Performance Confirmation of Constructed Geotechnical Facilities

Measurement of engineering performance is fundamental to empirical understanding, model development, and to the observational method. It is also expensive. Yet, how representative are field observations of the geotechnical conditions at work, and how informative are they of critical design assumptions? Interesting lessons for geotechnical practice are suggested from considering such issues, and from qualitative statistical reasoning about observing geotechnical performance on limited budgets. The discussion considers three interrelated topics: intuitive misperceptions about samples and sampling variations, the nature of uncertainty and randomness in modeling soil deposits, and simple lessons derived from statistical sampling theory.

Performance observations and measurements serve the same essential functions on geoenvironmental projects as they do on traditional geotechnical projects. The classical observational approach to geotechnical projects expounded upon by Terzaghi, Casagrande, Peck, and Whitman is as viable and necessary on geoenvironmental projects as it is on traditional projects. However, on geoenvironmental projects, there is often an added emphasis on observation and measurement for regulatory compliance and regulatory approval compared to traditional projects. Furthermore, the limitations and challenges of the classical observational approach associated with measurement of extremely small quantities, uncertainty with respect to mechanisms and phenomena of interest, and emotional issues related to project impacts may be exacerbated on geoenvironmental projects. Case histories are used to illustrate the role of performance observation and measurement on geoenvironmental projects. Monitored natural attenuation is discussed as the penultimate example of the application of the observational approach on geoenvironmental projects.

The Central Artery/Tunnel Project in Boston, Massachusetts is the largest, most complex highway project in American history. It involves the construction of 7.5 miles of highway through the center of the city with half the work occurring underground. Several conditions favored an extensive geotechnical monitoring program to provide engineers and contractors with up-to-date performance information. This paper describes the data acquisition and management processes used to collect, store, validate, reduce and report data from a large number of geotechnical instruments. It also describes some of the unique experiences presented by this challenging project.

Drilled shafts and other mixed or cast-in-place concrete deep foundation elements can be costly solutions. These foundations usually carry very high design loads, and often serve as a non-redundant, single load-carrying unit. These conditions have created a need for a high-level of quality assurence and control applied to each in-place constructed deep foundation element. The non-destructive testing method, Cross-Hole Sonic Logging (CSL), currently offers the most reliable technique for assessing the integrity of in-place constructed deep foundation elements. Recent years have seen progress in CSL instrumentation, taking advantage of the available computer technology. The software applications, however, have greatly fallen behind, thereby limiting the effectiveness and potential of the CSL method and deep foundations integrity testing in general. A new, original CSL testing system by the name of PISA (Pile Integrity Sonic Analyzer) makes use of an innovative software and data acquisition system, hence representing the state-of-the-art in deep foundation integrity testing. The PISA has the capability to show real-time graphical information during logging, including planar tomography, which can identify the boundaries of a compromised zone within the foundation element. The equipment operates completely in a Windows^TM graphical environment allowing alphanumeric and graphical reports to be generated directly into word processing software. The real-time graphical representation during logging and the ease of reporting enables immediate, extensive on-site evaluation and decision making. The PISA system was evaluated on several different construction sites. The obtained results from one particular site are presented, demonstrating the ease of use, accuracy of measurements and enhanced capabilities. The systems' abilities are shown to be superior to any other currently available commercial system.

The settlements of four pile supported mat foundations were monitored over a period of 14 years. The mats were supported by 88, 105, 706, and 708 piles. Pile lengths ranged from 17 m to 27.5 m. Load tests were performed on instrumented single piles to determine their load capacity and load distribution along the length. The subsurface conditions at the site consisted of 3m to 4.6 m of dredge soils fill underlain by overconsolidated, stiff to very stiff, clay, loose to medium dense sand and stiff to very stiff, overconsolidated clay. The observed settlements were compared with settlements calculated using four different theoretical methods. Based on the study, a simple correlation is proposed to compute pile group settlements using measured or computed single pile settlements.

