Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
18 This chapter describes the framework used to evaluate and test the recommended measures and presents examples of testing and validating these measures using agency data. In the context of this research, testing describes the process of assessing the sensitivity of the measure to changes in performance resulting from the application of preservation as compared to changes in the performance measure when no preservation is applied. For example, if a thin overlay affects the IRI, testing will compare the change in IRI due to the application of a thin overlay to the change in IRI during subsequent years if an overlay is not applied (i.e., a control case). On the other hand, if a chip seal does not immediately affect the IRI, comparing changes in IRI following the application of a chip seal to that of a control case would only reflect measurement error plus the small growth in IRI. Alternatively, validation refers to the process of assessing the measuresâ ability to capture changes in performance, service life, and LCC. Framework Development As stated earlier, pavement performance measures are metrics that measure specific aspects of the pavement in order to quantify the degree of achieving specific goals pertaining to perfor- mance. Implicit in this definition is the evaluation of the current condition and long-term trends in pavement condition in order to assess the decisions made to achieve specific goals for the asset (Cambridge Systematics 2006; Zietsman et al. 2011; Amekudzi and Meyer 2011; Simpson et al. 2013). The performance measure should be capable of quantifying the effects of treatments applied to the pavement surface. This section documents the development and evaluation of a framework to characterize the effects of applying preservation treatments. The framework is based on the methodology devel- oped for the Indiana DOT (Ong et al. 2010), which considers both the initial condition jump, which results from the application of the treatment, and how the applied treatment affects the long-term rate of deterioration. This approach is illustrated in Figure 6. The first step is to determine the immediate change in the pavement condition following the application of a pres- ervation treatment (i.e., initial condition jump). This jump (term CJ in Figure 6) is a function of treatment type and condition of the pavement prior to treatment application. The second step is to determine the performance of the treated pavement over time and compare its condi- tion to that of a control section (C â²t-Ct in Figure 6) at several points in time. These two elements provide the information needed to establish the effects of preservation treatments on pavement performance that can be used to assess service life and LCC. The analysis to evaluate the proposed framework used data from the LTPP program database or from studies that used LTPP data. For the asphalt pavements, the data used were extracted from the Specific Pavement Studies (SPS)-3 experiment, which compared various preservation C H A P T E R 3 Testing and Validation of Recommended Measures
Testing and Validation of Recommended Measures 19 treatment options to a control site (i.e., non-treatment deterioration). For the concrete pavements, data from the General Pavement Studies (GPS)-3 and GPS-4 experiments were the main source for determining the effects of diamond grinding and dowel bar retrofit (DBR), but no suitable control sections were identified to compare the long-term effects. Figure 7 illustrates the approach used for determining the initial condition jump after applica- tion of treatment. In this approach, the change in measured distress (e.g., ride, cracking, rutting, and faulting) after application of the treatment for each pavement segment is determined. It was found that the measured difference was a function of the initial condition (as shown in Figure 7) in some cases, but independent of the initial condition in others. Figure 8 illustrates the concept used for comparing the changes in performance over time due to the application of a preservation treatment. Changes in distress over time for both control and treatment sections are shown for three separate pavement segments. The differ- ences in distress between the treated and control sections were calculated at each measurement following the treatment application and over time. The top set of charts illustrates a situation of constant difference over time (the treated section deteriorates at the same rate as the control section). The middle set of charts illustrates decreasing difference over time (the treated section deteriorates at a faster rate than the control section) and an increasing difference over time is shown in the bottom set of charts (the treated section deteriorates at a slower rate than the control section). 0 20 40 60 80 100 0 3 6 9 12 15 18 21 Pa ve m en t C on di ti on Time Original Deterioration Deterioration Post Preservation tt C't - Ct Treatment Performance CJ Figure 6. Illustration of pavement preservation effect on pavement condition. Time D is tr es s Initial Distress Value C ha ng e in D is tr es s Initial Distress Change in Distress Evaluate for Many Cases Regression Relationship Distress Measurement in Single Year Figure 7. Illustration of the effect of treatment.
