U.S. Department of Transportation
Federal Highway Administration
1200 New Jersey Avenue, SE
Washington, DC 20590
202-366-4000


Skip to content U.S. Department of Transportation/Federal Highway AdministrationU.S. Department of Transportation/Federal Highway Administration

Bridges & Structures

  
<< PreviousContentsNext >>

Geotechnical Aspects of Pavements Reference Manual

Chapter 6.0 Pavement Structural Design And Performance

6.1 Introduction

Previous chapters have described in qualitative terms the many geotechnical factors influencing pavement design and performance, the wide range of geotechnical properties required as input to the design procedures, and the various methods for determining the values of these geotechnical inputs. Now it is time to evaluate quantitatively the importance of these factors and properties. These are the primary objectives of the present chapter:

  1. To illustrate via examples how the geotechnical properties described in Chapter 5 are incorporated in the pavement design calculations; and
  2. To highlight the effects of the geotechnical factors and inputs on pavement design and performance.

These objectives will be met through a series of design scenarios. First, a set of reference or baseline flexible and rigid pavement designs are developed for a hypothetical and simple project scenario. Then, the effects of various deviations from the baseline conditions will be investigated and quantified. These include:

  • Soft/weak subgrade conditions
  • Subgrade stabilization
  • Low quality base/subbase material
  • Drainage and water conditions
  • Shallow bedrock conditions

The design scenarios are intentionally highly idealized and simplified. Their point is to emphasize in quantitative terms how changes in geotechnical inputs affect the overall pavement design and performance. In a sense, these design scenarios are examples of the types of sensitivity studies one should perform during design to evaluate the importance of the various design inputs, especially with reference to the quality of the information used in their estimation.

All of the design scenarios described in this chapter are for new construction (or reconstruction). This is not to minimize the importance of rehabilitation design; as described in Chapter 1, most pavement design today is in fact for rehabilitation and not new construction. However, most structural rehabilitation designs focus on restoration of the surface layer, either through asphalt concrete overlays, concrete pavement restoration, or a combination of the two. For these types of scenarios, the geotechnical inputs are essentially the same as for new construction design - e.g., a subgrade is a subgrade whether it is beneath a new or an existing pavement. The principal things that change are the methods by which the geotechnical inputs are determined - e.g., MR backcalculated from FWD tests instead of measured in the laboratory. The ways that these geotechnical inputs are used in the design calculations and the effects that they have on the design pavement structure are similar for new construction, rehabilitation, and reconstruction designs.

Both the current 1993 AASHTO Design Guide and the forthcoming new design guide from NCHRP Project 1-37A are applied to the scenarios in this chapter. Summaries of each of these design procedures are provided in Appendices C and D, respectively. Calculations for the 1993 AASHTO Guide designs are based on simple spreadsheet evaluation of the flexible and rigid pavement design equations. The calculations for the mechanistic-empirical designs were performed using Final Report Release version 0.700 (4/7/2004) of the NCHRP 1-37A software. This is the final version of the software as submitted to NCHRP at the conclusion of Project 1-37A.

It is important to keep in mind the significant differences between the two design procedures. The 1993 AASHTO Design Guide is an empirical methodology in which typically the design traffic, environmental conditions, and maximum serviceability (performance) loss are specified, and the corresponding required pavement structure-typically described just in terms of layer thicknesses-is determined. The NCHRP 1-37A procedure is a mechanistic-empirical methodology in which the design traffic, environmental conditions, and pavement structure are specified, and the corresponding pavement performance vs. time is predicted. In the NCHRP 1-37A procedure, several trial designs generally need to be evaluated in an iterative fashion in order to find the one (or more than one) that meets the design performance requirements.

Typically, there are multiple pavement designs that can provide the required performance for any scenario. This is true for both the 1993 AASHTO and NCHRP 1-37A design methodologies. The final selection of the "best" design should be based upon life-cycle costs, constructability, and other issues. Crude economic evaluations can be made in terms of initial construction costs, although even this is difficult because of large region-to-region variations in unit costs.

One final note regarding units of measure in this chapter: FHWA policy is to report values in SI units in all reports, with the corresponding U.S. Customary equivalent in parentheses. This is not done here. All values for the design examples in this chapter are reported in U.S. Customary units. There are two important practical reasons that dictate this choice. First, the structural layer coefficients, empirical correlations, and other data in the 1993 AASHTO Design Guide were developed and are presented in U.S. Customary units only. Although many of these could be converted to SI units, the consequences would be confusing. For example, expressing flexible pavement layer thicknesses in millimeters would require changing the values of the structural layer coefficients to quantities that would be unfamiliar to pavement engineers (e.g., the a1 value for asphalt concrete would change from 0.44 to 0.017 if asphalt thickness were expressed in millimeters). Second, although an SI version of the NCHRP 1-37A software is planned, at the time of this writing, only the U.S. Customary version is available. All of the outputs from the software are expressed in U.S. Customary units, and, consequently, the inputs described in this report are left in U.S. Customary units as well for consistency. The general SI-U.S. Customary conversion table included at the beginning of this reference manual can be used, if necessary, for converting units in this chapter.

