Report of Investigations No by Jeffrey G. Paine and Michael R. Murphy*

Transcription

1

2 Report of Investigations No. 259 PAVEMENT DEFLECTION AND SEISMIC REFRACTION FOR DETERMINING BEDROCK TYPE, DEPTH, AND PHYSICAL PROPERTIES BENEATH ROADS by Jeffrey G. Paine and Michael R. Murphy* *Design Division, Pavements Section Texas Department of Transportation 125 East 11th Austin, Texas Bureau of Economic Geology Scott W. Tinker, Director The University of Texas at Austin Austin, Texas

3 Cover: Similarity of patterns evident on maps of (left) simplified geologic map of Texas and (right) average deflection of pavement by county. ii

4 CONTENTS ABSTRACT... 1 INTRODUCTION... 2 Physiographic Regions and Bedrock Types... 3 Importance of Depth to Bedrock on Deflection and Moduli Estimates... 5 Limitations of Estimating Rigid-Layer Depth from FWD Data Alone... 6 METHODS... 7 Bedrock Type and FWD Response... 7 Seismic Refraction... 8 RELATIONSHIP BETWEEN BEDROCK TYPE AND FWD RESPONSE North-Central Plains Site Site A: Texas 16, Archer and Young Counties Central Texas Uplift Sites Site B: Texas 16, Llano and Gillespie Counties Site C: Texas 71, Burnet County Edwards Plateau Site Site D: U.S. 29, Blanco and Hays Counties Gulf Coastal Plains Sites Site E: Texas 71, Bastrop County Site F: Texas 16, Jim Hogg and Zapata Counties Rock-Type Response BEDROCK DEPTHS FROM SEISMIC REFRACTION Pickle Research Campus Site Refraction Experiment PRC SPH Refraction Experiment PRC SPH Interpreted Strata Jacksboro Site Effect of Digital Filtering Refraction Analysis... 4 Interpreted Strata DISCUSSION POTENTIAL APPLICATIONS CONCLUSIONS ACKNOWLEDGMENTS REFERENCES APPENDIX A. Age, lithology, constituents, and thickness of geologic units iii

5 Figures 1. Falling-Weight Deflectometer with detectors deployed on pavement Average roadway deflection by county measured at Falling-Weight Deflectometer detector W Simplified geologic map of Texas Physiographic regions of Texas Soil-probe hammer and recording spread of spike-mounted geophones on the shoulder of Road D at the J. J. Pickle Research Campus, The University of Texas at Austin Map of seismic refraction test site at PRC Geophones mounted on steel plates and placed on the pavement of southbound U.S. 281, south of Jacksboro, Texas Map of the Jacksboro seismic refraction test site North-south cross section, showing geologic units, elevation, and W1 through W7 deflection along Texas 16 between reference mile markers 22 and 264, Archer and Young Counties, North Texas Average and individual deflections for rock types mapped along Texas 16 in Archer and Young Counties Average deflections for all rock units mapped along Texas 16 in Archer and Young Counties Average W7 deflection and W2:W7 deflection ratio by rock type along Texas 16, Archer and Young Counties North-south cross section, showing geologic units, elevation, and W1 through W7 deflection along Texas 16 between reference mile markers 45 and 49, Llano and Gillespie Counties, Central Texas Average and individual deflections for rock types mapped along Texas 16 in Llano and Gillespie Counties Average deflections for all rock units mapped along Texas 16 in Llano and Gillespie Counties Average W7 deflection and W2:W7 deflection ratio by rock type along Texas 16, Llano and Gillespie Counties West-east cross section, showing geologic units, elevation, and W1 through W7 deflection along Texas 71 between reference mile markers 528 and 542, Burnet County, Central Texas Average and individual deflections for rock types mapped along Texas 71 in Burnet County Average deflections for all rock units mapped along Texas 71 in Burnet County Average W7 deflection and W2:W7 deflection ratio by rock type along Texas 71, Burnet County West-east cross section, showing geologic units, elevation, and W1 through W7 deflection along U.S. 29 between reference markers 536 and 563, Blanco and Hays Counties, Central Texas Average and individual deflections for rock types mapped along U.S. 29 in Blanco and Hays Counties Average deflections for all rock units mapped along U.S. 29 in Blanco and Hays Counties Average W7 deflection and W2:W7 deflection ratio by rock type along U.S. 29, Blanco and Hays Counties West-east cross sections, showing geologic units, elevation, and W1 through W7 deflection along Texas 71 between reference mile markers 59 and 598, Bastrop County, southeast Texas iv

6 26. Average and individual deflections for rock types mapped along Texas 71 in Bastrop County Average deflections for all rock units mapped along Texas 71 in Bastrop County Average W7 deflection and W2:W7 deflection ratio by rock type along Texas 71, Bastrop County North-south cross section, showing geologic units, elevation, and W1 through W7 deflection along Texas 16 between reference mile markers 758 and 84, Jim Hogg and Zapata Counties, South Texas Average and individual deflections for rock types mapped along Texas 16 in Jim Hogg and Zapata Counties Average deflections for all rock units mapped along Texas 16 in Jim Hogg and Zapata Counties Average W7 deflection and W2:W7 deflection ratio by rock type along Texas 16, Llano and Gillespie Counties Composite W7 deflection and W2:W7 deflection ratio for individual rock types mapped along Texas 16, U.S. 29, and Texas 71 test sites Seismic response recorded by a 48-geophone recording spread with the FWD as a seismic source Field records from refraction test PRC SPH1 acquired by using a soil-probe hammer as a seismic source First-arrival times for refraction test PRC SPH1 for forward- and reversepropagating waves Apparent velocity and zero-offset time for forward data from refraction test PRC SPH Apparent velocity and zero-offset time for reverse data from refraction test PRC SPH Calculated layer velocities and thicknesses and apparent dips of layer interfaces for refraction tests PRC SPH1, PRC SPH2, and Jacksboro SPH Field records from refraction test PRC SPH2 acquired by using a soil-probe hammer as a seismic source Effect of digital filtering and amplification on a field record from the Jacksboro site acquired by using a soil-probe hammer as a seismic source, with neither digital filtering nor time-varying gain, time-varying gain but no digital filtering, 125-Hz low-cut filter without time-varying gain, and 125-Hz low-cut filter and time-varying gain applied Field records from the Jacksboro site acquired by using a soil-probe hammer as a seismic source First-arrival times for the Jacksboro site for forward- and reverse-propagating waves Apparent velocity and zero-offset time for forward refraction data from the Jacksboro site Apparent velocity and zero-offset time for reverse refraction data from the Jacksboro site Tables 1. Principal physiographic regions of Texas Deflection statistics for sites A through F Summary of refraction data collected at the Pickle Research Campus and Jacksboro sites v

7 ABSTRACT We examined the relationship between three data types geologic maps, measurements of pavement deflection under load, and seismic refraction data from diverse geologic settings in Texas to determine (1) whether geologic maps and seismic refraction data might be used to interpret deflections and assess pavement condition and (2) whether deflections and refraction data acquired on pavement might have geologic applications. Engineers assess pavement condition by applying a known load to a road and measuring vertical pavement deflection using the Falling-Weight Deflectometer (FWD). Our comparisons of deflections with mapped geologic units in four physiographic regions of Texas revealed differences in FWD response that are related to differences in either bedrock depths or physical properties of geologic units that range from Precambrian to Holocene in age and that include many different sedimentary, igneous, and metamorphic rocks. At the FWD sensor most distant (1.8 m) from the load, deflections are greatest where roads are underlain by clastic sedimentary rocks (sandstones, mudstones, and shales) and unconsolidated alluvium. Lowest deflections are measured over stiffer limestones and igneous and metamorphic rocks. Ratios calculated by dividing the deflections measured.3 m (W2 sensor) and 1.8 m (W7 sensor) from the load are better discriminators. These ratios are highest where pavement is underlain by rigid rock types such as granites, gneisses, and schists (ratios of 17:1 to 4:1), are intermediate where the underlying bedrock is limestone (1:1 to 27:1), and are lowest over sandstones, mudstones, and unconsolidated sediments (6:1 to 14:1). These results suggest that geologic maps are useful in FWD analysis and that FWD data alone might be used to predict rock type, allowing the FWD to be used in applications such as geologic mapping and sinkhole detection. We employed the FWD and a modified soil-probe hammer as seismic refraction sources to determine whether the correlation between rock type and pavement deflections is caused by differing rock properties or bedrock depth. These tests revealed that the FWD is a viable seismic source and that refraction data can be acquired on pavement. The refraction experiments and the rock type deflection relationship show that (1) pavement deflections can help the geologist make geologic maps, (2) geologic maps can help the engineer interpret pavement deflections, and (3) seismic refraction measurements can assist both geologist and engineer in determining bedrock depths, rock properties, and rock types. Keywords: bedrock, Falling-Weight Deflectometer, geologic mapping, pavement deflection, rock properties, seismic refraction 1

