Skip to contentU.S. Department of Transportation/Federal Highway Administration
Asset Management | Bridge Technology | Operations | Pavement
FHWA > Asset Management > Pavement Health Track > PHT Final Report > PHT Version 1.1 Output Validity Test

Enhancement of the Pavement Health Track (PHT) Analysis Tool Final Report

PHT Version 1.1 Output Validity Test

Introduction

In addition to the sensitivity analysis, an experimental test was developed to observe the distress propagation under a set of incremental loading conditions, pavement surface thickness, base type and four climate zones using a total of 40 LTPP unique sites from 20 states. This experiment predicts the distress values at the time of critical failure or at the time when it reports the maximum remaining service life. A complete summary of the projects selected for the analysis is provided in Appendix A of this document.

Experimental Design

The following steps are carried out before running the PHT tool for a visual data quality analysis to identify any obvious anomalies.

  1. New or recently rehabilitated pavement sections but showing at least one distress type with very high values or approaching to critical condition.
  2. Pavement has already went through 30 to 40 years of service life but with little or no sign of distress propagation over time or no rehabilitation activities during this time period
  3. Wrong distress units.
  4. Reported year of last improvement but no reported improvement.
  5. Group pavement sections that passes the visual inspection

The analysis results are summarized in three general areas.

  1. PHT sensitivity in predicting distress propagation under a set of incremental traffic loading condition as expressed by cumulative ESALs at the time of failure.
  2. PHT sensitivity in predicting distress propagation over time; and
  3. PHT sensitivity when applied to the sample data under national default conditions.

To validate the model sensitivity to loads and aging factor for each of the distress model of PHT tool, the following test assumptions and associated model run parameter designs were established.

Design One:
The purpose of this design is to test the distress propagation due to pavement loading.
  1. For a given pavement type, base design, and thickness the truck traffic load was increased using 5% growth per year and run the model under "no critical " distress constraints. The model runtime is 60 years. The 60 years is based on the observed maximum pavement overall life in the USA (interstate) that was built in the mid-50s and still in operation.
  2. The analysis assumed NO maintenance over the analysis period (60 years).
  3. Under the assumptions 1 and 2, create a set of experimental data points (or pavement section) by keeping all design variable fix but change only the following:
    1. Pavement thickness. For rigid pavement use 8, 10 and 12 inches; for new asphalt use 4, 6, and 8 inches; for composite overlay use 2, 3, and 4 inches.
    2. For each of these, change the design section by increasing the initial truck load by 5% and prepare 29 additional sections by repeating the first section but changing the truck loading by 5% from first section to next section. This process results 30 sections with a common pavement design and climate but with incremental loading.
    3. Reset initial distress data as null or assume a new pavement.
  4. Change the distress critical value to its maximum domain value so that pavement does not fails by PHT tool due to critical trigger value rather it last for entire analysis period (60 years)
  5. Prepare charts showing the pavement distress propagation under various pavements loading condition measured in terms of ESAL.
Design Two:
The purpose of this design is to test the distress propagation over time with constant loading.
  1. Use the samples developed under Design One.
  2. Run a unique sample with fixed loading, fixed design and climate by changing the analysis period in an increment of 5 years up to a maximum of 60 years. This analysis will generate distress value at the end of each analysis period for a total of 12 observations.
  3. Use three fixed loading conditions; truck volume 2,000/day; truck volume 5,000/day; and truck volume 20,000/day.

New HMA Evaluation

The diagrams shown in Figure 7 illustrate distress propagation as reported by PHT tool at the end of the 60 years of pavement service life assuming no rehabilitation and or maintenance were performed during the analysis period. The charts also show the contribution of loading effect on distress propagation by HMA thickness, by base type (Base 2 and 3), and by four climatic zones. Each data point represent the final distress value for a given pavement section for a given load after 60 years of pavement life. To estimate the multiple data points for a given section, all parameter matrices, pavement properties, climate zone are assumed constant except the truck volume that was increased by 5% for each data point.

