Processing math: 100%
English  /  中文
Javon Adams, Cassie Castorena, Y. Richard Kim. 2019: Construction quality acceptance performance-related specifications for chip seals. Journal of Traffic and Transportation Engineering (English Edition), 6(4): 337-348. DOI: 10.1016/j.jtte.2019.05.003
Citation: Javon Adams, Cassie Castorena, Y. Richard Kim. 2019: Construction quality acceptance performance-related specifications for chip seals. Journal of Traffic and Transportation Engineering (English Edition), 6(4): 337-348. DOI: 10.1016/j.jtte.2019.05.003

Construction quality acceptance performance-related specifications for chip seals

More Information
  • Author Bio:

    Dr. Javon Adams was a postdoctoral research scholar in the Department of Civil, Construction, and Environmental Engineering at North Carolina State University researching the development of mix designs and performance specifications for bituminous surface treatments. He currently serves as the Director of Engineering Transfer Programs in the College of Engineering's Office of Academic Affairs at North Carolina State University

    Dr. Cassie Castorena is an associate professor in the Department of Civil, Construction, and Environmental Engineering at North Carolina State University. She is interested in multi-scale characterization of asphalt materials, asphalt binder modification, asphalt pavement design, and asphalt pavement distress mechanisms

    Dr. Y. Richard Kim is the Jimmy D. Clark Distinguished University Professor and Alumni Association Distinguished Graduate Professor in the Department of Civil, Construction, and Environmental Engineering at North Carolina State University and Changjiang Scholar in the Department of Materials Science and Engineering at Chang'an University. Professor Kim's specialty areas include characterization and performance modeling of asphaltic materials and asphalt pavement systems, condition assessment of asphalt pavements, pavement evaluation by accelerated pavement testing, and pavement preservation. E-mail address: kim@ncsu.edu

  • Corresponding author:

    Y. Richard Kim, Tel.: +1 919 515 7758; fax: +1 919 515 7908

  • Peer review under responsibility of Periodical Offices of Chang'an University.

  • Received Date: November 26, 2018
  • Revised Date: March 28, 2019
  • Accepted Date: April 14, 2019
  • Available Online: September 28, 2022
  • Published Date: May 31, 2019
  • The research described herein details the establishment of a framework for construction quality acceptance performance-related specifications (PRS) for chip seal surface treatments. This paper establishes acceptance quality characteristics (AQCs) and relationships between the AQCs and related chip seal performance measures. This work is a critical step in developing PRS for constructed chip seal treatments and establishing suitable guidelines for the PRS. The main objective of this paper is to determine appropriate test methods to evaluate each defined AQC. The relationships that are established between the AQCs and chip seal performance are used to define performance ranges and threshold values for a particular performance measure. These performance ranges and threshold values then can be used to define pay factors for the constructed chip seal section. The threshold values and pay factors for each AQC described herein are preliminary recommendations and require further validation prior to implementation. However, this research demonstrates how these PRS would be utilized to ensure acceptable chip seal construction quality in the field.

    The developed construction quality acceptance PRS framework uses the percent within limits (PWL) concept to determine whether a chip seal lot passes the PRS threshold values and whether the contractor receives full pay, is subject to a pay penalty, or fails the construction PRS and must correct the chip seal within the first year in service, which constitutes the typical warranty period for contracted chip seal work.

    Finally, recommendations are made as to the next steps in the development and validation of these construction quality acceptance PRS that are needed for implementation by contractors and state roadway maintenance agencies.

    HIGHLIGHTS
    ● A framework is developed for the performance-related specification of chip seal surface treatments.
    ● Relationships are identified between acceptance quality characteristics (AQCs) and critical chip seal performance measures.
    ● Practical test methods are identified to measure the AQCs.
    ● Recommendations are made for the establishment and future implementation of pay factors for chip seal construction.
  • A framework for construction quality acceptance performance-related specifications (PRS) for chip seals aims to provide guidelines that govern the acceptance, appropriate penalty, or rejection of a chip seal surface treatment based on whether samples taken over a defined area (or lot) meet established performance test standards. The framework presented herein is intended as a starting point to address construction variability problems that have been observed in previous researches led by the authors, such as the North Carolina Department of Transportation (NCDOT) HWY-2008-04 project (Kim and Adams, 2011), where the measurements of the emulsion and aggregate application rates revealed significantly high rate variability in the field construction; that is, the measured rates were significantly different from the targeted design rates. The result of such application rate inaccuracies is that the chip seal experiences aggregate loss or bleeding. As many state highway maintenance agencies use independent contractors to construct their chip seals, it is critical that construction quality acceptance specifications are available that can be used to accept or reject a chip seal based on performance-related test methods. The following list provides an overview of the steps taken to develop this construction quality acceptance PRS framework:

    1) Select appropriate acceptance quality characteristics (AQCs) that are related to chip seal performance.

    2) Identify practical test methods and parameters that reliably capture the AQCs.

    3) Demonstrate the relationships between the AQCs and critical chip seal performance measures.

    4) Establish initial quality acceptance and rejection limits.

    5) Develop a sampling and measurement plan.

    6) Define pay adjustment factors.

