages of cement kiln dust (CKD) as a soil stabilizer, relative to the implementation of new Mechanistic-Empirical Pavement Design Guide (MEPDG). Cylindrical specimens were compacted and cured for 28 days in a moist room having a constant temperature and controlled humidity. After curing specimens were tested for resilient modulus (M_r), modulus of elasticity (M_E) and unconfined compressive strength (UCS). The study revealed that the values of M_r, M_E and UCS for the CKD-stabilized specimen increased with CKD amount. Reaction products due to chemical reactions are clearly observed in the soil voids based on the micrographs obtained using scanning electron microscope (SEM), and they are responsible for the increase in modulus and strength.

Many municipalities and private owners have faced increased pavement maintenance and replacement costs when lime modified clay subgrades prematurely fail. Preliminary laboratory test results determined that the typical lime treatment (7% by dry weight) used by the City of Garland, TX was approximately one-half the amount of lime necessary to permanently stabilize a high P.I. (39) clay. Further laboratory and a field testing program was then undertaken to determine if an economical alternative to 14% lime could be used to extend the durability of street subgrades. Four test sections were constructed using various combinations of lime to pre-treat the clays before stabilizing with cement. The testing program revealed that using combinations of lime and cement increased the typical subgrade durability based upon 4.8 to 5.7 times greater strengths after one year of exposure to in-place conditions. In addition, 28 day moist cured then saturated samples had 3.7 times greater compressive strength and 3.5 times better strength following 12 cycles of wetting and drying after 4 months of moist curing.

This paper describes an approach to cold-weather subgrade stabilization. Background information was derived from the open literature, ongoing research and discussions with stakeholders in industry, academia and the government. Traditional subgrade stabilization in road construction is defined herein as the use of unmodified lime, cement or fly ash in soil to improve the strength and overall performance of a pavement system. Many of the non-traditional additives currently on the market have been evaluated in recent years. Broadly, these may be categorized as asphalts, polymers, electrolytes, biochemical additives, FGD gypsum and lime/cement additives. Careful evaluation of each category in turn reveals that the most promising alternative to unmodified lime and cement is modified lime and cement. Specifically, rapidly maturing research in the field of cold weather concreting has demonstrated the ability for cementitious reactions to occur at low and sub-freezing temperatures when modified appropriately. Cementitious reactions are the same principle by which conventional lime and cement impart strength on subgrade soils. Naturally, straightforward research is required to transfer concreting technology to soil stabilization. To that end, an experimental campaign has been conducted. In particular, three chemical additives were selected and procured from the W.R. Grace Company, namely Polarset, Gilco and Daraccel. As a preliminary metric of performance, unconfined compression strength testing was conducted on mixtures with and without the chemical additives. At the levels tested, the range of additives increase the cost of conventional stabilization from approximately 10 to 50%, although subsequent research may lower those values still. The control (unmodified) mixture of soil and cement resulted in an average 1-day strength of 487.1 kPa (70.6 psi) when cured at 2 °C (35.6 °F). For the same mixture at the same curing temperature, the average 1-day strength increased to 1286.9, 1394.5 and 1079.2 kPa (186.5, 202.1 and 156.4 psi) for the Polarset, Gilco and Daraccel additives, respectively. These increased strengths at 2 °C are also approximately double that of the unmodified samples cured at 20 °C. These results are unique in that they represent the first application of cold-weather concreting technology to soil stabilization. While more work remains, these results suggest that the additives are promising.

A series of p-y and elastic 3-D finite element analyses have been performed to investigate the influence of limited soil improvement on lateral pile stiffness at soft soil sites. The purpose of the analyses is to provide guidance on the depth and lateral extent of soil improvement and improved soil properties required to significantly increase lateral pile stiffness. Modeled soil improvement methods consisted of treating the soft clay with cement and removal and replacement with dense cohesionless fill. Analyses with cement treated clay show lateral pile stiffness increased 2 times with a treated soil modulus 5 times greater than the soft soil and a depth and lateral extent of improvement of 2 and 5 pile diameters, respectively. Analyses with the dense cohesionless fill indicate an increase of lateral pile stiffness of at least 2 times requires a depth and lateral extent of soil improvement of 5 and 10 pile diameters, respectively. Limited soil improvement appears to be an economically viable option for improving foundation performance.

