Investigating the Relationships between the ElectricalResistivity Characteristics and the Volumetric

Properties of Asphalt MixturesSeyed Arash Forough1; Fereidoon Moghadas Nejad2; and Hasan Ziari3

Abstract: A novel approach is introduced for investigating the relationships between the electrical resistivity characteristics, Er, and thevolumetric properties of asphalt mixtures, i.e., bulk specific gravity,Gmb, air voids in mixture, Vbm, and voids in mineral aggregates, Vma. It iswell known the volumetric properties of asphalt mixtures are controlled by several variables, i.e., aggregate gradation, aggregate type, fillertype, bitumen type, and bitumen content. In this study, the effects of both variables of aggregate gradation and bitumen content are inves-tigated on the electrical resistivity of asphalt mixtures and some correlations are developed between the electrical resistivity and the volu-metric properties of laboratory compacted asphalt mixture specimens. For this purpose, 455 Marshall specimens containing calcareous typecrushed stone aggregate with 60=70 penetration asphalt binder were fabricated and tested at five different aggregate gradations and sevenvaried bitumen contents with 13 replicate specimens for each experimental combination. Statistical ANOVA and significance tests wereconducted on the test data and the effects of electrical resistivity on the volumetric properties were evaluated. Finally, some simplecorrelations were formed between the two data sets by using regression analysis. DOI: 10.1061/(ASCE)MT.1943-5533.0000735.© 2013 American Society of Civil Engineers.

CE Database subject headings: Asphalts; Mixtures; Material properties; Electrical resistivity.

Author keywords: Asphalt mixture; Volumetric properties; Electrical resistivity; Simple correlations.

Introduction

Outline

The most important volumetric properties of asphalt mixtures in-clude bulk specific gravity, Gmb, air voids in mixture, Vbm, andvoids in mineral aggregates, Vma. Many different methods havebeen developed to determine the Gmb of compacted hot mix asphalt(HMA) specimens (Buchanan 2000; Hall et al. 2001; Malpass andKholsla 2001). Some of these methods include dimensional analy-sis (DA), as defined by AASHTO T269 (2007b), the water dis-placement or saturated surface dry (SSD) method, as defined byASTM D2726 (2011) or AASHTO T166 (2012), and the paraffincoating (PC) method, as described by AASHTO T275 (2007a). Todetermine the Vbm of an asphalt mixture specimen, it is required totest the specimen by using the instruments related to maximumtheoretical gravity, Gmm (Rice method). In addition, to determinethe Vma in an asphalt mixture specimen, the tests related to Gmmand Vbm must be performed first, then the Vma can be calculated.

The primary scope of this study was to investigate the relation-ships between the electrical resistivity characteristics and the volu-metric properties of asphalt mixtures. For this purpose, a noveltesting method was proposed to determine the electrical resistivityof laboratory-fabricated or field-drilled asphalt mixture specimens(Forough 2002).

Electrical resistivity is one of the physical properties of amaterial that can be used to obtain other properties of the material.Generally, there are two ways to conduct electrical current througha material:1. The material is naturally an electrical conductor. In this case,

electrical current is conducted by numerous free electrons inthe material. Metals like iron or copper are examples of suchmaterials.

2. The material is naturally an electrical insulator (dielectric), butit can be changed to an electrical conductor by a chemicalcombination or an artificial method. In this case, electrical cur-rent is conducted electrolytically via the ionic conduction.Composite materials like cement paste, cement concrete, orsoil are examples of such materials.

Asphalt mixture is a heterogeneous and composite material,which comprises the following constituents: (1) coarse and fineaggregates, (2) filler, (3) bitumen, (4) air, and (5) other additivessuch as antistripping agents (Walubita 2006). Asphalt mixture, asa blend of aggregates and bitumen, is an electrical insulator(dielectric) material with a high electrical resistance of approx-imately 1013 Ω. Thus, no free electrons exist in the asphaltmixture to directly conduct the electrical current. However, ifan asphalt mixture specimen becomes an electrical conductorby an artificial method, it is possible to obtain its electrical re-sistivity within a reasonable range and to relate this electricalresistivity to the volumetric properties of the specimen (Forough2002).

1Ph.D. student, Highway Division, Dept. of Civil and EnvironmentalEngineering, Amirkabir Univ. of Technology, Tehran 15914, Iran (corre-sponding author). E-mail: a.forough@aut.ac.ir

2Associate Professor, Dept. of Civil and Environmental Engineering,Amirkabir Univ. of Technology, Tehran 15914, Iran. E-mail: moghadas@aut.ac.ir

3Associate Professor, Dept. of Civil Engineering, Iran Univ. of Scienceand Technology, Tehran 16846, Iran. E-mail: h.ziari@iust.ac.ir

Note. This manuscript was submitted on April 2, 2012; approved onNovember 13, 2012; published online on November 15, 2012. Discussionperiod open until April 1, 2014; separate discussions must be submitted forindividual papers. This paper is part of the Journal of Materials in CivilEngineering, Vol. 25, No. 11, November 1, 2013. © ASCE, ISSN 0899-1561/2013/11-1692-1702/$25.00.

1692 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / NOVEMBER 2013

J. Mater. Civ. Eng. 2013.25:1692-1702.

Dow

nloa

ded

from

asc

elib

rary

.org

by

MA

RR

IOT

T L

IB-U

NIV

OF

UT

on

11/2

6/14

. Cop

yrig

ht A

SCE

. For

per

sona

l use

onl

y; a

ll ri

ghts

res

erve

d.

http://dx.doi.org/10.1061/(ASCE)MT.1943-5533.0000735http://dx.doi.org/10.1061/(ASCE)MT.1943-5533.0000735http://dx.doi.org/10.1061/(ASCE)MT.1943-5533.0000735http://dx.doi.org/10.1061/(ASCE)MT.1943-5533.0000735

Literature Review

Comparatively little work has been undertaken on the electricalproperties of asphalt mixtures, especially on the relation of electri-cal resistivity to its volumetric properties. Almost all of the earlierwork in this field has been undertaken on the electrical conductivityof asphalt mixtures containing electrically conductive materialssuch as basalt aggregate, limestone powder, carbon black, graphiteparticles, and steel wool for generating heat by the passage of anelectrical current through the asphalt mixture to prevent the accu-mulation of frozen precipitation thereon [Zaleski et al., “Electri-cally conductive paving mixture,” U.S. Patent No. 20040062606(2004)]. It is theoretically feasible to improve the electrical conduc-tivity of asphalt mixtures by the addition of electrically conductiveadditives to thermoelectrically melt and remove snow and ice onasphalt pavement (Shaopeng et al. 2005; García et al. 2009). Fur-thermore, adding electrically conductive materials to an asphaltmixture has been previously used to monitor the fatigue damagemechanism (Shaopeng et al. 2006), softening temperature, storagemodulus, damping capacity, and electrical resistivity of modifiedasphalt mixtures (Sihai and Chung 2004). According to previousstudies (Wu et al. 2002a, b, 2003), the electrical conductivity ofthe asphalt mixture is proportional to the concentration of electri-cally conductive additives.

