Evaluation of PMS Recommendations for Flexible Pavement Rehabilitation Using RWD Deflection Data



Similar documents
Impacts of Increased Loading Due to Heavy Construction Traffic on Thin Pavements

CHAPTER 2 PAVEMENT MANAGEMENT SYSTEM

Expected Service Life and Performance Characteristics of HMA Pavements in LTPP

Pavement Management Systems

Rehabilitation by cracking and seating of concrete pavement optimized by FWD analysis

Evaluation of Trench Backfill at Highway Cross-Drain Pipes

Chapter 7: Pavement Rehabilitation 7-1 Asphalt Pavement Overlays 7-1 Surface Preparation Methods 7-2 Concrete Pavement Preparation 7-3 Recycling

5. Report Date GUIDELINES FOR EVALUATING ROUTINE OVERWEIGHT TRUCK ROUTES. 8. Performing Organization Report No. Emmanuel G. Fernando and Jeong-Ho Oh

Pavement Management Implementation: Success Stories for Maryland Counties

Date: 10/10/2014 Initiated by: AAS-100

Minimizing Reflective Cracking With Applications of the Rolling Dynamic Deflectometer and Overlay Tester

Thin Whitetopping Application at Williamsburg Regional Airport and Other Thin Whitetopping Airport Applications.

Pavement Rehabilitation Using Hot Mix Asphalt. - National Perspective -

Final Report Evaluation of I-Pave Low Volume Road Design Software

Asphalt Pavement Association Of Michigan Selecting the Right Mix Atrium Drive, Suite 202 Okemos, MI

Research Article Determination of Pavement Rehabilitation Activities through a Permutation Algorithm

SELECTING REHABILITATION STRATEGIES FOR FLEXIBLE PAVEMENTS IN TEXAS. Andrew J. Wimsatt, Ph.D., P.E. Fort Worth District Pavement Engineer

AN ANALYSIS PROCEDURE FOR LOAD-ZONING PAVEMENTS

The AASHO Road Test site (which eventually became part of I-80) at Ottawa, Illinois, was typical of northern climates (see Table 1).

This is the accepted version of this journal article:

Evaluation of Long-Lasting Perpetual Asphalt Pavement Using Life-Cycle Cost Analysis. Arnar Agustsson Sulton Azamov

CID STREET INFRASTRUCTURE ASSESSMENT

REGIONAL TRANSPORTATION COMMISSION OF WASHOE COUNTY

DRAFT MAINTAINING HOUSTON S STREETS REPAIR, REHABILITATION, RECONSTRUCTION. Using the Full Range of Tools for a Challenging Job

PERFORMANCE TESTING OF BITUMINOUS MIXES USING FALLING WEIGHT DEFLECTOMETER

RANDOM/GRID BLOCK CRACKING EXTREME SEVERITY. Transverse Cracking- Extreme Severity. Alligator Cracking- High. Severity

Transportation Infrastructure Asset Management

CITY COUNCIL AGENDA ITEM CITY OF SHORELINE, WASHINGTON

Application of the Endurance Limit Premise in Mechanistic-Empirical Based Pavement Design Procedures

Pavement Design. Guest Lecturer Dr. Sirous Alavi, P.E. SIERRA TRANSPORTATION Terminal Way, Suite 125 Reno, Nevada 89502

Road Asset Management

Decision Matrix for Pavement Preservation. Mike Mamlouk Arizona State University

Evaluation of Cement Treated Base Courses

Improving Rehabilitation Selection for Pavement Life Cycle Cost Analysis

LIFE-CYCLE COST COMPARISON FOR MUNICIPAL ROAD PAVEMENTS

APPENDIX B. I. Background Information

Advancements in GPR for a Sustainable Tomorrow

Road Asset Management Research Program In Finland

Reformulated Pavement Remaining Service Life Framework

Road Asset Management: Evolution and Trends

Rehabilitation Strategies for Bonded Concrete Overlays of Asphalt Pavements

AASHTO Subcomittee on Materials Biloxi, Mississippi August 4-10, 2012 Chris Abadie, P.E.

