Construction and Building Materials 25 (2011) 3480–3485

Contents lists available at ScienceDirect

Construction and Building Materials

journal homepage: www.elsevier.com/locate/conbuildmat

Comparison of test specimen preparation techniques for pervious

concrete pavements
Bradley J. Putman ⇑, Andrew I. Neptune
Department of Civil Engineering, Clemson University, Clemson, SC 29634, USA

a r t i c l e i n f o a b s t r a c t

Article history: The objective of this study was to evaluate different pervious concrete test specimen preparation tech-
Received 6 October 2010 niques in an effort to produce specimens having properties similar to in-place pervious concrete pave-
Received in revised form 28 December 2010 ment. Cylinders and slabs were cast using pervious concrete from three different paving projects using
Accepted 1 March 2011
different procedures. The comparisons of cast specimens to pavement cores were based on infiltration
Available online 25 March 2011
rate, density, and porosity. Of the cylinder consolidation procedures tested, the standard Proctor hammer
provided the least variability of results and yielded properties similar to the in-place pavement. However,
Keywords:
600 mm square slabs were even more consistent with the in-place pavement density and porosity.
Concrete
Pervious
Ó 2011 Elsevier Ltd. All rights reserved.
Porous
Permeable
Pavements
Testing
Porosity
Density
Specimen preparation

1. Introduction structure allows both water and air to percolate through the matrix
into the subsoil beneath. Because of the interconnected pores, per-
1.1. Background vious concrete reduces runoff, but also performs the role of a filter
by the degradation and entrapment of contaminants (e.g., oils and
Pervious concrete is a concrete mixture comprised of cement, debris) on and within the pervious concrete structure [2]. The size
controlled amounts of water, uniformly graded coarse aggregate, of these pores is affected by the gradation and type of the aggre-
little or no sand, and sometimes other additives. The uniform gate in the mix, the quantity of water and cement added, and the
coarse aggregate gradation in combination with low water content level of compaction.
(i.e., water-to-cement ratios ranging from 0.25 to 0.35) make for a Pervious concrete pavements also have greater acoustic absorp-
concrete mixture with void contents ranging from 11% to 35% [1,2]. tion properties than conventional concrete pavements, which
This allows for high water and air permeability. On the other hand, results in quieter pavements. This superior acoustic absorption is
the compressive, tensile, and flexural strength of pervious concrete due, again to the porous nature of the material. Neithalath et al.
mixtures tend to be lower than conventional concrete due to the studied the effect of various aggregate sizes and gradations of
high void ratio and lack of fine aggregate [3]. pervious concrete and found that the acoustic absorption of the
Although pervious concrete has been in existence since the concrete increased as the porosity and permeability of the material
middle of the 19th century, it was not until the 1980s that pervious increased [5].
concrete became popular in the United States. When it gained More recently, pervious concrete has the potential of being cer-
popularity in the US, pervious concrete was utilized in pavement tified for construction projects by the US Green Building Council’s
applications as a stormwater management solution. The high per- Leadership in Energy and Environmental Design (LEED) Green
meability of pervious concrete creates the ability to significantly Building Rating System, because of its environmental benefits
reduce stormwater runoff; therefore, it has been considered a along with the potential capability of lowering the heat island
stormwater Best Management Practice (BMP) [4]. The porous effect [1]. Urban areas tend to enclose large areas of impervious
pavements, which add to the level of heat. The concentrated
heat wave can be reduced by the open structure of pervious con-
⇑ Corresponding author. Tel.: +1 864 656 0374; fax: +1 864 656 2670. crete that allows air to flow through it. This has been exhibited
E-mail address: [email protected] (B.J. Putman). in limited studies of pervious concrete pavement applications

0950-0618/$ - see front matter Ó 2011 Elsevier Ltd. All rights reserved.
doi:10.1016/j.conbuildmat.2011.03.039
 B.J. Putman, A.I. Neptune / Construction and Building Materials 25 (2011) 3480–3485 3481

[6,7]. Additionally, the roots of plants and trees adjacent to these Table 1
pavements not only experience watering but also aeration [1]. Pervious concrete mix design used for the three projects. Admixture dosage rates
were based on manufacturer recommendations.
However, to realize all of the benefits of pervious concrete, it is
important that pervious concrete pavements be designed and con- Component Quantity per m3
structed properly. Because of the unique properties of pervious Type I/II cement 356 kg
concrete (fresh and hardened), it is important to pay close atten- #8 Crushed granite 1435 kg
tion to details during mix production, concrete placement, and cur- Water 123.8 L
Air 21.9%
ing. Placement of pervious concrete requires alternative methods
to those used for conventional concrete pavement construction. Admixtures Quantity per 100 kg of cement

