Performance Modeling of Rubberized Hot Mix Asphalt- A Literature Review (WO number 1)

Transcription

1 Performance Modeling of Rubberized Hot Mix Asphalt- A Literature Review (WO number 1) California Department of Resources Recycling and Recovery May 15, 2014 Contractor's Report Produced Under Contract By: Marissa Garcia and DingXin Cheng, PhD. California Pavement Preservation Center CSU, Chico

2 S T A T E O F C A L I F O R N I A Edmund G Brown Jr. Governor Matt Rodriguez Secretary, California Environmental Protection Agency DEPARTMENT OF RESOURCES RECYCLING AND RECOVERY Caroll Mortensen Director Department of Resources Recycling and Recovery Public Affairs Office 1001 I Street (MS 22-B) P.O. Box 4025 Sacramento, CA RECYCLE (California only) or (916) Publication # DRRR-XXXX-XXX [OPA editor will provide this] To conserve resources and reduce waste, CalRecycle reports are produced in electronic format only. If printing copies of this document, please consider use of recycled paper containing 100 percent postconsumer fiber and, where possible, please print images on both sides of the paper. Copyright 2014 by the California Department of Resources Recycling and Recovery (CalRecycle). All rights reserved. This publication, or parts thereof, may not be reproduced in any form without permission. Prepared as part of contract number DRR The California Department of Resources Recycling and Recovery (CalRecycle) does not discriminate on the basis of disability in access to its programs. CalRecycle publications are available in accessible formats upon request by calling the Public Affairs Office at (916) Persons with hearing impairments can reach CalRecycle through the California Relay Service, Disclaimer: This report was produced under contract by the California Pavement Preservation Center. The statements and conclusions contained in this report are those of the contractor and not necessarily those of the Department of Resources Recycling and Recovery (CalRecycle), its employees, or the State of California and should not be cited or quoted as official Department policy or direction. The state makes no warranty, expressed or implied, and assumes no liability for the information contained in the succeeding text. Any mention of commercial products or processes shall not be construed as an endorsement of such products or processes.

3 Table of Contents List of Tables... iii List of Figures... iv Acknowledgments... iv Executive Summary... vi Introduction... 1 Background... 1 Overall Project Objectives... 1 Purpose of Report... 2 Literature Review... 3 Colorado Remaining Service Life (RSL) Concept... 3 Minnesota Pavement Quality Index and Ride Quality Index, (3, 4)... 4 Minnesota Pavement Quality Index... 4 Minnesota Ride Quality Index (3)... 6 Mississippi Pavement Condition Rating System... 8 Original Flexible Pavement, Pavement Condition Rating (PCR) Model... 9 Overlaid Flexible Pavement, Pavement Condition Rating (PCR) Model MTC StreetSaver Pavement Condition Index, (6, 7)...10 New Jersey Stress Index, (8, 9)...12 Ohio Pavement Quality Index, (11,12)...14 PaveM Pavement Management System (13)...17 South Dakota Surface Condition Index, (14,15)...19 Stantec Pavement Management System, RoadMatrix TM, (16)...21 Arizona Pavement Distress Index (PDI), (17,18) Los Angeles County Pavement Condition Index (20) Texas DOT Deterioration Condition score, (21, 22)...24 Deterioration Condition Score for Future Predictions Added Life of Maintenance Treatments Washington Pavement Structural Condition Model, (23, 24)...28 All Structural Cracking (ACA) Wide Structural Cracking (ACW) Transverse Thermal Cracking Summary of Models...29 Contractor s Report i

4 Typical Performance Curves for Asphalt Rubber...30 Preliminary Conclusions and Recommendations...45 Conclusions...45 Recommendations...45 Abbreviations and Acronyms...47 Glossary of Terms...50 References...53 Contractor s Report ii

5 List of Tables Table 1 Conversion Table of TWD to SR... 5 Table 2 Bituminous Pavement SR Weighting Factors... 6 Table 3 Coefficients to Ohio Pavement Quality Index Equation Table 4 Severity and Extent Weight factors for Ohio PCR Table 5 Example MPD Values Table 6 Deduct Points from Fatigue Cracking and Patching Table 7 Deduct Points from Transverse Cracking Table 8 Deduct Points from Block Cracking Table 9 Curve parameter factors, β pq & ρ pq Contractor s Report iii

6 List of Figures Figure 1 Example of a Pavement Model Curve... 2 Figure 2 RQI Example 1 from Minnesota DOT... 7 Figure 3 RQI Example 2 from Minnesota DOT... 8 Figure 4 Typical Performance Curves for MTC StreetSaver Figure 5 New Jersey Pavement Distress Model Comparison (10) Figure 6 IRI Component of PQI Figure 7 Deduct Points from Rutting Figure 8 Deduct Points from IRI Figure 9 Pavement Condition Index deterioration curves for Los Angeles County Figure 10 The General Shape of the Pavement Performance Prediction Model used by TxDOT Figure 11 Texas Pavement models using Original Coefficients Figure 12 Pavement Deterioration curve of a RAC Overlay of an unknown depth on a Road with an Arterial Functional Class Figure 13 Pavement Deterioration curve of a RAC Overlay of an unknown depth on a Road with a Collector Functional Class Figure 14 Pavement Deterioration curve of a RAC Overlay of an unknown depth on a Road with a Residential Functional Class Figure 15 Pavement Deterioration curve of a 1.5 RAC Overlay on a Major Rural Collector Functional Class Figure 16 Pavement Deterioration curve of a 1.5 RAC Overlay on an Arterial Functional Class Figure 17 Pavement Deterioration curve of a 1.5 RAC Overlay on a Rural Local Functional Class Figure 18 Pavement Deterioration curve of a 1.75 RAC Overlay on an Arterial Functional Class Road.. 37 Figure 19 Pavement Deterioration curve of a 2 RAC Overlay on a Collector Functional Class Road Figure 20 Pavement Deterioration curve of a 2 RAC Overlay on an Arterial Functional Class Figure 21 Pavement Deterioration curve of a 2.5 RAC Overlay on a Collector Functional Class Figure 22 Pavement Deterioration curve of a 2.5 RAC Overlay on an Arterial Functional Class Figure 23 Pavement Deterioration curve of a 4 RAC Overlay on an Arterial Functional Class Figure 24 Pavement Deterioration curve of a Thin RAC Overlay with Fabric on a Major Rural Collector Functional Class Figure 25 Pavement Deterioration curve of a Dig out and 2 RAC Overlay on an Arterial Functional Class Contractor s Report iv

