Skip to contentUnited States Department of Transportation - Federal Highway AdministrationSearch FHWAFeedback

Geotechnical Engineering

 
<< PreviousContentsNext >>

Geotechnical Aspects of Pavements Reference Manual

Chapter 4.0 Geotechnical Exploration And Testing

4.1 Introduction

The purpose of the geotechnical subsurface investigation program for pavement design and construction is to obtain a thorough understanding of the subgrade conditions along the alignment that will constitute the foundation for support of the pavement structure. The specific emphasis of the subsurface investigation is to identify the impact of the subgrade conditions on the construction and performance of the pavement, characterize material from cut sections that may be used as subgrade fill, and to obtain design input parameters. The investigation may be accomplished through a variety of techniques, which may vary with geology, design methodology and associated design requirements, type of project and local experience. To assist agencies in achieving the stated purpose of subsurface investigation, this chapter presents the latest methodologies in the planning and execution of the various exploratory investigation methods for pavement projects. It is understood that the procedures discussed in this chapter are subject to local variations. Users are also referred to AASHTO R 13 and ASTM D 420, Conducting Geotechnical Subsurface Investigations and FHWA NHI-01-031 Subsurface Investigations, for additional guidance.

In Chapter 1, a simplistic subsurface exploration program consisting of uniformly spaced soil borings (i.e., systematic sampling) with SPT testing was mentioned as an antiquated method for determining the subsurface characteristics for pavement design. "Adequate for design and low cost" are often used in defense of this procedure. The cost-benefit of additional subsurface exploration is a subject that is often debated. This subject is now addressed in the new NCHRP 1-37A Design Guide. The guide allows for use of default values in the absence of sufficient data for characterizing the foundation, thus minimizing agency design costs, but at the increased risk of over- or under-designing the pavement structure.

In evaluating the cost-benefit of the level of subsurface investigation, all designers must recognize that the reliability and quality of the design will be directly related to the subsurface information obtained. The subsurface exploration program indeed controls the quality of the roadway system. A recent FHWA study indicated that a majority of all construction claims were related to inadequate subsurface information. With great certainty, inadequate information will lead to long-term problems with the roadway design. The cost of a subsurface exploration program is a few thousand dollars, while the cost of over-conservative designs or costly failures in terms of construction delays, construction extras, shortened design life, increased maintenance, and public inconvenience is typically in the hundreds of thousands of dollars.

Engineers should also consider that the actual amount of subgrade soil sampled and tested is typically on the order of one-millionth to one-billionth of the soil being investigated. Compare this with sampling and testing of other civil engineering materials. Sampling and testing of concrete is on the order of 1 sample (3 test specimens, or about ¼ cubic meter) every 40 cubic meters, which leads to 1 test in 100,000. Sampling and testing of asphalt is on the same order as concrete. Now consider that the variability in properties of these well-controlled, manufactured materials is much less than the properties of the subgrade, which often have coefficients of variation of well over 100% along the alignment. Again cost, not quality is usually the deciding factor. The quality of sampling can be overcome with conservative designs (as is often the case; e.g., AASHTO 1972). For example, laboratory tests are often run on soil samples in a weaker condition than in the ground, rather than running more tests on the full range of conditions that exist in the field. While this approach may provide a conservative value for design purposes, there are hidden costs in both conservatism and questionable reliability. Modern pavement design uses averages with reliability factors to account for uncertainty (AASHTO, 1993 and NCHRP 1-37A). However, sufficient sampling and testing are required to check the variability of design parameters to make sure that they are within the bounds of reliability factors; otherwise, on highly variable sites designs, they will not be conservative and on very uniform sites, they will still be over conservative.

The expense of conducting soil borings is certainly a detriment to obtaining subsurface information. However, exploration itself is not just doing borings. There is usually a significant amount of information available from alternate methods that can be performed prior to drilling to assist in optimizing boring and sampling locations (i.e., representative sampling). This is especially the case for reconstruction and rehabilitation projects. Significant gains in reliability can be made by investigating subgrade spatial variability in a pavement project and often at a cost reduction due to decreased reliance on samples. This chapter provides guidelines for a well-planned exploration program for pavement design, with alternate methods used to overcome sampling and testing deficiencies. Geotechnical exploration requirements for borrow materials (base, subbase, and subgrades) are also reviewed.