Nevada DOT Cold In-Place Recycling Federal Highway Administration National Review Close out meeting, August 25, 2005

PAVEMENT STRUCTURE DESIGN GUIDELINES

DEVELOPMENT OF A NATIONAL ROAD ASSET MANAGEMENT SYSTEM FOR THE KINGDOM OF SAUDI ARABIA. Hazim M Abdulwahid Abdullah Al Kahtani Mustafa Abdulaziz

RoadMatrix V3.1 has incorporated many engineering enhancements including the following:

CHAPTER 2 LCCA APPROACHES

Cost Considerations of In-Place Recycling as a Pavement Rehabilitation Alternative Brian Diefenderfer, PhD, PE Alex Apeagyei, PhD, PE

Cornell Asset Management Program - Roads & Streets (CAMP-RS) 2014

Using Accelerated Pavement Testing to Evaluate Permeable Interlocking Concrete Pavement Performance

Comparison of the Humboldt GeoGauge With In-Place Quasi-Static Plate Load Tests. Charles R. Nelson, P.E. Mike Sondag

Road maintenance policy based on an expert asset management system. - Concept and case study - Author Philippe LEPERT IFSTTAR Nantes

Rehabilitation for Concrete Pavements

A Pavement Management System for County Roads

Software Development (PAKPAVE) for Flexible Pavement Design

TAC Primer Pavement Asset Design and Management

Pavement Management Program

CHAPTERS PAVEMENT ENGINEERING CHAPTER 600 GENERAL ASPECTS

Transportation Asset Management Program 2009 Federal Aid PASER Road Survey. August 2009

Using Cloud-Computing to Promote Asset Management Best Practices A Ministry of Transportation Ontario Case Study

MODULE 3. Upon completion of this module the participants will be able to:

SECTION PAVEMENT DESIGN

FLH Asset Management

TAMS 2.1 User s Manual. Utah LTAP Center. Contact: Utah LTAP 4111 Old Main Hill Logan, UT

Bending, Forming and Flexing Printed Circuits

Life Cycle Cost Analysis (LCCA)

Airfield Pavement Rehabilitation Planning Choosing the right fix. Rob McLure, M.Eng., P.Eng. Senior Associate Hatch Mott MacDonald

Ohio Department of Transportation Division of Production Management Office of Geotechnical Engineering. Geotechnical Bulletin PLAN SUBGRADES

Chapter 4 - Lecture 1 Probability Density Functions and Cumul. Distribution Functions

LOAD TESTING OF INSTRUMENTED PAVEMENT SECTIONS

How To Improve Road Quality

Benefit Cost Models to Support Pavement Management Decisions

Chapter 5: Normal Probability Distributions - Solutions

Stat 20: Intro to Probability and Statistics

KENTUCKY TRANSPORTATION CABINET POLICY FOR FEDERALLY FUNDED PAVEMENT REHABILITATION AND PREVENTATIVE MAINTENANCE PROJECTS

Rehabilitation Strategies for Highway Pavements

PERFORMANCE EVALUATION SYSTEM FOR BITUMINOUS BINDERS

NDT to Identify Concrete Bridge Deck Deterioration

ASPHALT PAVEMENT RECYCLING

South African Pavement Design Method (SAPDM) Revision Status Report. 26 th RPF Meeting 6 November 2013 L Kannemeyer

Fly Ash Slurry Injection (FASI) of Bituminous Thermal Cracks

What you measure is what you get? a novel approach for specifying and controlling acoustic quality of road surfaces

How To Treat An Expansive Soil

Automated Pavement Distress Survey: A Review and A New Direction

Modeling pavement performance by combining field and experimental data

Glossary of Terms Ability Accommodation Adjusted validity/reliability coefficient Alternate forms Analysis of work Assessment Battery Bias

ACPT. TechBrief july 2010 FHWA-HIF Impact of Temperature Curling and Moisture Warping on Jointed Concrete Pavement Performance.

