Using Falling Weight Deflectometer Data With Mechanistic-Empirical Design and Analysis, Volume III: Guidelines for Deflection Testing, Analysis, and Interpretation

CHAPTER 2. Deflection Testing Guidelines

Introduction

Pavement deflection testing is a quick and easy way to assess the structural condition of an in-service pavement in a nondestructive manner. Over the years, a variety of deflection testing equipment has been used for this purpose, from simple beam-like devices affixed with mechanical dial gauges to more sophisticated equipment using laser-based technology. Nevertheless, all pavement deflection testing equipment basically operates in the same manner—a known load is applied to the pavement, and the resulting maximum surface deflection or an array of surface deflections located at fixed distances from the load, known as a deflection basin, are measured. Figure 1 is a schematic of a deflection basin.

Figure 1. Diagram. Typical pavement deflection basin.

This chapter reviews various deflection testing equipment, presents the reasons for conducting deflection testing, describes common deflection testing patterns, discusses important factors influencing deflection measurements, and provides guidelines for conducting deflection testing.

Deflection Testing Devices

In general, there are three primary methods for conducting deflection testing: static loading, steady-state loading, and impulse loading. The following subsections describe the fundamentals of each of these testing methods, their shortcomings, and their benefits.

Static Loading

The primary device used in the static loading method is the Benkelman beam. The Benkelman beam device is based on level arm principles, where the tip of the device is placed between the dual tires of a single axle loaded to 80 kN (18,000 lbf), and the tires are inflated to 480 to 550 kPa (70 to 80 lbf/inch²) (see figure 2). The operator records the dial measurement as the pavement rebounds from the weight of the axle as the truck is moved forward. Limitations of the Benkelman beam include its inability to measure a deflection basin and only the maximum surface deflection, its relatively labor-intensive requirements for use, and its slow rate of testing that requires traffic control for stopped conditions. Perhaps the primary benefit of the Benkelman beam is that it is relatively inexpensive.

Photo courtesy of John Harvey.
Figure 2. Photo. Benkelman beam.⁽³⁾

Steady-State Loading

In steady-state loading, a nonchanging vibration using a dynamic force generator is applied to the pavement surface, and deflections are measured using velocity transducers. Devices that incorporate steady-state loading (see figure 3) can measure deflection basin. Because of the lighter loading, steady-state deflection devices are suitable for thinner pavements. These devices require traffic control during deflection testing.

Photo courtesy of John Harvey.
Figure 3. Photo. Steady-state deflection device.⁽³⁾

Impulse Loading

Impulse loading is conducted by dropping weights at various drop heights to apply an impulse load, ranging from 6.7 to 120 kN (1,500 to 27,000 lbf), to the pavement surface. Deflections are measured using seismometers, velocity transducers, or accelerometers. Devices of this type—known as FWDs and are available through various manufacturers—are capable of measuring a deflection basin and more closely simulate truck traffic loading (see figure 4). As with steady-state testing devices, traffic control is required with FWDs.

Figure 4. Photo. Impulse loading or FWD device.

Because the majority of State transportation departments use the FWD for deflection testing, that device is the focus of this report.⁽⁴⁾

Purpose of Deflection Testing

The primary purpose of deflection testing is to determine the structural adequacy of an existing pavement and to assess its capability of handling future traffic loadings. As observed in the early work by Hveem, there is a strong correlation between pavement deflections (an indicator of the structural adequacy of the pavement) and the ability of the pavement to carry traffic loadings at a prescribed minimum level of service.⁽⁵⁾ This early work attempted to identify maximum deflection limits below which pavements were expected to perform well; these limits were based on experience and observations of performance of similar pavements. This concept quickly lent itself to overlay design, in which the required overlay thicknesses could be determined based on trying to reduce maximum pavement deflections to tolerable levels.

When complete deflection basins are available, deflection testing can provide key parameters for the existing pavement structure through backcalculation of the measured pavement responses. Specifically, for hot-mix asphalt (HMA) pavements, the elastic modulus (E) of the individual paving layers can be determined, along with the resilient modulus (M_R) of the subgrade. For portland cement concrete (PCC) pavements, the elastic modulus (E) of the PCC slab and the modulus of subgrade reaction (k) can be determined. In addition, deflection testing conducted on PCC pavements can be used to estimate the load transfer efficiency (LTE) across joints or cracks as well as for the identification of loss of support at slab corners.

