–-

NEW MEXICO DEPARTMENT OF TRANSPORTATION

RESEARCH BUREAU Innovation in Transportation

Prepared by: University of New Mexico Department of Civil Engineering Albuquerque, NM 87131

Prepared for: New Mexico Department of Transportation Research Bureau 7500 Pan American Freeway NE Albuquerque, NM 87109 In Cooperation with: The US Department of Transportation Federal Highway Administration

Report NM11MSC-04

NOVEMBER 2013

Addressing Permeability of Superpave Mixes in New Mexico

USDOT FHWA SUMMARY PAGE

1. Report No. NM11MSC-04

2. Recipient’s Catalog No.

3. Title and Subtitle Addressing Permeability of Superpave Mixes in New Mexico

4. Report Date November 19, 2013

5. Author(s): Rafiqul A. Tarefder and Mohiuddin Ahmad

6. Performing Organization Report No. 456-340F

7. Performing Organization Name and Address University of New Mexico Department of Civil Engineering MSC01 1070 1 University of New Mexico Albuquerque, NM 87131

8. Performing Organization Code 456A

9. Contract/Grant No. 456-340

10. Sponsoring Agency Name and Address Research Bureau 7500B Pan American Freeway PO Box 94690 Albuquerque, NM 87199-4690

11. Type of Report and Period Covered Final Report July 01, 2011 to November 19, 2013

12. Sponsoring Agency Code

13. Supplementary Notes This project is funded by New Mexico Department of Transportation

14. Abstract Limiting the presence of water inside an asphalt concrete (AC) can slow down the process involved in water diffusion, hydration, adhesion loss and other mechanisms of moisture damage. This study attempts to determine the permeability of New Mexico mixes to make a database of permeability and to compare these permeability with specification value set by other DOT’s. This permeability values are also used to determine a permeability specification for New Mexico. This study also attempts to find out the correlation of permeability with permeable pore, effective pore and total pore. Attempts are made to correlate field permeability with laboratory permeability. It is observed that field permeability doesn’t show good correlation with laboratory permeability. An analytical model is developed based on permeable pore and permeability relation and field permeability is predicted. The model predicts field permeability pretty well for pavements without Open Graded Friction Coarse (OGFC). Attempts are also made to see if permeability has any correlation with moisture damage. It is observed that pavements with higher moisture damage have higher permeability. Moisture damage increases with the increase of permeability. Finally, a specification for permeability is proposed for New Mexico pavements. 15. Key Words Permeability, NMDOT, IDT, TSR, MIST, AASHTO T283

16. Distribution Statement Available from NMDOT Research Bureau

17. Security Classification of this Report None

18. Security Classification of this page None

19. Number of Pages 104

20. Price N/A

THIS PAGE LEFT BLANK INTENTIONALLY

ADDRESSING PERMEABILITY OF SUPERPAVE MIXES IN NEW MEXICO

Final Report July 1, 2011 - November 19, 2013

A Report on Research Sponsored by:

Research Bureau New Mexico Department of Transportation

7500B Pan American Freeway NE PO Box 94690

Albuquerque, NM 87199-4690 (505)-841-9145

research.bureau@state.nm.us http://NMDOTResearch.com

Prepared by:

Rafiqul A. Tarefder and Mohiuddin Ahmad

Department of Civil Engineering 1 University of New Mexico

MSC01 1070 Albuquerque, N.M. 87131

© New Mexico Department of Transportation

i

PREFACE The research reported herein represents laboratory and field testing of asphalt pavements and cores for permeability and their relation with other mix properties and moisture damage and to develop a permeability specification.

NOTICE

The United States government and the State of New Mexico do not endorse products or manufacturers. Trade or manufactures’ names appear herein solely because they are considered essential to the object of this report. This information is available in alternative accessible formats. To obtain an alternative format, contact the NMDOT Research Bureau, 7500B Pan American Freeway NE, PO Box 94690, Albuquerque, NM 87199-4690, (505)-841-9145

DISCLAIMER

This report presents the results of research conducted by the authors and does not necessarily reflect the views of the New Mexico Department of Transportation. This report does not constitute a standard or specification.

ii

ACKNOWLEDGEMENTS

The authors would like to thank the project advocates Mr. James Gallegos and Mr. Parveez Anwar, Project Manager Keli Daniell, and the Project Technical Panel Mr. Garry Schubert, Bob Meyers, Jeff Mann, and Robert McCoy for their support and guidance. Thanks go to Luis Melgoza of FHWA district at Santa Fe. The authors would also like to thank John Gillentine and his field exploration crew from NMDOT for continuous support during field permeability testing and coring. The authors would like to thank Valerie McCoy, Naomi Waterman, and Ghazanfar Barlas for their assistance. The authors like to thank NMDOT Research Bureau for funding this study.

iii

TABLE OF CONTENTS INTRODUCTION .......................................................................................................................... 1

RESEARCH NEED .................................................................................................................... 1

Permeability Specification ...................................................................................................... 1

Correlation of Permeability with Different Pores and Mix Volumetric ................................. 1

Field and Laboratory Permeability ......................................................................................... 2

Permeability and Moisture Damage........................................................................................ 2

GOALS AND OBJECTIVES ..................................................................................................... 2

REPORT ORGANIZATION ...................................................................................................... 3

LITERATURE REVIEW ............................................................................................................... 4

INTRODUCTION ...................................................................................................................... 4

PERMEABILITY TEST METHODS ........................................................................................ 4

Concept of Permeability Testing ............................................................................................ 4

Laboratory Methods ................................................................................................................ 4

Field Methods ......................................................................................................................... 8

Arbitrary methods ................................................................................................................. 13

FACTORS AFFECTING PERMEABILITY ........................................................................... 13

ANISOTROPY ......................................................................................................................... 14

PREDICTING FIELD PERMEABILITY FROM TESTING IN LABORATORY ................. 15

Method A .............................................................................................................................. 16

Method B .............................................................................................................................. 16

STATE DOT’S REVIEW ......................................................................................................... 16

Results of Study .................................................................................................................... 17

METHODOLOGY AND SURVEY ............................................................................................. 18

EXPERIMENTAL PLAN ........................................................................................................ 18

SURVEY................................................................................................................................... 19

Survey Response Synthesis and Data Collection .................................................................. 19

PERMEABILITY TESTING........................................................................................................ 21

INTRODUCTION .................................................................................................................... 21

FIELD TESTING PROGRAM ................................................................................................. 21

Special Field Testing Issue ................................................................................................... 24

iv

Core Collection and Selection of Candidate Pavements for Laboratory Testing ................. 24

LABORATORY PERMEABILITY TESTS ............................................................................ 25

TEST RESULTS ....................................................................................................................... 25

Field Permeability Test Results ............................................................................................ 27

Laboratory Permeability Test Results ................................................................................... 28

Comparison of Permeability at Different Test Modes .......................................................... 29

GYRATORY COMPACTED SAMPLES ................................................................................ 30

CONCLUSIONS....................................................................................................................... 30

CORRELATION OF PERMEABILITY WITH VOLUMETRICS AND PORES ...................... 32

INTRODUCTION .................................................................................................................... 32

MIX VOLUMETRIC AND PERMEABILITY........................................................................ 32

Mix-volumetric ..................................................................................................................... 32

DIFFERENT TYPE OF PORES .............................................................................................. 34

Definition .............................................................................................................................. 34

Determination of Effective and Total Pores ......................................................................... 35

Permeable Pores by Tracer Method ...................................................................................... 36

SELECTION OF CANDIDATE PAVEMENT........................................................................ 38

RESULTS AND DISCUSSIONS ............................................................................................. 39

Different Pores and Their Correlation .................................................................................. 39

Correlation of Different Pores with Permeability ................................................................. 40

Quantification of Different Types of Pores .......................................................................... 40

CONCLUSIONS....................................................................................................................... 41

CORRELATION BETWEEN FIELD AND LABORATORY PERMEABILITY...................... 43

INTRODUCTION .................................................................................................................... 43

THEORETICAL DEVELOPMENT OF FIELD PERMEABLE PORES AND PERMEABILITY ..................................................................................................................... 43

Model 1 ................................................................................................................................. 45

Model 2 ................................................................................................................................. 46

RESULT AND DISCUSSIONS ............................................................................................... 48

Laboratory Measured and Predicted Permeability ................................................................ 48

Laboratory, Predicted and Field Measured Permeability ..................................................... 48

v

CONCLUSIONS....................................................................................................................... 49

PERMEABILITY AND MOISTURE DAMAGE........................................................................ 51

INTRODUCTION .................................................................................................................... 51

LABORATORY DETERMINATION OF MOISTURE SUSCEPTIBILITY BY CONDITIONING ..................................................................................................................... 51

AASHTO T 283 Wet Conditioning ...................................................................................... 51

MIST Wet Conditioning ....................................................................................................... 51

Tensile Strength Testing ....................................................................................................... 52

HAMBURG WHEEL TRACKER TEST ................................................................................. 53

RESULTS AND DISCUSSIONS ............................................................................................. 54

Field Damage with Field Permeability ................................................................................. 54

Field Damage with Laboratory Permeability of Full Depth Samples................................... 55

Field Damage with Laboratory Permeability of Samples Separated into Layers ................. 55

MIST Damage with Permeability ......................................................................................... 56

AASHTO T 283 Damage with Permeability ........................................................................ 58

MIST and AASHTO T283 Damage ..................................................................................... 59

Field Damage with Laboratory Damage ............................................................................... 60

Permeability and Moisture Damage of Laboratory Compacted Samples ............................. 61

Relation of Different Type of Pores with Moisture Damage ................................................ 62

Hamburg Wheel Tracking (WHT) Tests .............................................................................. 63

CONCLUSIONS....................................................................................................................... 67

PERMEABILITY SPECIFICATION ........................................................................................... 69

DEVELOPMENT OF PERMEABILITY SPECIFICATIONS ................................................ 70

CONCLUSIONS AND RECOMMENDATIONS ....................................................................... 74

CONCLUSIONS....................................................................................................................... 74

RECOMMENDATIONS .......................................................................................................... 74

REFERENCES ............................................................................................................................. 76

APPENDIX A ............................................................................................................................... 79

FIELD PERMEABILITY TEST RESULTS ............................................................................ 79

APPENDIX B ............................................................................................................................... 85

LABORATORY FULL DEPTH SAMPLE TEST RESULT ................................................... 85

vi

APPENDIX C ............................................................................................................................... 89

LABORATORY PERMEABILITY OF SAMPLES CUT INTO LAYERS ............................ 89

APPENDIX D ............................................................................................................................. 101

MOISTURE DAMAGE TEST RESULTS ............................................................................. 101

APPENDIX E ............................................................................................................................. 103

TRACER TEST DATA .......................................................................................................... 103

vii

LIST OF FIGURES

FIGURE 1 Florida Asphalt Permeameter ....................................................................................... 5 FIGURE 2 Karol Warner Permeameter .......................................................................................... 7 FIGURE 3 Pulse-Decay and Faster Pulse-Decay Method of Rock Permeability .......................... 8 FIGURE 4 Kentucky Air Induced Permeameter ............................................................................ 9 FIGURE 5 NCAT Field Permeameter .......................................................................................... 11 FIGURE 6 ROMUS Air Permeameter ......................................................................................... 12 FIGURE 7 Kuss Field Permeameter ............................................................................................. 13 FIGURE 8 Horizontal Permeability Diagram............................................................................... 15 FIGURE 9 Experimental Design .................................................................................................. 18 FIGURE 10 Field Permeability Test Set Up................................................................................. 22 FIGURE 11 Field Permeability Test Summary for US285 MP 140.53 ....................................... 23 FIGURE 12 Schematic of Falling Head Permeability Method .................................................... 24 FIGURE 13 Laboratory Permeability Test Set-up ........................................................................ 26 FIGURE 14 Field Permeability Results ........................................................................................ 28 FIGURE 15 Permeability of Full Depth Samples......................................................................... 28 FIGURE 16 Permeability of Different Layers .............................................................................. 29 FIGURE 17 Comparison of Permeabiliity at Different Test Modes ............................................ 29 FIGURE 18 Permeability vs Air Voids of Gyratory Compacted Samples ................................... 30 FIGURE 19 Gradation Curve for All Pavement Sections ............................................................ 34 FIGURE 20 Different Types of Pores Inside A AC Sample ........................................................ 34 FIGURE 21 Bulk Specific Gravity of a Cored Sample ................................................................ 35 FIGURE 22 Tracer Test Set-up .................................................................................................... 36 FIGURE 23 Breakthrough Curve for a Sample from NM344 at MP1.8. ..................................... 38 FIGURE 24 Correlation between Different Types of Pores ......................................................... 39 FIGURE 25 Correlation of Different Pores with Permeability .................................................... 40 FIGURE 26 Pore Quantification ................................................................................................... 40 FIGURE 27 Different Types of Pores in A Sample ..................................................................... 45 FIGURE 28 Laboratory and Predicted Permeability .................................................................... 48 FIGURE 29 Measured and Predicted Permeability ...................................................................... 49 FIGURE 30 Sample in the MIST Chamber .................................................................................. 52 FIGURE 31 IDT Testing in Progress............................................................................................ 52 FIGURE 32 Hamburg Wheel Tracking Tests ............................................................................... 53 FIGURE 33 Field Permeability at Different Damage Conditions ................................................ 54 FIGURE 34 Laboratory Full Depth Permeability at Different Damage Conditions .................... 55 FIGURE 35 Laboratory permeability of Separated Samples at Different Damage Conditions ... 56 FIGURE 36 Correlation of MIST TSR with Permeability at Different Modes ............................ 58 FIGURE 37 Correlation of AASHTO T283 TSR with Permeability at Different Test Modes .... 59 FIGURE 38 Comparison between AASHTO T 283 and MIST TSR ........................................... 60 FIGURE 39 TSR Values for Good and Bad Performing Pavement Sections .............................. 61

viii

FIGURE 40 Permeability and Moisture Damage of Labory Compacted Samples ...................... 62 FIGURE 41 Correlation of TSR with Dead End and Permeable Pores ........................................ 62 FIGURE 42 HWT Plots for Good Pavements .............................................................................. 64 FIGURE 43 HWT Plots for Bad Performing Pavements ............................................................. 67 FIGURE 44 Correlation of Permeability with Air Voids ............................................................. 69 FIGURE 45 Correlation of Average Permeability with Average Air Voids ................................ 69 FIGURE 46 Permeability Vs. Air Voids for Full Depth Samples ................................................ 70 FIGURE 47 Critical Permeability from Labortory Testing of Field Cores .................................. 71 FIGURE 48 Critical Permeability Values Obtained from Average Permeability Values ............ 71 FIGURE 49 Critical Permeability For Full Depth Samples ......................................................... 71

ix

LIST OF TABLES

TABLE 1 Different range of permeability and air voids. ............................................................. 14 TABLE 2 NMAS and related permeability values. ...................................................................... 14 TABLE 3 Participation of Districts from NMDOT. ..................................................................... 19 TABLE 4 Field permeability test locations. ................................................................................. 21 TABLE 5 Selected pavements for laboratory testing. .................................................................. 25 TABLE 6 Calculation of average permeability. ........................................................................... 26 TABLE 7 Mix data for selected pavement. .................................................................................. 32 TABLE 8 Pavement sections selected for tracer test. ................................................................... 38 TABLE 9 Moisture damage inflection point. ............................................................................... 53 TABLE 10 Rut depth limit ........................................................................................................... 53 TABLE 11 TSR calculation. ......................................................................................................... 57 TABLE 12 HWT data summary for good pavements. ................................................................. 65 TABLE 13 Summary of HWT test data for bad pavements. ........................................................ 67 TABLE 14 Different permeability limits for laboratory testing. .................................................. 72 TABLE 15 Critical permeability obtained from average permeability plot. ................................ 72 TABLE 16 Critical permeability values for full depth sample ..................................................... 72

1

INTRODUCTION

RESEARCH NEED

Permeability Specification

In the past, New Mexico Department of Transportation (NMDOT) has built asphalt pavements using dense-graded Hot Mix Asphalt (HMA) mixes designed by the old Marshall Method. The initial in-place total pores (voids) of those dense graded mixtures are controlled not to be higher than eight percent and never fall below three percent during the life of the pavement. Higher pore content makes the pavement structure more permeable to air and water. High permeability may cause enhanced aging and moisture damage of HMA as well as endanger the subgrade and base courses. Lower pore content causes rutting and shoving due to shear deformation of mix constituents under repeated traffic loading. Thus, pores in an asphalt concrete pavement have contrasting effects on its properties and performance. Studies have shown that a dense graded (Marshall) mix containing total pores below 8% has low permeability (less than 100 x 10-5 cm/sec), which prevents infiltration of water inside a pavement (1-2). As a result, very little to no attention has been given so far to permeability requirements of dense graded Marshall mixes nationally and in the state of New Mexico. In recent past, most of the state DOT’s including NMDOT, have adopted a new mix design system called “SuPerPave” (Superior Performing Pavement). Superpave uses more angular, fractured, and relatively larger size aggregates. Like Marshall and/or Hveem mixtures, though the initial in-place pores of dense graded Superpave mixtures are also controlled not to be higher than eight percent (a target of 7% with ±1% tolerance); the relatively coarser and angular nature of Superpave mixtures produces a greater number of permeable pores which is believed to cause higher permeability. Though high permeability is not desirable in a pavement structure as it maximize the water intrusion into the pavement, what is not known is whether NMDOT’s Superpave designed mixes are more permeable than comparable Marshall designed mixes that have been used for years and, most importantly, how permeable is too permeable, i.e. what is a reasonable maximum permeability for a dense graded asphalt concrete. In this study field and laboratory permeability tests are done on pavements from 16 pavements of New Mexico also on two different laboratory mixes (SP III and SP IV) which may be used as database of permeability for future studies. Also, using these data, a permeability specification is proposed for New Mexico.

Correlation of Permeability with Different Pores and Mix Volumetric

Currently most of the DOTs accept a mix design based on total pores (total pore and total air voids are used interchangeably in this report). Specifically, cores from new pavements are collected immediately after compaction to check the air voids. Logically, if total air voids are related to permeability, there is no need for a permeability specification in addition to the air voids specification. Total pore consists of three pores: permeable pore, dead-end pore, and isolated pore (3). Only pores that continue from the top to the bottom of a sample are responsible for permeability (i.e., conduction of flow) and therefore they are called permeable pores. Permeable pores and dead-end pores are accessible by water and therefore, the sum of these two

2

pores is also known as effective pores (4). It is possible that permeable pores have better correlation with permeability than total pores. This study attempts to determine whether permeability has better correlation with permeable pore or effective pore or total pore and if gradation has any effect on permeability.