20 Quantifying the Effects of Preservation Treatments on Pavement Performance Framework Evaluation The proposed framework was evaluated using information from the literature review (much of which used LTPP data) and data from the LTPP program. Literature Data Data available in the literature was evaluated to assess whether these data supported the recommended framework (i.e., whether the required elements have been identified in previous literature). Each treatment was evaluated based on two criteria: (1) effect of treatment on initial condition jump and (2) long-term performance effects from treatment. The latter was evaluated by comparing the performance curve of the treated pavement to that of a control (i.e., untreated) pavement. For example, Mamlouk and Dosa (2014) demonstrated that deterioration curves for roughness for pavements treated with chip seal and those for the control pavement are not expected to intersect. However, Hajj et al. (2011) showed that deterioration of pavements after slurry seal application, as measured by the PCI, is such that the deterioration curve of the treated pavement intersects that of the control section at a future point in time because the slurry seal sections deteriorated at a faster rate than the untreated sections. Other research (Mamlouk and Dosa 2014; Carvalho et al. 2011; Hajj et al. 2011) has used similar approaches as those recom- mended in this research. Analysis using LTPP Data Data from the LTPP were gathered for asphalt and concrete pavements to evaluate the recom- mended framework by determining whether the condition jump and changes in performance Time D is tr es s TimeC h an g e in D is tr es s Control Section Treatment Section TimeP av em en t C o n d it io n Time D is tr es s Time TimeP av em en t C o n d it io n Time D is tr es s Time TimeP av em en t C o n d it io n C h an g e in D is tr es s C h an g e in D is tr es s Figure 8. Illustration of the effect of treatment on distress and condition.
Testing and Validation of Recommended Measures 21 required by the framework could be calculated. Analyses were performed for asphalt pavements using roughness, rutting, fatigue cracking, longitudinal cracking, and transverse cracking data from the LTPP SPS-3 experiments to evaluate thin overlay, chip seal, and slurry seal treatments. Each of the treatments resulted in an immediate change in cracking of all types, generally reducing cracking to zero immediately following treatment application. The performance of cracking varied depending on the treatment and the type of cracking considered. The rate of cracking growth after slurry seals application remained the same as before treatment, but was reduced for some types of cracking after application of seals and thin overlays. Thin overlays caused an immediate change in rutting, but chip seals and slurry seals did not immediately affect rutting. In addition, no long-term changes in rutting were found for any of the treatments using the LTPP data. It was found that thin overlays were the only treatment that had an immediate effect on the IRI, but both chip seals and thin overlays affected the long-term IRI performance. Figure 9 shows the IRI measurements before and after the application of the chip seal. It can be seen that the measurements generally fall along the line of equality (i.e., no immediate change in condition). Mamlouk and Dosa (2014) noted that the long-term perfor- mance of chip seal sections was better than that of control sections, and the improvement in performance was dependent on the initial roughness of the pavement. Concrete pavement data were also extracted from the LTPP database to assess the immediate effects of preservation on pavement performance (differences in long-term performance between treated and control segments could not be evaluated because of the lack of control site data). The treatments evaluated were diamond grinding and diamond grinding with DBR (but not DBR independent of diamond grinding). Both treatments were found to reduce roughness and faulting, but neither treatment was shown to affect cracking. Figure 10 shows the change in wheel path faulting following diamond grinding. The literature provided limited information on the effect of preservation treatments on the long-term performance of concrete pavements. In summary, the literature review and the analyses conducted using LTPP data helped quantify the effect of certain preservation treatments on pavement performance. For asphalt pavements, the treatments evaluated were thin AC overlay, chip seal, and slurry seal. For concrete pave- ments, the treatments evaluated were diamond grinding and diamond grinding with DBR. The effects of specific treatments on performance measures varied depending on geography, climate, and other factors. These effects provided insights into the sensitivity of certain perfor- mance measures to the application of preservation treatments; a matrix of treatments and effects on performance measures, as determined based on LTPP and highway agency data, is presented Figure 9. Chip seal effect on roughness.