Two 20 m long, strain-gage instrumented, closed-toe, 350-mm Monotube piles with a 7 m long bottom section tapering down to a 200-mm diameter at the pile toe were driven into a deposit of silty sand of glacial origin. The piles were concreted and static loading tests were performed—one pile was tested three days after completion of driving and concreting and the other after twenty-three days. The static test established the former pile to 2,000 KN. A maximum load of 2,500 KN was applied to the pile tested after twenty-three days without reaching the ultimate resistance of the pile, confirming the experience of earlier tests at the site that the pile capacity increases with time due to set-up. The actual magnitude of the set-up was not established. Both piles were instrumented with vibrating wire strain gages. The measurements showed that the pile Young's modulus of the composite steel concrete was not a constant but a linear function of the imposed strain. The measured axial strains were used to determine the non-constant Young's modulus of the steel-concrete cross section, enabling calculation of the axial loads in the piles during the test. The test data showed that locked-in load (residual load) acted on the piles before the start of the loading tests. The point of equilibrium (neutral plane) between downward and upward acting residual load forces on the pile was determined to be close to where the pile cross section changed from a straight section to a tapered section. Because of the presence of residual load, the measured loads indicated, falsely, that the soil resistance acting against the tapered section pile was very small. The analysis determined the magnitude and distribution of the actual shaft and toe resistances as well as of the residual load proving that most of the pile capacity was developed in the bottom portion of the pile agreeing well with effective stress principles. In the tapered section, at the maximum applied test load, the resistance was governed by fully mobilized positive shaft resistance and a movement-dependent "toe resistance component". The analysis results suggest that the set-up was due to a stiffening of the soil and not to increase of shaft resistance. The following effective stress soil parameters were established for the two test piles, Piles 2 and 3: beta-coefficient equal to 0.4 and N_t-coefficients of 60 and 75, respectively.

As foundation design becomes more structured by codes the question of axial strength definition becomes more important Most deep foundation elements, when tested to soil failure, do not exhibit a clearly defined failure load. But the design procedure requires that a specific ultimate capacity be taken from the load test results. The load test result is also affected by the procedure used in load testing and particularly by the loading rate. Loading procedures are discussed and suggestions are made for standardization of these procedures. Failure load definition is also discussed and some commonly used methods are reviewed. It is suggested that more consideration should be given to the requirements of the structural engineer and that load testing and load test result evaluation methods should be standardized.

As part of a research project investigating time dependent pile capacity gain, heavily instrumented test pile cluster was installed at a bridge reconstruction site in Newbury, Massachusetts. The test pile cluster consisted of (i) three-instrumented test piles, (ii) a surrounding ground piezometer field, and (iii) a complex data acquisition array to monitor the instrumentation. An extensive testing program, which included various static, dynamic, and Statnamic tests, was carried out over a lengthy time period to determine the pile capacity with time. Electrical resistance, piezo-electric, piezo-resistive, and vibrating wire instruments were used throughout the cluster to record acceleration, displacement, pressure, and strain within the test pile cluster. Primary test pile instrumentation consisted of an alternating pattern of piezometers and strain gages. Auxiliary instrumentation within the piles included accelerometers and total pressure gages. The ground piezometer field consisted of a total of 10 vibrating wire pressure transducers installed within normally consolidated and silty sand soil layers. Four separate data acquisition systems were assembled into an on-site data acquisition array to monitor the instrumentation throughout the testing period. This paper details the maximum loading conditions that each individual test pile was subjected to over the course of the testing program and evaluates the performance of the individual gage models and designs, instrument types, and data acquisition systems. Conclusions are drawn regarding (i) the reliability of the various gages and gage types, (ii) the performance of similar gage types, (iii) the selected measurement frequencies for each gage type, and (iv) the performance of the data acquisition array during the harsh conditions of driving and testing over time.

Recent advances in tactile sensor technology have produced very thin sensors allowing the simultaneous measurement of normal pressure at multiple points. Such sensors were used under the base of a laboratory model of a rigid strip footing (76.2mm x 609.6mm). The developed contact stresses are peaked along the centerline and lower along the edges. This distribution, visually observed for the first time, matches well the- anticipated theoretical distribution and is quite different than the even distribution routinely assumed in design. The insight obtained by the observed contact stress distribution can assist in a better geotechnical and structural economical design of shallow foundations.