These parameters of the pavement layers and of the subgrade are used in pavement design procedures or in performance prediction models to estimate the remaining life or load-carrying capacity of the pavement. They can also be used in elastic layer or finite element programs to compute stresses and strains in the pavement structure and are also direct inputs in many overlay design procedures to determine the required overlay thickness needed for the current pavement condition and the anticipated future traffic loadings.

Deflection data can also be used in other ways to help characterize the condition of the existing pavement. For example, plots of deflection data along a pavement project can be examined for nonuniformity, which may suggest areas that require further investigation using other means, including destructive sampling and testing (see the section later in this report titled Computed Indices from Deflection Data). In addition, daily or seasonal deflection data can provide insight into a pavement’s response to environmental forces, including the effects of thermal curling, frozen support conditions, and asphalt stiffening. Some agencies also use deflection criteria to establish seasonal load restrictions for certain low-volume roads. Deflection testing has also seen some limited use as a means of monitoring the quality of a pavement during construction.⁽⁶⁾ Finally, a few agencies conduct deflection testing at the network level to provide a general indication of the structural capacity of the pavement structure.

Backcalculation of Deflection Data

As described previously, pavement deflection data can be analyzed in a number of ways to help provide detailed information about a specific pavement. Perhaps the most common use of deflection data is in the backcalculation process through which the fundamental engineering properties of the pavement structure, such as the modulus values of the paving layers and the subgrade, are determined. An underlying assumption in the backcalculation process is that a set of layer modulus values exists that produces the measured deflections under the applied load. It is important to note, however, that the solution may not be unique. To obtain good results, engineering judgment must be used to ensure that the modulus value selected for each layer is within a reasonable range for the material type. Backcalculation results can be highly variable owing to variability in pavement condition, subsurface condition, material properties, and pavement structure along the project.

Different backcalculation methodologies are employed for flexible and rigid pavements, but even for a specific pavement type, a number of different approaches can be used. Common procedures include iterative methods, closed-form solutions (currently available for two-layer pavement systems), and simultaneous equations (using nonlinear regression equations). However, varying results can be obtained from these approaches because of differences in the way the pavement structure is modeled. Chapter 3 provides more detailed information on recommended backcalculation procedures and approaches for both flexible and rigid pavements.

Computed Indices From Deflection Data

A number of deflection-based indices are often computed as a means of characterizing some aspect of the existing pavement structure. A few of the more common indices are described in the following subsections.

AREA Method

Hoffman and Thompson first introduced the AREA method to characterize the deflection basin for a simple two-parameter backcalculation procedure for flexible pavements, but its use has been expanded to rigid pavements as well.⁽⁷⁾ The AREA method represents the normalized area of a vertical slice through a deflection basin between the center of the test load and at varying radial distances from the test load. For a four-sensor configuration, the AREA method equation is shown in figure 5.

Figure 5. Equation. AREA method equation for a four-sensor configuration.

Where:

d₀ = Surface deflection at center of test load (inches).
d₁₂ = Surface deflection at a distance of 300 mm (12 inches) from load.
d₂₄ = Surface deflection at a distance of 600 mm (24 inches) from load.
d₃₆ = Surface deflection at a distance of 900 mm (36 inches) from load.

The AREA method equation for a seven-sensor configuration is shown in figure 6:

Figure 6. Equation. AREA method equation for a seven-sensor configuration.

Where:

d₈ = Surface deflection at a distance of 203 mm (8 inches) from load.
d₁₈ = Surface deflection at a distance of 457 mm (18 inches) from load.
d₆₀ = Surface deflection at a distance of 1,219 mm (48) inches from load.

Typical AREA values (four-sensor configuration) and D₀, the surface deflection at the center of test load (in mm (inches) are shown in table 1, while typical trends are shown in table 2.

¹Exceptions can occur.

As demonstrated in figure 7, plotting maximum deflection, AREA value, and subgrade can be used further for identifying areas needing further investigation, coring, or additional testing and analysis. In figure 7, the HMA layer thickness for the pavement section considered is greater than 150 mm (4 inches), which indicates a lower than expected AREA value has been determined for this thickness of pavement (refer to table 1 and table 2). Looking at the maximum center deflection, a higher deflection occurs over the first half of the project length and corresponds to lower subgrade modulus; conversely, lower maximum deflections are noted from milepost (MP) 211.50 to MP 211.05, with corresponding higher subgrade moduli. Coordinating the type of the plot shown in figure 7 with a pavement conditions survey can also be beneficial and assist in determining locations for any needed coring, boring, and additional material sampling.

©Washington State Department of Transportation.
1 mil = 0.0254 mm.
1 ft = 0.305 m.
1 ksi = 6,895 MPa.