United States Highway 1 (US1) is known as the Overseas Highway in the Florida Keys and is the main road serving these islands. This 2-lane highway traverses the Jewfish Creek Drawbridge and is the main throughway for travelers to and from the Keys. Florida DOT, motivated by hurricane preparedness and public safety, is replacing the drawbridge at Jewfish Creek with a fixed bridge and widening the roadway north of the bridge: The widening scheme involves improving the soft underlying soils at the site with a technique known as dry soil mixing (DSM). Choosing the appropriate binder depends on the soil to be mixed, and in various applications can be a variety of materials including cement, fly ash, slag, lime, and other additives. Likewise, no single piece of equipment will be best suited for all applications and soil types, and for this reason mixing tools are often developed for individual projects. Additional design issues revolved around quality assurance and quality control (QA/QC). A unique program and testing methods were developed by project engineers to ensure that the product met the owner's requirements and project specific specifications. The DSM process is presented from the viewpoint of the engineer and the contractor. Topics of discussion detail the steps engineers followed to determine the applicability of DSM, and include design considerations, laboratory and field testing, reagent binder design, construction methods/procedures, QA/QC, equipment, and performance monitoring.

The conventional slurry trench method of construction to form low-permeability vertical barriers of soil-bentonite (SB), cement-bentonite (CB) and slag-CB have been widely used. More recently, in situ mixing methods to construct vertical barriers, such as deep mixing method (DMM) and the Trench Remixing and Deep wall method (TRD) have found increased usage. For the first TRD project in the United States, the method has been used for construction of test cells and is under consideration for the construction of a soil mixed barrier wall to prevent the intrusion of salt water into fresh ground water aquifers at the Alamitos Gap in southern California. To fully investigate the technology and its suitability for the application under consideration, laboratory and field studies (including construction and evaluation of two test cells) were conducted. This paper presents the results of various material formulations upon properties such as hydraulic conductivity and unconfined compressive strength. Laboratory studies demonstrated that mixtures of site soils containing saline pore water and blends of clay-water slurry, slag and Portland cement could be developed having measured values of hydraulic conductivity less than 1 x 10^–7 cm/s and being resistant to long-term degradation in the presence of the saline ground water.

A former sludge disposal basin at a Superfund site in Colorado requires the installation of a vertical deep mixing groundwater cutoff wall as a part of the overall site remediation. This barrier wall will be approximately 8,030 square meters and will prevent the flow of groundwater through the contaminated soil. The barrier wall will extend to a low permeability layer at 13.7 meters below ground surface (bgs). Soil mixing was selected based on consideration of cost, less exposure of the highly odorous waste material at the surface, and elimination of off-site disposal of waste soil. A bench scale compatibility study was performed using onsite soils and groundwater for a leachate to evaluate the impacts of the on-site contaminants to the performance and constructability of the soil mixing. This paper discusses the process for designing the study, selection of test parameters, selection of appropriate low-permeability reagents used in the testing, and the test results. In order to obtain a low permeability barrier wall (≤1 x 10^–7 cm/sec), several slurries were created and subjected to initial compatibility testing. Acceptable slurries were then made with the groundwater and long-term hydraulic conductivity tests were set up for two different mixtures (soil-bentonite and soil-cement). The reagents used in each mixture were: 1. Soil-Bentonite: SW101(salt-resistant bentonite)/lignosulfate thinner 2. Soil-Cement: Blast furnace slag/Portland Cement I-II/bentonite/soda ash. The ratios were set for each of these reagent materials to meet the required permeability while still providing a flowable slurry or grout for deep soil mixing. The test results indicate that the optimum mixture, based on cost and the lowest permeability, is the soil-bentonite mix.