Shallow foundation has a potential to be a successful alternative to deep foundations at many highway bridge construction sites. To further encourage their uses, comprehensive performance data must be compiled and made available to the practicing civil engineers. A research team at Ohio University has completed a major study on the field and laboratory performance of highway bridge spread footing foundations for the Ohio Department of Transportation and the Federal Highway Administration. Majority of the spread footings placed over the bearing soil layers were instrumented with sensors to measure their overall performance. At one of the five bridge construction sites involved in the study, the upper zone of the bearing soil layer received a few different treatments due to localized unique subsurface conditions and changes in payment agreement between the Contractor and the State. This paper presents detailed field performance of the spread footings at this site, which rested on bearing soil layers with different finish methods. Although the amount of data obtained in the study was still limited, the field gave a preliminary indication that the surface finish method might not have much influence on the field performance of these spread footings.

Four 3.05m diameter permanent steel-cased shafts were laterally load-tested over water in the Columbia River, downstream of Bonneville Dam, Oregon. The site investigation into the subsurface gravels, sandy gravels and near mudline silts included pre-bored pressuremeter tests. Design life lateral loads arrive from the pile supported concrete fish flume superstructure being submerged during periods of high river flows, producing large repeat loads to the shafts. The industry standard LPILE code was employed with user defined P-y curves constructed from Pressuremeter data to predict the first cycle load deflection response. Predicted mudline deflection & slopes were close to the maximum structurally tolerated by the flume, 25m above the mudline. Full-scale lateral instrumented load testing on the production shafts validated the analysis and P-y curves up to the maximum 2.22MN. A power law model was used to evaluate the repeat load cycles in the load test and establish the long term accumulated cyclic deflection. Optional foundation improvements were also examined, including a 12.2m diameter concrete Drag Collar designed from the Pressuremeter data.

A six-story parking garage was designed to be supported by 23 m (75 foot) long steel pipe piles driven through loose silts and silty sands. Rammed Aggregate Piers with shaft lengths ranging between 2.1 m and 2.7 m (7 feet and 9 feet) were proposed and accepted as an alternative to the driven piles. In this system, highly densified aggregate piers are incorporated within the upper weak and compressible soils resulting in a composite bearing material that is substantially suffer than the unimproved soil, and on which shallow foundations can be supported with tolerable settlements. Methods for analyzing and predicting settlements of footings supported on aggregate pier-reinforced soils are discussed. Design parameters were verified by the performance of aggregate pier modulus load tests. Total settlements of footings were estimated to be less than 4 cm (1–1/2 inches). Measured footing settlements were less than 1.3 cm (0.5 inch). The implementation of the aggregate pier system saved the project owner $187,000.

Upper Huia concrete gravity dam was constructed in Auckland-New Zealand, 70 years ago. The dam was not designed using present standards for uplift, although a concrete cut-off wall was constructed as a means to reduce the uplift pressures beneath the dam. This paper presents a description of the installation of vibrating wire piezometers to continually monitor the effectiveness of the foundation cut-off wall. The piezometer readings showed that the cut-off wall was providing the reduction in pore water pressures required for maintaining the stability of the dam.

In the mid-1980s, Saturn Corporation built manufacturing facilities of over 604,000 m² (6,500,000 ft²) on a 8620 ha (1300 acre) site in Spring Hill, Tennessee for a total construction cost of over $3.5 billion. Although the facilities were constructed on gently sloping terrain, site grading requirements mandated cuts of up to 9 m (30 ft.) and fills of up to 15 m (50 ft.). Design criteria required that the subgrade under the entire plant be of a uniform consistency so that equipment could be relocated anywhere within the structure without concern for poor performance of the slab and subgrade. Saturn authorized a geotechnical study that included the construction of test fills to establish engineering and construction parameters for the construction of the engineered fill. During the construction of the test fills, the source of the shot rock fill, the lift thickness, the moisture content, and type of compaction equipment were varied while the unit weight of the fill was measured and plate load tests were conducted. This paper presents the results of the experimental work conducted at Saturn to establish the engineering and construction parameters of the fill used in the construction.