Field and Laboratory Permeability

Once a new pavement is constructed, permeability tests can be conducted on these new pavements in the field. Alternatively, permeability tests can be conducted in the laboratory on the cores collected from new pavements. Now the question whether the two quantities (field and laboratory permeability) are related or not. If they are related, only laboratory permeability tests can be performed, as cores are always collected from the new pavement to check the in-situ density. Field permeability value is then obtained using the laboratory permeability data. However, correlating laboratory permeability with field permeability may be very difficult. In the laboratory, a permeability test sample's sides are confined and flow is solely one-dimensional, which is different from the field permeability boundary condition. In this study, attempts are made to examine whether laboratory and field permeability can be correlated or not.

Permeability and Moisture Damage

Moisture damage of asphalt pavement is an important issue worldwide. In the United States, the problem is also severe, as 34 out of 50 states are suffering from some sort of moisture related distress (5). According to Mogawer et al. (2002) (6), 15 states reported moisture related problems out of 27 states surveyed. Moisture damage in asphalt can be prevented either by preventing water from entering an asphalt concrete (AC) or by improving the adhesion (strength) of the asphalt-aggregate system. Permeability is the way by which water can enter inside the pavement. If the permeability of pavement can be controlled, the moisture damage in AC can be prevented or reduced. Unfortunately, very few studies have been conducted correlating permeability and moisture damage to this day. In this study attempts are made to examine whether moisture damage is related to permeability or not. Moisture damage of a pavement can be evaluated by two ways, visual inspection and mechanical testing. Stripping generally occurs at interfaces and propagates upward. It is difficult to identify moisture damage by visual inspection only. Currently the AASHTO T 283 (7), Moisture Induced Sensitivity Testing (MIST), and Hamburg wheel test are widely used. These three methods are used in this study to determine laboratory moisture damage. A field survey is conducted to identify pavements which have moisture damage.

GOALS AND OBJECTIVES The primary goal of the research project is to develop a specification for permeability of Superpave mix designs which will result in extending the lifetime of pavements by reducing moisture related damage and pavement maintenance costs as well as by public safely. The specific objectives of this study are to:

1. Develop a database of permeability of New Mexico pavements and mixes and develop a permeability specification for New Mexico.

3

2. Determine the correlation of permeability with permeable pores, effective pores and dead end pores and mix gradation.

3. Determine correlation between field and laboratory permeability. 4. Find whether permeability is related to moisture damage or not.

REPORT ORGANIZATION This report consists of nine sections. Section 1 describes the research need and objectives of this study. Section 2 contains a literature review on different laboratory and field permeability test methods, factors affecting permeability and permeability practices of different states and countries. Section 3 contains an experimental plan or methodology for this study along with the survey to identify candidate pavements. Section 4 includes field and laboratory permeability testing performed during this study. Section 5 describes the determination of different kinds of pores in an asphalt concrete (AC) sample and their correlation with permeability. This section also includes the mix design data of all pavements retrieved in the laboratory. Section 6 includes the correlation between field and laboratory permeability. An analytical model is described in this section to determine field permeability by testing samples in the laboratory. Section 7 describes different moisture damage test methods and their correlation with permeability. Section 8 describes the proposed permeability specification for New Mexico. Section 9 summarizes the findings of this study.

4

LITERATURE REVIEW

INTRODUCTION The authors conducted a comprehensive search of databases and various information sources (like journals, research reports, and standards) to build a solid base of knowledge on current research and specifications. Particular emphasis was placed on permeability testing methods and permeability specifications for Superpave mix design. A review of the factors affecting permeability is also made.

PERMEABILITY TEST METHODS

Concept of Permeability Testing

Permeability or hydraulic conductivity of a material refers to its ability to transmit water through it. It is obtained from Darcy’s law: flow velocity is proportional to hydraulic gradient:

𝑉 ∝ 𝑖

𝑉 = 𝑘𝑖 (1)

where V = flow velocity (cm/s); i = flow gradient defined by head loss over sample length; k = permeability of the sample (cm/s). Therefore, discharge rate (Q) will be:

𝑄 = 𝐴𝑉 = 𝑘𝑖𝐴 (2)

where Q = discharge (cm3/s); A = cross section area (cm2).

Permeability of soil is determined either using falling head or constant head permeameter. Almost all permeameter are based on one of these two mechanisms. Permeability of asphalt is evaluated both in the field and in the laboratory. Different states or organizations use different instruments to determine permeability. Brief descriptions of these are given in the next sections.

Laboratory Methods

Florida Method of Permeability

Scope: This test method covers the laboratory determination of the water conductivity of compacted asphalt paving mixture sample. Figure 1 shows the Florida apparatus. The measurement provides an indication of water permeability of the sample as compared to those of other asphalt samples tested in the same manner. The procedure uses either laboratory compacted cylindrical specimens or field core samples obtained from existing pavements.

5

FIGURE 1 Florida Asphalt Permeameter

Summary of Test Method: A falling head permeability test apparatus is used to determine the rate of flow of water through the specimen. Water in a graduated cylinder is allowed to flow through a saturated asphalt sample and the interval of time taken to reach a known change in head is recorded. The coefficient of permeability of the asphalt sample is then determined based on Darcy’s Law. Significance and Use: This test method provides a means for determining water conductivity of water-saturated asphalt samples. It applies to one-dimensional, laminar flow of water. It is assumed that Darcy’s law is valid. Calculations: The coefficient of permeability, k, is determined using the following equation:

𝑘 = 𝑎𝐿𝐴𝑡𝑙𝑛 �ℎ1

ℎ2� × 𝑡𝑐 (3)

where k = coefficient of permeability, cm/s;

a = inside cross-sectional area of the burette, cm2;

L = average thickness of the test specimen, cm; A = average cross-sectional area of the test specimen, cm

2;

t = elapsed time between h1

and h2, s;

6

h1

= initial head across the test specimen, cm; h

2 = final head across the test specimen, cm;

tc

= temperature correction for viscosity of water.

Oklahoma Method of Permeability

Scope: This test method covers the laboratory determination of permeability of the compacted asphalt paving mixture sample. The sample is either laboratory compacted or collected field cores. Summary of the Test Method: Evacuate air from the sealing tube. Place the specimen on top of the lower plate. Place the sealing tube over the specimen. Insert the upper cap assembly it into the sealing tube. Install the clamp assemble onto the permeameter frame. Inflate the membrane and pour water in the graduated cylinder and shake to remove air. Fill again up to initial timing mark. Start the timing device and record total time taken by water to reach the lower meniscus. Permeability is determined using Eq. (3).

Karol-Warner Falling Head Permeability Device

Introduction: As a result of their work, the Florida Department of Transportation (DOT) developed a laboratory asphalt permeameter to measure the permeability of field cores. This device was further refined and then marketed by Karol-Warner and known as the “Karol-Warner falling head permeability device” and shown in Figure 2. Description: In this falling head permeability test, as outlined in the ASTM provisional standard PS-129, a saturated asphalt sample is sealed on the sides and placed under a column of water so that water can only flow through the sample. The time required for the water column to experience a specified change in elevation is determined. The test is repeated until four consecutive readings do not differ by more than ten percent. This process confirms that the sample was, in fact, saturated. Otherwise, it would be unclear whether movement of the water column was due to water infiltrating void spaces or actual flow through the sample (8). Calculation: The permeability for water is calculated using the following equation:

𝑘 = 𝑎𝐿𝐴𝑡𝑙𝑛 �ℎ1

ℎ2� (4)

The terms are explained in Eq. (3)

7

FIGURE 2 Karol Warner Permeameter

Texas Method of Permeability

Scope: Use this test procedure to measure the time of water flow through laboratory compacted Permeable Friction Course (PFC) specimens. Then determine permeability. Apparatus: Cylindrical laboratory permeameter, stop watch. Specimen: 6 in. diameter 4.5 in. height lab compacted sample or field core of 6 in. diameter and height not more than 4.5 in. Procedure: Place test specimen in the cylindrical laboratory permeameter. Secure rubber clamps around the permeameter at the top and bottom edges of the test specimen. Place the clamps such that the top and bottom edges are approximately in the middle of the location of the clamps. Tightly secure the clamps such that water does not flow around the test specimen. Water must flow through the test specimen. Fill the permeameter with water approximately 1–2 in. above the top marking on the pipette. Start the timing device when the water level reaches the top marking on the pipette. Stop the timing device when the water level reaches the bottom marking on the pipette. Record the time the water travelled from the top marking to the bottom marking. Calculate permeability by falling head formula as shown by Eq. (3).

Pulse-Decay Measurement Technique

Introduction: The system consists of an upstream reservoir of volume V1, a sample holder and a downstream reservoir of volume V2. A differential pressure transducer measures the pressure difference between the reservoirs and another transducer measures the pressure p2 in the downstream reservoir. No flow measurement device is required. Flow rate can be calculated from known volume of each reservoir, fluid compressibility and rate of pressure change. The device is shown in Figure 3. Procedure: High pressure is applied at upstream reservoir with all bulbs open. When pressure become uniform all over the system, upstream pressure is increased keeping bulbs in such way that pressure pulse enters into the sample. The upstream pressure will decrease with time and downstream pressure will increase until the pressure difference become zero. The pressure

8

difference is measured with time and plotted in a semi-log plot. The slope of the line could be obtained by:

|𝑆𝑙𝑜𝑝𝑒| = 6.805×10−5𝑓1𝐴𝑘𝐿𝜇𝐿(𝑐𝐿+𝑐𝑉1)

� 1𝑉1

+ 1𝑉2� (5)

𝑘𝐿 = −14696𝑚1𝜇𝐿(𝑐𝐿+𝑐𝑉1)

𝑓1𝐴�1𝑉1+ 1𝑉2�

(6)

where m = slope;

µL = viscosity of the liquid, cp; cL = liquid compressibility, psi-1; cv1 = compressibility of upstream reservoir, psi-1; f1 = mass-flow correction factor; A = cross section area of the cylindrical sample, cm2.

Faster pulse decay permeability measurement

Description: The method is similar to that described in previous method, except the pressure equilibrium step is eliminated and the test can be performed quicker than before. Two additional reservoirs are added to the previous instrument. These methods are used normally for rock permeability measurement.

FIGURE 3 Pulse-Decay and Faster Pulse-Decay Method of Rock Permeability

Field Methods

California Method

Scope: This method describes the procedure for determining the permeability of bituminous pavements and seal coats. Test Procedure: Dense graded asphalt concrete and seal coats: Draw a 152 mm diameter circle on pavement surface, extrude grease from caulking gun, push small amount of grease on pavement, fill cylinders with test solution, release the valve at the base of the special plastic graduated cylinder, start the stop watch and run solution from the special plastic graduated

9

cylinder onto the area within the grease ring, keeping this area constantly covered with about a 1.6 mm film of the solution for two minutes. Refill the special plastic cylinder from the polyethylene graduate if more solution is needed during the test. At the end of the 2-minute test period, determine the total amount of solution used. Calculation: Divide the total quantity of solution used during the test period by two and record the result as the relative permeability in mL/min.

Kentucky Method

Scope: This method describe the procedure for determining in place permeability of an AC sample using an air induced permeameter. This method is applicable for all nominal maximum size and gradation. Apparatus: Permeameter, Air compressor and caulking gun. Procedure: The apparatus is shown in Figure 4. Connect the air compressor to the multi venture vacuum cube. Apply approximately a one-half in. bead of silicon rubber caulk one in. inside the outer edge of sealing ring. Place the permeameter in the center of the area to be tested. Apply no more than 50 pounds on permeameter and twist. Open the bulb of multi-venturi vacuum cube to permit the flow of air. The reading on the digital vacuum will begin to increase. When this number reaches to peak, the test is finished and the bulb can be shut. Test time should not exceed 15 s. Record the highest reading attained by the permeameter by pressing the button marked “HI/LO”. Calculation: Permeability of the mat in ft/day may be calculated from the following equation:

𝑘 = 255757.53𝑉1.556 (7)

where k = permeability in ft/day; V = vacuum reading in mm Hg.

FIGURE 4 Kentucky Air Induced Permeameter

10

Texas Method

Scope: Use this empirical test procedure for PFC pavements under construction or on PFC roadways already constructed to test and verify that the compacted mixture has adequate permeability. Apparatus: Cylindrical field permeameter, stop watch and plumber’s putty. Procedure: Place the permeameter on the pavement surface, use putty to seal it with pavement. Fill the permeameter 1-2 in. more than the top marking. Measure time taken by water to travel from the top marking to the bottom marking. Time is normally less than 20 s for newly constructed pavement. We can calculate the permeability using falling head method.

NCAT Field Permeameter

Scope: This test method covers the in-place estimation of the water permeability of compacted hot mix asphalt (HMA) pavement.

A falling head permeability test is used to estimate the rate at which water flows into a compacted HMA pavement. The apparatus is shown in Figure 5. Water from a graduated standpipe is allowed to flow into a compacted HMA pavement and the interval of time taken to reach a known change in head loss is recorded. The coefficient of permeability of a compacted HMA pavement is then estimated based on Darcy’s Law. Significance and Use: This test method provides a means of estimating water permeability of compacted HMA pavements. The estimation of water permeability is based upon assumptions that the sample thickness is equal to the immediately underlying HMA pavement course thickness; the area of the tested sample is equal to the area of the permeameter from which water is allowed to penetrate the HMA pavement; one-dimensional flow; and laminar flow of the water. It is assumed that Darcy’s law is valid (9).

Calculation: Same formula used for any falling head method as shown by Eq. (3).

ROMUS Air Permeameter

Description: The ROMUS air permeameter uses air that is gathered from the atmosphere as the fluid to measure permeability as shown in Figure 6. The machine has a vacuum pump that operates on a rechargeable battery and depressurizes the tank to negative 24 in. of head. When the test is ready to begin, the air tank is pressurized to 24 in. of head if it is not already at the appropriate pressure. The air is then drawn through the pavement surface while a pressure sensor checks the pressure in the tank. At every 4 in. of pressure drop the time is recorded. The time is the only output for the device (10). Calculation: Permeability of the pavement is calculated from the following equation which is modified from falling head equation:

𝑘𝑤 = �𝐿𝑉𝜇𝜌𝑤𝑔𝑇𝐴𝑃𝑎𝜇𝑤

� 𝑙𝑛 �𝑝1𝑝2� (8)

11

where kw = hydraulic permeability (L/T); L = pavement layer thickness (L); V = volume of vacuum chamber (L3); μ = kinematic viscosity of air (M/LT); ρw = density of water (M/L3); g = gravitational acceleration (L/T2); T = time of head drop (T); A = area of being tested (L2); Pa = pressure (atmospheric) (F/L2); μw = kinematic viscosity of water (M/LT); p1 = initial pressure (L); p2 = final pressure (L)

FIGURE 5 NCAT Field Permeameter

12

FIGURE 6 ROMUS Air Permeameter

The Kuss Field Permeameter

Description: The Kuss Field Permeameter (KSFP) operates using the constant head approach. The device is shown in Figure 7. A patented gas-measurement system is used to measure the amount of air needed to replace the water in order to maintain a constant pressure head. When the test begins, water is allowed to flow from the standpipe and cover the pavement testing surface to a depth of approximately 1 in. A sensor is used to monitor the water level, and is connected to a flow valve in the flow meter box. As water infiltrates the pavement, the water level over the pavement drops, and the sensor alerts the flow valve, allowing air to enter the standpipe above the water column. This metered volume of air acts as a substitute for the head pressure originally applied by the water, thereby maintaining a constant head. A data acquisition system measures and records the flow rate of water through the pavement over time, and the permeability is calculated using the rate of flow of water into the pavement and the cross-sectional area of the pavement test section. The relationship is presented in the equation below (8).

𝑘 = 𝑄

60�2.54+𝐿𝐿 �𝐴

(9)

where k = coefficient of permeability, cm/s;

Q = flow rate, cm3/min; A = area of base plate, 1264.5cm2; L = pavement thickness.

Vacuum Pump

Air Tank of Known Volume Pressure Sensor

Timing Device

Check Valve

Seal to Pavement

Asphalt Surface

Main Valve

13

FIGURE 7 Kuss Field Permeameter

Arbitrary methods

Modified Kozeny-Carman Equation

𝑘 = 𝑉𝑎𝑚

𝑐.𝑆𝑎𝑔𝑔𝑡𝛾𝜇

(10)

where Vam = total air voidsγ = 9.79 kN/m3;

µ = fluid viscosity (11).

Arkansas Method

The permeability coefficient of an asphalt mixture can be estimated by the following equation (Arkansas Highway Transportation Department (1998)):

𝑘 = 1.38 × 10−7 × 3.92%𝑉𝑎 × 0.61𝑡 (11)

where %Va = Air Voids, expressed as a percentage; t = lift thickness (12).

FACTORS AFFECTING PERMEABILITY

Research conducted by the NCHRP showed an inverse relationship between VMA and mixture permeability, i.e. for a given air void content permeability decreased as the VMA increased (13). With the increase in coarse aggregate, permeability also increased. One reason for such increase is increase of interconnected air voids. Superpave mixes tend to have this trait as well as having a larger NMAS with lesser fines to fill void spaces which results in a substantial increase in permeability (14). Void contents as low as 4.4% have been shown adequate for 25 mm NMAS mixes and 7.7% being the maximum for NMAS mixes of 9.5 mm. Table 1 shows air permeability ranges at different air voids.

14

TABLE 1 Different range of permeability and air voids.

k (cm/s) Description of Permeability Description of Air Voids <10-4 Impervious-Impermeable <5%

10-4-10-2 Poor drainage-Permeable 5-7% 10-2 or higher Good Drainage-Very

Permeable >7%

Permeability increases with an increase of NMAS. Mallick, et al. (1) got the permeabilities for 6 percent air voids as shown in Table 2.

TABLE 2 NMAS and related permeability values.

NMAS(mm) Permeability(cm/s) 9.5 6 x 10-5

12.5 40 x 10-5

19.0 140 x 10-5

25.0 1200 x 10-5

Another factor that influences permeability is gradation. For coarse graded mixes, because of coarser particles, the void size is larger resulting more interconnected voids. According to Mallick et al (1), lift thickness (or t/NMAS) also affects permeability. As thickness (or t/NMAS) increases, the potential for interconnected air void decreases, resulting in a decrease in permeability.

ANISOTROPY

All the permeameters available could measure the vertical permeability only but in the field, water flows in all three direction (x, y and z axis). Hence, we need to determine the horizontal permeability to have a better prediction of actual field permeability. Little research focused on horizontal permeability. They used different approaches. A study was conducted by Kutay et al. (14). He used the modified Kozeny-Carman equation to predict HMA permeability:

𝑘 = 𝐶̅𝑛3

(1−𝑛)2�𝐷𝑠 �1 + 𝐺𝑠𝑏(𝑃𝑏−𝑃𝑏𝑎(1−𝑃𝑏))

𝐺𝑏(1−𝑃𝑏)��2 𝛾𝜇

(12)

where k = coefficient of permeability;

C = an empirical constant, usually 0.003; n = the porosity; Ds = the average particle size; Gb = the binder specific gravity; Pba = the percent of absorbed binder by weight of aggregate; Pb = percent of asphalt content by total weight of the mix; Gsb = the bulk specific gravity of the aggregate.