Figure 7. Chart. Maximum deflection, AREA, and subgrade modulus.⁽⁸⁾

F - 1 Shape Factor

The F - 1 shape factor represents the amount of deflection basin curvature and is inversely proportional to the ratio of the pavement stiffness to the subgrade stiffness.⁽⁹⁾ The F - 1 shape factor is defined by the equation in figure 8:

Figure 8. Equation. F - 1 shape factor definition.

Where:

D₁ = Surface deflection at a distance of 300 mm (12 inches) from load (mm (inches)).
D₂ = Surface deflection at a distance of 600 mm (24 inches) from load (mm (inches)).

Base Layer Index

The Base Layer Index (BLI), sometimes referred to as the Surface Curvature Index (SCI), gives an indication of the structural condition of the base layer.⁽¹⁰⁾ Figure 9 shows the equation for BLI.

Figure 9. Equation. BLI definition.

Where:

D₃₀₀ = Surface deflection at a distance of 300 mm (12 inches) from load (mm (inches)).

Middle Layer Index

The Middle Layer Index (MLI), also referred to as the Base Curvature Index (BCI), provides an indication of the subbase structural condition.⁽¹⁰⁾ Figure 10 shows the equation for MLI.

Figure 10. Equation. MLI definition.

Where:

D₆₀₀ = Surface deflection at a distance of 600 mm (24 inches) from load (mm (inches)).

Lower Layer Index

The Lower Layer Index (LLI), also referred to as the Base Damage Index (BDI), provides an indication of the structural condition of the subgrade layers.⁽¹⁰⁾ Figure 11 shows the equation for LLI.

Figure 11. Equation. LLI definition.

Where:

D₉₀₀ = Surface deflection at a distance of 900 mm (36 inches) from load (mm (inches)).

Radius of Curvature

The radius of curvature (RoC) was developed in South Africa and provides an indication of the structural condition of the surface and base course.⁽¹⁰⁾ Figure 12 shows the equation for RoC.

Figure 12. Equation. RoC definition.

Where:

L = 200 mm (8 inches).
D₂₀₀ = Surface deflection at a distance of 200 mm (8 inches).

For BLI, MLI, LLI, and RoC, Horak and Emery determined benchmark classification for various flexible pavement sections (see table 3).⁽¹⁰⁾

1 inch = 25.4 mm.

Surface Modulus

The plot of the surface modulus (E₀) can be used to provide an indication of the layer stiffness at different equivalent depths.⁽¹¹⁾ E₀, at an equivalent depth (r), approximates a combined modulus of the underlying layers. For values of r that are greater than the total pavement equivalent thickness, E₀ is approximately equal to the subgrade modulus. The equations for E₀ are shown in figure 13.

Figure 13. Equation. Surface modulus at center of loading plate (E₀) and at distance r (E_r).

Where:

E₀ = Surface modulus at the center of the loading plate (MPa (lbf/inch²)).
E_r = Surface modulus at a distance r (MPa (lbf/inch²)).
υ = Poisson’s ratio.
σ₀ = Contact pressure under the loading plate (MPa (lbf/inch²)).
a = Radius of loading plate (mm (inches)).
r = Distance from sensor to loading center (mm (inches)).
d_r = Deflection at distance r (mm (inches)).

The equation for equivalent depth is shown in figure 14:

Figure 14. Equation. Equivalent depth definition.

Where:

h_e,n = Equivalent depth (mm (inches)).
f_i = Factor (0.8-1.0, depending on the modular ratio, thickness, and number of layers).
h_i = Thickness of layer i (mm (inches)).
E_i = Stiffness modulus of layer i (MPa (lbf/inch²)).
E_n= Stiffness modulus of layer n (MPa (lbf/inch²)).

LTE

LTE is a parameter that can be computed from deflection testing to characterize the ability of joints and cracks in rigid pavements to effectively transmit load from one side of the joint or crack to the next (see figure 15). This can be done in the field with an FWD by applying a load on one side of the joint or crack and measuring the deflections on the loaded and unloaded slabs under that loading.

©National Highway Institute
1 mm = 39.3 mil.
δ_L = Deflection at loaded slab edge.
δ_U =Deflection at unloaded slab edge.

Figure 15. Diagram. Load transfer concept.⁽¹²⁾

The equation in figure 16 is used to express deflection-based LTE.

Figure 16. Equation. Deflection-based LTE.

Where:

LTE = Load transfer efficiency (percent).
β = d_0center/d_12center, slab bending correction factor.
d_u = Deflection on the unloaded slab (mm (inches)).
d_l = Deflection on the loaded slab (mm (inches)).