This report summarizes the work of Dr. G.P. Gould who studied instrumented performance of Earth Dams during a 3 year stay at the Bureau of Reclamation. Settlement and pore pressures were measured in over 20 earth dams from a period of the 1930s through the 1950s. Detailed measurements were made possible by using multiple cross arm settlement columns and hydraulic pieziometer arrays. Compression records indicated most of the consolidation occurs during construction, and that for soils compacted dry of optimum, most of the compression occurs by reduction of air voids. From the compression data, typical compressions indices were correlated to Unified Soil Classification System soil types. Three phase media theory was used to develop equations to predict pore air and pore water pressure development in compacted soils during compression. Compressible soils compacted wetter than 0.6 % of dry of optimum moisture, exhibited undesirably high pore water pressure increases during construction. On dams where soils were compacted to an average 1.5 to 0.5 % dry of optimum, there were no significant excess pore pressures. This study and several related studies performed during the 1950s at Reclamation constitute some of the most important contributions to dam engineering. By using this information, earth dam designers can confidently design earth dams, and gradually, instrumentation has been reduced in new earth dam construction.

The impacts of slope creep on two residences in the Vail ski area were evaluated by floor elevation and wall plumb monitoring, and hillside stability was evaluated by subsurface exploration, multi-depth piezometers and inclinometer casing surveys. Exploratory borings identified a relatively deep, confined water bearing layer with piezometeric levels between about 2 to 4 m (6 to 12 feet) above the ground surface, and a possible basal shear surface between 7 1/2 and 13 m (23 and 42 feet) below the ground surface uphill of the residences. Subsurface drainage by use of low angle (horizontal and declined) drains was installed to depths of 76 to 105 m (250 to 340 feet) in a fan shape below the properties. The piezometer water levels responded quickly after drain construction with a drop of about 14 m (45 feet) uphill of the residences and between about 5 to 10 m (18 to 33 feet) adjacent to the residences. Monitoring of the piezometers, inclinometers and buildings has continued to evaluate the effectiveness of slope creep mitigation. Based on the relatively large drops in the piezometric levels, and the subsequent monitoring results, subsurface drainage appears to have significantly improved the hillside stability and stopped or reduced the slope creep movements.

Ten earth pressure cells and two pore water pressure transducers have been installed along the face of a 30 foot tall spillway sidewall to measure lateral earth and pore water pressures over time in the compacted clayey backfill. The spillway is a new labyrinth structure constructed as part of the Tongue River Dam Rehabilitation Project in Big Horn County, Montana. The sidewall is a counterfort wall that was assumed to be rigid for purposes of lateral loading estimation and design. Wall rigidity arises from stiffness offered by the counterfort design and the massive roller compacted concrete fill upon which the wall is founded. The portion of the wall instrumented was backfilled with a lean clay constituting the core of the dam. Compaction was achieved with both jumping-jack hand compactors used immediately adjacent to the wall and a vibratory sheep's foot roller. The wall was designed to resist lateral loads induced by the self-weight of the clay, compaction of the backfill clay, water pressure due to soil saturation and swell pressures in the clay. Data collected to-date show the development of earth pressure as compaction proceeds up the wall and the decay or relaxation of this pressure with time. Compaction induced pressures are compared to values calculated using several design solutions. Steady state earth pressures, measured after relaxation is essentially complete, are compared to at-rest values for this material. The time for relaxation to occur is discussed in light of design and operation considerations.

Embankment preloading, in conjunction with prefabricated vertical (PV) drains, was used to accelerate primary consolidation and to eliminate most of the anticipated secondary settlements of the soft clay foundation due to loading from a proposed wastewater treatment plant (WWTP). Subsurface exploration conducted during preliminary design indicated the presence of up to 60-m thick deposit of soft to very soft clay underlying the site. Estimates of settlement due to filling of the existing lagoon and construction of the WWTP indicated that the soft clay deposits will undergo up to 150 cm of settlement. Approximately 80 cm of this settlement was estimated to occur in the first 10 years of the plant operation and an additional 60 cm was estimated to occur during the following 30 years. UP to 19 settlement plates were installed under the embankment fill to monitor the preload performance. The measured settlements were then used to analyze the behavior of the soft clay foundation, and to develop recommendations regarding the time for preload removal and subsequent construction of the proposed WWTP. This paper discusses the results of settlement analyses and highlights the procedure that was used to predict the magnitude and time-rate of settlements based on field measurements.