This paper examines the cone penetration testing (CPT) results prior to and after two passes of deep dynamic compaction (DDC) to investigate compaction effects on the loose saturated granular backfill between steel sheet pile walls and peat bogs. The site characterization results show that the backfill was significantly densified after each pass of compaction. Two factors, (1) the existence of soft soil layers, and (2) severe liquefaction after DDC, greatly reduced compaction effects. The depth to which significant densification took place ranged from 6.5 to 9.5 m after the first pass of DDC, and 7.5 to 11 m after two passes. Finally, this paper recommends parameters for the evaluation of densification depth based on the comparison of the results from this study with those obtained from the literature.

Ground modification has been used on many construction sites to density granular material and reduce potential settlement or the susceptibility to liquefaction. This paper evaluates the benefits derived through several ground improvement methods, deep dynamic compaction, vibro-compaction, and vibro-replacement. These ground modification procedures were performed on several study sites with the effectiveness of the methods evaluated by comparing the tip resistance of pre- and post-construction cone penetrometer test (CPT) soundings. The results provided by the treatment methods were also evaluated with respect to the amount of fines observed in the subsurface profile. The results of our evaluation are intended to expand the availability of information on the effectiveness of these ground modification methods.

A laboratory study is performed to develop a model for evaluating the extent of the disturbance zone created by the pile jetting process. The physical phenomenon of jetting was observed in three different sands. Observations indicated that installation of piles by jetting is from the simultaneous erosion of sand beneath the pile tip and transport of soil particles through an annulus created around the circumference of the pile up to the ground surface. The pile advances only after a sufficient area has been eroded beneath the tip to cause a bearing capacity failure, as side friction is reduced due to the return water and liquefaction of sand within the annulus. Given equal jetting parameters, the extent of the debris zone from the pile center was approximately 100% further for sand with smaller average particle sizes (D₅₀ = 0.15 mm) than for sand with larger average particle size (D₅₀ = 0.5 mm). An empirical model to provide dimensional characteristics of the disturbance zone is proposed based on data from the laboratory study. Further research into the effect of jetting and water currents on disturbance zone is warranted including monitoring of field installations to verify the proposed model.

The Phase II works of Al Khaleej Village required the raising of the site above flood levels and preloading to improve the underlying deposits. At an early stage in the project a large body of gypsum was discovered that had to be removed from site. To offset this major impact on project costs, the area of preload was re-evaluated by use of the electric piezocone (ESPPT) in conjunction with settlement analyses. Confidence in settlement analysis procedures was previously obtained from back analysis of preloaded test areas. Subsequent evaluation of monitoring results during the main works identified the use of the Burland and Burbidge method (1978) as the preferred approach for settlement analysis in coastal areas of the Middle East. The other main conclusion was that evaporite strata in the region could not always be expected to reflect the well documented sequence of deposits generally found in sabhka deposits around the coastal regions of the Arabian Peninsula.

How do we effectively manage the volume of materials generated from the dredging of our waterways near our major coastal cities? A large component of this problem is the available storage volume. The coarse material in dredged material (DM) disposal basins can already be mined for use as fine aggregate due to the depositional processes used in its placement. However, poor quality soils (fines) continue to accumulate in disposal areas and will ultimately result in long term storage capacity issues. Recent work with the U.S. Army Corps of Engineers (USACE)-Philadelphia District has sought ways to utilize these materials in construction. While DM traditionally has been hydraulically placed in disposal basins, its legitimate beneficial use in fill construction is emerging. The successful use of DM in construction will be linked to the reliability of quality control methods associated with DM placement and compaction, DM variability, and knowing the properties of DM and its blends with other media such as crushed curbside-collected glass (CG). In addition to unfamiliarity with DM properties, not being able to respond to DM variability in the field is a major obstacle to its use in mainstream construction. Given the use of "families of curves" for compaction control in standard fill construction, our collaboration with USACE sought to compare scarce DM compaction data to the widely-used Ohio Department of Transportation (ODOT) family of curves. The ODOT curves are often used to allow field technicians to address incoming fill material variations by performing rapid one-point field estimates of the maximum dry density (MDD) to verify lab values or to adjust compaction criteria. This paper proposes modifications to the ODOT curves for use with typical DM and CG-DM blends enabling improved field quality control.