In theory, the slab bending correction factor (β) is necessary because the deflections d₀ and d₁₂, measured 305 mm (12 inches) apart, would not be equal even if measured in the interior of the slab. However, this correction factor is somewhat controversial and is not always used.

The LTE definition given above is based on deflections, but LTE is sometimes defined in terms of stress as shown in figure 17.

Figure 17. Equation. Stress-based LTE.

Where:

LTE_σ = Stress LTE (percent).
σ_u = Corresponding stress at the joint of the unloaded slab (MPa (lbf/inch²)).
σ_l = Maximum stress at the joint of the loaded slab (MPa (lbf/inch²)).

Because deflections can be easily measured in the field, and because stress-based LTE is much more affected by geometry of the applied load than deflection LTE, the deflection-based LTE is the more commonly used expression for LTE.

The theoretical deflection-based LTE ranges from 0 percent (no deflection on the unloaded slab) to 100 percent (equal deflections on the loaded and unloaded slabs). Generally speaking, the following guidelines can be used to define different levels of deflection LTE:⁽¹⁾

• Excellent: 90 to 100 percent.
• Good: 75 to 89 percent.
• Fair: 50 to 74 percent.
• Poor: 25 to 49 percent.
• Very Poor: 0 to 24 percent.

Void Detection

Pumping of underlying foundation materials (i.e., base, subbase, and subgrade) from beneath a concrete slab can lead to loss of support or voids at slab corners. Although small (typically 0.25 mm (0.01 inches) or smaller), these voids can lead to significant pavement deterioration, such as faulting and corner breaks.

One method of detecting voids beneath concrete slabs is based on the analysis of corner deflections under variable loads.⁽¹³⁾ In this method, corner deflections are measured at three load levels, and the results are plotted to establish a load-deflection relationship at each corner, as shown in figure 18, which is adapted from figure III-5 in Joint Repair Methods for Portland Cement Concrete Pavements.⁽¹³⁾ The figure illustrates an example in which, for the approach joint, the load-deflection line crosses the x-axis close to 0 at 0.051 mm (0.002 inches). For the leave joint, the load-deflection line crosses the deflection axis at a much greater distance away from the origin, indicating greater deflections under the same load. A line crossing the deflection axis at a point greater than 0.076 mm (0.003 inches or 3 mil) suggests the potential for a void under the slab.

© National Academy of Sciences. Reproduced with permission
of the Transportation Research Board.
1 kip = 453.6 kg.
1 milli-inch (mil) = 0.0254 mm.

Figure 18. Graph. Example void detection plot using deflection data.⁽¹³⁾

To ensure that built-in curling of the concrete slab is not presenting a false indication of voids, deflection testing should not be conducted in the early morning when pavement slabs are typically exposed to negative temperature gradients. Higher midday temperatures should also be avoided during deflection testing to minimize the possibility of joint lockup and slab curl.

Factors Affecting Deflections

A number of factors affect the magnitude of measured pavement deflections, which can make the interpretation of deflection results difficult. To the extent possible, direct consideration of these factors should be an integral part of the deflection testing process so that the resultant deflection data are meaningful and representative of actual conditions. Recognizing and accounting for these factors before the establishment of a field testing program helps ensure the collection of useful deflection data that can be used in subsequent backcalculation analyses.⁽¹⁴⁾ The major factors that affect pavement deflections include pavement structure (type and thickness), pavement loading (load magnitude and type of loading), and climate (temperature and seasonal effects). Each of these is discussed briefly in the following subsections.

Pavement Structure

The deflection of a pavement represents an overall system response of the surface, base, and subbase layers, as well as the subgrade itself. Thus, the parameters of the surface layer (thickness and stiffness) and of the supporting layers (thickness and stiffness) all affect the magnitude of the measured deflections. Generally speaking, weaker systems deflect more than stronger systems under the same load, with the exact shape of the deflection basin related to the stiffness of the individual paving layers.⁽¹²⁾ Other pavement-related factors that can also affect deflections include the following:

• Testing near joints, edges, or cracks or in areas containing structural distress (such as alligator cracking), can produce higher deflections than testing at interior portions of the pavement.
• Random variations in pavement layer thickness can create variability in deflection.
• Variations in subgrade parameters and the presence of underlying rigid layers (such as bedrock/stiff layer or a high water table level) can provide significant variability in deflections.

Pavement Loading

One of the most obvious factors that affects pavement deflections is the magnitude of the applied load. Most modern deflection equipment can impose load levels from as little as 13 kN (3,000 lbf) to more than 245 kN (55,000 lbf), and it is important to target appropriate load levels for each application. The type of loading can also affect pavement deflection—a slow, static loading condition produces a different response than a rapid, dynamic loading condition. In general, the more rapid the loading, the shorter the load pulse, and the smaller the deflections.

