-
Evaluation of high-early strength PCC mixtures usedin full depth
repairs
Neeraj Buch a,*, Thomas J. Van Dam b, Karl Peterson b, Larry
Sutter c
a Department of Civil and Environmental Engineering, 3556
Engineering Building, Michigan State University, E. Lansing, MI
48824, USAb Department of Civil and Environmental Engineering,
Michigan Technological University, 1400 Townsend Drive, Houghton,
MI 49931, USA
c School of Technology, Michigan Technological University, 1400
Townsend Drive, Houghton, MI 49931, USA
Received 15 March 2006; received in revised form 31 July 2006;
accepted 25 September 2006Available online 21 November 2006
Abstract
The research described herein was conducted as a part of NCHRP
Project 18-04B, Early Opening-to-Trac Portland Cement Concretefor
Pavement Rehabilitation [Van Dam TJ. Durability of
Early-Opening-to-Trac portland cement concrete for pavement
rehabilita-tion. Final Report, NCHRP Web Document 76,
Transportation Research Board, National Research Council,
Washington, DC; 2005.].In the study fourteen dierent high-early
strength portland cement concrete (PCC) full-depth repair mixtures
(two replicates or batcheswere prepared for each mixture for a
total of 28 batches) were designed for a 68 h opening to trac
window. The fresh and hardenedconcrete properties, freeze-thaw
durability and microstructural properties of the concrete were
assessed as part of the laboratory testingprogram. As a result of
this assessment it was determined that high-early strength PCC
mixtures of adequate strength could be produced,but that
interactions between the various constituents could result in
durability problems in some mixtures, particularly those made
withType III cement and high-range water reducers. It is
recommended that durability-related testing be conducted on such
mixtures toensure longevity of the repair.! 2006 Elsevier Ltd. All
rights reserved.
Keywords: Microstructure; Freeze-thaw durability; Fast track;
High-early strength
1. Introduction
The research described herein was conducted as a partof NCHRP
Project 18-04B, Early Opening-to-Trac Port-land Cement Concrete for
Pavement Rehabilitation [1]. Withincreasing trac volume in urban
areas, delays duringpavement rehabilitation are becoming less
tolerated. Tominimize delays, state highway agencies (SHAs)
useearly-opening-to-trac (EOT) rehabilitation strategiesthat allow
work to be completed at night or during periodsof low trac.
Generally, portland cement concrete (PCC)used in these applications
is expected to gain sucientstrength to carry trac within 624 h
after placement.But concerns have been raised that such high-early
strengthmaterials lack durability, resulting in the need for
addi-
tional lane closure to replace failing repairs.
Rigorousrequirements for mix design and strength developmenthave
usually been stipulated for EOT concrete applica-tions, often with
limited consideration given to materialsand construction aspects
that influence long-term perfor-mance and durability.
The main dierence between high-early strength andnormal repair
concrete mixtures is that strength gain mustoccur much more rapidly
in the former, thus more cement,less water, a cocktail of
admixtures, and aids to retain heatare commonly used. Most SHAs
allow the use of eitherType I or Type III portland cement in their
high-earlystrength PCC mixtures. When specified, the minimumcement
contents varies from state to state, ranging from440 to 550 kg/m3
(740925 lb/yd3) for Type I cement and390490 kg/m3 (660825 lb/yd3)
for Type III cement. Inaddition, the use of accelerators is almost
universally spec-ified for mixtures containing Type I cement, with
the most
0950-0618/$ - see front matter ! 2006 Elsevier Ltd. All rights
reserved.doi:10.1016/j.conbuildmat.2006.09.006
* Corresponding author. Tel.: +1 517 432 0012; fax: +1 517 432
1827.E-mail address: [email protected] (N. Buch).
www.elsevier.com/locate/conbuildmat
Available online at www.sciencedirect.com
Construction and Building Materials 22 (2008) 162174
Constructionand Building
MATERIALS
Jonathas Iohanathan Felipe de Oliveira
-
common accelerator being calcium chloride. Some mix-tures
specified with Type III cement do not require theuse of an
accelerator, relying solely on the early strengthgain
characteristics of the cement. The water-to-cement(w/c) ratio for
the high-early strength PCC mixtures varieswidely, with maximum
allowable values ranging from 0.33to 0.49.
Based on a review of several SHAs construction speci-fications
it was found that the acceptability of a high-earlystrength PCC
mixture is commonly linked exclusively tostrength requirements. For
example, the time to openingcriteria used by SHAs varied from as
early as 4 h (Kansasand Ohio) to as late as 12 h (Maryland and
Minnesota). Insome cases, only a time to opening criteria was
specified,but in others, strength was used as the sole criteria
foropening. A few SHAs used both a strength and time toopening
criteria. Florida for example specifies that thecompressive
strength must exceed 21 MPa (3000 psi) in24 h while allowing a 6 h
time to opening. In the case ofNew York, the repair was opened to
trac once the surfacetemperature of the repair reached 65 "C (150
"F).
Research conducted under the Strategic HighwayResearch Project
(SHRP) found that an opening criteriausing a third-point modulus of
rupture of 2.1 MPa(300 psi) or a compressive strength of 13.8
MPa(2000 psi) was reasonable [2]. Recommendations made bythe
American Concrete Pavement Association (ACPA)are not based on
strength, but instead on opening time[3] and in the NHIs PCC
Pavement Evaluation and Reha-bilitation guidelines, both minimum
strength requirementsand time to opening requirements are
recognized as beingacceptable [4]. The flexural strength criteria
dier slightlyfrom that presented by SHRP, with a requirement of1.7
MPa (250 psi) for third-point loading and 2.1 MPa(300 psi) for
center-point loading.
In this paper the authors report on an extensive labora-tory
study in which 14 dierent high-early strength PCCmixtures (two
replicates or batches were made for eachmixture for a total of 28
batches) designed for 68 h open-
ing were produced and tested. The testing included assess-ment
of the properties of the fresh and hardened concrete,freeze-thaw
durability, and the microstructural character-ization of the
concrete.