For reasons of economics and tradition, geotechnical engineers obtain a limited number of data points for most parameters. Well known results from the theory of statistics indicate that uncertainties associated with the small number of points place severe limits on the inferences that can be drawn from the data. This has implications for the design of field instrumentation programs and for the use of field data in verifying computational models. A model of the strength of mine pillars, based on actual data but simplified for present purposes, illustrates the effects.

This paper highlights and discusses important statistical considerations in calibrating performance models with field data. These considerations include realistically modeling the mean and variance of the data, incorporating multiple sources of data in the calibration, and accounting for uncertainty in the calibrated model when making new predictions with it.

Geotechnical problems are unique and hard to pin down using exact mathematical models. In such environment there are numerous sources of ambiguity and imprecisely defined data. Conventionally, in the construction of mathematical models, the uncertainties are treated as random variables. Probability theory is perceived as the unique methodology to handle uncertainty in these parameters. Nevertheless, our preliminary investigations with geological and engineering geological maps, bore holes, field tests, sampling, laboratory work up to the analysis and design, are all combined with vague, indeterminist and to a great extent non statistical data. The application of decision theory and stochastic models in such an environment is linked up with many drawbacks and perplexing features. In this paper, the reliability evaluation of measured properties will be investigated applying the fuzzy- reliability techniques. Within this conception setting, the measured parameters are rendered into fuzzy variables and the structural safety is characterized by fuzziness. The procedure is proposed as an alternative or complementary postulation of reliability in geo-engineering. This approach will be illustrated with computational algorithms and a numerical example.

This paper discusses the critical speed (the `bow-wave' effect) and how it is influenced by soil stiffness. In order to reach the centers of major conurbation, modern high- speed rail lines are increasingly utilizing marginal land avoiding previously built housing areas. This land often has low soil stiffness. The critical speed of waves across such soils can approach the speed of modern high-speed trains. The large deflections caused can increase environmental noise and the track maintenance required and decrease ride quality. This paper discusses how the effect of train speeds can be predicted using computer modeling and an instrumented impact device. The modeling of possible remedies to the problem are discussed. A convolution teclmique is used to model axles travelling at different speeds and the effects of multiple axles. The resulting displacements are highly dependent on the amount of damping introduced into the system modeled. In-situ testing, laboratory testing and finite element modeling are compared and are shown to provide consistent results. The conclusion is that a powerful analytical tool for the prediction of the `bow-wave' effect has been devised.

"Capillary barrier" is the term for a layer of large-pored material (soil and/orgeosynthetic) placed in a finer-pored soil to 1) reduce unsaturated flow of soil water upwards and/or 2)reduce or prevent water from infiltrating from the overlying fine-pored unsaturated soil into the soilbelow the capillary barrier. In the latter case, the capillary barrier is sloped so that the infiltratingwater flows downwards along the interface between the capillary barrier and the overlying soil. Thesame physical principles govern the behavior of capillary barriers in both cases. This paper reviewscapillary theory and what is known about capillary barriers (made of soil and/or geosynthetics) in avariety of applications. This information is then applied to the potential use of geosynthetic capillarybarriers in pavements to reduce frost heave and to divert infiltrating water. The use of a geocompositecapillary barrier drain (GCBD), consisting of a drainage net sandwiched between geotextiles thatpossess certain hydraulic properties, appears promising. In addition to reducing frost heave, a GCBDmay also help drain the soil above it before it becomes fully saturated. It may also be designed to drainsaturated pavements at critical times of the year, such as spring thaw.

There is a broad interest and certain expectations in knowing the in- place engineering and performance properties of the compacted or stabilized soil. This has been the case since soil was thought of as an engineered structure. The processes of removing small numbers of specimens and then conditioning the specimens to represent the actual site are estimates at best. A great deal of skill is required to accurately prepare the specimens to approximate field conditions and provide meaningful and believable laboratory test results. Density based measurements have been surrogates for material modulus and structural stiffness. Available tests to measure the in-place properties are cumbersome, destructive and time consuming. Recently, a new device has been developed that does what the geotechnical community has desired, measure the in-place engineering and performance properties of the soil structure. Conceived and developed partially by funding from FHWA, it is now commercially available. Fast and immediate measurements of stiffness and modulus allow the compaction process to be controlled by the contractor without delay or outside feedback and should provide a more uniform structure. Control of the compaction process via stiffness and modulus permits construction to performance specifications derived directly from design. The use of one test method to link design, specification, construction, and maintenance appears to complete a loop that ultimately enables contractor warranties. This paper will examine the use of the device and the practical implementation of various functions or processes that the device offers.