Figure 4-1 provides a flow chart of the process for performing a geotechnical exploration and testing program. As shown in the flow chart, the steps for planning and performing a complete geotechnical and testing program include:

Subsurface Exploration StepsRelevance to Pavement Design
1) Establish the type of pavement construction.Whether new construction, reconstruction, or rehabilitation.
2) Search available information.To identify anticipated subsurface conditions at the vertical and horizontal location of the pavement section.
3) Perform site reconnaissance.To identify site conditions requiring special consideration.
4) Plan the exploration program for evaluation of the subsurface conditions and identification of the groundwater table, including methods to be used with consideration for using:
  • remote sensing,
  • geophysical investigations,
  • in-situ testing,
  • disturbed sampling, and
  • undisturbed sampling.
To identify and obtain:
  • more information on site conditions,
  • spatial distribution of subsurface conditions,
  • rapid evaluation of subsurface condition,
  • subgrade soils & classification test samples,
  • samples for resilient modulus tests and calibration of in-situ results.
5) Evaluate conceptual designs, examine subsurface drainage and determine sources for other geotechnical components (e.g., base and subbase materials).Identify requirements for subsurface drainage and subgrade stabilization requirements, as well as construction material properties.
6) Examine the boring logs, classification tests, soil profiles and plan view, then select representative soil layers for laboratory testing.Use the soil profile and plan view along the roadway alignment to determine resilient modulus or other design testing requirements for each influential soil strata encountered.

Each of these steps will be reviewed in the following sections of this chapter.

Figure 4-1. Geotechnical exploration and testing for pavement design.

Flow chart of the process for performing a geotechnical exploration and testing program for pavement design

4.2 Levels Of Geotechnical Exploration For Different Types Of Pavement Projects

There are three primary types of pavement construction projects. They are:

  • new construction,
  • reconstruction, and
  • rehabilitation.

Each of these pavement project types requires different considerations and a corresponding level of effort in the geotechnical exploration program.

4.2.1 New Pavement Construction

For new construction, the exploration program will require a complete evaluation of the subgrade, subbase, and base materials. Sources of materials will need to be identified and a complete subsurface exploration program will need to be performed to evaluate pavement support conditions. Prior to planning and initiating the investigation, the person responsible for planning the subsurface exploration program (i.e., the geotechnical engineer or engineer with geotechnical training) needs to obtain from the designers the type, load, and performance criteria, location, geometry and elevations of the proposed pavement sections. The locations and dimensions of cuts and fills, embankments, retaining structures, and substructure elements (e.g., utilities, culverts, storm water detention ponds, etc.) should be identified as accurately as practicable.

Also, for all new construction projects, samples from the subgrade soils immediately beneath the pavement section and from proposed cut soils to be used as subgrade fill will be required to obtain the design-input parameters for the specific design method used by the agency. Available site information (e.g., geological maps and United States Department of Agriculture Natural Resources Conservation Service's soil survey reports) as discussed in Section 4.3, site reconnaissance (see Section 4.4), air photos (see Section 4.5.3) and geophysical tests (see Section 4.5.4) can all prove beneficial in identify representative and critical sampling locations.

For all designs using AASHTO 1993 or NCHRP 1-37A, particularly for critical projects, repeated load resilient modulus tests are needed to evaluate the support characteristics and the effects of moisture changes on the resilient modulus of each supporting layer. The procedures, sample preparation and interpretation of the resilient modulus test are discussed in Chapter 5. For designs based on subgrade strength, either lab tests (e.g., CBR) as discussed in Chapter 5 or in-situ tests (e.g., DCP) as discussed later in Section 4.5.5 of this chapter can be used to determine the support characteristics of the subgrade.

Another key part of subsurface exploration is the identification and classification (through laboratory tests) of the subgrade soils in order to evaluate the vertical and horizontal variability of the subgrade and select appropriate representative design tests. Field identification along with classification through laboratory testing also provides information to determine stabilization requirements to improve the subgrade should additional support be required, as discussed in Chapter 7.

Location of the groundwater table is also an important aspect of the subsurface exploration program for new construction to evaluate water control issues (e.g., subgrade drainage requirements) with respect to both design and construction. Methods for locating the groundwater level are discussed in Section 4.5.6. Other construction issues include the identification of rock in the construction zone, rock rippability, and identification of soft or otherwise unsuitable materials to be removed from the subgrade. The location and rippability of rock can be determined by geophysical methods (e.g., seismic refraction), as discussed in section 4.5.4 and/or borings and rock core samples.