Unlike asphalt mixtures, various studies have been undertakenon the electrical properties of cement mortars, cement pastes, andparticularly cement concretes, especially on the variation of elec-trical resistivity during the stages of initial setting and subsequenthardening (Li et al. 2007; Sengul and Gjrov 2008; Xiao and Li2008; Koleva et al. 2008). The conduction of electricity by moistconcrete is expected to be essentially electrolytic. The electrolyticconduction of moist concrete is linked to the circulation of fluidthrough the pore network (Lataste et al. 2003). Thus, the electricalresistivity of concrete is sensitive to the volume of porosity and tothe degree of porous interconnectivity (Andrade et al. 2000). Moistconcrete behaves essentially as an electrolyte with a resistivity ofapproximately 104 Ω-cm, a value in the range of semiconductors.Oven-dried concrete has a resistivity in the order of 1011 Ω-cm, areasonably good insulator (Malhotra and Carino 2004).

During this research, the conduction of electricity through theasphalt mixture specimens was performed similar to the electrolyticconduction of cement concrete. This aim was possible by saturatingthe interconnected air voids of the specimens with an electrolytesolution. Asphalt mixture saturated with an electrolyte solutionis electrically conductive because its interconnected pore networkis partially or fully filled with the solution containing mobile ions.Thus, the electrical conductivity of such an asphalt mixture isrelated to the volume fraction of pores, the conductivity of the

pore solution, and the interconnectivity of the porosity. In fact,the electrical resistivity is a measure of this interconnectivity.

Experimental Program

Material Properties

In a laboratory test program, continuous aggregate gradation wasused to comply with the gradation specifications of the Iran High-way Asphalt Paving Code (Ministry of Road and TransportationResearch and Education Center 2011) for wearing course. The ag-gregate gradations used in this study and the gradation limits aregiven in Table 1 and plotted in Fig. 1. The aggregates were calca-reous type crushed stone obtained from a local quarry. The fillerwas portland cement (Type II) with the apparent specific gravityof 3.13 g=cm3. The 60=70 penetration bitumen, which correspondsto performance grade (PG) 64-16, obtained from a local oil refinery,was used as binder to prepare the asphalt mixture specimens. Thephysical properties of coarse and fine aggregates and bitumen sam-ples are given in Tables 2 and 3.

Specimen Preparation

Two variables, bitumen content and aggregate gradation, were se-lected for study in the experimental design. Seven bitumen contentsof 3.5 to 6.5% in 0.5% increments and five different aggregate gra-dations formed the coarsest, Gradation 1, to the finest, Gradation 5,as mentioned in Table 1, which were used to make 455 asphalt mix-ture specimens. Thirteen replicate specimens [one replicate as thefirst group for the additional tests on Gmb, five replicates as thesecond group for the volumetric (Gmb, Vbm, and Vma) tests, tworeplicates as the third group for the saturating tests, and five rep-licates as the fourth group for the electrical resistivity tests] werefabricated for each experimental combination in a randomized se-quence. A summary of the experimental design is shown in Table 4.

Laboratory asphalt mixture specimens were prepared accordingto the Marshall mix design procedure in a laboratory environment,utilizing 75 blows on each face representing heavy traffic condi-tions, as specified in ASTM D1559-76 (2000). All specimens werecompacted by using the same load by the standard Marshall ham-mer. In addition, because the similarity of the specimen dimensionswas important only for the electrical resistance measurements andthe electrical resistivity calculations, the specimens of the fourthgroup were prepared with the same dimensions, diameter, andheight. For this purpose, an extra control was performed on thesespecimens after the fabrication process. Because the same moldwas used in the fabrication process, all the specimens had the samediameters, 101.6 mm, but with different surface irregularities.

Table 1. Aggregate Gradations Used in this Study and Gradation Specifications

Sieve Total percent passing (%)

size (mm) Gradation specifications Gradation 1 Gradation 2 Gradation 3 Gradation 4 Gradation 5

12.5 100 100 100 100 100 1009.5 80–100 80 85 90 95 1004.75 55–75 55 60 65 70 752.36 35–50 35 39 43 46 500.6 18–29 18 21 23 26 290.3 13–23 13 16 18 21 230.15 8–16 8 10 12 14 160.075 4–10 4 6 7 9 10Pan — — — — — —

Note: Gradation 1 is the coarsest and Gradation 5 is the finest.

JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / NOVEMBER 2013 / 1693

J. Mater. Civ. Eng. 2013.25:1692-1702.

Dow

nloa

ded

from

asc

elib

rary

.org

by

MA

RR

IOT

T L

IB-U

NIV

OF

UT

on

11/2

6/14

. Cop

yrig

ht A

SCE

. For

per

sona

l use

onl

y; a

ll ri

ghts

res

erve

d.

However, the different aggregate gradations and bitumen contentsaffected the degree of compaction; it was found that the differentspecimens, even replicates, had different heights with the minimumvalue of 63 mm. Thus, it was decided to prepare all specimens ofthe fourth group to the same height, 63 mm, before the electricalresistance measurements. For this purpose, the specimens with aheight of more than 63 mm were cut (milled) to the same heightas the other specimens. Finally, all of the electrical resistance testswere performed on the specimens with the same dimensions.

First Group: Additional Gmb Tests

The SSD and PC methods are the most common procedures em-ployed to determine the Gmb of compacted HMA. However, bothmethods have some major drawbacks. One drawback of the SSDmethod is the tendency for different operators to obtain results thatare dissimilar when performing testing on the same materials, usingthe same equipment, and following the same procedures. Thisproblem seems to be related to the interconnected air voids thatare present in the specimen. Water may infiltrate differently intothe submerged specimen when replicate testing is performed. Also,more or less water may drain from the specimen when trying toobtain the SSD weight. According to the test procedure, the SSDmethod is only valid for the water absorption of less than 2% andthe procedure is not recommended for specimens that contain openor interconnecting air voids. Also, the reliability of the SSD methoddecreases with increasing depth of the surface irregularities and thepresence of interconnected air voids that are open to the surface ofthe solid (Troxler Electronic Laboratories, Inc. 2003). In addition,the SSD condition is very difficult to determine because it is subjectto individual interpretation about when a specimen is SSD; thus,the procedure is prone to variability and error. These problems sug-gest that the SSD method, because of its potential variability, maynot be the most consistent method to determine the Gmb of com-pacted HMA. However, this method of testing is a rapid methodwith very low costs.