8 INTRODUCTION Engineers routinely use Falling-Weight Deflectometers (FWD s; fig. 1) to assess pavement condition in Texas by measuring vertical deflection of pavement under a controlled load. Maps depicting average deflection by county (fig. 2) bear a striking resemblance to the generalized geologic map of Texas (fig. 3). Our goals in this study are to determine why this apparent relationship exists, how well it translates to the local scale, and how it might be exploited both to aid pavement analysis and suggest geologic uses of the FWD. Further, we wish to examine whether FWD data can be augmented by acquiring seismic refraction data on pavement and how this capability might enhance its value in pavement analysis and geologic applications. Once the relationship between pavement deflections, refraction measurements, and geologic maps is established, these instruments can be used to improve our knowledge of bedrock depths, distribution of geologic units, and the physical properties of geologic units throughout Texas where exposures are limited. FWD s (fig. 1) consist of a falling weight and a series of seven calibrated geophones, W1 through W7, at distances of, 1, 2, 3, 4, 5, and 6 ft (,.3,.6,.9, 1.2, 1.5, and 1.8 m) from the falling weight. The height of the weight drop can be selected to produce vertical loads of varying size. The vertical geophones, in contact with the pavement, measure pavement deflection following weight impact. Although the FWD records a time series of movement at each sensor, engineers commonly assess pavement condition by using the maximum deflections at each detector, normalized for drop load and temperature. In general, deflections measured close to the source are most affected by pavement condition; deflections measured at the longest offsets are more affected by underlying layers such as fill, soil, and bedrock. Physical properties of roadway layers that can be calculated from FWD data also depend on depth to bedrock (depth to rigid layer), which is generally not known. Rather than drill to bedrock at each FWD site to improve pavement assessment, we wish to determine whether geologic maps or seismic refraction methods can be used to estimate bedrock depths beneath roads. The empirical relationship between rock type and FWD deflections (figs. 2, 3), particularly evident at the longest offsets (W7 sensor, 6-ft [1.8-m] source-to-detector distance), suggests Figure 1. Falling-Weight Deflectometer with detectors deployed on pavement. 2

9 that geologic maps can be used to interpret FWD data and that FWD data might have geologic applications. Largest W7 deflections in Texas are observed along the coast and in the Panhandle, where geologic units are relatively young. Smallest W7 deflections are observed in Central Texas, where Precambrian igneous and metamorphic rocks and Cretaceous limestones are mapped. Outcrop trends of individual geologic units match average W7 deflection trends extending over many counties, including (1) the increased average deflections in East Texas on the Miocene Fleming and Oakville Formations and the Pliocene Willis Formation; (2) low deflections on the Cretaceous Trinity, Fredericksburg, and Lower Washita Groups in Central Texas; and (3) increased deflections that follow the Cretaceous Austin, Eagle Ford, Woodbine, Upper Washita, Navarro, and Taylor Groups in northeast Texas (figs. 2, 3). Physiographic Regions and Bedrock Types The relationship between FWD data and geologic units supports the subdivision of Texas into regions that have similar FWD response. Many Earth scientists have recognized physiographic regions that reflect differences in elevation, topography, geologic structure, and bedrock types (fig. 4; table 1). These seven Deflection (mils) < Amarillo > 2.5 A R2 Dallas El Paso Midland N B D C R1 Austin E Houston San Antonio 1 2 mi km F Corpus Christi G U L F O F M E X I C O Brownsville QAc26c Figure 2. Average roadway deflection by county measured at Falling-Weight Deflectometer detector W7. Also shown are locations of study areas A through F and refraction test sites R1 (Pickle Research Campus, Austin) and R2 (U.S. 281, Jacksboro). 3

10 El Paso N 1 2 Amarillo Midland 1 2 mi 3 km A B C F D Austin San Antonio Brownsville R2 R1 E Dallas Corpus Christi G U L F O F Houston M E X I C O Alluvium Quaternary undivided Beaumont Formation Lissie Formation Blackwater Draw Formation Willis Formation Ogallala Formation Goliad Formation Fleming and Oakville Formations Catahoula Formation Oligocene and Eocene undivided Jackson Group U. Claiborne Group L. Claiborne Group Wilcox and Midway Groups Navarro and Taylor Groups Austin, Eagle Ford, Woodbine and U. Washita Groups Fredericksburg and L. Washita Groups Trinity Group Cretaceous undivided Jurassic Triassic undivided Ochoan Series U. Guadalupian Series L. Guadalupian Series Leonardian Series Wolfcampian Series Permian undivided Virgilian Series Missourian Series Desmoinesian Series Atokan and Morrowan Series Mississippian, Devonian, and Ordovician undivided Cambrian Paleozoic undivided Precambrian undivided QAc27c Figure 3. Simplified geologic map of Texas. Adapted from Bureau of Economic Geology (1992). Also shown are locations of study areas A through F and refraction test sites R1 and R2. principal physiographic regions (Gulf Coastal Plains, Edwards Plateau, Grand and Blackland Prairies, Central Texas Uplift, North-Central Plains, High Plains, and Basin and Range) provide a framework for grouping rock types that influence FWD response. Bedrock types differ in each of the seven principal physiographic regions (figs. 3, 4; table 1). Over much of the Gulf Coastal Plains, unconsolidated and semiconsolidated sands, silts, and clays deposited along rivers and shorelines in the Cenozoic Era (within the last 66 million yr) form relatively weak substrates. Relatively young bedrock is also found in the High Plains, where unconsolidated to moderately cemented eolian and alluvial sand and silt formed the Blackwater Draw Formation during the Quaternary Period (<2 mya) and the Ogallala Formation during the Miocene to Pliocene Periods (24 to 2 mya). Limestone and dolomite deposited during the Cretaceous Period (144 to 66 mya) underlie the Edwards Plateau in Central Texas, forming strong substrates that are resistant to erosion. Sandier, calcareous deposits of similar age underlie the Grand Prairie, the northern extension of the Edwards Plateau. Westward-dipping limestone, sandstone, and shale dating to the late Paleozoic Era (32 to 245 mya) are found in the North- Central Plains. The oldest rocks in Texas are found in the Central Texas Uplift and the Basin and Range regions. In the Central Texas Uplift, mechanically strong, late Precambrian Era (2, to 1,2 mya) igneous and metamorphic rocks and Paleozoic Era (57 to 245 mya) sandstone, limestone, and shale crop out. In the Basin and Range, topographic highs composed of strong 4

11 N AMARILLO Gulf Coastal Plains Edwards Plateau/ Grand Prairie/ Blackland Prairie A R2 DALLAS Central Texas Uplift North-Central Plains High Plains EL PASO MIDLAND Basin and Range B C D R1 AUSTIN E SAN ANTONIO HOUSTON CORPUS CHRISTI mi 2 3 km F QAb8762c Figure 4. Physiographic regions of Texas. Adapted from Wermund (1996). Also shown are locations of study areas A through F and refraction test sites R1 and R2. igneous and metamorphic rocks are separated by downfaulted basins filled with younger sedimentary deposits that are generally weaker than the range-forming rocks. More detailed information on the distribution of geologic units is obtained from geologic maps produced at various scales. The most useful map series for statewide study is the Geologic Atlas of Texas. This series consists of 38 maps that cover the entire state at a scale of 1:25, and that have been compiled, published, and updated over the last several decades by the Bureau of Economic Geology (Bureau). Soil-type maps, published electronically and on paper by the Natural Resources Conservation Service of the U.S. Department of Agriculture, exist for most Texas counties. The information on soil distribution, grain size, soil depth, and surface slope contained in the maps and tables that make up these surveys, more detailed than that shown on geologic maps, may also be useful in the interpretation of FWD data. Importance of Depth to Bedrock on Deflection and Moduli Estimates McCullough (1969) identified the importance of using an accurate depth-to-rigid-layer estimate in the linear-elastic layered analysis of a pavement structure. He found that variations in deflection estimates can occur as a result of either the subgrade thickness or rigid-layer stiffness estimate, concluding that for subgrade thicknesses greater than 36 inches (>.9 m), the 5

12 Table 1. Principal physiographic regions of Texas. Adapted from Wermund (1996). Region Elevation range (ft) Topography Geologic structure Bedrock type Gulf Coastal Plains 1 Nearly flat to low, rolling terrain Nearly flat strata Unconsolidated deltaic sands and muds; chalks and marls Grand Prairie 45 1,25 Plains to low, stairstep hills Edwards Plateau 45 3, Flat upper surface with box canyons Gentle eastward dip Gentle southward dip Calcareous to sandy sedimentary rocks Limestones and dolomites Central Texas Uplift 8 2, Knobby plain Outward dip; faulted Igneous and metamorphic rocks North-Central Plains 9 3, Low north-south ridges Gentle westward dip Limestones, sandstones, shales High Plains 2,2 4,75 Southeastwardsloping prairies Gentle southeastward dip Windblown silt and fine sand Basin and Range 1,7 8,75 North-south mountains and basins Complex folding and faulting Igneous, metamorphic, and sedimentary rocks rigid layer stiffness may range between 3, and 2,, psi (2.1 and Pa) without affecting deflection estimates. Further, the effect of the depth to rigid layer on pavement stress is significant over the range of to 12 ft ( to 3.7 m) but is insignificant beyond 12 ft (3.7 m). Briggs and Nazarian (1989) studied the effect of depth-to-rigid-layer estimates on backcalculated moduli, which are used to predict pavement performance. They found that if the actual rigid-layer depth is half the assumed value, the accuracy of back-calculated moduli is reduced. If the stiffness estimate of the rigid layer is only about 1 times the subgrade stiffness and the depth-to-rigid layer is actually half the estimated value, back-calculated moduli will be totally inaccurate. In addition, if the depth to rigid layer is ignored or overestimated, the remaining pavement-life estimates will be greatly overestimated because subgrade stiffness is overestimated if the depth-to-rigid-layer estimate is greater than the actual value. During modulus back calculation, poor fit to the observed FWD deflections might be due to inaccurate depth-torigid-layer estimates rather than to the more common explanation, which invokes the use of a simplified linear-elastic model to characterize nonlinear or viscoelastic materials. Limitations of Estimating Rigid-Layer Depth from FWD Data Alone Rohde (1991) developed a method for estimating depth to rigid layer for use in the backcalculation program MODULUS. His method, based on previous work by Ullditz (1987), estimates the depth at which deflections diminish to zero within the subgrade layer when an FWD load is applied at the pavement surface. In this approach, the point of zero deflection is assumed to coincide with the rigid-layer depth. Ullditz (1987) noted that deflections beyond about twice the radius of the FWD load plate could be modeled equally well by either a point load or circular plate load. On the basis of this observation, Rohde and Smith (1991) used the Boussinesq equations 6