7 Acknowledgments We appreciate the financial support of the CalRecycle for providing the funding for this important and meaningful study. We would like to extend our gratitude to Nate Gauff and Bob Fujii of CalRecycle, who provided continuous support to this project. Lerose Lane also provided support on reviewing the document. Contractor s Report v

8 Executive Summary Performance models are used in most pavement management systems throughout the United States to predict future performance of various mix types. Models are generally available for hot mix asphalt (HMA), but not for rubberized asphalt concrete. The purpose of this literature review study was to identify the various performance curves used by agencies in their pavement management systems and to recommend a usable pavement performance model for rubberized asphalt concrete in order to confirm the benefits of asphalt rubber mixes verses conventional hot mix asphalt. Various models were evaluated to demonstrate the pavement performance over time and to predict future performance. The tasks of this work order were to: 1. Study and evaluate selected performance models to predict the long-term performance of hot mix asphalt materials through literature or by contacting various agencies that are using pavement management systems. 2. Review local government Rubberized Asphalt Concrete (RAC) project data to correlate field performance with material parameters. Recommend the most appropriate models for use by local agencies and states from California data. These findings should be of great benefit to the pavement management software used by agencies in California. The findings from this study included: A variety of models were evaluated to predict not only pavement distress, but also an overall combined indices such as Pavement Condition Index, Pavement Quality Index, Surface Distress Index, Cracking Index, etc. Most local agencies in California use MTC StreetSaver, Stantec s RoadMatrix, or MicroPaver. For this portion of the study, we focused our efforts using the StreetSaver program developed by the Metropolitan Transportation Commission. For the development of a Pavement Model for Asphalt Rubber, MTC StreetSaver s PCI Performance Model, Arizona DOT s PSR Performance Model, and PQI/SDI models by LA County and City of Sacramento will be used in the second report of Work Order 1. Contractor s Report vi

9 Introduction Background The California Department of Transportation (Caltrans) and hundreds of local agencies throughout the state of California use pavement management systems (PMS) to determine which strategies to use in their pavement maintenance and rehabilitation projects. Most PMSs use performance models to predict future asphalt concrete pavement condition and lifespan of the various treatments. PMS performance models were first developed in the 1980 s and much has changed in pavement technology since then, such as the use of new materials, which provide improved performance. It is now well-documented that rubberized hot mix asphalt (RAC) pavements, specifically asphalt rubber and terminal blend modified mixes, have better performance than conventional hot mix asphalt AC) in terms of resistance to reflective cracking, durability, skid-resistance, and lower noise. Based on these benefits, it is expected that the performance models predicting RAC overlay distresses will be significantly different than those of AC overlays; however, RAC performance models do not currently exist for most pavement management systems. Overall Project Objectives The objectives of this project are to: Develop RAC performance models that can be used in PM systems, and Demonstrate the added benefit in the performance of selected CalRecycle grant projects. Show monetary saving with RAC usage. The information shared through the reporting of this work will increase the use of RACRAC by documenting its superior performance via models that can be used in pavement management systems. These models will assist Caltrans, other state agencies, and local governments that utilize PMS in making their pavement Maintenance and Rehabilitation (M&R), materials and strategy selections in showing RAC to be the preferred strategy. The work orders issues for this project consist of the following: Work order 1. Collect information from local agencies and develop performance curves such as shown in Figure 1. Work order 2. Collect information from Caltrans and local agencies, and develop performance curves for asphalt rubber Work order 3. Collect materials from California projects and perform fatigue tests on conventional and asphalt rubber hot mix to quantity the increase in cracking resistance Work order 4. Disseminate the information through reports, papers, and presentations Contractor s Report 1

10 Pavement Condition Index This report is for the first part of the Work Order 1. Purpose of Report The purpose of this report was to review performance models used by various agencies in order to select a model that works for RAC. To our knowledge, none of the agencies have developed a mathematical performance model specifically for RAC, as shown in Figure Year Conventional Mix Mix RAC Rubber Mix Modified Mix Figure 1 Example of a Pavement Model Curve The long-term performance modeling and development of performance curves are needed to predict future performance and to perform life cycle cost analysis (LCCA). These analytical tools are not currently available so this work order was designed to: 1. Review local government Rubberized Asphalt Concrete (RAC) project data to correlate field performance to assist in the development of performance models for local agencies. 2. Develop performance models to predict the long-term performance of rubberized asphalt materials through field studies. Contractor s Report 2