The PC method addresses the water absorption problems inher-ent in the SSD method. The HMA specimen is coated with paraffinto determine its exact volume; thus, the Gmb testing is performedwith the highest accuracy. However, the use of paraffin is expen-sive, time-consuming, and awkward to perform (Buchanan 2000).The PC method also may limit any further evaluation of the speci-men after the Gmb testing is completed.

To address the previously mentioned problems and to capturethe advantages of both of the preceding methods, a procedurewas proposed in this study in which all the specimens of the firstgroup (35 specimens) were tested for Gmb by using both the SSDand PC methods. After this, an adjusting factor was determined foreach experimental combination of aggregate gradation and bitumencontent to convert the inaccurate results of the SSD method to theaccurate results of the PC method.

60

80

100

pass

ing

Gradation 1

G d i 5

No.

100

No.

200

No.

50

No.

30

No.

8

No.

4

3/8

1/2

0

20

40

Perc

ent p

Sieve size (0.45 power)

Gradation 5

Gradation 2

Gradation 3

Gradation 4

Fig. 1. Aggregate gradations used in this study

Table 2. Physical Properties of Coarse and Fine Aggregates

Test value

Coarse Fine

Property[ASTM C127-04

(2004)][ASTM C128-04

(2004)]

Bulk specific gravity (g=cm3) 2.52 2.27Apparent specific gravity (g=cm3) 2.63 2.63Water absorption (%) 1.62 5.93

Table 3. Physical Properties of Bitumen Samples

Property Test value Standard

Penetration at 25°C (1=10 mm) 66 ASTM D5-97 (1997)Ductility at 25°C (cm) >100 ASTM D113-99 (1999)Loss on heating (%) 0.30 ASTM D680 (1965)Specific gravity at 25°C (g=cm3) 0.95 ASTM D7076 (2010)Softening point (°C) 52 ASTM D36-95 (2013)Flash point (°C) 303 ASTM D92-02 (2002)

Table 4. Experimental Design Used in this Study for Testing

Experimental variables Number of levels Variable levels

Bitumen content (%) 7 3.5, 4, 4.5, 5, 5.5, 6, 6.5Aggregate gradation 5 Coarsest to finestReplication 1 (first group) 1 Gmb additional testsReplication 2 (second group) 5 Volumetric testsReplication 3 (third group) 2 Saturating testsReplication 4 (fourth group) 5 Electrical resistivity tests

1694 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / NOVEMBER 2013

J. Mater. Civ. Eng. 2013.25:1692-1702.

Dow

nloa

ded

from

asc

elib

rary

.org

by

MA

RR

IOT

T L

IB-U

NIV

OF

UT

on

11/2

6/14

. Cop

yrig

ht A

SCE

. For

per

sona

l use

onl

y; a

ll ri

ghts

res

erve

d.

Second Group: Volumetric Tests

The tests required to determine the volumetric properties (Gmb,Vbm, and Vma) of the asphalt mixtures were conducted on thespecimens of the second group. For this purpose, all specimensof the second group (175 specimens) were tested for Gmb by usingthe SSD method and the results were converted to the accurateGmb of the PC method by using the adjusting factors. Finally,all specimens of the second group were tested for Gmm by usingthe Rice method and the properties of both Vbm and Vma werecalculated.

Third Group: Saturating Tests

Asphalt mixture becomes an electrical conductor material by satu-rating its interconnected Vbm by using an electrolyte solution. Theelectrolyte solution used in this study was NaCl (sodium chloride inpure water). In addition, it was necessary to find the best (mini-mum) concentration of the solution. Asphalt mixture, as an insu-lator (dielectric) material, has a high value of electrical resistance inthe order of 1013 Ω. However, because the digital resistivity meterused in this study was able to capture and record the resistance val-ues of 2 × 102 to 2 × 106 Ω, the electrical resistance of the spec-imens after the saturating process was within the mentioned limit.On the other hand, asphalt mixtures with different aggregate gra-dations and bitumen contents have different values of electricalresistance. Therefore, the minimum concentration of the NaCl sol-ution should be selected in such a way that the highest electricalresistance that results in the specimens with the finest gradation,Gradation 5, and the highest bitumen content, 6.5%, are withinthe mentioned limit. It is clear in such a condition that the electricalresistance values of the other specimens will be certainly within thelimit. To reach this condition the minimum amount of sodium chlo-ride was added to a given volume of pure water. For this purpose,different amounts of sodium chloride in 1 L of pure water weretested; it was found that 100 g of sodium chloride in 1 L of purewater (corresponding to the concentration of 9.1%) resulted in theminimum concentration for the solution.

There were two ways to saturate the asphalt mixture specimenswith the electrolyte solution: the specimens could be immersed inthe solution under an open atmosphere or under a high pressure.In this study, both ways were tested on the specimens of the thirdgroup to find the fastest method to saturate the specimens ofthe fourth group. For this purpose, weight variation was used asthe criterion for saturating the specimens. This means that eachspecimen was determined to be saturated when there was no moreincrease in its weight. Assuming this criterion, 35 out of 70 spec-imens of the third group were immersed in the NaCl solution underopen atmosphere and their weight variations were monitored versustime. All of the specimens were saturated within a long time of10 days under this condition. To decrease the necessary time forsaturating the interconnected Vbm of the specimens, it was neces-sary to test the specimens under a higher pressure. For this purpose,an apparatus was constructed. The apparatus consisted of a pressurevessel equipped with a pressure gauge, an air solution container, anair pump, two high pressure connective hoses, and a pressure con-trol valve. Fig. 2 shows the configuration of the apparatus used inthis study for saturating the interconnected Vbm of the specimensunder high pressure.

The pressure vessel was able to sustain a pressure less than orequal to 15 psi. Therefore, the remaining 35 specimens of the thirdgroup were placed in the pressure vessel filled with the NaCl sol-ution under a pressure of 15 psi and their weight variations weremonitored versus time. All of the specimens were saturated withina shorter, but still lengthy, time of 24 h under this condition.

However, if the pressure vessel was able to sustain a pressure morethan 15 psi, the required testing time would have been significantlydecreased.

Fourth Group: Electrical Resistivity Tests

After the saturating process under the pressure of 15 psi, all spec-imens of the fourth group similarly reached the SSD conditionunder the same operator. Therefore, no electricity was allowedto conduct on the wet surface of the specimens. Electrical resistancemeasurements were performed by using another apparatus con-structed for this purpose. The apparatus consisted of a mast frame,a main shaft, a rotational handle, a ring and gauge set, a specimenholder, two electrodes, and a digital resistance meter. Fig. 3 showsthe configuration of the apparatus used in this study for the elec-trical resistance measurements.