This figure contains twelve scatter graphs, three each for IRI, Rutting, Fatigue Cracking, and Transverse Cracking. The data are plotted for HMA thicknesses of four inches, six inches, and eight inches in four climate zone categories numbered 1 through 4 and representing an analysis period of 60 years. The trends for IRI show all test pavements will exceed the critical value at 60 years. The trends for rutting show that all pavement section types will experience significant rutting during the pavement service life of 60 years. The trends for fatigue cracking show a high sensitivity of distress propagation due to pavement thickness and type. The trends for transverse cracking show distress propagation is independent of loading and that most pavement will experience significant transverse cracks over the analysis period.
Figure 7. New HMA Distress Propagation Due to Incremental Traffic Loading
IRI

Assuming an IRI critical trigger value of 170 in/mile, all test pavements regardless of thickness, design and the climate condition will exceed the critical value at 60 years. The charts also show that for every million of ESAL, the rate of IRI deterioration for 4 " HMA pavement is 300% higher for then the 8 " pavement and 180% higher than the 6 " pavement. The average rate of IRI deterioration for per million ESAL loading is approximately 0.75 in/mile, 0.4 in/mile, and 0.25 in/mile for 4 ", 6 " and 8 " pavement respectively. Form mechanistic point of view, the model sensitivity is reasonable but in reality, the pavement construction is not perfect and neither are the causes that accumulate the IRI for a given pavement section. The PHT model demonstrates low sensitivity on IRI propagation as a function of traffic loading and therefore reporting a higher RSL forecast.

This behavior of IRI models confirms that model parameters are more tied to pavement's mechanistic properties and hardly any effect on empirical properties and reporting low IRI as a function of cumulative ESAL loading. This observation is also consistent with low statistical parameter reported in Table 4.

Fatigue Cracking

By definition fatigue cracking is a series of interconnected cracks caused by fatigue failure of the HMA surface under repeated traffic loading. In thin pavements, cracking initiates at the bottom of the HMA layer where the tensile stress is the highest then propagates to the surface as one or more longitudinal cracks. This is commonly referred to as "bottom-up " or "classical " fatigue cracking. In thick pavements, the cracks most likely initiate from the top in areas of high localized tensile stresses resulting from tire-pavement interaction and asphalt binder aging referred to as "top-down " cracking. This mechanistic behavior of forming fatigue crack may explain some degree of such a variation. However, based on the in-service-pavement in the United States, the fatigue cracks developed earlier than what PHT tool predicting under such a high load condition.

The fatigue cracking charts shown in Figure 7 show high sensitivity of distress propagation due to pavement thickness and type of base used. The PHT Analysis Tool reports that for a typical HMA pavement with 6-inch asphalt thickness, more than 140 million ESAL are needed before the pavement reaches its critical value. The rate of distress propagation from a 4-inch pavement to 8-inch pavement with aggregate base is also extremely sensitive. When under a low traffic condition, it may take more than 60 years before pavement can show any sign of fatigue cracking. For HMA with an asphalt and cement treated base, the distress is non-responsive to traffic load. In an ideal, pure mechanistic condition it can be said that due to the cement treated base, bottom up cracking is completely checked thus become non-responsive to loading; however, for thick pavement this theory does not hold and some degree of top-down cracking must appears as loading increases due to localized tensile stresses as well as binder aging. This observation confirms significant biased on mechanistic material properties and stress and strain relationship and lack of empirical adjustment to the model. Calibrating the model coefficient with empirical data can bring the model that is more aligned with the observed in-service pavement conditions.

Transverse Cracking

Distress propagation of transverse cracking is independent of loading. The charts shown in Figure 7 show that most of the pavement will experience significant transverse cracks over the analysis period. Out of the four distresses, transverse cracking is the critical distresses that will prevent the pavement from have a service life more than 60 years.

Rutting

The rutting charts shown in Figure 7 show that, regardless of base type and traffic loading, each pavement section will experience significant rutting during the 60 years of the pavement service life. The model also shows the difference in the rate of distress propagation under different climate, pavement thickness and base type. The lowest rate of distress propagation is observed for 8-inch pavement in under dry non-freeze climate condition. The PHT tools response to rutting under the different loading conditions is more aligned with the empirical evidence as observed in the site condition and performing reasonably compare to the IRI and Fatigue cracking distresses. The data results generated using the PHT tool also demonstrates that the new calibrated models carried out under MEPDG version 1.0 can significantly improve the PHT predictive capability. The charts show distress propagations with loading that are very consistence with in-service pavement.

Conclusion

The PHT analysis on for new HMA pavement shows that out of the four HMA models analyzed under this research, both the Transverse Cracking and Rutting models are more likely aligned with the in-service pavement. However both the Fatigue Cracking and IRI model shows very slow distress propagation over a long analysis period and less responsible to loading specifically for fatigue cracking. Therefore, both the IRI and Fatigue Cracking models need to be calibrated with empirical data to establish the creditability of the PHT tool's application.