4.2.2 Reconstruction

For pavement reconstruction projects, such as roadway replacement, full depth reclamation, or road widening, information may already exist on the subgrade support conditions from historical subsurface investigations. Existing borings should be carefully evaluated with respect to design elevation of the new facility. A survey of the type, severity, and amount of visible distress on the surface of the existing pavement (i.e., a condition survey as described in the NHI, 1998, "Techniques for Pavement Rehabilitation" Participant's Manual) can also indicate local issues that need a more extensive evaluation. However, an additional limited subsurface investigation is usually advisable to validate the pavement design calculations and design for weak subgrade conditions, if present. It is also likely that resilient modulus, CBR or other design input values used by agencies would need to be obtained for the existing materials using current procedures. Test methods used by the agency often change over time (e.g., lab CBR versus field CBR). Previous data may also not be valid for current conditions (e.g., traffic). Water in old pavements can often result in poorer subgrade conditions than originally encountered. Drainage features, or lack thereof, in the existing pavement and their functionality should be examined. Again, subgrade soil identification and classification will be required to provide information on subgrade variability and assist in selection of soils to be tested.

It is possible to determine the value of reworking the subgrade (i.e., scarifying, drying, and recompacting) if results indicate stiffness and/or subgrade strength values are below expected or typical values. This comparison can be made by examining the resilient modulus of undisturbed tube samples obtained to verify backcalculated moduli to that of a recompacted specimen remolded to some prescribed level of density and moisture content. For example, this comparison may ultimately lead to the need for underdrain installation in order to reduce and maintain lower moisture levels in the subgrade.

Subsurface investigation on reconstruction projects can usually be facilitated by using non-destructive tests (NDT) (a.k.a. geophysical methods) performed over the old pavement (or shoulder section for road widening) with one or more of the variety of methods presented in Section 4.5. For example, resilient modulus properties can best be obtained from non-destructive geophysical methods, such as falling weight deflectometer (FWD) tests and back calculating elastic moduli to characterize the existing structure and foundation soils needed for design. This approach is suggested because it provides data on the response characteristics of the in-situ soils and conditions. Back calculation of layer elastic moduli from deflection basin data is discussed later in Section 4.5.4 of this chapter. These results can be supported by laboratory tests on samples obtained from a minimal subsurface exploration program (described in Section 4.5). Old pavement layer thickness (i.e., asphalt or concrete, base and/or subbase) should also be obtained during sampling to provide information for back-calculation of the modulus values.

For designs based on subgrade strength (e.g., CBR), in-situ tests (e.g., Dynamic Cone Penetrometer (DCP), field CBR, and other methods as described in Section 4.7) can be performed to obtain a rapid assessment of the variability in subgrade strength and to determine design strength values via correlations. Some samples should still be taken to perform laboratory tests and confirm in-situ test correlation values. Geophysical test results (e.g., FWD, Ground Penetrating Radar (GPR), and others described in Section 4.5.4) can also be used to assist in locating borings.

The potential sources of new base and subbase materials will need be identified and laboratory tests performed to obtain resilient modulus, CBR or other design values, unless catalogued values exist for these engineered materials. For pavement reclamation or recycling projects, composite samples should be obtained from the field and test specimens constituted following the procedures outlined in Chapter 5 to obtain design input values. The subgrade soils will also need to be evaluated for their ability to support construction activities, such as rubblize-and-roll type construction.

4.2.3 Rehabilitation

As discussed in Chapter 3, rehabilitation projects include a number of strategies, including overlays, rubbilization, and crack and seat. The details required for the subsurface investigation of pavement rehabilitation projects depends on a number of variables:

  • The condition of the pavement to be rehabilitated (e.g., pavement rutting, cracking, riding surface uniformity and roughness, surface distress, surface deflection under traffic, presence of water, etc., as described in the condition survey section of NHI, 1998, "Techniques for Pavement Rehabilitation" Participants Manual.)
  • If the facility is distressed, the type, severity and extent of distress (pavement distress, pavement failures, crack-type pattern, deep-seated failures, settlement, drainage and water flow, and collapse condition) (see NHI, 1998, "Techniques for Pavement Rehabilitation" Participants Manual) should be quantified. Rutting and fatigue cracking are often associated with subgrade issues and general require coring, drilling, and sampling to diagnose the cause of these conditions.
  • Techniques to be considered for rehabilitation.
  • Whether the facility will be returned to its original and as-built condition, or whether it will be upgraded, for example, by adding another lane to a pavement. If facilities will be upgraded, the proposed geometry, location, new loads and structure changes (e.g., added culverts) must be considered in the investigation.
  • The required performance period of the rehabilitated pavement section.