The specimen holder can be used for either laboratory fabricatedor field-drilled specimens. The electrodes were made of brass witha diameter of 101.6 mm, equal to the diameter of the specimens.The main shaft was used to apply uniform axial compression oneach side of the specimens through the brass electrodes by clock-wise rotation of the handle. In addition, because the electricalresistance measurements should be conducted under the same con-ditions on all specimens of the fourth group, the ring and gauge setwas used to verify the equality of axial compression applied to thespecimens. The measurement test was set up at room temperatureof 25°C. To ensure perfect contact between the electrodes and thespecimens and to minimize the measurement error as a result of thejagged and rough texture of the asphalt mixture specimens, espe-cially those with coarse gradations and low bitumen contents, anelectrical conductor material with a very low level of electrical re-sistance was used to fill the gaps between the electrodes and thespecimens. This material provided a full contact area on the crosssection of the specimens with no effect on their electrical resistance.For this purpose, cement grout in a liquid-plastic form was used.

Fig. 2. Configuration of the apparatus used for saturating the speci-mens under pressure of 15 psi

JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / NOVEMBER 2013 / 1695

J. Mater. Civ. Eng. 2013.25:1692-1702.

Dow

nloa

ded

from

asc

elib

rary

.org

by

MA

RR

IOT

T L

IB-U

NIV

OF

UT

on

11/2

6/14

. Cop

yrig

ht A

SCE

. For

per

sona

l use

onl

y; a

ll ri

ghts

res

erve

d.

When water was added to cement, the resulting cement paste had aninitial setting and a subsequent hardening. The cement paste in theinitial setting phase was a kind of liquid-plastic material that has alow level of electrical resistance. In this study, the total contact re-sistance between the brass electrodes and the cement grout was lessthan 1 Ω, which was negligible with respect to the high resistancevalues of the specimens: more than 1,000 Ω. If the cement groutwas allowed to cure more over time, it would have hardened and itselectrical resistance would increase. Therefore, the cement groutwas used before any hardening.

After the electrical resistance measurements, the electrical resis-tivity, Er, of each specimen was obtained by using the electricalresistance and dimensions of the specimen: both the length in

the current direction and the cross-sectional area perpendicularto the current direction.

Tests Results and Discussion

Volumetric Test Results

The volumetric properties (Gmb, Vbm, and Vma) of the asphalt mix-ture specimens versus bitumen content, Pb, for the five gradationsare shown in Figs. 4–6. These figures show the average results ofthe five replicate specimens for each experimental combination.Therefore, each figure represents the results of 175 specimens.Table 5 shows the coefficient of variation (COV) values for allof the volumetric properties. If 30% is used as the threshold, theCOVs of the volumetric test results would be judged as acceptable,indicating that the repeatability and variability of the volumetrictests are reasonable (Walubita 2010). This table shows that thespecimens with the coarsest gradation, Gradation 1, and the lowestbitumen content, 3.5%, exhibited the highest variability comparedwith the other mixtures. This was not surprising because of thegreater heterogeneity and poor Vbm distribution of these mixtures,arising predominantly from the coarser gradation and the lowerbitumen content, and thus, the poor workability of the mixtures.According to Table 5, the COVof Vbm in the mixtures with a bitu-men content of 3.5% and Gradation 1 was 10.63%, versus an over-all average of 6.80% for all of the mixtures considered in the study.In general, the finely graded mixtures with higher bitumen contentswere observed to exhibit lower variability in terms of all volumetrictest results than the densely to coarsely graded mixtures with lowbitumen contents. The coarsely graded mixtures with low bitumencontents were generally associated with high variability. Low vari-ability of the finely graded mixtures with high bitumen contentswas attributed to their good workability characteristics, which al-low easy specimen fabrication, mixture homogeneity, and uniformVbm distribution. The coarsely graded mixtures with low bitumencontents, on the other hand, are comparatively difficult to workwith, and it is equally difficult to maintain a consistent and uniformVbm distribution in the specimens.

Comparing the material properties, the rank order of increasingvariability in terms of the COV was Gmb, Vma, and Vbm. The COVranges were from 0.18 to 1.13% for Gmb, 1.71 to 10.63% for Vbm,and 1.13 to 7.53% for Vma. However, it was apparent in this study

Fig. 3. Configuration of the apparatus used for the electrical resistancemeasurements

2.30

2.35

2.40

ty, G

mb

(gr/

cm³)

Gradation 1

Gradation 2

Gradation 3

Gradation 4

Gradation 5

R² = 0 960

2.15

2.20

2.25

3 3.5 4 4.5 5 5.5 6 6.5 7

Bul

k sp

ecif

ic g

ravi

t

Bitumen Content, Pb (%)

R² = 0.960

R² = 0.946

R² = 0.948

R² = 0.982

R² = 0.944

Fig. 4. Gmb versus Pb for the five gradations

1696 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / NOVEMBER 2013

J. Mater. Civ. Eng. 2013.25:1692-1702.

Dow

nloa

ded

from

asc

elib

rary

.org

by

MA

RR

IOT

T L

IB-U

NIV

OF

UT

on

11/2

6/14

. Cop

yrig

ht A

SCE

. For

per

sona

l use

onl

y; a

ll ri

ghts

res

erve

d.

that variability in all of the test results depended on the variability ofVbm, which is ultimately a function of mixture workability and thespecimen fabrication process. Thus, proper specimen preparationand maintenance of consistency in the distribution of Vbm contentsis one critical approach to minimizing variability in the volumetrictest results.

The test results in Figs. 4–6 indicate that all of the volumetricproperties were sensitive to both variables of aggregate gradationand bitumen content; they have very strong nonlinear quadratic re-lationships with these variables. These observations agree with pre-vious studies, which are well established in the existing literature.

Saturating Test Results

The weights of the penetrating solution into the interconnected Vbmof the specimens versus Pb for the five gradations under openatmosphere and a pressure of 15 psi are shown in Figs. 7 and 8,respectively.

The statistical analysis was conducted by using a pairedsamples t-test with a reliability level of 95%; the results are givenin Table 6. This table indicates a significant level of agreement, or

correlation, between the two data sets, suggesting that no statisti-cally significant difference exists between these two methods at95% reliability level.

Figs. 7 and 8 show that the weight of the penetrating solutiondecreases as bitumen content increases under a constant aggregategradation. This is because the weight of the penetrating solution isdependent on the ratio of interconnected Vbm to the total Vbm,which decreases as the bitumen content increases. In addition, thesefigures show that the weight of the penetrating solution decreases asgradation becomes finer under constant bitumen content. However,the slope of this decrease in the mixtures with Gradation 5 is notsimilar to those for the other gradations. In these mixtures, theweight of the penetrating solution decreases more for the low bitu-men contents and less for the high contents.