Climate

Temperature is a very important factor that must be considered as part of any pavement deflection testing program. In HMA pavements, the stiffness of the asphalt layer decreases as the temperature increases, resulting in larger deflections. Therefore, correction of the measured deflections to a standard temperature (commonly 21 °C (70 °F)) is required to perform meaningful interpretations of the data. Deflections on PCC pavements are also affected by temperature because differences in temperature between the top and bottom of the slab cause the slab to curl either upward (i.e., when the slab surface is cooler than the slab bottom) or downward (i.e., when the slab surface is warmer than the slab bottom). If basin testing is conducted when the slab is curled down or if the corner testing is conducted when the slab is curled up, the slab could be unsupported and greater deflections may result. Temperature also affects joint and crack behavior in PCC pavements. Warmer temperatures cause the slabs to expand and, coupled with slab curling effects, may lock up the joints. Deflection testing conducted at joints when they are locked up results in lower joint deflections and higher load transfer efficiencies, which are misleading regarding the overall load transfer capabilities of the joint.

Testing Season

Seasonal variations in temperature and moisture conditions also affect pavement deflection response. Generally speaking, deflections are greatest in the spring because of saturated conditions and reduced pavement support and are lowest in the winter when the underlying layers and subgrade are frozen. PCC pavements are less affected by seasonal variations in support conditions.

FWD Testing Guidelines

The guidelines discussed in the following subsections are related to the physical testing equipment configuration (such as sensor locations and load levels), as well as the type and location of deflection data that are obtained during FWD testing. The discussion of equipment configuration is as generic as possible but may reflect specific capabilities found in the Dynatest® FWD equipment because this equipment is used in the Long-Term Pavement Performance (LTPP) Program.⁽¹⁵⁾

Sensor Configuration

The LTPP Program’s nine-sensor configuration is recommended for most routine roadway testing, but other configurations are also acceptable as long as the sensor configuration is known when analyzing the deflection data. The advantage to nine sensors is the ability to perform PCC joint or crack LTE testing without relocating a sensor from the HMA testing configuration. Table 4 presents commonly used seven- and nine-sensor LTPP configurations.⁽¹⁴⁾

N/A = Not applicable.

Number of Drops and Load Levels

The LTPP Program recommends multiple drops at different load levels for both HMA and PCC pavements. The different load levels vary the mass of the weight package or release it from different heights. The designated drop heights, target load, acceptable load range, and drop sequence for each pavement type are summarized in table 5.

^aThe last drop of each recorded set contains full load history data.
^bSeating drop data are not recorded in project data; drop is performed at height 3.
^cHeight 1 is not used for testing PCC pavements.
N/A = Not applicable.

The multiple drops per load level allows checking of uniformity (or variation) of the applied load and deflections. Multiple load levels also allow evaluation of nonlinear material behavior and, for PCC pavements, can be used to evaluate the potential for voids beneath slab corners. The LTPP Program testing protocol also requires seating drops (data are not collected) and a complete time history of the drop is required for the fourth drop in the testing sequence.

In deflection testing outside of the LTPP data collection program, multiple drops at each load level are often not performed so that testing productivity is increased and lane closure times are reduced. ASTM D4694, “Standard Test Method for Deflections with a Falling-Weight-Type Impulse Load Device,” recommends that at least two drops be performed, whereas the American Association of State Highway and Transportation Officials (AASHTO) T 256, Standard Method of Test for Pavement Deflection Measurements and ASTM D4695, “Standard Guide for General Pavement Deflection Measurements” suggests one or more drops at any load level.^(16-18) AASHTO T 256 and ASTM D4695 also indicate that seating drops should be recorded for the analysis of pavement conditioning and further suggests that multiple load levels be used to evaluate nonlinear behavior.^(17,18)

Based on review of the various testing protocols and studies, a test sequence of four drops at varying load magnitudes is recommended. The first drop should be a seating drop, and the next three drops should be recorded data at 27-, 40-, and 53-kN (6,000-, 9,000-, and 12,000-lbf) target loadings. This test sequence reduces the time at each test location, allows assessment of nonlinear material behavior, and can be used for evaluating the potential for voids under PCC pavements, but it does not allow for repeatability analysis. Moreover, the use of the 27- to 53-kN (6,000-12,000-lbf) load range is recommended because heavier loadings often result in lower backcalculated moduli for granular and subgrade materials.⁽¹⁾

The LTPP Program drop sequence presented in table 5 should be considered for some test locations to provide repeatability analysis, such as at the beginning of testing, end of testing, and every 100 test locations (or a minimum of 3 repeatability test locations per project). Recording the time history for at least the last drop is recommended.