Transcription:

Evaluation of PMS Recommendations for Flexible Pavement Rehabilitation Using RWD Deflection Data Zhongjie Zhang Pavement & Geotech Research Administrator Louisiana Transportation Research Center Louisiana Department of Transportation & Development 1 Gourrier Ave., Baton Rouge, LA 00 e-mail: doc.zhang@la.gov Kevin Gaspard Senior Pavement Research Engineer Louisiana Transportation Research Center Louisiana Department of Transportation & Development 1 Gourrier Ave., Baton Rouge, LA 00 e-mail: kevin.gaspard@la.gov Mostafa A. Elseifi Assistant Professor Department of Civil and Environmental Engineering Louisiana State University 0 Patrick Taylor Hall, Baton Rouge, LA 00 e-mail: elseifi@lsu.edu Submitted to: rd Transportation Research Board Annual Meeting January 1-1, 01 Washington, D.C.

1 1 1 1 1 1 0 1 0 1 ABSTRACT This paper presents a simple and direct approach for evaluating the criteria currently used by LADOTD PMS that makes the recommendation for AC pavement treatments using RWD deflection data. A new index based on RWD data is proposed, which is using the mean and variance of RWD deflection data over 0.1 mile pavement segment. To do this, a database of pavement segments with the length of 0.1 mile is generated with the new RWD index, pavement surface distress indices, such as cracking, rutting, roughness, and asphalt layer thickness on each segment. All the segments in the database were sorted according to their AC layer thickness and divided into various subgroups. The cumulated distribution function (CDF) of the new RWD index in each subgroup was generated so that various percentiles can be determined with the selection of a threshold percentile being made as the boundary between structural and functional rehabilitations. All the segments within each thickness group were further sorted according to various criteria of surface distress indices used to make recommendations for pavement treatments, respectively, and the histogram of each criterion was generated and compared with the selected threshold percentiles of the RWD index. It can be concluded from the analytical results that in term of the deflection based structural index, none of the criteria currently used by the LADOTD PMS can avoid Type I or II mistake when used to make pavement treatment recommendations, i.e., structural rehabilitation to pavements with sound pavement structure and functional rehabilitation to pavements with damaged and weak pavement structure. Because the severity of the problem is not trivial, a future PMS should consider both pavement structural indices and surface distress indices currently available when making recommendations for pavement preservation and rehabilitation. The RWD testing technology and data is one of the most promising candidates to fulfill this need. Keywords: Pavement rehabilitation, Pavement structural index, Pavement surface distress index, PMS, RWD, Treatment selection criterion Word Count Text,00 Figures 1x0 Tables x0 Total,00

0 1 0 0 INTRODUCTION The current pavement management system (PMS) collects various pavement surface distress data and generates the corresponding distresss indices as pavement condition indicators. Then, based on certain empirical criteria, the PMS will make recommendation on pavement preservation and rehabilitation to pavement engineers for actions. Using the state of Louisiana as an example, the PMS in the Louisiana Department of Transportationn and Development (LADOTD) hires an independent consultant for data collection on a two year collectionn cycle, which includes pavement distresses such as cracking, rutting, roughness, etc. for asphalt pavements. Five pavement distress indices, alligator cracks (ALCR), random cracks (RNDM), patching (PTCH), rutting (Rut), and roughness (RUFF, International Roughness Index based) are then derived accordingly (1, ). Table 1 specifies pavement treatments with corresponding ranges of the indices. Treatments 1 and are generally structurally related, treatmentt can be either structurally or functionally related, with the latter being the most probable, and treatments and are functionally related (1,, ). Table 1 Distress Indices and Treatment Criteria 1 Due to the limitation of each individual index, these five indices,, however, are seldom used solely on an individual basis. In LADOTD PMS, a composite criterion system is adopted and Figure 1 shows one examplee of such system for a local road. The use of the system is indicated by Figure and can be explainedd as follows.