13 and the relationship between surface deflection and the inverse of radial offset (1/R) to estimate the depth at which a zero deflection would occur. Four regression equations were developed that predict depth to rigid layer as a function of the zero-deflection point, surface-layer thickness and various FWD deflection basin parameters (Surface Curvature Index [W1 W2], Base Damage Index [W2 W3] and Base Curvature Index [W3 W4]). Regression equations have been incorporated into the MODULUS program as a means of estimating depth to rigid layer. On the basis of the Mahoney and others (1993) evaluation, the Rohde equations were adopted for use in the EVERCALC modulus backcalculation program used by the Washington State Department of Transportation. Mahoney and others (1995) further determined, on the basis of a study performed at the PACCAR research center, that subsurface units other than bedrock can act as a rigid layer. For example, a high water table that produces saturated subgrade conditions could be recognized as a rigid layer by using Rohde s methods. If saturated subgrade conditions exist, an assumed rigid-layer stiffness of 4, to 5, psi (.28 to Pa) may be more appropriate in a linear-elastic analysis (Mahoney and others, 1995). METHODS Our approach to understanding the relationship between FWD response and bedrock type and depth was to examine road segments in different physiographic regions. We first examined the relationship between existing Texas Department of Transportation (TxDOT) FWD data and mapped geologic units along the highway segments. We then collected seismic refraction data to investigate their effectiveness in determining bedrock depths to maximum depths of about 6 m beneath pavement. The six highway segments analyzed (figs. 2 through 4) are located in the southern and interior Gulf Coastal Plains (2 segments), Edwards Plateau (1 segment), the Central Texas Uplift (2 segments), and the North- Central Plains (1 segment). We use metric (SI) units throughout this report except in reporting highway and FWD data. Original highway-related data are in English units, which are listed first and are followed by the metric equivalent in parentheses. Pavement deflections measured by the FWD are traditionally reported in mils, which represent.1 inch or 25.4 µm. To increase readability, we do not give metric equivalents for pavement deflections. Bedrock Type and FWD Response To determine whether a quantifiable relationship exists between bedrock and FWD response beyond what is apparent from the similarity of the geologic map of Texas and the countyaverage W7 deflection, we selected six highway segments in different parts of Texas. For each highway segment, we (1) obtained FWD deflections from TxDOT adjusted to a common 9,-lb (4,82-kg) load, (2) plotted FWD locations on U.S. Geological Survey 7.5-minute quadrangle maps, and (3) determined what geologic unit underlies the highway at each FWD site from 1:25,- scale geologic maps published by the Bureau. These data formed a data base that includes highway name and location, geologic unit, elevation, and normalized deflection for each FWD site. We then analyzed the data to investigate how bedrock influences FWD response. Plots of elevation, rock type, and deflection versus distance along the highway show how deflections relate to different geologic units beneath the highway and to changes in elevation and relief. When the 7

14 deflection data are sorted by rock type, we can calculate the average deflection series for a given bedrock type, determine how the deflection series vary, and decide whether bedrock types have distinctive deflection series. If deflection series have different slopes, we can calculate deflection ratios for near- and far-offset detectors to further discriminate rock types. Seismic Refraction Seismic refraction is a geophysical method (Telford and others, 1976; Milsom, 1989) for determining compressional-wave velocities of materials at various depths below the land surface. In the shallow subsurface, seismic refraction is commonly used to measure depth to the water table or to bedrock (the rigid layer beneath soil and weathered bedrock). Compressional-wave Figure 5. Soil-probe hammer and recording spread of spike-mounted geophones on the shoulder of Road D at the J. J. Pickle Research Campus (PRC), The University of Texas at Austin. velocities increase downward in most geologic settings, where relatively dry soil (compressionalwave velocities ranging from 3 to 7 m/s) is underlain by saturated soil at the water table (compressional velocities of ~1,5 m/s) or by unweathered bedrock (compressional velocities commonly >2, m/s, depending on rock type). These typically abrupt, downward increases in wave velocity refract surface-generated seismic waves along the interface between the units. The refracted waves generate wavefronts that propagate back to the surface, where they are detected by motion sensors (geophones). The time delay between seismic-source impact and first seismic arrivals (first break) at known geophone distances allows us to calculate compressional velocities and thicknesses of near-surface layers, the calculations in turn allowing us to estimate a depth to the water table or to bedrock. In general, exploration depth increases with distance between the source and detector. For shallow investigations, the detector spread should extend from within a short distance of the source to four or more times the desired maximum exploration depth. This spread allows us to measure enough arrivals of both the direct wave (traveling in the surface layer only) and the critically refracted wave (traveling along the water table or at the interface between the surface layer and bedrock) to calculate accurate velocities for these layers. To test the usefulness of the seismic refraction method on pavement, we recorded seismic refraction data using 48 4-Hz geophones, a 48-channel seismograph, and 2 seismic sources (the FWD and a modified soil-probe hammer) at the J. J. Pickle Research Campus, The University of Texas at Austin (PRC), and on U.S. Highway 281 near Jacksboro. Spread length, geophone spacing, and seismic-source selection depend on target depths, ambient seismic noise, ground conditions, and desired lateral resolution. For typical pavement settings, a sledge hammer, a modified soil-probe hammer, or the FWD itself can be suitable sources. We used both the FWD and the soil-probe hammer in our tests. We picked first compressional-wave arrivals using the Seismic Processing Workshop software package. We calculated true seismic velocities, layer thicknesses, and apparent dip angles using the slope-intercept method (Palmer, 1986). 8

15 Road D Geophone 1 Geophone 48 N 2 ft 5 m FWD source point Soil-probe-hammer source point Geophone QAc4612c Figure 6. Map of seismic refraction test site at PRC (site R1, fig. 2). At the PRC site (site R1, figs. 2 through 4), we deployed a recording spread on the road shoulder using geophones mounted on spikes (fig. 5). Geophones were spaced at.3-m intervals for a total recording spread length of 14.3 m (fig. 6). The FWD was operated on the pavement, offset about 1 m north of the east-west recording spread. The soil-probe hammer was operated both on the pavement.4 m north of the recording spread and on the shoulder inline with the recording spread. Seismic pulses were recorded with the sources located at the center and ends of the recording spread (fig. 6). At the Jacksboro site (site R2, figs. 2 through 4), we deployed the recording spread on the pavement using geophones threaded onto steel plates (fig. 7). The recording spread, covering a distance of 23.5 m at a geophone interval of.5 m, was laid out on the inside, southbound lane of U.S. Highway 281 (fig. 8). FWD seismic pulses were recorded from source locations offset.9 m west of the north-south recording spread; soil-probehammer pulses were recorded from locations along the recording spread. For both sources, source points were at the center and ends of the recording spread (fig. 8). At both sites, a seismograph sampling interval of.1 s allowed precise first-arrival Figure 7. Geophones mounted on steel plates and placed on the pavement of southbound U.S. 281, south of Jacksboro, Texas. 9

16 times to be picked. At a propagation velocity of 5 m/s, a seismic pulse travels 5 cm in.1 s. A longer sample interval, such as the.1 s typical of deeper refraction surveys, translates to.5 m of wave propagation between samples. Sample intervals this long cause unacceptable errors in arrival-time picks, which in turn cause erroneous layer-depth calculations. Spatial aliasing of the recorded seismic pulse was prevented by our ensuring that the geophone spacing was much shorter than the compressional-wave wavelengths, which are 5 to 3 m at 1 Hz. Recording was initiated by an electronic switch mounted to the seismic source. Seismic data were recorded for.2 s after source impact. Jacksboro (4 km) U.S. 281 South Geophone 48 Geophone 1 N FWD source point Soil-probe-hammer source point Geophone Figure 8. Map of the Jacksboro seismic refraction test site (site R2, fig. 2). 1 ft 5 m QAc4613(a)c 1

