Use of Geophysical Methods in Construction

A variety of remote sensing, surface geophysical, borehole geophysical and other non-destructive methods can be used to determine conditions of roads and bridges. Some of these methods can be utilized to determine subsurface conditions prior to construction. Some can be applied to QC measurements during construction and many can be applied after construction to determine as-built conditions, as well as degradation. The benefit of such measurements are that they are non-destructive, provide in-situ measurements of a wide range of physical properties, sample larger areas or volumes, provide continuous measurements in some cases, and provide faster measurements. All of this results in a greater sample density, which can more readily identify uniform conditions as well as locate anomalous conditions. Once anomalies conditions are identified, those areas requiring further tests, borings or repairs can be accurately and quickly located. These methods can also provide temporal measurements (detecting changes in conditions with time). Such data can be used in a database to guide management decisions for maintenance and repairs. Maintenance and repairs on roads, bridge decks and bridge scour is becoming ever important as our infrastructure ages. A variety of methods can be used to increase the efficiency and technical effectiveness of such inspections. This paper provides an overview of those methods that can and have been applied to characterizing conditions of roads and bridges.

Time Domain Reflectometry (TDR) is a promising new technology for control of compaction now that a procedure has been developed to include density measurement in addition to water content measurement. The TDR method employs the transmission of electromagnetic waves in a "cable" of known length where soil is the dielectric medium between the leads of the cable. Because the length of the "cable" is known, it is possible to determine the dielectric constant of the soil. The dielectric constant is strongly affected by the water content and density of soil but is relatives unaffected by soil texture and salt content. A simple equation relates dielectric constant to density and water content. By conducting two tests, one insitu, and one on the same soil compacted in a mold, both water content and density can be determined as long as the water content of the two specimens are the same. The testing procedures are straight forward and simple. Specially designed equipment makes the test practical and robust. Examples show that accuracy is quite good. Ongoing work is examining applicability of this method for soils with large particle sizes and for soils with high activity, additives, and stabilizers. This paper also previews some ongoing developments of theoretical modeling that allow for a more complete understanding of electromagnetic wave propagation in soil.

Ultrasonic methods implemented in various types of integrated seismic devices can be successfully used in quality control and condition assessment of bridge decks. Of special interest are three ultrasonic techniques: ultrasonic body-wave (UBW), ultrasonic surface-wave (USW) and impact echo (IE). While the first two are used in a deck material characterization, the IE method is primarily used in evaluation of a corrosion induced delamination. The biggest advantage of the IE method over a curent practice of chain dragging is that it allows detection of zones of delamination at various stages: from initial to progressed and developed, thus enabling better prediction of deterioration processes in the deck. To improve automation, accuracy and interpretation of results of the IE technique, the testing is simulated by the finite element method. Numerous finite element results confirm the ability of the IE method to accurately measure theposition, extent and the stage of delamination. The finite element method was successfully used in simulation of two probable scenarios of delamination progression: expansion\growth of a single small delamination, and progressive linking of several smaller delaminations. Elements that can be used in recognition of the two processes are described and can be used in long term monitoring and prediction of deterioration processes. Issues related to finite element modeling of wave propagation in bridge decks and finite element analysis results interpretation are discussed too.

The electrical resistivity method, which is one of the most widely practiced geophysical prospecting methods, has in the past usually been applied to groundwater and underground resource exploration. Its scope has recently been extended to civil & environmental areas because of new developments that allow it to image, underground structures effectively. The applicability of the electrical resistivity imaging techniques to civil and environmental problems has been demonstrated in two case studies in Korea. First, the electrical resistivity surface imaging technique was performed in a tunnel construction site, to map the underground geological structure of a large area in a short period of time. The dimension of the weak zones was estimated and the support pattern of the tunnel was adjusted. The result was a successful excavation of the tunnel. Second, electrical resistivity tomography (ERT) was implemented in a research tunnel site to monitor the movement of brine after it was injected in a test borehole. A pair of low-conductivity zones which was interpreted as the pathway of the groundwater flow was clearly imaged.