Figure 1 Example of a Composite Criterion System for Local Road 0 1 Figure Flow Chart for Treatment Selection If the alligator crack index is less than, the patch index is less than, or the roughness index is less than, the corresponding pavement should be reconstructed. Otherwise, if the random crack index is less than 0, the alligatorr crack index is less than, the patch index is less than 0, or the roughness index is less than, the corresponding pavement should get medium overlay.. If the pavement distress indices do not meet the criteria of reconstruction and medium overlay, while the random crack index is less than 0, the alligator crack index is less than 0, the patch index is less than 0, the

1 1 1 1 1 1 0 1 0 1 0 1 rut index is less than, or the roughness index is less than 0, conduct thin overlay to the corresponding pavement. However, if none of above exists, the criterion for surface treatment is the random crack index being less than and the criterion for micro surfacing is the rut index being less than 0. As indicated by reviewing the above example, all the recommendation criteria are based on pavement surface distresses, which may or may not be the good indicator on the condition of pavement structures. Consequently, there is a possibility that two types of recommendation errors will be made, i.e., structural rehabilitation to pavements with sound pavement structure (Type I error) and functional rehabilitation to pavements with damaged and weak pavement structure (Type II error). Both cases will lead to the situation that highway maintenance funds are not spent cost effectively. Therefore, we need a tool to evaluate the recommendations made according to the above procedure and criteria to avoid or mediate these two types of errors. Fortunately, the pavement deflection data collected by the rolling weight deflectometer (RWD) has made such evaluation possible. RWD TESTING AND DATA (, ) A RWD measures surface deflections of in-service pavement at traffic speeds by comparing pavement profiles of surface in both un-deflected and deflected states. Due to its potential to be used to evaluate pavement structures network wide, LADOTD contracted with Applied Research Associate (ARA) in 00 and conducted a pioneer study using RWD and collecting the deflection data of asphalt pavement in District 0 of Louisiana with a total of around1,0 miles. The Louisiana testing used the latest upgraded measurement system of ARA with an,000 lb. standard load by a regular dual-tire assembly over the rear single axle. In order to evaluate the pavement structures tested by RWD, LADOTD also contracted with Fugro Consultants for collecting coring, falling weight deflectometer (FWD), and ground penetration radar (GPR) testing on the RWD tested pavements. More details on the testing, data quality, and processing as well as other information of the study can be found in the cited reference of this paper (, ). RWD collects pavement deflection data at an interval of 0. inch and a minimum length of unit for analysis has to be determined. As Elseifi, et al (, ) indicated, the random error (indicated by standard deviation (SD)) of average pavement defection will reduce as the increase of averaging length. An averaging length of ft. (0.1 mile) reduced the random error to approximately 1 mil, which contains approximately,0 deflection data points. Also in the PMS of LADOTD, all the pavement surface distress indices are the average values based on 0.1 mile of pavement segment (1). Therefore in this study, a 0.1 mile pavement segment was used as an analytic unit and each unit had a mean deflection with its standard deviation, which were provided by ARA and calculated out of the,0 relevant deflection data points within each 0.1 mile of pavement segment.

1 1 1 0 1 ANALYTICAL DATABASE A pavement segment database was established in thiss study using the data from the PMS of LADOTD, RWD, and pavement structure information, the element of which is 0.1 mile of pavement segment. Each element of the database has the informationn of asphalt thickness, surface deflection with standard deviation,, and surface distress indices discussed before. RWD INDICES A structural deflection index of RWD is needed to conduct the comparison analysis with other surface distress indices. Elseifi, et al (, ) has defined a parameter known as the RWD Index (RI), as follows, to correlatee with the effective structure number (SN eff ) of existing asphalt pavements: RI = Avg. RWD deflection * SD of RWD deflection (1) where, RI = RWD Index (mils ); Avg. RWD deflection = average deflection (mils) measured on a road segment with a length of 1. miles or longer (1 0.1-mile pavement segments or more); and SD of RWD deflection = standard deviation (mils) off average RWD deflections from those 0.1-mile pavement segments (standard deviation of means) ). This index was correlated reasonably well with the effective pavement structural number (SN ef ff) determined by falling weight Deflectometer (FWD) since RWD deflections reflected the deterioration of the pavement structure through both an increasee in the magnitude of the deflection and an increase in the scattering and variability of the deflection measurements (, ). Figure shows the phenomenon with the data used in that research effort. Figure Standard Deviation Increased with Deflection (one data point in the figure represents a 1. mile pavement segment)