22 Quantifying the Effects of Preservation Treatments on Pavement Performance later in this chapter. More importantly, the findings from the literature and the analysis of LTPP data validated the approach to assessing the performance measures using the recommended evaluation framework. Gathering Highway Agency Data Having established the appropriateness of the proposed framework to assess the effect of pres- ervation treatments on pavement performance, the next step was to gather data from highway agencies to use in testing and validating the recommended performance measures. Data were obtained from eight state DOTs (Georgia, Idaho, Maryland, Ohio, Texas, Utah, Virginia, and Washington). These states were selected for the following reasons: ⢠Adequate mileage of asphalt, concrete, and composite pavements. ⢠Use of most common preservation treatments. ⢠Familiarity with recommended performance measures, especially the individual pavement condition measures. ⢠Availability of data, information, and processes to support the testing and validation. ⢠Adequate regional distribution and coverage of climatic conditions. The data gathered from Georgia DOT included many years of condition data, but no cor- responding construction history to link with preservation activities, and therefore it could not be used in the testing and validating efforts. The research team then conducted a webinar for the representatives of the highway agencies to (1) present relevant project background information and (2) review the testing and validation effort, with special emphasis on the projectâs data needs. Subsequent to the webinar, the research team held follow-up conference calls with each of the eight agencies to (1) review the data requirements and (2) establish how those data would be provided to the research team. The following are some of the topics covered: ⢠Type of pavements available for evaluation. ⢠Type of treatments available for evaluation. ⢠Availability of information on when and where preservation treatments were placed (the research team can get construction history/information from another source). ⢠The performance measures and/or pavement condition data used by the agency. ⢠Availability of pavement condition data and performance models for asphalt and concrete treatments. Figure 10. Change in wheel path faulting resulting from diamond grinding.
Testing and Validation of Recommended Measures 23 ⢠Performance models ability to predict future values for individual distresses or for composite measures. ⢠Likely availability of data to allow identifying initial condition jump and long-term perfor- mance history information. ⢠Availability of data from control sites that have not received preservation and could be used as a reference for comparing long-term performance of preservation treated pavements to non-treated ones. ⢠Parts of the network recommended for use in this research. ⢠Agency willingness to provide relevant information for each segment of network including pavement condition as a function of time, location (start and end point), pavement structure, traffic, functional class, type of treatment, date of application, condition data, surface prepa- ration prior to the treatment, treatment details, such as type of chip seal, and condition of pavement prior to treatment application. Seven highway agencies involved in the webinar and follow-up calls indicated availability of sufficient data related to two or more preservation treatments. Table 1 lists the data provided by each agency and the approximate number of years of data provided. Ohio and Washington provided their entire database of construction and pavement condition; Idaho gave the research team access (read only) to their pavement management system. Although each highway agency provided pavement condition data, some agencies did not provide data for specific distresses. For example, Idaho and Ohio perform windshield surveys for cracking, and provided condition index values that reflect cracking, but no measured crack- ing data. Also, Ohio provided data on roughness and a composite performance measure but no rutting data. In addition, issues were encountered with the supplied data. For example, in one case the naming convention of the pavement segments in the database changed from one year to the other, and there was no method for identifying the same segments after the change. Therefore, the data used for evaluating the performance measures was limited to only those data available after the changes in the naming convention. Another issue encountered was the variations of how the begin and end points of pavement segments were defined and reported throughout the years. For example, in many cases it was found that a treatment was applied to parts of two consecutive pavement segments that were defined in the following year as new segments. Therefore, condition over time could not be tracked directly using this data. Potential issues that could be encountered during implementation of the proposed perfor- mance measures are addressed in the guide. Examples of these issues include data accuracy and Treatments MD-SHA VDOT WSDOT IDDOT UTDOT TXDOT OHDOT Number of Years of Condition Data 4 7 30+ (Entire Database) 15 2 10 30+ (Entire Database) Thin AC OL â â â â â Chip Seal â â Microsurfacing â â â â Slurry Seals â â â â â â AC Patching â â â â â â AC Crack Sealing â â â â â Diamond Grind â â â â â â DBR â â â â â â indicates data was provided â indicates that no data was provided Table 1. Data provided by highway agencies.
