Embed Size (px)
Citation preview
to contents
SP
EC
IAL
RE
PO
RT
99-2
0
Performance Testing ofHot-Mix Asphalt AggregatesVincent C. Janoo and Charles Korhonen December 1999
US Army Corpsof Engineers®
Cold Regions Research &Engineering Laboratory
to contents
Abstract: Hot-mix asphalt (HMA) pavements are sub-ject to thermal cracking, fatigue cracking, rutting, strip-ping, raveling, and freeze–thaw damage. Some ofthese distresses are directly affected by the choice ofaggregates. Particle shape, surface texture, particlesize, pore structure, and particle strength are the most
common characteristics cited for controlling rutting andfor maintaining adequate skid resistance. A literaturereview was conducted to evaluate commonly used andpotential test methods for evaluating hot-mix aggre-gates in term of pavement performance.
How to get copies of CRREL technical publications:
Department of Defense personnel and contractors may order reports through the Defense Technical Information Center:DTIC-BR SUITE 09448725 JOHN J KINGMAN RDFT BELVOIR VA 22060-6218Telephone 1 800 225 3842E-mail [email protected]
[email protected] http://www.dtic.mil/
All others may order reports through the National Technical Information Service:NTIS5285 PORT ROYAL RDSPRINGFIELD VA 22161Telephone 1 703 487 4650
1 703 487 4639 (TDD for the hearing-impaired)E-mail [email protected] http://www.ntis.gov/index.html
A complete list of all CRREL technical publications is available fromUSACRREL (CEERD-IM-HL)72 LYME RDHANOVER NH 03755-1290Telephone 1 603 646 4338E-mail [email protected]
For information on all aspects of the Cold Regions Research and Engineering Laboratory, visit our World Wide Web site:http://www.crrel.usace.army.mil
to contents
Special Report 99-20
Performance Testing ofHot-Mix Asphalt AggregatesVincent C. Janoo and Charles Korhonen December 1999
Prepared for
NEW HAMPSHIRE DEPARTMENT OF TRANSPORTATION
Approved for public release; distribution is unlimited.
US Army Corpsof Engineers®
Cold Regions Research &Engineering Laboratory
to contents
PREFACE
This report was prepared by Dr. Vincent Janoo and Charles Korhonen, Research CivilEngineers, Civil Engineering Research Division, U.S. Army Cold Regions Research andEngineering Laboratory (CRREL). The report was prepared in cooperation with the NewHampshire Department of Transportation and the U.S. Department of Transportation,Federal Highway Administration.
Funding for this project was provided by the New Hampshire Department ofTransportation through an intergovernmental cooperative agreement.
The work was conducted at CRREL, and the authors wish to acknowledge thecontributions of the following people for successful completion of this project: Technicalreview of the report was provided by Sally Shoop of CRREL, Richard Lane of the NewHampshire Department of Transportation, and David Hall of the Federal HighwayAdministration. The authors wish to express their gratitude to Lynette Barna, CivilEngineering Research Division, for preparation and editorial changes of the document.Assistance from the CRREL Technical Information Branch was provided by Dawn Bodenin preparation of the graphics used in this report and Maria Bergstad and David Cate inediting and publishing assistance. Recognition is also extended to Bruce Ashley of theLogistics Management Office for reproduction services.
The contents of this report reflect the views of the authors, who are responsible for thefacts and the accuracy of the data presented herein. The contents do not necessarilyreflect the official views or policies of the New Hampshire Department of Transportationor the Federal Highway Administration at the time of publication. This report does notconstitute a standard, specification, or regulation.
ii
to contents
CONTENTS
Preface ......................................................................................................................................... 1Introduction ................................................................................................................................ 1Parameters and tests indicative of rutting performance ...................................................... 2
Particle index ......................................................................................................................... 4Rugosity ................................................................................................................................. 5Uncompacted voids in aggregates ..................................................................................... 7Time index.............................................................................................................................. 9Characterization summary .................................................................................................. 9
Parameters influencing skid resistance .................................................................................. 9Roughness and texture ......................................................................................................... 10Macrotexture characterization ............................................................................................ 12Microstructure characterization .......................................................................................... 13Friction measurement devices ............................................................................................ 13Polish–wear testing devices ................................................................................................ 15British Accelerated Wear and Polishing Device ............................................................... 15Small-wheel circular track polishing machine ................................................................. 16Projection method ................................................................................................................. 16Summary of text equipment ................................................................................................ 17
Evaluating frost resistance of pavement aggregates ............................................................ 17Soundness test ....................................................................................................................... 17Unconfined freeze–thaw test ............................................................................................... 18Confined rapid freeze–thaw test ........................................................................................ 18Hydraulic fracture test ......................................................................................................... 18Cryogenic test ........................................................................................................................ 19Slow freeze test ...................................................................................................................... 19VPI single-cycle freeze test .................................................................................................. 19Iowa Pore Index test ............................................................................................................. 19
Summary and recommendations ............................................................................................ 20Literature cited ........................................................................................................................... 20Abstract ....................................................................................................................................... 23
ILLUSTRATIONS
Figure1. Influence of aggregate roundness on rutting potential of asphalt concrete
mixtures ........................................................................................................................ 22. Aggregate classification chart ........................................................................................ 33. Roundness chart for 16- to 32-mm aggregates ............................................................. 34. Degree of angularity ........................................................................................................ 45. Comparison of paraticle index values for rounded and crushed coarse aggregates .. 46. Particle index for different aggregate gradations ........................................................ 5
iii
to contents
Figure7. Description of pouring device ........................................................................................ 68. Effect of aggregate type on specific rugosity ............................................................... 69. NAA particle shape and texture void contents vs. percent natural sand in fine
aggregate ...................................................................................................................... 710. Test apparatus for ASTM C 1252 and modified ASTM C 1252 ................................. 811. Modified NAA particle shape and texture and void contents .................................. 812. QMOT time index test apparatus .................................................................................. 913. Road surfacing textural characteristics ......................................................................... 1014. Correlation between macrotexture, microtexture, skid resistance, and speed ....... 1015. Pavement surface characteristics ................................................................................... 1116. Effect of pavement texture on skid resistance.............................................................. 1117. Containers of test materials for sand patch test .......................................................... 1218. Circular patch of test material ........................................................................................ 1219. Measurement of diameter of sand patch ...................................................................... 1220. British portable skid resistance tester ............................................................................ 1321. Relationship between British pendulum number and Illinois skid trailer number ... 1422. Correlation between variable-speed number and skid trailer number at 40 mph ..... 1423. British wear and polishing device ................................................................................. 1524. Water projection device ................................................................................................... 1625. Specimen for projection method .................................................................................... 1626. Correlation between on-road British pendulum number and polishing by
projection coefficient Cpp ............................................................................................ 17
TABLE
Table1. Recommended aggregate tests ....................................................................................... 1
iv
to contents
INTRODUCTION
When one thinks of hot-mix asphalt (HMA)pavement distresses, the following come to mind:thermal cracking, rutting, fatigue cracking, ravel-ing, and moisture-induced damage (stripping).Some of these, such as low temperature andfatigue cracking, are more directly related to theproperties of the asphalt cement. The aggregatesplay a minor role in crack retardation once a crackhas formed. Rutting and moisture-induced dam-age are not only dependent on the asphalt contentbut on the characteristics of the aggregates in themixture as well, and thus a proper selection ofaggregates can reduce rutting and moisture-induced damage to HMA. For example, ruttingcan be reduced by the use of large aggregates,and/or angular rough coarse and fine aggre-gates. Raveling is more of a mixture-proportion-ing problem, i.e., the amount of asphalt cement inthe mixture is critical. Low asphalt content canaccelerate the raveling of an HMAmixture. Another distress, notusually mentioned, is the amountof change in skid resistance. Hereagain, a proper choice of aggre-gate surface characteristics andshape qualities can significantlyenhance skid resistance.