2. Experimental program
2.1. Experimental matrix
To evaluate the properties of the high-early strengthPCC
mixtures a factorial experimental design approachwas adopted,
allowing for two or more levels to be setfor each based on
knowledge of concrete technology as itpertains to high-early
strength PCC mixtures, concentrat-ing on variables of greatest
interest. The material combina-tions are presented in Table 1,
where an alpha-numericdesignation (A-1 through N-14) was assigned
to eachmixture.
The seven independent variables considered in the studyhave
either 2 or 3 levels. In total, 288 material combina-tions are
possible, of which 14 material/curing combina-tions were evaluated.
Two batches were made for eachmixture. Thus, a total of 28 batches
of PCC mixtures weremade and evaluated. The ranges for the seven
independentvariables were carefully selected after considering the
infor-mation contained in the NCHRP 18-04A Final Report
[5],numerous published articles, specifications used by variousSHAs
in the construction of high-early strength PCC repairmaterials. The
first six variables (cement type, cement fac-tor, w/c ratio, coarse
aggregate type, accelerator type, andwater reducer type) are
justified for inclusion in this studyas each is a common variable
in the mix design/proportion-ing process, as is evidenced by their
inclusion in most SHAspecifications. Further, each is suspected to
have a directeect on the microstructure of portland cement
concreteand must be considered when studying the durability ofEOT
PCC materials.
The justification for the inclusion of the final
variable,curing, was thought to potentially have a significant
impact
Table 1Summary of high-early strength PCC mixture designs used
in laboratory study
Combination 1 2 3 4 5 6 7
Cement type Cement factor w/c ratio Coarse aggregate Accelerator
Water reducer Curing
A-1 Type I 425 kg/m3 (716 lb/yd3) 0.4 Carbonate Non-chloride No
23 "C (73 "F)B-2 Type I 525 kg/m3 (885 lb/yd3) 0.4 Carbonate
Non-chloride No 23 "C (73 "F)C-3 Type I 525 kg/m3 (885 lb/yd3) 0.4
Carbonate Non-chloride No Heated blanketD-4 Type I 525 kg/m3 (885
lb/yd3) 0.4 Siliceous Non-chloride No Heated blanketE-5 Type I 425
kg/m3 (716 lb/yd3) 0.4 Siliceous Non-chloride No 23 "C (73 "F)F-6
Type I 525 kg/m3 (885 lb/yd3) 0.4 Carbonate Non-chloride No 23 "C
(73 "F)G-7 Type I 525 kg/m3 (885 lb/yd3) 0.4 Carbonate Non-chloride
Type F 23 "C (73 "F)H-8 Type I 525 kg/m3 (885 lb/yd3) 0.4 Carbonate
No Type E 23 "C (73 "F)I-9 Type I 425 kg/m3 (716 lb/yd3) 0.4
Carbonate Calcium chloride No 23 "C (73 "F)J-10 Type I 525 kg/m3
(885 lb/yd3) 0.4 Carbonate Calcium chloride No 23 "C (73 "F)K-11
Type III 425 kg/m3 (716 lb/yd3) 0.4 Carbonate Non-chloride Type F
23 "C (73 "F)L-12 Type III 525 kg/m3 (885 lb/yd3) 0.4 Carbonate
Non-chloride Type F 23 "C (73 "F)M-13 Type III 525 kg/m3 (885
lb/yd3) 0.4 Carbonate Non-chloride Type F Heated blanketN-14 Type
III 425 kg/m3 (716 lb/yd3) 0.4 Carbonate Calcium chloride Type F 23
"C (73 "F)
N. Buch et al. / Construction and Building Materials 22 (2008)
162174 163
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
-
on the durability of EOT PCC repairs. It has been
welldemonstrated in field and laboratory studies that manyof these
PCC mixtures obtain relatively high temperatures(in excess of 65 "C
[150 "F]) in the first few hours afterplacement due to the high
heat of hydration that occurs.This could lead to thermally induced
stress and/or possiblealteration of the primary hydration products,
negativelyaecting durability.
2.2. Specimen preparation
Laboratory specimens were prepared in accordance withAASHTO T
126 at the Michigan Department of Transpor-tations (MDOT)
Construction and Technology Labora-tory in Lansing, Michigan. The
fresh concrete was mixedin a 0.4 m3 (14 ft3) capacity Lancaster
mixer, ensuring thatall specimens needed for a given mixture were
produced ina single batch, thereby reducing experimental
variability.As noted, two batches were made of each material
combi-nation. Furthermore, the order in which batches were
pre-pared was randomized to avoid systematic error. Table 2lists
the number of specimens that were fabricated fromeach batch and how
they were tested, including cylinders,beams, prisms, and rings.
All coarse aggregates were sieved into standard size frac-tions
and then combined to the desired grading meetingASTM C33
specifications. The aggregate volume fractionhad to change for the
various mixtures as cement factorand w/c ratio varied, but the
proportion of coarse aggre-gate to fine aggregate was held
constant. The aggregateswere brought to SSD condition prior to
mixing to ensureaccuracy in the resulting w/c ratio. Molds were
strippedfrom the specimens 6 h after casting.
2.3. Tests conducted on laboratory prepared specimens
Once the specimens were molded, various tests wereconducted to
assess the characteristics of interest. The sametesting procedures
were followed for all mixtures. Asshown in Table 2, the material
characterization tests wereclassified as follows:
! Properties of fresh concrete.! Freeze-thaw durability.!
Microstructural characterization.