17 RELATIONSHIP BETWEEN BEDROCK TYPE AND FWD RESPONSE Physical properties of sedimentary, igneous, and metamorphic rocks, including density, wavepropagation velocities, and elastic parameters, have been shown in numerous field and laboratory experiments to vary widely (Press, 1966). For geologic maps to be useful in the interpretation of FWD data and for FWD data to be useful in geologic applications, FWD deflections should show a relationship to mapped rock type. This relationship may be the result of a similarity in bedrock depths for a given bedrock type or a similarity in physical properties of a given bedrock type. The four physiographic regions where we examined bedrock type and FWD response in detail are underlain by geologic units that range from Precambrian rocks as old as 2 b.y. to recently deposited Holocene sediments. Sedimentary bedrock types include (1) unconsolidated gravel, silt, sand, and clay along streams in each of the regions studied; (2) chemically precipitated limestones and dolomites in the Edwards Plateau, Central Texas Uplift, and North-Central Plains; and (3) lithified to semiconsolidated sandstone and shale in each of the regions studied. Igneous bedrock types include granites that crop out in the Central Texas Uplift. Metamorphic rocks, including gneisses and schists, are also mapped in the Central Texas Uplift. North-Central Plains Site Compared with those in the rest of Texas, county-average deflections in the North-Central Plains physiographic region are moderate, ranging from 1.1 to 2. mils (figs. 2, 4). Lithified sedimentary bedrock types common in this region include Paleozoic limestone, sandstone, and shale (table 1). Unconsolidated sediments are common along the major rivers (Colorado, Brazos, Trinity, and Red Rivers) and numerous smaller streams that cross the region. Site A: Texas 16, Archer and Young Counties Site A extends 7 km along Texas 16 in Archer and Young Counties. Average W7 deflections are between 1.6 and 2. mils for Archer County and between 1.1 and 1.5 mils for Young County (fig. 2). Geologic units mapped along this highway segment include lithified Paleozoic sandstones, limestones, and mudstones and unconsolidated Quaternary stream deposits (fig. 9; app. A). FWD data on 87 locations along this highway segment show a wide range of deflections for each detector (fig. 9). W7 deflections average about 1. mil (table 2), which is lower than the reported average for Archer and Young Counties. The calculated average for Texas 16 is higher than average deflections calculated for sites in the Central Texas Uplift and Edwards Plateau regions (sites B, C, and D) and lower than calculated averages for the Gulf Coastal Plains sites (sites E and F), in agreement with the map of countywide average deflections. For many of the 11 geologic units mapped along this highway segment, FWD data show considerable overlap in observed deflection ranges (fig. 1). For example, W7 deflections over the Markley Mudstone range from.8 to more than 2. mils; Markley Sandstone deflections range from.5 to 2. mils (fig. 1b, 1c). Other rock units having more than a few measured deflections have similarly broad ranges. Deflection averages calculated for the geologic units mapped at site A decrease from between 1 and 4 mils at W1 to between.5 and 2 mils at W7 (figs. 1, 11). Deflection series that have high near-offset deflections also tend to have high far-offset deflections. Geologic units over which relatively small average W7 deflections (less than 1. mil) were measured are the Thrifty-Graham and Kisinger Sandstones at.5 mil, the Ranger and Home Creek Limestones at.6 to.8 mil, 11

18 Table 2. Deflection statistics (normalized to 9,-lb [4,82-kg] load) for sites A through F (fig. 2). Statistics shown in mils. Site A: Texas 16, Archer and Young Counties, n=87. Statistic W1 W2 W3 W4 W5 W6 W7 Average Standard deviation Maximum Minimum Site B: Texas 16, Llano and Gillespie Counties, n=69. Statistic W1 W2 W3 W4 W5 W6 W7 Average Standard deviation Maximum Minimum Site C: Texas 71, Burnet County, n=3. Statistic W1 W2 W3 W4 W5 W6 W7 Average Standard deviation Maximum Minimum Site D: U.S. 29, Blanco and Hays Counties, n=52. Statistic W1 W2 W3 W4 W5 W6 W7 Average Standard deviation Maximum Minimum Site E: Texas 71, Bastrop County, n=34. Statistic W1 W2 W3 W4 W5 W6 W7 Average Standard deviation Maximum Minimum Site F: Texas 16, Jim Hogg and Zapata Counties, n=89. Statistic W1 W2 W3 W4 W5 W6 W7 Average Standard deviation Maximum Minimum and the Gonzales Creek Sandstone at.9 mil (fig. 12a). Relatively large average W7 deflections were measured over the Bunger Limestone (1.8 mils), the Markley Mudstone (1.3 mils), and the Ivan Limestone (1.2 mils). Ratios calculated for average deflections at different detectors can help remove the covariance of near- and far-offset deflections and better reveal bedrock effects. We calculated the W2:W7 ratio (fig. 12b) because W2 should have the 12

19 North South Elevation (ft) Deflection (mils) 1 1 W1 W2 W3 W4 W5 W6 W Reference marker GEOLOGIC UNIT Qal alluvium IPtg Thrifty-Graham Sandstone IPk Kisinger Sandstone IPPm Markley Sandstone IPtg Thrifty-Graham Mudstone IPhc Home Creek Limestone IPPm Markley Mudstone IPbu Bunger Limestone IPcc Colony Creek Shale IPi Ivan Limestone IPgc Gonzales Creek Sandstone IPr Ranger Limestone QAc844c Figure 9. North-south cross section, showing geologic units, elevation, and W1 through W7 deflection along Texas 16 between reference mile markers 22 and 264, Archer and Young Counties, North Texas (site A, figs. 2 through 4). largest source- and pavement-related deflection component and W7 should have the largest bedrock-related deflection. Ratios calculated for North-Central Plains geologic units range from 8 to 14, increasing for units that have large W2 deflections for a given W7 deflection. All other factors being equal, rigid geologic units should have higher W2:W7 ratios than less rigid ones. The Home Creek Limestone, Kisinger Sandstone, and Thrifty-Graham Mudstone have low ratios (less rigid); the Markley Sandstone, Ranger Limestone, and Thrifty-Graham Sandstone have relatively high ratios (more rigid). These ratios are lower than those calculated for geologic units in the Central Texas Uplift and Edwards Plateau and higher than those in the Gulf Coastal Plains, as they should be if bedrock type affects FWD response. Central Texas Uplift Sites The Central Texas Uplift, underlain by Precambrian igneous and metamorphic rocks, Paleozoic and Mesozoic sedimentary rocks, and unconsolidated Quaternary sediments (table 1), 13

20 (a) (b) (c) 1. Qal alluvium IPPm Markley Mudstone IPPm Markley Sandstone Deflection (mils) (d) (e) (f) 1. IPtg Thrifty-Graham Mudstone IPbu Bunger Limestone IPtg Thrifty-Graham Sandstone Deflection (mils) IPi Ivan Limestone IPr Ranger Limestone (g) 1. (h) IPgc Gonzales Creek Sandstone IPhc Home Creek Limestone (i) IPk Kisinger Sandstone Deflection (mils) Detector Detector Detector QAc846(a+b)c-rev. Figure 1. Average (heavy line[s]) and individual (light lines) deflections for rock types mapped along Texas 16 in Archer and Young Counties. 14

21 Deflection (mils) Detector Qal alluvium IPPm Markley Mudstone IPPm Markley Sandstone IPPtg Thrifty-Graham Mudstone IPPtg Thrifty-Graham Sandstone IPi Ivan Limestone IPbu Bunger Limestone IPgc Gonzales Creek Sandstone IPhc Home Creek Limestone IPk Kisinger Sandstone IPr Ranger Limestone QAc845c Figure 11. Average deflections for all rock units mapped along Texas 16 in Archer and Young Counties. (a) Qal alluvium IPPm Markley Mudstone IPPm Markley Sandstone IPPtg Thrifty- Graham Mudstone IPi Ivan Limestone IPPtg Thrifty- Graham Sandstone IPbu Bunger Limestone IPgc Gonzales Creek Sandstone IPhc Home Creek Limestone IPk Kisinger Sandstone Rock unit Rock unit IPr Ranger Limestone (b) Qal alluvium IPPm Markley Mudstone IPPm Markley Sandstone IPPtg Thrifty- Graham Mudstone IPi Ivan Limestone IPPtg Thrifty- Graham Sandstone IPbu Bunger Limestone IPgc Gonzales Creek Sandstone IPhc Home Creek Limestone IPk Kisinger Sandstone IPr Ranger Limestone Average W7 deflection (mils) W2:W7 deflection ratio QAc847c covers the smallest area of any physiographic region (fig. 4). County-average deflections in this region of typically rigid bedrock types are the lowest in Texas, ranging from less than 1 to 1.5 mils (fig. 2). Two study sites are located in this region (sites B and C, fig. 4). Figure 12. (a) Average W7 deflection and (b) W2:W7 deflection ratio by rock type along Texas 16, Archer and Young Counties. 15

22 North South Elevation (ft) Deflection (mils) 1 1 W1 W2 W3 W4 W5 W6 W Reference marker Qal alluvium Kft Fort Terrett Limestone Kh Hensell Sand GEOLOGIC UNIT pcy younger granites pctm Town Mountain Granite pcps Packsaddle Schist Crh Hickory Sandstone pcvs Valley Spring Gneiss QAc848c Figure 13. North-south cross section, showing geologic units, elevation, and W1 through W7 deflection along Texas 16 between reference mile markers 45 and 49, Llano and Gillespie Counties, Central Texas (site B, figs. 2 through 4). Site B: Texas 16, Llano and Gillespie Counties This segment of Texas 16 begins south of Llano and extends 61 km to Fredericksburg. It is mostly underlain by Precambrian metamorphic (Packsaddle Schist and Valley Spring Gneiss) and igneous (Town Mountain Granite) rocks and the Cretaceous Hensell Sand (fig. 13). A few occurrences of younger granites, Cambrian Hickory Sandstone, Cretaceous Fort Terrett Limestone, and Quaternary stream deposits are mapped along the highway (fig. 13; app. A). Younger geologic units are found at relatively high elevations on the south part of the segment; older igneous and metamorphic rocks are found at relatively low elevations on the north part of the segment (fig. 13). Average W7 deflections for both Llano and Gillespie Counties are less than 1. mil (fig. 2), reflecting the abundance of rigid bedrock in the Central Texas Uplift. We analyzed FWD data from 69 sites along this highway segment (table 2), which has the highest average W1 deflection of any of the study sites (34.7 mils) but a low average W7 deflection (.9 mil). When the data are grouped by geologic unit (fig. 14), they show that sites having large 16

23 (a) (b) (c) 1. Qal alluvium Kft Fort Terrett Limestone Kh Hensell Sand Deflection (mils) (d) (e) (f) 1. Crh Hickory Sandstone pcy younger granites pctm Town Mountain Granite Deflection (mils) QAc85c (g) 1. Deflection (mils) pcps Packsaddle Schist (h) pcvs Valley Spring Gneiss Figure 14. Average (heavy line) and individual (light lines) deflections for rock types mapped along Texas 16 in Llano and Gillespie Counties Detector Detector 17