Although aggregates consti-tute approximately 90% of anHMA mixture, there is no perfor-mance grading of aggregatessimilar to the Strategic HighwayResearch Program (SHRP) PGsystem for asphalt cement.Many ASTM and AASHTO indextests attempt to characterize thequalities of aggregates neededfor HMA mixtures; they measuresize and gradation, aggregate
Performance Testing of Hot-Mix Asphalt Aggregates
VINCENT C. JANOO AND CHARLES KORHONEN
cleanliness, toughness/hardness, durability sound-ness, surface texture, particle shape, absorption,and affinity for asphalt. These tests however, donot give any clear indication of the performanceof the aggregates in HMA with respect to the dis-tresses mentioned above. SHRP realized the im-portance of aggregate properties on the perfor-mance of HMA pavements, but due to a lack offunds and time, the selection of aggregate proper-ties for the SUPERPAVE mixes was based on“expert consensus.” SHRP retained some previ-ously used aggregate tests, such as determiningthe proportion of flat and elongated particles,sand equivalent tests for determining the amountof plastic fines and dust in the fine aggregates,and the shape and angularity of both the coarseand fine aggregates. No tests were recommendedfor skid resistance.
Realizing this shortfall, the National Coopera-tive Highway Research Program (NCHRP) con-ducted a study on the relationship between exist-
Table 1. Recommended aggregate tests (Kandhal and Parker 1998).
Aggregate Fatigue Permanent Moisture-inducedproperties cracking deformation Raveling damage (stripping)
Gradation and size x x
Uncompacted void contentof coarse aggregates x x
Uncompacted void contentof fine aggregates x
Methylene blue tests offine aggregates x
Methylene blue test ofP200 materials x
Micro-Deval tests x
Magnesium sulfatesoundness tests x
Particle size analysis ofP200 material x x
to contents
ing aggregate tests and asphalt concrete perfor-mance (Kandhal and Parker 1998). The study,based only on laboratory evaluation, recom-mended a set of aggregate tests that relate to rut-ting, fatigue cracking, raveling, and strippingperformance of HMA pavements (Table 1).
This report discusses test methods proposed byNCHRP for rutting susceptibility and focuses onthe two areas not covered by the NCHRP report:skid resistance and freeze–thaw durability charac-terization. The report also focuses on aggregatecharacterization and not on test methods thatinvolve mixture testing.
PARAMETERS AND TESTS INDICATIVEOF RUTTING PERFORMANCE
Particle shape, surface texture, particle size,pore structure, and particle strength appear to bethe most common characteristics cited for con-trolling rutting.
Permanent flow or rutting from traffic loadingis due to densification and plastic flow of theHMA mixture at high temperatures. Factors suchas asphalt content, asphalt grade, air voids andaggregate characteristics, construction practices,temperature, and increase in traffic load/repeti-tions all have an influence on the rutting potentialof a mixture. Although all these factors are impor-tant, the effect of aggregates, which comprise upto 90% of the mixture, plays a significant role incontrolling rutting, as seen by the recent intro-
duction of stone matrix asphalt (SMA) mixturesin the USA.
To minimize rutting, aggregate interlock iscrucial. The shape, angularity, and surface textureof the aggregate affect the interlock. However, thisinterlock appears to be critical for the fine aggre-gates, as noted by Uge and Van de Loo (1974),who reported that use of rounded coarse aggre-gate with crushed fine aggregate also producedrut-resistant HMA mixtures (Fig. 1). A discussionof the various factors that affect pavement ruttingcan be found in Janoo (1990).
The quantification of the shape, angularity,and surface texture of an aggregate is difficult butnot impossible. Several methods involve eitherdirect measurements of aggregate or indirectinference from aggregate properties or from mix-ture testing. For example, for coarse aggregates,engineers have developed visual classificationmethods to characterize shape and angularity orindex tests that use engineering properties suchas porosity to speculate on the shape, angularity,and surface texture of aggregates. In the indextest developed by engineers, however, it is notpossible to quantify separately the shape, theangularity, or the surface texture. Usually theyare lumped together as geometric irregularities.Details on quantification of the shape, angularity,and surface texture of aggregates can be found inJanoo (1998). For coarse aggregates, geologistshave a sophisticated system that involves physi-cal measurement of the aggregates. As used by
2
Figure 1. Influence of aggregate roundness on rutting potential of asphalt con-crete mixtures.
Sbit, Bitumen Stiffness Modulus (N/m2)
103 104 105 106
Gm
ix, S
hear
Stif
fnes
s M
odul
us (
N/m
2)
106
107
4x108
30% Crushed Materials (0/2 mm)70% Rounded Materials (2/12 mm)
100% Crushed Aggregates
100% Rounded Aggregates
25% Rounded Materials, 75% Crushed Materials
(0.08/2 mm) (2/12 mm)
Gmix =τ mix
γmix
to contents
petrologists, the shape of a coarse aggregate canbe described by its length, width, thickness, sphe-ricity, roundness, and angularity. Aggregate clas-sification charts to evaluate shape, roundness,and angularity of aggregates are shown in Figures2, 3, and 4. Surface texture is more difficult to deter-mine. Traces of magnified surfaces have beenused to describe surface texture (Wright 1955).Others, such as Barksdale and Itani (1994), devel-oped roughness scales to describe the texture ofaggregates.
For fine aggregates, shape and angularity canbe described with image analysis. Computer-
based image analysis has been developed for anal-yzing aggregate shape, angularity, and rough-ness. A feasibility study conducted for the FederalHighway Administration (FHWA) on this tech-nology (Wilson et al. 1995) showed that imageanalysis is a viable tool for distinguishing theshape and angularity of fine sands (manufac-tured vs. natural). The method involves first cap-turing magnified images of the aggregates usinga high-resolution video camera, and then usingan image analysis program to identify and separatethe objects and trace the edges of the aggregates.Based on the traces, algorithms in the programsare used to determine the different characteristicsof the aggregates. The Quebec Ministry of Trans-portation (QMOT) routinely uses image analysisof the HMA fine aggregates to distinguish round-ness and angularity of the fine aggregates. Addi-tional descriptions of this method can be found inJanoo (1998).
An alternate approach taken by engineers is toinfer these characteristics from the mass propertiesof the aggregates. Several indices such as angu-larity number, time index, particle index, andrugosity have been identified in the literature.
For coarse aggregates, the angularity number(AN) developed by Shergold (1953) is recom-mended by the British Standards Institution(1989) for indexing the angularity of natural andcrushed aggregates used in concrete. This tech-nology can be easily transferred to HMA. Sher-gold found that when the aggregates were com-pacted in a prescribed fashion, the percentage of
Figure 3. Roundness chart for 16- to 32-mm aggregates.
Roundness = 0.1 Roundness = 0.2 Roundness = 0.3
Roundness = 0.6 Roundness = 0.7 Roundness = 0.8
Roundness = 0.4 Roundness = 0.5
Roundness = 0.9 Broken Pebbles
0.3 0.40.4
0.50.5
0.4
0.8
1.0
0.6
0.4
0.2
0.00.80.0 0.2 0.4 0.6 1.0
Inte
rmed
iate
/Lon
g =
q
Short/Intermediate = p
F=0.
33F=
0.67
F=1.0
F=1.5
F=3.0 ψ = 0.4
ψ = 0.5
ψ = 0.6
ψ = 0.8
ψ = 0.9
ψ = 0.95
DISC
RODBLADE
CUBIC
ψ = 0.7
Figure 2. Aggregate classification chart.
3
to contents
voids in the aggregate mass decreased as theaggregates became more rounded. He also foundthat as the amount of round gravel increased in amixture of natural and crushed aggregates, thepercentage of voids also decreased. Using roundgravel as a reference point, he characterized theangularity of all other aggregates as the differ-ence between the percent voids and 33%. AN isfound to range between 0 and 12:
AN = percent voids – 33. (1)
Particle indexAnother index for characterizing the coarse
and fine aggregates in an HMA mixture is the par-ticle index (Ia). The particle index is based on theconcept that the shape, angularity, and surfacetexture of uniformly (single) sized aggregateaffects not only the void ratio but also the rate atwhich the voids change when compacted in astandard mold (Huang 1962). The particle index(Ia) is calculated using eq 2:
4
0–99 100–199 200–299 300–399 400–499 500–599 600–699 700–799
800–899 900–999 1000–1099 1100–1199 1200–1299 1300–1399 1400–1499 1500–1599
Figure 4. Degree of angularity.
Figure 5. Comparison of particle index values for rounded and crushed coarseaggregates.
16
No. 4 – 6
Par
ticle
Inde
x
Gradation of Samples
14
12
10
8
6
4
2
6 – 9 9 – 13 13 – 19 19 – 25
Crushed Aggregate
Natural Rounded Aggregate
to contents
Ia = 1.25V10 – 0.2550 – 32 (2)
where V10 = % of voids in aggregates at 10 strokesper layer and V50 = % of voids in aggregates at 50strokes per layer.