Geotechnical engineers have struggled for years with the difficult task of characterizing geologic conditions that affect the design and stability of various transportation infrastructure, including road cuts, tunnels, and bridge abutments. Commonly, designs are based on data from borings that may or may not reflect the true character or three-dimensional continuity of the materials present and may also not detect the presence of localized structural defects (faults, shear zones, slip planes) or previous excavation activity. Geophysical techniques such as seismic reflection, seismic refraction, and ground penetrating radar have been used with some success to expand the knowledge of subsurface conditions prior to excavation however; these techniques often interfere with the construction process and are time-intensive in terms of data collection and interpretation. To improve the quality, continuity, and tirneliness of geotechnical data for site characterization, NSA Engineering, Inc., has developed a true, three- dimensional tomographic imaging system known as Rock Vision3D^TM. This system is based on the principle that acoustic waves travel through different materials at different speeds and with different attenuation rates. By recording the waveforms from known acoustic sources, three-dimensional tomograms inferring the relative competence of the materials that the waves travel through can be constructed. Although tomographic techniques are not new, advancements in the technology that allow for near-real-time imaging, make tomography a practical tool for geotechnical site characterization, both in the design stages and as excavation proceeds. With the expanded subsurface information that volumetric tomography provides, the risks and expenses associated with change orders are greatly reduced. This publication describes tomographic imaging as implemented in the RockVision3D^TM software package. Included in this discussion are applications of both seismic velocity tomography and seismic reflection tomography systems. Although attenuation tomography also can provide information on subsurface conditions, especially water saturation, we limit our discussion here to velocity and reflection tomography. Several case studies are presented describing applications of volumetric tomographic imaging for highway and tunnel site characterization.

The spectral-analyses-of-surface-waves (SASW), developed in the early eighties, has constituted an important step, in the use of surface waves for definition of shear wave velocity profiles without intrusion. The SASW testing procedure was designed to minimize the contribution of higher Rayleigh modes and thus assumes that the average dispersion curve is representative of the fundamental Rayleigh mode. A number of numerical studies have however demonstrated that this basic assumption is valid only when the shear wave velocity varies regularly with depth. This study demonstrates, based in numerical simulations and an experimental case, that the multi-Rayleigh mode problem can occur in various situations even if the shear wave velocity increases regularly with depth. Techniques are proposed in order to identify and separate the energy of higher modes. The applicability of those techniques is demonstrated with simulated and experimental cases.

This paper investigates the feasibility of site characterization via spectral analysis of horizontally-polarized shear (SH) waves as a means to filter the effects of pore water for saturated soil conditions. Using a multi-layered wave propagation model to simulate the proposed testing configuration revolving around a torsionally vibrating loading plate, it is shown that the use of generalized Love (i.e. SH body and surface) waves as a sounding tool precludes the influence of the soil's Poisson's ratio, thus reducing the number of parameters relevant for seismic data interpretation and ruling out the existence of compressional waves. Preliminary numerical simulations indicate that the dispersion curves for generalized Love waves may be sufficiently sensitive to stratigraphic variations to serve as a basis for the development of an SH alternative to the spectral analysis of surface waves commonly associated with vertically-polarized ground motion.

Seismic wave propagation methods provide in situ wave velocity profiles from which various elastic moduli can be determined. The tests are applicable to a wide range of civil engineering materials, including soil and portland cement concrete (PCC). Wave velocities obtained from seismic testing can also be used to calculate other salient design properties such as Poisson's ratio. Although the accuracy of seismic surveying is well researched and accepted by the engineering community, little research has been reported on uncertainty or precision of parameters obtained from the tests. In this study, a large sample of replicate field data was collected from multiple borehole and surface tests. Measurements of vertical (SV) shear waves were obtained at two independent soil sites. Seismic surveying was conducted at one PCC site to obtain measurements of compression (P) and Rayleigh waves. From this data, variation in seismic wave velocities collected from typical seismic testing was determined. In addition, variation in seismic wave velocities was used with error propagation techniques to evaluate uncertainty of Poisson's ratio. Based upon data presented within this study, coefficients of variation of SV, P, and Rayleigh wave velocities indicate a high degree of precision for these velocities. Also, the coefficient of variation for Poisson's ratio attests to the high precision of determining this property with seismic surveying techniques.