15

Here the sample was divided into three layers and porosity of each sub-layer was estimated using the Corelok device. The porosity in the top and bottom layers was more than middle layer. Al-Omari et al. (15) obtained the same character using X-ray CT images. They found good correlation of horizontal permeability with the average of top and bottom layer porosity and the vertical permeability with middle layer porosity. Hence, the average porosity of top and bottom layer was used in this equation to determine horizontal permeability and the porosity at mid layer was used to determine vertical permeability. Harris et al. (16) developed a device and method to determine horizontal permeability. He derived the following equation to determine horizontal permeability:

𝑘ℎ =𝑟𝑝2(ℎ0−ℎ𝑡)ln (𝑟0𝑟𝑖

)

𝐿.(ℎ0+ℎ𝑡) (13)

where rp = radius of standpipe;

r0 = outer radius of sample; ri = inner radius of the sample; h0 = initial head; ht = final head; t = time; L = length of the sample. These parameters are illustrated in Figure 8.

FIGURE 8 Horizontal Permeability Diagram

The horizontal and vertical permeability were determined, and then compared with finite element analysis [Phase2 6.0 software]. He finally got the ratio of vertical to horizontal permeability is around 3.2. He also found that, permeability decreases with increase of saturation. Masad et al. (17) used SEEP/W finite element software to determine permeability of pavement.

PREDICTING FIELD PERMEABILITY FROM LABORATORY TESTING

Kanitpong et al. (18) performed many field tests for permeability and laboratory tests of laboratory compacted sample of the same mix design to relate field and lab permeability values. He also correlated field permeability by field test method with field permeability by lab testing. He gave the following correlation of field permeability and lab permeability of field cores.

16

𝑃8 > 45%, 𝐹𝑖𝑒𝑙𝑑 𝐾 = 8.34 × 𝑒0.055(𝐿𝑎𝑏 𝑘) 𝑅2 = 0.80 (14) 40 < 𝑃8 < 45%, 𝐹𝑖𝑒𝑙𝑑 𝐾 = 21.68 × 𝑒0.039(𝐿𝑎𝑏 𝑘) 𝑅2 = 0.49 (15)

where P8 is the percent passing through No.8 sieve. No correlation was found for other mixtures. To determine correlation between field permeability with lab compacted sample permeability, he described two methods.

Method A

First, bulk specific gravity (Gmb) of field core is determined. Then amount of material needed to produce a laboratory sample of same height and density were determined using the following equation:

𝑊𝑡 = 𝐺𝑚𝑏 × 𝑡 × 𝐴1000

(16) where Wt = amount of material in gm;

t = height or thickness of the field cores (mm); A = cross section area of the specimen (mm2).

The SGC sample was compacted using the Superpave Gyratory Compactor, fixing the height. The permeability of these samples were predicted and found less than that obtained from field cores. And the relation is described by,

𝑦 = 22.24 + 0.734𝑥 (17)

where y = lab permeability of field cores; x = lab permeability of SGC sample.

Method B

Here mixes are collected from field and Ndes was fixed for all samples. So, densities of the samples were different. Then, permeability of the samples was determined using ASTM D5084. Permeability vs. density curve was plotted for SGC samples of method B. From this plot, permeability corresponding to field density was determined. Finally, predicted permeability and measured permeability of field cores were plotted on a graph which gives almost a straight line of slope 1. That indicates the permeability determined by method B was equal to the permeability of field cores.

STATE DOTS REVIEW

As required, a review of all 50 states DOTs is on-going for standards and specifications for permeability of Superpave mixes. The research team has preliminarily reviewed a number of DOTs specifications. Most of the works done by 49 other states on HMA permeability were studied to understand the importance of the works as well as the kind and nature of the research conducted. Standard specifications and different research conducted on Superpave permeability by 50 states including New Mexico were the main resource for this study. Fortunately, answers to several important questions were achieved during this literature review.

17

Results of Study

From this study, it is found that 18 states have problems on their pavements due to high permeability. To solve this problem, 4 DOTs already included permeability in their standard specifications, 10 DOTs already have their own methods and instruments to test for permeability, 12 DOTs performed test in the field for permeability and 13 DOTs performed tests in the lab to determine permeability and its relation with other factors such as air voids, density, lift thickness, NMAS etc. The upper permeability limit set by most of the DOTs is 125E-5 cm/s and the lower limit is 0 cm/s.

18

METHODOLOGY AND SURVEY

EXPERIMENTAL PLAN

The experimental plan for this study is shown in Figure 9. The field survey was conducted in cooperation with New Mexico Department of Transportation (NMDOT). The main objective of the field survey was to identify a set of bad (moisture damaged) and good (undamaged) performing asphalt pavement sections.

Field permeability tests were performed on those sections using an NCAT field asphalt permeameter. In New Mexico, most of the pavements are surfaced with open graded friction course (OGFC). Permeability testing and field coring was conducted on the shoulder, which is not typically treated with OGFC.

FIGURE 9 Experimental Design

The Florida apparatus using the falling head permeability test method determined laboratory permeability values of full depth cores. Ideally, full depth laboratory permeability should be compared to field permeability. The cores were separated by layers using a laboratory wet saw. Permeability tests were performed on the samples representing individual layers.

Moisture damage on samples separated by layers was evaluated by a ratio of the indirect tensile strength of a set of wet samples to that of a set of dry samples. In this study, wet conditioning was performed by two methods: AASHTO T 283 and a recently developed MIST device. The difference between AASHTO T283 and MIST conditioning is that AASHTO T 283 causes

Survey: Selection of Candidate Pavements

Field Permeability Test and Coring

Laboratory Testing

Laboratory Testing on Full Depth Cored Samples Samples Separated into Layers

Laboratory Tests for Permeability

Dry Conditioning

AASHTO T 283 MIST

Good Performing

Bad Performing

IDT

Moist Conditioning

IDT

19

damage to the AC sample by freeze-thaw action and MIST causes damage to the AC sample by increasing and releasing pore pressure inside the sample. In both cases, temperature is 60 °C.

SURVEY

The survey that was developed and presented during 1st quarter of this project was distributed to 42 NMDOT employees from all six of the NMDOT districts. The survey was followed by phone calls to spur the response. Below are the two questions in the survey. A. Is there a section (or project) of a roadway/pavement that shows moisture wicking in your

district? Basically, do you know about a pavement section that does not get dried up in a few hours (say couple of hours) after a short rainfall? If such, what is causing the above problem (high permeability, stripping, roadway drainage problem, moisture damage, etc.)?

B. Can you mention couple of pavement sections that are performing very well in terms of permeability or drainage or moisture damage or stripping issue? Milepost or location will be helpful.

Survey Response Synthesis and Data Collection

Except for District 1 and 4, all districts answered the survey questions. Based on their answers, pavements with high permeability or showing visual stripping are classified as bad performing pavement. On the other hand, pavements with low permeability or no visual stripping are grouped as good performing pavements. The list of good and bad performing sections is shown in Table 3.

TABLE 3 Participation of Districts from NMDOT.

District Low Permeability/ Good performing Pavement

High Permeability / Bad performing Pavement

District 2 US 285 MP 115-MP 205 US 70 MP 268 - MP 301 District 3 I-40 from Coors to Unser I-25 from south of Budaghers

north to Santa Fe county line (District 3 boundary)

District 6 C.N. ESG5B66, US 491, M.P. 59.0 – 67.7, San Juan County

I-40 mile markers 18 – 22

C.N. G1436ER, I-40, M.P. 126.2 – 130.7, Cibola County

C.N. 6100430, NM 264, M.P. 10.6 – 13.1, McKinley County, Roadway Rehab with WMA

Dist. 5 US 285, MP 284 – 290 NM 14, MP 47 - 50

US 84, MP 233 – 238 NM 344, NM 2 – 14

20

THIS PAGE LEFT BLANK INTENTIONALLY

21

PERMEABILITY TESTING

INTRODUCTION

As the permeability values of New Mexico pavements are not known, it is necessary to determine them. Permeability is determined both in the field and in the laboratory. These values are compared with specification limit set by different states. These values are also used to determine a permeability specification for New Mexico pavements.

FIELD TESTING PROGRAM

Based on the survey, field permeability tests were performed on the following locations,

TABLE 4 Field permeability test locations.

Pavements Location US285 MP126.23

MP140.53 MP152

MP285.25 MP285.5

US70 MP289.26 MP282.2 MP272.67

US491 MP60.9 MP60.5 MP60.7

I40 MP335.5 US264 MP10 MN14 MP46.80NB

MP46.80SB MP46.9SB

NM344 MP1.80 MP1.82 MP1.84

US84 MP235.8 MP235.9 MP236

I40 MP23.1 The procedure for the field testing followed by the UNM research team includes a field inspection of drainage, measurement and sketch of the lane markings and characteristics of the section including width of lanes, shoulder, wheel path location and distance from the shoulder stripe. The drainage is determined by the measurement of the slope and inspection of the crown

22

design. The team then lays down a template around the tape marks from the FWD testing that the NMDOT team has performed at the section location. The holes are marked with a construction crayon and the team determines which will be used for field testing and proceeds to testing. After the tests are performed, the NMDOT team cores the 24 holes that were marked by the UNM research team. 12- 6 in. diameter cores and 12- 4 in. diameter cores were taken and cataloged by the NMDOT team. Figure 10 shows the field permeability test set up. Figure 11 shows schematics of a section from District 2 that was tested March 14-15, 2012.

(a) Pavement marked to test permeability (b) Field permeability testing

FIGURE 10 Field Permeability Test Set Up

The NCAT field asphalt permeameter is used to determine field permeability value using the following formula:

𝑘 = 𝑎𝐿𝐴𝑡𝑙𝑛 �ℎ1

ℎ2� (18)

where k = coefficient of permeability, cm/s;

a = inside cross-sectional area of the standpipe, cm2; L = lift thickness of HMA, cm; A = base area of the permeameter, cm2; t = elapsed time between h1 and h2, s; h1 = initial head, cm; h2 = final head, cm. All parameters are shown in Figure 12.

23

FIGURE 11 Field Permeability Test Summary for US285 MP 140.53

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

District-2 Section-2 US-285 MP-140.53 Test date: 03/13/2012 Color: Ash Not Aged No Crack No Damage Direction: Southbound Coming from: Vaughn Going to: Roswell Shoulder Slope: 0.126 Road Slope: 0.021 OGFC: Yes

6˝Field Core Locations

4˝ Field Core Locations

Location of field permeability testing

55˝ 21˝ 6˝ 27˝ 9˝ 9˝ 21˝

Traffic Direction

Test Number Core Location

01 01

02 03

03 08

04 17

05 22

06 24

Results:

Permeability = 1.14x10-2 cm/s

Shou

lder

24

FIGURE 12 Schematic of Falling Head Permeability Method

In the field, a crack-free spot is selected on which to place the permeameter. This area is then cleaned using a broom. Next, a wax ring is placed at the base of the permeameter, which is then placed on the cleaned area. The base is pushed downward so that the wax ring is attached to the pavement without any water leaking. Three 10 lb. weights are then placed on the permeameter base to resist it from uplift pressure by the water. Water is poured inside the tube and it is allowed to flow for a few minutes to saturate the pavement. Figure 10(b) shows a field permeability test set-up. Though Eq. (18) is used for field permeability value, flow occurs both in vertical and lateral directions in the field. An impermeable or incompatible middle layer can direct the flow along the lateral direction. The permeability test was conducted around noon time. It is possible that a thermal gradient exists along the thickness of the Hot Mix Asphalt (HMA) pavements, which may affect the flow. However, no temperature correction is considered for field permeability calculations.

Special Field Testing Issue

Due to a few observations and issues experienced during the initial field testing, a few modifications have been made to the plans for field testing going forward. Testing on section 2 of the first pavement in District 2 was conducted on an HMA that had recently received a new OGFC and other rehabilitation. This section was very porous and exhibited characteristics of low friction. The first tests that were attempted led the research team to believe that the testing would be unsuccessful due to the OGFC. The OGFC was cut out by the NMDOT drill core team and the UNM team chipped it out in order to test the permeability of the underlying HMA. While this method allowed the team to test the section and get data that can be used, it is the goal of the team to maintain consistent methodology and obtain results that are reliable and may not be discarded later; therefore, no further sections was selected where a recent OGFC or any other rehab of the pavement has occurred. Further tests were performed on shoulders.

Core Collection and Selection of Candidate Pavements for Laboratory Testing

A total of 24 cores were collected from each of the 23 locations. The cores were inspected to determine whether stripping existed at the interface of the layers. Depending on the visual

Sample area, A

h

h

Tube area, a

Length, L

25

inspection, the pavements were classified as good or bad performing as shown in Table 5. The number assigned to each pavement and number of layer each pavement has are also shown in this table.

TABLE 5 Selected pavements for laboratory testing.

Good Performing Pavements (Not showing Moisture Damage)

Bad Performing pavements (Showing Moisture Damage)

Pavement location ID Number of layers

Pavement location ID Number of layers

US285 MP126.3 1 2 US70 MP289.26 9 3 US285 MP285.25 2 3 US70 MP282.2 10 3 NM344 MP1.80 3 1 US70 MP272.67 11 3 NM14 MP46.8 4 3 NM264 MP10 12 1 US491 MP60.7 5 3 US491 MP60.9 13 3 US491 MP60.5 6 3 US285 MP140.53 14 2 NM14 MP46.9 7 3 NM344 MP1.82 15 1 NM344 MP1.84 8 2 I40 MP23.1 16 3

LABORATORY PERMEABILITY TESTS

The Florida apparatus was used for laboratory permeability measurement. It is repeatable, available, and easy to use. Flow in this permeameter is one-dimensional. It uses the same equation as described during field permeability testing. A correction factor for water temperature is used in Eq. (18). Figure 13 shows the permeability test set-up in the laboratory. Both ends of the cored sample were smoothed using a wet saw. Before placing a sample in the laboratory permeameter, it was saturated using the CorelokTM device. In the CorelokTM, a sample is vacuum sealed in a plastic bag, which is cut open under water to saturate the sample. This method has shown to be more effective than flask-vacuum saturation (19). Petroleum jelly is applied on the curve surface of the sample to make lateral surface impermeable and to have a good seal so that no water flow through the interface of sample and permeameter. The sample was then placed inside a cylinder enclosed by a membrane and pressurized to confine the side so that water flows in vertical direction only. The temperature of the water was recorded. The water level was recorded at different time intervals.

TEST RESULTS

At each location, six permeability tests are performed. The average of six permeability values is calculated as shown in Table 6. From each location, cores are collected. Three full depth and nine samples separated at layers are tested for permeability. The average permeability is also determined as shown in Table 6. Table 6 shows data for pavement section 9 only. Similar calculations are made for all other pavements.

26

FIGURE 13 Laboratory Permeability Test Set-up

TABLE 6 Calculation of average permeability.

Field Permeability Tests

Pavement Section Test No Permeability(cm/s) Average Standard Deviation

9

1 48.1E-5

160.1E-5

92E-5

2 201.1E-5 3 92.4E-5 4 260.8E-5 5 100.4E-5 6 257.7E-5

Laboratory Permeability of Full Depth Samples

Pavement Section Test No Permeability(cm/s) Average Standard Deviation

9

1 5.46E-06

5.38E-05 6.1E-5

2 7.22E-06 3 3.41E-06 4 1.35E-04 5 1.27E-04 6 4.54E-05

27

Laboratory Permeability of Sample Cut into Layers Pavement Section Layer Test No Permeability(cm/s) Average Standard

Deviation

9

Top

1 5.2E-5

32.5E-5 19.3E-5

2 13.6E-5 3 47.0E-5 4 31.6E-5 5 21.9E-5 6 58.6E-5 7 55.6E-5 8 17.1E-5 9 42.3E-5

Middle

1 37.2E-5

36.4E-5

0.5E-3

2 57.6E-5 3 169.3E-5 4 3.9E-5 5 1.2E-5 6 0.7E-5 7 10.0E-5 8 44.0E-5 9 3.5E-5

Bottom

1 .9E-6

92.9E-5 127E-5

2 9.4E-5 3 66.7E-5 4 192.6E-5 5 383.2E-5 6 136.2E-5 7 3.6E-5 8 33.5E-5 9 11.2E-5

Field Permeability Test Results

The field permeability values range from 0 to 1208 × 10-5 cm/s with an average of 178 × 10-

5 cm/s as shown in Figure 14. The maximum allowable one dimensional permeability is 125 × 10-5 cm/s, according to many DOTs. Only 5 out of 16 pavements’ permeabilities are higher than that limit. Therefore, New Mexico pavements are not too permeable.

28

FIGURE 14 Field Permeability Results

Laboratory Permeability Test Results

Full Depth Sample

Figure 15 shows the average laboratory permeability of the full depth sample. Permeability of almost all samples is almost equal to zero. The permeameter available in the market cannot be used for a full depth sample as the samples are too long. Therefore a custom made permeameter is used. Only two pavements have very high permeability, as they are single layered. Excluding this two pavements, other pavements have average permeability of 3.15 × 10-5 cm/s.

FIGURE 15 Permeability of Full Depth Samples

Samples Separated at Layers

Samples are separated at the interface using a wet saw and permeability of each layer is determined and shown in Figure 16. It is observed that 8 of the 16 pavements have higher permeability of top layer than other layers. The average permeability of top layer is 65 × 10-5

0

200

400

600

800

1000

1200

1400

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Perm

eabi

lity

(x10

-5 c

m/s

)

Pavement sections

05

101520253035404550

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Perm

eabi

lity

(x 1

0-5 cm

/s)

Pavement sections

29

cm/s, middle layer is 56 × 10-5 cm/s and bottom layer is 35 × 10-5 cm/s. Only 4 out of 39 mixes have permeability higher than 125 × 10-5 cm/s.

FIGURE 16 Permeability of Different Layers

Comparison of Permeability at Different Test Modes

Average permeability values of full depth samples, samples separated by layers and field results were compared for all roads and shown in Figure 17. In most cases, field permeability > top layer permeability > full depth permeability. In the field, water moves in all directions whereas in the lab, the flow is one-dimensional. For the full depth sample, the interface and discontinuity of interconnected voids between layers retards the flow. For a few locations, field permeability is less than lab permeability. In these locations, the middle and bottom layers are almost impermeable. This chokes the vertical flow. Hence, field permeability here is caused only by lateral flow. For the single layered sample, the field and the laboratory tests gave almost identical results. The full depth sample has the lowest and field permeability has the highest value, however, no correlation between them can be made.