Testing Location

FWD testing locations generally consist of basin tests for flexible and rigid pavements and tests at joints (either midpanel along the joint or at the slab corner) or cracks for rigid pavements. Basin tests are used for backcalculating pavement layer parameters and are generally taken at nondistressed areas for flexible pavement and at midpanel (nondistressed) locations for rigid pavements. However, the MEPDG recommends that FWD testing be conducted at distressed HMA areas as well to determine the “damaged” modulus.⁽²⁾

FWD testing is generally performed in the outermost lane (adjacent to the shoulder) for roadways with multiple lanes in one direction. The LTPP Program developed 11 test plans based on the experiment type (general or specific) and pavement type.⁽¹⁴⁾ Testing layouts similar to test plans 4 and 5 of the LTPP data collection guidelines (see figure 19) are recommended. Note that flexible pavement testing includes two lanes of basin tests, one midlane, and one in the outer wheelpath, while rigid pavement testing is also conducted at midlane and in the outer wheelpath but also includes load transfer and corner testing in addition to basin tests. For two-lane roadways, consideration can be given to staggering the test points by directional lane, assuming that the traffic levels are directionally similar. This can provide efficient testing coverage of the pavement project but does require additional traffic control planning and setup.

1 m = 3.28 ft.
P₁, P₂, P₃ = Pass through mid-lane, pavement edge, and outer wheel path, respectively.
F₁, F₃ = Measurement location along P₁ and P₃, respectively.
J₁, J₂, J₃, J₄, J₅ = Measurement location along P₁—mid-panel, along P₂—corner, along P₂— mid-panel, along P₃—joint approach, and along P₃—joint leave, respectively.
CL = Center line.

Figure 19. Diagram. Illustration of flexible and rigid pavement test plans.⁽¹⁴⁾

AASHTO T 256 and ASTM D4695 indicate testing can be conducted either at midlane or in the outer wheelpath or both, and for rigid pavements suggests that a minimum of 25 percent of the joints associated with the basin tests should be tested.^(17,18) Furthermore, for a detailed project analysis of a rigid pavement, AASHTO T 256 and ASTM D4695 recommend a closer basin testing interval and that all joints corresponding to basin tests should be tested.^(17,18)

Testing Increments

Testing increments are typically different for network-level and project-level evaluation. Network-level testing is commonly performed to obtain a general indication of the load-carrying capacity of the pavement structure as a means of identifying and prioritizing projects for maintenance and rehabilitation. Studies by several highway agencies suggest that testing intervals of between two and three points per 1.6 km (per mi) are adequate for network-level analyses.^(19,20)

For project-level testing, much closer testing intervals are required to better characterize the pavement structure. The 11 testing plans developed for the LTPP Program testing show testing intervals of 7.6, 15.2, or 30.5 m (25, 50, or 100 ft) for flexible pavements and intervals of every 10 or 20 slabs for rigid pavements.⁽¹⁴⁾ (Note that if a crack is present near midpanel, the slab is considered two effective slabs.) However, these intervals are for relatively short pavement test sections (generally 150 m (500 ft) or less). More universal guidance is offered by the MEPDG, AASHTO T 256, and ASTM D4695, which suggests basin test spacing of 30 to 150 m (100 to 500 ft) for project-level investigations.^(1,17,18) For joint testing, the MEPDG recommends that testing should be performed across joints (or cracks) every 30 to 150 m (100 to 500 ft) and also suggests that depending on the length of the project and the availability of resources, the increment can be increased to every 305 m (1,000 ft).⁽¹⁾ In addition, it is also recommended that a minimum of 12 to 15 tests be conducted per uniform test section.^(11,18)

Temperature Measurements

Temperature measurements should be collected during FWD testing. Because HMA is a temperature-dependent material, the modulus obtained during backcalculation represents the material’s temperature at the time of testing. Having accurate temperature data helps determine the correction factor to apply to the backcalculated HMA modulus to obtain a value at a standard temperature (typically 21 °C (70 °F)) for use in design.