    Determining the appropriate AQCs and establishing performance relationships were critical steps in developing this study's PRS framework. A key objective was to establish relationships between the AQCs and the pavement performance of a chip seal. In the process of developing these relationships, certain ranges of a particular performance measure could be correlated to the threshold value of the selected AQC. Examples of such correlation were established by collecting and analyzing relevant data through:

    • Analyzing performance-related trends in historical data obtained from existing projects.

    • Establishing relationships between AQCs and performance measures.

    • Applying engineering judgment and statistical analyses.

    Once the relationships between the AQCs and performance measures were established, and the methods for measuring the AQCs were identified, the limits or thresholds of the AQCs for acceptable levels of performance were determined. General guidelines were provided to determine such limits for AQCs based on performance thresholds. The successful adoption of these construction quality acceptance PRS requires the use of objective test methods that can measure performance-related parameters. The process of specifying these test methods considers the following concepts:

    • Minimizing the impact on user delays.

    • Collecting and processing data in a timely manner.

    • Placing emphasis on nondestructive testing techniques.

    Table 1 details the AQCs that have been identified and presents the performance measures that relate to each AQC, associated test methods, and the specific parameter measured in the field or laboratory for the various AQCs.

    Table  1.  Overview of acceptance quality characteristics for the PRS.
    Acceptance quality characteristics Related performance measure Test method Test parameter
    Emulsion-aggregate adhesive strength Aggregate loss Vialit test (Lab) Aggregate loss (%)
    Gradation Aggregate loss Gradation analysis of Vialit samples (Lab) Performance uniformity coefficient
    Emulsion application rate (EAR) Aggregate loss and bleeding Ignition oven: Vialit samples Difference from target emulsion application rate (EAR) (%)
    Aggregate application rate (AAR) Aggregate loss and bleeding Ignition oven: Vialit samples Difference from target aggregate application rate (AAR) (%)
     | Show Table
    DownLoad: CSV

    Aggregate loss is the primary distress in chip seals at intermediate temperatures. One of the main causes of aggregate loss is a lack of adhesive bond strength between the aggregate and emulsion such that significant cover aggregate is lost upon traffic loading. The adhesive bond between the aggregate and emulsion is a function of the construction practices employed during the seal construction. Construction-related factors, such as the time between the application of the aggregate layer onto the emulsion and the first rolling pass, the type of compaction effort applied, the number of roller passes, and the curing time allowed prior to traffic opening, can affect the adhesive bond that should form between the aggregate and emulsion and thus the aggregate loss observed (Lee and Kim, 2008). The Vialit testing of extracted field samples in these PRS directly measures the strength of the adhesive bond that is formed during the construction of the chip seal treatment. This AQC (i.e., the emulsion-aggregate adhesive strength) was determined to be critical to aggregate loss performance during the National Cooperative Highway Research Program (NCHRP) 9–50 project research conducted by the authors (detailed in NCHRP Report 837) (Kim et al., 2016).

    The performance uniformity coefficient (PUC) is a performance indicator of aggregate gradation and gives an indication of the uniformity, or lack thereof, of the aggregate being analyzed. In chip seal surface treatments, gradations that are more uniform perform better than those that are less uniform in terms of aggregate loss and bleeding failure criteria. This AQC, i.e., the PUC of the aggregate, affects the bleeding and aggregate loss performance of the chip seal surface treatment being constructed. The concept of the PUC is founded on principles that are based on McLeod's chip seal failure criterion. Essentially, McLeod's premise that 70% embedment is the ideal embedment for chip seal surface treatments is implemented in the PUC definition. The PUC is the ratio of the "percentage passing" at a given embedment depth (PEM) to the "percentage passing" at twice the embedment depth (P2EM) in a sieve analysis curve (Lee, 2007, McLeod, 1969).

    The emulsion application rate (EAR) and aggregate application rate (AAR) are AQCs that are both critical to the performance of chip seal surface treatments. Previous findings by the authors indicated that considerable variability often exists between the measured and designed EAR and AAR in field chip seal construction (Kim and Adams, 2011). To determine the actual field EARs and AARs for a constructed chip seal, samples are extracted from the chip seal sections after construction and then transported to the laboratory to undergo ignition oven testing to validate the EARs and AARs used for each field section. The ignition oven tests are carried out in accordance with the AASHTO T 308 specifications for asphalt mixtures to determine the amount of asphalt residue that was burned off during the ignition oven test (which then can be converted to an emulsion rate using the residual asphalt content for the emulsion) and the amount of applied aggregate initially applied over the sample area. These measured rates are then compared against the targeted design application rates to identify the amount of construction rate variability.

    In summary, for the four AQCs defined in Table 1, the authors observed various relationships between the AQCs and functional performance measures and then developed preliminary limits that require further validation prior to implementation.

    One of the most critical performance measures for a chip seal surface treatment is aggregate loss/retention. The strength of the adhesive bond that forms between the emulsion and aggregate used in chip seal construction is vital to the ability of the chip seal to retain aggregate under field traffic loading. Therefore, the authors measured the strength of that bond directly by employing Vialit aggregate loss impact loading tests for specimens extracted directly from constructed field sections. The authors have used this method successfully in numerous past field construction efforts to collect field aggregate loss data (Adams and Kim, 2014, Im, 2013).