Selection of the rehabilitation alternative will partly depend on the condition assessment. NHI, 1998, "Techniques for Pavement Rehabilitation" Participants Manual covers condition surveys and selection of techniques for pavement rehabilitation. Information from the subsurface program performed for the original pavement design should also be reviewed. However, as with reconstruction projects, some additional corings and borings will need to be performed to evaluate the condition and properties of the of the pavement surface and subgrade support materials. Pavements are frequently cored at 150 - 300 m (500 - 1000 ft) intervals for rehabilitation projects. The core holes in the pavement also provide access to investigate the in-situ and disturbed properties of the base, subbase, and subgrade materials. Samples can be taken and/or in-situ tests (e.g., DCP) can be used to indicate structural properties, as well as layer thickness.

Geophysical tests (e.g., FWD, GPR, and others described in Section 4.5.4) can be used to assist in locating coring and boring locations, especially if the base is highly contaminated or there are indications of subgrade problems. Otherwise, the frequency of corings and borings should be increased. As with reconstruction projects, rehabilitation projects can use FWD methods and associated back-calculated elastic modulus to characterize the existing structure and foundation. Again, the FWD method is covered in Section 4.5.4 and back-calculation of layer elastic moduli from deflection basin data is discussed in Chapter 5. FWD results can also be correlated with strength design values (e.g., CBR). A limited subsurface drilling and sampling program can then be used to confirm the back-calculated resilient modulus values and/or correlation with other strength design parameters. The layer thickness of each pavement component (i.e., surface layer, base, and or subbase layer) is critical for back-calculation of modulus values.

4.2.4 Subsurface Exploration Program Objectives

As stated in the NCHRP 1-37A Design Guide, the objective of subsurface investigation or field exploration is to obtain sufficient subsurface data to permit the selection of the types, locations, and principal dimensions of foundations for all roadways comprising the proposed project, thus providing adequate information to estimate their costs. More importantly, these explorations should identify the site in sufficient detail for the development of feasible and cost-effective pavement design and construction.

As outlined in the FHWA Soils and Foundation Workshop manual (FHWA NHI-00-045), the subsurface exploration program should obtain sufficient subsurface information and samples necessary to define soil and rock subsurface conditions as follows:

  1. Statigraphy (for evaluating the areal extent of subgrade features)
    1. Physical description and extent of each stratum
    2. Thickness and elevation of various locations of top and bottom of each stratum
  2. For cohesive soils (identify soils in each stratum, as described in Section 4.7, to assess the relative value for pavement support and anticipated construction issues, e.g., stabilization requirements)
    1. Natural moisture contents
    2. Atterberg limits
    3. Presence of organic materials
    4. Evidence of desiccation or previous soil disturbance, shearing, or slickensides
    5. Swelling characteristics
    6. Shear strength
    7. Compressibility
  3. For granular soils (identify soils in each stratum, as described in Section 4.7, to assess the relative value for pavement support and use in the pavement structure)
    1. In-situ density (average and range)
    2. Grain-size distribution (gradations)
    3. Presence of organic materials
  4. Groundwater (for each aquifer within zone of influence on construction and pavement support, especially in cut sections as detailed in Section 4.5)
    1. Piezometric surface over site area, existing, past, and probable range in future
    2. Perched water table
  5. Bedrock (and presence of boulders) (within the zone of influence on construction and pavement support as detailed in Section 4.5)
    1. Depth over entire site
    2. Type of rock
    3. Extent and character of weathering
    4. Joints, including distribution, spacing, whether open or closed, and joint infilling
    5. Faults
    6. Solution effects in limestone or other soluble rocks
    7. Core recovery and soundness (RQD)
    8. Ripability

4.3 Search Available Information

The next step in the investigation process is to collect and analyze all existing data. A complete and thorough investigation of the topographic and subsurface conditions must be made prior to planning the field exploration program so that it is clear where the pavement subgrade will begin and to identify the type of soils anticipated within the zone of influence of the pavement. The extent of the site investigation and the type of exploration required will depend on this information. ("If you do not know what you should be looking for in a site investigation, you are not likely to find much of value." Quote from noted speaker at the 8th Rankine Lecture). Simply locating borings without this information is like sticking a needle in your arm blindfolded and hoping to hit the vein. A little sleuthing can greatly assist in gaining an understanding of the site and planning the appropriate exploration program.