24 1. (a) Qal alluvium 1.18 Kft Fort Terrett Limestone 1.24 Deflection (mils) Rock unit Kh Hensell Sand Crh Hickory Sandstone pcy younger granites pctm Town Mountain Granite pcps Packsaddle Schist pcvs Valley Spring Gneiss Detector Qal alluvium Kft Fort Terrett Limestone Kh Hensell Sand Crh Hickory Sandstone pcy younger granites pctm Town Mountain Granite pcps Packsaddle Schist pcvs Valley Spring Gneiss QAc849c Figure 15. Average deflections for all rock units mapped along Texas 16 in Llano and Gillespie Counties. near-offset deflections generally also have large far-offset deflections and that there is more variation within a geologic unit than there is between average deflections of each rock type. Although average W7 deflections calculated for the Hensell Sand are higher than those for the Town Mountain Granite and the Valley Spring Gneiss, the range in individual W7 deflections observed for these rock types is similar:.4 to 2 mils for the Hensell Sand,.3 to 1.1 mils for the Town Mountain Granite, and.3 to 1.8 mils for the Valley Spring Gneiss (fig. 14). (b) Rock unit Qal alluvium Kft Fort Terrett Limestone Kh Hensell Sand Crh Hickory Sandstone pcy younger granites pctm Town Mountain Granite pcps Packsaddle Schist pcvs Valley Spring Gneiss Average W7 deflection (mils) W2:W7 deflection ratio QAc851c Figure 16. (a) Average W7 deflection and (b) W2:W7 deflection ratio by rock type along Texas 16, Llano and Gillespie Counties. Statistically, Precambrian igneous and metamorphic rocks have low-average W7 deflections that range from.5 mil for younger granites to.9 mil for the Packsaddle Schist (figs. 15, 16a). Higher average W7 deflections, ranging from 1.1 to 1.2 mils, are calculated for Cretaceous and younger sedimentary units. W7 averages for geo- 18

25 1 West East 12 8 Elevation (ft) Deflection (mils) 1 1 W1 W2 W3 W4 W5 W6 W Reference marker GEOLOGIC UNIT Qal alluvium Ksy Sycamore Sand Kgru Upper Glen Rose Limestone IPmf Marble Falls Limestone Kgrl Lower Glen Rose Limestone Oh Honeycut Limestone Kh Hensell Sand Og Gorman Limestone QAc86c Figure 17. West-east cross section, showing geologic units, elevation, and W1 through W7 deflection along Texas 71 between reference mile markers 528 and 542, Burnet County, Central Texas (site C, figs. 2 through 4). logic units at site B are similar to those calculated for geologic units in the North-Central Plains. The W2:W7 ratio provides better discrimination of rock type for site B. Very high ratios are calculated for the rigid rock units (fig. 16b): between 17 and 4 for granites, metamorphic rocks, and the Fort Terrett Limestone. Lower ratios are calculated for the clastic sedimentary units, ranging from 13 to 14 for the Hickory Sandstone and Hensell Sand and 11 for Quaternary stream deposits. Ratios for the most common units encountered along Texas 16, higher than those observed in the North-Central Plains and Gulf Coastal Plains, are similar to ratios calculated for the Edwards Plateau. Site C: Texas 71, Burnet County This 23-km-long segment is located in eastern Burnet County. Paleozoic limestones mapped in the west are replaced eastward by Cretaceous sands and limestones, representing a transition from typical Central Texas Uplift units to typical Edwards Plateau units (fig. 17). Average W7 deflection for Burnet County is less than 1 mil, the lowest category (fig. 2). 19

26 (a) (b) (c) 1. Qal alluvium Kgru Glen Rose Limestone (upper) Kgrl Glen Rose Limestone (lower) Deflection (mils) (d) (e) (f) 1. Kh Hensell Sand Ksy Sycamore Sand IPmf Marble Falls Limestone Deflection (mils) (g) 1. Deflection (mils) Oh Honeycut Limestone (h) Og Gorman Limestone Detector QAc862c Figure 18. Average (heavy line) and individual (light lines) deflections for rock types mapped along Texas 71 in Burnet County Detector Detector 2

27 1. (a) Qal alluvium Kgru Upper Glen Rose Limestone Deflection (mils) Detector Qal alluvium Kgru Upper Glen Rose Limestone Kgrl Lower Glen Rose Limestone Kh Hensell Sand Ksy Sycamore Sand IPmf Marble Falls Limestone Oh Honeycut Limestone Og Gorman Limestone QAc861c Figure 19. Average deflections for all rock units mapped along Texas 71 in Burnet County. Rock unit Rock unit (b) Kgrl Lower Glen Rose Limestone Kh Hensell Sand Ksy Sycamore Sand IPmf Marble Falls Limestone Oh Honeycut Limestone Og Gorman Limestone Qal alluvium Kgru Upper Glen Rose Limestone Kgrl Lower Glen Rose Limestone Kh Hensell Sand Ksy Sycamore Sand IPmf Marble Falls Limestone Oh Honeycut Limestone Og Gorman Limestone Average W7 deflection (mils) W2:W7 deflection ratio QAc863c FWD data from 3 locations along this highway segment (table 2) indicate that average deflections at each offset are low. Average W7 deflection is less than.6 mil, which falls within the indicated county-average category (fig. 2). Most of the FWD measurements were acquired over the Ordovician Honeycut Limestone, for which individual W7 deflections ranged from less than to.8 mil (fig. 18g). The average W7 deflections for all but two rock types fall within this range, including Quaternary stream deposits, Cretaceous Upper Glen Rose Limestone and Figure 2. (a) Average W7 deflection and (b) W2:W7 deflection ratio by rock type along Texas 71, Burnet County. Hensell Sand, and Ordovician Gorman Limestone (figs. 19, 2a). Two units having higher average W7 deflections than the range observed for the Honeycut Limestone were the Cretaceous Sycamore Sand (1.5 mils) and the Pennsylvanian Permian Marble Falls Limestone (1.1 mils). 21

28 Average W7 values for all other units were below.7 mil. High W2:W7 ratios (between 13 and 27, fig. 2b), indicating a rapid decrease in deflection as offset increases and probably a relatively stiff or shallow bedrock, were calculated for the Honeycut Limestone, the Upper Glen Rose Limestone, and the Gorman Limestone. Intermediate ratios (9 to 1) were calculated for the small number of examples over the Marble Falls Limestone, the Sycamore and Hensell Sands, and Quaternary stream deposits. The most common geologic unit, the Honeycut Limestone, has a ratio that is similar to that of other rigid units in the Central Texas Uplift and Edwards Plateau regions and is higher than those in the North-Central and Gulf Coastal Plains regions. Edwards Plateau Site Average W7 deflections for counties within the Edwards Plateau are below 1.5 mils (figs. 2, 4), similar to those in the Central Texas Uplift counties and the lowest in Texas. Relatively rigid Cretaceous limestones and dolomites are the most common bedrock types across the Edwards Plateau (table 1). Unconsolidated gravel, sand, and clay are common along numerous streams and rivers that dissect the plateau (fig. 3). One study site (site D, fig. 4) is located in the central part of the Edwards Plateau. Site D: U.S. 29, Blanco and Hays Counties This segment of U.S. 29 extends 43 km across eastern Blanco and northern Hays Counties. Average W7 deflections for these counties are very low (fig. 2). Only four geologic units are mapped: Cretaceous lithified sedimentary rocks that include the Upper and Lower Glen Rose and Fort Terrett Limestones and unconsolidated Quaternary stream deposits (app. A). Upper and Lower Glen Rose Limestones are the most common geologic units. The Lower Glen Rose is found in the west part of the segment, and the Upper Glen Rose crops out at the higher elevations common in the east part of the segment (fig. 21). Younger, unconsolidated deposits are found in local topographic lows. FWD measurements acquired at 52 sites have the lowest average deflections of the six sites (table 2). Average deflections for the W5, W6, and W7 detectors are each below 1. mil. Deflections observed at detector W7 for the Upper Glen Rose Limestone, the most common geologic unit along this segment, range from less than to.9 mil. This range spans the observed range of W7 deflections for each geologic unit (fig. 22). Upper and Lower Glen Rose Limestones have similar deflection averages for each offset (fig. 23) and very low W7 averages (<.4 mil, fig. 24a). Deflections for the one Fort Terrett Limestone example are even lower than those of the Glen Rose units for detectors W5, W6, and W7. Highest deflections are observed for the Quaternary stream deposits, although W7 values for this unit are quite low (.6 mil) relative to those of similar deposits in other regions, perhaps because of roadway stiffness. W2:W7 ratios for these geologic units appear to remove the road-stiffness effect (fig. 24b). The Fort Terrett and Upper and Lower Glen Rose Limestones all have high ratios (19 to 24) that are comparable to rigid geologic units in the Central Texas Uplift. The W2:W7 ratio for Quaternary stream deposits is near 1, which is within the range of 7 to 12 observed for similar deposits in other regions. Gulf Coastal Plains Sites The Texas coastal plain, which slopes toward the Gulf of Mexico from the Edwards Plateau (fig. 4), is the largest and geologically youngest of the major physiographic regions. Bedrock types in this region are all Cenozoic sedimentary deposits that are variably lithified (table 1). This region has the highest county-average W7 deflections in Texas, ranging from 1.1 to more than 2.5 mils (fig. 2). The two sites studied in this region (sites E and F, fig. 4) are located on the central and south parts of the upper coastal plain. Site E: Texas 71, Bastrop County Site E is a 13-km segment of Texas 71 on the central part of the upper Gulf Coastal Plains (fig. 4). Average W7 deflections for Bastrop County are between 1.1 and 1.5 mils, representing the 22