2.4. Properties of fresh concrete
Testing of fresh concrete was conducted in accordancewith the
relevant test methods, and the results used to ver-ify workability,
air content, and establish maturity trends.This testing was
necessary to ensure that the mixtures pro-duced and tested could be
constructed, and thus have prac-tical application. The properties
of fresh concrete measuredfor each mixture include the slump
(AASHTO T 119), aircontent (AASHTO T 152), unit weight (AASHTO
T121), and maturity (ASTM C 1074). The measured air con-tents were
also compared to the air-void system parametersobtained from the
hardened concrete as determined byASTM C 457, as it is known that a
specified air contentof fresh concrete does not necessarily equate
to adequateair-void system parameters to protect against
damageincurred due to freezing and thawing. The establishedmaturity
trends are also valuable in that they provide abasis for early
opening criteria, as well as an understandingof the heat of
hydration characteristics for the variousmixtures.
2.5. Freeze-thaw durability
Two freeze-thaw durability tests were conducted onspecimens that
were cured for 28-days. AASHTO T 161Procedure A was used to assess
the durability of concretesubjected to cyclic freezing and thawing.
Specimens weretested to 300 cycles in strict accordance to the
temperaturecycling regime specified in the test method and the
dilationwas measured.
Of almost equal value is that the microstructure of thetest
specimens was characterized after testing. The degreeof
microcracking in the AASHTO T 161 specimens wasassessed. This
permitted evaluation of whether significantalterations attributable
to the environmental conditioning
Table 2Test details associated with the laboratory evaluation
including number of specimens required for each batch
Test attribute Test name/Equipment Measured property No. of
specimens Specification
Properties of freshconcrete
Slump Concrete workability One per batch AASHTO T 119-93Air
content Total air content of fresh concrete One per batch AASHTO T
152-93Maturity and time of setting Time of curing and temperature
of
strength gain/time and depth2 ASTM C 1074/
AASHTO T131-93Freeze-thaw
durabilityResistance of concrete to freezingand thawing
Loss of concrete stiness due to freeze-thaw damage
2 AASHTO T 161-93
Microstructuralcharacterization
Compressive strength Concrete strength 8 AASHTO T 22-92Flexural
strength Concrete flexural strength 4 AASHTO T 97-86Stereo optical
microscopy Air-void system parameters 2 ASTM C 457 and C
856Petrographic optical microscopywith UV dye impregnation
Microstructure analysis, reactionproducts, and evidence of
deterioration
Same specimens asstereo microscopy
ASTM C 856
Scanning electron microscopy(SEM)
Microstructure analysis, elementalanalysis
Same specimens asstereo microscopy
ASTM C 856
164 N. Buch et al. / Construction and Building Materials 22
(2008) 162174
-
occurred. This is important since identification of the fail-ure
mechanism is necessary to address any deficienciesand to verify
that the distress observed was indeed due tofreezing and
thawing.
2.6. Microstructural characterization
Five separate test methods were conducted under
themicrostructural characterization test attribute. The firsttwo
tests are compressive strength (AASHTO T 22) andflexural strength
(AASHTO T 97). Compressive strengthtesting was conducted for the
mixtures at 6 h, 8 h, 24 h,and 28 days. Flexural strength testing
was conducted at6 h and 8 h.
Microscopic evaluation was conducted to gain anunderstanding of
how the microstructure varied from mix-ture to mixture and to
determine how the microstructuralfeatures impacted material
performance. In this way,improved knowledge on how to produce
high-earlystrength PCC mixtures having a durable microstructurewas
obtained. Further, the results of the
microstructuralcharacterization were correlated, to the degree
possible,with the freeze-thaw durability testing in an attempt to
linkthe results of simple laboratory tests to the more
advancedtesting.
Specifically, ASTM C 457 was used to determine the air-void
system parameters from polished slabs observed usingthe stereo
optical microscope. The air-void system param-eters include the air
content, spacing factor, specific sur-face, and paste to air ratio.
Petrographic evaluation inaccordance with ASTM C 856 was also
conducted on thinsections. Due to the nature of petrographic
analysis, a sub-jective rating system was developed to rate the
degree ofpaste inhomogeniety (on a scale of 1 to 3, with 1
beinghomogenous and 3 being highly inhomogenous) and degreeof
microcracking (on a scale of 1 to 3, with 1 being free
ofmicrocracks and 3 indicating a high degree of microcrack-ing). As
described, beams that had been subjected toAASHTO T 161 were
sectioned and the degree of micro-cracking measured (reported as
l/mm2) using a high-reso-lution flatbed scanner.
3. Laboratory test results and discussion
3.1. Properties of fresh concrete
Table 3 summarizes the descriptive statistics associatedwith
fresh concrete tests. The target slump for the PCCmixtures was 150
mm. With the exception of mixture G-7, which has slump values in
excess of 250 mm (10 in.),all other mixture resulted in acceptable
workability.Mixture G-7 was a high cement content (525 kg/m3),
highw/c ratio (0.40), calcium chloride accelerated mixture witha
Type F HRWR, making it dicult to create the desiredmixture with a
lower slump. The fresh concrete unit weightranged from 2.15 to 2.30
Mg/m3. The target air content forvarious mixtures ranged from 4.5%
to 7.5%.
The distribution of the laboratory measured air contentof fresh
concrete for the various mixtures is shown inFig. 1. For the most
part, the measured air contents werewithin the desired range with a
few notable exceptions.One mixture that was a bit problematic,
having air contentslightly higher than desired was E-5, which was a
lowcement content (425 kg/m3), high w/c ratio (0.40) mixturemade
with a non-chloride accelerator. Of greater concernto durability
was that three of the eight mixtures made withType III cement had
low measured fresh concrete air con-tent. All mixtures containing
Type III cement were madeusing a Type F HRWR since the fineness of
the cementmade it impossible to adequately mix these mixtures
with-out it. The high and low values for air observed in the
mix-tures contributed to overall higher variability as reflected
inthe coecient of variability for this parameter (Table 3).