An extensive amount of information can be obtained from a review of literature about the site. There are a number of very helpful sources of data that can and should be used in planning subsurface investigations. Review of this information can often minimize surprises in the field, assist in determining boring locations and depths, and provide very valuable geologic and historical information, which may have to be included in the exploration report.

The first information to obtain is prior agency subsurface investigations (historical data) at or near the project site, especially for rehabilitation and reconstruction projects. To determine its value, this data should be carefully evaluated with respect to location, elevation, and site variability. Also, in review of data, be aware that test methods change over time. For example, SPT values 20 to 30 years ago were much less efficient than today, as discussed in Section 4.5.5. Prior construction and records of structural performance problems at the site (e.g., excessive seepage, unpredicted settlement, and other information) should also be reviewed. Some of this information may only be available in anecdotal forms. For rehabilitation and reconstruction projects, contact agency maintenance personnel and discuss their observations and work along the project alignment. The more serious construction and/or maintenance problems should be investigated, documented if possible, and evaluated by the engineer.

In this initial stage of site exploration, for new pavement projects, the major geologic processes that have affected the project site should also be identified. Geology will be a key factor to allow the organization and interpretation of findings. For example, if the pavement alignment is through an ancient lakebed, only a few representative borings will be required to evaluate the pavement subgrade. However, in highly variable geologic conditions, additional borings (i.e., in excess of the normal minimum) should be anticipated. Geological information is especially beneficial in pavement design and construction to identify the presence and types of shallow rock, rock outcrops, and rock excavation requirements. Geological information can readily be obtained from U.S. Geological Survey (USGS) maps, reports, publications and websites (http://www.usgs.gov/), and State Geological Survey maps and publications.

Soils deposited by a particular process assume characteristic topographic features, called landforms, which can be readily identified by a geotechnical specialist or geologist. A landform contains soils with generally similar engineering properties and typically extends irregularly over wide areas of a project alignment. The soil may be further described as a residual or transported soil. A residual soil has been formed at a location by the in-place decomposition of the parent material (sedimentary, igneous, or metamorphic rock). Residual soils often contain a structure and lose strength when disturbed. A transported soil was formed at one location and has been transported by exterior forces (e.g., water, wind, or glaciers). Alluvium soils are transported by water, loess type soils are transported by wind, and tills are transported by ice. Transported soils (especially alluvium and loess) are often fine grained and are usually characterized as poorly draining, compressible when saturated, and frost susceptible (i.e., not the most desirable soils for supporting pavement systems). Sources of information for determining landform boundaries and their functional uses are given in Table 4-1.

Table 4-1. Sources of topographic & geologic data for identifying landform boundaries.
SourceFunctional Use
Topographic maps prepared by the United States Coast and Geodetic Survey (USCGS).Determine depth of borings required to evaluate pavement subgrade; determine access for exploration equipment; identify physical features, and find landform boundaries.
County agricultural soil maps and reports prepared by the U.S. Department of Agriculture's Natural Resources Conservation Service (a list of published soil surveys is issued annually, some of which are available on the web at http://soils.usda.gov/survey/online_surveys/).Provide an overview of the spatial variability of the soil series within a county.
U.S. Geological Survey (USGS) maps, reports, publications and websites (http://www.usgs.gov/), and State Geological Survey maps and publications.Type, depth, and orientation of rock formations that may influence pavement design and construction.
State flood zone maps prepared by state or U.S. Geological Survey or the Federal Emergency Management Agency (FEMA: http://www.fema.gov/) can be obtained from local or regional offices of these agencies.Indicate deposition and extent of alluvial soils, natural flow of groundwater, and potential high groundwater levels (as well as danger to crews in rain events).
Groundwater resource or water supply bulletins (USGS or State agency).Estimate general soils data shown, and indicate anticipated location of groundwater with respect to pavement grade elevation.
Air photos prepared by the United States Geologic Survey (USGS) and others (e.g., state agencies).Detailed physical relief shown; flag major problems. By studying older maps, reworked landforms from development activities can be identified along the alignment, e.g., buried streambed or old landfill.
Construction plans for nearby structures (public agency).Foundation type and old borings shown.