    Fig. 1 shows the relationship between the Vialit test aggregate loss results and the bitumen bond strength (BBS) that was measured using a PATTI (pneumatic adhesive tensile testing instrument) to test the emulsions in accordance with AASHTO TP 91, determining asphalt binder bond strength by means of the bitumen bond strength test. The data show a relationship between the Vialit test aggregate loss results and the bond strength values for a variety of modified and unmodified emulsion types at different test temperatures, and prove that Vialit test aggregate loss results can be used effectively as an AQC to capture the bond strength between the aggregate and emulsion used in a chip seal.

    Figure  1.  Vialit test aggregate loss performance vs. Bitumen bond strength.

    The relationship shown in Fig. 1 suggests that the BBS value might be a suitable candidate AQC to characterize the chip seal's resistance to aggregate loss. However, the BBS values shown in Fig. 1 were obtained using the substrate of the specific aggregate used in the Vialit test. The effort that would be required to perform a BBS test for each project suggests that the Vialit test, which uses chip seal samples obtained directly from actual pavements, is a more practical way to determine the aggregate retention for a specific project and to evaluate adhesive bond strength than BBS testing. The Vialit test also provides a more direct measure of the performance measure (i.e., aggregate loss) that is related to the adhesive bond strength than the BBS test. The results presented in Fig. 1 also show that the Vialit aggregate loss test is able to differentiate effectively between modified and unmodified emulsions at multiple intermediate temperatures.

    To develop the Vialit test aggregate loss thresholds for the PRS, limits are needed to be derived based on the traffic demand expected for a constructed chip seal section. For example, roadways with higher traffic levels often have higher speeds, and vehicles are thus more susceptible to windshield damage due to aggregate loss than is the case for lower volume roads. Therefore, an acceptable aggregate loss threshold value would be lower (i.e., more restrictive) for higher traffic levels than lower traffic levels. Conversely, at lower traffic levels, the aggregate loss threshold should be less restrictive compared to the threshold at higher traffic levels. The PRS provide threshold values for three different traffic levels, as defined in Table 2.

    Table  2.  Traffic level definitions for the PRS.
    Traffic level Average annual daily traffic (AADT) (in vehicles)
    Low ≤500
    Medium 500 < AADT < 2500
    High 2500–20, 000
     | Show Table
    DownLoad: CSV

    These traffic levels are consistent with the recommended traffic levels defined in the NCHRP 9–50 project, which proposed PRS for emulsions used in chip seal treatments (Kim et al., 2016). In the NCHRP 9–50 project, the authors recommended 20, 000 vehicles as the upper average annual daily traffic (AADT) limit for high-volume traffic based on a study of high-volume traffic chip seal practices across the United States as well as the authors' own experience. In California, Colorado, and Montana, for example, chip seals are commonly constructed at AADT counts that can exceed 20, 000 vehicles (Gransberg and James, 2005). Also, the authors have constructed chip seals at an AADT above 15, 000 vehicles with no reported performance problems (Kim and Im, 2015). However, the performance of chip seals constructed at high traffic volumes is heavily dependent on local factors, such as climate, traffic speed, aggregate quality, contractor's experience, equipment, etc. Therefore, the high-volume traffic upper limit is conservatively set at 20, 000 vehicles for these construction quality acceptance PRS.

    To develop Vialit aggregate loss limits for the PRS, an aggregate loss limit that can differentiate between acceptable and unacceptable mixture performance was needed. Two aggregate loss limits were adopted from previous research studies that were based on laboratory and field chip seal experiments. The first limit is the maximum allowable aggregate loss limit for the lowest traffic level. The Alaska Department of Transportation defines "acceptable" field aggregate loss as 10% or less for any traffic situation where a chip seal is constructed (McHattie, 2001). This 10% aggregate loss limit also has been found in previous research to characterize acceptable aggregate loss for third-scale model mobile load simulator (MMLS3) testing (Kim and Adams, 2011, Lee, 2007, Lee and Kim, 2009). These earlier research studies found that, if a chip seal exhibits 10% aggregate loss in the laboratory, it is likely to exhibit significant aggregate loss (based on visual inspection) in the field.

    However, these previous research studies also showed that Vialit-tested specimens exhibit more aggregate loss than MMLS3-tested specimens when testing chip seals constructed and tested under the same conditions. The relationship between the MMLS3 test results and Vialit test aggregate loss results was examined in research detailed in a previous NCDOT project final report (Kim and Im, 2015). The researchers found a relationship between the MMLS3 and Vialit mixture test performance for both modified and unmodified binders. The mixture specimens that were Vialit-tested and MMLS3-tested for performance were extracted directly from chip seal field sections constructed by an experienced NCDOT chip sealing crew. Table 3 summarizes the relationships observed between the Vialit-tested and MMLS3-tested chip seal samples at 25 ℃.

    Table  3.  Relationship between vialit and MMLS3 aggregate loss test results at 25 ℃ for chip seals (Kim and Im, 2015).
    Average Vialit agg. loss (%) Average MMLS3 agg. loss (%) Vialit to MMLS3 ratio MMLS3 agg. loss limit (%) Vialit agg. loss limit (%)
    Modified emulsions 11 8 1.375 10 10 × 1.375 = 13.75
    Unmodified emulsions 22 11 2 10 10 × 2 = 20
     | Show Table
    DownLoad: CSV

    The results presented in Table 3 indicate that, for unmodified emulsions, which often are used in low-volume traffic situations, the Vialit test aggregate loss is double the aggregate loss caused by MMLS3 testing. Therefore, based on the 10% MMLS3 aggregate loss threshold established for the MMLS3 and the field, the highest allowable equivalent Vialit test aggregate loss is 20%. Therefore, a 20% maximum aggregate loss threshold is suggested as the low-volume traffic limit for the Vialit test in these PRS, as this value is roughly equivalent to the 10% aggregate loss limit established for the field and MMLS3 wheel loading. Aggregate loss should not exceed this 20% Vialit test limit even at the lowest traffic level that a chip seal experiences, as the windshield damage and reduction in skid resistance associated with excessive aggregate loss beyond this limit can be highly hazardous for vehicles, drivers, pedestrians, etc.