6.2 Baseline Designs

The baseline designs for flexible and rigid pavements are intended to provide very simple and ordinary reference cases that can be used as the basis for subsequent exploration of the effects of various geotechnical inputs. The design scenario is based on the following assumptions:

  • New construction
  • Simple pavement structure
    • Hot mix asphalt concrete (HMA) over crushed stone graded aggregate base (GAB) over subgrade (SG) for flexible pavements
    • Jointed plain concrete pavement (JPCP) over graded aggregate base (GAB) over subgrade (SG) for rigid pavements
  • Excellent, non-erodable base material (AASHTO A-1-a crushed stone)
  • Nonexpansive subgrade
  • Benign environmental conditions - e.g., no frost heave/thaw or expansive soils
  • Good drainage
  • Simple traffic conditions - e.g., no traffic growth over design period

Reference pavement designs consistent with these assumptions have been developed for a hypothetical new arterial highway outside of College Park, MD. The roadway is assumed to have two lanes in each direction and significant truck traffic consisting primarily of Class 9 5-axle tractor-trailer units. The subgrade conditions are a non-expansive silty clay subgrade (AASHTO A-7-5/USCS MH material), a deep groundwater table, and no shallow bedrock. Environmental conditions in the Mid-Atlantic region are mild, so frost heave/thaw is not a design issue, and seasonal variations of the unbound material properties are expected to be minor.

Table 6-1 provides some typical in-place initial construction unit costs for paving materials in Maryland. These costs will be used for rough economic evaluations of the designs developed in this chapter.

Table 6-1. Typical in-place unit material costs for use in example design problems (MDSHA, 2002).
MaterialReasonable RangeTypical Unit PriceTypical Unit Price
Hot mix asphalt concrete (12.5 mm PG 64-22)$30-$50/ton$36/ton$14,250/lane-mi-in
Portland cement concrete (PCC) without steel$110-$180/cy$144/cy$28,200/lane-mi-in
Graded aggregate base$24-$60/cy$42/cy$8,200/lane-mi-in
6.2.1 1993 AASHTO Design
Flexible Pavement

The baseline flexible pavement design is a three-layer system consisting of an asphalt concrete (AC) surface layer over a nonstabilized graded aggregate base (GAB) layer over subgrade (SG). The input parameters for the baseline design using the 1993 AASHTO flexible procedure for new pavements are summarized in Table 6-2. Refer to Chapter 5 for detailed explanations of all input parameters and the methods available for their determination.

Table 6-2. Input parameters for 1993 AASHTO flexible pavement baseline design.
Input ParameterDesign ValueNotes
Initial service life15 years1
Traffic (W18)6.1x106 ESALs2
Reliability90%3
Reliability factor (ZR)-1.282 
Overall standard error (So)0.451
Allowable serviceability deterioration (ΔPSI)1.74
Subgrade resilient modulus (MR)7,500 psi5
Granular base typeAASHTO A-1-a1
Granular base layer coefficient (a2)0.186
Granular base drainage coefficient (m2)1.07
Asphalt concrete layer coefficient (a1)0.441

Notes:

  1. Typical value for flexible pavement design.
  2. Consistent with more detailed traffic input in the NCHRP 1-37A design (Section 6.2.2).
  3. Typical value for a principal arterial (AASHTO, 1993).
  4. Typical value for flexible pavements. No serviceability reduction for swelling or frost heave.
  5. Consistent with the Level 3 default input for this soil class in the NCHRP 1-37A design after adjustment for seasonal effects (Section 6.2.2).
  6. Corresponds to an MR value of 40,000 psi, which is consistent with the Level 3 default input for this soil class in the NCHRP 1-37A design (Section 6.2.2).
  7. Representative of good drainage and moderate (5-25%) saturation conditions; matches value typically used by the Maryland State Highway Administration for design.

The methodology by which the input parameters in Table 6-2 are used to determine the final structural design in the 1993 AASHTO Guide is described in Appendix C. The calculations are sufficiently straightforward that they can be easily performed using a spreadsheet. The key output from the 1993 AASHTO design methodology is the required pavement structure, which is determined as follows:

  • Required overall structural number SN = 4.61
  • Required structural number for asphalt concrete surface layer SN1 = 2.35
  • Required minimum thickness of asphalt D1 = SN1 / a1 = 5.3 inches1
  • Remaining structural number required for granular base layer SN2 = SN - D1a1 = 2.28
  • Required thickness of granular base inches D2 = SN2 / m2a2 = 12.7 inches1

Since the ratio of the layer coefficients (a1 / a2 = 0.44 / 0.18 = 2.44) is greater than the ratio of the associated in-place unit costs per lane-mile-inch of thickness in Table 6-1 ($14,250 / $8,200 = 1.74), there is an economic benefit from substituting granular base thickness with additional asphalt in this design - i.e., replacing 2.4 inches of granular base with an additional 1.0 inch of asphalt concrete is both structurally feasible (at least in terms of the 1993 AASHTO Guide) and economically beneficial (at least in terms of initial construction costs) since it would result in a savings of about $5400 per lane mile at the same SN value. However, in order to avoid complicating comparisons between the various design scenarios later in this chapter, the baseline flexible pavement structure will be kept at 5.3 inches of AC over 12.7 inches of GAB (or 5.5 inches of AC over 13 inches of GAB after rounding).

Although the 15-year initial service life specified for this scenario is typical for flexible pavements and equal to the values used in the design examples in the 1993 AASHTO Guide, current trends are toward longer life or "perpetual" pavement designs. The required pavement section for a 30-year initial service life based on the 1993 AASHTO Guide is 6.0 inches of AC over 13.7 inches of GAB. The "premium" for an additional 15 years of pavement life is thus only about three-quarters of an inch of asphalt and one inch of crushed stone base.