29 1 West East Elevation (ft) Deflection (mils) 1 1 W1 W2 W3 W4 W5 W6 W Reference marker Qal alluvium Kft Fort Terrett Limestone GEOLOGIC UNIT Kgru Upper Glen Rose Limestone Kgrl Lower Glen Rose Limestone QAc856c Figure 21. West-east cross section, showing geologic units, elevation, and W1 through W7 deflection along U.S. 29 between reference markers 536 and 563, Blanco and Hays Counties, Central Texas (site D, figs. 2 through 4). low end of the range observed for all coastal-plain counties (fig. 2). Mapped geologic units are old relative to deposits closer to the Gulf of Mexico (fig. 3). They include Cretaceous clay and marl, Eocene mudstones, sandstones, unconsolidated clay and sand, unconsolidated Quaternary gravel, sand, and clay (fig. 25; app. A). Average deflections calculated from both sides of the roadway at 17 locations (34 deflection series) are higher for the W3 to W7 detectors than they are for any other study site (table 2). Average W7 deflection is 1.5 mils, which is in the range calculated for Bastrop County (fig. 2). There are large variations in deflections measured at individual detectors for some geologic units (W7 deflection is between.5 and 4 mils for the Hooper Mudstone, fig. 26) and relatively small variations in other geologic units (W7 deflection ranges from.9 to 1.5 mils for the Simsboro Sand). Most geologic units have large average deflections relative to geologic units in other physiographic regions (figs. 27, 28a). The largest W7 deflections, greater than 2 mils, were recorded at locations mapped as Cretaceous clay and marl units and Quaternary stream deposits. Intermediate W7 deflections of 1.1 to 1.9 mils were observed over the Midway Group, Hooper Mudstone, Simsboro Sand, and Quaternary gravel found at the highest topographic positions along the roadway (fig. 25). The smallest W7 deflections were measured over the Calvert Bluff Mudstone. 23

30 (a) 1. Qal alluvium (b) Kft Fort Terrett Limestone Deflection (mils) (c) 1. Kgru Glen Rose Limestone (Upper) (d) Kgrl Glen Rose Limestone (Lower) Deflection (mils) Figure 22. Average (heavy line) and individual (light lines) deflections for rock types mapped along U.S. 29 in Blanco and Hays Counties Detector Detector QAc858c W2:W7 ratios also suggest that the Calvert Bluff Mudstone is the most rigid of the geologic units mapped at this site (fig. 28b). The W2:W7 ratio for this unit is about 15, well below that of the most rigid units in the Central Texas Uplift and Edwards Plateau regions, but comparable to that of similar lithified sedimentary rocks in the North-Central Plains. Ratios for all other geologic units at site E are below 1, indicating materials with low rigidity (fig. 28b). Site F: Texas 16, Jim Hogg and Zapata Counties Site F is a 75-km-long segment of Texas 16 in the Rio Grande Valley on the south part of the Gulf Coastal Plains (fig. 4). Average W7 deflections for the counties crossed by this segment of Texas 16 are moderate to high, ranging from 1.1 to 1.5 mils for Jim Hogg County and 1.6 to 2. mils for Zapata County (fig. 2). Geologic units 24

31 1. (a) Qal alluvium.6 Deflection (mils) Detector Qal alluvium Kft Fort Terrett Limestone Kgru Upper Glen Rose Limestone Kgrl Lower Glen Rose Limestone QAc857c Figure 23. Average deflections for all rock units mapped along U.S. 29 in Blanco and Hays Counties. Rock unit Rock unit (b) Kft Fort Terrett Limestone Kgru Upper Glen Rose Limestone Kgrl Lower Glen Rose Limestone Qal alluvium Kft Fort Terrett Limestone Kgru Upper Glen Rose Limestone Average W7 deflection (mils) mapped along this roadway are variably lithified Cenozoic sedimentary deposits that include Eocene sandstone and clay formations, Miocene to Oligocene mudstones, Pliocene clay, and younger Quaternary eolian and alluvial sediments (fig. 29, app. A). Average deflections calculated from FWD data are either the highest or second-highest values for all segments (table 2). Average W7 deflection is 1.4 mils, a value that is within the deflection range reported for these counties (fig. 2). Individual deflections at all detectors are relatively high, particularly where Catahoula and Frio mudstones and Jackson sandstones are mapped (fig. 29). Ranges of individual deflections are large; despite differences in the average deflections Kgrl Lower Glen Rose Limestone W2:W7 deflection ratio QAc859c Figure 24. (a) Average W7 deflection and (b) W2:W7 deflection ratio by rock type along U.S. 29, Blanco and Hays Counties. for each geologic unit, many individual deflections collected over one rock type fall within a deflection range recorded for another rock type (fig. 3). W7 deflections measured over Jackson sandstones range from.8 to nearly 3. mils, a 25

32 1 West East 6 2 Elevation (ft) Deflection (mils) 1 1 W1 W2 W3 W4 W5 W6 W Reference marker Qt alluvium Qhg high gravels GEOLOGIC UNIT Eh Hooper Mudstone Emi Midway Group Ecb Calvert Bluff Mudstone Esb Simsboro Sand Kknm Kemp Clay, Corsicana and Marlbrook Marl QAc864c Figure 25. West-east cross sections, showing geologic units, elevation, and W1 through W7 deflection along Texas 71 between reference mile markers 59 and 598, Bastrop County, southeast Texas (site E, figs. 2 through 4). range that is similar to that measured for Catahoula and Frio mudstones and Laredo sandstones. Lower, but overlapping, W7 deflections are observed for the Goliad Formation (.5 to 2 mils) and the Quaternary sand sheet (.7 to 2 mils). Average deflections for each geologic unit are relatively high at all offsets (fig. 31). Highest average W7 deflections (1.6 to 1.8 mils) are found over unconsolidated Quaternary stream deposits, Jackson sandstones, and Catahoula and Frio mudstones. Lowest average W7 deflections are calculated for segments over areas where Quaternary windblown sands and the Pliocene Goliad Formation are mapped. W2:W7 ratios occupy a narrow range between about 8 and 13 (fig. 32b). These relatively low values are similar to ratios calculated over stream deposits in other physiographic regions, suggesting that much of the coastal plain is underlain by materials of low rigidity. Ratios below 1, indicating the weakest material, were calculated for Quaternary stream deposits, the Yegua Clay, and the Laredo Sandstone. The Pliocene Goliad Formation (W2:W7 = 13) is the most rigid sedimentary deposit along this segment. 26

33 (a) (b) (c) 1. Qt alluvial terrace Qhg high gravels Ecb Calvert Bluff Mudstone Deflection (mils) (d) (e) (f) 1. Esb Simsboro Sand Eh Hooper Mudstone Emi Midway Group Deflection (mils) (g) 1. Kknm Kemp Clay, Corsicana and Marlbrook Marls Detector Detector QAc866c Figure 26. Average (heavy line) and individual (light lines) deflections for rock types mapped along Texas 71 in Bastrop County. Deflection (mils) Detector 27

34 1. (a) Qt alluvial terrace 2.11 Qhg high gravels 1.19 Deflection (mils) Detector Qt alluvial terrace Qhg Quaternary high gravels Ecb Calvert Bluff Mudstone Esb Simsboro Sand Eh Hooper Mudstone Emi Midway Group Clay Rock unit Rock unit (b) Ecb Calvert Bluff Mudstone Esb Simsboro Sand Eh Hooper Mudstone Emi Midway Group Kknm Kemp Clay, Corsicana and Marlbrook Marls Qt alluvial terrace Qhg high gravels Ecb Calvert Bluff Mudstone Esb Simsboro Sand Eh Hooper Mudstone Average W7 deflection (mils) Kknm Kemp Clay, Corsicana and Marlbrook Marl QAc865c Figure 27. Average deflections for all rock units mapped along Texas 71 in Bastrop County. Emi Midway Group Kknm Kemp Clay, Corsicana and Marlbrook Marls W2:W7 deflection ratio QAc867c Rock-Type Response We further examined average FWD response for principal rock types by grouping individual geologic units into these basic types regardless of physiographic region. Principal rock types mapped are (1) unconsolidated sedimentary deposit, (2) mudstone, (3) sandstone, (4) limestone, (5) granite, and (6) metamorphic gneiss and schist. Each geologic unit mapped along the six test segments has been classified as one of these geologic types. Figure 28. (a) Average W7 deflection and (b) W2:W7 deflection ratio by rock type along Texas 71, Bastrop County. W7 averages and W2:W7 ratios for each individual rock type define ranges of observed values for the principal rock types (fig. 33). FWD response along roads built over igneous and metamorphic rocks such as granite, gneiss, and schist have small average W7 deflections (.5 to.9 mil) and high to very high W2:W7 ratios (>17). 28