The results (Fig. 5) indicate that the test methodis capable of distinguishing the differencebetween natural rounded and rough, angularaggregates by the increasing particle index value.
The test is time-consuming because the parti-cle index is developed for individual sieve sizeaggregates. The Michigan Department of Trans-portation (Michigan DOT 1983) studied the effect
of combining the different sieve sizes on the parti-cle index. The effectiveness of the index in differen-tiating between natural and crushed aggregateswas consistent for all combinations (Fig. 6).
RugosityAnother method based on flow of aggregates
through a given sized orifice can also be used todifferentiate physical aggregate properties. Theinitial work with HMA aggregates was done byTons and Goetz (1968), who developed the pack-ing volume concept to characterize the shape,angularity, and roughness of the aggregates usedin bituminous mixtures. The test was developedfor both the coarse (12 mm) and fine fractions.They found that although the shape of the parti-cle could be quantified separately, it was difficult
to separate the interaction of angularity androughness on aggregate performance. They pro-posed that the effect of both angularity androughness be combined into a single term called‘rugosity.’
Later, Ishai and Tons (1971) developed anindex to describe this rugosity, called the specificrugosity (Srv). Srv will be approximately equal tozero for smooth spherical particles:
S
VVrv
sr
p
px
ap
GG
=
= −
100 100 1
where Srv = specific rugosity (%)Vsr = volume between the packing vol-
ume membrane and the volume ofmacro- and microsurface voids
Vp = packing volume of the particleGpx = packing specific gravityGap = apparent specific gravity.
The apparent specific gravity can be calculatedusing ASTM C 127-88 (1995), Standard Test for Spe-cific Gravity and Absorption of Coarse Aggregate(AASHTO T 85-85 1998). Gpx is determined fromthe pouring test developed by Ishai and Tons(1977). The pouring test involves taking twosingle-sized samples and pouring them into astandard container using a standard procedure.One of the particles used is a standard (smooth,
16
Par
ticle
Inde
x
14
12
10
8
6
4
2
No. 4 – 9C-4
No. 4 – 13C-3
No. 4 – 19C-2
No. 4 – 25C-1
MeanStandard Deviation
CrushedAggregate
Natural RoundedAggregate
Gradation of Samples
Figure 6. Particle index for different aggregate gradations.
(3)
5
to contents6
Figure 7. Schematic description ofpouring device.
1/2 – 5/8 in.No. 3 – 4No. 8 – 10No. 20 – 30No. 60 – 80
Ave
rage
Pac
king
Vol
ume
of a
Par
ticle
(in o
ne s
ize
frac
tion,
Vp
- cm
3 )
1
10–1
10–2
10–3
10–4
10–5
10–6
0 4 8 12 16 20 24 28 32 36
Specfic Rugosity, Srv (%)
Beach Pebble
NaturalGravel
Crushed Gravel
Sandstone
Figure 8. Effect of aggregate type onspecific rugosity.
φh
Container
H
b
c
Supports
Bin
Funnel
D
Bin Shutter
Large Pan forParticle Collection
a
Bin diameterFunnel orifice diameterBin heightAggregate headPouring heightContainter diameterContainer height
=======
DacbHφh
to contents
spherical glass beads) with a known packing spe-cific gravity, Gps. The other sample is the test par-ticles for which Gpx is sought. Gpx is a function ofthe ratio of the weight of the test particles to thestandard particles:
G
WW
Gpxx
sps=
ΣΣ
. (4)
A schematic of the pouring test is given in Fig-ure 7. Typical results on the effect of different aggre-gates on the specific rugosity are shown in Figure8. The specific rugosity increases as the angu-larity and roughness increase. One drawback ofthis test is that it is time-consuming because testshave to be done on individual sieve sizes.
Uncompacted voids in aggregates(ASTM C 1252)
This test method is similar to the specific rug-osity test described above. The exception is thatthe aggregates need not be separated into varioussieve sizes. The test for fine aggregates has beenstandardized by ASTM C 1252 (1995), StandardTest Methods for Uncompacted Void Content of FineAggregate (as Influenced by Particle Shape, SurfaceTexture, and Grading). Initially developed by theNational Aggregate Association (NAA), itinvolves allowing the fine aggregates to fall freelyinto a calibrated cylinder from a specified height.The excess material is removed and weighed. The
7
weight of aggregate together with the bulk specificgravity of the aggregate are used to calculate theuncompacted void content:
UCV
VM
GV
= ×cyl
sb
cyl100 (5)
where UCV = uncompacted void content (%)Vcyl = volume of cylinder (cm3)
M = mass of aggregate in cylinder (g)Gsb = bulk specific gravity of aggregates.
Figure 9 shows the UCV increasing with increas-ing percent of crushed fine particles. Method A inFigure 9 refers to a standard gradation of the fineaggregates (i.e., the gradation of selected sievesizes), and method C refers to the gradation of theminus no. 4 portion of the as-received material.The standard gradation in method A consists of agradation made from a known weight of materialfrom specific sieves.
Aldrich (1996) modified this test for determin-ing the uncompacted void content of coarseaggregates. A schematic of both the fine andcoarse aggregate test apparatus is shown in Fig-ure 10. The test coarse aggregate passed the 19-mm sieve and was retained on the no. 4 sieve. Atest protocol similar to that used in ASTM C 1252(1995) was developed for the coarse aggregates.Utilizing method 1, 5000 g of as-received materialis used, or individual aggregate sizes ranging
Figure 9. NAA particle shape and texture void contents vs. percent natural sand in fineaggregate
NA
A U
ncom
pact
ed V
oid
Con
tent
s
48
44
40
36
32
Percent Natural Sand in Fine Aggregate
0 20 40 60 80 100
Method AR2 = 0.963Method CR2 = 0.868
to contents8
60°
Fine Aggregate
102 mm
13 mm 114 mm
60°
Coarse Aggregate
152 mm
102 mm 114 mm
Figure 10. Test apparatus for ASTM C 1252 and modifiedASTM C 1252.
Mod
ified
NA
A –
Voi
d C
onte
nts
52
48
44
40
36
Percent Crushed Coarse Particles
0 20 40 60 80 100
As Prepared
R2 = 0.936
a.
Figure 11. Modified NAAparticle shape and texturevoid contents vs. (a) percentcrushed particles (Aldrich1996), and (b) particle indexvalues of coarse aggregate.
Mod
ified
NA
A –
Voi
d C
onte
nts
52
48
44
40
368 10 12 14 16 18
As Prepared
R2 = 0.936
b.
Particle Index Values – Coarse Aggregate
to contents
from 19 to 25 mm, 25 to 9.5 mm, and 9.5 mm to no.4 are tested separately and the UCV averagedfrom the individual fraction sizes. As shown inFigure 11a, the modified NAA method developedby Aldrich (1996) for the coarse fraction of theHMA mixture clearly indicates that the UCV in-creases with increasing percent of crushed coarseparticles. In addition, Aldrich (1996) showed agood correlation between the results from themodified NAA test for coarse aggregates and theparticle index value (Figure 11b). Overall, this is afairly simple test to conduct, and correlation be-tween rutting potential and aggregate character-ization can be easily developed.
Time indexSimilar in concept to the uncompacted void
tests, a test method developed by the QuebecMinistry of Transportation characterizes the angu-larity and surface texture of aggregates using therate of flow of the aggregate through a known-diameter opening (Fig. 12). Additional details onthis apparatus can be found in Janoo (1998).
The index, the flow coefficient (Ce), is a func-tion of the time it takes 7 kg of material to movethrough a 60-mm-diameter opening for materialpassing the 20-mm sieve and retained on the4-mm sieve. A different size is used for the fineaggregates. In addition, the flow coefficientdepends on the bulk specific gravity. Based on
limited tests, the flow coefficient for crushedgravel was 92.2 and for crushed stone was 115(Janoo 1998).
Characterization summaryIn summary, it was found that several methods
are available for characterizing the shape, angu-larity, and surface texture of HMA aggregates. It isdifficult to separate the effect of individual fea-tures of the aggregate, with the exception of petro-graphic methods. All of the tests tend to providethe effect of the overall shape, angularity, androughness. The NAA and modified NAA testsseem the easiest to use and implement. The timeflow test used by the Quebec Ministry of Trans-portation also appears to be easy to implement.