Rigid Pavement

The baseline rigid pavement design is a three layer JPCP system consisting of a Portland cement concrete (PCC) slab over a nonstabilized graded aggregate base (GAB) layer over subgrade (SG). The input parameters for the baseline design using the 1993 AASHTO rigid procedure for new pavements are summarized in Table 6-3. Refer to Chapter 5 for detailed explanations of all input parameters and the methods available for their determination. The rigid pavement design inputs are consistent with those used for the baseline flexible pavement design.

Table 6-3. Input parameters for 1993 AASHTO rigid pavement baseline design.
Input ParameterDesign ValueNotes
Initial service life25 years1
Traffic (W18)16.4x106 ESALs2
Reliability90%3
Reliability factor (ZR)-1.282 
Overall standard error (So)0.351
Allowable serviceability deterioration (ΔPSI)1.94
Terminal serviceability level (pt)2.51
Subgrade resilient modulus (MR)7,500 psi5
Granular subbase typeAASHTO A-1-a1
Granular subbase resilient modulus (ESB)40,000 psi6
Drainage coefficient (Cd)1.07
Loss of Support (LS)2.08
PCC modulus of rupture (Sc)690 psi1
PCC modulus of elasticity (Ec)4.4x106 psi1
Joint load transfer coefficient (J)2.89

Notes:

  1. Typical value for rigid pavement design.
  2. Consistent with more detailed traffic input in the NCHRP 1-37A design (Section 6.2.2).
  3. Typical value for a principal arterial (AASHTO, 1993).
  4. Typical value for rigid pavements. No serviceability reduction for swelling or frost heave.
  5. Consistent with the Level 3 default input for this soil class in the NCHRP 1-37A design after adjustment for seasonal effects (Section 6.2.2).
  6. Consistent with the Level 3 default input for this soil class in the NCHRP 1-37A design (Section 6.2.2).
  7. Representative of good drainage and moderate (5-25%) saturation conditions.
  8. Within AASHTO-recommended range for unbound granular materials.
  9. Typical value for JPCP with tied PCC shoulders and dowelled joints.

The methodology by which the input parameters in Table 6-3 are used to determine the final structural design in the 1993 AASHTO Guide is described in Appendix C. The calculations are sufficiently straightforward that they can be easily performed using a spreadsheet. The key output from the 1993 AASHTO design methodology is the required pavement structure. This is determined from the input parameters in Table 6-3 and the following additional intermediate steps:

  • Assume a 6-inch design thickness for the granular subbase. This value is typical for rigid pavements on reasonably competent subgrades.
  • Determine the composite modulus of subgrade reaction k representing the combined stiffness of the subgrade and the subgrade layer. For a 6-inch subbase thickness and unbound moduli as given in Table 6-3, k = 423 pci.
  • Correct k for loss of support to determine the design modulus of subgrade reaction keff. (NOTE: No shallow bedrock correction is required for the assumed subgrade conditions.) For k = 423 pci and LS = 2, keff = 38 pci.
  • Determine the required slab thickness D = 10.4 inches from the rigid pavement design equation.2

The final design selection based upon the design inputs in Table 6-3 is therefore a 10.4 inch PCC slab over 6 inches of GAB (or 10.5 inches of PCC over 6 inches of GAB after rounding).

6.2.2 NCHRP 1-37A Design
Flexible Pavement

Consistent with the 1993 AASHTO design, the baseline flexible pavement structure for the NCHRP 1-37A design methodology is a three-layer new construction consisting of an asphalt concrete (AC) surface layer over a nonstabilized graded aggregate base (GAB) layer over subgrade (SG). However, the input parameters required for the NCHRP 1-37A methodology are considerably more extensive than those for the 1993 AASHTO Design Guide. The NCHRP 1-37A design inputs are summarized in Table 6-4. Refer to Chapter 5 for detailed explanations of all input parameters and to Appendix D for a summary of the NCHRP 1-37A design procedure. All of the inputs correspond to Level 3 quality, and the default values provided within the NCHRP 1-37A software are used wherever appropriate.

The NCHRP 1-37A procedure requires the evaluation of several trial pavement sections in order to find the design that best meets the performance requirements. The baseline pavement structure from the 1993 AASHTO design procedure can be conveniently taken as the initial trial section. The predicted rutting performance for a trial section corresponding to the 1993 AASHTO design of 5.3 inches of AC over 12.7 inches of GAB is shown in Figure 6-1; total rutting (after adjustment for reliability) at the end of the 15-year initial service life is 0.646 inches. Fatigue and thermal cracking are negligible for this design scenario, and rutting is the controlling distress type. The design limit for predicted total rutting is an explicit input in the NCHRP 1-37A procedure that would, in general, be set by individual agency policy. For the examples in this chapter, however, the design limit for total rutting is taken as the predicted rutting for the 1993 AASHTO design section in order to make the 1993 AASHTO and NCHRP 1-37A designs equivalent for the baseline conditions. The design limit for total rutting (after adjustment for reliability and rounding) is thus 0.65 inches, which is slightly less than the 0.75 inch default value in the NCHRP 1-37A software.

Figure 6-1. Predicted rutting performance for NCHRP 1-37A baseline flexible pavement design.

GRAPH: Graph showing predicted rutting performance for a trial section corresponding to the 1993 AASHTO design of 5.3 inches of AC over 12.7 inches of GAB (caption of figure refers to this as NCHRP 1-37A baseline flexible pavement design).