New JPCP Evaluation

The chart diagrams shown in Figure 8 illustrates distress propagation of new JPCP as reported by PHT tool at the end of the 60 years of pavement service life assuming no rehabilitation and or maintenance were performed during the analysis period. The charts also show the contribution of loading effect on distress propagation by rigid pavement thickness, by base type, and by the four climatic zones.

This figure contains nine scatter graphs, three each for IRI, Faulting, and Transverse Cracking. The data are plotted for pavement thicknesses of eight inches, ten inches, and twelve inches in four climate zone categories numbered 1 through 4 and representing an analysis period of 60 years. The trends for IRI show accelerated distress propagation for climatic zone 1 and slower propagation in other zones. The trends for faulting show that except for climatic zone 4, most pavement shows faulting at or above critical distress over the analysis period. The trends for transverse cracking show that distress propagation is highly responsive to traffic loading and reaches above critical distress value at the end of the analysis period.
Figure 8. New JPCP Distress Propagation Due to Incremental Traffic Loading
IRI

The analysis shows accelerated distress propagation for climatic zone 1 and slower propagation in other climatic zones. For the other three climatic zones with a traffic loading less than a cumulative ESAL of 40 million over a analysis period of 60 years, the IRI distresses for most of the test pavement sections remain below the critical distress of 170 inches/mile.

The IRI propagation charts in Figure 8 also demonstrate a comparatively slow deterioration rate for typical 10-inch pavement for climatic zone 2 and 3, and little or no sign of distress propagation for pavement located in zone 4. The IRI remains below the critical distress for a cumulative ESAL loading equivalent to 35,000 trucks/day over a 60 year pavement life for the 12-inch rigid pavement for all climatic zones except for the wet-freeze zone 1.

Faulting

The faulting analysis assumed dowel bars at the joints. Except for the climatic zone 4, most of the pavement shows faulting at or above critical distress over the 60 years analysis period with cumulative ESAL in excess of 100 million ESAL.

Since the cause of faulting is mainly due to the difference in elevation across a joint or crack usually associated with undoweled joint construction as well as base and subbase strength, a non-stabilized aggregate base, as shown in the chart, is more sensitive to developing faulting than a cement or asphalt stabilized base.

The overall distress propagation shown in the faulting charts in Figure 8 is mostly aligned with a typical in-service 8-inch and 10-inch JPCP pavement with similar design properties in climatic zones 1 and 2. Additional calibration of this model should be able to eliminate any observed noises in the charts for pavement sections in the climatic zone 3 and 4.

Cracking

The distress propagation of percent of slab cracking as reported by the PHT tools shows it is highly responsive to traffic loading and reaches beyond the critical distress value at the end of the analysis period. The exception is those pavement sections located in the climactic zone 4 where the distress propagation is comparatively slow and does not reach at the critical point until the pavement section experiences a cumulative loading of 240 million ESAL. Overall, the cracking distress is responsive to traffic load.

Conclusion

The PHT results show a slower overall distress propagation due to traffic loading and have little effect in pavement sections located in the climatic zone 4. The slower progression of reported distresses is also consistence with the sensitivity analysis and the IRI is shown to be less sensitive to loading. The result outcome is very consistent with low statistical parameter reported in Table 4.

AC/AC Evaluation

The chart diagrams shown in Figure 8 illustrate distress propagation as reported by PHT under a cumulative ESAL loading of 60 years of pavement service life. The charts also show the contribution of loading effect on distress propagation by AC/AC pavement thickness and by the four climatic zones.

Conclusion

The charts show the similar distress propagation pattern as of HMA pavement

This figure contains twelve scatter graphs, three each for IRI, Rutting, Fatigue Cracking, and Transverse Cracking. The data are plotted for pavement thicknesses of two inches, three inches, and four inches in four climate zone categories numbered 1 through 4 and representing an analysis period of 60 years. The trends for IRI show all test pavements will exceed the critical value at 60 years. The trends for rutting show that all pavement section types will experience significant rutting during the pavement service life of 60 years. The trends for fatigue cracking show a high sensitivity of distress propagation due to pavement thickness and type. The trends for transverse cracking show distress propagation is independent of loading and that most pavement will experience significant transverse cracks over the analysis period. These results show distress propagation patterns that are similar to that observed for HMA pavement.
Figure 9. AC/AC Distress Propagation Due to Incremental Traffic Loading
Updated: 02/19/2014