Rammed Aggregate Pier soil reinforcing elements are commonly used to reinforce compressible soils below shallow spread footings to control settlement and increase allowable foundation bearing pressures. Rammed Aggregate Pier elements are also used to support concrete floor slabs in lieu of deep foundations or massive excavation and recompaction. The piers reduce total and differential settlements because of their high strength and high stiffness characteristics. Due to the variation in pier stiffness with respect to in-situ soil stiffness, the commonly-used assumption of uniform sub-grade support is invalidated. The dissimilar slab support conditions, consisting of high stiffness at the pier locations and relatively low stiffness between the piers, leads to the development of bending moments and shear stresses within the slabs under applied loads. In an attempt to quantify the bending moment and shear stress conditions that develop in relatively thin structural slabs supported by Geopier Rammed Aggregate Piers, a series of finite element analyses were performed for various pier geometries and soil and pier stiffness conditions under uniform slab loading conditions. This paper presents the results of these analyses and provides design charts that may be used to estimate required concrete slab thicknesses for a uniformly distributed loading condition. Although it is recognized that a uniformly distributed loading analysis may not capture the critical load case for the design of the slab, this paper is of significance because it presents simple design charts that may be used to estimate preliminary floor slab thicknesses for a variety of span-to-depth ratios.

Design methods used to estimate the settlement below Geopier Rammed Aggregate Pier — supported footings rely on the performance and interpretation of a modulus test conducted on a single Rammed Aggregate Pier element. The test is used to determine the spring stiffness of a single element. However, actual footings are generally supported by multiple elements and the interaction effects between groups of elements are not simulated in the modulus tests. This paper presents the results of research performed to investigate the interaction effects between groups of Rammed Aggregate Piers installed at test facilities at the University of Utah and Iowa State University. The full-scale field test programs include load tests for individual elements, footings supported by groups of elements, and footings not supported by reinforcing elements. The results of the research demonstrate the increase in bearing capacity that is afforded by the installations of the piers. The results also indicate that a group interaction factor of 1.0 is appropriate for design. This paper is of particular significance because it is used to verify design procedures for a soil reinforcement method widely used in the United States.

Stone columns installed by the Vibro Replacement technique have been used in the United States specifically for liquefaction purposes since 1976, when they were first utilized for the Santa Barbara wastewater treatment plant. The method is ideally suited for seismic applications as it provides densification, reinforcement and drainage. Firstly, the depth vibrator repeatedly induces localized liquefaction in granular soils and achieves densification through vibrations and displacement. Secondly, crushed rock is used as backfill to form a stone column, providing a reinforcing element. And thirdly, the stone column can act as a drain to dissipate pore pressures and maintain the integrity of the stone column during a seismic event. Traditionally, performance specifications in terms of minimum or average SPT or CPT criteria have been used. However, as design peak ground accelerations have increased in many parts of the country, and finer grained soils get analyzed for liquefaction, it can be advantageous to specify Vibro Replacement stone columns to meet a given static, seismic or total settlement based on the post-treatment SPT/CPTs. A stone column foundation for a Costco store in Cypress, California, designed to meet a combined static and seismic differential settlement of 1 inch in 100 feet, is presented herein. Before and after CPT tests were used to compute deformations. Samples obtained from SPT tests were used for fines and clay content corrections.

This paper presents two case histories describing the use of Rammed Aggregate Pier soil reinforcing elements to stabilize slope failures. The analytical methods used in the design solutions are presented along with a description of the construction sequence and post-slide repair, instrumentation, and performance. This paper is of significance because it describes how a simple and cost-effective solution may be effectively implemented to stabilize unstable slopes.