24 Quantifying the Effects of Preservation Treatments on Pavement Performance reliability, measurement errors and biases, and completeness of the data (e.g., missing measure- ments and data elements). Testing and Validation of Performance Measures This section describes the testing and validation of the recommended pavement performance measures. The testing and validation made use of data collected from highway agencies. The first and second parts of this section address the testing and validation of the performance measuresâboth individual and composite pavement conditionâfor asphalt treatments and concrete pavement preservation treatments, respectively. The last part of the section addresses the testing and validation using LCC as a performance measure. Each highway agency that provided data has different practices and techniques associated with condition assessment. For example, rutting measurement methods varied from the multi-point systems method in some states to scanning lasers in others. These differences in data collection prac- tices are expected to influence the perceived effect of pavement preservation, which is expressed by performance measures. Therefore, the trends of these measures were evaluated to determine whether appropriate conclusions could be made when analyzing data from multiple states. Analysis of agency data has shown relatively large year-to-year variance (known as time series variance) in the performance measures. This variance can result from equipment measurement errors, differences in location referencing from year to year, variations in data interpretation (e.g., a crack may appear more or less severe in a subsequent year because of climatic conditions), or other sources. Therefore, traditional and advanced statistical techniques that account for such variance (Carroll et al. 2006) were used to analyze the data. In general, using datasets with a large number of measurements allows rational conclusions to be reached in spite of the time series variance. However, because of this variance, measurements viewed as outliers can influence the trend if relatively few measurements for a given distress and treatment are available. Multiple cases involving data seeming as outliers were investigated. These data were not removed from the analysis in order to not artificially assign higher confidence to the measurements. Absent control sections to evaluate treatment effect over time, the relationship between the initial value of a distress and the distress growth rate over time was investigated. Prior research has shown a relationship between prior condition and treatment performance (e.g., Mamlouk and Dosa 2014; Carvalho et al. 2011). Demonstrating such relationship using highway agency data would confirm the performance measureâs ability to capture the preservation treatment performance. Using data from untreated pavement segments with similar initial conditions as control segments was considered. However, untreated segments are generally better performing pavements that do not require treatments, and thus the comparison would bias the findings. Therefore, control sections should be selected from pavements that were candidates for preser- vation, but did not receive a treatment. Otherwise a typical pavement deterioration model may be assumed to represent the control behavior. Asphalt Pavements Roughness The use of roughness as a measure of the effectiveness of thin asphalt overlays, chip seals, and microsurfacings was evaluated using data from six state highway agencies. Only thin asphalt overlays had an immediate effect on roughness values (chip seals and microsurfacing had no effect). The change in roughness due to the application of a thin asphalt
Testing and Validation of Recommended Measures 25 overlay using data from Virginia DOT (VDOT) and Ohio DOT (ODOT) is shown in Figure 11. The equation shown in the figure presents the change in IRI as defined by the IRI after treatment minus the IRI before treatment. When comparing the VDOT and ODOT data it was found that ODOT data exhibited rela- tively larger variance in the year-to-year measurements. To consider the variance, the change in IRI from ODOT was calculated as the average of multiple measured IRI values before the thin overlay minus the average of multiple IRI values following the overlay (with no more than 3 years of data in each average). However, because the VDOT data showed less year-to-year variance, the data were evaluated using the values directly before treatment and directly after treatment. The variance combined with the relative few data points significantly influenced the model before the averaging was considered. As noted in Mamlouk and Dosa (2014), chip seals can slow the progression of roughness based on the initial condition of the pavement. However, the network-level roughness measure- ments obtained from some highway agencies could not be used to confirm this trend due to the large variability in the data for the same site and availability of data for a limited number of years. Two methods were used to evaluate the growth of roughness following a chip seal application in Virginia. In the first method, the rate of roughness growth as defined by the slope of best fit line through the roughness measurements versus time following the treatment is used as a measure of the change in roughness. Figure 12 shows the average rate of roughness growth over Figure 11. Effect of thin asphalt overlay on roughness. Figure 12. Roughness growth versus the initial roughness for chip seals in Virginia.
