Verification, Refinement, and Applicability of Long-Term Pavement Performance Vehicle Classification Rules

Chapter 5. Sensitivity of Pavement Design Models to Differences in Selected Vehicle Classification Rule sets

Discussion of Findings From AASHTO 93

The same sensitivity analysis of selected pavement design structures to changes in vehicle classification and truck volume owing to use of different vehicle classification rule sets was repeated using the AASHTO 93 pavement design models. The traffic data used in the MEPDG designs were converted into equivalent single-axle loads (ESAL) following the guidelines provided in the AASHTO 93 Interim Guide. For the base scenario, the pavements were designed using traffic inputs obtained by the LTPP vehicle classification rule set. The same pavement sections were then analyzed using the ESALs adjusted based on truck volume and VCD computed using alternative vehicle classification rule sets. All traffic inputs were summarized in table 15 and table 16. The same material types used in the MEPDG analysis for subgrade, base, and surface layers were used in the AASHTO 93 analysis. Pavement design input parameters and the results of sensitivity analyses are summarized in table 20. The AASHTO 93 design guide uses serviceability as a performance indicator and design criterion. The predicted serviceability model is a logarithmic function of structural capacity, traffic, and other variables.

Table 20. Summary of pavement life predictions from AASHTO 93 sensitivity analyses.

AC = Asphalt Concrete

PCC = Portland Cement Concrete

LTPP = Long-Term Pavement Performance

The ROPA designs for both pavement types were found to be the most sensitive to variations in traffic classification rules, which is similar to what was found in the MEPDG analyses. However, the sensitivity of the AASHTO 93 models to differences in traffic characterization was less significant than the sensitivity of the MEPDG models. Moreover, the variations in vehicle classification rule sets did not result in differences in design surface layer thickness of more than 0.5 inches for either pavement type or road functional class. Figure 23 and figure 24 show an example of performance sensitivity to variations in the traffic scheme of AASHTO 93 designs for rigid pavements. Figure 25 and figure 26 show the same sensitivity for flexible pavements. In all four figures, the impact of increased volume of heavy trucks, represented by the Washington WIM rule set, is more evident and more substantial for ROPAs than it is for RIs.

PSI = Present Serviceability Index

Figure 23. Graph. AASHTO 93 performance predictions for wet-no freeze condition for rigid pavements: ROPAs.

PSI = Present Serviceability Index

Figure 24. Graph. AASHTO 93 performance predictions for wet-no freeze condition for rigid pavements: RIs.

PSI = Present Serviceability Index

Figure 25. Graph. AASHTO 93 performance predictions for wet-no freeze condition for flexible pavements: ROPAs.

PSI = Present Serviceability Index

Figure 26. Graph. AASHTO 93 performance predictions for wet-no freeze condition for flexible pavements: RIs.

From the sensitivity analysis results, it is possible to conclude that, for high-volume roads such as RIs, the structural capacity of the pavement—given by the structural number for flexible pavements and PCC slab thickness for rigid pavements—has a dominant effect over small variations in traffic, such as the variation resulting from the three vehicle classification rule sets tested. In this study, percentile variations in truck volumes due to the application of different classification rules were much more significant for low-volume roads compared with high-volume roads, contributing to higher sensitivity of pavement design model outcomes for low truck volume designs.

Conclusions

High Truck Volume Roads with High Presence of Class 9 Trucks

The analysis results indicate that, for pavement designs typical for RIs that are subjected to a high volume of heavy truck traffic (and which typically consist of 75 percent or more of Class 9 vehicles), differences in vehicle classification rule sets, compared with the LTPP classification rule set, are likely to have very little practical impact. For these cases, a combination of vehicle classification data collected using a non-LTPP vehicle classification rule set with load spectra obtained from data collected using the LTPP classification rule set should not result in significant errors in either MEPDG or AASHTO 93 design and analysis outcomes, provided that the load spectra shapes (relative percentages of loaded and unloaded axles) accurately describe the expected traffic loading at the site.

Low Truck Volume Roads With Low Presence of Class 9 Trucks

The same conclusion as above (no significant errors in pavement design and analysis outcomes) applies for ROPA designs with moderate-to-low truck traffic (200 to 300 AADTT per lane) and low presence of Class 9 vehicles (20 percent or less) for classification rule sets that result in a decrease of overall loads (such as the California WIM algorithm) compared with vehicle classification data collected using the LTPP classification rules. Such classification systems are likely to implement additional weight-based rules to classify lightweight multi-axle vehicles in lower classes.

For these roads, vehicle classification rules that result in an increase in traffic loads when combined with load spectra collected using the LTPP classification rule set can lead to moderate differences in MEPDG predictions if vehicle classification data from a non-LTPP vehicle classification rule set are combined with load spectra obtained from data collected using the LTPP classification rule set. Vehicle classification rule sets that lead to an increase in total truck loads for pavement design when combined with load spectra collected using the LTPP classification rule set are those that classify lightweight multi-axle vehicles in heavy truck classes (Class 8 or above) and algorithms that classify lightweight two-axle pickup trucks as Class 5 instead of Class 3. However, to see the significant difference, the increase in traffic load has to be quite significant. For example, in the analysis conducted in this investigation, the total volume of trucks (in Classes 4 through 13) increased by 35 percent with an increase in heavy trucks (in Classes 6 through 13) of 59.4 percent.

The differences in MEPDG pavement design life predictions were up to 3 years for flexible pavements and up to 8 years for rigid pavements. The difference in sensitivity between the two pavement types was primarily because different distress modes of failure were observed and these distresses evolved differently over time. In the case of the AASHTO 93 designs, given the nature of the serviceability index used as an indicator of performance, flexible pavements were more sensitive to increases in traffic loads than rigid pavements, where the differences in service life observed were up to 3 and 1.9 years, respectively.

Of all the test cases, only the MEPDG rigid design for low truck volume roads resulted in differences exceeding 0.5 inches in design slab thickness. For this type of pavement design, if the LTPP SPS TPF data are to be used as surrogate axle load spectra or default for the site, a detailed site-specific analysis of truck types observed at the site is recommended to determine whether a significant number of trucks in different classes would be shifted and if a significant percentage of additional vehicles would be added to the total truck volume estimate owing to differences in the vehicle classification rules. This analysis would require field observations of trucks or local knowledge of truck types typical for the site and an evaluation of how these truck types would be classified using the LTPP and State classification rule sets. For example, if no pickup trucks pulling light trailers are observed at the site, no additional error in Class 8 vehicle counts is expected, even if the two classification rule sets theoretically would classify this vehicle type differently.

Disclaimer

The analyses presented in this section cover a limited number of traffic scenarios (volume, class, load spectra). Based on the traffic data obtained from SPS TPF sites, it is expected that the analyses presented in this chapter are likely to represent the worst case scenarios in terms of predicted pavement design differences. However, the percentage of misclassified vehicles depends greatly on a specific vehicle stream observed at a site, creating an indefinite number of misclassification scenarios. It is not practical to test the MEPDG sensitivity to all possible combinations of site-specific vehicle streams and vehicle classification rule sets.