The maturity results followed expected trends. However,it should
be noted that some of the mixtures (C-3, D-4, andM-13) were cured
at higher ambient temperatures (65 "C),contributing to a noticeable
increase in maturity at a givenage. Within a category, there are
little noticeable dierencesdue to dierent cement content, w/c
ratio, accelerator type,or type of cement.
3.2. Freeze-thaw durability
The resistance of concrete to damage due to freezing andthawing
(AASHTO T161) test results indicate that the dila-tion values
varied greatly from mixture to mixture. In gen-eral, approximately
20% of the specimens exceeded thethreshold of 0.02 mm/mm dilation.
Most of these speci-mens were made with Type III cement, which
would ini-tially suggest that the cement type might play a role.
Butsince all of the Type III cement mixtures contained theType F
HRWR as well, this conclusion cannot be drawndirectly. Furthermore,
mixture N-14 made with Type IIIcement did not show exceptionally
high dilations. This par-ticular mixture contained the calcium
chloride acceleratorinstead of the non-chloride accelerator. The
distributionof average dilation for the fourteen mixture designs is
illus-trated in Fig. 2. The negative dilations indicate that someof
the specimens actually experienced small levels ofshrinkage. This
is normal, indicating that no damage hasoccurred and that the
durability factor in these cases wouldbe 100.
The argument that the Type F HRWR may be contrib-uting to high
dilation values is supported by the results
Table 3Descriptive statistics of the fresh concrete
properties
Mixture property Average Standarddeviation
Coecient ofvariation (%)
Slump (mm) 115.20 66.12 57.40Air content (%) 5.66 1.02
18.10Maturity 8 h 385.8 68.22 17.70
24 h 1160.4 262.54 22.6Unit weight
(Mg/m3)2.186 0.05 2.27
N. Buch et al. / Construction and Building Materials 22 (2008)
162174 165
-
from mixture G-7-A, which is the only mixture made withType I
cement and the Type F HRWR and the non-chlo-ride accelerator. The
dosage of the Type F admixturewas higher for G-7-A than G-7-B
(G-7-B was made at alater date, and the admixture dosage was
reduced to tryto correct the high slump), whereas the dosage of air
entrai-ner was increased (both G-7-A and G-7-B had measuredair
contents of 5.2%). Similar results were observed forone of the Type
III mixtures where the first replicate (K-11-A) was made with a
higher HRWR dosage and lowerair entrainer dosage than the second
replicate (K-11-B).This suggests that an interaction between the
variousadmixtures may have negatively aected the air-void sys-tem
in these high cement content, low w/c ratio mixtures.
3.3. Microstructural characterization
The microstructural characterization of the mixturesincluded
compressive and flexural strength testing, stereo
optical microscopy, petrographic microscopy, electronmicroscopy,
and scanning with a flatbed scanner to assesscrack length in
specimens that were subjected to freeze-thaw testing.
Fig. 3a and b summarizes the strength variability withina batch
of concrete mixture design. The descriptive statis-tics (average
and standard deviation) were computed basedon four replicates. A
summary of the strength data fromthe various PCC mixtures is
presented in Figs. 4a throughc. The initial design criteria
included a minimum compres-sive strength of 13.8 MPa (2000 psi) and
a flexural strengthof 2.1 MPa (300 psi) within the 68 h opening
time criteria.Of the 28 batches 19 did not achieve the desired 6-h
com-pressive strength, with three of the mixtures that achievedthe
desired early strength undergoing high temperaturecuring. This
demonstrates the importance of increased tem-perature in achieving
early strength. Interestingly, the fourbatches cured at laboratory
ambient temperatures thatmade the minimum compressive strength
criteria were from
-0.02
-0.01
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
A-1
-A
B-2
-A
C-3
-A
D-4
-A
E-5-
A
F-6-
A
G-7
-A
H-8
-A
I-9-A
J-10
-A
K-1
1-A
L-12
-A
M-1
3-A
N-1
4-A
Mixture ID
Aver
age
dila
tion
, m
m/m
m
Replicate 1 Replicate 2 Dilation threshold
Fig. 2. Average dilation for the high-early strength
mixtures.
0
1
2
3
4
5
6
7
8
9
A-1 B-2 C-3 D-4 E-5 F-6 G-7 H-8 I-9 J-10 K-11 L-12 M-13
N-14Mixture ID
Avg.
% a
ir co
nten
t
Fig. 1. Air content for the high-early strength mixtures (dashed
lines indicated desired target range).
166 N. Buch et al. / Construction and Building Materials 22
(2008) 162174
-
two mixtures made with Type I cement, one of which waslow cement
content, high w/c ratio and the other highcement content, low w/c
ratio. Within 8 h, only five batchesfrom three mixtures (all made
with Type I cement) did notmeet the minimum compressive strength
requirement. Ofthe mixtures that failed to meet the minimum
compressivestrength criteria within 8 h, one (G-7) used a Type
Eadmixture whereas the other (H-8) used a Type F admix-ture. The
fifth batch (J-10-B) had unexpectedly lowstrengths.
The flexural strength results were slightly better, with 12of
the 28 batches passing the 6-h criteria of 2.1 MPa(300 psi). These
were the same mixtures that passed thecompressive strength
criteria, with the addition of bothbatches from mixture K-11 and
one from L-12. These lattertwo mixtures were made with Type III
cement. Only fourbatches did not meet the flexural strength
criteria at 8 h,three of which were from mixtures G-7 and H-8, and
onewas J-10-B, which is the same batch that had low compres-sive
strengths.
These results indicate that meeting high early strengthcriteria
is not assured at 6 h, although the heat generatedduring hydration
assists in this process. Further, admix-tures can have a
significant eect on early strength gain,with the type of water
reducer and accelerator playing arole. Ultimately, only a few
mixtures had diculty inachieving the desired strength within 8 h,
and these wereprimarily mixtures made with Type I cement and a
waterreducer.