11 Literature Review As a part of this report, a detailed literature review was conducted to identify performance models for possible use in this study. The research found a total of eleven agencies that had models that might be used for developing a RAC model. They were selected based on the flexibility to modify their models for our purposes, or inclusion of age in their parameters. These models come from a list of over twenty different pavement management systems and are currently used by various state departments of transportation. The in-practice methods were used to evaluate pavement performance models from the Arizona, Colorado, California, Minnesota, Mississippi, New Jersey, Ohio, South Dakota, Texas, and Washington Departments of Transportation. An important side note is that almost all of the state agencies use the International Roughness Index (IRI) as a pavement performance quality indicator. IRI is a roughness quality in units of in/mile or mm/m. IRI is a roughness quality that relates overall ride quality and surface condition. If the equations do not specifically have IRI in them, they have developed a variable in the equation that is substituted for it. In addition, all of these pavement performance systems have some similarities. IRI is the one unifying quality that is comparable among all the states researched. Pavement condition in terms of the severity level, weight factors, and other distresses types can be more difficult to compare. This review presents a summary of the models used by the various states for their pavement management systems. Colorado Remaining Service Life (RSL) Concept Colorado DOT developed distress values from combining quality indicators for an Individual Index value. The Individual Index is used to quantify the road quality. ( ) (1) where, = an index used to quantify road quality, i.e. IRI = Average distress value = Statewide maximum and minimum distress values for a road quality = Remaining Service Life= Threshold Age= Age in which an individual index is equal to 50 The Remaining Service Life Concept used by the Colorado Department of Transportation (CDOT) allows them to predict the time in which the pavement deteriorates into a poor condition. The CDOT system equates an Individual Index value of 50 as a failed surface. The age in which the individual index is 50 is the Threshold Age. This is on a scale of 1 to 100. Colorado has statewide minimum and maximum distresses to use in this equation as well as categories of Remaining Service Life (RSL). An RSL greater than 10 years is rated as good, 6-10 years is fair, Contractor s Report 3

12 and less than 6 is rated as poor service life. The road quality indicators evaluated by CDOT are IRI, rutting, fatigue cracking, transverse cracking, and longitudinal cracking. Minnesota Pavement Quality Index and Ride Quality Index, (3, 4) Minnesota DOT developed a pavement model which includes Pavement Quality Index (PQI) and a Ride Quality Index (RQI) rating system as shown below. This is a combined rating system that includes Pavement Serviceability Rating (PSR), a function of ride quality, and Surface Rating, which is a function of pavement distresses including cracking and rutting. Minnesota Pavement Quality Index (2) (3) (4) (5) where, = Pavement Serviceability Rating = International Roughness Index = Weighting Factor = Total Weighted Distress = Surface Rating = Pavement Quality Index Minnesota DOT uses a multi-step process to evaluate pavement conditions for their Pavement Quality Index (PQI). As stated above, it combines the Pavement Serviceability Rating (PSR), which uses IRI, and a subjective Surface Rating (SR) system. The SR system has a table that shows the different weighting based on type and severity of distress. After a summation of various weighted distresses, the Total Weighted Distress (TWD) can be converted into a Surface Rating (SR) score as shown in Table 1. These scores are on a scale of 0.0 to 4.0, in which the lower the percentage of TWD gives a higher SR. The PSR is on a rating scale of 0.0 to 5.0, in which 5.0 means perfect conditions. The PQI is on a scale from 0.0 to 4.5 in which 4.5 means no defects. Contractor s Report 4

13 Table 1 shows a conversion table for TWD to SR values while Table 2 provides a summary of the weighting factors used by MNDOT for each of the distress types. Table 1 Conversion Table of TWD to SR Total Weighted % SR Contractor s Report 5

14 Table 2 Bituminous Pavement SR Weighting Factors Distress Type Severity Weighting Factor Transverse Cracking Low 0.01 Medium 0.10 High 0.20 Longitudinal Cracking Low 0.02 Medium 0.03 High 0.04 Longitudinal Joint Deterioration Low 0.02 Medium 0.03 High 0.04 Multiple (Block) Cracking Alligator Cracking Rutting Raveling & Weathering Patching Minnesota Ride Quality Index (3) A secondary measure Minnesota uses to consider the overall quality of the pavement is Ride Quality Index (RQI). RQI is a function of Age and is shown in Equation 6. (6) (7) where, = Initial Ride Quality = Coefficients In order to continually improve their database, Minnesota DOT updates the regression coefficients by adding RQI data collected by inspectors each year. Figures 2 and 3 show examples of deterioration of RQI. The constants used for Example 1 are: a = , b = , c = 0.611, and RQI i = 3.8. This yields a deterioration of 3.0 from 2014 to However a treatment should be done before the RQI becomes 0.1, an unacceptable Ride Quality. For Example 2 the constants are a=38.238, b= and c=1.047 with an initial RQI = 3.8 yields. As you compare the effects of the constants you can see that they have a major effect on the life of the pavement in terms of ride quality. Contractor s Report 6

15 Figure 2 RQI Example 1 from Minnesota DOT Contractor s Report 7

16 Figure 3 RQI Example 2 from Minnesota DOT Mississippi Pavement Condition Rating System Mississippi DOT developed a model for Pavement Condition Rating (PCR) as a function of ride quality and various distress types. They have not developed models for asphalt rubber. ( ) ( ) (8) where, = Pavement Condition Rating = Measured IRI mm/m = Maximum possible deduct points due to distress ( =205 for flexible pavement) = Actual total of deduct points = Constants Contractor s Report 8