Table 6-4. Input parameters for NCHRP 1-37A flexible pavement baseline design.
Input ParameterDesign ValueNotes
General Information
Design life15 years1
Base/subbase construction monthSeptember2
Pavement construction monthSeptember2
Traffic open monthOctober2
Site/Project Identification
Functional classPrincipal Arterials - Others 
Analysis Parameters
Initial IRI63 in./mi2
Terminal IRI172 in./mi2
Alligator cracking limit25%2
Total rutting limit0.65 inSee text
Reliability90% 
Traffic
Initial two-way AADTT2000 
Number of lanes in design direction2 
Percent of trucks in design direction50%2
Percent of trucks in design lane95%2
Operational speed55 mph2
Monthly adjustment1.0 throughout2
Vehicle class distributionLevel 3 defaultsTable 6-5
Hourly distributionLevel 3 defaultTable 6-6
Traffic growth factor0% 
Axle load distribution factorsLevel 3 defaultsTable 6-7
Mean wheel location from edge18 in2
Traffic wander standard deviation10 in2
Design lane width12 ft2
Number of axles per truckLevel 3 defaultsTable 6-8
Average axle outside width8.5 ft2
Dual tire spacing12 in2
Tire pressure120 psi2
Tandem axle spacing51.6 in2
Tridem axle spacing49.2 in2
Quad axle spacing49.22
Climate
Latitude38.98° 
Longitude-76.94° 
Elevation48 ft 
Depth of water table20 ft 
College Park, MD climate dataGenerated3
Thermal Cracking
Average AC tensile strength at 14°F366.5 psi4
Creep test duration100 sec4
Creep complianceLevel 3 defaults4
Mixture VMA14.1%5
Aggregate coefficient of thermal contraction5x10-6/°F4
Drainage and Surface Properties
Surface shortwave absorptivity0.852
Infiltrationn/a2
Drainage path lengthn/a2
Pavement cross slopen/a2
AC Surface Layer
Cumulative % retained on 3/4 inch sieve45
Cumulative % retained on 3/8 inch sieve395
Cumulative % retained on #4 sieve595
% passing #200 sieve35
Asphalt binder gradePG 64-225
Reference temperature70°F2
Effective binder content10.1%5
Air voids4.0%5
Total unit weight151 pcf5
Poisson's ratio0.352
Thermal conductivity0.67 BTU/hr-ft-°F2
Heat capacity0.23 BUT/lb-°F2
Granular Base Layer
Unbound material typeAASHTO A-1-a 
Analysis typeICM Inputs 
Poisson's ratio0.352
Coefficient of lateral pressure K00.52
Modulus40,000 psi2,6
Plasticity index1% 
% passing #200 sieve3 
% passing #4 sieve20 
D608 mm 
Compaction stateCompacted2
Maximum dry unit weight122.2 pcf2
Specific gravity of solids2.662
Saturated hydraulic conductivity263 ft/hr2
Optimum gravimetric water content11.1%2
Calculated degree of saturation82%2
SWCC parameter af11.1 psi2
SWCC parameter bf1.832
SWCC parameter cf0.512
SWCC parameter hr361 psi2
Compacted Subgrade (top 6 inches)
Unbound material typeAASHTO A-7-5 
Analysis typeICM Inputs 
Poisson's ratio0.352
Coefficient of lateral pressure K00.52
Modulus12,000 psi2,6
Plasticity index30%2
% passing #200 sieve852
% passing #4 sieve992
D600.01 mm2
Compaction stateCompacted 
Maximum dry unit weight97.1 pcf2
Specific gravity of solids2.752
Saturated hydraulic conductivity3.25x10-5 ft/hr2
Optimum gravimetric water content24.8%2
Calculated degree of saturation88.9%2
SWCC parameter af301 psi2
SWCC parameter bf0.9952
SWCC parameter cf0.7322
SWCC parameter hr1.57x104 psi2
Natural Subgrade (beneath top 6 inches)
Unbound material typeAASHTO A-7-5 
Compaction stateUncompacted 
Maximum dry unit weight87.4 pcf2
(other properties same as for compacted subgrade)
Distress Potential
Block crackingNone2
Sealed longitudinal cracks outside wheel pathNone2

Notes:

  1. Typical initial service life for flexible pavement design.
  2. Level 3 default/calculated/derived value from NCHRP 1-37A software.
  3. Based on interpolated climate histories at IAD, DCA, and BWI airports.
  4. Level 3 default/calculated/derived values from NCHRP 1-37A software for baseline AC mixture properties. Thermal cracking is not expected for the baseline design. However, these values are included here because they will be used in subsequent design scenarios.
  5. Based on a Maryland State Highway Administration 19.0mm Superpave mix design.
  6. Default input value at optimum moisture and density conditions before adjustment for seasonal effects (adjustment performed internally within the NCHRP 1-37A software).
Table 6-5. AADTT distribution by truck class
(Level 3 defaults for Principal Arterials - Others).
Class 41.3%
Class 58.5%
Class 62.8%
Class 70.3%
Class 87.6%
Class 974.0%
Class 101.2%
Class 113.4%
Class 120.6%
Class 130.3%
Table 6-6. Hourly truck traffic distribution
(Level 3 defaults for Principal Arterials - Others).
By period beginning:
Midnight2.3%Noon5.9%
1:00 am2.3%1:00 pm5.9%
2:00 am2.3%2:00 pm5.9%
3:00 am2.3%3:00 pm5.9%
4:00 am2.3%4:00 pm4.6%
5:00 am2.3%5:00 pm4.6%
6:00 am5.0%6:00 pm4.6%
7:00 am5.0%7:00 pm4.6%
8:00 am5.0%8:00 pm3.1%
9:00 am5.0%9:00 pm3.1%
10:00 am5.9%10:00 pm3.1%
11:00 am5.9%11:00 pm3.1%
Table 6-7. Truck axle load distributions: Percentage of axle loads by truck class for single axle configurations (Level 3 defaults).
Axle Weight (lbs)Truck Class
45678910111213
30001.8010.052.472.1411.651.743.643.556.688.88
40000.9613.211.780.555.371.371.242.912.292.67
50002.9116.423.452.427.842.842.365.194.873.81
60003.9910.613.952.706.993.533.385.275.865.23
70006.809.226.703.217.994.935.186.325.976.03
800011.478.278.455.819.638.438.356.988.868.10
900011.307.1211.855.269.9313.6713.858.089.588.35
1000010.975.8513.577.398.5117.6817.359.689.9410.69
110009.884.5312.136.856.4716.7116.218.558.5910.69
120008.543.469.487.425.1911.5710.277.297.1111.11
130007.332.566.838.993.996.096.527.165.877.32
140005.551.925.058.153.383.523.945.656.613.78
150004.231.543.747.772.731.912.334.774.553.10
160003.111.192.666.842.191.551.574.353.632.58
170002.540.901.925.671.831.101.073.562.561.52
180001.980.681.434.631.530.880.713.022.001.32
190001.530.521.073.501.160.730.532.061.541.00
200001.190.400.822.640.970.530.321.630.980.83
210001.160.310.641.900.610.380.291.270.710.64
220000.660.310.491.310.550.250.190.760.510.38
230000.560.180.380.970.360.170.150.590.290.52
240000.370.140.260.670.260.130.170.410.270.22
250000.310.150.240.430.190.080.090.250.190.13
260000.180.120.131.180.160.060.050.140.150.26
270000.180.080.130.260.110.040.030.210.120.28
280000.140.050.080.170.080.030.020.070.080.12
290000.080.050.080.170.050.020.030.090.090.13
300000.050.020.050.080.040.010.020.060.020.05
310000.040.020.030.720.040.010.030.030.030.05
320000.040.020.030.060.120.010.010.040.010.08
330000.040.020.030.030.010.010.020.010.010.06
340000.030.020.020.030.020.010.010.000.010.02
350000.020.020.010.020.020.000.010.000.000.01
360000.020.020.010.020.010.010.000.000.000.01
370000.010.010.010.010.010.000.010.000.010.01
380000.010.010.010.010.000.000.000.020.010.01
390000.010.000.010.010.010.000.010.010.000.01
400000.010.000.010.010.000.000.040.020.000.00
410000.000.000.000.000.000.000.000.000.000.00
Total %100.0100.0100.0100.0100.0100.0100.0100.0100.0100.0
Table 6-8. Truck axle distribution (Level 3 defaults).
Vehicle ClassSingle AxleTandem AxleTridem AxleQuad Axle
Class 41.620.390.000.00
Class 52.000.000.000.00
Class 61.020.990.000.00
Class 71.000.260.830.00
Class 82.380.670.000.00
Class 91.131.930.000.00
Class 101.191.090.890.00
Class 114.290.260.060.00
Class 123.521.140.060.00
Class 132.152.130.350.00
Rigid Pavement

Consistent with the 1993 AASHTO design, the baseline rigid pavement structure for the NCHRP 1-37A design methodology is a three-layer JPCP construction consisting of a Portland cement concrete (PCC) slab over a nonstabilized graded aggregate base (GAB) over subgrade (SG). However, the input parameters required for the NCHRP 1-37A methodology are considerably more extensive than for the 1993 AASHTO Design Guide. The NCHRP 1-37A design inputs are summarized in Table 6-9. Refer to Chapter 5 for detailed explanations of all input parameters and to Appendix D for a summary of the NCHRP 1-37A design methodology. All of the inputs correspond to Level 3 quality, and the default values provided within the NCHRP 1-37A software are used wherever appropriate.

The NCHRP 1-37A procedure requires the evaluation of several trial pavement sections in order to find the design that best meets the performance requirements. The baseline pavement structure from the 1993 AASHTO design procedure can be conveniently taken as the initial trial section. The predicted faulting performance for a trial section corresponding to the 1993 AASHTO Design section of 10.4 inches of PCC over 6.0 inches of GAB is shown in Figure 6-2; total faulting (after adjustment for reliability) at the end of the 25-year initial service life is 0.117 inches. Transverse fatigue cracking is negligible for this design scenario, and faulting is the controlling distress type. The design limit for predicted faulting, which is an explicit input in the NCHRP 1-37A procedure, would in general be set by individual agency policy. For the examples in this chapter, however, the design faulting limit is taken as the predicted faulting for the 1993 AASHTO design section. This is done in order to make the 1993 AASHTO and NCHRP 1-37A designs equivalent for the baseline conditions. The design limit for faulting (after adjustment for reliability and rounding) is thus 0.12 inches, which coincidentally equals the default value in the NCHRP 1-37A software. Note that initial service life for the baseline rigid pavement is 25.5 years after rounding of the faulting limit.

Figure 6-2. Predicted faulting performance for NCHRP 1-37A baseline rigid pavement design.

GRAPH: Graph showing predicted faulting performance for a trial section corresponding to the 1993 AASHTO Design section of 10.4 inches of PCC over 6.0 inches of GAB (caption of figure refers to this as NCHRP 1-37A baseline rigid pavement design). The solid line is total faulting after adjustment for reliability during the 25-year initial service life.