In addition, as aggregate gradation becomes finer fromGradation 4 toward Gradation 5 under a constant bitumen content,the Vbm increases suddenly, but the interconnected Vbm decreasescontinuously. On the other hand, the weight of the penetratingsolution decreases in this limit. Therefore, the weight of the pen-etrating solution is not dependent to the Vbm; in fact, it is a functionof the interconnected Vbm.

8.00

10.00

12.00

14.00

re, V

bm(%

)

Gradation 1

Gradation 2

Gradation 3

Gradation 4

Gradation 5

0.00

2.00

4.00

6.00

3 3.5 4 4.5 5 5.5 6 6.5 7

Voi

ds in

mix

tur

Bitumen Content, Pb (%)

R² = 0.989

R² = 0.982

R² = 0.986

R² = 0.997

R² = 0.997

Fig. 5. Vbm versus Pb for the five gradations

12 00

13.00

14.00

15.00

rega

tes,

Vm

a(%

)

Gradation 1

Gradation 2

Gradation 3

Gradation 4

Gradation 5

R² = 0 946

9.00

10.00

11.00

12.00

3 3.5 4 4.5 5 5.5 6 6.5 7

Voi

ds in

min

eral

agg

r

Bitumen Content, Pb (%)

R = 0.946

R² = 0.937

R² = 0.951

R² = 0.961

R² = 0.967

Fig. 6. Vma versus Pb for the five gradations

JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / NOVEMBER 2013 / 1697

J. Mater. Civ. Eng. 2013.25:1692-1702.

Dow

nloa

ded

from

asc

elib

rary

.org

by

MA

RR

IOT

T L

IB-U

NIV

OF

UT

on

11/2

6/14

. Cop

yrig

ht A

SCE

. For

per

sona

l use

onl

y; a

ll ri

ghts

res

erve

d.

Electrical Resistivity Test Results

The electrical resistivity test results versus Pb for the five grada-tions are shown in Fig. 9. This figure shows the average resultsof the five replicate specimens for each experimental combination.Therefore, it represents the results of 175 specimens. Because thevariation of the electrical resistivity values versus both variables ofaggregate gradation and bitumen content strongly follows an expo-nential pattern, the natural logarithm of electrical resistivity, LEr,was used instead of Er in Fig. 9. Table 7 shows the COVof the Erfor the five replicate specimens for each experimental combination.If 30% is used again as the threshold, all COV results in this tableare acceptable, indicating that the repeatability and variability of theelectrical resistivity tests are reasonable.

Fig. 9 shows that LEr increases linearly as bitumen content in-creases under a constant gradation. This is because the intercon-nected Vbm and weight of the penetrating solution decrease asbitumen content increases under a constant gradation. In addition,it is clearly evident from Fig. 9 that LEr decreases linearly as

gradation becomes finer from Gradation 1 toward Gradation 5under constant bitumen content. As aggregate gradation becomesfiner, the interconnected Vbm and weight of the penetrating solutiondecrease continuously; thus, LEr is expected to increase. However,Fig. 9 shows that the results do not agree with this expectation dueto the following reason. The filler used in this study was cement,which becomes cement paste, a good electrical conductor, by theaddition of water. Moreover, the weight percent of the cementin the mixtures increases as aggregate gradation becomes finer.When the NaCl solution penetrates the asphalt mixture and com-bines with the cement, the electrical conductivity of the mixtureincreases and its electrical resistivity decreases by the resulting ce-ment paste. Therefore, as aggregate gradation becomes finer andthe weight percent of the filler increases, the effect of high electricalconductivity of cement paste, in the initial setting phase beforeany hardening, is more significant on the electrical conductivity/resistivity of the asphalt mixture specimens.

To examine this hypothesis, limestone was used as a new fillerto fabricate five extra specimens with different gradations under aconstant bitumen content of 5.5%. These specimens were tested forelectrical resistance measurements after the saturating process.Unlike the previous trend for specimens with the cement filler,the specimens with the limestone filler showed more LEr for thefine gradations than those for the coarse gradations. In other words,based on the results of a limited laboratory experiment on the spec-imens with limestone filler, it was determined that the effect ofcement paste in the initial setting phase before any hardeningprevails on the electrical conductivity/resistivity of the asphaltmixture specimens. However, more specimens with different aggre-gate gradations and bitumen contents are needed to investigate thereal effects of limestone filler on the electrical resistivity character-istics of asphalt mixtures and to reliably support the precedinghypothesis.

Statistical Analyses

Correlation Analysis among the Volumetric Properties

Correlation analysis was performed by using the Pearson cor-relation method to determine the significance levels for therelationships among the volumetric properties; the results are givenin Table 8. This table shows that each volumetric property has asignificant correlation with the others at the 0.01 level. According

Table 5. COV of Volumetric Test Results

PropertyPb(%)

Gradation1

Gradation2

Gradation3

Gradation4

Gradation5

3.5 1.13 0.96 0.77 0.72 0.664.0 0.96 0.62 0.59 0.55 0.484.5 1.00 0.88 0.85 0.48 0.41

Gmb 5.0 0.97 0.57 0.77 0.68 0.665.5 0.75 0.65 0.39 0.67 0.256.0 0.86 0.45 0.38 0.27 0.366.5 0.90 0.41 0.31 0.23 0.183.5 10.63 9.52 8.22 6.34 4.684.0 10.31 7.33 7.32 6.29 4.004.5 8.51 8.53 7.93 7.36 3.97

Vbm 5.0 7.61 4.25 7.67 7.49 2.825.5 8.54 9.39 8.34 5.79 3.376.0 9.89 7.71 6.52 6.29 4.786.5 7.89 6.47 5.23 5.30 1.713.5 7.53 6.48 5.35 4.94 3.824.0 6.79 4.51 4.49 4.18 2.944.5 7.17 6.72 6.75 3.98 2.58

Vma 5.0 7.02 4.39 6.16 5.73 4.245.5 5.91 5.69 3.61 6.37 1.746.0 6.33 3.63 3.30 2.43 2.386.5 6.26 3.07 2.42 1.95 1.13

20.0

25.0

30.0

35.0

netr

atin

g so

luti

onm

osph

ere

(gr)

Gradation 1

Gradation 2

Gradation 3

Gradation 4

Gradation 5

R² 0 992

0.0

5.0

10.0

15.0

3 3.5 4 4.5 5 5.5 6 6.5 7

Wei

ght o

f th

e pe

nun

der

open

atm

Bitumen content, Pb (%)

R² = 0.992

R² = 0.990

R² = 0.988

R² = 0.996

R² = 0.992

Fig. 7. Weight of the penetrating solution versus Pb for the five gradations under open atmosphere

1698 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / NOVEMBER 2013

J. Mater. Civ. Eng. 2013.25:1692-1702.

Dow

nloa

ded

from

asc

elib

rary

.org

by

MA

RR

IOT

T L

IB-U

NIV

OF

UT

on

11/2

6/14

. Cop

yrig

ht A

SCE

. For

per

sona

l use

onl

y; a

ll ri

ghts

res

erve

d.

to the results of this table, Gmb has a negative correlation with bothproperties of Vbm and Vma and these two properties have a positivecorrelation with each other.