26 Quantifying the Effects of Preservation Treatments on Pavement Performance 5 years following the treatment for the different sites as a function of the initial roughness. The figure indicates that the average roughness growth over 5 years following the treatment is effectively zero, regardless of the initial roughness. However, LTPP data showed a reduction in the progression of roughness that is dependent on the initial roughness of the pavement (Mamlouk and Dosa 2014). This difference in these findings appears to be caused by the large variability in data (as shown in Figure 12) and the evaluation of data for only a few years following chip seal application. The second method is the approach by Mamlouk and Dosa (2014) to investigate whether the rate of roughness growth is affected by the initial roughness value. In order to account for the variability in the initial value, the intercept of the best fit line to the measurements versus time (with the pre-treatment measurement at year 0) was taken as the initial roughness. However, no relationship was found between the intercept of the best fit line and the slope of the best fit line, indicating that the rate of roughness growth is not affected by the initial roughness value. The average rate of roughness growth following the application of thin asphalt overlays in Virginia was investigated. The slope of the best fit line for the rate of roughness growth versus initial roughness showed that the rate of roughness growth over the years following the thin asphalt overlay was effectively zero (i.e., change in roughness performance following a thin asphalt overlay is unrelated to the roughness just prior to the thin asphalt overlay). To better understand the effect of the measurement variance in the results, roughness values following chip seal applications in Texas were evaluated and the results showed an evident year to year variance in the measurements. It was found that the differences between two roughness measuresâone immediately following the chip seal application and one between five and nine years following the chip seal applicationâare zero on average but with large variance. This sug- gests the pavement roughness did not change on average over the time frame; there is a prob- ability of an increase or decrease in roughness by 50 in./mile. Cracking The three treatments showed a significant immediate reduction in cracking values. Figures 13 and 14 show the change in transverse and longitudinal cracking, respectively, one year and two years following the application of a chip seal in Virginia. Cracking following treatment effectively is zero. Cracking performance after the application of preservation treatments was investigated using data from Virginia and Texas. The slope of the best fit line through the cracking measurements Figure 13. Change in transverse cracking for chip seals in Virginia.
Testing and Validation of Recommended Measures 27 was calculated for 3 to 9 years following the treatment, to assess whether the rate of crack growth is a function of the measured pretreatment cracking. Figure 15 shows the average rate of growth of fatigue cracking versus the pretreatment fatigue cracking for chip seals in Texas. It shows that the growth rate of fatigue cracking is not related to the pretreatment fatigue cracking. The cracking growth rate is the slope of the line fit through consecutive cracking measures following the treatment. Although the mean value of the growth rate is effectively zero, there is a positive growth rate when the pretreatment fatigue cracking is less than 10 percent. The performance of longitudinal and transverse cracking measures was also evaluated for chip seals in Texas and thin asphalt overlays and chip seals in Virginia. The analysis results showed that the rate of growth of transverse cracks for chip seals in Texas and thin asphalt overlay in Virginia were related to the pretreatment transverse cracking (Figure 16 and Figure 17), but there was no statistically significant relationship for transverse cracking following chip seal application in Virginia. The evaluation of cracking performance showed no significant change over time for the crack- ing most associated with loading (fatigue cracking). However, cracking associated with aging and environmental considerations (transverse) showed growth as a function of the initial condition. Figure 14. Change in longitudinal cracking for chip seal in Virginia. Figure 15. Fatigue crack growth following chip seals in Texas.
28 Quantifying the Effects of Preservation Treatments on Pavement Performance Rutting Rutting data for asphalt surfaced pavements was obtained from five highway agencies, and the trend of rutting following the treatments was evaluated. Only thin asphalt overlays data showed reduced rutting overall. The change in rutting due to thin asphalt overlays in Virginia is shown in Figure 18. The average reduction is parallel to the line of maximum improvement with an offset of approximately 0.1 inches, which is the value after a thin asphalt overlay is applied. Only thin asphalt overlays applied in Virginia showed significant relationship between the initial rutting and long-term rutting performance with rut growth occurring at higher rates when the overlay was placed on pavements with higher values of initial rutting. Composite Index Each of the agencies that submitted distress data also included composite indexes. This section presents the evaluation of these indexes for Virginia and Ohio. The composite index from the VDOT was selected for analysis because of the availability of many years of data, and Ohioâs pavement condition rating was selected because it is similarly structured to other composite indexes used by other states. The composite index developed by VDOT is the Critical Condition Index (CCI); details regarding the calculation of the CCI are readily available in other resources (McGhee 2002). Figure 16. Transverse crack growth rate following chip seals in Texas. Figure 17. Transverse crack growth rate following thin asphalt overlays in Virginia.