PARAMETERS INFLUENCINGSKID RESISTANCE
For HMA pavements, skid resistance dependson the friction developed between the pavementsurface and the tire. This friction is dependent onthe microscopic and macroscopic roughness of thepavement surface, the polish–wear characteristicsof the aggregates, and the ability of the surface todrain (Beaton 1976). Marek (1972) reported that50% of the initial skid resistance is lost during thefirst two years of pavement service and that thesingle most important factor that affects the reduc-tion of skid resistance is the polishing characteris-tics of the aggregates in an HMA mix. Other fac-tors that affect the skid resistance of the pavementsurface are wetness of the surface, seasonal varia-tion, and temperature.
The skid resistance of a wide range of dry sur-faces is high and fairly constant. When these sur-faces get wet, however, the skid resistance dropsand is very dependent on the surface type. Thereis seasonal variation: the skid resistance is higheron wet roads during the winter than in the sum-mer (Hosking and Woodford 1976) because inwinter the road surface is contaminated by sand-ing and salting and the removal of fine materialincreases the roughness of the microtexture. In thesummer, continuous abrasion of the macrotextureproduces fines that coat the microtexture andreduce the skid resistance.
The friction of the surface is indexed from fieldtests following various standardized tests (ASTME 274 [1994] and E 1551[1994]). The procedure isbased on measurements of the frictional forcedeveloped between a standard tire at a standardinflation pressure and the wetted pavement sur-Figure 12. QMOT time index test apparatus.
9
to contents
face. Since the frictional force between the tire andthe pavement surface is a function of the vehiclespeed, it is usually reported at a speed of 65 km/hr. The frictional coefficient is multiplied by 100and reported as the skid number (SN).
Roughness and textureThe microscopic and macroscopic roughness
are characterized by the texture of the aggregatesurface and by the overall roughness of the pave-ment surface due to protrusion of the aggregatesurfaces, respectively. Macro- and microtextures areillustrated in Figure 13. Macrotexture has the great-
est influence on the change in friction with speed.Figure 14 illustrates the effect of the macro- andmicrotextures on skid resistance as a function ofspeed. Clearly, to maintain a constant high skidresistance value at various speed levels, the pave-ment surface has to have both good macro- andmicrotextures. The change in the texture dependson the aggregate resistance to fragmentation, wear,and polishing. Aggregate fragmentation and weardepend on the toughness and hardness of theaggregate minerals and the aggregate itself. Pol-ishing depends on the difference in hardness of the
different minerals present in the aggregate (Trem-blay et al. 1995).
The macroscopic roughness is provided byspacing between the aggregate from 0.5 to 50 mm(horizontal) and 0.2 to 10 mm (vertical) (Donbav-and 1989). This spacing provides channels forrapid drainage of water from the surface and isimportant at vehicle speeds greater than 50 km/hr (Donbavand 1989). Therefore, at speeds higherthan 50 km/hr, the skid resistance depends pri-marily on the spacing provided by the coarseaggregates (Beaton 1976).
The microtexture refers to the irregularities on
the surface of the coarse and fine aggregates. It isalso affected by the amount of fines in the HMAmix (Tremblay et al. 1995). The effect of differenttypes of pavement surfaces on the macro- andmicrotextures is illustrated in Figure 15. Theterms viscous and dynamic fluid pressure allevi-ation in Figure 15 relate to the natural flow andflow under the impact of a moving tire. The effectof texture on skid resistance is shown in Figure16. Several observations can be made. First, themicrotexture can provide adequate skid resis-tance at lower speeds, such as in urban areas (line
10
Good Macrotexture, Good Microtexture
Poor Microtexture, Good Macrotexture
Good Microtexture, Poor Macrotexture
Poor Microtexture, Poor Macrotexture
Ski
d R
esis
tanc
e
Speed
Figure 14. Correlation between macrotexture, microtexture, skidresistance, and speed (Tremblay 1995).
MacrotextureMicrotexture
BINDER
Figure 13. Road surfacing textural characteristics (Tremblay 1995).
Good Macrotexture, Good Microtexture
Poor Microtexture, Good Macrotexture
Good Microtexture, Poor Macrotexture
Poor Microtexture, Poor Macrotexture
Speed
Ski
d R
esis
tanc
e
to contents
3). However, it is important at all speeds and is avital component for overall skid resistance. Sec-ond, the use of rounded coarse or fine aggregatesresults in low skid resistance values (sample 2and 4 in Figure 16).
Several laboratory tests are available to studythe change in macro- and microtextures underrepeated loading in the laboratory. Methods formeasuring texture can be categorized into volu-metric, profile, topography, contact, drainage,and miscellaneous measurements (Rose et al.1972). Brief descriptions of some of the tests areprovided here. Details and references to theseand other devices can be found in Rose et al.(1972).
It must be realized that different test methodsproduce different results. It may not be possibleto correlate results from one test to another. Themost common test used today for determiningthe macrotexture depth is the sand patch test,described in the next subsection.
Volumetric measurements are commonly usedfor macrotexture depth determinations. Theseinclude the sand patch, modified sand patch,vibrating sand patch, sand track, grease smear,and silicone putty tests. These tests differ in thetest materials and the patch geometry.
Figure 16. Effect of pavement texture on skidresistance.
Figure 15. Pavement surface charac-teristics.
PavementTexture
Fluid PressureAlleviation
Micro Macro Viscous Dynamic
Smooth Low Low Poor Poor
Smooth Stones Low Medium Poor Good
SandpaperyHigh Low Excellent Poor
Fractured StonesHigh Medium Excellent Good
GroovesLow High Fair Excellent
PorousMedium High Good Excellent
GroovesHigh High Excellent Excellent
Speed, mph
Ski
d N
umbe
r, S
N
00
20 40 60 80
20
40
60
80
5
3
4
21
Smooth
Fine Textured, Rounded
Fine Textured, Gritty
Coarse Textured, Rounded
Coarse Textured, Gritty
1
2
3
4
5
11
to contents
Profile measurement tests that could be adaptedfor the laboratory include the texturemeter, lineartraverse, light, and wear/roughness meter tests.The texturemeter, developed by the Texas Trans-portation Institute, consists of a series of evenlyspaced vertical parallels in a frame. With theexception of two, the rods can move verticallyand are independent of one another. A string isattached to the movable end and to the frame.The texturemeter produces a straight line on asmooth surface and a dial indicator has been cali-brated to zero. If there are any irregularities, thestring produces a zig-zag line and results in a dialreading greater than zero. The reading is propor-tional to the coarseness of the macrotexture: thecoarser the surface, the higher the reading.
Other methods that use microscopes and/orlight are the linear traverse, light, and the wearand roughness meter. The linear traverse methoddeveloped by the Kansas Highway Commissionuses a motorized lathe and a stereo microscopewith the shaft of a potentiometer attached to themicroscope focusing shaft. The specimen ismoved transversely under the microscope. Theoperator keeps the surface under constant focus,thus changing the potentiometer voltage. Thisresults in a tracing of the surface texture.
Surface texture can also be determined by lightsectioning. In this method, a beam of light ispassed through a slit at an angle. Based on thereflection of the light, the apparent profile heightcan be determined mathematically.
The wear/roughness meter also uses light tomeasure the mean wear height and mean texture.The data from this device provide a surface plotof maximum depth and distribution of the peaksof the observed surface. Details on this device canbe found in Rosenthal et al. (1969).
Texture can also be quantified with topographicmeasurements of pavement surface using stereophotography. Pairs of photographs are taken
using a specially designed camera. With the aidof a stereo comparator and a parallax bar, relativeheights can be measured using stereo photos ofsuccessive points on the surface.
Other methods such as casting, molding, ink,and photographic emulsion prints have beenused to obtain the surface texture. Detailed studyof the magnified cast or print is then made in thelaboratory.
Macrotexture characterization
Sand patch testThe sand patch test (ASTM E 965 1995) is used
to assess the average macrotexture depth of apavement surface. Values over 0.80 mm are con-sidered excellent. Below 0.60 mm, the macrotex-ture is inadequate and can lead to extremely slip-pery conditions (Tremblay et al. 1995).