Relationships between LEr and the VolumetricProperties

The volumetric test results versus LEr for the five gradations areshown in Figs. 10–12. These figures show that under a constantgradation, Gmb increases and Vma decreases as LEr increases up

to a peak point; next, as LEr increases again, Gmb decreases andVma increases. However, under a constant gradation, Vbm de-creases continuously as LEr increases. These results show thatthere is a significant direct linear relationship between Pb andLEr. On the other hand, as gradation becomes finer fromGradation 1 toward Gradation 4 under constant LEr, Gmb in-creases and the properties of both Vma and Vbm decrease. Afterthis, as gradation becomes finer again from Gradation 4 towardGradation 5, Gmb decreases and the properties of both Vma andVbm suddenly increase.

2 00

3.00

4.00

5.00

6.00

7.00

ectr

ical

resi

stiv

ity,

LE

r

.m)

Gradation 1

Gradation 2

Gradation 3

Gradation 4

Gradation 5

R² = 0 924

-3.00

-2.00

-1.00

0.00

1.00

2.00

3 3.5 4 4.5 5 5.5 6 6.5 7

Nat

ural

loga

rith

m o

f el

e(k

Bitumen Content, Pb (%)

R² = 0.924

R² = 0.929

R² = 0.921

R² = 0.928

R² = 0.956

Fig. 9. Er versus Pb for the five gradations

20.0

25.0

30.0

35.0

netr

atin

g so

luti

one

of 1

5Psi

(gr

)

Gradation 1

Gradation 2

Gradation 3

Gradation 4

Gradation 5

R² 0 990

0.0

5.0

10.0

15.0

3 3.5 4 4.5 5 5.5 6 6.5 7

Wei

ght o

f th

e pe

nun

der

pres

sure

Bitumen content, Pb (%)

R² = 0.990

R² = 0.992

R² = 0.989

R² = 0.997

R² = 0.993

Fig. 8. Weight of the penetrating solution versus Pb for the five gradations under pressure of 15 psi

Table 6. Paired Sample t-Test for Weights of Penetrating Solution underDifferent Pressures

Paired differences at 95%confidence level

Mean SDStandarderror mean t df

Sig.(2-tailed)

−0.5171 0.3092 5.226 × 10−2 −9.896 34 0.000Note: SD = standard deviation; df = degree of freedom; Sig. =significance level.

Table 7. COV of Er Test Results

Pb (%) Gradation 1 Gradation 2 Gradation 3 Gradation 4 Gradation 5

3.5 15.61 16.64 7.24 7.69 7.234.0 12.37 17.22 15.46 8.21 9.584.5 10.89 15.86 16.08 5.12 11.625.0 13.65 12.96 19.20 6.17 10.435.5 8.64 10.81 5.45 9.41 6.246.0 9.78 9.84 4.89 9.35 5.536.5 12.29 5.97 6.68 6.22 5.85

JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / NOVEMBER 2013 / 1699

J. Mater. Civ. Eng. 2013.25:1692-1702.

Dow

nloa

ded

from

asc

elib

rary

.org

by

MA

RR

IOT

T L

IB-U

NIV

OF

UT

on

11/2

6/14

. Cop

yrig

ht A

SCE

. For

per

sona

l use

onl

y; a

ll ri

ghts

res

erve

d.

The test results in Figs. 10–12 indicate that all volumetric prop-erties are sensitive to LEr and have very strong nonlinear quadraticrelationships with this physical property of the asphalt mixtures.

Because of these very strong nonlinear quadratic relationshipsbetween LEr and the volumetric properties, the Pearson linear cor-relation method cannot be used to determine the nonlinear corre-lation coefficients between them. Therefore, the relationshipsbetween LEr and all of the volumetric properties of the asphaltmixture specimens are evaluated by using statistical ANOVAand significance tests. For this purpose, one-way ANOVAwas con-ducted at 95% confidence level on Gmb, Vbm, and Vma by using theSPSS statistical package. The Fisher numbers and probability ofsignificance for LEr at the given confidence level were calculatedfor each volumetric property and presented in Table 9. Based on the

selected confidence level, 95%, when the calculated probability forLEr is smaller than 5%, it shows a significant difference from themean, which may result from some other factors rather than purechance. In addition, a larger Fisher number corresponding to aprobability less than 5% increases the significance level of LErfor the volumetric properties. Table 9 indicates that all of the rela-tionships between LEr and the volumetric properties of the asphaltmixture specimens are significant at 95% confidence level. The sig-nificance levels can be compared for the volumetric properties byusing the calculated Fisher numbers. Accordingly, Vbm is the mostaffected property by LEr, followed by Gmb and the Vma, respec-tively. In other words, Vbm is strongly dependent on LEr havingthe largest Fisher number, 30.264, among the other volumetricproperties. Similarly, Gmb also proves a strong dependence on LErand LEr shows also a significant influence on Vma at 95% confi-dence level.

Correlations between LEr and the VolumetricProperties

Regression analyses were performed on the test data to developsome simple correlations between LEr and all of the volumetric

2.30

2.35

2.40

ravi

ty, G

mb

(gr/

cm³)

Gradation 1

Gradation 2

Gradation 3

Gradation 4

Gradation 5

R² = 0.973

R² 0 961

2.15

2.20

2.25

-4 -3 -2 -1 0 1 2 3 4 5 6

Bul

k sp

ecif

ic g

r

Natural logarithm of electrical resistivity, LEr (kΩ.m)

R² = 0.961

R² = 0.959

R² = 0.947

R² = 0.982

Fig. 10. Gmb versus LEr for the five gradations

Table 8. Correlation Analysis among Volumetric Properties

Property Gmb Vbm Vma

Gmb 1.000 −0.927a −0.792aVbm −0.927a 1.000 0.527aVma −0.792a 0.527a 1.000aIndicates that correlation is significant at the 0.01 level (two-tailed).