FWD testing on PCC pavements must consider the temperature at the time the testing is conducted. Ideally, testing should be performed at a time (typically night or early morning hours) when the slab is in a neutral or flat condition (that is, the edges or center are not potentially lifted off the base). However, this may be impractical for an agency that must test many kilometers (miles) of pavement every day. In general, deflection testing on PCC pavements should be conducted when the ambient temperature is below 27 °C (80 °F). Although the backcalculation procedures for PCC pavements do not currently incorporate temperature corrections, temperature measurements are also useful in evaluating backcalculation results for PCC pavements, particularly in terms of whether the slabs are exhibiting any curling that may be affecting the results. In addition, knowledge of the temperature conditions at the time of testing assists in evaluating LTE data.

Air and Surface Temperature

Air and pavement surface temperatures should be recorded at each test location, and most FWD equipment has temperature sensors and operating software that record the data automatically. Air and surface temperatures can be used in procedures to estimate the mean temperature of the pavement but direct measurement is generally preferred.⁽⁹⁾ The daily average temperatures for the 5 days preceding testing should also be obtained, particularly if the air and surface temperatures will be used to predict the mean pavement temperature.

Temperature Gradients

The LTPP Program testing includes measuring the temperature gradient within the pavement surface layer.⁽¹⁴⁾ This is accomplished by drilling holes to varying depths and measuring the temperatures with thermometers. The LTPP Program uses up to five holes drilled to depths summarized in table 6. (Note that hole depths that extend into the unbound layer may be eliminated, and the deepest hole should be drilled 25.4 mm (1 inch) above the bottom of the bound layer.)

Holes are generally drilled at one end of the project section in the outer wheelpath, and temperature readings are obtained at the beginning and end of the testing, as well as at selected intervals. Although the LTPP Program recommends retrieving temperatures every 30 min, this may not be practical given the time restraints of many project site closures; temperature readings every 1 h are recommended as a more practical interval. A minimum of three temperature readings, roughly correlated with the beginning, middle, and end of testing, should be obtained for smaller projects with shorter testing times.

When direct measurement of the temperature gradient is performed, the air and surface temperatures should also be taken at the temperature holes. This allows correlation of the air and surface temperatures at each test location to the measured mean pavement temperature.

Joint/Crack Opening

The LTPP Program recommends collecting joint (and crack) width measurements at a minimum of 25 percent of the joint (or crack) deflection testing locations; however, if time allows, measurement at all testing locations is preferred.⁽¹⁴⁾ For joints, the sawcut width is measured, and for cracks, the width of the full-depth crack (not necessarily the surface width) is measured. Joint/crack measurements can be reviewed during the analysis of LTE. In general terms, a tight joint/crack should have higher LTEs.

Safety Guidelines

Safety during FWD testing applies to the operation of the equipment and working in (or adjacent to) moving traffic. The FWD includes high-pressure hydraulics, electronics, and heavy moving parts that create many potential work hazards. Equipment manufacturers provide extensive documentation on the operation and maintenance of the testing equipment, and it is strongly recommended that operators become familiar with the documented materials and are well trained with the equipment.

Working around moving traffic can be a hazardous situation regardless of the work activity. Traffic control measures and work zone requirements must adhere to the guidelines of the governing agency.

Summary of FWD Testing Recommendations

Table 7 provides an overall summary of the FWD testing recommendations described in the previous subsections. Recommendations include sensor configuration, load levels and drops, testing locations, testing increments, and temperature measurements.

Data Checks

Types of Errors

The following data checks should be enabled in the FWD data collection software to flag certain conditions suggestive of errors or problems:⁽¹⁴⁾

• Roll-off (i.e., deflection sensor does not return to near 0 within 60 ms of the trigger activation): The roll-off error can occur when there is poor contact between the pavement surface and the deflection sensor or when the magnitude of the deflection approaches the resolution of the geophone.
  
  
• Non-decreasing deflections (i.e., deflection measurements do not decrease with increasing distance from the load plate): This error may occur if a transverse crack or other discontinuity exists between two adjacent geophones or on very stiff PCC pavements if the deflection difference is less than the random error inherent to the deflection sensor. If neither condition exists, this error can typically occur owing to poor seating between the sensor and the pavement surface.
  
  
• Overflow (i.e., measured deflections exceed the range of the deflection sensor): In general, this error is only expected with extremely weak pavements or when testing at extremely high load levels. (See next section for a discussion on dealing with deflection errors.)
  
  
• Load variation (the peak load for multiple drops at the same height varies by more than ± (0.18 kN + 0.02 kN × applied load) (± (40.5-lbf + 0.02 × applied load)): This condition may occur on an extremely weak pavement if the FWD testing has damaged the pavement structure, when testing during a spring thaw or if the load plate is not properly seated on the pavement surface. (See next section for a discussion on dealing with load errors.)
  