FIGURE 17 Comparison of Permeabiliity at Different Test Modes

050

100150200250300350400450

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

Perm

eabi

lity

(x10

-5 c

m/s

)

Pavement sections

Top LayerMiddle LayerBottom Layer

0

50

100

150

200

250

1 2 3 4 5 6 7 15 9 10 11 12 13 14 14 16Perm

eabi

lity

(x10

-5 c

m/s

)

Road sections

Field Permeability(cm/s) Lab k top layer (cm/s) Lab k full depth k (cm/s)

30

GYRATORY COMPACTED SAMPLES

SP III and SP IV mixes of New Mexico were used to prepare gyratory compacted samples. A total of 21 samples of 2 in. height were prepared and tested for permeability. Air voids ranges from 4 to 8% were used. It is observed that, samples with air voids less than 6% have zero permeability as shown in Figure 18. All pavements are compacted to air voids of 4 to 7%. With time, they are expected to be compacted more by wheel loading. All pavements are supposed to be impermeable; however, they are not impermeable. The gyratory compacted samples do not represent the field in terms of connectivity of air voids. To determine pavement’s permeability, tests on field collected cores is the only option.

FIGURE 18 Permeability vs Air Voids of Gyratory Compacted Samples

CONCLUSIONS

The following conclusions can be made from this study:

• The average field permeability of selected New Mexico pavements is 178 × 10-5 cm/s which is little bit more than specification limit set by other pavements (125 × 10-5 cm/s). Twelve out of sixteen pavements have permeability less than 125 × 10-5 cm/s.

• Laboratory permeability of full depth samples is very low because of different interfaces and discontinuity of flow line. Only two single layered pavements have very high permeability. In terms of permeability, it is better to compact a pavement in multiple layers compared to single layer.

• For layered samples, a number of 39 mixes were tested. Only 4 out of 39 mixes have permeability less than 125 × 10-5 cm/s.

• New Mexico pavements have permeability within tolerable limit.

0102030405060708090

100

0 2 4 6 8 10 12 14

Perm

eabi

lity

(×10

-5 c

m/s

)

Air voids (%)

31

THIS PAGE LEFT BLANK INTENTIONALLY

32

CORRELATION OF PERMEABILITY WITH VOLUMETRICS AND PORES

INTRODUCTION

Water flow occurs through the pores of s AC sample. Therefore, permeability should depend on pores: both on quantity and connectivity. Past studies show a poor correlation between permeability and total pores. The correlation is better for effective pores. The correlation between permeability and permeable pores is not known and this is the main objective of this study. The three types of pores (total, effective and permeable) are further used to determine the three distinct types of pores: permeable, dead-end and isolated.

MIX VOLUMETRIC AND PERMEABILITY

Mix Volumetric

The core samples were tested in the laboratory for bulk specific gravity (Gmb), maximum specific gravity of loose mix (Gmm), asphalt content (% AC) by ignition oven, total pores and gradation. Total pores vary from 6% to 10% and asphalt content varies from 4% to 7% as shown in Table 7. Mixes from all locations have 19 mm Nominal Maximum Size (NMAS) aggregate. The power charts are shown in Figure 19(a) to (d). Most of the mixes have 19 mm NMAS aggregate. Good performing sections have an almost equal portion of the power gradation curve above and below the maximum density line. Bad performing sections have more portions below the maximum density line. Therefore, good performing pavement sections contain more fine aggregates than bad performing pavement sections. Bad performing pavements maintain a regular “S” shape whereas for good performing pavements the lower portion of the curve becomes parallel to the maximum density line. This indicates, for good sections, the proportion of fine aggregates is equal to the proportion that produces maximum density.

TABLE 7 Mix data for selected pavement.

Pavement Sections Gmb Gmm VA AC Pavement Sections Gmb Gmm VA AC

US285 MP 126.3TL 2.37 2.44 3.07 4.54 US70 MP272.67ML 2.22 2.45 9.25 4.04

US285 MP 126.3BL 2.19 2.37 7.60 4.93 US70 MP272.67BL 2.33 2.41 3.33 3.31

US285 MP285.25TL 2.41 2.58 6.39 6.13 NM264 MP10TL 2.32 2.47 6.05 4.61

US285 MP285.25ML 2.43 2.58 5.78 5.80 US285 MP140.53TL 2.27 2.47 8.10 6.40

US285 MP285.25BL 2.41 2.52 4.18 6.05 NM344 MP1.82TL 2.26 2.50 9.77 5.93

NM344 MP1.80TL 2.28 2.51 9.03 5.30 US285 MP152TL 2.25 2.45 7.88 5.94

NM14 MP46.8TL 2.28 2.49 8.30 5.60 US285 MP152ML 2.22 2.43 8.77

33

Pavement Sections Gmb Gmm VA AC Pavement Sections Gmb Gmm VA AC

NM14 MP46.8ML 2.33 2.44 4.42 5.50 US285 MP152BL 2.20 2.45 10.13 4.64

NM14 MP46.8BL 2.32 2.47 6.33 5.30 US491 MP60.5TL 2.21 2.47 10.36 5.42

US491 MP60.7TL 2.34 2.47 5.19 5.60 US491 MP60.5ML 2.31 2.44 5.29 5.50

US491 MP60.7ML 2.33 2.48 6.27 5.00 US491 MP60.5BL 2.31 2.44 5.29 5.51

US491 MP60.7BL 2.33 2.47 5.85 5.80 US491 MP60.9TL 2.33 2.46 5.35 5.51

NM344 MP1.84TL 2.31 2.46 6.10 4.63 US491 MP60.9ML 2.33 2.47 5.61 5.44

NM344 MP1.84BL 2.35 2.46 4.65 6.07 US491 MP60.9BL 2.34 2.44 4.02 5.63

US70 MP289.26TL 2.28 2.45 7.13 6.60 NM14 MP46.8SBTL 2.30 2.435 5.71 6.62

US70 MP289.26ML 2.23 2.46 9.29 6.99 NM14 MP46.8SBML 2.33 2.51 7.17 4.56

US70 MP289.26BL 2.22 2.45 9.23 7.13 NM14 MP46.8SBBL 2.29 2.47 7.39 5.79

US70 MP282.2TL 2.24 2.45 8.59 4.13 NM14 MP46.9SBTL 2.16 2.51 13.70 2.68

US70 MP282.2ML 2.21 2.46 10.30 4.10 NM14 MP46.9SBML 2.33 2.44 4.51 5.49

US70 MP282.2BL 2.29 2.43 5.97 4.52 NM14 MP46.9SBBL 2.33 2.46 5.12 6.55

US70 MP272.67TL 2.24 2.45 8.64 4.72 US285 MP140.53BL 2.08 2.39 12.97 5.67

Here, TL = top layer, ML = middle layer and BL = bottom layer

Good performing pavements

12.5 9.5 4.75 2 0.425 0.075 0

10

20

30

40

50

60

70

80

90

100

Perc

ent p

assi

ng

Sieve size raised to 0.45 power (a) 12.5 mm NMAS

Max density line

US285MP126.3BL

US285MP140.53BL

US491MP60.9ML

US491MP60.9BL

19 12.5 9.5 4.75 2 0.425 0.075

0

10

20

30

40

50

60

70

80

90

100

Perc

ent p

assi

ng

Sieve size raised to 0.45 power (b) 19 mm NMAS

Max density lineUS285MP126.3TLUS285MP140.53TLUS285MP285.25TLUS285MP285.25MLUS285MP285.25BLUS491MP60.9TLUS491MP60.5TLUS491MP60.5MLUS491MP60.5BLUS491MP60.7TLUS491MP60.7MLUS491MP60.7BLNM14MP46.8TLNM14MP46.8MLNM14MP46.8BLNM14MP46.9TLNM14MP46.9MLNM14MP46.9BL

34

Bad performing pavements

FIGURE 19 Gradation Curve for All Pavement Sections

DIFFERENT TYPE OF PORES

Definition

The total air voids in a sample is termed as total pores. It consists of three distinct types of pores: permeable pores, isolated pores, and dead end pores. Pores that continue from top to bottom of the sample are termed as permeable pores, as they convey water from one end of the sample to the other end. Dead end pores start from one end or permeable pore of the sample but don’t travel to the other end. Therefore, these pores are accessible by water but water inside these voids doesn’t move. Isolated pores have no connection to permeable pores or dead end pores or to any side of the sample. They are not accessible by water. Permeable and dead end pores are accessible by water and termed together as effective pores. These pores are shown in Figure 20.

FIGURE 20 Different Types of Pores Inside A AC Sample

12.5 9.5 4.75 2 0.425 0.075 0

10

20

30

40

50

60

70

80

90

100Pe

rcen

t pas

sing

Sieve size raised to 0.45 power (c) 12.5 mm NMAS

Max density line

US264MP10TL

19 12.5 9.5 4.75 2 0.425 0.075 0

10

20

30

40

50

60

70

80

90

100

Perc

ent p

assi

ng

Sieve size raised to 0.45 power (d) 19 mm NMAS

Max density line

US70MP289.26TL

US70MP289.26ML

US70MP289.26BL

US70MP282.2TL

US70MP282.2ML

US70MP282.2BL

US70MP272.67TL

US70MP272.67ML

US70MP272.67BL

NM344MP1.82TL

NM344MP1.8TL

NM344MP1.84

25

Permeable pores

Isolated pores

Dead end pores

Effe

ctiv

e po

res

35

Determination of Effective and Total Pores

Bulk specific gravity of a cored sample is determined using the CorelokTM device as per ASTM D 6752. The dry weight of the sample is measured. The sample is placed inside a plastic bag of known weight and specific gravity and placed inside the CorelokTM device. Air from the Corelok chamber is sucked out and the polybag is sealed as shown in Figure 21. The underwater weight of the sample with polybag is measured. The bag is cut open under water and the weight is recorded again. Bulk specific gravity of the sample is calculated from Eq. (19):

𝐺𝑚𝑏 = 𝐴𝐴+𝐷−𝐵−𝐷/𝐹

(19)

where A = sample dry weight in air; D = polybag weight in air; B = weight of sealed sample in water; F = specific gravity of polybag at 25 °C.

Apparent specific gravity is determined using Eq. (20):

𝐺𝑚𝑎 = 𝐴𝐴+𝐷−𝐶−𝐷/𝐹

(20) where C = weight of sealed sample and polybag cut under water.

(a) Vacuum sealed sample inside a polybag (b) Weight of sealed sample under water

FIGURE 21 Bulk Specific Gravity of a Cored Sample

Effective pores are determined using Eq. (21) as follows:

𝑛𝑒 = 𝐺𝑚𝑎−𝐺𝑚𝑏𝐺𝑚𝑎

(21)

After all tests are done on a sample, it is placed inside the oven to prepare a loose mix. Maximum specific gravity is determined using a CorelokTM device as per ASTM D6857. The weight of loose mix and polybag is measured. The loose mix is placed inside the polybag and

36

vacuum sealed. The bag is cut open under water and water saturates the mix. The weight is measured. Maximum specific gravity is then determined by Eq. (22):

𝐺𝑚𝑚 = 𝐴𝐴+𝐷−𝐶−𝐷/𝐹

(22)

where A = loose mix dry weight in air; D = polybag weight in air; C = weight of saturated sample and polybag in water; F = specific gravity of polybag at 25 °C.

Finally, total pores are determined using Eq. (23) as follows:

𝑛 = 𝐺𝑚𝑚−𝐺𝑚𝑏𝐺𝑚𝑚

(23)

Permeable Pores by Tracer Method

The laboratory setup for the tracer test is shown in Figure 22. Essentially, a salt concentration-measuring meter was added to a falling head permeameter. Before placing a sample in the laboratory permeameter, it was saturated using the CorelokTM device. The sample was then placed inside a cylinder enclosed by a membrane and pressurized to confine the sides so that water flows in vertical direction only. Salt water of a known concentration was poured inside the standpipe. The salt concentration of outflow was measured using a salt concentration meter. Constant head was maintained by continuously pouring water at inflow and discharge was measured.

FIGURE 22 Tracer Test Set-up

Sample

Tube

Salt-meter

37

Darcy’s velocity is determined using Eq. (24) and permeability is determined using Eq. (25):

𝑞 = 𝑄𝐴

(24)

𝑘𝑙 = 𝑞𝐿ℎ

= 𝑄𝐿𝐴ℎ

(25)

where q = Darcy’s velocity, cm/s; Q = discharge rate, cm3/s; A = cross sectional area of the sample, cm2; kl = laboratory permeability, L = length of the sample, cm; h = head, cm.

For the one-dimensional tracer, the outflow concentration was determined by Eq. (26) (20):

𝐶𝐶0

= 12

[1 ± 𝑒𝑟 𝑓 � 𝐿−𝑣𝑡2𝑠𝑞𝑟𝑡(𝐷𝑡)�] (26)

where erf = an error function;

c = outflow salt concentration, %; c0 = inflow salt concentration, %; v = mean velocity of the tracer, cm/s; t = time, s.

c/c0 = 0.5 occurrs when one pore volume of solution passed through the sample. The time required for c/c0 = 0.5 is known as break through time, tb. Laboratory permeable pores (npl) can be calculated from Eq. (27) (21):

𝑛𝑝𝑙 = 𝑡𝑏 𝑄𝐴𝐿

(27) The terms of this equation are defined before.

Numerical Example 1

To determine permeability and permeable pores: For a core sample collected from New Mexico Highway 344 (NM344) (MP1.80), the output from the salt concentration meter is shown in Figure 23. Salt concentration ratio c/c0 is plotted against time. The ratio increases as time passes. Data is fitted to a sigmoidal ‘S’ shape curve. Breakthrough time from the graph can be found to be 58 s. The measured Q is 0.645 cm3/s. Sample diameter, length and constant head are 14.2 cm, 8.1 cm, and 53.5 cm respectively.

38

FIGURE 23 Breakthrough Curve for a Sample from NM344 at MP1.8.

Thus, plugging these values in Eq. (25) and Eq. (27): 𝑘𝑙 = 𝑄𝐿

𝐴ℎ= 0.645×8.1

158×53.5= 61.8×10-5 cm/s

𝑛𝑝𝑙 = 𝑡𝑏 𝑄

𝐴𝐿= 58×0.645

158×8.1 = 2.915%

SELECTION OF CANDIDATE PAVEMENT

For moisture damage and permeability analysis, sixteen pavements were used. Among them, nine pavement sections have high permeability. For the tracer test, the sample needs to be permeable. The permeable nine pavement sections shown in Table 8 were used for tracer testing and further analysis with this method.

TABLE 8 Pavement sections selected for tracer test.

Pavement Sections Number of layer used (Top to bottom)

US285 MP140.53 1

US70 MP289.26 2

US70 MP 282.2 2

US70 MP272.67 2

US491 MP60.5 1

NM14 MP46.9 1

NM344 MP1.8 1

NM344 MP1.82 1

NM344 MP1.84 1

0

0.1

0.2

0.3

0.4

0.5

0.6

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80

Salt

conc

entra

tion

ratio

, C/C

0

Time (s)

39

RESULTS AND DISCUSSIONS

Different Pores and Their Correlation

For each location, three samples were tested for permeable pores (npl) in the laboratory using the tracer method. Average value of the three samples were determined and correlated to effective pores (ne) and total pores (n). As shown in Figure 24(a), the solid line represents the correlation between permeable pores and total pores and the dotted line represents the correlation between permeable pores and effective pores. Overall, npl doesn’t show good correlation with n, because total pores contain isolated and dead-end pores. The relationship between npl and ne is better as ne doesn’t include isolated pores. The two regression lines are almost parallel as shown in Figure 24(a). That is, the rate of increase of npl with total pore and effective pores are the same. Figure 24(b) shows the effect of sample size on permeable pores. It is observed that all samples have higher permeable pore for 6 in. diameter than 4 in. diameter. As total pores stays constant, other type of pores decrease with increase of sample size. There is a good linear relation between npl1 and npl2. The dotted line in this figure indicates npl1 = npl2 (equal line). All points lay below the equal line indicating npl1 > npl2.

(a) Correlation of permeable pore with effective or total pores

(b) Correlation between permeable pores

FIGURE 24 Correlation between Different Types of Pores

R² = 0.3804 R² = 0.6118

02468

101214

0 2 4 6 8 10 12 14 16

n pl (

%)

n or ne (%)

npl vs. nnpl vs. ne

R² = 0.8299

0

2

4

6

8

10

12

14

0 2 4 6 8 10 12 14

n pl2

(%)

npl1 (%)

40

Correlation of Different Pores with Permeability

Figure 25 shows correlation of permeability with total pores (n), effective pores (ne) and permeable pores (np). Permeability shows very poor correlation with total pores as it includes isolated and dead-end pores, which has no contribution to flow. Effective pores show a better correlation with permeability. As permeability is directly controlled by permeable pores, there exists a good correlation between permeability and permeable pores.

FIGURE 25 Correlation of Different Pores with Permeability

Quantification of Different Types of Pores

The three distinct type of pores of AC samples from all 9 locations are evaluated and shown in Figure 26. It is observed that all samples have more permeable pores than the other two types of pores. It is worth mentioning that all 9 locations were selected based on high permeability. The pore quantity may be different for low permeable samples. As we see from the Figure 26, some of the samples have higher amount of isolated pores (i.e. US70 MP 289.26TL). For samples with high pore content, i.e. US70 MP272.67BL, isolated or dead end pores become negligible.

FIGURE 26 Pore Quantification

R² = 0.2298 R² = 0.5332

R² = 0.6677

0

500

1000

1500

2000

2500

3000

0 2 4 6 8 10 12 14

k l (

x10-5

cm

/s)

n, ne or npl

n vs. kl ne vs. kl npl vs. kl

0.002.004.006.008.00

10.0012.0014.00

Diff

eren

t typ

e of

por

es (%

)

Pavement sections

isolated porespermeable poresdead-end pores

41

CONCLUSIONS

The study presented in this section can be summarized as below:

• The permeable pores of 6 in. diameter samples are higher than 4 in. diameter samples. That is, permeable pores increase with the increase of sample radius.

• Permeable pores show better correlation with effective pores than with total pores. • Permeability has a better correlation with permeable pores rather than effective or total

pores. • Permeable pores are the maximum among all pores for highly permeable samples. The

scenario may be different for low permeable samples, which has not been done in this study.

42

THIS PAGE LEFT BLANK INTENTIONALLY

43

CORRELATION BETWEEN FIELD AND LABORATORY PERMEABILITY

INTRODUCTION

Permeability is determined both in the field and laboratory. However, it may be difficult to do a field permeability set up due to inadequate field crew, traffic control or bad weather. A field permeability test may give the wrong prediction because of inadequate saturation, OGFC or different coating at the interface. Also, the length of flow path in a field test is unknown. Cores are collected and transported to the laboratory for density testing immediately after the compaction. It is easy to determine permeability of field cores in the laboratory during density testing. It is necessary to check if there exists any correlation between field or laboratory permeability or to develop a method to determine field permeability by testing cores in the laboratory.