    The appropriateness of the 20% aggregate loss limit was substantiated by field performance findings. Chip seal sections were constructed using both modified and unmodified emulsions in the same lane at a single construction location to remove all variables except for emulsion type from the field study. The specimens that were Vialit-tested and MMLS3-tested to obtain the data shown in Table 3 were extracted from these same sections so that field versus laboratory aggregate loss also could be evaluated. The chip seal section that used modified emulsion did not exhibit any aggregate loss-related problems, and the specimens extracted from the modified field sections had Vialit aggregate loss below 20%. The unmodified chip seal section displayed significant aggregate loss in the field, as shown in Fig. 2. The specimens extracted from the unmodified field section exhibited Vialit test aggregate loss in the lab above 20% (as shown in Table 3). This combination of field and lab findings supports the 20% aggregate loss threshold as an appropriate limit for characterizing aggregate loss performance for the PRS.

    Figure  2.  Images of aggregate loss observed at the unmodified CRS-2 field section in Durham, NC.

    Fig. 3 shows the low-volume traffic aggregate loss limit of 20% plotted against the aggregate loss performance for both modified and unmodified emulsions at two different test temperatures. The figure indicates that both the poor-performing emulsion and the two unmodified emulsions with the lowest bond strength values show the most aggregate loss and are right at the low-volume traffic threshold of 20% aggregate loss. Note that the poor-performing emulsion was an emulsion that was intentionally altered by the emulsion supplier to be poor-performing in terms of aggregate retention, but would still meet all current emulsion specifications.

    Figure  3.  Low-volume traffic aggregate loss limit.

    The high-volume traffic aggregate loss limit for use in the development of the PRS was found through a combination of research findings and results reported in the literature. In the report, Chip Seals for High Traffic Pavements (Shuler, 1990), Shuler recommends that polymer-modified emulsions should be used in high-volume traffic situations. Pasquini et al. (2014) also demonstrated that the use of polymer-modified binders in chip seal applications resulted in better performance. In addition, Kim and Im (2015) found through field chip seal experiments that modified binders should be used exclusively for chip seal surface treatments in high-volume traffic situations. In Kim and Im's study, single-seal and triple-seal field validation sections were constructed in the same lane of a roadway in North Carolina with an AADT count of 5000 vehicles (i.e., high-volume traffic). These chip seal sections were visited after the first year of traffic loading for monitoring. The field validation sections clearly showed that the modified binder outperformed the unmodified binder on the same high-volume roadway, as the roadway constructed with modified binder exhibited no performance problems, whereas the roadway constructed with unmodified binder experienced significant aggregate loss.

    Therefore, to determine the high-volume traffic performance limit for the PRS, the Vialit test lab data presented in Fig. 4 were utilized to select the aggregate loss threshold that could distinguish between the modified emulsion, which is known to perform well in high-volume traffic situations, and the unmodified emulsion, which is not recommended for use in high-volume traffic situations. The high-volume traffic aggregate loss limit of 15% was selected as the aggregate loss threshold that can distinguish modified binders from unmodified binders in terms of performance, as shown in Fig. 4.

    Figure  4.  High-volume traffic aggregate loss limit.

    With the aggregate loss threshold limits for low-volume traffic (i.e., 20% aggregate loss) and high-volume traffic (i.e., 15% aggregate loss) established, the medium-volume traffic limit was selected as the average of the low-volume and high-volume traffic limits (i.e., 17.5% aggregate loss). This medium-volume traffic limit is shown in Fig. 5 plotted against the Vialit test performance data.

    Figure  5.  Medium-volume traffic aggregate loss limit.

    Table 4 summarizes the performance limits established for the Vialit test aggregate loss AQC according to the traffic level at the chip sealing location.

    Table  4.  Summary of Vialit test performance limits based on traffic level.
    Traffic level AADT Vialit test aggregate loss performance threshold (%)
    Low ≤500 20.0
    Medium 500 < AADT < 2500 17.5
    High 2500–20, 000 15.0
     | Show Table
    DownLoad: CSV

    Gradation is a significant AQC that is related directly to the performance of chip seal treatments. The AQC parameter that is representative of the effect of gradation on performance is the PUC. In chip seal surface treatments, gradations that are more uniform perform better than those that are less uniform in terms of aggregate loss and bleeding failure criteria (Adams and Kim, 2014, Lee and Kim, 2008).

    For these PRS, the PUC was demonstrated to be an AQC that is related directly to aggregate loss. The relationships between the PUC and aggregate loss are shown in Fig. 6 for granite aggregate and Fig. 7 for lightweight aggregate. Each data point in the figures is the average of nine chip seal specimens that were traffic-loaded using the MMLS3. All of the specimens were fabricated using a single CRS-2L emulsion with the residual binder rate of 67%. The optimum EAR was found to be 0.9 l/m2 and the optimum AAR was found to be 8.7 kg/m2 using a performance-based mix design method (Adams and Kim, 2014). Note that these mix design optimum rates are specific to the granite aggregate used in these experiments.