Testing and Validation of Recommended Measures 29 The CCI is presented on a 100 point scale with 100 being the best possible score and 0 being the worst possible score. To calculate the CCI, two different indexes are calculated from the data collected during the distress survey, the load-related distress rating (LDR) and the non-load- related distress rating (NDR), and the lower value of the two is defined as the CCI. The LDR is calculated by estimating deduct values for each load-related distress that are deducted from 100. The distresses used in the LDR are alligator cracking, patching, potholes, delamination, and rutting. The NDR considers deduct values for non-load-related distresses: block cracking, patching, and longitudinal cracking out of wheel path, transverse cracking, reflection cracking, and bleeding. The immediate improvement in the CCI was evaluated for both chip seals and thin asphalt overlays. For chip seals (Figure 19), the CCI was improved on average 28 points. For thin asphalt overlays (Figure 20) the CCI values were set back to 100 upon treatment appli- cation. The maximum possible improvement lines in these figures represent the CCI change required for the CCI before treatment to reach 100. Figure 19 shows that for CCI values before treatment above 80, the application of the chip seal lowered the CCI value. This could be attributed to errors in the data for the individual distresses that are reflected in the calculation of the CCI. Additionally, the distresses are mapped to deduct values, which are then subtracted from the LDR or NDR. Because these deduct values are logarithmic, the difference in deduct values between 0 and 10 percent Figure 18. Change in rutting from thin asphalt overlays in Virginia. Figure 19. Immediate change in CCI due to chip seal application.
30 Quantifying the Effects of Preservation Treatments on Pavement Performance distress is much larger than the difference between 10 and 20 percent distress when reflected in the composite index. Thus, if the errors in individual distresses are randomly distributed, errors in the distresses that result in a perceived improvement in the distresses will affect the LDR and NDR calculations differently than errors that result in a perceived degradation in the distresses. An increase in the CCI from 80 to 90, for example, requires a larger change in the measured distresses than a decrease in CCI from 80 to 70, which can explain the reduction for high CCI values. Seven years of performance data for the 2,940.1 mile long pavement segments that were overlaid with a thin asphalt overlay were evaluated. It was found that there was little change in the CCI over the seven year time period. Therefore, the CCI could not be used as a measure for evaluating the effect of thin asphalt overlays on performance. Several other states provided data relative to their composite indexes, and it was found that the change in that index immediately following the application of a preservation treatment closely matched the change in cracking conditions. For example, the ODOT uses a PCR that ranges from 0 (very poor) to 100 (very good) to describe the condition of the pavement (Simpson et al. 2013). The methodology translates the severity and extent of pavement distresses into deduct values that are subtracted from 100 to obtain the PCR. The qualitative values of severity and extent of cracking are obtained from the windshield survey. Crack sealing had almost no effect on the PCR, but microsurfacing, chip seals, and thin asphalt overlays reset the PCR to 90, 95, and 98, respectively. Rigid Pavements Only WSDOT provided sufficient information regarding preservation of rigid pavements, and the effect of diamond grinding with and without DBR on several measures was evaluated using this data. Roughness WSDOT data showed that diamond grinding with and without DBR resulted in an immedi- ate decrease in roughness, with a greater decrease occurring when DBR was performed with the diamond grinding. The results of the change in roughness due to diamond grinding with and without DBR are shown in Figure 21. Analysis of pavement roughness over time showed that, in both cases (diamond grinding with and without DBR), the growth of roughness over time was constant and independent of the roughness before preservation. Figure 20. Immediate change in CCI due to thin asphalt overlay application.
Testing and Validation of Recommended Measures 31 Cracking It was found that diamond grinding with and without DBR had no immediate effect on the measured cracking, and the rate of growth of cracking following the treatment was not related to the initial cracking values. Many of the cracking measurements indicated a negative rate of crack growth following the diamond grind which may be due to variance in the measurements or unrecorded maintenance (e.g., crack sealing) that resulted in a decrease in cracking. Faulting Diamond grinding with and without DBR significantly reduced faulting in both cases; to more or less as shown in Figure 22. However, no relationship was found between the pre-treatment faulting, and the fault growth rate following preservation. Composite Index WSDOT has also developed a set of composite indexes for rigid pavements to indicate when grinding, DBR, or reconstruction is needed. The reconstruction index is based on cracking and faulting, the DBR index is based on faulting with a maximum limit placed on cracking, and the grinding index is based on faulting, rutting, and roughness. Diamond grinding without DBR improved each of the three indexes, with the largest improvement occurring in the grinding index [Figure 23 (left)]. The DBR index is only calculated when DBR has not been previously performed on the pavement. Therefore, for the case of diamond grind with DBR, only the grinding and reconstruction indexes were evaluated. Both the reconstruction index and the grinding index showed improvement following diamond grinding with DBR, with the more improvement shown in the grinding index [Figure 23 (right)]. The relationship between the rate of change over time and the composite indexes and their values before preservation was evaluated. No relation between the rate of change of the pre-preservation index value or other variables (e.g., traffic) was found. Figure 21. Effect of diamond grinding on roughness. Figure 22. Effect of diamond grinding on faulting.