  
• Deflection variation (the load-normalized peak deflections from repeat drops varies by more than ± (2 μm + 0.01μm × deflection) (± (0.079 mil + 0.01× measured deflection)): This error can occur for pavement structures that are unaffected by FWD testing (e.g., extremely weak pavements or the unbound layers of the pavement are saturated). In addition, it can be caused by poor seating of either the load plate or the deflection sensors, or possibly by vibration caused by heavy equipment (e.g., trucks traveling in an adjacent lane) in the vicinity of the FWD test. (See next subsection for discussion on dealing with deflection errors.)

Addressing Deflection Errors

When deflection errors are encountered during FWD testing, the following steps are recommended to resolve the issue:⁽¹⁴⁾

1. Verify the condition of FWD by ensuring the deflection sensor(s) is seated securely to the sensor holder(s), all screws holding the sensor magnet and sensor holder are tight, and the holder springs and foam bushing are in good shape. If multiple sensors have errors, check all analog connections.
   
   
2. Verify the pavement condition by ensuring the sensor holder is not resting on a loose stone or crack.
   
   
3. Reject the original data and repeat the test without moving the FWD.
   
   
4. If the error persists and the FWD can be repositioned, move forward 0.6 m (2 ft) and retest. If the error still persists, accept the data and note that the error could not be resolved.
   
   
5. If the error persists and the FWD cannot be repositioned (e.g., load transfer test), accept the test and note that the error could not be resolved.

Addressing Load Errors

If load errors are experienced, the following steps are recommended to resolve the issue:⁽¹⁴⁾

1. Reject the data and retest without repositioning the FWD.
   
   
2. If the error persists, check all analog connections to ensure the weight/height targets are tight, raise the load plate and ensure the swivel moves easily, and ensure the rubber sheet and pavement surface beneath the load plate are clear of debris.
   
   
3. Reject the data and repeat the test without repositioning the FWD.
   
   
4. If the error persists and the FWD can be repositioned, move forward 0.6 m (2 ft) and retest. If the error still persists, accept the data and note that the error could not be resolved.
   
   
5. If error persists and the FWD cannot be repositioned (e.g., load transfer test), accept the test and note that the error could not be resolved.

Additional Sources for Deflection Testing Guidelines

In addition to the LTPP Program (Federal Highway Administration (FHWA)), AASHTO, and ASTM International documents that are cited as primary source documents in this chapter, a number of additional sources provide guidance on FWD testing and data collection. These include documents prepared by NCHRP, the Department of Defense, the Federal Aviation Administration, and the European Commission Directorate General Transport. (See references 4, 21, 22, and 11.) In addition, many highway agencies have developed their own custom FWD testing procedures and protocols.

FWD Calibration

Routine FWD calibration is a vital component to ensure accurate loading and deflection measurements. As outlined in AASHTO R32-09, FWD calibration should include the following:⁽²³⁾

• Annual calibration of the load cell and deflection sensors using an independently calibrated reference device (referred to as reference calibration): Deflection sensors are also compared with each other (referred to as relative calibration). Annual calibration should also be conducted as soon as possible after load cell or deflection sensor replacement. Annual calibration is performed by a certified technician.
  
  
• Monthly relative calibration of the deflection sensors: Monthly deflection sensor calibration is conducted using a relative calibration stand supplied by the FWD manufacturer and is different than the relative calibration conducted during annual calibration. Relative calibration should also be conducted immediately after replacement of a deflection sensor. Relative calibration does not require a certified technician.

Summary

This chapter presents an overview of deflection testing. Pavement deflection testing is recognized as a reliable, quick, and inexpensive method for determining the structural condition of existing pavements. Specifically, deflection measurements can be used for backcalculating the elastic moduli of the pavement structural layers and for estimating the load-carrying capacity of both HMA and PCC pavements. In addition, in PCC pavements, loss of support at slab corners can be identified and evaluation of the joint or crack load transfer can be performed using deflection testing.

Pavement deflections represent an overall system response of the pavement structure and subgrade soil to an applied load. The major factors that affect pavement deflections can be grouped into categories of pavement structure (type and thickness), pavement loading (load magnitude and type of loading), and climate (temperature and seasonal effects). Consideration of these factors should be an integral part of the deflection testing process so that the resultant deflection data are meaningful and representative of actual conditions.

Overall recommendations for setting up a FWD testing program are presented, including sensor configuration, loading levels and drop sequencing, testing locations and intervals, and temperature measurements. In addition, the types of errors commonly encountered during FWD testing are briefly described, along with ways of addressing these items during the testing program.