35 1 North South 8 3 Elevation (ft) W1 Deflection (mils) 1 1 W2 W3 W4 W5 W6 W Reference marker Qal alluvium Qs sand sheet GEOLOGIC UNIT Ej Jackson Sandstone Ey Yegua Clay Pg Goliad Formation MOcf Catahoula, Frio Formations El Laredo Sandstone QAc852c Figure 29. North-south cross section, showing geologic units, elevation, and W1 through W7 deflection along Texas 16 between reference mile markers 758 and 84, Jim Hogg and Zapata Counties, South Texas (site F, figs. 2 through 4). Siliciclastic sedimentary units such as sandstone and mudstone have low W2:W7 ratios of between 9 and 15, but sandstone has smaller average W7 deflections than does mudstone. Unconsolidated sediments, including Quaternary alluvium and older uncemented sand and gravel, exhibit a wide range of W7 deflections (.6 to 2.3 mils), along with very low W2:W7 ratios (6 to 14) that are similar to those observed for lithified siliciclastic rocks. Although sandstone, mudstone, and unconsolidated sedimentary deposits have similar W2:W7 ratios, average W7 deflections provide a basis for discriminating among these types. FWD response along highways underlain by limestone is the most variable of that of the principal rock types (fig. 33). Some limestones have the lowest observed W7 deflections (<.5 mil) and high W2:W7 ratios (15 to 27); other limestones have W7 deflections as large as 1.8 mils and W2:W7 ratios as low as 5, values that are comparable to those of siliciclastic units. These higher W7 deflections and lower W2:W7 ratios probably indicate common clay-rich units within larger sections of limestone, weathered limestone, or greater depth to bedrock. 29

36 (a) (b) (c) 1. Qal alluvium Qs sand sheet Pg Goliad Formation Deflection (mils) (d) (e) (f) 1. MOcf Catahoula, Frio Ej Jackson Sandstone Formations Ey Yegua Clay Deflection (mils) (g) 1. El Laredo Sandstone Detector Detector QAc854c Figure 3. Average (heavy line) and individual (light lines) deflections for rock types mapped along Texas 16 in Jim Hogg and Zapata Counties. Deflection (mils) Detector 3

37 1. (a) Qal alluvium 1.78 Qs sand sheet 1.9 Deflection (mils) Qal alluvium Qs sand sheet Detector Pg Goliad Formation MOcf Catahoula, Frio Formations Ej Jackson Sandstone Ey Yegua Clay El Laredo Sandstone QAc853c Figure 31. Average deflections for all rock units mapped along Texas 16 in Jim Hogg and Zapata Counties. Rock unit Rock unit (b) Pg Goliad Formation MOcf Catahoula, Frio Formations Ej Jackson Sandstone Ey Yegua Clay El Laredo Sandstone Qal alluvium Qs sand sheet Pg Goliad Formation MOcf Catahoula, Frio Formations Ej Jackson Sandstone Ey Yegua Clay El Laredo Sandstone Average W7 deflection (mils) W2:W7 deflection ratio QAc855c Figure 32. (a) Average W7 deflection and (b) W2:W7 deflection ratio by rock type along Texas 16, Llano and Gillespie Counties. 31

38 4 Igneous and metamorphic 35 Unconsolidated Mudstone 3 Sandstone Limestone Limestone 25 Granite W2:W7 ratio 2 Metamorphic 15 Mudstone Unconsolidated 1 Sandstone Average W7 deflection (mils) QAc4628c Figure 33. Composite W7 deflection and W2:W7 deflection ratio for individual rock types mapped along Texas 16, U.S. 29, and Texas 71 test sites (sites A through F, figs. 2 through 4). 32

39 BEDROCK DEPTHS FROM SEISMIC REFRACTION We collected seismic refraction data at two sites (fig. 2) to (1) prove that refraction data can be acquired on roads, (2) investigate how useful refraction data might be in resolving FWD ambiguity in determining bedrock type and depth, and (3) measure key physical properties of material underlying roads. Pickle Research Campus Site We acquired seismic refraction data in September 1997 along Road D on the Pickle Research Campus (PRC) at The University of Texas at Austin (figs. 5, 6). Road D is an asphaltpavement road laid on an unknown thickness of road base over residual sediments and Upper Cretaceous limestone of the Austin Group (Garner and Young, 1976). Both the FWD and soil-probe hammer were used as seismic sources. Several seismic wave types are evident in a field record collected at PRC using the FWD as a seismic source (fig. 34). Types of ground motion detected by the geophones during the first 6 ms following impact of the FWD weight against the pavement include (1) high-amplitude, lowfrequency, and slowly propagating surface waves (lower left of field record, less than 28 m/s propagation velocity); (2) a direct wave, which is the first recorded signal at geophones that are less than 4 m from the source; (3) a critically refracted arrival, representing the first recorded signal at geophones greater than 4 m from the source (3, m/s propagation velocity); and (4) a reflected wave that has a hyperbolic shape, arriving at about 2 ms at the source location and about 3 ms at the maximum offset. Spectral analyses of ground motion detected by near-source geophones indicate that the FWD produces an impulse with frequencies between about 2 and 1 2 Distance from source (ft) Direct arrival Refracted arrival Figure 34. Seismic response recorded by a 48-geophone recording spread with the FWD as a seismic source. Visible phases include the direct arrival, a critically refracted arrival from the underlying rigid layer, longwavelength, low-frequency surface waves, and reflected compressional waves. Data recorded at site R1 (fig. 2). Time (ms) 3 Surface waves QAc4629c 33

40 (a) Geophone (b) Geophone Time (s) Time (s).2.2 QAC4614c (c) Geophone Figure 35. Field records from refraction test PRC SPH1 (site R1, fig. 2) acquired by using a soil-probe hammer as a seismic source. Source was located on the road shoulder at the (a) west end, (b) east end, and (c) center of the recording spread. Records displayed by using a 125-Hz low-cut filter and time-varying gain applied. Geophone spacing.3 m. Time (s).2 2 Hz, which is a useful range for seismic refraction and shallow-reflection investigations. Refraction Experiment PRC SPH1 In this test, both the soil-probe-hammer source and the 48 geophones were located on the shoulder of PRC Road D. Displays of filtered and amplified seismic energy from impulses at the west and east ends and the center of the recording spread (fig. 35) reveal the presence of a slowly propagating direct arrival at geophones nearest the source and a faster, critically refracted arrival at geophones farther from the source. 34

41 Reverse offset (m) End shot (forward) Center shot (forward) Center shot (reverse) Arrival time (s).15.1 End shot (reverse) Forward offset (m) QAc4618c Figure 36. First-arrival times for refraction test PRC SPH1 for forward- (eastward) and reverse- (westward) propagating waves. First-arrival times for each trace plotted against distance from the source can be segregated into arrivals measured when the source was west of the geophone (assigned the forward direction, fig. 36) and when the source was east of the geophone (the reverse direction). In the forward direction, arrivals group by distance into two linear segments (fig. 36). The group located closest to the source (between and about 6 m forward offset) are arrivals from the direct wave. Arrivals at greater distances belong to the compressional wave that is critically refracted by a higher velocity layer at some depth beneath the surface. The slope of a line fit to these arrival times ( x/ t) is the apparent velocity of the critically refracted wave. Extrapolating this line to zero offset gives the intercept time, which is used along with the direct and refracted velocities to calculate the depth of the refractor. Arrivals in the reverse direction can be interpreted similarly, but calculated velocities and intercept times may differ from those calculated in the forward direction if dipping layers are present. Rather than qualitatively choosing arrival-time layer assignments by viewing a time-versusdistance plot (fig. 36), we chose to examine velocity-versus-distance and intercept-time-versusdistance relationships to allow a more rigorous definition of layer assignments for forward and reverse propagation directions (figs. 37, 38). By calculating best-fit velocities progressively (gradually increasing the offset range included in the calculation), we could quantify the effect of changing layer assignments and choose the optimal offset range (figs. 37a, 38a). Similarly, the effect of changing offset ranges on calculated intercept times can be assessed (figs. 37b, 38b). Ideally, cutoff distances between arrivals assigned to the direct wave and arrivals assigned to the critically refracted wave can be consistently chosen in this manner. In the forward direction, calculated velocities for arrivals between the source and increasingly distant geophones (fig. 37a) increase to about 5 m/s by 1-m offset and remain near that velocity to an offset distance of 3 m. Including arrival times from more distant geophones in the velocity calculation causes the velocities to increase progressively with distance, suggesting that only arrivals between the source and 3 m 35

42 (a) 75 Sensors with offsets equal to or less than cutoff distance (a) 75 Sensors with offsets equal to or less than cutoff distance Velocity (m/s) 5 25 Sensors with offsets equal to or greater than cutoff distance Velocity (m/s) 5 25 Sensors with offsets equal to or greater than cutoff distance (b).2 (b).2 Refraction intercept time (s) Refraction intercept time (s) Source sensor cutoff distance (m) QAc4621c Figure 37. (a) Apparent velocity and (b) zero-offset time for forward data from refraction test PRC SPH1. Velocities and intercepts are calculated by assigning arrivals from various offset ranges to the direct or refracted arrival. Velocities and intercepts (black boxes) that are calculated from arrivals at geophones located between the source and progressively increasing source receiver distances are used to pick the optimal velocity and offset range for arrivals assigned to the direct wave. Velocities and intercepts (open boxes) that are calculated from arrivals at geophones located between the maximum source receiver distance and progressively decreasing source receiver distances are used to pick the optimal velocity, intercept time, and offset range for arrivals assigned to the critically refracted wave. belong to the direct arrival. This interpretation is confirmed by calculating zero-offset intercept times (fig. 37b), which begin increasing from the expected value of s when arrivals from geophones at source receiver distances greater than Source sensor cutoff distance (m) QAc4622c Figure 38. (a) Apparent velocity and (b) zero-offset time for reverse data from refraction test PRC SPH1. Velocities and intercepts are calculated by assigning arrivals from various offset ranges to the direct or refracted arrival. 3 m are included in the calculation. By assigning all arrivals at offsets of 3 m or less to the direct wave, we calculate a layer-1 velocity of 5 m/s (table 3). To make layer-2 assignments in a two-layer solution, velocity and intercept time can be calculated by using arrival times from geophones between the maximum source receiver distance and those progressively closer to the source. For forward data, calculated velocities reach a maximum when arrivals at geophones at distances greater than about 6 m are included in the calculation (fig. 37a). Calculated intercept times increase with increasing minimum source 36