The test involves placing a known volume ofmaterial on the dry surface of the test specimen.Figure 17 shows the predetermined volume oftest material in the container and test specimen.The test material is made from solid round glassspheres graded between the no. 60 and no. 80sieves. This material is then spread into a circularpatch with a disk (Fig. 18). The diameter of thepatch is measured (Fig. 19). At least four meas-urements are made, and the average diameter isused to determine the average surface macrotex-ture depth (MATXd). The macrotexture depth iscalculated from the following equation:
MATX
VD
d = 42π
(6)
where V is the test material volume and D is theaverage diameter of area covered by the testmaterial.
12
Figure 17. Containers of test materials for sand patchtest.
Figure 18. Circular patch of test material.
to contents
Surface drainage testThe amount of time for a surface to drain can
also be used as an indicator of the surface texture.A drainage meter (Moore 1966), 13 mm in diame-ter and 13 mm in height containing a known vol-ume of water under atmospheric pressure, isplaced on the pavement surface. A rubber ring isglued to the bottom of the cylinder and acts simi-larly to tire treads on the pavement surface. Theamount of time taken for the water to flow out of thetank is related to the macrotexture of the pave-ment surface. High flow rates indicate high macro-texture depths.
Microstructure characterizationThe change in the microstructure of a pavement
surface is difficult to quantify and can be inferredfrom the changes in its frictional properties. This isdone using an accelerated polishing device and afriction measurement system. The frictional prop-
erties of the pavement surface are usually deter-mined in the laboratory using either the Britishportable skid resistance tester or the North Caro-lina State University variable-speed frictiontester. Both these test methods are standardizedby ASTM.
With respect to accelerated polishing, the testsare mostly done on wheel tracking systems.Examples of such tests are the small-wheel circu-lar track wear and polishing machine, the Georgiaskid tester, and the British polishing wheel. Someof these devices were fabricated by individualstate DOTs or by universities and are not avail-able commercially. The French have developed apolishing method using high-pressure water on apavement surface. The equipment is less complexthan the circular wheel and appears to correlatewell with changes in the frictional properties dueto polishing. The following discussion will startwith the two friction testers currently standard-ized by ASTM followed by descriptions of severalaccelerated polishing devices.
Friction measurement devices
British portable skid resistance testerThe British skid resistance tester (BSRT) (Fig.
20), also called the British pendulum skid tester,can be used both in the laboratory and in thefield. The tester consists of a pendulum that isfree to swing through 180°. A rubber shoe slider isattached to the bottom end of the pendulum. Inthe position shown in Figure 20, the pendulumcontains potential energy. When released, thepotential energy is converted to kinetic energythat is dissipated by friction on the rubber shoe,which slides over the surface. A graduated scaleon the pendulum registers the maximum swing
Figure 20. British portable skid resistance tester.
Figure 19. Measurement of diameter of sand patch.
13
to contents14
of the pendulum. Test method ASTM E 303 (1995),Standard Test Method for Measuring Surface Fric-tional Properties Using the British Pendulum Tester,uses the BSRT to measure the frictional propertiesof the microtexture.
The result from the BSRT is the British pendu-lum number (BPN), which relates well to skidnumbers obtained in the field (Fig. 21). (The termClass I in Figure 21 refers to an Illinois DOT roadclassification system.) Sometimes, the BPN valueis converted and reported as a polished stonevalue (PSV). The PSV value is the BPN value atthe end of 6 hours of testing.
The advantages of this device are that it is por-table, has a low initial cost, and can test in differ-ent orientations. Its disadvantages include thatresults from coarse macrotexture are question-able, it can only simulate low-speed skidding(780 m/hr), and it requires laborious calibration.To maintain consistency between differenttesters, a calibration process is used with a stan-dard aluminum surface. Good agreement hasbeen reported between two testers calibratedwith this procedure.
NCSU variable-speed friction testerThe variable-speed friction tester (VST) is sim-
ilar to the BSRT except that it has a locked-wheelsmooth rubber tire at the bottom of the pendu-lum. It can also be used in the laboratory and inthe field. One advantage of this device is that itcan be used to simulate variable vehicle speed,something the British skid resistance tester can-
not. Another advantage of this device over theBSRT is that it can be used to determine the fric-tional properties of coarse and open gradedmixes. As with the BSRT, this test has been stan-dardized by ASTM under E 707-90 (1995).
As with the BSRT, the friction value is readfrom the scale and is presented as the variable
60
0
Illin
ois
Ski
d N
o.
BPN
R = 0.90
50
40
30
20
10
0
10 20 30 40 50 60 70
Class I
Figure 21. Relationship between British pendulum number and Illinois skidtrailer number.
100
20
Var
iabl
e-S
peed
Tes
ter
Num
ber,
VS
N
Skid Trailer Number, SN
80
60
40
20
0
40 60 800
Figure 22. Correlation between variable-speed numberand skid trailer number at 40 mph.
to contents
speed number (VSN). The VSN relates well to theskid numbers collected from the field (Fig. 22).This device is not as commonly available as theBSRT.
Polish–wear testing devicesWith respect to testing, five characteristics need
to be evaluated for selection of skid-resistantaggregates: texture, shape, mineral constituents,chemical composition, and gradation (Beaton1976). The shape (angularity) of the aggregatescan be determined with the tests discussed in theprevious section. Mineral constituents can bedetermined from petrographic analysis. Trem-blay et al. (1995) concluded that aggregate type(mineral constituents) plays a significant role inmaintaining skid resistance and that the skidresistance from limestone/dolomite aggregatestends to wear quickly and thus decrease rapidly.Volcanic aggregates also showed low but stableresistance to skidding, but their resistance towear was high, and pavement surfaces with thesetype of aggregates tended to maintain their macro-texture over time. Sandstone made from quartz,feldspar, and clay minerals showed the highestresistance to polishing. Marek (1972) came to sim-ilar conclusions and reported ‘polish resistant’aggregates included siliceous gravels, granites,diabase, quartzite, sandstones, and expandedshales. This corresponds to the hardness of theminerals that comprise these rocks.
The mineral constituents give information onthe wear-polish characteristics of the aggregatesand in turn on the texture of the pavement sur-face. However, if used alone, the information hasto be indexed to field performance. The acidinsoluble test (ASTM D 3042 1995) can also beused to characterize the wear-polish characteris-tics of limestone aggregates, which can exhibitskid resistance ranging from extremely slipperyto very good (Marek 1972) due to their mineralogi-cal and textural properties. The interpretationand use of the test results are controversial (Bea-ton 1976), however, and the acid insoluble test isrecommended for laboratory evaluation and not asa primary indicator of wear-polish.
Currently, NHDOT uses the Los Angeles abra-sion test to evaluate the wear characteristics ofHMA aggregates. This test is described in AASHTOT-96-94 (1998), Standard Test Method for Resistanceto Degradation of Small Size Coarse Aggregate byAbrasion and Impact in the Los Angeles Machine. Itinvolves placement of a known weight of materialin a drum. Several steel spheres of known weights
are placed in the drum together with the aggre-gates. The drum is then rotated, and the steel ballcreates an impact load on the aggregates. After500 revolutions, the material finer than the no. 12sieve is weighed and any loss is noted. The L.A.abrasion number is the percent passing the no. 12sieve. No standard value is mandated, and moststates tend to limit it to 40. The L.A. abrasionnumber does not correlate well with field perfor-mance. High values for slag and limestone aggre-gates have been generated from this test, andpavements constructed with these aggregateshave shown good performance in the field.
British Accelerated Wear and Polishing DeviceThe British Accelerated Wear and Polishing
Device (BAWPD) was designed to acceleratewear and polish of a pavement surface under apneumatic tire approximately 20 cm in diameterat an inflation pressure of 310 kPa. The speed ofthe tire is approximately 420 m/hr (Fig. 23).
Fourteen aggregate test specimens aremounted on a 41-cm-diameter and 6-cm-widewheel. One drawback with this method is that the
Figure 23. British wear and polishing device.
15
to contents
samples are curved and need to be made manual-ly. The wheel is pressed onto the surface with anormal load of 391 N. It usually takes about 6 hrto complete the test. The tire is free to rotate on itsaxis and is driven by the friction between the gritand the pavement samples. Coarse grit is used toaccelerate the wear and fine grit to accelerate pol-ishing. Water is used to wet the pavement sur-face. Additional details on the test equipment andprocedure can be found in the British standardprocedure BS 812 (British Standards Institution1989).
Small-wheel circular track polishing machineThe small-wheel circular track polishing
machine is similar to the BAWPD discussedabove (ASTM E 660 1995), but the track is in thehorizontal plane and can hold up to 12 circular,15-cm-diameter test specimens. The test track is91 cm in diameter and has four smooth wheelsdriven at a rate of 30 rpm around the test track.This device is currently being updated by NorthCarolina State University. Instead of a circulartrack, the device will be a linear track.