8.00

10.00

12.00

14.00

re, V

bm(%

)

Gradation 1

Gradation 2

Gradation 3

Gradation 4

Gradation 5

0.00

2.00

4.00

6.00

-3 -2 -1 0 1 2 3 4 5 6

Voi

ds in

mix

tur

Natrural logarithm of electrical resistivity, LEr (kΩ.m)

R² = 0.970

R² = 0.962

R² = 0.967

R² = 0.948

R² = 0.981

Fig. 11. Vbm versus LEr for the five gradations

1700 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / NOVEMBER 2013

J. Mater. Civ. Eng. 2013.25:1692-1702.

Dow

nloa

ded

from

asc

elib

rary

.org

by

MA

RR

IOT

T L

IB-U

NIV

OF

UT

on

11/2

6/14

. Cop

yrig

ht A

SCE

. For

per

sona

l use

onl

y; a

ll ri

ghts

res

erve

d.

properties of the asphalt mixture specimens. Because variations ofall volumetric properties versus LEr strongly follow nonlinearquadratic patterns and because LEr has significant effects on allvolumetric properties, it was decided to develop simple correlationsbetween LEr and the volumetric properties for each gradationbased on the general quadratic formulation ofaðLErÞ2þbðLErÞþc.Table 10 shows the regression coefficients (a, b, and c) and thecoefficients of determination for each simple correlation betweenLEr and all of the volumetric properties under different gradations.

Conclusion

This research was conducted to investigate the relationshipsbetween the electrical resistivity and the volumetric properties ofasphalt mixtures, i.e., Gmb, Vbm, and Vma. For this purpose, all

volumetric properties of the mixtures were evaluated at variousaggregate gradations and bitumen contents and electrical resistivitymeasurements were performed on the specimens by using a noveltesting method. Correlation analysis, statistical ANOVA, and sig-nificance tests were performed on the test data; finally, some simplecorrelations were formed between LEr and the volumetric proper-ties of the asphalt mixture specimens for each aggregate gradationby the method of regression analysis by using the SPSS statisticalsoftware.

Although the simple correlations developed for the tested spec-imens were characterized by high coefficients of determination, thegeneral application of these correlations to asphalt mixtures withdifferent aggregate types, aggregate gradations, aggregate absorp-tions, filler types, and bitumen types from those tested would notlikely yield accurate results. More research is needed to account forthe effects of these important variables on the correlations betweenLEr and the volumetric properties of asphalt mixtures.

Although the saturating process in the proposed approach re-quires a significant time, 24 h, more suitable instruments than thoseused in this study for saturating the specimens can decrease therequired testing time. In other words, if the pressure vessel is ableto sustain a pressure of more than 15 psi, the required testing timefor saturating the interconnected air voids of the specimens will besignificantly decreased. Although both the saturating and resistivitytesting require special equipment, implementation of the proposedapproach will be met with no obstacles. In fact, one of the majoradvantages of this approach is the low cost of the special equipmentrequired for electrical resistivity measurements and saturation ofthe specimens.

References

AASHTO. (2007a). “Standard method of test for bulk specific gravity ofcompacted bituminous mixtures using paraffin-coated specimens.”T275, Washington, DC.

AASHTO. (2007b). “Standard method of test for percent air voids in com-pacted dense and open asphalt mixtures.” T269, Washington, DC.

AASHTO. (2012). “Standard method of test for bulk specific gravity ofcompacted hot mix asphalt using saturated surface-dry specimens.”T166, Washington, DC.

Andrade, C., Alonso, C., Arteaga, A., and Tanner, P. (2000). “Methodologybased on the electrical resistivity for calculation of reinforcement

12

13.00

14.00

15.00

greg

ates

, Vm

a(%

)

Gradation 1

Gradation 2

Gradation 3

Gradation 4

Gradation 5

R² = 0 712

9.00

10.00

11.00

-3 -2 -1 0 1 2 3 4 5 6

Voi

ds in

min

eral

agg

Natural logarithm of electrical resistivity, LEr (kΩ.m)

R = 0.712

R² = 0.731

R² = 0.680

R² = 0.774

R² = 0.924

12.00

Fig. 12. Vma versus LEr for the five gradations

Table 10. Regression Analysis for Each Simple Correlation

Property Gradation a b c r2

1 −0.0004 0.015 2.229 0.9732 −0.0017 0.021 2.263 0.961

Gmb 3 −0.0021 0.023 2.280 0.9594 −0.0034 0.026 2.307 0.9475 −0.0022 0.017 2.240 0.9821 0.020 −1.258 9.718 0.9702 0.034 −1.236 8.060 0.962

Vbm 3 0.064 −1.381 7.136 0.9674 0.099 −1.452 6.038 0.9485 0.080 −1.178 7.826 0.9811 0.068 −0.454 12.761 0.7122 0.075 −0.434 11.608 0.731

Vma 3 0.099 −0.478 10.796 0.6804 0.124 −0.476 10.177 0.7745 0.078 −0.302 13.166 0.924

Table 9. ANOVA for Volumetric Properties at 95% Confidence Level

Property

LEr

F Probability > F

Gmb 30.264 0.000Vbm 76.225 0.000Vma 3.464 0.043

JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / NOVEMBER 2013 / 1701

J. Mater. Civ. Eng. 2013.25:1692-1702.

Dow

nloa

ded

from

asc

elib

rary

.org

by

MA

RR

IOT

T L

IB-U

NIV

OF

UT

on

11/2

6/14

. Cop

yrig

ht A

SCE

. For

per

sona

l use

onl

y; a

ll ri

ghts

res

erve

d.

service life.” 5th CANMET/ACI Int. Conf., American Concrete Institute(ACI), Farmington Hills, MI, 899–915.

ASTM. (1965). “Methods of testing and tolerances for certain fine staplecotton gray goods.” D680, West Conshohocken, PA.

ASTM. (1997). “Standard test method for penetration of bituminousmaterials.” D5-97, West Conshohocken, PA.

ASTM. (1999). “Standard test method for ductility of bituminous materi-als.” D113-99, West Conshohocken, PA.

ASTM. (2000). “Standard test method for resistance to plastic flow ofbituminous mixtures using Marshall apparatus.” D1559-76, WestConshohocken, PA.

ASTM. (2002). “Standard test method for flash and fire points byCleveland open cup tester.” D92-02, West Conshohocken, PA.

ASTM. (2004a). “Standard test method for density, relative density(specific gravity), and absorption of coarse aggregate.” C127-04, WestConshohocken, PA.

ASTM. (2004b). “Standard test method for density, relative density(specific gravity), and absorption of fine aggregate.” C128-04, WestConshohocken, PA.

ASTM. (2010). “Standard test method for measurement of shives in rettedflax.” D7076, West Conshohocken, PA.

ASTM. (2011). “Standard test method for bulk specific gravity and densityof non-absorptive compacted bituminous mixtures.” D2726, WestConshohocken, PA.

ASTM. (2013). “Standard test method for volatile alcohols in waterby direct aqueous-injection gas chromatography.” D3695, WestConshohocken, PA.