Intrusion of surface aggregate into the underlying soil of hike and bike trails creates undesirable conditions for trail users and exacerbates the problem of trail rutting. A study is underway to evaluate the effectiveness of geotextile separators for rails-to-trails applications. Non-woven needle-punched and spunbonded geotextiles were placed between the subsoil and the aggregate surface in three test sections along a trail. The test sections are being monitored for visible intrusion of aggregate into the subsoil, drainage of surface water off the trail, and enhanced performance and durability of the aggregate surface compared to control sections. Destructive samples of the geotextile separator and aggregate were taken and tested periodically to evaluate field performance of the geotextiles. The characteristics of the surface aggregate and geotextile separators, installation of the three test sections, and results of performance monitoring and laboratory testing activities are described herein. The findings indicate the geotextile separators to be performing well and the test sections have maintained a well-drained, smooth surface after one and a half years. The performance of these sections vastly exceeded that of the control (no separator) sections.

On a rural road near Ada, Oklahoma, four test sections were constructed with treated subbases. Treated subbases in three sections were constructed using Portland cement kiln dust (CKD) from three different sources and one was constructed using granular quick lime. The purpose of the study was to compare the effectiveness of the three CKD sources and granular quick lime for modifying and stabilizing moderate to highly plastic soils. To evaluate the effectiveness of the various additives, treated soil was collected immediately before compaction to prepare unconfined compression test (UCT) samples that were used to assess strength gain after 7 and 28 days of curing. In addition, dynamic cone penetration (DCP) tests were conducted at different times after compaction, and falling weight deflectometer (FWD) tests were conducted on the finished flexible pavement. A comparison of results indicates there is agreement between the trend of cone index (CI) values from DCP tests and unconfined compression strength data along the roadway alignment. Furthermore, the DCP results give an indication of the variation in strength with depth, an advantage that the other forms of testing lack. Unconfined compression test results and DCP data do not correlate as well with the FWD data, apparently because the FWD results are less sensitive to differences in the treated subbase moduli for each section. This study shows that a reliable comparison of stabilized subbase courses for different test sections requires field testing to account for variations in field compaction and soil type, and laboratory testing to assess the relative effect of the soil additives on strength.

The layer moduli of track substructure components are important factors in the evaluation of track condition and estimation of maintenance costs. Layer moduli can be estimated from in-situ tests such as the cone penetration test (CPT); however, for layers of open graded aggregate, such as railway ballast with large individual particles, the results of the CPT are significantly influenced by individual particles and may not reflect the behavior of the mass. A method of evaluating layer properties using downhole measurement of layer deformation is presented in this paper. Tests were conducted on track using downhole measurements of layer deformation under both quasi-static test loads and passing traffic. The applied loads were measured by instrumenting the track superstructure. The top of rail load and substructure deformation data were used with a geotechnical model of track to backcalculate the moduli of individual layers. The process is similar to the procedure for processing data from the falling weight deflectometer; however, data from actual traffic can be evaluated. The results are compared to the CPT results from the same site and also to expected results for similar materials.

Verification of the success or failure of grouting is rarely possible as an afterthought when a project is complete. Also, verification often is difficult and expensive. If potentially effective methods of verification are considered right along with the project and its objectives during the planning stage, there is a good chance of coming up with a reasonably effective verification scheme. The paper outlines example projects where performance was its own verification. It also covers areas where direct measurement and monitoring can prove effectiveness, where use of tracer materials will track performance, and where locating the grout seems to be the only way to ascertain performance and verify quality. Two projects are described where complex and combined methods were used, methods which showed some of the possibilities which exist if the proper fact gathering is designed into the system from the start.