17 For AC pavements: = , = , = 1.55, = 205 The above equation shows the calculation of PCR at a specific inspection time. Mississippi uses a combination of an IRI with distress deduct points from inspection. The constants and weigh the PCR so that distress deductions have a greater impact the rating system. In addition, there are different constants for different types of pavements. There are constants and maximum deduction points for PCC, AC, and composite pavements. It also should be noted that 12 mm/m (760 in/ mi) is the maximum acceptable IRI value for this equation. The other distresses considered in this model for asphalt pavement are: Alligator cracking Block cracking Edge raveling Longitudinal fatigue cracking Reflection cracking Transverse cracking Patch/Patch deterioration Potholes Rutting Shoving Bleeding Polished Aggregate Raveling within traveled way Lane to shoulder drop-off Water bleeding & pumping Oregon Department of Transportation (ODOT) strongly considered adopting this Mississippi model for their state. Mississippi also has an additional set of equations that shows the deterioration of the pavement over time. They are categorized based on the type of pavement and by treatment. Original Flexible Pavement, Pavement Condition Rating (PCR) Model The approach used to predict PCR is shown in the equation below; ( ) (9) where, = Pavement Condition Rating = The number of year since the last rehabilitation or construction activity = Cumulative Equivalent Single Axel Load, in million = Modified Structural Number Contractor s Report 9

18 This PCR model is modified to take into consideration asphalt pavement in its first iteration only. New flexible pavements can use this model. Unlike the previous equation, this equation is used show the deterioration over time not just the condition of the pavement at the point of inspection. Overlaid Flexible Pavement, Pavement Condition Rating (PCR) Model ( ) (10) where, = Pavement Condition Rating = The number of year since the last rehabilitation or construction activity = Cumulative Equivalent Single Axel Load, in millions = Modified Structural Number = Thickness of last overlay, in inches = 10, Categorical Variable for AC Overlay This PCR model is modified to account for the difference in deterioration based on the fact that it is a treatment overlaid on a deteriorated pavement. There are additional models for composite and other types of PCC pavements. Mississippi DOT is currently in the process of updating their performance model and pavement management system in order to better suit their requirements. MTC StreetSaver Pavement Condition Index, (6, 7) The model used by MTC is shown below. It is used for asphalt concrete (AC) and Portland Concrete Cement (PCC) pavement types. It is adapted to each pavement type through the change in modifier and constants; however, it does not have constants for asphalt rubber pavements at the present time. ( ( )) (11) where =Projected Pavement Condition Index Value = Years since the last major rehabilitation/reconstruction activity = A projection modifier, initially set to 0 = Regression constant that controls the age in which the curve is asymptomatic Contractor s Report 10

19 Pavement Condition Index (PCI) = Regression constant that controls how sharply the PCI family curve bends = Regression constant that controls the age at which the inflection point in the PCI family curve = A projection modifier, initially set to 1 The Metropolitan Transportation Commission uses the Pavement Management Program called StreetSaver TM. StreetSaver has many components to it, in both pavement evaluation and cost effectiveness. For its pavement evaluation component an s-curve is used to more accurately portray the pavement deterioration over time. The regression constants, α, β, and ρ, used to control the graphs are dependent upon functional class and type of pavement. The projection modifiers, χ and Shift, are dependent on different maintenance treatments and are customized on a street section level. Values for χ and Shift start at 0 and 1 respectively but can be changed posttreatment. Typically, a new family curve is created when each major rehabilitation or reconstruction treatment is done to a section. However, as a method of comparison within StreetSaver two curves are projected, a family curve and an adjusted curve. This adjusted curve is changed due to inspections and maintenance treatments, which make the curve more accurately reflecting the performance and life of the pavement. Examples of performance curves are discussed in the next section of this report. Examples of the current typical curves for MTC StreetSaver are shown in Figure 4. These do not account for the difference between Asphalt Rubber treatments and conventional asphalt treatments Age (Years) AC Arterial AC Collector AC Residential AC/AC Arterial AC/AC Collector AC/AC Residential Figure 4 Typical Performance Curves for MTC StreetSaver Contractor s Report 11

20 New Jersey Stress Index, (8, 9) New Jersey DOT developed pavement deterioration curves for asphalt pavements using the following relationship: ( ) (12) (13) (14) where, = A distress weight for each type of distress, j = Surface Distress Index = Normalized IRI = International Roughness Index Equations evaluate the current state of pavement while they do not take into account the deterioration of the pavement. New Jersey DOT uses two separate systems to evaluate pavement in their state. They use an International Roughness Index (IRI) and Surface Distress Index (SDI). The IRI value that is typically reported in mm/m or in/mile is normalized for comparison purposes and use in the Final Pavement Rating. These systems are both on a scale of 0 to 5 where 5 means perfect conditions and 0 is deficient or very poor conditions. The system uses a series of terminal and trigger values to evaluate the accuracy of their system and for use in management. The terminal values for IRI Norm 1.68, 2.03, and 2.05 for asphalt, concrete, and composite pavement respectively. The trigger values used at NJDOT are categorized by the use of the road. This means that there are different trigger values for different uses. The trigger values for Interstate routes are an IRI Norm and SDI 3.5 and a Rut Depth of 0.5 in or greater. The trigger values for other statewide routes are IRI Norm and SDI 3.0 and a Rut Depth of 0.5 in or greater. There is also a Final Pavement Rating (FPR), which is used in benefit analysis that is evaluated using these three steps: 1. If both IRI Norm and SDI are > 2.51 then they are weighted at 50% each 2. If IRI Norm and/or SDI are < 2.00 then the lower of the two ratings is weighted at 100% 3. If the lowest value of IRI Norm and/ or SDI is 2.00 and 2.50 then the lower number is weighted at 75% and the other value is weighted at 25% Contractor s Report 12