THEORETICAL DEVELOPMENT OF FIELD PERMEABLE PORES AND PERMEABILITY

Total pores in an asphalt concrete sample can be divided into eight broad categories as shown in Figure 27(a). They are, a = pores that continue from top to bottom, b = dead-end pores at the top of the sample, c = pores that continue from top to side, d = pores that continue from bottom to side, e = dead-end pores at the bottom, f = pores that continue from side to side, g = dead-end pores at the side, and h = isolated pores. When a one-dimensional permeability test is performed on such sample (Figure 27(a)), water flow occurs because of pores shown as “a”, that is, the pores that continue from top to bottom of the sample. In the field, a fraction of “c” contributes to the flow through lateral flow. Therefore, permeable pores responsible for flow in the field are:

𝑛𝑝𝑓 = 𝑎 + 𝑓(𝑐) (28) where npf = field permeable pores. Assuming a field core with radius r has some permeable pores (a and c) at the top surface and they are distributed uniformly. If β = is the percent of curve surface pores that travel to top. Then, pores (Al) that continue from the side to the top of the sample can be determined using Eq. (29):

𝐴𝑙 = 2𝜋𝑟𝐿𝛽 (29) where L is the sample length. Therefore, pores responsible for vertical flow (Apl ) can be determined by the following Eq. (30):

𝐴𝑝𝑙 = 𝜋𝑟2𝛼 − 2𝜋𝑟𝐿𝛽 (30)

44

where α = percent of the surface top area occupied by c and a pores, πr2α = flow area through c and a pores. In volumetric units (dividing Eq. (30) by πr2), laboratory permeable pores are:

𝑛𝑝𝑙 = 𝛼 − 2𝛾𝑟

(31) where npl = laboratory permeable pores and γ = Lβ, a factor. Solving Eq. (31) for two different sample sizes yields:

𝛼 = 𝑟1𝑛𝑝𝑙1−𝑟2𝑛𝑝𝑙2𝑟1−𝑟2

(32)

𝛾 = (𝑛𝑝𝑙1−𝑛𝑝𝑙2)𝑟1𝑟22(𝑟1−𝑟2)

(33) Hall and Ng (22) have shown that the intensity of pores that continue from the side to the top of the sample is not uniformly distributed on the side. The intensity is higher near the top than the bottom. In this study, a triangular distribution of pore intensity may be assumed on the side of the sample (Figure 27(b)). In addition, β-pores near the top of the sample are less likely to continue to the bottom of the sample. In other words, β-pores near the top have small contribution to permeability. Therefore, permeable pore intensity can be assumed to be minimum (0.0) at the top and maximum (1.0) at the bottom. This distribution is shown in Figure 27(b) schematically. For a differential length dy as shown in Figure 27(b), lateral pores that continue from side to top can be expressed as:

𝑑𝐴𝑙 = �4𝜋𝑟𝛽𝐿

× 𝑦� 𝑑𝑦 (34)

The probability of these pores to continue to the bottom is:

𝑝 = 1𝐿

(𝐿 − 𝑦) (35)

Therefore, part of dAl that is permeable in case of field testing is:

𝑑𝐴𝑙𝑝 = 1

𝐿(𝐿 − 𝑦) × �4𝜋𝑟𝛽

𝐿× 𝑦� 𝑑𝑦 (36)

Integration of Eq. (36) yields the total portion of the lateral pores that continues from the top to the bottom and given below:

𝐴𝑙𝑝 = 2

3𝜋𝑟(𝛽𝐿) = 2

3𝜋𝑟𝛾 (37)

45

(a) Pores distribution inside a cored sample (b) β (left triangle) and the probability distribution (right triangle) on the side of the sample.

FIGURE 27 Different Types of Pores in A Sample

Therefore, the total permeable pores are (adding Eq. (30) with Eq. (37)):

𝐴𝑝 = 𝐴𝑝𝑙 + 𝐴𝑙𝑝 = 𝜋𝑟2𝛼 − 2𝜋𝑟𝛾 + 2

3𝜋𝑟𝛾 = 𝜋𝑟2𝛼 − 4

3𝜋𝑟𝛾 (38)

In the volumetric unit:

𝑛𝑝𝑓 = 𝛼 − 4𝛾3𝑟

(39)

where npf = field permeable pores, γ and r were defined previously.

Model 1

Assuming permeability increases exponentially with base e and permeable pores as an exponent, it can be written:

𝑘 ∝ 𝑒𝑚×𝑛𝑝 (40)

where m is a constant and np is permeable pores. m is determined by solving Eq. (40) for two different sample sizes as follows:

𝑘𝑙1𝑘𝑙2

= 𝑒𝑚×𝑛𝑝𝑙1

𝑒𝑚×𝑛𝑝𝑙2 𝑚 = 1

𝑛𝑝𝑙1−𝑛𝑝𝑙2𝑙𝑛 𝑘𝑙1

𝑘𝑙2 (41)

dy

y

p

L-y

4πrβ

1

a b c

d e

f

g h

46

where kl1 = laboratory measured permeability of a 6 in. diameter sample, kl2 = laboratory measured permeability of a 4 in. diameter sample cored from the 6 in. diameter sample, npl1 = laboratory measured permeable pores of 6 in. diameter sample, and npl2 = laboratory measured permeable pores of 4 in. diameter sample cored from the 6 in. diameter sample. Finally, field permeability (kf) can be predicted using Eq. (40) and (41):

𝑘𝑓𝑘𝑙

= 𝑒𝑚×𝑛𝑝𝑓

𝑒𝑚×𝑛𝑝𝑙

𝑘𝑓 = 𝑒𝑚(𝑛𝑝𝑓−𝑛𝑝𝑙)𝑘𝑙 (42)

Model 2

Assuming permeability varies exponentially with permeable pores as a base:

𝑘 ∝ 𝑛𝑝𝜇 (43)

where µ is a constant and np is permeable pores. μ is determined by solving Eq. (43) for two different sample sizes as follows:

𝑘𝑙1𝑘𝑙2

=𝑛𝑝𝑙1𝜇

𝑛𝑝𝑙2𝜇

µ =𝑙𝑛 (

𝑘𝑙1𝑘𝑙2

)

𝑙𝑛 (𝑛𝑝𝑙1𝑛𝑝𝑙2

) (44)

Finally, field permeability (kf) can be predicted by solving Eq. (43) for field and laboratory permeability and permeable pores:

𝑘𝑓𝑘𝑙

=𝑛𝑝𝑓𝜇

𝑛𝑝𝑙𝜇

𝑘𝑓 = �𝑛𝑝𝑓𝑛𝑝𝑙�µ𝑘𝑙 (45)

Numerical Example 2

For a sample from NM344 MP1.80: For a sample of radius r1 = 7.05 cm, 𝑘𝑙1 = 61.8 × 10−5 𝑐𝑚

𝑠 𝑎𝑛𝑑 𝑛𝑝𝑙1 = 2.915% (using the

procedure described in Numerical Example 1)

47

For a sample of radius r2 = 5 cm cored from previous sample, 𝑘𝑙2 = 47.4 × 10−5 𝑐𝑚

𝑠 𝑎𝑛𝑑 𝑛𝑝𝑙2 = 2.086% (as the procedure described in Numerical Example

1) Using Eq. (32) and (33): 𝛼 = 𝑟1𝑛𝑝𝑙1−𝑟2𝑛𝑝𝑙2

𝑟1−𝑟2= (7.05×2.915−5×2.086)

7.05−5= 4.94%

𝛾 = (𝑛𝑝𝑙1−𝑛𝑝𝑙2)𝑟1𝑟2

2(𝑟1−𝑟2)= (2.915−2.086)×7.05×5

2(7.05−5)= 7.125

Using Eq. (24), field permeable pores are: 𝑛𝑝𝑓 = 𝛼 − 4𝛾

3𝑟= 4.94 − 4×7.125

3×7.05= 3.59%

Model 1

The constant m is calculated using Eq. (41): 𝑚 = 1

𝑛𝑝𝑙1−𝑛𝑝𝑙2𝑙𝑛 𝑘𝑙1

𝑘𝑙2= 1

2.915−2.086ln �61.8

47.4� = 0.32

Therefore, field permeability can be predicted from Eq. (42): 𝑘𝑓 = 𝑒𝑚(𝑛𝑝𝑓−𝑛𝑝𝑙)𝑘𝑙 = 𝑒0.32(3.59−2.915) × 61.8 × 10−5 = 76.7 × 10−5 cm/s

Model 2

The quantity μ can be calculated from Eq. (44):

µ =𝑙𝑛 (

𝑘𝑙1𝑘𝑙2

)

𝑙𝑛 (𝑛𝑝𝑙1𝑛𝑝𝑙2

)=

𝑙𝑛 (61.847.4)

𝑙𝑛 (2.9152.086)

= 0.793

Using Eq. (45):

𝑘𝑓 = �𝑛𝑝𝑓𝑛𝑝𝑙�µ𝑘𝑙 = � 3.59

2.915�0.793

61.8 × 10−5 = 72.9 × 10−5 cm/s

Tests on two more samples were performed and average permeability of three of the samples was predicted as 391.5 × 10−5 cm/s which is much more than field permeability determined on that location (69.7 × 10−5 cm/s). Laboratory permeability for this location is 210.2 × 10−5 cm/s. Therefore, field permeability measurement gives an unsatisfactory (as it should be grater due to flow) and predicted permeability yields a satisfactory result (more than the laboratory

48

value indicating some contribution from lateral flow). A detailed discussion will be introduced in a later section.

RESULT AND DISCUSSIONS

Laboratory Measured and Predicted Permeability

Figure 28 describes the correlation between laboratory permeability (kl) with predicted field permeability (kf). The theory presented previously here can determine field permeability from measuring laboratory permeability alone. This will reduce the efforts needed for coring and measuring of np.

FIGURE 28 Laboratory and Predicted Permeability

Laboratory, Predicted and Field Measured Permeability

Figure 29 compares laboratory permeability, estimated permeability and field permeability. Field tests on the first four sections were performed on the driving lane, which was overlaid by an OGFC layer. Tests on the rest of the pavements were done on the shoulders without OGFC. As expected, measured field permeability results on the first four pavements are higher than the predicted permeability. For the other five pavements, field permeability results are lower compared to the estimated permeability. Measured field permeability values don’t show any regular relationship with laboratory permeability, but estimated permeability does. It can be stated that the permeability estimation from the newly developed theory gives better prediction of the field permeability of the compacted mix. To accommodate the field conditions (OGFC, coats e.g. tack coat, seal coat, etc.) a shift factor may be used with predicted permeability to make it equal to measured permeability. For pavements with OGFC, the ratio of measured to predicted permeability varies to a range of 4 and 22. Therefore, measured and predicted permeability cannot be correlated in this case. For pavements without OGFC, the ratio varies to a narrow range of 0.12 to 0.22 with an average of 0.17. A regression analysis on predicted and measure value by analysis of variance (ANOVA)

y = 1.3294x + 21.831 R² = 0.9723

0

500

1000

1500

2000

2500

0 500 1000 1500 2000 2500

k f (x

10-5

cm

/s)

kl (x10-5 cm/s)

kf vs. kl

49

gives an intercept of 9.43×10-5 cm/s with slope 0.132. The R2 value in this case 0.95 which is well enough. Finally, the modified permeability is predicted using this regression analysis and shown in Figure 38 by bar chart and line plot. The mean of error and standard deviation of error are 12.3% and 6.4% which are fair enough.

FIGURE 29 Measured and Predicted Permeability

CONCLUSIONS

Results of this study can be summarized as below:

• Field and laboratory permeability do not show any correlation. • The model proposed initially gives field permeability of an ideal pavement which has no

OGFC, fully saturated, single layer with no coating etc. In reality, no pavement is ideal therefore, the model cannot be used directly.

• A shift factor is used with the prediction which gives results very close to the measured permeability, but it works only for pavements without OGFC. For pavements with OGFC, the model does not work.

0100200300400500600700800900

1000

Perm

eabi

lity

(×10

-5 c

m/s

)

Pavement sections

Laboratory kPredicted kMeasured kModefied predicted k

Pavements with OGFC Pavements without OGFC

020406080

100120140160

0 20 40 60 80 100 120 140 160

Pred

icte

d pe

rmea

bilit

y (x

10-5

cm

/s)

Field measured permeability (x 10-5 cm/s)

50

THIS PAGE LEFT BLANK INTENTIONALLY

51

PERMEABILITY AND MOISTURE DAMAGE

INTRODUCTION

This section describes whether permeability is related to moisture damage or not. Permeability is determined as described in section four using the NCAT field permeameter and the Florida laboratory permeameter. Moisture damage is initially predicted by visual inspection of the field then visual stripping of the cores. Three mechanical approaches are used in the laboratory to determine moisture susceptibility. These methods are AASHTO T 283, Moisture Induced Susceptibility Test (MIST), and Hamburg wheel tracking test. All of these methods are compared among themselves and also correlated with different permeabilities.

LABORATORY DETERMINATION OF MOISTURE SUSCEPTIBILITY BY CONDITIONING

Moisture damage is measured by the ratio of the wet to the dry sample’s IDT. For dry conditioning, the dry sample was placed inside a Ziploc bag and placed under water at 25 °C (77 °F) for two hours. For wet conditioning, the AASHTO T 283 method and MIST devices were used.

AASHTO T 283 Wet Conditioning

In AASHTO T 283 method, an asphalt core is saturated using a vibro-deairator device that applies vibration and suction simultaneously. After saturation, the sample is placed inside a moist Ziploc bag and placed in a refrigerator for 16 hours at -23 °C (0 °F) for freezing. The sample is thawed in a hot water bath at a temperature of 60 °C (140 °F) for 24 hours followed by two hours of conditioning at 25 °C (77 °F) (24). Thus, in the AASHTO T 283 conditioning process, water is forced to enter inside the sample during saturation and to increase in volume during freezing. The increased volume of water causes increased pressure inside the pores of the sample causing damage. Thawing by hot water for 24 hours also contributes to the softening of the binder, mastic and samples.

MIST Wet Conditioning

In this method, a core sample is placed inside the MIST chamber filled with water as shown in Figure 30. A bladder inside the watertight chamber is used to increase and decrease the chamber pressure. In this study, the chamber temperature was set at 60 °C (77 °F) with a chamber pressure of 40 psi. A total of 3500 cycles of pressure increase and release were applied to the cored samples. As the number of cycles increase, the air inside the sample is replaced gradually with water. At certain cycles of intervals, air bubbles are released through the opening of the top of the chamber lid. Water inside the pores is pressurized to cause damage in the cores. After completion of MIST conditioning, the sample is placed under water at room temperature for about 2 hours for further conditioning.

52

FIGURE 30 Sample in the MIST Chamber

Tensile Strength Testing

Dry and wet conditioned samples are loaded diametrically to fail in tension as shown in Figure 31. The load is applied at 50 mm/min rate and the peak value of the load is recorded. The indirect tensile strength is calculated using Eq. (46):

𝐼𝐷𝑇 = 20𝑃𝜋𝐷𝐿

(46) where IDT = indirect tensile strength (kPa);

P = peak force needed to crack the sample diagonally, recorded from the compression testing device (Newton);

D = diameter of the sample (cm); L = length of the sample (cm).

FIGURE 31 IDT Testing in Progress

53

Tensile Strength Ratio (TSR) of wet to dry samples is calculated using Eq. (47):

𝑇𝑆𝑅 = 𝐼𝐷𝑇𝑤𝑒𝑡𝐼𝐷𝑇𝑑𝑟𝑦

(47)

where IDTwet = Average IDT of three wet conditioned samples and IDTdry = Average IDT of three dry conditioned samples.

HAMBURG WHEEL TRACKER TEST

Hamburg wheel tracker was first used in Germany in mid 70s. Cylindrical or slab samples are placed in the mold. Metal wheels of 158lbs weight roll over the sample for 20000 cycles. The device is shown in Figure 32. The rut depth is measured at different cycle intervals. After several cycles, the rut depth increases suddenly which is a measure of moisture damage susceptibility. Several states have their own specification for Hamburg wheel testing as shown in Table 9 and 10 for moisture damage inflection point and rutting potential respectively.

FIGURE 32 Hamburg Wheel Tracking Tests

TABLE 9 Moisture damage inflection point.

DOT Stripping inflection limit

Colorado 10000

California 5000 (Conventional binder)

10000 (Polymer modified binder)

TABLE 10 Rut depth limit.

DOT/ Agency Max. rut depth Number of cycles Hamburg, Germany 4 20000 Colorado 4 10000

54

DOT/ Agency Max. rut depth Number of cycles 10 20000 Texas 12.5 10000 (PG-64 or less) 12.5 15000(PG-70) 12.5 20000 (PG-76 or more) California 8 20000 (Conventional binder) 11 20000 (Polymer modified binder)

RESULTS AND DISCUSSIONS

Field Damage with Field Permeability

Individual and average field permeability values of pavement section 9 are shown in Table 11. Similar averaging is done for the other 15 pavement sections and is plotted in Figure 33. Figure 33(a) shows permeability of the good performing sections and Figure 33(b) shows permeability of the bad performing sections. It can be seen that the average permeability of the eight good performing sections is 62.7 × 10-5 cm/s which is much less than the average permeability of the bad performing pavements, 298 × 10-5 cm/s. This was expected. From Figure 33(a), it can be seen that pavement 5 has higher permeability than 125 × 10-5 cm/s, which is the required limit set by many DOTs (Ahmad and Tarefder 2013). From Figure 33(b), permeability of some pavement sections are less than 125 × 10-5 cm/s because field permeability depends on lot of factors suck as, air voids of each layer, continuity of continuous air voids through different layers, track coat, seal coat, base permeability, flow direction, etc. As for the example, if any layer below the top layer is impermeable, the permeameter will measure only the lateral flow yielding less permeability. There may be a higher permeable layer, though, and more or less field permeability of the individual pavement is not related to visual stripping of the pavement. In general, pavements with higher field permeability exhibit higher field damage in most of the cases.

FIGURE 33 Field Permeability at Different Damage Conditions

0

50

100

150

200

250

300

9 10 11 12 13 14 15 16

Perm

eabi

lity

(x10

-5 c

m/s

)

(b) Bad performing pavements

Ave. k = 298×10-5

cm/s

k= 125×10-5

0

50

100

150

200

250

300

1 2 3 4 5 6 7 8

Perm

eabi

lity

(x10

-5 c

m/s

)

(a) Good performing pavement sections

Ave k = 62.7E-5

k = 125×10-5 cm/s

55

Field Damage with Laboratory Permeability of Full Depth Samples

Individual and average laboratory permeability results of full depth samples of pavement section 9 are also shown in Table 9. Similarly, average permeability for all other pavement sections are determined and plotted in Figure 34. Figure 34(a) shows the full depth permeability of eight good performing pavement sections and Figure 34(b) shows the full depth permeability values for eight bad performing sections. It can be seen that the average permeability of the good performing sections is 4.15 × 10-5 cm/s, which is much lower than the average permeability of the bad performing sections (69.7 × 10-5 cm/s). This was expected. Because of the presence of the interface and the heterogeneous nature of the different layers, full depth permeability is very low and sometimes zero. One permeability value in Figure 34(a) is higher compared to other good performing sections, and permeability of some bad performing sections in Figure 34(b) is almost zero. This is unexpected. Therefore, pavements with higher full depth permeability undergo more in-situ damage with some exceptions.