One of the most common complaints about grouting is that quantities cannot be estimated in advance, to a reasonable level of accuracy. Some feel that cost of grouting always overruns the estimate and by a lot. A second common complaint is that after a significant sum of money had been spent it became obvious that the type of grouting used just was not going to work. Both of these complaints originate with the planning of the grouting program. Planning begins with the subsurface geotechnical investigation. A good investigation is essential to the success of a grouting project. While the contractor's execution of the contract is a significant factor, a reputable and knowledgeable contractor will be more likely to be successful with a well written specification and appropriate and sufficient subsurface information. For a design engineer to write a good specification, knowledge of the appropriate grouting methods and thorough subsurface information is essential. This paper will focus primarily on grouting in soil but may have application in rock grouting.

Verification of jet grouting to stabilize an existing stone masonry abutment by increasing the abutment's bearing area and mass was necessary to assure proper performance. Project constraints such as: maintaining bi-directional vehicular traffic over the bridge, maintaining uninterrupted rail traffic under the bridge, maintaining and protecting eight underground utility lines through the work, and ensuring safety of the bridge abutment, the roadway and nearby structures from the high-pressure grout limited the techniques and verification methods available. The jet grouting created a soilcrete mass to enlarge the existing abutment by overlapping individual grout columns interlocked with the existing stone masonry. The jet grouting was completed under continuous full-time observation by the design engineer to verify compliance with the intent of the project specifications. To verify that the grouting was not damaging to the existing structures, the abutment, roadway and sidewalk were instrumented to continuously monitor movement during jet grouting. Test columns were constructed prior to production grouting using feeler pipes to assess grout coverage, column diameter and select the proper jetting pressure, grout flow rate and monitor extraction speed. Core samples were taken to verify the integrity and properties of the soilcrete. Laboratory density and compression tests were performed on selected core samples of the soilcrete. All tests exceeded project requirements. The core borings through the soilcrete mass met the project's core quality requirements and demonstrated the effectiveness of the jet grouting despite some unanticipated rubble in the abutment backfill.

The Nova Dutra project consisted of the design and construction of Segmental Retaining Walls (SRW) over very soft (0 to 1 blow count) organic silt and clay deposit 9 m in depth. The walls were needed to assist with the widening of a main transportation route leading to Sao Paulo, Brazil. The SRW section which was monitored is 8.2 m in height and approximately 150 m long and rests on a geogrid reinforced load transfer platform which in turn is supported by jet grout columns 1.2m in diameter spaced 3 m center to center. The soil used in the reinforced zone consisted of a low plasticity fine-grained soil with an effective friction angle of 26° and unit weight of 19 kN/m³. Monitoring methods included of piezometers within the reinforced zone, settlement plates above and below the load transfer platform, and survey points on the face of the SRW. This paper discusses the selection of instrumentation and results of monitoring. The measured results indicate that the SRW performed well.

Obtaining accurate and timely subsurface information during site characterization and construction phases is a key to successful completion of constructed geotechnical systems. Recent technological advancements in computers and algorithms have resulted in seismic tomography methods becoming a standard engineering tool during geotechnical investigations for highways, bridges, and tunnels. The RockVision3D^TM seismic tomography system, developed by NSA Engineering, Inc., is a high-tech "through-the-earth CAT scan" that provides on-site true three-dimensional images of subsurface conditions that exhibit a significant change in seismic velocity or attenuation. The system uses newly developed algorithms for data processing and provides images with greater detail and accuracy than traditional tomography methods. The system has been used worldwide to characterize the subsurface and to evaluate the performance of various constructed support systems. This paper presents case studies of the RockVision3D^TM applicability in geotechnical site characterization and design confirmation for the following civil projects: 1) Location of a rock tunnel beneath an interstate highway in Colorado, 2) Detection of abandoned coal mine entries beneath a retaining wall foundation in Illinois, 3) Evaluation of the effectiveness of grouting a karstic limestone bridge pier foundation in Missouri, and 4) Assessment of highway settlement causes and determination of the effectiveness of remedial grouting in Colorado.

A case history is presented in this paper for an excavation in Boston marine clay supported by soil mix walls and earth anchored tiebacks. Large lateral deformation occurred when the excavation reached a depth of 13.4 m. Excavation unloading combined with the dissipation of excess pore pressures is shown to account for undrained shear strength reduction and the observed excavation performance. Design practices are examined in light of this case history, and the lessons learned are used to recommend improvements in evaluating the base stability of deep clay excavations.