1 1 1 1 1 1 1 1 0 1 0 1 Figure shows the relationship among the RI, deflection, and standard deviation, which indicates that RI is dominatedd by the deflection and modified by the standardd deviation. This information was used in this paper to guide the search for a new RWD index. In the study for this paper, since all the data points are based on the 0.1 mile of pavement segment, the RI would be dominated by the standard d deviation off deflection if the RI defined by Equation 1 is used again, as indicated by Figure, which is not reasonable. Therefore, a new definition of index based on RWD data, Zone RWD Index (ZRI), was used as follows. ZRI = averagee RWD deflection * fourth root of variance () where, ZRI is in mils / ; the average of deflection andd variance are based on the 0.1mile of pavement segment. The relationship among the ZRI, deflection, and fourth root of variance is shown in Figure, which is quite similar to Figure. The reason to use the new definition of RWD index is due to the fact that the standard deviation based on the 0.1 mile pavement segment is much higher than the one used in Elseifi s study discussedd previously, whichh is actually the standardd deviation of the meanss of 0.1 mile pavement segments. In general, the variation in the stiffness off pavement structures will be affected by the deterioration of pavement structures, differentt base type, variation of subgrade condition, variation of surface distresses such as cracking, rutting, etc., and the current RWD testing technology. The last factor is demonstrated in Figure, whichh shows a very slight but recognizable correlation between the deflection and standard deviation of 0.1 mile pavement segments. This is contradictory too our experience with and observation on in-situ pavement data (). On the other hand, it can be proved that RI and ZRI are mathematically related with respect to the variation of pavement deflections. 0 Figure Relationship among RI, Deflection, and Standard Deviation (one data point in the figure represents a 1. milee pavement segment)

Figure Relationship among RI, Deflection, and Standard Deviation, to AC Group (one data point in the figure represents a 0..1 mile pavement segment) 1 1 Figure Relationship among RI, Deflection, and Standardd Deviation, to AC Group (one data point in the figure represents a 0.1 mile pavement segment)

1 1 1 1 1 1 1 1 0 1 1 0 0 1 Figure Standard Deviation Increased with Deflection, to AC Group (one data point in the figure represents a 0.1 mile pavementt segment) The meaning of ZRI can be interpreted as follows. A high ZRI indicates a high pavement deflection with poor structure uniformity; a low pavement surface deflection means a high traffic loading capacity. This has been proved by Elseifi (ett al) s work published previously (, ) ). Based on various expressions evaluated during the course of that work, the following relationship between SN and RWD-measured parameters was the most accurate: where, RI = RWD Index = Avg. RWD deflection * SD of RWD deflection as defined previously in Equation 1, SN = Pavement Structural Number, and RWD = Avg. RWD deflection in mils. Historically, Asphalt Institute recommended the following formula for AC overlay design ():.. SN.1.RI RI. Where, ESALr = remaining life in equivalent single axle loadings (ESALs) and δ rd = rebound deflection, inches. Therefore, without considering the AC thickness, the deflection of AC pavement itself can be an indicator of traffic loading capacity of pavements. CUMULATED DENSITY FUNCTION N (CDF) OF ZRI AND ITS PERCENTILES 0..1RWD 0.0 () ()

1 1 1 1 1 1 0 1 0 1 Although the deflection of AC pavement itself can be an indicator of traffic loading capacity, the direct comparison of deflections from different structures and thicknesses is not very useful in practice due to the fact that they were designed to address different traffic loadings and environments. On the other hand, the direct comparison of pavement surface deflections with same pavement structure will not only distinguish their capability to handle traffic loading, but also establish the ranking of pavement structure conditions in a similar circumstance and so the priority list of candidates for actions with limited funding. Therefore, all the elements of 0.1 mile pavement segments in the established database were sorted according to the AC layer thickness as indicated by Table, which shows the number of data points in each thickness group. Table Various AC Thickness Groups and Data Thickness Group Number of Data Points 0-1, 1,0 1, 0 0 1 For a pavement network in a local district of Louisiana, the pavement of a local road normally follows a same typical structure design unless it has a unique traffic pattern. This allows us to assume that pavement with same asphalt thickness should have same pavement structure. In District 0 of Louisiana, the dominant base course should be soil cement (00 psi target value for strength) or cement treated base ( psi target value for strength, but with thicker layer). For some very old low volume roads, they might still use the sand-clay-gravel as the base. Since we cannot further distinguish them, another assumption was made here to treat them equally and their difference for traffic loading upholding was the credit given to the asphalt layer over them. This assumption will cause the extra variation in pavement surface deflection within each asphalt thickness groups, but seems reasonable since the subgrade of roadways varies anyway. It can be further investigated in the future if possible. A systematic approach for analysis is then conducted on these sorted subgroups and demonstrated here using the data of to AC thickness group. A cumulative distribution function (CDF) curve was developed for the ZRI of each subgroup with different percentiles being determined, as shown in Figure.