FIGURE 34 Laboratory Full Depth Permeability at Different Damage Conditions

Field Damage with Laboratory Permeability of Samples Separated into Layers

The permeability values for samples separated into the layers of section 9 are shown in Table 11. Similarly, average permeability values for all other pavement sections are shown in Figure 35. Figure 35(a) shows the permeability values for good performing sections and Figure 35(b) shows the permeability values for bad performing sections. Three of the values are too high and are not shown fully. Otherwise, other bars will be invisible. The average of all permeability values for good performing sections is 22.9 × 10-5 cm/s, which is much lower than the average of all permeability values for bad performing sections (78.6 × 10-5 cm/s), as expected. All permeability values of good performing sections are less than 125 × 10-5 cm/s. Some permeability values of the bad performing sections are less than 125 × 10-5 cm/s. For good performing sections, four top layer permeability values are almost zero. For other sections, the top layer has the highest and middle layer has the lowest permeability. Here, the average permeability for the top layer is 39 × 10-5 cm/s, for the middle layer is 18.8 × 10-5 cm/s, and for the bottom layer is 7.4 × 10-5 cm/s. Pavements with zero permeability of the top layer, water can’t enter into the pavement, resulting

0102030405060708090

100

1 2 3 4 5 6 7 8

Perm

eabi

lity

(×10

-5 c

m/s

)

(a) Good performing sections

Ave. k =4.15×10-5cm/s 0

102030405060708090

100

9 10 11 12 13 14 15

Perm

eabi

lity

(×10

-5 c

m/s

)

(b) Bad performing sections

Ave. k = 69.7× 10-5 cm/s

56

in less interaction between the pavement and water and less damage. For higher top layer permeability, it acts as OGFC. It provides good drainage and less damage occurs. Most of the bad performing pavements have higher permeability values for all layers. The average permeability for the top layer is 72.8 × 10-5 cm/s, for the middle layer is 97.3 × 10-5 cm/s, and for the bottom layer is 70.6 × 10-5 cm/s. Here, water saturates the HMA and base very quickly after precipitation. This may damage the pavement severely due to the pumping action. Therefore, pavement with zero or higher top layer permeability compared to other layers exhibits less in-situ damage and pavement with high permeability to all layers exhibits more field damage.

FIGURE 35 Laboratory permeability of Separated Samples at Different Damage Conditions

MIST Damage with Permeability

The calculation for MIST TSR for pavement section 9 is shown in Table 11. The average TSR of all layers is determined to compare with field and laboratory full depth permeability. The TSR for all other pavement sections are calculated similarly and plotted in Figure 36. Figure 36(a) shows the relation between MIST TSR and field permeability, Figure 36(b) shows the relation

0

50

100

150

200

250

300

Top Layer Middle Layer Bottom Layer

Perm

eabi

lity

(×10

-5 c

m/s

)

(a) Different layers of good performing pavement section

1

2

3

4

5

6

7

8Ave. k =22.9×10-5 cm/s

k = 125 × 10-5 cm/s

0

50

100

150

200

250

300

Top Layer Middle Layer Bottom Layer

Perm

eabi

lity

(x10

-5 c

m/s

)

(b) Different layers of bad performing sections

9

10

11

12

13

14

15

16

Ave. k = 78.6 × 10-5 cm/s

k = 125 × 10-5 cm/s

57

between MIST TSR and laboratory full depth permeability, and Figure 36(c) shows the correlation between MIST TSR and samples separated into layers. None of the figures show good correlation between TSR and permeability. This can be explained as follows. MIST cyclic pressure acts not only inside the pore but also on the sides. When a less permeable sample is placed inside the MIST chamber at 60°C temperature with 40 psi pressure for 3500 cycles, it gets soft due to temperature and damaged due to cyclic pressure on the surfaces of the soft sample. Some samples show less permeability but more damage, which is unexpected. Therefore, MIST damage cannot be correlated with permeability.

TABLE 11 TSR calculation.

TSR Calculation for AASHTO T283 Conditioned Samples

Pavement Section

Dry Samples

ID

IDT (psi)

Average IDT (psi)

T 283 Conditioned samples ID

IDT (psi)

Average IDT (psi)

TSR Average

9

Top layer

1 239.09 190.36

4 197.32 146.31 0.77

0.82

2 175.68 5 115.70 3 156.33 6 125.92

Middle layer

1 189.31 189.23

4 141.41 132.66 0.7 2 185.44 5 138.20

3 193.93 6 118.39

Bottom layer

1 81.16 91.25

4 142.58 92.84 1 2 101.33 5 48.04

3 79.18 6 87.90

TSR Calculation for MIST Conditioned Samples

Pavement Section

Dry Samples

ID

IDT (psi)

Average IDT (psi)

MIST Conditioned samples ID

IDT (psi)

Average IDT (psi)

TSR Average

9

Top layer

1 239.09 190.36

7 157.57 168.64 0.89

0.87

2 175.68 8 166.31 3 156.33 9 182.03

Middle layer

1 189.31 189.23

7 109.01 135.12 0.71 2 185.44 8 130.25

3 193.93 9 166.10

Bottom layer

1 81.16 91.25

7 85.01 91.07 1 2 101.33 8 93.90

3 79.18 9 94.32

58

FIGURE 36 Correlation of MIST TSR with Permeability at Different Modes

AASHTO T 283 Damage with Permeability

The calculation for AASHTO T 283 TSR for pavement section 9 is shown in Table 11. Average TSR of all the layers is determined to compare with field and laboratory full depth permeability. TSR for all other sections are calculated similarly and plotted in Figure 37. Figure 37(a) shows the relation between AASHTO T 283 TSR and field permeability, Figure 37(b) shows the relation between AASHTO T 283 TSR and laboratory full depth permeability, and Figure 37(c) shows the correlation between AASHTO T 283 TSR and samples separated into layers. In case of AASHTO T 283, the increase of water volume in the voids due to freezing exerts internal pressure on the sample. The more permeable voids indicate that more water can get in. This eventually increases the internal pressure due to freezing and more damage occurs. Field permeability tests and permeability of full depth samples doesn’t show any relation with damage,

R² = 0.0886

00.10.20.30.40.50.60.70.80.9

1

0 50 100 150 200 250 300 350

MIS

T TS

R

(c) Permeability of layered samples (×10-5 cm/s)

R² = 0.0166

00.10.20.30.40.50.60.70.80.9

1

0 100 200 300 400 500

MIS

T TS

R

(a) Field permeability (×10-5 cm/s)

R² = 0.0811

00.10.20.30.40.50.60.70.80.9

1

0 10 20 30

MIS

T TS

R

(b) Laboratory full depth k (×10-5 cm/s)

59

as shown in Figure 37(a) and Figure 37(b) respectively. From laboratory permeability tests on layered samples, a good correlation of TSR with permeability was obtained and is shown in Figure 37(c). This was expected. Here, for zero permeability, TSR is around 0.80, decreasing to 0.6 for permeability 150 × 10-5 cm/s. Therefore, AASHTO T 283 TSR decreases with an increase of permeability and there exists a better correlation (R2 = 0.38) than MIST for laboratory testing on samples cut into layers.

FIGURE 37 Correlation of AASHTO T283 TSR with Permeability at Different Test Modes

MIST and AASHTO T283 Damage

Figure 38 compares TSR values obtained by AASHTO T 283 and MIST. MIST TSR values are higher than AASHTO T 283 TSR for most of the samples. For a few samples, AASHTO T283 TSR is higher than MIST TSR. This can be explained as follows: MIST is independent of permeability and AASHTO T 283 TSR increases with a decrease of permeability. Therefore, for

R² = 0.3808

00.10.20.30.40.50.60.70.80.9

1

0 50 100 150 200 250 300 350

TSR

Permeabillity (× 10-5 cm/s)

R² = 0.0349

0

0.2

0.4

0.6

0.8

1

0 100 200 300

AA

SHTO

T 2

83 T

SR

(a) Field permeability (×10-5 cm/s)

R² = 0.1721

0

0.2

0.4

0.6

0.8

1

0 5 10 15 20 25

AA

SHTO

T 2

83 T

SR

(b) Full depth permeability (×10-5 cm/s)

60

the less permeable sample, it is possible that the AASHTO T 283 TSR will be higher than the MIST TSR.

FIGURE 38 Comparison between AASHTO T 283 and MIST TSR

Field Damage with Laboratory Damage

Figure 39 shows the TSR values for good and bad performing sections. Figure 39(a) and 39(b) are for AASHTO T 283 conditioned samples and Figure 39(c) and 39(d) are for MIST conditioned samples. All good performing sections are supposed to have TSR values more than equal to 0.8 and bad performing sections should have less. However, the results shows TSR<0.8 for 11 mix from good performing sections and TSR>0.8 for 3 bad performing sections, in case of AASHTO T 283. A similar scenario for MIST is observed. This is unexpected. This happens because MIST and AASHTO T 283 sometimes give an inaccurate prediction. Permeable porosity might not be uniformly distributed all through the sample. If there is less porosity alone the line of loading during IDT, a higher TSR value might be obtained, although, higher damage at other locations of the sample might occur. The opposite is also possible. Therefore, the MIST or the AASHTO T 283 method does not give an accurate prediction of moisture susceptibility of HMA.

0

0.2

0.4

0.6

0.8

1

TSR

Pavement Sections ( TL = Top layer, ML = Middle Layer, BL = Bottom Layer)

T 283 TSR MIST TSR

00.10.20.30.40.50.60.70.80.9

1

Top layer Middle Layer Bottom layer

TSR

Different layers

1

2

3

4

5

6

7

80.000.100.200.300.400.500.600.700.800.901.00

Top layer Middle layer Bottom layer

TSR

Different layers

9

10

11

12

13

14

15

16

(a) AASHTO T283 TSR for good performing sections

(b) AASHTO T 283 TSR for bad sections

61

FIGURE 39 TSR Values for Good and Bad Performing Pavement Sections

Permeability and Moisture Damage of Laboratory Compacted Samples

Permeability and moisture damage relationships for laboratory compacted samples are shown in Figure 40(a). Neither of the MIST or AASHTO T 283 shows any correlation with permeability. AASHTO T 283 and MIST TSR values are compared in Figure 40(b). In case of field collected cores, MIST TSR values were always higher than AASHTO T 283 TSR. For, laboratory compacted samples, these two TSR values are almost similar. A paired t-test yields a p value 0.45. That is, for laboratory compacted samples, AASHTO T 283 and MIST cause the same amount of damage.

(a) Permeability vs. TSR for laboratory compacted samples

00.10.20.30.40.50.60.70.80.9

1

Top Layer Middle layer Bottom layer

TSR

Different layers

1

2

3

4

5

6

7

80.000.100.200.300.400.500.600.700.800.901.00

Top layer Middle layer Bottom layer

TSR

Different layers

9

10

11

12

13

14

15

16

R² = 0.0101

R² = 0.1089

0.00

0.20

0.40

0.60

0.80

1.00

1.20

0 5 10 15 20 25 30 35 40 45 50

TSR

Permeability (x 10-5 cm/s )

(c) MIST TSR for good pavement sections (d) MIST TSR for bad pavement sections

62

(a) Comparision of AASHTO T 283 and MIST TSR for laboratory compacted samples

FIGURE 40 Permeability and Moisture Damage of Labory Compacted Samples

Relation of Different Type of Pores with Moisture Damage

Permeable pores drain out water quickly. Hence, it may reduce the ‘moisture damage due to pumping action’. On the other hand, dead-end pores hold the water for a long time after precipitation. Therefore, they may have greater impact on moisture damage due to the pumping action. Isolated pores have no contribution on moisture or moisture related damage. Figure 41 shows the correlation of TSR (indication of moisture damage) with permeable pores and dead-end pores. As expected, moisture damage increases with the increase of dead-end pores and decreases with the increase of permeable pores. In all cases, the MIST line is above the AASHTO T 283 line, indicating less damage during MIST conditioning. The correlation is not good as other factors like asphalt type, aggregate type and gradation were not maintained constant.

FIGURE 41 Correlation of TSR with Dead End and Permeable Pores

0.00

0.20

0.40

0.60

0.80

1.00

1.20

1 2 3 4 5 6 7

TSR

Axis Title

AASHTO T 283

MIST

SP III SP IV

R² = 0.0393

R² = 0.1604

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

0 1 2 3 4

TSR

Dead-end pores (%)

MIST TSR vs. dead-end voids

AASHTO T 283 TSR vs. dead-end airvoids

R² = 0.5019

R² = 0.6122

0.30

0.40

0.50

0.60

0.70

0.80

0.90

1.00

4 6 8 10 12

TSR

Permeable pores (%)

MIST TSR vs. Pearmeable pores

AASHTO T 283 TSR vs. permeablepores

63

Hamburg Wheel Tracking (WHT) Tests

The plot for rut depth vs. number of wheel passes for 8 good performing pavements is shown in Figure 42. For each location, 4 samples were tested. On each sample, rut depth is measured at 6 equidistant locations. Each line in the plot is the average of these six readings. Table 12 summarizes the data obtained from the plots. It can be seen that all the good pavements satisfies the stripping potential criteria set by other DOTs (discussed in this section before). That is, the pavements are not susceptible to moisture damage. Only pavement section 3 fails to satisfy the rutting criteria. Similarly, plots for bad performing pavements are shown in Figure 43 and summarized in Table 13. It is observed that only one of the bad pavements has rutting and stripping potential. It is unexpected. Therefore, HWT doesn’t give a good prediction of a pavement’s moisture susceptibility.

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 2 >20000 2 1.6 >20000 3 3 >20000 4 4 >20000

Average 2.6 >20000 Ave. initial compaction = 1 mm

(a) Pavement section 1

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 10 >20000 2 15.5 >20000 3 15.5 17500 4 19.5 18000

Average 15.5 17500 Ave. initial compaction = 1.5 mm

(b) Pavement section 3

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 2.5 >20000 2 4 >20000 3 2.5 >20000 4 3.7 >20000

Average 3.1 >20000 Average initial compaction = 1 mm

-20-18-16-14-12-10-8-6-4-20

0 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

-20-18-16-14-12-10-8-6-4-20

0 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

-20-18-16-14-12-10-8-6-4-20

0 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

64

(c) Pavement section 4

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 6.3 >20000 2 5.3 >20000 3 2.8 >20000 4 2.8 >20000

Average 4.2 >20000 Average initial compaction = 2 mm

(d) Pavement section 5

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 5.9 >20000 2 11.4 >20000 3 9 >20000 4 7.2 >20000

Average 8.1 >20000 Average initial compaction = 6 mm

(e) Pavement section 6

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 7.8 >20000 2 11 >20000 3 10.5 >20000 4 8 >20000

Average 8.8 >20000 Average initial compaction = 3.8 mm

Comment: Sample 3 was neglected to find Ave.

(f) Pavement section 7

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 4.9 >20000 2 10 >20000 3 9.8 >20000 4 >20 >20000

Average 8 >20000 Average initial compaction = 1 mm

Comment: Sample 4 was neglected to find Ave.

(g) Pavement section 8 FIGURE 42 HWT Plots for Good Pavements

-20-18-16-14-12-10-8-6-4-20

0 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

-20

-15

-10

-5

00 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

-20-18-16-14-12-10-8-6-4-20

0 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

-20-18-16-14-12-10-8-6-4-20

0 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

65

TABLE 12 HWT data summary for good pavements.

Pavement Section

Initial Compaction

(mm)

Rut depth (mm)

Inflection point

Comment

1 1 2.6 >20000 ok 3 1.5 15.5 17500 Rut depth- not ok

Stripping potential –ok 4 1 3.1 >20000 Ok 5 2 4.2 >20000 Ok 6 6 8.1 >20000 Ok 7 3.8 8.8 >20000 Ok 8 1 8 >20000 Ok

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 5 >20000 2 5 >20000 3 5 >20000 4 17.5 >20000

Average 5 >20000 Average initial compaction = 1 mm

Comment: Sample 4 was neglected to find Ave.

(a) Pavement section 9

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 7 >20000 2 17.5 >20000 3 7.5 >20000 4 8 >20000

Average 7.5 >20000 Average initial compaction = 1 mm

Comment: Sample 2 was neglected to find Ave.

(b) Pavement section 10

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 2.3 >20000 2 3.3 >20000 3 4.6 >20000 4 7 >20000

Average 3.4 >20000 Average initial compaction = 1 mm

Comment: Sample 4 was neglected to find Ave.

-20

-15

-10

-5

00 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

-20

-15

-10

-5

00 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

-20

-15

-10

-5

00 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

66

(c) Pavement section 11

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 6 >10000 2 7 >10000 3 7 >10000 4 8.6 >10000

Average 7 >10000 Average initial compaction = 1 mm

(d) Pavement section 12

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 5 >20000 2 7.6 >20000 3 4.5 >20000 4 7.6 >20000

Average 6.1 >20000 Average initial compaction = 1 mm

(e) Pavement section 13

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 12 >20000 2 11 >20000 3 5 >20000 4 9.8 >20000

Average 8.2 >20000 Average initial compaction = 1 mm

Comment: Sample 1 was neglected to find Ave.

(f) Pavement section 14

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 7 >20000 2 10.8 >20000 3 11 >20000 4 12 >20000

Average 10.2 >20000 Average initial compaction = 1 mm

(g) Pavement section 15

-10-9-8-7-6-5-4-3-2-10

0 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of cycles

Sample 1 Sample 2Sample 3 Sample 4Average

-20

-15

-10

-5

00 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

-20-18-16-14-12-10-8-6-4-20

0 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

-20

-15

-10

-5

00 5000 10000 15000 20000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

67

Sample no.

Rut depth for 20000 passes (mm)

Inflection point (Passes)

1 >24 ≈5000 2 >12 >10000 3 >28 ≈5000 4 >24 ≈5000

Average >20 ≈5000 Average initial compaction < 1 mm

(h) Pavement section 16

FIGURE 43 HWT Plots for Bad Performing Pavements

TABLE 13 Summary of HWT test data for bad pavements.

Pavement Section

Initial Compaction

(mm)

Rut depth (mm)

Inflection point

Comment

9 1 5 >20000 Ok 10 1 7.5 >20000 Ok 11 1 3.4 >20000 Ok 12 1 7 >10000 Ok 13 1 6.1 >20000 Ok 14 1 8.2 >20000 Ok 15 1 10.2 >20000 Ok 16 1 >20 ≈5000 Fail

CONCLUSIONS

The following conclusion can be made from this section:

• Bad performing pavements have higher permeability than good performing pavements. Permeability increases the potential of moisture damage.