The Exploratory Studies Facility is currently under construction at Yucca Mountain, Nevada. The purpose of the Exploratory Studies Facility is to characterize rock strata for its suitability to construct a nuclear waste repository. At the current stage, a considerable amount of underground work has been performed providing direct access to the rock mass interior. This access facilitates verification of the initial estimates of the magnitudes of rock strata parameters and allows for evaluation of the ground support system(s) performance based on the field-acquired data. As mandated by federal law (10 CFR 60), the engineering design at the Yucca Mountain Site Characterization Project must be supported by evaluation of the performance of the underground openings based on field-acquired evidence. This paper presents a discussion of selected elements of the Performance Confirmation Program that were designed to provide verifiable evidence of the in situ performance of the underground openings and ground support system through measurements performed in the field. The two aspects of the Performance Confirmation Program are presented, namely, the results of ground deformation monitoring and the results from strain gage measurements of steel set loads. The deformation monitoring analysis presents a semi-empirical approach for analyzing the initial, unmeasured portion of the tunnel deformation, and for predicting future opening displacements. The steel set loads analysis presents the development of a numerical model that provides the ability to examine various steel set support concepts and to optimize the performance of the steel set given the operational constraints in the field.

The analyses and performance of a soldier-pile-tremie-concrete (SPTC) wall with a significant unbraced length is presented. The wall was constructed as part of the excavation support system of the Ramp D tunnel jacking pit built for Contract C09A4 of the Central Artery/Tunnel Project in Boston, Massachusetts. This wall is unique in that it supported a cut with a height of approximately 19.5 m (64 ft) with only two brace levels above the subgrade with a clear span between the subgrade and lower brace level of 14.3 m (47 ft) above the excavation subgrade. The excavation subgrade consisted of Boston Blue Clay, the top 6.4 m (21 ft) of which was improved by jet grouting within the footprint of the jacking pit. The SPTC wall was also designed to resist train loads located about 6.1m (20 ft) behind the wall. The SPTC wall was socketed into the Boston argillite bedrock. Other prominent features of this support wall include the structural details of the two internal braces. The upper level brace consisted of a horizontal steel truss, while the lower level brace was a 4.88-m(16-foot)-wide, 1.22-m(4-foot)-deep post-tensioned reinforced concrete beam. Details of these two braces are described herein. The behavior of the wall was predicted using SOILSTRUCT during the design phase. A comparison of the predicted and measured wall deflections is provided.

Three geogrid-reinforced, modular block retaining walls were constructed as part of a terminal expansion project. The largest of the three walls, designated the primary wall, is nominally 4.7 m tall and 541 m long. It was designed to withstand relatively high, live, surcharge loading (48 kPa) and Zone 2 seismic loading. Construction of the walls was completed in mid-1992. The new container storage area was completed and placed into service in 1994. Presently (mid-1999) the walls are in excellent condition with no reported performance problems. However, the walls have experienced no seismic loading and only light surcharge loads have been applied behind the top of the primary wall.

The Drift Scale Test (DST) is an integral part of the U.S. Department of Energy site investigations program to characterize Yucca Mountain and determine its suitability for the permanent disposal of spent nuclear fuel and high-level nuclear waste. The DST is a large-scale, long-term in situ thermal test consisting of a 47.5-meter long drift, 50 wing heaters, and 9 floor heaters. Thermal- mechanical-hydrological-chemical (TMHC) responses of the rock are measured by approximately 3,500 sensors installed in 147 boreholes drilled into the test block. The heating phase of the test, which started on December 3, 1997, is scheduled to continue for four years. It will be followed by a cooling phase of similar duration. During the heating phase, temperature in approximately 10,000 cubic meters of rock will exceed 100 °C. The primary objective of the DST is to gain a more in-depth understanding of TMHC processes anticipated to exist in the local rock mass surrounding the potential repository at Yucca Mountain. The numerical simulations of the DST is an integral part of the development and enhancement of coupled-process models. The following three types of models were used in the analyses: thermal-hydrological (TH), thermal-hydrological-mechanical (THM), and thermal- hydrological-chemical (THC). These analyses and ongoing comparisons to the measured responses have advance the understanding of the complex coupled processes anticipated in the host rock at Yucca Mountain.