Table 6-9. Input parameters for NCHRP 1-37A rigid pavement baseline design.
Input ParameterDesign ValueNotes
General Information
Initial service life25 years1
Pavement construction monthSeptember2
Traffic open monthOctober2
Site/Project Identification
Functional classPrincipal Arterials - Others 
Analysis Parameters
Initial IRI63 in./mi2
Terminal IRI172 in./mi2
Transverse cracking (% slabs cracked)15%2
Mean joint faulting0.12 inSee text
Reliability90% 
Traffic
Initial two-way AADTT2000 
Number of lanes in design direction2 
Percent of trucks in design direction50%2
Percent of trucks in design lane95%2
Operational speed55 mph2
Monthly adjustment1.0 throughout2
Vehicle class distributionLevel 3 defaultsTable 6-5
Hourly distributionLevel 3 defaultsTable 6-6
Traffic growth factor0% 
Axle load distribution factorsLevel 3 defaultsTable 6-7
Mean wheel location from edge18 in2
Traffic wander standard deviation10 in2
Design lane width12 ft2
Number of axles per truckLevel 3 defaultsTable 6-8
Average axle outside width8.5 ft2
Dual tire spacing12 in2
Tire pressure120 psi2
Tandem axle spacing51.6 in2
Tridem axle spacing49.2 in2
Quad axle spacing49.22
Wheelbase spacingLevel 3 defaultsTable 6-10
Climate
Latitude38.98° 
Longitude-76.94° 
Elevation48 ft 
Depth of water table20 ft 
College Park, MD climate dataGenerated3
Design Features
Permanent curl/warp effective temperature difference-10°F2
Joint spacing15 ft2
Dowel bar diameter1 in2
Dowel bar spacing12 in2
Edge supportWidened slab 
Slab width14 ft 
Bond at PCC-base interfaceUnbonded 
Base erodibility index4 (Fairly Erodable) 
Drainage and Surface Properties
Surface shortwave absorptivity0.852
InfiltrationMinor (10%)2
Drainage path length12 ft2
Pavement cross slope2%2
PCC Surface Layer
Unit weight150 pcf2
Poisson's ratio0.22
Coefficient of thermal expansion5.5x10-6/°F2
Thermal conductivity1.25 BTU/hr-ft-°F2
Heat capacity0.28 BTU/lb-°F2
Cement type Type12
Cement content600 lb/yd32
Water/cement ratio0.422
Aggregate typeLimestone2
PCC zero-stress temperature120 °F4
Ultimate shrinkage at 40% relative humidity632 με4
Reversible shrinkage50%2
Time to develop 50% of ultimate shrinkage35 days2
28-day PCC modulus of rupture690 psi2
28-day PCC elastic modulus4.4x106 psi4
Granular Base Layer
Unbound material typeAASHTO A-1-a 
Analysis typeICM Inputs 
Poisson's ratio0.352
Coefficient of lateral pressure K00.52
Modulus40,000 psi2,5
Plasticity index1% 
% passing #200 sieve3 
% passing #4 sieve20 
D608 mm 
Compaction stateCompacted2
Maximum dry unit weight122.2 pcf2
Specific gravity of solids2.662
Saturated hydraulic conductivity263 ft/hr2
Optimum gravimetric water content11.1%2
Calculated degree of saturation82%2
SWCC parameter af11.1 psi2
SWCC parameter bf1.832
SWCC parameter cf0.512
SWCC parameter hr361 psi2
Compacted Subgrade (top 6 inches)
Unbound material typeAASHTO A-7-5 
Analysis typeICM Inputs 
Poisson's ratio0.352
Coefficient of lateral pressure K00.52
Modulus12,000 psi2,5
Plasticity index30%2
% passing #200 sieve852
% passing #4 sieve992
D600.01 mm2
Compaction stateCompacted 
Maximum dry unit weight97.1 pcf2
Specific gravity of solids2.752
Saturated hydraulic conductivity3.25x10-5 ft/hr2
Optimum gravimetric water content24.8%2
Calculated degree of saturation88.9%2
SWCC parameter af301 psi2
SWCC parameter bf0.9952
SWCC parameter cf0.7322
SWCC parameter hr1.57x104 psi2
Natural Subgrade (beneath top 6 inches)
Unbound material typeAASHTO A-7-5 
Compaction stateUncompacted 
Maximum dry unit weight87.4 pcf2
(Other properties same as for compacted subgrade)

Notes:

  1. Typical initial service life for rigid pavement design.
  2. Level 3 default/calculated/derived value from NCHRP 1-37A software.
  3. Based on interpolated climate histories at IAD, DCA, and BWI airports.
  4. Level 3 default/calculated/derived values from NCHRP 1-37A software for baseline PCC mixture properties.
  5. Default input value at optimum moisture and density conditions before adjustment for seasonal effects (adjustment performed internally within the NCHRP 1-37A software).
Table 6-10. Wheelbase spacing distribution (Level 3 defaults).
 ShortMediumLong
Average Axle Spacing (ft)121518
Percent of trucks33%33%34%
6.2.3 Summary

Baseline flexible and rigid pavement designs were developed using the 1993 AASHTO Design Guide and the input parameters in Table 6-2 and Table 6-3. The final designs are 5.3 inches of AC over 12.7 inches of GAB (5.5"/13" after rounding) and 10.4 inches of PCC over 6 inches of GAB (10.5"/6" after rounding), respectively. Initial construction costs for these designs, based on the unit cost data in Table 6-1, are $180,000 and $342,000 per line-mile, respectively ($185,000 and $345,000 per lane-mile using rounded layer thicknesses).