• MIST TSR doesn’t show any kind of correlation with permeability. AASHTO T 283 TSR shows a better correlation with permeability.

• MIST does less damage to a sample than AASHTOO T 283 does. • Moisture damage increases with the increase of dead-end pores and decrease with the

increases of permeable pores. • Hamburg wheel tracking does not show the stripping potential properly.

-30

-25

-20

-15

-10

-5

00 2000 4000 6000 8000

Rut

dep

th (m

m)

Number of passes

Sample 1 Sample 2Sample 3 Sample 4Average

68

THIS PAGE LEFT BLANK INTENTIONALLY

69

PERMEABILITY SPECIFICATION

Permeability increases with the increase of air voids with a higher gradient as shown in Figure 44. This is because increase in air voids also increases the number of interconnected air voids as well as the diameter of the flow path. Both of these increase the flow of water. The correlation between average permeability with air voids is almost similar and shown in Figure 45. The full depth permeability test yields a very small permeability value, which may be considered less significant and the plot is shown in Figure 45. For this plot, data from not only good or bad pavements but also all the permeability tests done on cores from New Mexico pavements are included (424 data points).

FIGURE 44 Correlation of Permeability with Air Voids

FIGURE 45 Correlation of Average Permeability with Average Air Voids

y = 3E-10e1.489x R² = 0.5279

00000E+0

200E-5

400E-5

600E-5

800E-5

1000E-5

0 2 4 6 8 10 12 14

Perm

eabi

lity

(cm

/s)

%air void

R² = 0.569

00000E+0

200E-5

400E-5

600E-5

800E-5

1000E-5

0 2 4 6 8 10 12

Perm

eabi

lity

(cm

/s)

% Air voids

70

FIGURE 46 Permeability Vs. Air Voids for Full Depth Samples

DEVELOPMENT OF PERMEABILITY SPECIFICATIONS

Critical permeability is determined based on critical air voids which is determined as follows: • Permeability vs. air voids are plotted on a graph. • Tangents are drawn at the initial and final part of the regression line of the plot. • A perpendicular is drawn from the intersection of the tangents to the regression line. • The point of intersection between regression line and the perpendicular line gives the

critical air voids. • Before critical air voids, an increase of air voids yields more isolated air voids than

interconnected air voids. Hence, permeability doesn’t increase that much. (with slope <1) • After the critical air void, increase of air void yields more interconnected air voids than

isolated air voids. Hence, permeability increases abruptly. • Permeability at this air void is the maximum allowable permeability or critical

permeability.

Figure 47 describes the determination of critical permeability for laboratory testing of field cores. Permeability corresponding to 4% and 7% air voids is also determined. The values are shown in Table 12. The same procedure is applied for average permeability, full depth permeability and field permeability. The critical value, upper limit and lower limit for average laboratory tests are shown in Table 13 and Figure 46, for full depth tests are shown in Table 14 and Figure 47. Field test results don’t give good correlation between permeability and air voids. Hence, critical permeability value from field results cannot be obtained.

R² = 0.5705

00000E+0

5E-5

10E-5

15E-5

20E-5

25E-5

30E-5

0 1 2 3 4 5 6 7 8 9 10

Perm

eabi

lity

(cm

/s)

% Air voids

71

FIGURE 47 Critical Permeability from Labortory Testing of Field Cores

FIGURE 48 Critical Permeability Values Obtained from Average Permeability Values

FIGURE 49 Critical Permeability For Full Depth Samples

72

TABLE 14 Different permeability limits for laboratory testing.

Condition % Air voids Permeability (cm/s)

Critical 10 90x10-5

Upper Limit 7 1x10-5

Lower Limit 4 1.15x10-7

TABLE 15 Critical permeability obtained from average permeability plot.

Condition % Air voids Permeability (cm/s)

Critical 9.2 125x10-5

Upper Limit 7 6.6x10-5

Lower Limit 4 1.8x10-7

TABLE 16 Critical permeability values for full depth sample

Condition % Air voids Permeability (cm/s)

Critical 7.6 6x10-5

Upper Limit 7 2x10-5

Lower Limit 4 8x10-8

Critical permeability values for Oklahoma, Virginia and Florida is 125x10-5 cm/s. The average permeability plot in this study gives the same value. Hence, critical permeability value for New Mexico is proposed 125x10-5 cm/s.

73

THIS PAGE LEFT BLANK INTENTIONALLY

74

CONCLUSIONS AND RECOMMENDATIONS

CONCLUSIONS

The following conclusions can be made from this study:

• Both field and laboratory permeabilities of New Mexico pavements are very close to specification limits. Field permeability is higher than laboratory permeability for pavement with OGFC. For pavements without OGFC, the scenario is opposite. The permeability of full length sample is almost zero. For laboratory compacted samples, permeability is zero for sample with less than 6% air void.

• The combination of a permeameter with a salt meter can be used to determine permeable pores of AC samples. Asphalt concrete sample’s permeability varies in radial direction due to increase in permeable pore in radial direction. Permeability shows a good relationship with permeable pores, rather than total pores.

• Field permeability is always higher than the laboratory permeability except for single layered pavements. Field permeability cannot be correlated with laboratory permeability on the basis of experiments only. Therefore, an analytical model is developed to determine field permeability by testing cores in the laboratory. The analytical model predicts field permeability of asphalt pavements reasonably well if the pavement is not covered with OGFC.

• All types of permeability of bad performing pavements are much higher than the permeabilities of good performing pavements. Hence, permeability increases the potential of moisture damage. MIST TSR shows a bad correlation with permeability. The AASHTO T 283 TSR shows a better correlation with permeability. Moisture damage increases with the increase of permeability.

• The maximum allowable permeability for New Mexico pavements is obtained as 125 × 10-5 cm/s, which is exactly equal to the permeability specification proposed by other states.

RECOMMENDATIONS

The following studies can be made further to establish this study: • Hamburg wheel tracker may be used to determine the moisture susceptibility of the mix. As it applies cyclic loading and wheel pressure similar to the field, this prediction may be more useful and comparable with other test methods. • A small single layer test pavement can be constructed over a highly permeable sand layer with the facility of saturation. Field permeability tests on those pavements may give a very

75

accurate prediction, as there is no OGFC, tack coat, seal coat or discontinuity of flow path. Laboratory permeability of the cores collected from this pavement can be used to verify the model proposed in this study.

76

REFERENCES

1. Mallick, R. B., L. A. Cooley, Jr., M. R. Teto, R. L. Bradbury, and D. Peabody. An Evaluation of Factors Affecting Permeability of Superpave Designed Pavements. NCAT Report 03-02, 2003.

2. Cooley, Jr., L. A., B. D. Prowell, and E. R. Brown. Issues Pertaining to the Permeability

Characteristics of Coarse Graded SuperPave Mixes. Annual Meeting of the Association of Asphalt Paving Technologists, 2002.

3. Tarefder, R. A., L. White, and M. Zaman. Neural Network Model for Asphalt Concrete

Permeability. Journal of Materials in Civil Engineering, Vol. 17, No. 1, pp. 19-27, 2005. 4. Bear, J. Dynamics of Fluids in Porous Media. American Elsevier Publishing Company.

Inc., 1972. 5. Hicks, R. G. Moisture Damage in Asphalt Concrete: Synthesis of Highway Practice,

NCHRP Report 175, 1991. 6. Mogawer, W.S., Mallick, Teto, M.R., and Crockford, W.C. Evaluation of Permeability of

Superpave Mixes. NETCR 34, New England Transportation Consortium, North Dartmouth, MA, Project No. NETC 00-2, 2002.

7. AASHTO Standard T 283. Standard Method of Test for Resistance of Compacted Hot

Mix Asphalt (HMA) to Moisture-Induced Damage. AASHTO Guide, Washington, DC, 2007.

8. Williams, S.G. A Comprehensive Study of Field Permeability Using the Vacuum

Permeameter. MBTC-2054, 2006. 9. Cooley Jr., L.A. Permeability Of Superpave Mixtures: Evaluation of Field Permeameters,

NCAT Report No. 99-1, 1991. 10. Retzer, N. Permeability Research With the ROMUS Air Permeameter. Colorado

Department of Transportation, Final Report, Report No. CDOT-2008-5, 2008. 11. Masad, E., B. Birgisson, A. Al-Omari, and A. Cooley. Analysis Of Permeability And

Fluid Flow In Asphalt Mixes. A Paper Submitted to the 82nd Annual Transportation Research Board for Presentation and Publication, 2002.

12. Willoughby, K. A., J. P. Mahoney, L. M. Pierce, J. S. Uhlmeyer, K. W. Anderson, S. A.

Read, S. T. Muench, T. R. Thompson and R. Moore 2001 “Construction-Related Asphalt Concrete Pavement Temperature Differentials and the Corresponding Density Differentials” Research report for Washington DOT.

77

13. Brown, E. R., M. R. Hainin, A. Cooley, and G. Hurley. Relationship of Air Voids, Lift Thickness, and Permeability in Hot Mix Asphalt Pavements. NCHRP Report 531, 2004.

14. Kutay, M.E., Aydilek, A.H., Masad, E., and Harman, T. Computational and Experimental

Evaluation of Hydraulic Conductivity Anisotropy in Hot Mix Asphalt. International Journal of Pavement Engineering, Vol. 8, Issue 1, pp. 29-43, 2007.

15. Al-Omari, A., L. Tashman, E. Masad, A. Cooley, T. Harman. Proposed Methodology for

Predicting HMA Permeability. J. Assoc. Asphalt Pav. Technology, 71, 2002. 16. Harris, C.H. Hot Mix Asphalt Permeability: Tester Size Effect and Anisotropy. MS

thesis, Virginia Polytechnic Institute and State University, 2007. 17. Masad, E., Castelblanco, A., and Birgisson, B. Effects of air void size distribution, pore

pressure, and bond energy on moisture damage. J. of Testing and Evaluation, 34(1), 15–23, 2006.

18. Kanitpong, K., H. Bahia, J. Russell, and R. Schmitt. Predicting Field Permeability from

Testing Hot-Mix Asphalt Specimens Produced by Superpave Gyratory Compactor. Journal of Transportation Research Board, No.1929, pp. 52-58, 2005.

19. Tarefder, R. A., M. M. Zaman, and K. Hobson. Evaluating the CoreLokTM Measurement

of Bulk Specific Gravity for Hot Mix Asphalt Samples. ASTM Journal of Testing and Evaluation, Vol. 30, No. 4, pp. 274-282, 2002.

20. Stephens, D. B., K. Hsu, M. A. Prieksat, M. D. Ankeny, N. Blandford, T. L. Roth, J. A.

Kelsey, and J. R. Whitworth. A Comparison of Estimated and Calculated Effective Porosity. Hydrology Journal, Vol.6, pp. 156-165, 1998.

21. Jury, W.A., and R. Horton. Soil Physics. 6th Edition, John Wiley and Sons, Inc., 2004. 22. Hall, K. D., and H. G. Ng. Development of Void Pathway Test for Investigating Void

Interconnectivity in Compacted Hot-Mix Asphalt concrete. Journal of Transportation Research Board, Volume 1767, pp. 40-47, 2001.

78

THIS PAGE IS LEFT BLANK INTENTIONALLY

79

APPENDIX A

FIELD PERMEABILITY TEST RESULTS

Field permeability test results from all 16 pavements sections are presented in this appendix.

80

TABLE A.1 Field permeability test results.

Road Section Test No Permeability(cm/s) US 285 MP126.3 1 9.22E-06

2 3.07E-06 3 9.7E-5 4 40.6E-5 5 139.0E-5 6 96500.0E-10 7 119.0E-5 MP140.53 1 819.0E-5 2 815.0E-5 3 815.0E-5 4 898.0E-5 5 2610.0E-5 6 1290.0E-5 MP285.25 1 00000.0E+0 2 00000.0E+0 3 00000.0E+0 4 00000.0E+0 5 00000.0E+0 6 00000.0E+0 MP285.5 1 00000.0E+0 2 00000.0E+0 3 00000.0E+0 4 00000.0E+0 5 00000.0E+0 6 00000.0E+0 US70 MP282.2 1 717.6E-5 2 721.0E-5 3 591.1E-5 4 601.1E-5 5 848.3E-5 6 221.8E-5 MP289.3 1 48.1E-5 2 201.1E-5 3 92.4E-5 4 260.8E-5 5 100.4E-5 6 257.7E-5 US491 MP60.5 1 150.3E-5 2 68.2E-5

81

3 00000.0E+0 4 00000.0E+0 5 00000.0E+0 6 82.0E-5 MP60.7 1 115.0E-5 2 138.2E-5 3 112.2E-5 4 123.0E-5 5 169.2E-5 6 154.9E-5 MP60.9 1 105.7E-5 2 50.5E-5 3 52.1E-5 4 18.0E-5 5 30.6E-5 6 45.5E-5 US264 MP10 1 51.7E-5 2 90.7E-5 3 41.6E-5 4 77.5E-5 5 103.2E-5 6 208.4E-5 I40MP335.5 DL 1 11.2E-5 2 17.2E-5 3 16.4E-5 4 11.8E-5 5 12.5E-5 6 16.7E-5 Shoulder 1 345.2E-5 2 1230.0E-5 3 761.0E-5 4 257.7E-5 5 137.5E-5 6 137.5E-5 I40MP142 1st Lift DL WB 1 3625.4E-5 2 1878.4E-5 3 1799.4E-5 4 3356.3E-5 5 1736.0E-5 6 1331.1E-5 1st lift S WB 1 4476.3E-5 2 5391.1E-5 3 4369.5E-5

82

4 3467.0E-5 5 4562.3E-5 6 4436.5E-5 2nd Lift DL WB 1 9007.3E-5 2 12400.8E-5 3 10575.1E-5 4 11003.8E-5 5 12841.2E-5 6 12331.6E-5 2nd lift S WB 1 2055.34E-05 2 1673.94E-05 3 1483.32E-05 4 1922.74E-05 5 1681.20E-05 6 2502.31E-05 3rd lift DL left WB 1 13367.92E-05 2 12620.65E-05 3 13865.87E-05 4 12212.73E-05 5 12706.17E-05 6 12037.43E-05 3rd lift DL right WB 1 22994.87E-05 2 24925.70E-05 3 26227.24E-05 4 19937.57E-05 5 37365.94E-05 6 22152.85E-05 Full depth EB 1 3939.3E-5 2 3027.5E-5 3 2671.7E-5 4 1772.1E-5 5 2213.0E-5 6 3497.2E-5 NM14 MP146.8 1 1.02E-03 2 8.02E-04 3 5.83E-04 4 1.49E-03 5 7.04E-04 6 7.05E-04 MP146.9 1 8.53E-05 2 2.49E-04 3 2.35E-04 4 1.48E-04

83

5 0.00E+00 6 4.97E-04 MP146.8SB 1 1.05E-03 2 1.56E-03 3 1.88E-03 4 9.41E-04 5 5.82E-03 6 5.92E-04 NM344 MP1.80 1 82.7E-5 2 82.5E-5 3 46.8E-5 4 46.4E-5 5 65.4E-5 6 94.2E-5 MP1.82 1 58.1E-5 2 70.7E-5 3 43.8E-5 4 22.5E-5 5 94.7E-5 6 24.9E-5 MP1.84 1 33.1E-5 2 14.9E-5 3 24.7E-5 4 122.6E-5 5 28.2E-5 6 57.8E-5 US84 MP235.8 1 1.29E-04 2 1.47E-03 3 3.33E-04 4 1.84E-03 5 1.10E-03 6 7.95E-04 MP235.9 1 6.26E-04 2 9.67E-04 3 2.08E-04 4 2.38E-04 5 1.30E-04 6 2.76E-04 MP236 1 2.37E-05 2 2.80E-04 3 2.29E-05 4 4.27E-04 5 8.38E-05

84

6 1.16E-04 I40 MP23.1 1 1.62E-04 2 0.00E+00 3 0.00E+00 4 6.56E-05 5 0.00E+00 6 6.60E-05

85

APPENDIX B

LABORATORY FULL DEPTH SAMPLE TEST RESULT

Laboratory permeability tests results on full depth samples from all pavements are shown in this appendix.

86

TABLE B.1 Laboratory permeability test results on full depth samples.

Road Section Sample ID Permeability(cm/s) US285 MP285.25 2 0.00E+00 3 0.00E+00 8 0.00E+00 9 0.00E+00 13 0.00E+00 20 0.00E+00 MP140.53 1 159.4E-5 19 176.5E-5 20 438.4E-5 MP152 21 4.25E-05 MP126.23 1 0.00E+00 3 0.00E+00 8 0.00E+00 11 0.00E+00 US70 MP272.67 1 0.00E+00 2 0.00E+00 7 0.00E+00 19 0.00E+00 20 0.00E+00 21 0.00E+00 MP289.26 1 5.46E-06 2 7.22E-06 3 3.41E-06 8 1.35E-04 15 1.27E-04 24 4.54E-05 MP282.2 13 0.00E+00 19 6.21E-05 14 .6516E-5 US491 MP60.9 1 0.00E+00 2 0.00E+00 7 9.54E-06 9 9.45E-05 14 0.00E+00 18 0.00E+00 MP60.5 7 0.00E+00 8 0.00E+00 9 0.00E+00 14 0.00E+00 15 2.04E-05

87

MP60.7 7 7.74E-05 8 0.00E+00 9 0.00E+00 13 0.00E+00 14 0.00E+00 15 0.00E+00 I40 MP335.5S 1 151.7E-5 2 177.0E-5 3 275.9E-5 4 167.0E-5 5 181.5E-5 MP335.5DL 9 10.6E-5 10 27.1E-5 11 4.1E-5 13 18.0E-5 14 6.5E-5 US264 MP10 11 1.03E-06 16 7.70E-07 17 3.37E-06 18 3.50E-06 20 4.33E-06 23 5.80E-06 NM344 MP1.8 1 7.38E-04 3 1.66E-03 8 5.88E-04 14 3.83E-03 15 1.05E-02 21 4.55E-03 MP1.82 1 3.17E-04 2 2.09E-05 7 3.56E-04 8 2.93E-04 9 4.94E-05 21 3.29E-04 MP1.84 7 5.65E-05 8 3.67E-06 9 4.05E-06 15 6.77E-06 19 3.39E-05 20 8.14E-05 NM14 MP46.80NB 2 6.21E-05 3 1.51E-05 7 0.00E+00

88

9 0.00E+00 13 0.00E+00 15 5.99E-05 MP46.80SB 3 0.00E+00 7 0.00E+00 9 0.00E+00 8 0.00E+00 15 0.00E+00 13 0.00E+00 MP46.9SB 2 0.00E+00 3 0.00E+00 7 0.00E+00 15 0.00E+00 20 0.00E+00 21 0.00E+00

89

APPENDIX C

LABORATORY PERMEABILITY OF SAMPLES CUT INTO LAYERS

Laboratory permeability test results on samples separated into layers of different mixes are presented in this appendix.