21 Shown below in Figure 5 is a curve comparison of the default distress index model, measure points of inspection, and a site-specific distress model. The site-specific model takes into account traffic index, AADT, age, and type of pavement. It is unknown if this model is applicable to asphalt rubber. Figure 5 New Jersey Pavement Distress Model Comparison (10) Through contact with a representative from NJDOT, CP2C was able to gather the pavement deterioration models and constants for New Jersey. The equations for IRI and SDI are as follows: (15) where, ( ) = Surface Distress Index = International Roughness Index = Age of pavement in years since last treatment = Constants A sigmoidal (i.e. S-shaped) form is for New Jersey which is also present with in most state transportation agency This S-shape model form has a greater degree of flexibility in describing the deterioration of a section. Currently New Jersey DOT uses the constants: for IRI A= , B= , and C= For the SDI model the constants used are: A= , B = , and C= These are subject to change from year to year given the addition of new information. (16) Contractor s Report 13

22 Ohio Pavement Quality Index, (11, 12) Ohio DOT developed pavement deterioration curves for asphalt pavements as follows. This pavement deterioration curve does not take into account the change in quality of pavement over time it is only an analysis of the present state of the pavement. (17) (18) where, =Pavement Quality Index = Pavement Condition Rating = International Roughness Index = Coefficients = Weight factors Table 3 summarizes the coefficients used in Ohio pavement index and its unit adaptations. Table 3 Coefficients to Ohio Pavement Quality Index Equation Coefficient Priority System w/ IRI in metric (mm/m) Priority System w/ IRI in US Customary (in/mi) General System a b Figure 6 demonstrates how IRI deducts from PQI it graphically represents the a(iri) b portion of the PQI. The priority and general lines are the different analysis methods used by Ohio and the related coefficients are seen in Table 3. Contractor s Report 14

23 Figure 6 IRI Component of PQI Table 4 summarizes the weight factors of the various distress types it also includes a description of the distress severity and extent types. These factors combine to become the PCR portion of the PQI Equations. Contractor s Report 15

24 Table 4 Severity and Extent Weight factors for Ohio PCR Distress Distress Weight Severity Weight (SW) Extent Weight Raveling Low Medium High Occasional Frequent Extensive Slight Loss of Sand Bleeding Not Rated Open Texture 1 Rough or Pitted 0.5 <20% % 1 >50% Bit. & Agg. Visible 1 Black Surface 0.6 <10% % 1 >30% Patching < 1ft < 1 yd 2 1 > yd <10/mile /mile 1 >20/mile Debonding Depth < 1" Area < 1 yd D<1", A> 1yd 2 D>1", A< 1yd 2 1 Depth > 1" <5/mile /mile 1 >10/mile Area > 1 yd Crack Seal Deficiency 5 Severity Level Not Considered (SW=1) 0.5 <50% 0.8 >50% 1 No Sealant Rutting /8" - 3/8" 0.7 3/8"-3/4" 1 >3/4" 0.6 <20% % 1 >50% Potholes Depth < 1" Area < 1 yd D<1", A> 1yd 2 D>1", A< 1yd 2 1 Depth > 1" <20% /mile 1 >10/mile Area > 1 yd Wheel Track Cracking Single/Multi. Cracks < 1/4" 0.7 Multi. Cracks > 1/4" 1 Alligator > 1/4" w/ spall 0.5 <20% % 1 >50% Block & Transverse Cracking 'x6' or Trans. Crack 0.7 6'x6' to 3'x3' 1 < 3'x3' 0.5 <20% % 1 >50% Longitudinal Cracking Single, <1/4", No spall 0.7 Single/multi. 1/4-1" some spall 1 Multi >1" w/ spall 0.5 <50 per 100' per 100' 1 >150 per 100' Edge Cracking Tight, < 1/4" 0.7 > 1/4" some spall 1 > 1/4" moderate spall 0.5 <20% % 1 >50% Thermal Cracking < 1/4" 0.7 1/4" - 1" 1 >1" 0.5 CS > CS CS < Contractor s Report