0 1 0 1 Figure CDF of ZRI in mils /, to AC Group With the constraint of same asphalt thickness (i.e. assumed same pavement structure as discussed previously), a percentile of ZRI in its CDFF also indicates the relative ranking of a pavement unit compared to the rest of pavement units with the same structure in the capacity to take traffic loading. For example, a 0 percentile of ZRI means that the pavement unit has a stiffness (traffic load capacity) better than the rest of 0% pavement units, but worse than the other 0% pavement units. Theoretically, a percentile of ZRI can be so determined that structural rehab (reconstruction or medium overlay) should be applied to pavement sections with ZRI larger than that percentile, i.e. with higher surface deflection and worse structure uniformity. On the other hand, functional rehab (thin overlay micro surfacing, surface treatment, or do-nothing) should be applied to pavement sections with lesss surface deflection and better structural uniformity. This threshold percentile is empirical in nature and can be determined by the data of historical pavement preservation and rehabilitation projects. For the purpose of demonstration in this paper, the 0% percentile, which is the ZRI of as indicated in Figure, is assumed as such a threshold percentile since the data from the District 0 includes both the good and bad pavements structurally. Therefore, structural rehabilitation could only apply to the pavements with ZRI values lager and functional rehabilitation should only apply to the pavements with ZRI values less than. Figuree shows the histogram of ZRI for the to AC group.

1 1 0 1 Figure Histogram of ZRI in mils /, to AC Group COMPARISON OF ZRI WITH SURFACE DISTRESS INDICES The data points in the to AC group were further sorted according to the individual surface distress index shown in Table 1 to be compared with the threshold percentile determined from the distribution of ZRI. However, one extra point needs to be made on the merit of ZRI before further discussing the comparison. Figures and show the CDFs of ZRI and deflection using the data sorted according to the ranges of the alligator crack index, ALCR, respectively. It can be seen thatt ZRI better distinguishes between the pavements with and without severe alligator cracking than deflection alone does, whichh is the contribution of the variance component of ZRI.

1 Figure ZRI in mils / and Alligator Cracks, to AC Group ZRI 0 1 0 1 Figure Deflection in mils and Alligatorr Cracks, to AC Group Figure 1 shows the histograms of ZRI sorted according to the ranges of ALCR with corresponding Types I and III errors when using the alligator crack index for treatment recommendations. For example, for the reconstructio on group with ALCR <, there are totally 1 pavement units, but units among them (1%) have the ZRI values less than, the 0 percentile of ZRI CDF, whichh only needss functional rehab. Therefore, it can be observed that if the deflection data of RWD can be used as the indicator of AC pavement structural conditionn and the threshold percentile is reasonable, the pavement treatment recommendations based on the alligator crack index can cause Types I and II errors. In this case, Type I errors for reconstruction and mediumm overlay are 1% and %, respectively; Type II errors for thin overlay andd do nothing are % and %, respectively.