90

TABLE C.1 Permeability test results of layered samples.

Road Section Layer Sample ID

Permeability (cm/s)

US285 MP140.53 Top 2 241.6E-5 3 384.5E-5 4 193.6E-5 5 218.0E-5 7 282.2E-5 8 363.6E-5 9 353.2E-5 10 570.7E-5 11 597.1E-5 12 533.2E-5 13 331.8E-5 14 278.9E-5 15 348.2E-5 16 193.7E-5 17 236.8E-5 18 344.1E-5 21 146.0E-5 22 302.3E-5 23 122.2E-5 24 363.9E-5 Bottom 2 293.3E-5 4 213.2E-5 7 597.0E-5 8 587.1E-5 14 202.5E-5 15 212.5E-5 16 73.3E-5 17 98.2E-5 21 525.8E-5 22 270.1E-5 MP 152 Top 1 32.3E-5 2 32.5E-5 4 23.1E-5 5 47.1E-5 6 44.2E-5 7 81.5E-5 8 178.6E-5

91

9 159.0E-5 10 128.2E-5 11 56.6E-5 12 31.0E-5 13 7.7E-5 14 .7E-5 15 1.6E-5 16 .9E-5 17 1.1E-5 19 44750.1E-10 20 .5E-5 23 1.3E-5 24 1.1E-5 Middle 1 73.1E-5 2 36.8E-5 4 67.2E-5 5 87.9E-5 6 127.2E-5 7 26.8E-5 8 46.1E-5 9 41.6E-5 10 77.6E-5 11 239.0E-5 12 104.9E-5 13 59.0E-5 14 44.0E-5 15 35.4E-5 16 62.2E-5 17 84.4E-5 19 37.8E-5 20 19.4E-5 23 50.9E-5 24 150.4E-5 Bottom 1 257.3E-5 2 143.7E-5 4 295.6E-5 5 191.4E-5 6 100.5E-5 7 55.7E-5 8 355.1E-5 9 109.0E-5 10 194.8E-5 11 322.5E-5

92

12 703.8E-5 13 719.1E-5 14 605.1E-5 15 806.4E-5 16 1011.7E-5 17 275.9E-5 19 1722.2E-5 20 1655.1E-5 23 1839.5E-5 24 1305.4E-5 MP126.23 Top Layer 15 00000.0E+0 17 00000.0E+0 18 00000.0E+0 19 00000.0E+0 20 00000.0E+0 21 00000.0E+0 22 00000.0E+0 23 00000.0E+0 24 00000.0E+0 Bottom

Layer 15 00000.0E+0

17 00000.0E+0 18 33.6E-5 19 97820.8E-10 20 00000.0E+0 21 64800.3E-10 22 30072.6E-10 23 19.2E-5 24 19.2E-5 MP285.25 Top Layer 4 00000.0E+0 5 00000.0E+0 6 00000.0E+0 12 00000.0E+0 16 00000.0E+0 17 00000.0E+0 22 00000.0E+0 23 00000.0E+0 24 00000.0E+0 Middle

Layer 4 00000.0E+0

5 00000.0E+0 6 00000.0E+0 12 00000.0E+0 16 00000.0E+0

93

17 00000.0E+0 22 00000.0E+0 23 00000.0E+0 24 00000.0E+0 Bottom

Layer 4 00000.0E+0

5 00000.0E+0 6 00000.0E+0 12 00000.0E+0 16 00000.0E+0 17 00000.0E+0 22 00000.0E+0 23 00000.0E+0 24 00000.0E+0

US70 MP289.26 Top 4 5.2E-5 7 13.6E-5 10 47.0E-5 13 31.6E-5 14 21.9E-5 16 58.6E-5 20 55.6E-5 22 17.1E-5 23 42.3E-5 Middle 4 37.2E-5 7 57.6E-5 10 169.3E-5 13 3.9E-5 14 1.2E-5 16 .7E-5 20 10.0E-5 22 44.0E-5 23 3.5E-5 Bottom 4 .9E-6 7 9.4E-5 10 66.7E-5 13 192.6E-5 14 383.2E-5 16 136.2E-5 20 3.6E-5 22 33.5E-5 23 11.2E-5 MP282.2 Top 1 43.3E-5 2 32.8E-5

94

3 63.7E-5 4 39.7E-5 5 165.7E-5 7 247.4E-5 8 206.1E-5 9 524.7E-5 11 114.3E-5 15 18.3E-5 16 6.1E-5 18 7.7E-5 20 9.7E-5 21 61.5E-5 22 44.0E-5 23 24.3E-5 24 34.8E-5 Middle 1 722.1E-5 2 644.6E-5 3 332.3E-5 4 382.5E-5 5 918.5E-5 7 735.8E-5 8 539.0E-5 9 857.5E-5 11 620.9E-5 15 111.7E-5 16 19.3E-5 18 155.6E-5 20 222.2E-5 21 87.0E-5 22 222.5E-5 23 56.0E-5 24 104.4E-5 Bottom 1 00000.0E+0 2 8.4E-7 3 1.5E-7 4 00000.0E+0 5 00000.0E+0 7 00000.0E+0 8 00000.0E+0 9 00000.0E+0 11 00000.0E+0 15 111.7E-5 16 00000.0E+0

95

18 19.5E-5 20 2.7E-5 21 1.9E-6 22 00000.0E+0 23 00000.0E+0 24 2.7E-5 MP272.67 Top Layer 4 51.5E-5 5 30.2E-5 10 9.2E-5 11 49.5E-5 17 29.0E-5 18 8.9E-5 23 5.5E-5 24 13.1E-5 Middle

Layer 4 16.4E-5

5 60.5E-5 10 27.4E-5 11 00000.0E+0 17 72.6E-5 18 1226.6E-5 23 978.1E-5 24 1616.3E-5 Bottom

Layer 4 00000.0E+0

5 3.8E-5 10 00000.0E+0 11 51.3E-5 17 00000.0E+0 18 00000.0E+0 23 00000.0E+0 24 00000.0E+0

US491 MP60.9 Top Layer 5 18.9E-5 6 48.8E-5 10 63.9E-5 11 00000.0E+0 16 00000.0E+0 18 00000.0E+0 22 00000.0E+0 23 00000.0E+0 24 1.5E-5 Middle

Layer 5 00000.0E+0

6 00000.0E+0

96

10 00000.0E+0 11 00000.0E+0 16 00000.0E+0 18 00000.0E+0 22 00000.0E+0 23 00000.0E+0 24 00000.0E+0 Bottom

Layer 5 10.2E-5

6 3.5E-5 10 17.2E-5 11 8.6E-5 16 1.8E-5 18 00000.0E+0 22 00000.0E+0 23 00000.0E+0 24 7.4E-5 MP60.5 Top Layer 5 153.3E-5 6 252.4E-5 10 165.5E-5 12 120.9E-5 16 1508.9E-5 17 300.0E-5 18 474.8E-5 22 1587.4E-5 24 830.6E-5 Middle

Layer 5 2.5E-5

6 00000.0E+0 10 2.7E-5 12 00000.0E+0 16 27.4E-5 17 00000.0E+0 18 3.1E-5 22 34.2E-5 24 14.8E-5 Bottom

Layer 5 00000.0E+0

6 00000.0E+0 10 00000.0E+0 12 00000.0E+0 16 00000.0E+0 17 00000.0E+0 18 8.41E-05

97

22 00000.0E+0 24 00000.0E+0 MP60.7 Top Layer 4 00000.0E+0 5 00000.0E+0 6 00000.0E+0 10 00000.0E+0 11 00000.0E+0 12 00000.0E+0 16 21.2E-5 17 44.1E-5 23 218.0E-5 Middle

Layer 4 00000.0E+0

5 00000.0E+0 6 00000.0E+0 10 11294.6E-10 11 139.4E-5 12 191.3E-5 16 177.5E-5 17 34.1E-5 23 3.3E-5 Bottom

Layer 4 00000.0E+0

5 99528.5E-10 6 1.8E-5 10 81.5E-5 11 00000.0E+0 12 50.2E-5 16 5.4E-5 17 3.9E-5 23 1.4E-5

US264 MP10 Middle Layer

1 22.0E-5

2 3.6E-5 3 40.9E-5 8 21.7E-5 13 4.8E-5 15 9649.1E-10 19 00000.0E+0 20 00000.0E+0 21 00000.0E+0

I40 MP335.5DL Top Layer 9 151.7E-5 10 27.1E-5 11 4.1E-5

98

13 18.0E-5 14 6.5E-5 MP335.5S Top Layer 1 151.7E-5 2 177.0E-5 3 275.9E-5 4 167.0E-5 5 181.5E-5

NM14 MP46.8NB Top Layer 4 0.00E+00 6 0.00E+00 10 0.00E+00 11 1.00E-05 12 0.00E+00 17 8.19E-04 18 1.17E-03 22 2.90E-04 24 9.27E-04 Middle

Layer 4 2.73E-05

6 0.00E+00 10 2.98E-05 11 4.57E-05 12 2.98E-04 17 0.00E+00 18 1.16E-04 22 3.32E-04 24 3.95E-04 Bottom

layer 4 1.65E-05

6 0.00E+00 10 0.00E+00 11 8.87E-06 12 0.00E+00 17 3.35E-05 18 9.06E-05 22 3.90E-05 24 3.95E-04 MP46.8SB Top Layer 4 2.66E-05 5 6.98E-05 6 0.00E+00 11 16 6.96E-05 17 0.00E+00 18 0.00E+00 22 3.40E-04

99

23 2.22E-03 Middle

Layer 4 0.00E+00

5 0.00E+00 6 5.76E-05 11 7.34E-04 16 2.21E-04 17 4.04E-04 18 3.17E-04 22 1.33E-04 23 1.33E-03 Bottom

Layer 4 2.26E-04

5 2.13E-04 6 3.38E-05 11 1.96E-03 16 9.51E-05 17 1.38E-04 18 2.48E-05 22 3.35E-04 23 3.36E-04 MP46.9SB Top Layer 4 6.05E-04 5 4.97E-04 6 0.00E+00 11 1.09E-03 16 17 18 1.59E-03 22 1.72E-03 23 1.80E-03 Middle

Layer 4 1.36E-05

5 0.00E+00 6 0.00E+00 11 0.00E+00 16 0.00E+00 17 0.00E+00 18 0.00E+00 22 0.00E+00 23 0.00E+00 Bottom

Layer 4 0.00E+00

5 0.00E+00 6 0.00E+00 11 0.00E+00

100

16 0.00E+00 17 0.00E+00 18 0.00E+00 22 0.00E+00 23 0.00E+00

NM344 MP1.80 Top Layer 4 3.21E-04 5 1.09E-03 11 7.12E-04 12 2.28E-03 16 1.20E-03 17 4.00E-03 18 2.97E-03 23 1.30E-03 24 1.81E-03 MP1.82 Top Layer 4 1.27E-04 6 3.08E-04 11 1.48E-03 12 9.38E-04 17 1.64E-03 18 5.90E-04 22 1.21E-03 23 4.75E-04 24 2.97E-04 MP1.84 Top Layer 4 1.21E-03 6 1.39E-05 10 3.70E-05 11 0.00E+00 16 2.71E-05 17 1.04E-03 18 1.62E-04 23 7.03E-06 24 2.40E-05 Bottom

layer 4 0.00E+00

6 0.00E+00 10 1.51E-04 11 3.80E-04 16 0.00E+00 17 4.91E-05 18 3.57E-06 23 4.44E-06 24 0.00E+00

101

APPENDIX D

MOISTURE DAMAGE TEST RESULTS

The data collected from AASHTO T283 and MIST test are presented in this appendix.

102

TABLE D.1 Moisture damage testing data.

DRY IDT

MIST IDT

T283 MOIST IDT

MIST TSR

T283 TSR

US285 MP285.25 TL 94.01 78.01 78.63 0.83 0.84 US285 MP285.25 ML 91.14 57.91 69.38 0.64 0.76 US285 MP285.25 BL 78.13 44.99 67.24 0.58 0.86 NM14 MP46.8NB TL 134.06 120.69 93.20 0.90 0.70 NM14 MP46.8NB ML 139.34 119.61 76.37 0.86 0.55 NM14 MP46.8NB BL 136.99 111.29 71.61 0.81 0.52 NM14 MP46.8SB TL 157.42 140.46 121.10 0.89 0.77 NM14 MP46.8SB ML 131.37 124.35 122.71 0.95 0.93 NM14 MP46.8SB BL 127.54 98.15 96.87 0.77 0.76 NM14 MP46.9SB TL 110.60 87.92 69.27 0.79 0.63 NM14 MP46.9SB ML 138.76 138.60 143.16 1.00 1.03 NM14 MP46.9SB BL 132.39 113.51 122.67 0.86 0.93 NM344 MP1.8 TL 213.35 156.48 108.10 0.73 0.51 NM344 MP1.82 TL 183.25 149.52 127.95 0.82 0.70 NM344 MP1.84TL 248.10 189.07 154.84 0.76 0.62 NM344 MP1.84BL 136.76 122.14 86.23 0.89 0.63 US285MP126.23BL 67.04 50.01 38.34 0.75 0.57 US70MP272.67TL 118.92 106.55 99.10 0.90 0.83 US70MP272.67BL 126.11 135.15 127.14 1.07 1.01 US491MP60.9TL 187.59 166.01 179.59 0.88 0.96 US491MP60.9BL 210.98 164.89 170.62 0.78 0.81 US491MP60.5ML 199.91 152.48 169.67 0.76 0.85 US491MP60.5BL 170.12 132.93 167.40 0.78 0.98 US491MP60.7TL 197.11 167.18 189.18 0.85 0.96 US491MP60.7ML 179.23 135.19 152.12 0.75 0.85 US491MP60.7BL 213.53 139.50 169.30 0.65 0.79 I40MP335.5DL 183.03 104.77 84.55 0.57 0.46 I40MP335.5S 143.06 91.49 66.36 0.64 0.46 US264MP10ML 174.18 124.44 110.76 0.71 0.64 US285MP140.53TL 207.45 158.64 164.18 0.70 0.58 US70MP289.26TL 190.36 168.64 146.31 0.76 0.66 US70MP289.26ML 189.68 135.12 132.66 0.70 0.55 US70MP289.26BL 91.25 94.12 92.84 0.88 0.61 US285MP152TL 101.73 82.00 63.20 0.72 0.56 US285MP152ML 85.45 62.69 52.30 0.82 0.49 US285MP152BL 76.73 58.96 40.38 0.76 0.79 US70MP282.2TL 118.83 110.41 76.61 0.89 0.77 US70MP282.2ML 131.98 114.59 72.49 0.71 0.70 US70MP282.2BL 107.76 88.20 86.31 1.03 1.02

103

APPENDIX E

TRACER TEST DATA

This appendix contains the results obtained from the tracer test.

104

Table E.1 Tracer test results.

6 in diameter sample 4in diameter sample Pavement sections Total

Pores Effective pores

Permeable pores

Permeability Permeable pores

Permeability

US285MP140.53 S1 7.38 6.95 6.03 141.6E-5 5.40 105.9E-5 US285MP140.53 S2 8.87 6.74 4.77 136.4E-5 4.15 124.9E-5 US285MP140.53 S3 8.61 8.34 6.99 261.1E-5 5.00 209.0E-5 US70MP289.26 TL S1 7.79 3.79 3.13 8.6E-5 2.38 7.6E-5 US70MP289.26 TL S2 7.42 6.98 5.52 25.9E-5 3.90 23.4E-5 US70MP289.26 TL S3 7.38 2.31 1.93 4.4E-5 1.83 3.9E-5 US70MP289.26 BL S1 10.21 9.49 3.43 11.4E-5 2.51 4.4E-5 US70MP289.26 BL S2 8.46 7.87 4.00 34.0E-5 2.77 21.7E-5 US70MP289.26 BL S3 8.39 7.60 3.83 34.9E-5 1.75 31.6E-5 US70MP282.2 TL S1 7.70 6.81 3.67 19.4E-5 3.11 12.7E-5 US70MP282.2 TL S2 7.35 7.00 5.58 12.4E-5 4.91 11.7E-5 US70MP282.2 TL S3 7.12 6.64 6.51 23.0E-5 4.82 22.8E-5 US70MP282.2 BL S1 8.08 7.28 4.23 31.3E-5 2.80 23.0E-5 US70MP282.2 BL S2 7.11 4.64 2.67 9.6E-5 0.98 8.7E-5 US70MP282.2 BL S3 7.97 7.18 4.55 68.5E-5 2.32 23.3E-5 US70MP272.67 TL S1 8.12 7.27 5.01 71.0E-5 4.14 49.3E-5 US70MP272.67 TL S2 8.90 3.15 2.94 9.4E-5 2.69 5.6E-5 US70MP272.67 TL S3 8.23 7.16 6.17 75.2E-5 5.65 36.1E-5 US70MP272.67 BL S1 13.70 13.42 9.24 898.2E-5 6.17 507.4E-5 US70MP272.67 BL S2 13.89 13.50 10.99 847.9E-5 10.05 535.4E-5 US70MP272.67 BL S3 5.16 4.78 1.85 21.3E-5 1.37 18.1E-5 US491MP60.5 S1 13.00 13.00 12.53 1650.8E-5 12.21 1238.1E-5 US491MP60.5 S2 9.21 9.21 7.22 201.2E-5 4.53 74.7E-5 US491MP60.5 S3 8.69 8.69 6.85 128.3E-5 4.41 123.9E-5 NM14MP46.9 S1 10.39 6.47 4.19 20.9E-5 3.13 23.8E-5 NM14MP46.9 S2 13.41 12.72 5.85 106.6E-5 5.42 47.0E-5 NM14MP46.9 S3 11.02 9.91 6.18 88.9E-5 5.57 90.3E-5 NM344MP1.8 S1 8.73 8.23 4.86 335.2E-5 4.71 172.6E-5 NM344MP1.8 S2 7.68 6.93 2.91 61.6E-5 2.09 47.4E-5 NM344MP1.8 S3 7.33 6.84 5.20 233.8E-5 4.36 86.8E-5 NM344MP1.82 S1 8.94 8.94 8.20 646.6E-5 7.96 392.3E-5 NM344MP1.82 S2 7.34 6.82 4.52 123.6E-5 4.07 49.0E-5 NM344MP1.82 S3 7.87 7.41 5.83 280.1E-5 4.04 182.2E-5 NM344MP1.84 S1 5.80 5.80 5.04 53.5E-5 3.55 50.5E-5 NM344MP1.84 S2 6.18 6.18 5.95 183.4E-5 5.50 121.2E-5 NM344MP1.84 S3 6.01 5.89 4.93 78.0E-5 4.57 51.0E-5

105

New Mexico Department of Transportation RESEARCH BUREAU 7500B Pan American Freeway NE PO Box 94690 Albuquerque, NM 87199-4690 Tel: (505) 841-9145