25 Ohio Department of Transportation (ODOT) in their report Development of a Composite Pavement Performance Index recommended a combination of linear and power functions where the power function is IRI and the linear portion is concerning PCR. There was an alternative that was considered to be more flexible to suit the needs of ODOT but not entirely practical. This flexible and more accurate PQI model was a six-parameter polynomial. This model however needed a great deal more data than other models to be accurate in the field. The end result of the research done for ODOT is to use the combined function but narrow it down with two different systems depending on the priority level of the road. The constants in the equation were solved in order to fit a certain fixed criteria, a set intercept and an acceptable minimum PQI. The statistical analysis was done using TableCurve 3D Software in order to analyze the three dimensional data. PaveM Pavement Management System (13) Caltrans has developed interim models for flexible pavements using the following models. Please note they are predicting individual distress, not a combined index. Models were developed for ride, cracking, and mean depth profile (an indicator of raveling). The model used for Average IRI is: (19) where = Average International Roughness Index (IRI) in units of in/mile = Initial IRI value = Rate of change parameter = Rate of change parameter = Age of pavement in years since last treatment One of the performance variables that Pave-M uses to evaluate pavements is the international roughness index (IRI). The IRI is a measure of ride quality that is internationally recognized and for the purposes of Pave-M the range of acceptable values is inches per mile. The model used for Wheel Path Cracking Value (WPCV) is: (20) where, = age of pavement in years since last treatment = Constants The Wheelpath Cracking Value represents the percentage of wheelpaths that have cracking or have been patched. This is calculated using both right and left wheelpaths in the direction of travel. The cracking value is comprised of Wheelpath Advanced Cracking and patching. An Contractor s Report 17

26 advanced crack is a crack that has a length ratio greater than 1.6. This ratio is the feet of cracking per foot of the wheelpath in the direction of travel. The model used for Flexible Total Cracking Value (FTCV) where, ( ) = age of pavement in years since last treatment = Constants (21) Flexible Total Cracking includes all cracks in the pavement both, inside and outside the wheelpath. Cracks that are wider than a quarter of an inch are counted as double the crack because of their increased ability to let water into the pavement structure. The last model for Mean Profile Depth (MPD) is (22) where, = age of pavement in years since last treatment = Constants The Mean Profile Depth is a measure of how far stones protrude from the surface of the pavement. This can be done with a high-speed laser over a wheel path of a segment of roadway. If a profile depth is too low, this can contribute to a lower skid resistance. If a profile depth is too high, this can indicate raveling. The table below gives approximate MPD values for different conditions in flexible pavements as a frame of reference. Table 5 Example MPD Values MPD (mm) Conditions HMA with very low texture New HMA New RAC-G New RAC-O yr.-old RAC-O; old HMA yr.-old RAC-O; 10 yr. old OGAC Heavy raveling, finer chip seal Very heavy raveling, coarser chip seals Contractor s Report 18

27 South Dakota Surface Condition Index, (14, 15) South Dakota Department of Transportation (SDDOT) uses a composite index, which has been adapted for PCC pavements and AC pavements. This Surface Condition index only includes the present condition of pavement and not a performance curve. The adaption of this index system to a suitable performance curve is being researched by the SDDOT. (23) where, = Composite Index ( lowest individual index and 0.00) = Mean of all contributing individual indexes σ = Standard Deviation of the mean of the contributing individual indexes South Dakota Department of Transportation (SDDOT) uses the Surface Condition Index (SCI), which uses a 0-5 scale. In order to create this Surface condition index it is compiled of multiple indexes. The individual distress indexes for flexible pavement are fatigue cracking and patching, transverse cracking, block cracking, rutting, and roughness (IRI). The tables and figures below are the deduction points that SCI uses. Table 6 Deduct Points from Fatigue Cracking and Patching Fatigue cracking and Patching SEVERITY EXTENT Low Medium High Extreme Low Medium High Table 7 Deduct Points from Transverse Cracking Transverse Cracking SEVERITY EXTENT Low Medium High Low Medium High Contractor s Report 19

28 Table 8 Deduct Points from Block Cracking Block Cracking SEVERITY EXTENT Low Medium High Low Medium High Figure 7 Deduct Points from Rutting Contractor s Report 20

29 Figure 8 Deduct Points from IRI Stantec Pavement Management System, RoadMatrix TM, (16) Stantec RoadMatrix is a pavement management system that uses a combination of indexes to evaluate the pavement and plan maintenance for municipalities. The RoadMatrix program is used by various cities and counties throughout the United States and Canada. The RoadMatrix program is customizable and uses the same model for each municipality; although some factors are modified to better fit each municipalities needs. An example of a factor that could be modified is the Pavement Quality Index/Pavement Distress Index constitutes a pavement in need of rehabilitation or reconstruction. Depending on the availability of funds and total number of roads this may need to be modified in order to realistically maintain each municipality s pavement network. The following describes the models used by AZDOT, The City of Sacramento, and the Count of Los Angeles Arizona Pavement Distress Index (PDI), (17, 18) Stantec Consulting Corp. developed for Arizona Department of Transportation a model to predict PDI, which is a function of cracking, rutting, flushing, and patching. Ride is not considered in the calculation. They use a modified version of the Stantec Road Matrix Software. (24) (25) Contractor s Report 21

30 (26) (27) (28) where, = Cracking Index = Rutting Index = Flushing Index = Patching Index Arizona Department of Transportation (ADOT) used a combination of linear and power functions to develop their curve. This Pavement Distress Index puts emphasis on four categories of distresses: cracking, patching, flushing, and rutting in order to quantify the distress quality cracking and patching are measured as a percentage of the area, where 0% represents perfect conditions. Rutting is a measured in inches, in which the maximum acceptable rut depth for use in the equation is 2.0 in. If the rutting of a pavement exceeds 2.0 in then the value to be used for the rutting distance is automatically 100%. Flushing is evaluated on a scale between 0 and 5, where 5 represents perfect conditions. The individual distresses are normalized on an increasing scale of 0.0 to The Pavement Distress Index (PDI), itself is on an increasing scale from 0.0 to 5.0. The equation weights rutting and cracking more heavily than flushing or patching. ADOT also developed a Pavement Serviceability Rating for Maintenance and Rehabilitation Treatments ( ) (29) where, = Initial Conditions = The number of year since the last rehabilitation or construction activity = Coefficients that define the model shape = Coefficients that define the model shape = Coefficients that define the model shape A sigmoidal (i.e. S-shaped) form is used within the Highway Pavement Management Application (HPMA) for modeling the pavement performance. This model form has a greater degree of flexibility in describing the deterioration of a section. The flexibility of the sigmoid allows the models produced to be concave, convex, S-shaped, or almost linear. This has historically Contractor s Report 22