These baseline designs were then analyzed using the NCHRP 1-37A procedures and the input parameters in Table 6-4 and Table 6-9 to determine the corresponding distress levels at the end of initial service life. Permanent deformations are the controlling distress type for the flexible pavement scenario; the predicted total rutting (after adjustment for reliability) for the baseline flexible pavement section is 0.65 inches, as compared to the 0.75-inch default value in the NCHRP 1-37A software. Joint faulting is the controlling distress type for the rigid pavement scenario; the predicted joint faulting (after adjustment for reliability) for the baseline rigid pavement design is 0.12 inches, identical to the default value in the NCHRP 1-37A software. Note that the baseline design scenarios are not greatly different from the pavement conditions at the AASHO Road Test (except perhaps for climate), and therefore the 1993 AASHTO designs should be in general agreement with those from the more sophisticated NCHRP 1-37A methodology. Discrepancies between the design procedures should become more pronounced as conditions increasingly deviate from the AASHO Road Test conditions.

6.3 Soft Subgrade

The design scenario for this case is identical to the baseline conditions in Section 6.2, except for a much softer and weaker subgrade. The subgrade is now postulated to be a very soft high plasticity clay (AASHTO A-7-6, USCS CH) with MR = 6000 psi at optimum moisture and density before adjustment for seasonal effects. Note that this MR value is even lower than the NCHRP 1-37A default values for an A-7-6/CH material in order to accentuate the effects of low subgrade stiffness. The groundwater depth is left unchanged from the baseline scenario in order to focus on the subgrade stiffness effect. Intuitively, more substantial pavement sections are expected for this scenario as compared to the baseline conditions to achieve the same level of pavement performance. This can be achieved by increasing the thicknesses of the AC/PCC/granular base layers, increasing the quality of the AC/PCC/granular base materials, stabilizing the granular base layer, treating the soft subgrade soil, or some combination of these design modifications. In order to keep the comparisons among scenarios simple, only increases in AC or PCC thickness will be considered here.

6.3.1 1993 AASHTO Design
Flexible Pavement

The only modification to the 1993 AASHTO flexible pavement baseline inputs (Table 6-2) required to simulate the soft subgrade condition is a reduction of the seasonally-adjusted subgrade resilient modulus MR from 7,500 to 3,800 psi3. The W18 traffic capacity for baseline flexible pavement section under the soft subgrade conditions is only 1.25x106 ESALs, corresponding to an 80% decrease in initial service life.

The required pavement structure for the soft subgrade condition is determined from the 1993 AASHTO design procedure, as follows (assuming changes only in the AC layer thickness):

  • Required overall structural number SN = 5.76 (compared to 4.61 for the baseline conditions)
  • Structural number provided by granular base (same thickness D2 as in baseline design) SN2 = m2a2D2 = (1.0)(0.18)(12.7) = 2.28
  • Required asphalt structural number SN1 = SN - SN2 = 5.76 - 2.28 = 3.48
  • Required asphalt layer thickness D1 = SN1 / a1 = 7.9 inches

The design for the soft subgrade condition is thus 7.9 inches of AC over 12.7 inches of GAB (before rounding). This represents a 50% increase in AC thickness as compared to the 5.3 inches in the baseline design, which, in turn, translates to a 20% initial construction cost increase of about $37K per lane-mile (using the typical unit cost data in Table 6-1).

Rigid Pavement

The only modification to the 1993 AASHTO rigid pavement baseline inputs (Table 6-3) required to simulate the soft subgrade condition is a reduction of the seasonally-adjusted subgrade resilient modulus MR from 7,500 to 3,800 psi. The W18 traffic capacity for the baseline rigid pavement section under the soft subgrade conditions is reduced to 15.5x106 ESALs, corresponding to a 6% decrease in initial service life.

The reduction in foundation stiffness in this scenario has a direct effect on the design modulus of subgrade reaction keff, which decreases from its original value of 38 pci for baseline conditions to a value of 27 pci for the soft subgrade case. However, the required slab thickness is relatively insensitive to this reduction in foundation stiffness, increasing only 0.1 inches for a final design of 10.5 inches of PCC over 6 inches of GAB for the soft subgrade condition. Note that after rounding to the nearest half-inch, this design is identical to the rigid pavement design for the baseline conditions.

A common constructability concern under these soft in-situ soil conditions is the requirement for a stable working platform. The 6-inch granular subbase is unlikely to provide adequate stability. Consequently, a realistic final design would require either a thicker granular subbase, a separate granular working platform (not included in the structural design calculations), and/or subgrade improvement (see Chapter 7) for constructability.

6.3.2 NCHRP 1-37A Design
Flexible Pavement

Changing the subgrade soil type to A-7-6 changes many of the other Level 3 default inputs for the subgrade in the NCHRP 1-37A design methodology. The altered input parameters for the soft subgrade condition are summarized in Table 6-11. Figure 6-3 summarizes the predicted rutting vs. time for the baseline flexible pavement section (5.3" AC over 12.7" granular base); the time to the 0.65 inch total rutting design limit is only 93 months (7.75 years), corresponding to a 48% decrease in initial service life due to the soft subgrade conditions.

The trial designs (assuming only increases in AC thickness) and their corresponding predicted performance at end of the initial service life are listed in Table 6-12. Rutting is again the critical distress mode controlling the design in all cases; the design limit of 0.65 inches for total rutting is based on the performance of the baseline pavement section, as described previously in Section 6.2.2. Interpolating among the results in Table 6-12, the final flexible pavement design section for the soft subgrade conditions consists of 7.9 inches of AC over 12.7 inches of GAB. This design section is identical to that obtained from the 1993 AASHTO Design Guide for this scenario.