32 Quantifying the Effects of Preservation Treatments on Pavement Performance Life Cycle Costs LCC are a critical part of the framework for selecting the most appropriate preservation treat- ment. Once the effects of preservation treatments on the condition and performance of a pavement are known, these effects are combined with direct costs in order to estimate LCC for two scenarios: (1) the preservation is applied in the specified year, and (2) the preservation is not applied. The scenario with the LLCC or the one with higher benefits normalized by costs is the preferred choice. To evaluate the effect of preservation on LCC, an analysis was performed using the relation- ships shown in Figures 11 and 18 and cost data provided by the Maryland SHA. The EUAC were compared for the reconstructed pavement segments; one segment will serve as a control pave- ment (i.e., receives no overlay) and the other segment will receive a thin asphalt overlay 10 years after reconstruction. The analysis period for each segment will be the time required for the segments to reach its service life (i.e., the two segments will have different analysis periods). The length of the analysis period was taken as the time until the remaining service life of the pavement reached zero as calculated using the Maryland SHAâs criteria (Rada et al. 2016). In this method, the time until each distress reaches a defined threshold is calculated, and the remaining service life of the pavement is the lowest of these times. For illustration, a three per- cent discount rate was applied to future costs and the following assumptions were made: (a) the jump in roughness due to a thin asphalt overlay is that shown in Figure 11, (b) cracking indexes were reset to 100, and (c) rutting was reset to 0.1 inch (per the relationship shown in Figure 18). The Maryland SHA deterioration curves for IRI, friction number, functional cracking index, and structural cracking index were assumed in this example, and the data for treatment costs provided by the Maryland SHA was used. The results of the analysis showed that the functional cracking index was the criteria control- ling the remaining service life calculations (i.e., the thresholds for cracking were reached before those for roughness or friction). The threshold values (zero life condition) defined by the Mary- land SHA are 250 in./mile for IRI, 45 for the SCI, 30 for the FCI, and 35 for FN. In the case of only reconstruction in the first year, the remaining service life was calculated as 25 years, and the EUAC was calculated to be $46,336. When the thin asphalt overlay was placed in year 10, the remaining service life was extended by 10 years, and the EUAC was calculated as $41,788. Thus, the EUAC showed the effect of preservation on cost, making it appropriate for use as a measure. Summary of Findings This chapter summarized the results of the analyses performed using both LTPP and high- way agency data, and other results found in the literature. Table 2 summarizes the performance measures expected to be affected by specific treatments, and their expected effect on the condi- tion jump and/or in the change in long-term performance. Figure 23. Change in grinding index after diamond grinding.
Testing and Validation of Recommended Measures 33 In some cases the results summarized in Table 2 contradict what was found in the literature. For example, Labi et al. (2007) reported an improvement in IRI on Indiana DOT roads follow- ing microsurfacing. However, data from the Maryland, Ohio, and Utah DOTs showed no such improvement. Also, while changes in performance were identified only in a few cases in this analysis, the condition jump was consistently identified for several treatments and performance measures. The results presented in this chapter show that the selected performance measures reflect the effects of pavement preservation, and are appropriate for use as performance measures. Although the data collected from state highway agencies generally contained a significant level of time series variance, the methodology proposed in this project captured the effect of preserva- tion in most cases. Pavement Type Preservation Treatment Effect on Roughness (IRI) Effect on Cracking (at least one cracking type) Effect on Rutting Effect on Faulting Initial Condition Change Long-Term Condition Change Initial Condition Change Long-Term Condition Change Initial Condition Change Long-Term Condition Change Initial Condition Change Long-Term Condition Change Asphalt Thin Asphalt Overlay Yes Yes Yes Yes Yes Yes Chip Seal None Yes Yes Yes None None Microsurfacing None - Yes - - - Concrete Diamond Grinding â No DBR Yes - None - Yes - Diamond Grinding â With DBR Yes - None - Yes - - â Not assessed None â No effect demonstrated. Yes â Effect demonstrated using data from highway agency, LTPP and/or literature. Table 2. Effect of preservation treatments on performance measures.