31 produced curves that sufficiently fit the data and describe performance. The purpose of this PSR is to demonstrate the change in PDI over time due to the fact that PDI is a present analysis tool. It can only be done on a case-by-case basis. Currently Arizona DOT uses the constants: A=15, B=20, and C=1.1.These are subject to change from year to year given the addition of new information as well as subject to change when adapted from one state to another. City of Sacramento Pavement Quality Index, (19) The City of Sacramento evaluates their pavement using the Stantec Model much like the ADOT. The key difference is the inclusion of different indexes. The City of Sacramento uses the Pavement Quality Index. The indexes that combine to get this overall Quality are the Surface Distress Index (SDI), Riding Comfort Index (RCI), and Structural Adequacy Index (SAI). Each index is on a scale from 0 to 100. The City of Sacramento s SDI is equivalent to Arizona s PDI, although SDI accounts for many more types of distresses. (30) (31) (32) (33) Although the database records previous points and analyzes present information, it does not use the same sort of future prediction of pavement quality like the one ADOT does with their Pavement Serviceability Rating (PSR). Los Angeles County Pavement Condition Index (20) Los Angeles County is modified to suit their needs although little detail is known on the exact components that combine to create their Pavement Condition Index (PCI). The setup of the PCI is most similar to Arizona s PCI. Figure 9 below shows the various deterioration curves that are dependent on that thickness of pavement, traffic level, strength of subgrade. Contractor s Report 23

32 Figure 9 Pavement Condition Index deterioration curves for Los Angeles County. Texas DOT Deterioration Condition score, (21, 22) Texas Department of Transportation (TxDOT) uses an S-curve for their pavement performance curve. They use additional equation to encompass the future life and added life from pavement treatments. This model is also not adapted to Asphalt Rubber. { ( ) (34) (35) (36) ( ) (37) Contractor s Report 24

33 where, =Utility value (ranges from 0.0 to 1.0, 1.0 being most useful) =Maximum loss factor =Slope factor =Prolongation factor =Utility value of Ride Quality (ranges from 0.0 to 1.0) =Density of each distress time on pavement =Distress Score (scale of 0 to 100) =Condition Score (scale of 0 to 100) Texas Department of Transportation (TxDOT) uses a multistep process to come up with a Condition Score (CS) this is a value from that indicates overall condition of the pavement at the time of the evaluation. The Distress Scores are individual scores based on different types of distresses. The factors α, β, and ρ control the location of the inflection point and the slope of the curve on the given point of the utility curve. TxDOT uses deep rutting, shallow rutting, patching, failures, block cracking, alligator cracking, longitudinal cracking and transverse cracking for utility values of asphalt concrete pavement. Recently, Texas DOT has studied the pavement condition of various Texas metropolitan areas in order to have a more accurate curve. This was done by modifying the Utility Score values, which are related to surface distresses, and taking into account the way that speed and traffic affect the deterioration of pavement. Deterioration Condition Score for Future Predictions ( ) (38) [ ( ) ] (39) where, = prolongation factor = Age since last construction = Maximum loss factor =Condition Score at year i = Condition Score right after treatment =Density of each distress time on pavement = The slope factor = Traffic loading coefficient = Climate region coefficient = Subgrade coefficient & curve parameters Contractor s Report 25

34 Figure 10 demonstrates the general shape of the Pavement Performance Prediction Model. This model is the one proposed for use by TxDOT. The difference from the previous model is the inclusion of the χ, ε, σ, and ρ factors in the numerator of the exponent. These factors have cater to each individual environment (i.e. heavy traffic desert climate or low traffic Class 2 Sub-base case) Figure 10 The General Shape of the Pavement Performance Prediction Model used by TxDOT M&R Treatment Category Table 9 Curve parameter factors, β pq & ρ pq Low Traffic β pq Medium Traffic High Traffic Low Traffic ρ pq Medium Traffic High Traffic Preventative Maintenance Light Rehabilitation Medium Rehabilitation High Rehabilitation The above equations are adapted to include age into the equation for future predictions rather than the present condition. These equations are heavily dependent on the type of treatment and the traffic conditions in the area of the treatment. With this new addition to the model TxDOT is able Contractor s Report 26

35 to predict the year in which they need to treat a road. This year occurs when the CS becomes lower than 70. The table above shows the original figures for β pq & ρ pq and while the effect on the model by these factors are shown in the Figure 11 below. These figures are from the original model prior to the proposed values from Figure 11 Texas Pavement models using Original Coefficients. Added Life of Maintenance Treatments [ ( )] (40) where, =Added life of pavement due to treatment q = 70, Minimum Condition Score (meaning anything less than 70 requires and M&R treatment) = Increase in Condition Score due to the M&R Treatment for each traffic level =Condition Score before treatment The Added Life equation allows TxDOT to portray the extension of the service life of a pavement due to a maintenance treatment. This added life model is often used in combination with a Cost to Contractor s Report 27