One of the more valuable sources of landform information for pavement design and construction are soil survey maps produced by the U.S. Department of Agriculture, Natural Resources Conservation Service, in cooperation with state agricultural experiment stations and other Federal and State agencies. The county soil maps provide an overview of the spatial variability of the soil series within a county. These are well-researched maps and provide detailed information for shallow surficial deposits, especially valuable for pavements at or near original surface grade. They may also show frost penetration depths, drainage characteristics, and USCS soil types. Knowledge of the regional geomorphology (i.e., the origin of landforms and types of soils in the region and the pedologic soil series definitions) is required to take full advantage of these maps. Such information will be of help in planning soil exploration activities. Plotting the pavement alignment on a USDA map and/or a USGS map can be extremely helpful. Figure 4-2 shows an example for a section of the Main Highway project.

Figure 4-2. Soil Survey information along the Main Highway pavement alignment.

Photo showing a plot of the Main Highway project alignment on an USDA /USGS soil survey map.

The majority of the above information can be obtained from commercial sources (i.e., duplicating services) or U.S. and state government, local or regional offices. Specific sources (toll-free phone numbers, addresses, etc.) for flood and geologic maps, aerial photographs, USDA soil surveys, can be very quickly identified through the Internet (e.g., at the websites listed in Table 4-1).

4.4 Perform Site Reconnaissance

A very important step in planning the subsurface exploration program is to visit the site with the project plans (i.e., a plan-in-hand site visit). It is imperative that the engineer responsible for exploration, and, if possible, the project design engineer, conduct a reconnaissance visit to the project site to develop an appreciation of the geotechnical, topographic, and geological features of the site and become knowledgeable of access and working conditions. A plan-in-hand site visit is a good opportunity to learn about:

  • design and construction plans.
  • general site conditions including special issues and local features, such as lakes and streams, exploration and construction equipment accessibility.
  • surficial geologic and geomorphologic reconnaissance for mapping stratigraphic exposures and outcrops and identifying problematic surficial features, such as organic deposits.
  • type and condition of existing pavements at or in the vicinity of the project.
  • traffic control requirements during field investigations (a key factor in the type of exploration, especially for reconstruction and rehabilitation projects).
  • location of underground and overhead utilities for locating in-situ tests and borings. (For pavement rehab projects, the presence of underground utilities may also support the use of non-destructive geophysical methods to assist in identifying old utility locations.
  • adjacent land use (schools, churches, research facilities, etc.).
  • restrictions on working hours (e.g., noise issues), which may affect the type of exploration, as well as the type of construction.
  • right-of-way constraints, which may limit boring locations.
  • environmental issues (e.g., old service stations for road widening projects).
  • escarpments, outcrops, erosion features, and surface settlement.
  • flood levels (as they relate to the elevation of the pavement and potential drainage issues.
  • benchmarks and other reference points to aid in the location of borehole.
  • subsurface soil and rock conditions from exposed cuts in adjacent works.

For reconstruction or rehabilitation projects, the site reconnaissance should include a condition survey of the existing pavement as detailed in NHI (1998) "Techniques for Pavement Rehabilitation." During this initial inspection of the project, the design engineer, preferably accompanied by the maintenance engineer, should determine the scope of the primary field survey, begin to assess the potential distress mechanisms, and identify the candidate rehabilitation alternatives. As part of this activity, subjective information on distress, road roughness, surface friction, and moisture/drainage problems should be gathered. Unless traffic volume is a hazard, this type of data can be collected without any traffic control, through both "windshield" and road shoulder observations. In addition, an initial assessment of traffic control options (both during the primary field survey and during rehabilitation construction), obstructions, and safety aspects should be made during this visit.

<< PreviousContentsNext >>

Contact:

 
Updated: 01/15/2014
 

FHWA
United States Department of Transportation - Federal Highway Administration