Traditionally, nonintrusive techniques used to characterize soils have been based on P-wave refraction/reflection methods. However, near-surface unconsolidated soils are oftentimes water-saturated, and when groundwater is present at a site, the velocity of theP-waves is rnore related to the compressibility of the pore, water than to the matrix of the unconsolidated soils. Conversely, SH-waves are directly relatable to the soil matrix. This makes SH-wave refraction/reflection methods effective in site characterizations where groundwater is present. SH-wave methods have been used extensively in site characterization and subsurface imaging for earthquake hazard assessments in the central United States and western Oregon. Comparison of SH-wave investigations with geotechnical investigations shows that SH-wave refraction/reflection techniques are viable and cost-effective for engineering site characterization.

Using a program of mostly non-destructive-testing, the foundation capacity of a nearly 100-year old bridge was determined at a fraction of the cost of what a conventional approach would have cost. Initial proposals to determine the foundation size, depth, and integrity using an invasive/destructive approach of test pits, borings, and coring had substantial costs. The municipality accepted an alternative proposal to perform a combination of non-destructive testing and limited destructive testing. The program incorporated Ground Penetrating Radar (GPR), dispersive technology, and limited borings. GPR was used to determine the reinforcing in the abutments, thickness of the abutment, and if abutments and piers had footings. Dispersive wave technology was used to determine the thickness of the abutment, the depth of abutment, and the general quality of the concrete. Conventional borings were done to sample the concrete, calibrate the NDE, and determine the founding materials. Using this combination of complementary NDE and limited destructive testing, very few assumptions were necessary to determine the foundation capacity and integrity of the nearly 100-year old foundations at a fraction of the cost of an invasive/destructive evaluation.

Three different types of seismic techniques were tested to determine the lengths of pile(s) and drilled shafts of bridge bents in overconsolidated clay soils in the Houston, Texas, area. These methods were the seismic-wave reflection survey (SWRS), the transient forced vibration survey (TFVS) and the parallel seismic survey (PSS). They were conducted at bridge bents located along I-45, southeast of downtown of Houston, and on existing half-scale foundations at the University of Houston National Geotechnical Experimentation Site (NGES-UH). The purpose of the tests was to ascertain whether any or all of these methods would be reliable in determining the lengths of old or poorly documented foundations whose lengths are not known in overconsolidated clays of the type typically found in Houston, Texas. All three methods measured the depths to the toes of pile groups or individual drilled shafts to an accuracy acceptable for most bridge maintenance purposes, within approximately 0.3 m (1.0 ft) – 1.1 m (3.6 ft) of the actual pile lengths shown on the as-built plans. The most accurate prediction was achieved with the SWRS test, while the least accurate prediction was achieved with the TFVS test.

This case study describes the use of apparent conductivity surveys to determine the location and extent of voids created by coal mining in the shallow subsurface. The use of apparent conductivity was undertaken as part of the construction of a new cell at a regional solid waste landfill located in southeast Kansas. A. 1 meter thick coal seam was present at a depth of approximately 9 meters and had been mined in the early 1900's by both the strip and "room and pillar" methods, leaving many "rooms" in the subsurface. Excavation of the material above the rooms posed a significant risk to equipment and personnel, and the presence of these large voids beneath the landfill slopes raised serious slope stability concerns. Given the nature of the facility, voids within the slope could have caused a loss of support leading to excessive liner deformation and could also have provided a pathway for contaminant transport. It was therefore necessary to confirm that all of the voids beneath the slopes were identified for remedial action. An apparent conductivity survey was determined to be the most appropriate method for locating the voids and determining their extent. Measurements were taken along the slope using coil spacings of 10 and 20 meters. Based on the resulting data, it was determined that the presence of voids caused measurable changes in the apparent conductivity of the subsurface, and that all of the voids present beneath the slope had been identified. The use of apparent conductivity was considered so successful that a more extensive survey was performed over the remaining 35 undeveloped acres of the property. This survey detected a number of apparent conductivity lows, indicating that much of the remaining portion of the property had been mined. This survey enabled the owner and contractor to develop an appropriate excavation procedure prior to development.