    Figure  6.  Aggregate loss measured from MMLS3 testing vs. PUC for granite 78M aggregate.
    Figure  7.  Aggregate loss measured from MMLS3 testing vs. PUC for lightweight aggregate.

    After attempting various models to fit the data, the cumulative distribution form of the skewed logistic function was found to provide a good fit for the relationship between aggregate loss and the PUC for a given EAR. The model is defined in Eq. (1) and the model parameters are visually represented in Fig. 8.

    %AggLoss=%AggLossu(1+(PUCa)(log2/b))b/log2 (1)

    where %AggLossu is the asymptotic value that the aggregate loss approaches at high PUC values, a is allocation parameter, b is a shape parameter.

    Figure  8.  Illustration of the three cumulative distribution function parameters.

    Using Eq. (1) to predict the percentage of aggregate loss while minimizing error, Fig. 6 and 7 show that the model predictions fit the measured data for the PUC and the percentage of aggregate loss for the MMLS3-loaded chip seal specimens for the granite aggregate and lightweight aggregate, respectively.

    Fig. 6 shows that, as the PUC increases and the gradation becomes less uniform, the aggregate loss increases. This trend is shown for different EARs as a percentage of the optimum design EAR for the granite 78M aggregate used in these experiments. The data show that, at a low PUC value below 20 (i.e., greater aggregate uniformity), decreasing the EAR below 100% of the optimum rate has less effect on the aggregate loss. Conversely, the aggregate loss at PUC values above 50 becomes high at very low EARs.

    Fig. 7 shows a linear trend for the lightweight aggregate. Also, the magnitude of the aggregate loss is much less for the lightweight aggregate than for the granite 78M at all EARs, even when comparing the aggregate loss percentage at the respective optimum EAR. Even at high PUC values, the aggregate loss percentage is low for the lightweight aggregate, indicating that lightweight aggregate is a superior material for chip sealing compared to granite aggregate. In addition to the lower magnitude of aggregate loss shown in Fig. 7, the significantly lower density (compared to other coarse aggregate such as granite) of the manufactured coarse lightweight aggregate particles mitigates the risk of windshield damage due to aggregate loss. Therefore, the use of superior materials such as lightweight aggregate is proven to reduce the risk of aggregate loss significantly, and its usage should be encouraged in the PRS if other performance properties of lightweight aggregate, such as strength and resistance to abrasion, are satisfactory. Development of incentives for the use of superior materials is a recommendation for future research.

    The data shown in Fig. 6 and 7 for the two different aggregate types serve as evidence of the appropriateness of the PUC as an effective AQC that relates to aggregate loss performance. After using the cumulative distribution model to predict the percentage of aggregate loss from the PUC, appropriate values for the a and b parameters were found such that these parameters could be held constant. Holding a and b constant, the model was used to solve for the %AggLossu parameter, which represents the asymptotic behavior of the PUC versus aggregate loss curves. The %AggLossu parameter is the critical model parameter for the purposes of this analysis because the asymptotic aggregate loss value is the critical performance measure for chip seal treatments in the PRS. In Fig. 9, the %AggLossu is plotted as a function of the changing EAR as a percentage of the optimum EAR. This figure shows that the %AggLossu has a relationship with the percentage of the optimum rate such that, as the EAR decreases below the optimum rate, the %AggLossu model parameter (or asymptotic aggregate loss prediction) increases, as expected.

    Figure  9.  Model parameter %AggLossu plotted as a function of percentage of optimum EAR.

    Fig. 9-11 respectively present the three model parameters, %AggLossu, a, and b, plotted as functions of the percentage of the optimum EAR.

    Figure  10.  Model parameter a plotted as a function of percentage of optimum EAR.
    Figure  11.  Model parameter b plotted as a function of percentage of optimum EAR.

    The approach used to determine the PUC threshold limit for the construction acceptance quality PRS is based on the concept that 100% of the optimum EAR (which is based on the performance-based mix design) yields the appropriate baseline for aggregate loss. This performance-based mix design has been proven to minimize simultaneously both the potential for aggregate loss and bleeding problems in chip seal mixtures (Adams and Kim, 2014). Fig. 12 presents the approach for developing threshold values for the PUC.

    Figure  12.  Approach to developing threshold values for the PUC AQC based on design rates.

    The asymptotic aggregate loss for chip seal specimens designed at 100% of the optimum EAR and at the optimum AAR of 8.4 kg/m2 yields an aggregate loss limit of just above 8%, which is below the 10% aggregate loss threshold typically used to assess chip seal aggregate loss performance (McHattie, 2001). Fig. 12 presents a visual representation of this concept. The horizontal dashed line represents the asymptotic percentage of the aggregate loss value for 100% of the optimum EAR curve. The points at which the curves for the other percentages of the optimum EAR cross this horizontal dashed line reveal the appropriate PUC threshold values for the respective curves. Using this approach, the maximum PUC (or minimum allowable aggregate gradation uniformity) for satisfactory performance is a function of how close the measured EAR is to the optimum EAR. This approach is practical, because a chip seal with an EAR below the optimum EAR requires a better, more uniform aggregate to obtain the same aggregate loss/retention results as a seal with an EAR that meets the design requirements. For example, the approach shown in Fig. 12 indicates that, for the green curve developed for chip seals with a measured EAR at 85% of the optimum EAR of 0.9 l/m2, the maximum allowable PUC is approximately 32, whereas the red curve, which represents chip seals at 55% of the optimum EAR has a more restricted maximum allowable PUC of approximately 24 to indicate acceptable performance.