Projection methodA new accelerated polishing method using
high-pressure water was developed jointly byLaboratoire des Chaussées of the Ministére desTransport du Québec (MTQ) and the French Lab-oratoire Ponts et Chaussées (LRPC). This methoduses high-pressure water (10 MPa) and a fineabrasive pointed at an angle of 40° to the pave-ment surface. The sample is placed on a table thatis computer-controlled to move in increments of0.25 mm in the XY direction (Fig. 24). The waterand abrasive are collected in the chambers belowthe test device. The fine abrasive is recycled forfuture testing. The sample is 150 mm by 100 mm(Fig. 25) and can be manufactured mechanicallyusing a rolling wheel compactor or a kneadingcompactor. The test takes 3 hr per sample.
The sample is removed and its friction proper-ties are determined using the British skid resis-tance tester. The frictional properties obtainedfrom the tester are identified as the polishing byprojection coefficient (Cpp). Figure 26 showsa reasonable relationship between the Cpp andon-road BPN values.
16
Figure 24. Water projection device. Figure 25. Specimen for projection method.
to contents
Summary of test equipmentIn summary, the British skid resistance tester is
commonly used to characterize the changing fric-tional properties of pavement surface. In the U.S.,laboratory-accelerated polishing tests are uncom-mon. The frictional values (BPN) are related to fieldmeasurements of skid number. In other countries,such as Canada, laboratory accelerated tests onaggregates are routinely conducted using theBritish wear and polish device or using the pro-jection method (in Quebec).
EVALUATING FROST RESISTANCEOF PAVEMENT AGGREGATES
The durability of aggregate depends on theamount of water that freezes within it and on itsability to resist and/or accommodate the result-ing expansion. Thus, durability is related to porevolume—more precisely to pore size distribu-tion—and to the aggregate’s modulus. Typically,only those pores that are narrow enough to absorbwater and at the same time are large enough thatwater can freeze at normal temperatures are theproblem.
Numerous tests have been proposed over theyears to study the durability of constructionmaterials exposed to freezing and thawing, butpavement performance is still the best measure ofaggregate quality. This section reviews testingprocedures for those most closely related to test-
ing and selecting aggregates for use in coldregions pavements and identifies where addi-tional research is needed.
Soundness testMost highway departments rely on the sulfate
soundness test to determine the frost resistance ofaggregate. This is one of the earliest laboratorytests used to predict the durability of aggregates;it was mentioned in the 1800s and, after somemodifications, it has become today’s AASHTOT 104 (1990) (ASTM C 88 1981). Aggregates arerepeatedly soaked in either sodium sulfate ormagnesium sulfate solutions and are then driedin an oven. The test simulates the expansive forceof water freezing within the aggregate pores bygrowing sulfate crystals in the pores.
The appeal of this test is its simplicity and thespeed at which it can be conducted. The majorproblem is that it does not duplicate the freezingprocess in rock. Crystal growth during oven dry-ing creates internal expansion that can dilate theaggregate, but the internal stresses that developdue to freezing water are not solely related tocrystal (ice) growth. A major cause of stress dur-ing freezing is hydraulic pressure produced bywater expelled from the freeze front due toexpansion of the ice against the pore walls. More-over, it is well known that the temperature atwhich water freezes within any porous materialvaries with the size of the pore: the smaller the
Figure 26. Correlation between on-road British pendulum number and polishingby projection coefficient Cpp (Tremblay 1995).
80
70
60
50
40
30
BP
N
0.3 0.4 0.5 0.6 0.7
Cpp
17
to contents
pore, the lower the freezing temperature. Sinceaggregate includes a range of pore sizes, not allpores freeze at one temperature, whereas in thesulfate test, all pores, regardless of size, experi-ence expansive forces caused by crystal growth.Consequently, the mechanism of internal destruc-tion caused by the sulfate soundness test does notduplicate that of natural freezing. The literaturecontains examples of aggregate that failed thesulfate test but performed well in service and viceversa.
Unconfined freeze–thaw test (AASHTO T 103)Modern refrigeration equipment provides an
alternative to the sulfate soundness test for mea-suring aggregate durability. This equipment wasfirst used to test the aggregate in an unconfinedcondition; that is, not confined within concretemixtures as it is normally found. Traditionally,unconfined freeze–thaw tests have either evalu-ated the damage caused by a certain number offreeze–thaw cycles or they have determined thenumber of freeze–thaw cycles needed to cause acertain amount of damage. The basic procedure,AASHTO T 103 (1990), is to subject aggregate torepeated cycles of freezing and thawing. Varia-tions on this procedure have included vacuumsaturation of aggregate, as opposed to merely sub-merging the aggregate in water; alcohol–water, asopposed to water, as the wetting medium to aidpenetration of pores; soaking the aggregate in saltsolutions, as opposed to just water, to increasefrost damage; and varying the freezing and thaw-ing rates. Despite these and other variations onthe freeze–thaw test procedure, the resultsobtained by unconfined freezing and thawing ofaggregate have not always provided good corre-lation to service life.
The problem is that freezing unconfinedaggregate is not the same as freezing aggregateconfined in concrete. Though the freezing processis the same in either case, the interaction betweenthe aggregate and its immediate surroundingscreates the difference between the two testingconditions. As ice forms in a pore of an aggregate,the approximate 9% volume change that accom-panies this process squeezes unfrozen wateragainst the walls of the pore. This hydraulic pres-sure, in the case of unconfined aggregate, can behandled if the aggregate can elastically deform,or if the water can escape into a nearby emptyvoid or to the outside boundary of the aggregate.For confined aggregates, the outside pressurerelief depends on the ability of the concrete
matrix to absorb the escaping water into emptypores. In addition, even if the aggregate can dilatewithout cracking, pavement damage can only beavoided if the pavement matrix can also expandwithout cracking. Thus, aggregate that can elasti-cally accommodate water frozen within its porescould show no effect from unconfined testing butcould show considerable damage when tested inconcrete if the concrete matrix was relativelyimpermeable or if it was unable to accommodatethe dilation of the aggregate. Aggregate that wetsby capillarity may take on more water whenexposed to a free water surface than when con-fined in a pavement, particularly if the matrix hassmaller pores than those in the aggregate, inwhich case water will be preferentially absorbedby the matrix. In this case, the aggregate wouldlikely suffer more damage when tested uncon-fined than when tested confined.
Confined rapid freeze–thaw test(ASTM C 666 A and B)
A popular test procedure for evaluating thefrost durability of concrete and aggregate isASTM C 666 (1990) and its many variants aroundthe world. It consists of two methods: method A,which freezes and thaws concrete in water, andmethod B, which freezes concrete in air andthaws it in water. One freeze–thaw cycle lastsfrom 2 to 5 hr, and the test can last up to 300 ormore cycles. The correlation between this test andfield performance has not always been good. Themajor weakness with this test is a very rapid cool-ing rate compared with that in nature. Thehydraulic pressures that might be caused by a rela-tively slowly advancing freeze front in nature areprobably much lower than the unrealisticallyfast-moving freeze fronts caused by this test.Another objection to this test is that up to 5 monthscan pass before results are available. Though thistest has several shortcomings, it will continue tobe popular until a better method is available.
Hydraulic fracture testRecently, a new test was developed to simulate
the hydraulic forces generated by freezing waterinside aggregate without having to freeze theaggregate. In this test, aggregate submerged inwater is subjected to high pressures. Upon suddenrelease of the pressure, air compressed within thepores forces water to flow away from the pore inmuch the same way as freezing causes unfrozenwater to flow away from a freeze front. The factthat this creates very rapid pressure changes and
18
to contents
that the entire aggregate is pressurized, asopposed to only freezing sites, suggests that thistest could be quite different from natural freez-ing. Generally, aggregates that can withstandmore than 100 pressure cycles are quite durable,and those that produce fracture in 5% of theaggregate particles in less than 50 pressure cyclestend to exhibit poor frost resistance. Though thistest has shown reasonable correlation to field per-formance in limited testing, more experience isneeded to assess it.