In this paper an analysis is presented for the field observation of large amplification in vertical motion at a reclaimed site subjected to the 1995 Kobe earthquake. The observation is considered to be associated with the amplification of P-waves caused by incomplete saturation of shallow soils. Saturation effects of soils on P-wave velocity and Poisson's ratio are discussed to show evidence of partial saturation in shallow layers at the array site. An analytical study is described to demonstrate substantial influence of saturation on the amplification of a soil layer to vertical earthquake excitation. Furthermore, a numerical simulation of the vertical motion at the realistic site considered is performed by incorporating saturation effects of shallow soils. The present study indicates that vertical motions may significantly be affected by saturation of surface soils, suggesting that one may need to carefully examine this soil condition in the study of vertical site amplification.

A method utilizing surface refraction seismic geophysical measurements to estimate earthwork factors involving large rock cuts in weathered granites and associated embankments constructed from these materials has been developed. Estimating overall unit weights of a large volume of in-situ rock mass, including fractures or other discontinuities and very weak material that cannot be adequately sampled by standard coring methods, is a major feature of this earthwork factor method. The method uses Percolation Theory (PT) concepts of modulus porosity relationships to relate rock mass modulus values to rock mass porosity values. Surface refraction seismic measurements are utilized to estimate rock mass low strain modulus values. Recovered rock core unconfined compressive strengths are used to estimate rock particle modulus values and unit weights are used to determine corresponding porosities. The PT modulus-porosity relationship is shown to be consistent with the testing of recovered rock core and with test results of weathered granites reported in the literature. Estimates of unit weights of rock masses to be excavated are made based on the measured seismic velocity profiles through project rock cuts. Estimates of embankment fill unit weights are made using seismic velocities obtained through existing fills and from moisture-density relationship (Proctor test) results of materials sampled on the projects. Earthwork factors are determined from the ratios of these in-situ rock mass unit weights and existing or anticipated embankment fill unit weights.

An abandoned coal mine under Interstate 70 in eastern Ohio collapsed and created a large crater along the highway a few years ago. A myriad of unnmapped features under the State's roadways threatens similar localized collapses. These catastrophic failures may cause loss of life and always cause traffic disruption and require extensive repair expenses. Identifying the zones, which are likely to fail, would allow remedial action to prevent future failure. Methods to locate these zones need to be inexpensive enough to allow widespread use but focused enough to target very specific problem areas. Geophysical survey techniques have the potential for meeting these requirements. The principal objective of the present research study was to develop a reconnaissance geophysical survey technique to delineate the zones that can be further investigated by borings or by localized geophysical studies to determine the nature and probability of significant failures. Accordingly, several different geophysical techniques were adopted during the summer of 1998. The results obtained from this study show that seismic refraction, resistivity imaging and surface wave profiling have the most potential for meeting the project objectives and should be studied further in phase two.

Boreholes, test pits, and geophysical seismic refraction surveys were used to estimate quantities of soil excavation and rock requiring ripping or blasting for site grading at a power plant project construction site. These quantities were compared to the actual excavated, ripped and blasted quantities measured during construction, including comparison of the relative positions of the 3D surfaces generated using the results of the different test methods and the actual measured surfaces. This paper presents the results of the comparison, and discusses the relative merits of the site investigation methods used to develop the quantity estimates.