36 Added Life Equation (CAL q ) by the TxDOT in order to make decisions concerning cost benefits of alternative pavement treatments. Washington Pavement Structural Condition Model, (23, 24) Washington Department of Transportation (WsDOT) uses a combination of a Power and Linear function much like Ohio and Mississippi. The PSC model is a multi-step process that is not adapted to describe the change in pavement structural condition over time. (41) (42) (43) (44) (45) (46) where, = Alligator (Fatigue) Cracking = % of area of the wheel path that has hairline cracks = % of area of the wheel path that is spalled cracks = % of area of the wheel path that is spalled cracks and pumping of fines = Longitudinal Cracking = % of length with cracking less than ¼ in wide = % of length with cracking more than ¼ in wide = % of length with cracking that is spalled = Transverse Cracking = Number of cracks that are less than ¼ in wide = Number of cracks that are more than ¼ in wide = Number of cracks that are spalled = Patching = BST Patching = Blade (cold mix) patching Contractor s Report 28

37 = Full depth patching =Pavement Structural Condition =Equivalent Cracking All Structural Cracking (ACA) (47) Wide Structural Cracking (ACW) ( ) (48) Transverse Thermal Cracking (49) Washington Department of Transportation (WSDOT) uses the pavement structural condition (PSC) to evaluate flexible pavement in the state. For all evaluations it is assumed that a 100 ft. lane that is representative of the road segment is selected. It is important to note that in this 100 ft lane there are two wheel paths so that there is a total of 200 ft. of wheel path, this is important when conducting a survey where there are alligator cracks. Other than alligator cracks the PSR is characterized by longitudinal cracks, transverse cracks and patching. In addition to this PSR system that WS DOT uses in order to be compliant with the Highway Development and Management System (HDM-4) there are three equations that convert the PSR system to HDM-4 system. The three qualities that the HDM-4 system looks at are all structural cracking, wide structural cracking, and transverse thermal cracking. The conversion to HDM-4 is necessary for more analysis on a global level, as it is program developed, maintained and update through the World Bank Organization. Summary of Models Every State or Local Agency has a different set of requirements for their pavement modeling system; as such there are many models that the local agencies use to convey the same information. Many of the contrast between models include: how they classify distress, present analysis versus age analysis, the inclusion of Functional Classes, inclusion of pavement structural analysis, inclusion of IRI, and linear vs. nonlinear modeling. Every model has its advantages and disadvantages, but some models could be combined to create a better, more comprehensive model. Contractor s Report 29

38 Typical Performance Curves for Asphalt Rubber MTC currently has pavement deterioration curves for conventional asphalt pavement (AC) and Portland Concrete Cement (PCC) pavement in their respective functional classes. Due to the rise in usage of asphalt rubber pavements in California, there is a need to develop pavement deterioration curves for this product. Although asphalt rubber pavements are not a new product in California, they recently became more prevalent because of the implementation of AB 338 which requires State and local agencies to increase their rubber usage for roadways. Information has been gathered concerning the various rubber projects in the California Bay Area. These projects were chosen knowing the agencies received grants through the CalRecycle Rubberized Pavement Grant Program. However, there is still a great deal of information missing. This is mainly due to lack of information since the pavement treatments have been placed. Judging from the list of rubber projects provided to us through CalRecycle s list of grants, the first rubber pavement grant project entered into the database started in The majority of projects were placed in 2009 and 2010, which makes it difficult to develop a pavement deterioration curve. As such, we will need to identify projects placed 10 to 20 years ago, which have not had major maintenance treatments. This will be described in more detail in the companion project for WO1 titled, Performance Modeling of Rubberized Hot Mix Asphalt for California Local Agencies. Most of the older projects placed by local agencies are in Southern California. There is very little detail in StreetSaver concerning many of the known CalRecycle projects. Not much information about the treatments is known other than the name and date that are listed in StreetSaver. Because StreetSaver has preset PCI changes post-treatment, not all modifications in treatment are accurately portrayed. During the literature review process, the authors made initial attempt to develop pavement deterioration curves using typical data from MTC. These deterioration curves show the assumed life cycle of the pavement treatment and its condition of over time. Contractor s Report 30

39 PCI Value Figure 12 is a good example of the data trend of rubber treatment. It is slightly above the predicted performance curve of but without any major dips or change in slope. It however has very few data points, not enough to draw a definite conclusion PCI Curve of Arterial Roads with RAC Time (Years) MECARS 15 MECARS 18 MECARS 30 LELAND Performance Curve of AC/AC Arterial Figure 12 Pavement Deterioration curve of a RAC Overlay of an unknown depth on a Road with an Arterial Functional Class Contractor s Report 31

40 PCI Value Figure 13 is a graph that displays some of the lack of data concerning the exact treatment done to pavement. As is seen above there is a dip in PCI at year 3 followed by an abrupt return to a PCI value of 100. However not all of the streets follow this pattern. This could be due to many different reasons such as an additional unrecorded maintenance treatment or a poor quality inspection. PCI Curve of Collector Roads with RHMA Overlay Time (Years) THIRD 21 GRAND 35 GRAND 40 WILLOW 40 MCGLNC 10 MCGLNC 20 MCGLNC 30 Performance Curve of AC/AC Collector Performance Curve of O-AC/AC Collector Figure 13 Pavement Deterioration curve of a RAC Overlay of an unknown depth on a Road with a Collector Functional Class Contractor s Report 32