Cryogenic testThe cryogenic test (Korhonen and Charest
1995) examines the efficacy of cycling aggregatebetween hot water and liquid nitrogen. Modifiedfrom AASHTO T 103 (1990), the test consists ofimmersing water-saturated aggregate in liquidnitrogen for about 1 minute followed by a2-minute immersion in hot water. Within 10freeze–thaw cycles, aggregates susceptible tofrost damage are readily identified. The test,however, is not capable of ranking aggregates ofmoderate to good performance.
Slow freeze test (ASTM C 671)It has been documented that cooling rates in
the field rarely exceed a few degrees per hour, butthe ASTM C 666 (confined rapid freeze–thaw) testsubjects concrete to much higher cooling rates.ASTM C 671 (1981), on the other hand, consists ofcooling concrete at a rate of –2.8°C per hour. Thespecimens are held in a 2°C bath for two weeksbefore being frozen. Though this test has showngood correlation with field performance, it is verytime consuming.
VPI single-cycle freeze testIn the VPI single-cycle freeze test, a concrete
beam is subjected to –18°C air while lengthchange and temperature are measured over a 4-hrperiod. The test is rapid compared with ASTM C671 (1981) and has been successful in identifyingvery durable and nondurable aggregate. Othertesting is needed to rank aggregates of moderatedurability.
Iowa Pore Index testThe Iowa Pore Index test acknowledges the
importance of pore size in aggregate durability. Itconsists of measuring the amount of water thatenters aggregate submerged in water when pres-surized to 241 kPa. The amount of waterabsorbed during the first minute of the test is con-
sidered to be a measure of the beneficial voids inthe aggregate. A second reading, taken after 15minutes, indicates the relative portion of voidsthat are frost susceptible. Large amounts of sec-ondary water indicate that the aggregate is sus-ceptible to frost damage. This test seems to givegood correlation to ASTM C 666 (1990) results,though there is not always a good correlationwith field performance.
Assuming that the first minute represents ben-eficial voids appears to be a shortcoming of thistest. In nature, the first pores to fill with water arethose that are small enough to exhibit capillarity.The larger “beneficial” voids, provided they arenot connected to the surface, will not fill withwater unless they are under hydrostatic pressure(or gravity drainage). Thus, a 24-hr soak followedby additional wetting under pressure should beinvestigated as a modification to this test.
In summary, the experience of transportationofficials, pavement engineers, researchers, andothers is that, whether aggregates are confinedwithin Portland cement concrete (PCC) or withinasphalt cement concrete (ACC) pavements, theyare not always immune to damage caused by coldweather. D-cracking, a common distress in PCCpavements, is the disintegration of coarse aggre-gates by frost action, and stripping, a commondistress in HMA pavements, is the disbonding ofcoarse aggregates from the asphalt matrix duringcold weather. Both deterioration mechanismslead to pavement damage whose root causeappears to be excessive volume changes causedby the freezing of saturated aggregate. When waterfreezes inside particles of rock, the tendency is forthe rock to expand, which can be great enough tocrack the pavement matrix or to rupture theaggregate or both.
Currently, the best way to measure volumeexpansion of aggregate is when it is confined with-in concrete. In this manner, stresses from hydraulicforces within discrete particles of aggregate andthose from water escaping from the aggregateinto the surrounding matrix are accounted for.However, this test is time-consuming—making,curing, and testing specimens requires consider-able effort—and does not always correlate to fieldperformance.
A simpler method is to measure the volumeexpansion of unconfined aggregate. Though thisis different from being confined in concrete, itwould identify those aggregates that expandunusually or that expel a large amount of water. Itwould also avoid the variabilities associated with
19
to contents
consistency in fabricating laboratory specimens,and it would measure an important responsemechanism of aggregate to freezing. Sneck et al.(1972) and Davison (1982) describe such a test.They placed rock particles into a glass containerfitted with a calibrated tube and filled with alco-hol, and they observed that the volume of alcoholabruptly changed each time the container wascooled below the freezing point of water. It doesnot appear that this test was pursued any further.Another variation is to consider freezing aggre-gate in alcohol while it is suspended from a scale.Any volume change should be noted as a changein weight.
SUMMARY AND RECOMMENDATIONS
Hot-mix asphalt aggregates play a significantrole in controlling pavement distresses such asrutting, skid resistance, and freeze–thaw deteriora-tion. This report provides a state-of-the-art sum-mary of existing and new, promising laboratorytest methods and devices for characterizing theaggregate.
Several laboratory tests for rutting were identi-fied, ranging from sophisticated image analysisto determination of flow time of aggregates. TheNAA uncompacted void value test appears to besimple and has been found to correlate well withaggregate shape, angularity, and texture. Theflow time index method used by the Quebec Min-istry of Transportation also appears to show greatpromise as a simple tool for evaluating the sameproperties.
The macro- and microtexture of the pavementsurface appears to control the skid resistance ofthe pavement surface. The sand patch test is com-monly used to characterize the macrotexture ofthe pavement surface.
The frictional properties of microstructure aredetermined using frictional testers. The most com-mon one is the British skid resistance tester. Thewear and polishing characteristics of the aggre-gates can be determined by several wheel polish-ing devices. Quebec MOT uses high-pressurewater projection to conduct accelerated polish-ing, and friction results from the test appear tocorrespond well to those obtained from the Brit-ish pendulum number.
For freeze–thaw durability, aggregate expan-sion was identified as the major contributor topavement frost damage. A variation to the expan-sion measurement of unconfined aggregatedescribed by Sneck et al. (1972) and Davison
(1982), appears to be a quick and simple approachto indexing aggregates based on propensity toexpand during freezing.
LITERATURE CITED
AASHTO T 85-85 (1998) Specific Gravity andAbsorption of Coarse Aggregate. Washington D.C.:American Association of State Highway Trans-portation Officials.AASHTO T 96-94 (1998) Standard Test Method forResistance to Degradation of Small Size Coarse Aggre-gate by Abrasion and Impact in the Los AngelesMachine.Washington, D.C.: American Associationof Highway Transportation Officials.AASHTO T 103-83 (1990) Standard Method of Testfor Soundness of Aggregates by Freezing and Thaw-ing. Washington D.C.: American Association ofState Highway Transportation Officials.AASHTO T 104-86 (1990) Standard Method of Testfor Soundness of Aggregate by Use of Sodium Sulfateor Magnesium Sulfate. Washington D.C.: AmericanAssociation of State Highway TransportationOfficials.Aldrich, A.C. (1996) Influence of aggregates gra-dation and particle shape/texture on pavementdeformation of hot-mix asphalt pavements. U.S.Army Corps of Engineers, Waterways ExperimentStation, Technical Report GL-96-1.ASTM C 88-76 (1981) Standard Test Method forSoundness of Aggregates by Use of Sodium Sulfate orMagnesium Sulfate. Annual Book of ASTM Stan-dards, Part 14. Philadelphia: American Society forTesting and Materials.ASTM C 127-88 (1995) Standard Test Method forSpecific Gravity and Absorption of Coarse Aggregate.Annual Book of ASTM Standards, Part 14. Phila-delphia: American Society for Testing and Mate-rials.ASTM C 666-84 (1990) Standard Test Method forResistance of Concrete to Rapid Freezing and Thawing.Annual Book of ASTM Standards, Part 14. Philadel-phia: American Society for Testing and Materials.ASTM C 671-77 (1981) Standard Test Method forCritical Dilation of Concrete Specimens Subjected toFreezing. Annual Book of ASTM Standards, Part14. Philadelphia: American Society for Testingand Materials.ASTM C 1252-93 (1995) Standard Test Methods forUncompacted Void Content of Fine Aggregate (as In-fluenced by Particle Shape, Surface Texture, and Grad-ing). Annual Book of ASTM Standards, Part 14.Philadelphia: American Society for Testing andMaterials.