Buchanan, M. S. (2000). “An evaluation of selected methods for measuringthe bulk specific gravity of compacted hot mix asphalt (HMA) mixes.”J. Assoc. Asphalt Paving Technol., 69, 608–634.

Forough, S. A. (2002). “Predicting the results of Marshall test using theelectrical resistivity of asphalt mixtures.”M.Sc. thesis, Amirkabir Univ.of Technology, Polytechnic of Tehran, Iran.

García, Á., Schlangen, E., van de Ven, M., and Liu, Q. (2009). “Electricalconductivity of asphalt mortar containing conductive fibers and fillers.”Constr. Build. Mater., 23(10), 3175–3181.

Hall, K. D., Fances, G. T., and Williams, S. G. (2001). “Examinationof operator variability for selected methods for measuring bulkspecific gravity of hot-mix asphalt concrete.” Transportation ResearchRecord 1761, Transportation Research Board, Washington, DC,81–85.

Koleva, D. A., Copuroglu, O., van Breugel, K., Ye, G., and de Wit, J. H. W.(2008). “Electrical resistivity and microstructural properties of concretematerials in conditions of current flow.” Cem. Concr. Compos., 30(8),731–744.

Lataste, J. F., Sirieix, C., Breysse, D., and Frappa, M. (2003). “Electricalresistivity measurement applied to cracking assessment on reinforced

concrete structures in civil engineering.” NDT and E Int., 36(6),383–394.

Li, Z., Xiao, L., and Xiaosheng, W. (2007). “Determination of concretesetting time using electrical resistivity measurement.” J. Mater. Civ.Eng., 19(5), 423–427.

Malhotra, N., and Carino, J. (2004). “Magnetic/electrical methods.”Chapter 10, Handbook on nondestructive testing of concrete, Taylor& Francis, London.

Malpass, G. A., and Kholsla, N. P. (2001). “Evaluation of gamma ray tech-nology for the direct measurement of bulk specific gravity of compactedasphalt concrete mixtures.” J. Assoc. Asphalt Paving Technol., 70,352–367.

Ministry of Road and Transportation Research and Education Center.(2011). “Iran highway asphalt paving code.” Pub. No. 234, Iran.

Sengul, O., and Gjrov, O. E. (2008). “Electrical resistivity measurementsfor quality control during concrete construction.” ACI Mater. J., 105(6),541–547.

Shaopeng, W., Liantong, M., Zhonghe, S., and Zheng, Chen. (2005).“Investigation of the conductivity of asphalt concrete containing con-ductive fillers.” Carbon, 43(7), 1358–1363.

Shaopeng, W., Xiaoming, L., Liantong, M., and Qunshan, Y. (2006).“Research of self-monitoring mechanism of electrically conductiveasphalt-based composite.” Key Eng. Mater., 236–328(2), 1499–1502.

Sihai, W., and Chung, D. D. L. (2004). “Effects of carbon black on thethermal, mechanical and electrical properties of pitch-matrix compo-sites.” Carbon, 42(12–13), 2393–2397.

SPSS [Computer software]. Armonk, New York, IBM Corporation.Troxler Electronic Laboratories, Inc. (2003). Troxler Electronic Laborato-

ries, Research Triangle Park, NC, 〈http://www.troxlerlabs.com/3660apps.html〉 (Dec. 10, 2003).

Walubita, L. F. (2006). “Comparison of fatigue analysis approaches for pre-dicting fatigue lives of hot-mix asphalt concrete (HMAC) mixtures.”Ph.D. dissertation, Texas A&M Univ., College Station, Texas, 〈http://handle.tamu.edu/1969.1/3898〉 (Jun. 07, 2012).

Walubita, L. F. (2010). “Characterizing the ductility and fatigue crack re-sistance potential of asphalt mixes based on the laboratory direct tensilestrength test.” J. South African Inst. Civ. Eng. (SAICE), 52(2), 31–40.

Wu, S. P., Mo, L. T., and Shui, Z. H. (2002a). “Improvement of electricalproperties of asphalt concrete.” J. Wuhan. Univ. Technol., Mater. Sci.Ed., 17(4), 69–72.

Wu, S. P., Mo, L. T., and Shui, Z. H. (2002b). “Preparation of electricallyconductive asphalt concrete.” J. Wuhan. Univ. Technol., Transp. Eng.Ed., 26(5), 566–569.

Wu, S. P., Mo, L. T., and Shui, Z. H. (2003). “Piezoresistivity of graphitemodified asphalt-based composites.” Key Eng. Mater., 249, 391–395.

Xiao, L., and Li, Z. (2008). “Early-age hydration of fresh concrete moni-tored by non-contact electrical resistivity measurement.” Cem. Concr.Res., 38(3), 312–319.

1702 / JOURNAL OF MATERIALS IN CIVIL ENGINEERING © ASCE / NOVEMBER 2013

J. Mater. Civ. Eng. 2013.25:1692-1702.

Dow

nloa

ded

from

asc

elib

rary

.org

by

MA

RR

IOT

T L

IB-U

NIV

OF

UT

on

11/2

6/14

. Cop

yrig

ht A

SCE

. For

per

sona

l use

onl

y; a

ll ri

ghts

res

erve

d.

http://dx.doi.org/10.1016/j.conbuildmat.2009.06.014http://dx.doi.org/10.1016/j.cemconcomp.2008.04.001http://dx.doi.org/10.1016/j.cemconcomp.2008.04.001http://dx.doi.org/10.1016/S0963-8695(03)00013-6http://dx.doi.org/10.1016/S0963-8695(03)00013-6http://dx.doi.org/10.1061/(ASCE)0899-1561(2007)19:5(423)http://dx.doi.org/10.1061/(ASCE)0899-1561(2007)19:5(423)http://dx.doi.org/10.1016/j.carbon.2004.12.033http://dx.doi.org/10.1016/j.carbon.2004.04.005http://www.troxlerlabs.com/3660apps.htmlhttp://www.troxlerlabs.com/3660apps.htmlhttp://www.troxlerlabs.com/3660apps.htmlhttp://www.troxlerlabs.com/3660apps.htmlhttp://www.troxlerlabs.com/3660apps.htmlhttp://handle.tamu.edu/1969.1/3898http://handle.tamu.edu/1969.1/3898http://handle.tamu.edu/1969.1/3898http://handle.tamu.edu/1969.1/3898http://handle.tamu.edu/1969.1/3898http://dx.doi.org/10.1007/BF02838422http://dx.doi.org/10.1007/BF02838422http://dx.doi.org/10.4028/www.scientific.net/KEM.249http://dx.doi.org/10.1016/j.cemconres.2007.09.027http://dx.doi.org/10.1016/j.cemconres.2007.09.027