Fig. 4c shows the relationship between early age com-pressive
and flexural strengths for all mixtures. It is recom-mended that if
the concrete compressive strength is to beused during construction
to estimate flexural strength, therelationship should be
established for the actual job mix.
Stereo optical microscopy was used in accordance withASTM C 457
to collect air-void system parameters thatrelate to the ability of
the concrete to resist damage fromcyclic freezing and thawing in a
saturated state. The desiredair content was 6 1.5%. Based on ASTM C
457, theresults of which are included in Table 4, the average
Fig. 3. Concrete mixture strength variability (sample size = 4)
(a) Concrete mixture variability Compressive strength (b) Concrete
mixture variability Flexural strength.
N. Buch et al. / Construction and Building Materials 22 (2008)
162174 167
-
measured air content of the hardened concrete was 5.64%.In Fig.
5, the dashed horizontal lines indicate the desiredrange for air
content. In general, most mixtures had aircontents that fell within
the desired range. The most obvi-ous deviation was in three of the
four mixtures made withType III cement (L-12, M-13, and N-14),
which had aircontents significantly below the desired value and
mixturesD-4, E-5 and K-11 exhibit air contents greater than
7.5%.However, the lower air voids are typically more criticalto the
concrete freeze-thaw durability, whereas higher aircontent can
impact strength.
In general, the total air content of a mixture is notthought to
be nearly as important in protecting the pasteagainst freeze-thaw
damage as is the size and spacing ofthe entrained air bubbles. The
most common parameterused in North America to assess the spacing of
air bubblesis the spacing factor, which is the average distance
from apoint in the paste to the surface of the nearest air
bubble.As shown in Table 4, the average spacing factor is0.1591 mm
(0.006263 in.). The average value is below theASTM C 457 criterion
set for protecting the paste againstdamage (0.200 mm [0.008 in.]).
Further, it can be seen inFig. 6 that all of the spacing factors
exceeding the limitingcriterion (as indicated by the dashed
horizontal line) are inmixtures made with Type III cement and the
Type FHRWR, even though one batch made with Type III cementand Type
F HRWR had acceptable spacing factors (K-11-B). These findings
further illustrate that the combination ofconstituents, the fine
cement, and the admixtures used
made it dicult to create the desired air-void system.
Inter-estingly, most of these mixtures had acceptable fresh
aircontents (although not all), illustrating the need to
betterassess the air-void system in fresh concrete for such
mix-tures. Fig. 7a illustrates the relationship between air
bubblespacing factor and dilation and Fig. 7b illustrates the
rela-tionship between air bubble spacing and the correlatedASTM C
666 durability factor (DF) based on the followingrelationship
developed and used by the Michigan Depart-ment of
Transportation:
DF e0:01630:2772%dilation2
0:02016
After stereo optical microscope evaluation, thin sectionswere
made to assess the microstructural characteristics ofthe concrete.
This evaluation was conducted on specimensthat were kept at
laboratory ambient conditions for longerthan 56 days. Those made
strictly for petrographic analysiswere initially moist cured at 23
"C for seven days. Using thepetrographic microscope, the paste
homogeneity andmicrocracking was assessed on a scale of 13 for
eachbatch. High levels of inhomogeneity indicate that the den-sity
of the hydrated cement paste varies due to diculties inuniformly
dispersing the cement grains and/or in obtaininguniform hydration.
The results of this analysis are shown inTable 4. The mixtures
rated with the highest degree ofinhomogeneity were those made with
Type III cement,although some batches made with Type III cement
hadgood paste uniformity (e.g. N-14-A). In comparing
0
10
20
30
40
50
60
70
80A
-1-A
A-1
-B
B-2-
A
B-2
-B
C-3-
A
C-3-
B
D-4
-A
D-4
-B
E-5-
A
E-5-
B
F-6-
A
F-6-
B
G-7-
A
G-7-
B
H-8
-A
H-8
-B
I-9-A
I-9-B
J-10
-A
J-10
-B
K-11
-A
K-1
1-B
L-12
-A
L-12
-B
M-13
-A
M-13
-B
N-14
-A
N-1
4-B
Mixture ID
Com
pres
sive
stre
ngt
h, M
Pa
6 hour tes t 8 hour tes t 28 days tes t Threshold s trength
0
0.5
1
1.5
2
2.5
3
A-1 B-2 C-3 D-4 E-5 F-6 G-7 H-8 I-9 J-10 K-11 L-12 M-13 N-14
Mixture ID
Avg.
flex
stre
ngt
h, M
Pa
6-8 hours Opening criteria
00.5
11.5
22.5
33.5
4
0 5 10 15 20 25 30 35
Average compressive strength, MPa.
Aver
age
fle
xu
rals
tren
gth,
MPa
.
8-hr tes ts 6-hr tes ts Linear (6-hr tes ts ) Linear (8-hr tes
ts )
a b
c
Fig. 4. Measured compressive and flexural strengths for the
various high-early strength mixtures. (a) Average compressive
strength data for the high-earlystrength mixtures (dashed line
indicates the opening criteria). (b) Average flexural strength data
for the high-early strength mixtures (dashed line indicatesthe
opening criteria). (c) Early age comparison between compressive and
flexural strength results.
168 N. Buch et al. / Construction and Building Materials 22
(2008) 162174
-
mixtures made with Type I cement, the eect of the Type FHRWR is
observed in that mixture G-7 had greater pasteuniformity than B-2.
Similar benefits were not noted withthe Type E water reducer, where
moderate inhomogeneitywas observed both in the 68 h mixture
(H-8).
A small degree of microcracking is a common feature inconcrete,
and thus is not in and of itself of great conse-quence. A rating of
1 indicated an absence of micro-cracking. Moderate microcracking
(rating of 2),although not desirable, is thought not to indicate a
prob-lem with the mixture and would not be assumed to have agreat
impact on the durability of the concrete. But thesevere
microcracking (rating of 3) noted in all of thesemixtures is out of
the ordinary, and may be an indicatorof potential problems in the
future for these high earlystrength materials.