Table 6-11. Modified input parameters for NCHRP 1-37A flexible pavement design: soft subgrade scenario.
Input ParameterDesign ValueNotes
Compacted Subgrade (top 6 inches)
Unbound material typeAASHTO A-7-6 
Analysis typeICM Inputs 
Poisson's ratio0.351
Coefficient of lateral pressure K00.51
Modulus6,000 psi2,3
Plasticity index40%1
% passing #200 sieve901
% passing #4 sieve991
D600.01 mm1
Compaction stateCompacted 
Maximum dry unit weight91.31
Specific gravity of solids2.771
Saturated hydraulic conductivity3.25x10-5 ft/hr1
Optimum gravimetric water content28.8%1
Calculated degree of saturation89.4%1
SWCC parameter af750 psi1
SWCC parameter bf0.9111
SWCC parameter cf0.7721
SWCC parameter hr4.75x104 psi1
Natural Subgrade (beneath top 6 inches)
Unbound material typeAASHTO A-7-6 
Compaction stateUncompacted 
Maximum dry unit weight82.2 pcf2
(Other properties same as for compacted subgrade)

Notes:

  1. Level 3 default/calculated/derived value from NCHRP 1-37A software.
  2. Set artificially low to simulate a soft subgrade condition.
  3. Input value at optimum moisture and density conditions before adjustment for seasonal effects (adjustment performed internally within the NCHRP 1-37A software).

Figure 6-3. Predicted rutting performance for soft subgrade scenario.

GRAPH: Graph showing predicted rutting versus time for the NCHRP 1-37A baseline flexible pavement section when A-7-6 is used for the subgrade soil type to model a soft subgrade. The uppermost horizontal dashed line corresponds to the total rutting design limit. A comparison of the uppermost solid line (rutting for soft subgrade model) to the uppermost dashed line (rutting for the baseline subgrade) shows an initial service life of 93 months for the soft subgrade versus more than 180 months for the baseline subgrade (a 48% decrease).

Table 6-12. Trial cross sections for NCHRP 1-37A flexible pavement design: soft subgrade scenario.
AC Thickness
(in.)
Base Thickness
(in.)
Total Rutting
(in.)
5.312.70.765
6.012.70.730
8.012.70.643
10.012.70.572
Design Limit:0.65
Rigid Pavement

Changing the subgrade soil type to A-7-6 changes many of the other Level 3 default inputs for the subgrade in the NCHRP 1-37A design methodology. The altered input parameters for the rigid pavement soft subgrade condition are the same as those summarized earlier in Table 6-11 for the corresponding flexible design condition. Figure 6-4 summarizes the predicted faulting vs. time for the baseline rigid pavement section (10.4" PCC over 6.0" granular base); the time to the 0.12 inch faulting design limit is only 22.2 years, corresponding to a 13% decrease in initial service life due to the soft subgrade conditions.

The trial designs (assuming only increases in PCC slab thickness) and their corresponding predicted performance at end of design life are listed in Table 6-13. Faulting is again the critical distress mode controlling the design in all cases; the design limit of 0.12 inches for faulting is based on the performance of the baseline pavement section, as described previously in Section 6.2.2. Interpolating among the results in Table 6-13, the final rigid pavement design section for the soft subgrade conditions consists of a 10.9 inch PCC slab over 6.0 inches of GAB. This slab thickness is 0.5 inch (5%) greater than that obtained from the 1993 AASHTO Design Guide for this scenario; this corresponds to an initial construction cost increase of $14K (8%) per lane-mile.

Again, the subgrade soil is so soft and weak in this scenario that some additional design features may be required to provide a stable working platform during construction. The 6 inch granular subbase is unlikely by itself to provide an adequate working platform.

Figure 6-4. Predicted faulting performance for soft subgrade scenario.

GRAPH: Graph showing predicted faulting versus time for the NCHRP 1-37A baseline rigid pavement section when A-7-6 is used for the subgrade soil type to model a soft subgrade. The uppermost dashed line corresponds to the allowable faulting limit. The upper solid line shows an initial service life of 22.2 years or a decrease of 13% for a soft subgrade compared to the baseline conditions.

Table 6-13. Trial cross sections for NCHRP 1-37A rigid pavement design: soft subgrade scenario.
PCC Thickness
(in.)
Base Thickness
(in.)
Faulting
(in.)
10.460.131
10.760.125
11.060.118
Design Limit:0.12

Notes

  1. The 1993 AASHTO Design Guide recommends rounding the asphalt layer thickness to the nearest half inch and unbound layer thicknesses to the nearest inch. However, all layer thicknesses are rounded to the nearest 0.1 inch in this chapter to make the comparisons between the various scenarios more meaningful. Return to Text
  2. The 1993 AASHTO Design Guide recommends rounding the slab thickness to the nearest inch (nearest half inch if controlled grade slip form pavers are used). However, all slab thicknesses are rounded to the nearest 0.1 inch in this chapter to make the comparisons between the various scenarios more meaningful. Return to Text
  3. Based on the results from the NCHRP 1-37A analyses for these conditions. Return to Text
<< PreviousContentsNext >>
Updated: 06/27/2017
Federal Highway Administration | 1200 New Jersey Avenue, SE | Washington, DC 20590 | 202-366-4000