43 Table 3. Summary of refraction data collected at the Pickle Research Campus and Jacksboro sites. Velocity, depth, and apparent dip calculated by using the slope intercept method (Palmer, 1986). Pickle Research Campus Jacksboro site Forward direction West to east South to north Experiment PRC SPH1 PRC SPH2 Jacksboro SPH1 Source surface Shoulder Pavement Pavement Sensor surface Shoulder Shoulder Pavement Layer-1 velocity (m/s) Forward Reverse Calculated Layer-2 velocity (m/s) Forward 3, , ,83.7 Reverse 2, ,42.4 3,546.3 Calculated 3, , ,62.6 Refractor intercept time (s) Forward Reverse Calculated layer-1 thickness (m) Apparent refractor dip (degrees) 1.2 west.6 west 3.3 north receiver distance (fig. 37b), suggesting that arrivals at geophones less than about 6 m from the source belong to the direct wave or an intermediate refractor. Using arrival times from geophones at distances greater than 6 m results in a calculated apparent velocity of 3,796 m/s and an intercept time of.89 s (table 3). Time and distance relationships (fig. 36) suggest a crossover distance separating direct from refracted arrivals of nearly 6 m for seismic energy propagating from east to west (reverse data). Calculated velocities for arrivals at geophones nearest the source increase to 5 m/s by a distance of 1 m from the source, remaining at that velocity to a maximum source receiver distance of 5 m (fig. 38a). Intercept times, which should be zero for direct-wave arrivals, begin increasing as source receiver distances increase beyond 5 m (fig. 38b). Nonzero intercepts for the reverse data suggest either a short delay (<.1 s) between the source impact and the onset of recording or the presence of a very shallow, low-velocity refractor. The velocity calculated for layer 1 in the reverse direction is 57 m/s, slightly higher than that calculated for the forward direction (table 3). The apparent velocity of the critically refracted arrival reaches a plateau when arrivals from geophones beyond 7 m from the source are included in the velocity calculation (fig. 38a). Higher velocities calculated for greater threshold distances suggest that arrivals from deeper, higher velocity layers have been included in the analysis. Using 7 m as the cutoff distance, the apparent velocity of the critically refracted wave is 2,911 m/s. Its extrapolated intercept time is.87 s (table 3). To use the slope intercept method (Palmer, 1986) to calculate refractor depth, we must know apparent velocities for the direct and critically refracted arrivals in the forward and reverse directions and the intercept times for the critically 37

44 refracted arrivals. Using the values mentioned earlier, we calculate the true direct-wave velocity (layer 1) to be 53 m/s, the true layer-2 velocity to be 3,295 m/s, and the thickness of layer 1 to be 2.26 m (table 3). The interface between layers 1 and 2 has an apparent dip of 1.2 westward. Calculated depths to layer 2 are 2.27 m beneath at the west end and 2.21 m at the east end of the recording spread (fig. 39a). Refraction Experiment PRC SPH2 Site and acquisition parameters for this experiment are the same as they were for PRC SPH1, except that the soil-probe hammer was offset from the receiver spread on the paved road (fig. 6) for us to examine the influence pavement might have on refraction data. The geophones remained embedded in the shoulder. Filtered and amplified ground motion recorded by the geophones, with the source located at the west end, east end, and center of the recording spread (fig. 4), is similar to that recorded for experiment PRC SPH1. The velocity calculated for the direct arrival is 531 m/s, which is the velocity of layer 1 (table 3). Because the interface between layers 1 and 2 has a small apparent dip to the west, the apparent layer-2 velocity is higher in the forward direction (eastward propagation) than it is in the reverse direction (westward propagation). Layer-2 velocity, which represents the propagation speed of the critically refracted wave, is calculated to be 3,244 m/s, similar to that obtained in experiment PRC SPH1 (table 3). Using the intercept times of.74 s at the west end of the spread and.73 s at the east end of the recording spread, we calculate depths to layer 2 to be 1.99 m at the west end and 1.97 m at the east end (fig. 39b). Interpreted Strata Measured layer-1 compressional-wave velocities of about 5 m/s are within the 3- to 9- m/s range reported for dry, unconsolidated material (Press, 1966; Wylie, 1969), suggesting that layer 1 consists largely of road base and surficial sediments above the bedrock contact. Higher velocities measured for layer 2, reaching nearly 3,25 m/s, are consistent with those expected for relatively soft limestone (Press, 1966) such as the Cretaceous Austin Chalk. The Austin Chalk, the mapped geologic unit at the site (Garner and Young, 1976), is exposed in nearby ditches. Jacksboro Site In May 1998, we acquired seismic refraction data on U.S. Highway 281 southbound, south of Jacksboro, Texas (site R2, fig. 2). Data were acquired with the seismic source and geophones on the inner-lane pavement (fig. 8). Geologic maps of the site show the Pennsylvanian Ranger Limestone as the surface geologic unit (Hentz and Brown, 1987); a thin veneer of surficial sediments mantles the Ranger Limestone in fields adjacent to the highway. The soil-probe hammer was used as the seismic source. Effect of Digital Filtering Data quality is generally similar at the Jacksboro and PRC sites (compare figs. 35a, 4a, 41d). Noise levels are somewhat higher at the Jacksboro site because of increased road traffic and poorer coupling between the ground and the plate-mounted geophones, but data quality is sufficient to detect direct and critically refracted arrivals. Accurate picking of arrival times is enhanced by digital filtering and amplification of the recorded signals (fig. 41). Low-frequency and high-amplitude surface waves dominate the field record without filtering or amplification (fig. 41a), but direct and critically refracted arrivals are apparent in the record. Application of timevarying gain balances the amplitude along each segment of a geophone trace and makes first breaks more visible (fig. 41b), yet low-frequency seismic noise in the early part of the record makes accurate picking of arrival times difficult. Applying a digital filter that removes seismic energy having frequencies below 125 Hz removes much of the low-frequency noise, revealing a weaker, earlier arrival than the critically refracted arrival evident in the unfiltered record (fig. 41c). Balancing the amplitude of the filtered record by using time-varying gain does not greatly aid the picking of first breaks, but it does reveal reflected seismic energy later in the field record (fig. 41d). Filtered and amplified field records were used 38

45 (a) West East (b) (c) Depth (m) Depth (m) Depth (m) West South Layer 1 Vp = 53 m/s Apparent dip = 1.2 westward Layer 2 Vp = 3295 m/s 5 m Layer 1 Vp = 531 m/s Apparent dip =.6 westward Layer 2 Vp = 3244 m/s 5 m 2 ft 2 ft Layer 1 Vp = 57 m/s East Apparent dip = 3.3 northward North 3 Layer 2 Vp = 2621 m/s 4 2 ft 5 m QAc4627c Figure 39. Calculated layer velocities and thicknesses and apparent dips of layer interfaces for refraction tests (a) PRC SPH1, (b) PRC SPH2, and (c) Jacksboro SPH1. 39

46 (a) Geophone (b) Geophone Time (s) Time (s).2.2 QAc4615c (c) Geophone Figure 4. Field records from refraction test PRC SPH2 (site R1, fig. 2) acquired by using a soil-probe hammer as a seismic source. Source was located on the pavement at the (a) west end, (b) east end, and (c) center of the recording spread. Records displayed with a 125-Hz low-cut filter and timevarying gain applied. Geophone spacing is.3 m. Time (s).2 for refraction analysis at the Jacksboro site (fig. 42). Refraction Analysis First-arrival times picked for the forward and reverse end shots and for the reverse part of the center shot form two groups, each with a linear trend (fig. 43). For the forward data, in which the seismic source was south of the recording geophones, near-source geophones (5 m or less from the source) record direct seismic waves that have an apparent velocity of about 585 m/s (table 3; fig. 44a) and intercept the time axis at 4

47 (a) Geophone (b) Geophone Time (s) Time (s).2.2 (c) Geophone (d) Geophone Time (s) Time (s).2.2 QAc4616c Figure 41. Effect of digital filtering and amplification on a field record from the Jacksboro site (site R2, fig. 2) acquired by using a soil-probe hammer as a seismic source. Record displayed with (a) neither digital filtering nor time-varying gain, (b) time-varying gain (2-ms window) but no digital filtering, (c) 125-Hz low-cut filter without time-varying gain, and (d) 125-Hz low-cut filter and time-varying gain applied. Geophone spacing is.5 m. s (fig. 44b). The critically refracted wave is the first arrival at source receiver distances greater than about 7 m. Including all geophones 7 m and farther from the source in the velocity and intercept calculation, the apparent velocity of the critically refracted wave becomes 2,84 m/s 41

48 (a) Geophone (b) Geophone Time (s) Time (s).2.2 QAc4617c (c) Geophone Figure 42. Field records from the Jacksboro site acquired by using a soil-probe hammer as a seismic source. Source and geophones were located on the pavement of U.S. Highway 281 at the (a) south end, (b) north end, and (c) center of the recording spread. Records displayed with a 125-Hz low-cut filter and time-varying gain applied. Geophone spacing is.5 m. Time (s).2 (fig. 44a; table 3), and its intercept time.66 s (fig. 44b). For the reverse data, in which the source was north of the recording geophones, apparent velocity decreases as the maximum included source receiver distance increases to about 6 m, remains relatively constant to a maximum included distance of about 9 m, and increases as geophones beyond 9-m source receiver distance are included (fig. 45a). Over this same distance range, intercept times decrease to near zero when arrival times at geophones at source receiver dis- 42