These same thin sections were evaluated using theESEM in
backscatter electron mode. As was seen usingthe stereo optical
microscope, very dierent air-void sys-tems were observed between
the mixtures prepared usinga Type I cement and those with a Type
III cement, withair bubbles being far more abundant and smaller in
theType I cement mixtures. This observation is illustrated inFig.
8. It is also observed that the hydrated cement pastewas more
uniform in the Type III cement mixtures, primar-ily a result of the
finer cement, which left fewer unhydratedcement grains. The SEM
images also revealed that thehydrated cement paste in mixtures made
with the non-chlo-ride accelerator was distinctly more uniform than
that inthe mixtures made with calcium chloride.
The final microstructural observation was the estimationof crack
length per unit area of concrete subjected tofreeze-thaw testing
(AASHTO T 161) as summarized inTable 4. It is observed that
significant microcracking waspresent in the high-early strength
concrete mixtures. Thisis not surprising as significant
microcracking was originallyobserved in the mixtures using
petrographic means.
Table 4Summary of microstructural assessment of high-early
strength mixtures
Mixture Stereoopticalmicroscopeobservations
Petrographicmicroscoperatings
Crackingin freeze-thawbeams(lm/mm2)
Percentair
Spacingfactor(mm)
Pastehomogeneity
Microcracking
A-1-A 5.04 0.1519 2 3 15.670A-1-B 4.68 0.1252 2 3 0.546B-2-A
5.33 0.1427 2 3 33.713B-2-B 5.55 0.1383 2 3 21.368C-3-A 6.48 0.1225
2 3 52.232C-3-B 6.12 0.1206 2 3 42.735D-4-A 9.94 0.1393 2 3
12.821D-4-B 5.91 0.1447 1 3 15.195E-5-A 7.42 0.1360 2 3 12.583E-5-B
9.44 0.1091 2 3 4.036F-6-A 4.47 0.1774 1 3 33.476F-6-B 5.26 0.1584
1 3 17.094G-7-A 6.34 0.1570 1 3 30.627G-7-B 6.48 0.1401 1 3
29.677H-8-A 6.20 0.1103 2 3 23.742H-8-B 4.03 0.1179 2 3 32.051I-9-A
4.54 0.1075 1 3 8.072I-9-B 4.47 0.1121 2 3 8.072J-10-A 5.55 0.1178
1 3 30.627J-10-B 5.91 0.1171 1 3 28.965K-11-A 6.20 0.2320 2 3
17.094K-11-B 9.65 0.1364 2 3 3.561L-12-A 3.67 0.2867 2 3
26.353L-12-B 4.39 0.1997 3 3 41.311M-13-A 4.25 0.2398 3 3
42.735M-13-B 3.89 0.2298 3 3 53.656N-14-A 3.60 0.2585 1 3
0.000N-14-B 3.17 0.2252 2 3 16.144
Average 5.642 0.15908 1.8 3 23.3627St. Dev 1.7624 0.051278 0.63
0 15.00881COV 31.2% 32.2% 35.3% 0 64.2%
0.00
1.00
2.00
3.00
4.00
5.00
6.00
7.00
8.00
9.00
A-1 B-2 C-3 D-4 E-5 F-6 G-7 H-8 I-9 J-10 K-11 L-12 M-13 N-14
Mixture ID
AST
M C4
57 m
easu
red
air,
%
Fig. 5. Air content measured using ASTM C 457 (dashed lines
indicated desired target range).
N. Buch et al. / Construction and Building Materials 22 (2008)
162174 169
-
Microcracking correlated in an imprecise way to the mea-sured
dilation values, but when cracking is plotted againstcement factor
for all mixtures as in Fig. 9, it is easilyobserved that more
cracking is measured in higher cementfactor mixtures. This
observation likely reflects that there ismore paste present in such
mixtures thus more cracking,but it may also suggest high cement
content mixtures aremore susceptible to microcracking.
4. Statistical analysis
To study the eect of each individual factor (single fac-tor
eects) on a test result, all other mixture factors must bekept
constant. Any resulting dierence in the test result canthen be
attributed to the variation of the factor understudy. For example,
to study the eect of an increase ordecrease in a mixtures cement
factor, test results for mix-
tures A and B can be compared. Both used Type I cement,have a
0.40 w/c ratio, contain carbonate coarse aggregate,utilized a
non-chloride accelerator with no water reducerand were cured at 23
"C. The only dierence between thetwo mixtures is that they use a
dierent cement factor, withmixture A containing 425 kg/m3 of cement
whereas mix-ture B used 525 kg/m3 of cement. Table 5 lists the
possiblemixture pairs that can be used to study the eect of
eachmixture factor. For each test conducted, the mean, vari-ance
and 95% confidence intervals for each factor was cal-culated. To
determine if a mixture variable aects a certaintest result,
p-values are calculated. A p-value of less than0.05 indicates that
the mixture variable has a significantimpact on that test
result.
Varying the cement type (mixtures G and L) in the mix-ture from
Type I to Type III cement had a significantimpact on the (i) air
void volume (p = 0.0232), (ii) concrete
0.0000
0.0500
0.1000
0.1500
0.2000
0.2500
0.3000
A-1 B-2 C-3 D-4 E-5 F-6 G-7 H-8 I-9 J-10 K-11 L-12 M-13
N-14Mixture ID
Spac
ing
fact
or, m
m.
Spacing factor criterion = 0.2000 mm
Fig. 6. Spacing factor measured using ASTM C 457.
-0.01
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.0000 0.0500 0.1000 0.1500 0.2000 0.2500 0.3000 0.3500Spacing
factor,mm
Dila
tion
, m
m/m
m
exceed spacing factor threshold
exceed dilation threshold
G -7A
N-14B
N-14A
L -12A
L -12 B
M-13B
K -11 AM-13A
Fig. 7a. Relationship between spacing factor and dilation due to
freezing and thawing.