    Also, the graph presented in Fig. 12 indicates the margin for error at various PUC values. If a contractor constructs a chip seal using aggregate that has a PUC below 20, the effect of the EAR is less than if the contractor uses an aggregate with a high PUC around 50. With a PUC of approximately 50, if the measured EAR is 70% or less of the optimum EAR, aggregate loss problems likely will occur. In the past, the authors have observed measured field-constructed EAR values that are below the design optimum EAR by over 30% for the same aggregate and emulsion materials used in these experiments (Adams and Kim, 2014).

    Fig. 12 also reveals that the sensitivity of the aggregate loss/retention to the PUC parameter (i.e., gradation) is related to the EAR and, more specifically, to how close the measured EAR is to the design optimum EAR. This observation helped to provide threshold values for the aggregate and emulsion utilized in this research; however, more research is needed in the future to establish PUC threshold values that are verified as appropriate for a wider range of emulsion and aggregate combinations.

    The steps for establishing threshold limits for the AQCs and quality measures are as follows.

    (1) Determine AQC performance relationships. The relationships between each AQC and performance have been established. The aggregate loss (measured from the Vialit test) and the gradation (represented by the PUC) allow the prediction of the key performance measures that are associated with chip seal treatments.

    (2) Set specification limits. Initial specification limits for the AQCs have been determined based on laboratory and field performance data as well as engineering judgment. The research conducted in this study provides an example and framework for the construction quality acceptance PRS, although final test limits require refinement and validation prior to implementation.

    (3) Decide on a quality measure. The recommended quality measure for chip seals was decided as the "percent within limits" (PWL).

    (4) Define acceptance quality limits (AQLs). The upper AQL for chip seal treatments is recommended to be a PWL of 90 for this chip seal demonstration example, based on typical AQL values. That is, 90% of samples from a lot must pass the AQC specification limit to receive 100% pay.

    (5) Define rejection quality limit (RQL). The RQL for this chip seal demonstration example was determined to be a PWL of 60, based on the typical range of RQL values. That is, 60% of samples from a lot must pass the specification limit to be eligible for reduced pay.

    The PWLs were calculated for chip seal field sections used in previous NCDOT projects that were constructed on roadways with various traffic levels to exemplify how the concept would work in practice. Table 5 presents the results for the Vialit test aggregate loss tests that were conducted to determine the PWL values. These PWL values were then used to determine whether the contractor for a sample lot would have received full pay, reduced/partial pay, or if the work would have been rejected based on the construction PRS framework.

    Table  5.  Vialit test aggregate loss PWL values.
    Section ID Traffic volume Upper spec. limit Avg. loss Std. Dev. Q Sample size PWL Pay conclusion
    MD-1 4500 15.0 11.0 2.54 1.57 9 95.2 Full pay
    MD-2 4500 15.0 8.4 1.46 4.52 9 100.0 Full pay
    MD-3 4500 15.0 12.2 2.10 1.33 9 91.4 Full pay
    MD-7 4500 15.0 2.8 0.84 14.52 9 100.0 Full pay
    MD-8 4500 15.0 4.2 1.03 10.49 9 100.0 Full pay
    MD-9 4500 15.0 1.8 1.40 9.43 9 100.0 Full pay
    MD-10 1000 17.5 7.3 1.67 6.11 9 100.0 Full pay
    MD-11 1000 17.5 9.2 3.10 2.68 9 100.0 Full Pay
    MD-12 1000 17.5 13.6 1.75 2.23 9 99.7 Full Pay
    MDV-1 2000 17.5 10.2 1.10 6.64 9 100.0 Full Pay
    MDV-2 2000 17.5 16.7 2.70 0.30 9 59.5 < 60 lot rejected
     | Show Table
    DownLoad: CSV

    In Table 5, Q is the quality index function which was developed to calculate party risks for inspection plans for the purpose of using sample data to estimate the proportion of a normal population that lies outside a given interval (Lieberman and Resnikoff, 1955). Q is defined as the difference between the upper limit and the mean of the sample data divided by the standard deviation. In this context, the upper limit is the maximum aggregate loss allowable for the given traffic volume with the average and standard deviation coming from the aggregate loss tests of the samples taken from the chip seal field section.

    The risks associated with incorrectly accepting or rejecting a lot are related to sample size. The procedure for developing guidelines for a sampling and measurement plan for chip seal treatments is as follows.

    (1) Determine which party performs acceptance testing. This decision must be agreed upon by the contractor and the state bituminous maintenance agency.

    (2) Determine the type of acceptance plan to be used. Stratified random sampling, which is a modified version of random sampling commonly used in pavement construction acceptance sampling, is recommended. This sampling involves dividing lots into several sublots of equal size (Freeman and Grogan, 1998). Random sampling then occurs within each sublot with stratification, thus ensuring that the samples are distributed evenly throughout the entire sublot. Freeman and Grogan (1998) outline the three rules of stratified random sampling as follows.