1 1 1 1 1 1 Figure 1 Alligator Crack Index, Type I & II Errors, to AC Group Obviously, the percentages of errors will vary if the different threshold percentile is used. Increasing the percentile willl increase the Type I error, but reduce the Type II error. Vice versa is also true. Similar analyses were conducted on the data sorted according to other four surface distress indices criteria shown in Table 1 and the composite criterion in Figure 1. The summary of the results are shown in Table. Table Summary of Types I and II Errors for AC Group to 1 1 1

1 Table Summary of Types I and II Errors for AC Group to (cont.) 1 1 0 1 1 It can be concluded from Table that in term of the deflection based structural index, ZRI, none of the criteria currently used by the LADOTD PMS can avoid Types I or II mistakes when used to make pavement treatment recommendations, which shows the value and merit of RWD testing technology. Figure 1 shows the distributions of all the data points in the to AC group with ZRI, which confirms the same conclusion. The same comparison analysis is conducted for all the asphalt thickness groups in Table. Table shows the summary of Types I and II errors for all AC groups using the composite criteria of LADOTD PMS, which indicates the severity of the mis- recommendation problem in term of the deflection based structural index. 0 1 Figure 1 Distribution of Data Points with ZRI, to AC Group

1 Table Summary of Types I and II Errors for All AC Groups Table Summary of Types I and II Errorss for All AC Groups (cont.) 1 1 1 CONCLUSIONS This paper presents a simple and direct approach for evaluating the criteria currently used by LADOTD PMS to make the recommendation for AC pavement treatments using RWD deflection data. The approach analyzed the data of pavement distresses, RWD, and

1 1 1 1 1 1 0 1 0 1 0 1 thicknesses, which were collected over the entire asphalt road network (about 1,0 miles) in District of Louisiana using corresponding standard testing protocol. Based on the results of this analysis, the following conclusions are drawn: In term of the deflection based structural index, treatment recommendations for AC pavements made by current PMS cannot avoid Types I and II errors, i.e., recommending structural rehabilitation to pavements with sound pavement structure and functional rehabilitation to pavements with damaged and weak pavement structure. Future PMS should consider pavement structural indices in addition to the current surface distress indices when making recommendations for pavement preservation and rehabilitation. RWD testing technology and data is one of the most promising candidates to fulfill this need. Although the above conclusions are based on the analytical results of data from 0.1mile pavement segments and highway agencies rarely produce a list of treatments for 0.1 mile long pavement segments, the correct or proper treatment recommendations over 0.1 mile long pavement segments are still a necessity or fundamental requirement for a PMS to make proper treatment recommendations for pavement projects with various lengths. ACKNOWLEDGEMENTS The authors express their gratitude to the engineers and staff of LADOTD PMS, ARA, and Fugro Consultants for comment, help, and support. Fabiola Campoblanco and Parker Sanderson conducted the data processing and prepared the tables and figures for this paper. REFERENCES 1. Khattak, M. J., Baladi, G. Y., Sun, X., Veazey, J., and Landry, C., Development of Uniform Sections for PMS Inventory and Application. FHWA/LA.0/0, 00, Louisiana Transportation Research Center, Baton Rouge, LA.. Khattak, M.J., Baladi, G.Y., Zhang, Z., and Ismail, S., "A Review of the Pavement Management System of the State of Louisiana - Phase I." Transportation Research Record: Journal of the Transportation Research Board, 0, Washington D.C., 00.. Gaspard, K, Zhang, Z., and Elseifi, M.A., Integrating Rolling Wheel Deflectometer Stiffness Measurements into Pavement Management Systems using Multivariate Statistical Methods and Fuzzy Logic. Submitted to Journal of the Transportation Research Board, Washington D.C., 01. Elseifi, M. A., Abdel-Khalek, A., Gaspard, K., Zhang, Z., and Ismail, S.. Evaluation of the Rolling Wheel Deflectometer as a Structural Pavement Assessment Tool in Louisiana, Proceedings of the ASCE T&DI Congress 0, Integrated Transportation and Development for a Better Tomorrow, Chicago, IL, 0, -.

1. Elseifi, M. A., Abdel-Khalek, A. M., and Dasari, K., Implementation of Rolling Wheel Deflectometer (RWD) in PMS and Pavement Preservation. FHWA/., 01, Louisiana Transportation Research Center, Baton Rouge, LA.. Khattak, M. J., Baladi, G. Y., and Sun, X., Development of Index Based Pavement Performance Models for Pavement Management System (PMS) of LADOTD. FHWA/LA.0/0, 00, Louisiana Transportation Research Center, Baton Rouge, LA.. Asphalt Overlays for Highway and Street Rehabilitation, The Asphalt Institute,