20
to contents
ASTM D 3042-92 (1995) Standard test method forinsoluble residue in carbonate aggregates. Roadand Paving Materials; Pavement Management Tech-nologies, 04(03).ASTM E 274-90 (1994) Standard test method forskid resistance of paved surfaces using a full-scale tire. Road and Paving Materials; PavementManagement Technologies, 04(03).ASTM E 303-93 (1995) Standard test method formeasuring surface frictional properties using theBritish Pendulum Tester. Road and Paving Materi-als; Pavement Management Technologies, 04(03).ASTM E 660-90 (1995) Standard practice for accel-erated polishing of aggregates or pavement sur-faces using a small-wheel, circular track polish-ing machine. Road and Paving Materials; PavementManagement Technologies, 04(03).ASTM E 707-90 (1995) Standard test method forskid resistance measurements using the NorthCarolina State University variable-speed frictiontester. Road and Paving Materials; Pavement Man-agement Technologies, 04(03).ASTM E 965-87 (1995) Standard test method formeasuring the macrotexture depth using a volu-metric technique. Road and Paving Materials; Pave-ment Management Technologies, 04(03).ASTM E 1551-93a (1994) Standard specificationfor special purpose, smooth-tread tire, operatedon fixed braking slip continuous friction measur-ing equipment. Road and Paving Materials; Pave-ment Management Technologies, 04(03).Barksdale, R.D., and S.Y. Itani (1994) Influenceof aggregate shape on base behavior. Transporta-tion Research Record, 1227: 171–182.Beaton, J.L. (1976) Providing skid resistant pave-ments. Transportation Research Record, 622.British Standards Institution (1989) Testing aggre-gates. BS 812.Davison, J.I. (1982) Volume change due to freez-ing in plastic masonry mortar. In Masonry: Materi-als, Properties, and Performance (J.G. Borchelt, Ed.).American Society for Testing and Materials,ASTM STP 778, p. 27–37.Donbavand, J. (1989) Skidding resistance of roadsurfaces—Implications for New Zealand. RoadResearch Unit, Transit New Zealand, Wellington,New Zealand, Road Research Unit Bulletin 81.Horne, W.B. (1974) Elements affecting runwaytraction. SAE paper 740496, p. 29.Hosking, J.R., and G.C. Woodford (1976) Meas-urement of skidding resistance, Part II. Factorsaffecting the slipperiness of a road surface. TRRLReport 738, 1976.Huang, E.Y. (1962) A test for evaluating the geo-
metric characteristics of coarse aggregate parti-cles. Proceedings, American Society of Testing andMaterials, 62: 1223–1242.Ishai, I., and E. Tons (1971) Aggregate factors inbituminous mixture designs. University of Michi-gan, Ann Arbor, Report 335140-1-F.Ishai, I., and E. Tons (1977) Concept and testmethod for a unified characterization of the geo-metric irregularity of aggregate particles. Journalof Testing and Evaluation, 5(1): 3–15.Janoo, V.C. (1990) Use of soft grade asphalts inairfields and highway pavements in cold regions.USA Cold Regions Research and EngineeringLaboratory, Special Report 90-12.Janoo, V.C. (1998) Quantification of shape, angu-larity, and surface texture of base course mater-ials. USA Cold Regions Research and Engineer-ing Laboratory, Special Report 98-1.Kandhal, P.S., and F. Parker, Jr. (1998) Aggregatetests related to asphalt concrete performance inpavements. National Cooperative HighwayResearch Program, Transportation ResearchBoard, NCHRP Report 405,Korhonen, C., and B. Charest (1995) Assessingcryogenic testing of aggregates for concrete pave-ments. USA Cold Regions Research and Engi-neering Laboratory, Special Report 95-4.Krumbein, W.C. (1941) Measurement and geo-logical significance of shape and roundness ofsedimentary particles. Journal of Sedimentary Pet-rology, 11(2): 64–72.Lees, G. (1964) The measurement of particleshape and its influence in engineering materials.Journal of the British Granite and Whinestone Federa-tion, London, 4(2): 1–22, 1964.Marek, C.R. (1972) Review, selection and calibra-tion of accelerated wear and skid resistance test-ing equipment. Illinois Cooperative HighwayResearch Program, University of Illinois at Urbana-Champaign, Interim report – Phase 1, ProjectIHR-406.Moore, D.F. (1966) Prediction of skid-resistancegradient and drainage characteristics for pave-ments. Highway Research Record, 131: 181–203.Michigan DOT (1983) Evaluation of the straightline gradation chart and the particle index test.Research Report No. R-1210.Rose, J.G., J.W. Hutchinson, and B.M. Gallaway(1972) Summary and analysis of the attributes ofmethods of surface texture measurements. In Pro-ceedings, Symposium on Skid Resistance of HighwayPavements, ASTM.Rosenthal P., F.R. Haselton, K.D. Bird, and P.J.Joseph (1969) Evaluation of studded tires, perfor-
21
to contents
mance data and pavement wear measurements.National Cooperative Highway Research Pro-gram, Highway Research Board, Report 61,Shergold, F.A. (1953) The percentage of voids incompacted gravel as a measure of its angularity.Magazine of Concrete Research, 5(13): 3–10.Sneck, T., et al. (1972) Winter masonry, buildingtechnology and community development. Tech-nical Research Centre of Finland, Publication 1,Helsinki.Tons, E., and W.H. Goetz (1968) Packing volumeconcepts for aggregates. Highway Research Record,236: 76–96.Tremblay, G., S. Julien, A. Leclerc, and B. Auger
22
(1995) The role of aggregates in road surfacingtexture and skid resistance. In Proceedings, Trans-portation Association of Canada, Victoria, BritishColumbia.Uge, P., and P. Van de Loo (1974) Permanentdeformation of asphalt mixes. Canadian TechnicalAsphalt Association, 29: 307–342.Wilson, J.D., L.D. Klotz, and C. Nagaraj (1995)Automated measurement of aggregate indices ofshape. Federal Highway Administration ReportFHWA RD-95-116.Wright, P.J.F. (1955) A method of measuring thesurface texture of aggregate. Magazine of ConcreteResearch, 7(21): 151–160.
to contents
1. REPORT DATE (DD-MM-YY) 2. REPORT TYPE 3. DATES COVERED (From - To)
4. TITLE AND SUBTITLE 5a. CONTRACT NUMBER
5b. GRANT NUMBER
5c. PROGRAM ELEMENT NUMBER
6. AUTHOR(S) 5d. PROJECT NUMBER
5e. TASK NUMBER
5f. WORK UNIT NUMBER
7. PERFORMING ORGANIZATION NAME(S) AND ADDRESS(ES) 8. PERFORMING ORGANIZATION REPORT NUMBER
9. SPONSORING/MONITORING AGENCY NAME(S) AND ADDRESS(ES) 10. SPONSOR / MONITOR’S ACRONYM(S)
11. SPONSOR / MONITOR’S REPORT NUMBER(S)
12. DISTRIBUTION / AVAILABILITY STATEMENT
13. SUPPLEMENTARY NOTES
14. ABSTRACT
15. SUBJECT TERMS
16. SECURITY CLASSIFICATION OF: 17. LIMITATION OF 18. NUMBER 19a. NAME OF RESPONSIBLE PERSONOF ABSTRACT OF PAGES
a. REPORT b. ABSTRACT c. THIS PAGE 19b. TELEPHONE NUMBER (include area code)
Standard Form 298 (Rev. 8-98)Prescribed by ANSI Std. 239.18
REPORT DOCUMENTATION PAGEPublic reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining thedata needed, and completing and reviewing this collection of information. Send comments regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducingthis burden to Department of Defense, Washington Headquarters Services, Directorate for Information Operations and Reports (0704-0188), 1215 Jefferson Davis Highway, Suite 1204, Arlington, VA 22202-4302.Respondents should be aware that notwithstanding any other provision of law, no person shall be subject to any penalty for failing to comply with a collection of information if it does not display a currently validOMB control number. PLEASE DO NOT RETURN YOUR FORM TO THE ABOVE ADDRESS.
Form ApprovedOMB No. 0704-0188
Approved for public release; distribution is unlimited.
Available from NTIS, Springfield, Virginia 22161.
December 1999
Performance Testing of Hot-Mix Asphalt Aggregates
Vincent C. Janoo and Charles Korhonen
U.S. Army Cold Regions Research and Engineering Laboratory72 Lyme Road Special Report 99-20Hanover, New Hampshire 03755-1290
New Hampshire Department of Transportation
Hot-mix asphalt (HMA) pavements are subject to thermal cracking, fatigue cracking, rutting, stripping, raveling, andfreeze–thaw damage. Some of these distresses are directly affected by the choice of aggregates. Particle shape, surfacetexture, particle size, pore structure, and particle strength are the most common characteristics cited for controlling ruttingand for maintaining adequate skid resistance. A literature review was conducted to evaluate commonly used and poten-tial test methods for evaluating hot-mix aggregates in term of pavement performance.
Aggregates Hot-mix asphalt RuttingFreeze–thaw Laboratory testing Skid resistance
UL UL UL 27