170 N. Buch et al. / Construction and Building Materials 22
(2008) 162174
-
Fig. 8. Dierence in air-void system from Type I (Mixture E-19)
cement on left and Type III (Mixture M-27) on the right.
Magnification 50.
0
10
20
30
40
50
60
70
80
90
100
0.0000 0.0500 0.1000 0.1500 0.2000 0.2500 0.3000 0.3500Spacing
Factor (mm)
Dur
abili
ty F
act
or
ex ceed s ac in factor th resh old
D ura b ility F actor th res h old
M-13AK -11A
M-13B
L -12BG -7A
L -12A
N-14A
Fig. 7b. Relationship between spacing factor and calculated
durability factor due to freezing and thawing.
0
10
20
30
40
50
60
400 425 450 475 500 525 550 575 600Cement factor, kg/m3
Cra
ck le
ngt
h, m
icro
ns/m
m2
Fig. 9. Cement factor versus crack length per unit area for all
mixtures.
N. Buch et al. / Construction and Building Materials 22 (2008)
162174 171
-
workability as determined by the slump test (p = 0.0343)and
(iii) 68 h mechanical properties (p = 0.01). Air voidvolume was
found to be higher in mix G (type I) than inmix L (type III).
By changing the cement factor from 425 to 525 kg/m3
(mixtures A versus B), significant impacts occurred in the(i)
concrete workability as determined by the slump test(p = 0.0136);
and (ii) mechanical properties (p = 0.005).
Similarly, comparing the eect of cement factor of 425versus 525
kg/m3 in calcium chloride accelerated mixtures(mixtures I versus J:
Type I cement, calcium-chloride accel-erator and no water reducer)
resulted in significantlyimpacting the (i) air void volume (p =
0.0141); (ii) concreteworkability (p = 0.0250) and (iii) concrete
mechanicalproperties (p = 0.0310).
Varying the w/c ratio (mixture B versus F) from 0.36to 0.40
caused significant eect on the (i) mechanicalproperties of the
hardened concrete (p = 0.02); (ii) work-ability of the fresh
concrete (p = 0.0246); and (iii) freeze-thaw dilation (p = 0.0249).
The majority of these mechan-ical tests are either strength (22,
24, 25, 26 and 27) ordensity (30, 31, 33 and 34) tests. The other
test results thatare significant to a change in w/c ratio are
coecient ofthermal expansion (1), freeze-thaw dilation (19) and
slump(20). The coecient of thermal expansion, freeze-thawdilation,
significant strengths and bulk densities of mixtureF (w/c = 0.36)
were higher than those of mix B (w/c =0.40).
The eects of coarse aggregate (mixtures A versus E; Cversus D)
were significant on the (i) density of the hardenedconcrete
specimen (p = 0.0298) regardless of cement factorand curing regime;
(ii) air-void spacing factor (p = 0.0190)if the cement of 525 kg/m3
is used and cured at a tempera-ture of 65 "C. In this case, the
air-void spacing factor islower while the slump is higher when
carbonate aggregate(71-47) is used (Mix C) in comparison to
siliceous aggre-gate (95-10) is used (Mix D). On the other hand,
the coarseaggregate does not have an eect upon these tests if
thecement factor is 425 kg/m3 and the curing temperature is23 "C
(Mixes A and E).
The use of dierent accelerators with a cement factor of425 kg/m3
of Type I cement with no water reducer (mix-tures A and I)
significantly (p < 0.05) impacted concreteworkability and
hardened mechanical properties. Howeverthe same concrete properties
were not significantly aectedwhen using the two dierent
accelerators if 525 kg/m3 of aType I cement is used with no water
reducer or if 525 kg/m3 of Type III cement is used with a Type F
high-rangewater reducer.
Accelerator type is an important factor for the thermaland
strength properties of concrete if the mixtures aremade using 425
kg/m3 of a Type I cement with type Fwater reducer (mixtures K and
N) and it is not importantfor the other mixture combinations. Mix K
(non-chlorideaccelerator) resulted in a higher coecient of
thermalexpansion than mix N (calcium chloride accelerator).Mix N on
the other hand tested to have a higher 24 h T
able5
Mixture
pairsusedin
68hopeningtime
Factor
Mixes
used
Cem
enttype
Cem
entfactor(kg/m3)
w/c
Ratio
Coarse
aggregate
Accelerator
Water
reducer
Curingtemperature
Cem
enttype
GandL
I/III
525
0.40
Carbonate
Non-Chl
F23
"CCem
entfactor
AandB
I425/525
0.40
Carbonate
Non-Chl
No
23"C
IandJ
ICaC
lNo
KandL
III
Non-Chl
Fw/c
Ratio
BandF
I525
0.40/0.36
Carbonate
Non-Chl
No
23"C
Coarse
aggregate
AandE
I425
0.40
Carb/Silic
Non-Chl
No
23"C
CandD
525
65"C
Accelerator
AandI
I425
0.40
Carbonate
Non-Chl/CaC
lNo
23"C
BandJ
I525
No
KandN
III
425
FWater
reducer
BandG
I525
0.40
Carbonate
Non-Chl
No/F
23"C
Curingtemperature
BandC
I525
0.40
Carbonate
Non-Chl
No
23"C/65"C
LandM
III
F
172 N. Buch et al. / Construction and Building Materials 22
(2008) 162174
-
compressive strength and eective diusion coecient thandid mix
K.
The eect of water reducers was investigated by compar-ing
mixtures B and G. The statistical analysis of the dataindicated
that the admixture significantly (p < 0.05)impacted the air-void
system and the mechanical propertiesof the hardened concrete. Mix
B, which used no waterreducing admixture, resulted in better
air-void system char-acteristics and significant increases in
strength than did mixG which utilized a Type F water reducing
admixture.