    ① An equal number of samples are taken from each sublot.

    ② The sublots are of equal size.

    ③ Samples are selected randomly within each sublot.

    (3) Develop verification sampling and testing procedures. Verification sampling is a standard procedure and is used to verify the accuracy of quality acceptance test results. The decision whether to use split or independent sampling depends on the goals of the particular agency. For this example, it is assumed that the agency or independent third party will measure the Vialit aggregate loss at the recommended sampling frequency for each sublot for verification. In practice, it is appropriate that the agency's verification test methods are used solely for verification and that acceptance methods proposed by the contractor must first be compared to the results of the agency verification tests.

    (4) Select the appropriate verification sampling frequency. The verification sampling frequency of the agency should be approximately 10% of the acceptance sampling rate of the contractor. In practice, the verification testing frequency is decided for economic, rather than statistical, reasons. This decision must be agreed upon by the agency and contractor.

    (5) Determine lot size and sample size. The evaluation of the aggregate loss AQC, for example, involves the extraction of field samples for Vialit aggregate loss testing in a temperature-controlled laboratory environment. Therefore, lots and sublots should be defined logically as segmented lengths of a project. For aggregate loss testing, the recommended lot size is 5000 feet long. Sublot lengths from which stratified random sampling should be performed are recommended to be 100 feet. Naturally, the risks associated with sampling depend on sample size. Based on previous chip seal field research results, it is recommended that a sample size of nine is used for each sublot for the Vialit test aggregate loss AQC (Kim and Adams, 2011). However, for practical purposes, it is not reasonable to take nine samples from each sublot throughout an entire lot, because this effort would be time-consuming and transporting so many samples for laboratory testing would be difficult. Therefore, it is recommended that three sublots are selected randomly from the lot for sampling. Individual agencies may increase the number of sublots sampled to minimize risk if sufficient resources and personnel become available.

    For the evaluation of the PUC AQC, current agency practices include gradation measurements taken at the aggregate quarry to ensure that the aggregate specified in the chip seal contract meets the gradation requirements. Therefore, as the PUC is determined directly from the aggregate gradation, no field sampling is required. Current quality control checks of gradation should be maintained, and the PUC can be checked using the gradation data.

    Pay adjustment factors are necessary for quality acceptance plans when developing PRS. However, establishing pay reduction factors to determine partial pay using typical approaches that are based on reduction of service life concepts is not appropriate for performance measures such as aggregate loss, which is the most critical distress for chip seal treatments (Lee, 2007). For instance, most aggregate loss occurs within the first days and weeks in service, and aggregate loss early in the life of the seal does not necessarily lead to a reduction in the overall service life of the seal, as observed in field sections constructed in previous research (Im, 2013). Therefore, the authors obtained the opinion of pavement maintenance practitioners with state highway agencies to provide pay adjustment factor recommendations as a starting point. These recommendations should be validated in a rigorous manner prior to implementation.

    Issues associated with aggregate loss include vehicle damage claims that may occur and the resultant public perception of chip seal treatments as an ineffective treatment alternative. The other problem with aggregate loss is that it is one of the leading causes of pavement bleeding (Lawson, 2006). Given the established maximum specification thresholds of 20%, 17.5%, and 15% aggregate loss for low, medium, and high traffic volumes, respectively, a set of samples for a lot that fails to meet the AQL of 90, but exceeds the RQL of 60, should receive only partial pay.

    A relationship between aggregate loss and pay factors could not be developed fully in this study based on existing data because quantifying the effect of aggregate loss on bleeding, as well as on public perceptions/satisfaction with the quality of the seal work, is difficult. However, simply as a starting point for these PRS, the authors surveyed pavement maintenance practitioners from state highway agencies to obtain recommendations about reasonable partial pay factors for PWLs ranging from 60 to 90. The survey results were averaged and rounded to the nearest 5%. Table 6 presents the results as a function of the PWL.

    Table  6.  Pay factors for aggregate loss AQC.
    PWL range (%) Pay reduction
    90–100 Full pay
    75–90 25% pay reduction
    60–75 50% pay reduction
    0–60 Reject; no pay
     | Show Table
    DownLoad: CSV

    Also, the survey respondents unanimously recommended, and the opinion of the researchers supports, that the contracted party also should be responsible for addressing any vehicle damage claims at no cost to the state highway agency. The factors listed in Table 6 are the recommendations from experienced bituminous supervisors with previous experience overseeing chip sealing operations for state highway agencies and those of the authors. However, final pay reduction factors should be adjusted in accordance with additional research findings and agreed upon by all parties. Table 6 provides an example of how pay/penalty factors could be employed for the aggregate loss AQC based on the demonstrated performance of samples from a lot tested within this PRS framework.

    The research conducted in the development of this framework for construction quality acceptance PRS for chip seals resulted in the following conclusions.

    • Vialit testing of extracted field chip seal samples can effectively assess the aggregate retention potential of chip seals for different aggregate types, binder types, design rates, and traffic levels. Additionally, ignition oven testing of Vialit samples can reveal possible material application rate variability in constructed chip seals.

    • A relationship exists between the gradation (as represented by the PUC) that is measured for aggregate and performance measures such as aggregate loss. The PUC can be used to ensure that the aggregate selected for the chip seal meets the gradation requirements that are related to acceptable performance in chip seals.