The eects of curing on the various fresh and hardenedconcrete
properties were investigated by comparing mix-tures B and C and
mixtures L and M. The dierent curingtemperatures significantly (p
< 0.05) impacted the paste-to-air ratio and air-void spacing
factor, freeze-thaw beamdilation, and the 28-day compressive
strength when TypeI cement with no water reducer was used. The
curing tem-perature eect was significant for the 6 and 8 h
compressivestrength tests when Type III cement was used with a
TypeF water reducer.
Table 6 summarizes the level of significance for eachmixture
variable on the properties of the fresh and hard-ened concrete
mixtures.
! An Important mix factor is significant to that test nomatter
what other combinations of mix variables areused.
! A factor that is important under Certain Conditionsis
important to that test for some of the mixtures butnot in every
case where that factor was varied.
! A factor that is Not Important is not significant tothat
particular test in any of the tested mixtures.
5. Conclusions
The results presented in this paper are part of a compre-hensive
laboratory and field study sponsored by theNCHRP that was
undertaken to evaluate the fresh, hard-ened, volumetric and
microstructural properties of high-early strength concrete used in
concrete pavement rehabil-itation. Based on the results of the
experimental programand statistical analysis the following
conclusions can bedrawn:
1. For the specific cements used, the cement type had anotable
impact on various measures of early concretestrength, with Type III
cements producing higher
strengths at a given age than Type I cements. Thisincreased rate
of hydration was also evidenced in thematurity readings.
2. As expected, the coarse aggregate type had a very
signif-icant impact on measures of density, with denser aggre-gate
resulting in denser concrete. Mixtures made withthe gravel coarse
aggregate had the highest coecientof thermal expansion, whereas
mixtures made with thecarbonate aggregate had the least scaling.
Aggregatetype also aected some of the strength properties ofthe
mixtures, most notably compressive strength.
3. In many instances, the use of the finer Type III cement,in
conjunction with the Type F HRWR, negatively influ-enced the
air-void system parameters, most acutelywhen calcium chloride
accelerator was used. Thisresulted in generally poorer freeze-thaw
performance.It is noted that only single sources of Type III
cementand Type F HRWR were used, and thus no generalizedconclusions
regarding the suitability of Type III cementand/or Type F HRWR can
be drawn.
4. All the high-early strength mixtures had significantlyhigh
levels of microcracking as assessed viewing thin sec-tions under a
petrographic microscope. High cementcontent mixtures had relatively
greater amounts ofmicrocracking, likely due to the increased paste
content.
5. The use of the Type F HRWR in the mixtures reducedthe air
content and negatively impacted a number ofthe air-void system
parameters, resulting in poorerfreeze-thaw performance. It also
negatively impactedthe early strength characteristics of the
concrete whenType I cement was used. Mixtures containing Type EWR
exhibited poor paste homogeneity and very lowearly strengths.
6. Due to the unique characteristics of these materials andthe
increased potential for interactions between the var-ious mixture
constituents, it is recommended that trans-portation agencies apply
a thorough testing regimeduring the mixture design process and
employ a QC/QA program during construction to ensure the
produc-tion of durable high-early strength mixtures.
Acknowledgements
The authors acknowledge the financial support givenby the
National Cooperative Highway Research Programand the assistance
provided by Dr. Amir Hanna, theNCHRP Senior Program Ocer overseeing
this work
Table 6Factor significance for concrete properties
Concrete property Important Factors important under certain
conditions Not important
Workability (slump) Cement type, w/c ratio, water reducer Cement
factor, coarse aggregate, accelerator Remaining factorsPhysical
(density) w/c ratio, coarse aggregate Cement factor Remaining
factorsAir-void system Cement type, water reducer Coarse aggregate,
cement factor, accelerator, curing regime Remaining
factorsFreeze-thaw dilation w/c ratio Curing regime Remaining
factorsStrength Water reducer, cement type, w/c ratio Coarse
aggregate type, cement factor, curing regime Accelerator
N. Buch et al. / Construction and Building Materials 22 (2008)
162174 173
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
Jonathas Iohanathan Felipe de Oliveira
-
described. Furthermore, the authors acknowledge MDOTfor
providing invaluable assistance with the concrete mix-ture design
development, specimen fabrication and test-ing.
References
[1] Van Dam TJ. Durability of Early-Opening-to-Trac
portlandcement concrete for pavement rehabilitation. Final Report,
NCHRPWeb Document 76, Transportation Research Board,
NationalResearch Council, Washington, DC; 2005.
[2] Whiting D, Nagi M, Okamoto P, Yu T, Peshkin D, Smith K, et
al.Optimization of highway concrete technology. SHRP-C-373,
Strategic
Highway Research Program, National Research Council,
Washington,DC; 1994.
[3] American Concrete Pavement Association (ACPA). Concrete
pavingtechnology guidelines for full-depth repair. TB002-02P,
Skokie,Illinois; 1995.
[4] Hoerner TE, Smith KD, Yu HT, Peshkin DG, Wade MJ.
PCCPavement Evaluation and Rehabilitation, Reference Manual,
NHICourse Number 131062, Federal Highway Administration,
USDepartment of Transportation; October 2001. 410pp.
[5] Zollinger DG. Assessment of durability performance of
Early-Opening-to-Trac portland cement concrete. Final Report,
NCHRPProject 18-04A, National Cooperative Highway Research
Program,Transportation Research Board, National Research Council,
Wash-ington, DC; 1998.
174 N. Buch et al. / Construction and Building Materials 22
(2008) 162174
Evaluation of high-early strength PCC mixtures used in full
depth repairsIntroductionExperimental programExperimental
matrixSpecimen preparationTests conducted on laboratory prepared
specimensProperties of fresh concreteFreeze-thaw
durabilityMicrostructural characterization
Laboratory test results and discussionProperties of fresh
concreteFreeze-thaw durabilityMicrostructural characterization
Statistical analysisConclusionsAcknowledgementsReferences