    • Preliminary PWL values for each AQC can be used to determine if the contractor for a lot will receive full pay (AQC > 90), partial pay (60 < AQC < 90), or no pay (AQC < 60) for a constructed chip seal.

    • Survey results and engineering judgment provided guidance for preliminary pay adjustment factors for the AQCs. These preliminary pay adjustment factors require further consideration.

    • The construction quality acceptance PRS framework recommends that extracted Vialit samples should be used to measure rate variability and to assess aggregate loss and that the PUC, as an indicator of the uniformity of the aggregate gradation, should be addressed on a pass/fail basis during regular quality control testing of quarry material.

    • Overall, the construction quality acceptance PRS framework discussed in this paper provides protocols that can be used to address construction variability problems that have been observed during past chip seal construction research field efforts led by the authors.

    • The authors have developed a construction quality acceptance PRS framework that provides guidance for test procedures that can help identify construction-related problems in the future. These PRS are intended to provide practical solutions to help pavement management agencies minimize the risk of performance problems in contracted chip seal work.

    The following recommendations are made for future research.

    • Field-validate the construction quality acceptance PRS framework provided in this report by extracting samples from sections nationwide (as the findings provided herein reflect chip seal constructed only in North Carolina) and conducting ignition oven and Vialit tests using those samples. Also, monitor the field performance of those sections to adjust the recommended preliminary threshold values that translate to acceptable performance.

    • Obtain additional performance and cost data to establish and refine pay adjustment factors based on the percentage of aggregate loss observed from Vialit tests of field samples.

    • Finalize the sampling and measurement plan for measuring critical AQCs that are deemed practical by both state pavement maintenance agencies and the contracted personnel who typically construct chip seals. The goal here is to strike an appropriate balance between not testing enough, which increases risk, and testing too rigorously, which can lead to logistical and practicality problems in terms of increased traffic closings and specimen testing time.

    The authors would like to thank the North Carolina Department of Transportation for its support of the research through project HWY-2015-19 that led to the findings reported in this paper.

    Conflict of interest

    The authors do not have any conflict of interest with other entities or researchers.

  • Adams, J., Kim, Y.R., 2014. Mean profile depth analysis of field and laboratory traffic-loaded chip seal surface treatments. International Journal of Pavement Engineering 15(7), 645-656. doi: 10.1080/10298436.2013.851790
    Freeman, R., Grogan, W., 1998. Statistical Acceptance Plan for Asphalt Pavement Construction. GL-98-7. U.S. Army Corps of Engineers, Washington DC.
    Gransberg, D.D., James, D.M.B., 2005. reportChip Seal Best Practices. NCHRP Synthesis Report 342. Transportation Research Board, Washington DC.
    Im, J.H., 2013. report Performance Evaluation of Chip Seals for High Volume Roads Using Polymer-Modified Emulsions and Optimized Construction Procedures (PhD thesis). North Carolina State University, Raleigh.
    Kim, Y.R., Adams, J., Castorena, C., et al., 2016. reportPerformance-Related Specifications for Emulsified Asphaltic Binders Used in Preservation Surface Treatments. NCHRP Report 837. Transportation Research Board, National Research Council, Washington DC.
    Kim, Y.R., Adams, J., 2011. Development of a New Chip Seal Mix Design Method. HWY-2008-04. North Carolina Department of Transportation, Raleigh.
    Kim, Y.R., Im, J.H., 2015. Extending the Use of Chip Seals to High Volume Roads by Using Polymer-Modified Emulsions and Optimized Construction Procedures. FHWA/NC/2011-03. North Carolina Department of Transportation, Raleigh.
    Lawson, W.D., 2006. Maintenance Solutions for Bleeding and Flushed Pavements. 0-5230, Texas Department of Transportation, Austin.
    Lee, J., 2007. reportPerformance-Based Evaluation of Asphalt Surface Treatments Using the Third Scale Model Mobile Loading Simulator (PhD thesis). North Carolina State University, Raleigh.
    Lee, J., Kim, Y.R., 2008. Performance-based uniformity coefficient of chip seal aggregate. Transportation Research Record 2108, 53-60. doi: 10.3745/KIPSTB.2008.15-B.1.53
    Lee, J., Kim, Y.R., 2009. Quantifying the Benefits of Improved Rolling of Chip Seals. HWY-2006-06. FHWA, Washington DC.
    Lieberman, G.J., Resnikoff, G.J., 1955. Sampling plans for inspection by variables. Journal of the American Statistical Association 50(270), 457-516.
    McHattie, R.L., 2001. Asphalt Surface Treatment Guide. FHWA-AK-RD-01-03. Alaska Department of Transportation and Public Facilities, Fairbanks.
    McLeod, N.W., 1969. A general method of design for seal coats and surface treatments. Proceedings of the Association of Asphalt Paving Technologists 38, 537-628.
    Pasquini, E., Bonati, A., Giuliani, F., et al., 2014. Advanced characterization of clear chip seals. Journal of Testing and Evaluation 42(5) 1213-1227.
    Shuler, S., 1990. Chip seal for high traffic pavements. Transportation Research Record 1259, 24-33.

Catalog

    Figures(12)  /  Tables(6)

    Article views (13) PDF downloads (3) Cited by()
    Related

    /

    DownLoad:  Full-Size Img  PowerPoint
    Return
    Return