The low slump requires that the concrete be consolidated by roll- Low-range water reducer 130 mL
ing in most cases. The goal of compaction is to seat the aggregate Retarder 33 mL
Hydration stabilizer 196 mL
and enhance the bond between aggregates, but it is important that
excessive compaction effort not be used that would cause the voids
to collapse thus reducing the porosity [8].
As the placement of pervious concrete in the field requires techniques to determine which procedures would best represent
unique procedures compared to conventional concrete, it is also the in situ properties of pervious concrete pavements. Additionally,
important to consider the procedures used to create pervious con- the relationships between pavement infiltration rate, porosity and
crete test specimens in either the laboratory or field. Casting pervi- density were evaluated.
ous concrete cylinders using the same procedures outlined for
conventional concrete can potentially result in an inaccurate repre- 2. Materials and methods
sentation of the in situ properties of a pervious concrete pavement.
Research conducted by Rizvi et al. evaluated various compac- To accomplish the objective of this study, test specimens were cast using multi-
ple techniques and compared to cores taken from the pavements from which the
tion methods for pervious concrete cylinders (150 mm diameter
samples were prepared.
by 300 mm height) with the goal of finding a procedure(s) that
could be used to create specimens closely representing the pervi-
2.1. Materials
ous concrete in the field [9]. The different consolidation techniques
that were studied included rodding and a 2.5 kg Proctor hammer. During this research project, the research team cast specimens that were com-
These samples were prepared from the identical mix design as pared to the in-place pavement for three different pervious concrete paving pro-
the field concrete and were cured under laboratory conditions. jects. All of the pavements were placed by the same contractor and used the
The properties evaluated included compressive strength, perme- same mix design and materials supplied by the same ready-mix concrete supplier
(Table 1). Each of the three pavements is a parking lot and the specifics are summa-
ability, and air voids. It was determined that the samples prepared
rized in Table 2.
in two layers with ten blows of a standard Proctor hammer (2.5 kg)
per layer resulted in the closest properties to that of field pervious
2.2. Methods
concrete. However, none of the samples were within the range of
the void content for the field cores, but the void contents were 2.2.1. Pavement placement procedures
within the generally anticipated range of values for pervious All three pavements were placed on top of crushed stone base that had been
concrete. prepared ahead of time. Concrete was delivered to the site in front discharge transit
trucks. Each truck contained approximately 3.8 m3 for Projects A and C and approx-
Mahboub et al. also compared the properties of pervious
imately 5.4 m3 for Project B. The concrete was discharged from the chute in an arc
concrete cylinders prepared using different methods to the proper- as the truck backed up over the stone base. The workers moved the mix with come-
ties of in-place pervious concrete pavement [10]. Specifically, they alongs as needed, but not to excess. The crew then used a hydraulic roller screed
evaluated two different methods: rodding in accordance with ASTM that spun in the opposite direction than it was pulled. Following the screed, the
C192 and a custom built pneumatic press that applied a compactive fresh concrete was cross-rolled using a 1 m long roller having a 150 mm diameter.
As placement progressed, transverse joints were formed with a rolling joint cutter
effort of 70 kPa uniformly over the 100 mm diameter cylinders. Test every 3.5–4.5 m depending on the pavement configuration. Additionally, after the
results indicated that the cylinders compacted using the pneumatic fresh concrete was screeded to the proper grade, the fresh concrete was sprayed
press yielded statistically similar properties (compressive strength, with soy bean oil to aid curing and act as a sealer. Finally, the finished concrete
permeability, and porosity) as the pavement cores. The rodded was immediately covered with plastic for 7 days for curing purposes.
specimens, on the other hand had compressive strength values that
were significantly higher than the field cores and porosity values 2.2.2. Specimen preparation (cylinders and slabs)
that were significantly lower than the field cores. During placement of the pervious concrete pavements, 150 mm diameter by
300 mm tall cylinders were cast for all three projects. Concrete was sampled from
at least two trucks from each project to make cylinders. Cylinders were prepared
1.2. Objectives using different consolidation techniques in an effort to identify methods that would
best replicate the properties of in-place pervious concrete. To this end, eight differ-
ent methods were evaluated as summarized in Table 3. These methods included the
The main objective of this study was to compare the properties use of either a 15.9 mm diameter steel rod (standard tamping rod), a standard Proc-
of pervious concrete specimens prepared using a variety of tor hammer (2.5 kg), and dropping the mold on a concrete surface from a height of

Table 2
Summary of pervious concrete pavement project locations.

Project A Project B Project C

Location Aiken, SC Augusta, GA Aiken, SC
Use Public works parking spaces Restaurant overflow parking lot Public parking lot
Design pervious concrete thickness 150 mm 100 mm 150 mm
Pervious concrete area 400 m2 1035 m2 860 m2
Design stone base thickness 150 mm 200 mm 150 mm
Subgrade material Sandy Clay Sandy
Placement date March 2010 April 2010 May 2010
3482 B.J. Putman, A.I. Neptune / Construction and Building Materials 25 (2011) 3480–3485

Table 3 The cores that were obtained from the pavements (Projects A, B, and C) were
Consolidation techniques used to prepare cylinders. also cut into equal slices, but the thickness of the slices for each core varied based
on the thickness of the core. Each slice was then tested in the same manner as the
Consolidation method No. of No. of consolidation cylinders and cores from the slabs (porosity and density).
layers repetitions
P-2-5 Standard Proctor hammer 2 5
P-2-10 Standard Proctor hammer 2 10 3. Results and discussion
R-2-5 15.9 mm Tamping rod 2 5
R-2-10 15.9 mm Tamping rod 2 10 3.1. In-place pavement properties
R-2-15 15.9 mm Tamping rod 2 15
R-2-25 15.9 mm Tamping rod 2 25
R-3-25 15.9 mm Tamping rod 3 25 Fig. 2 summarizes the results of the infiltration rate of each of
D-2-10 50 mm Drop 2 10 the pavements. The results indicate that the infiltration rate varies
widely throughout the pavement. This is fairly typical of pervious
concrete pavements because of the potential sources of variability
50 mm. After consolidation, the concrete was screeded flush with the mold and during construction such as: variation in mix properties from truck
then a plastic cap was placed on the mold. The cylinders were placed on a level sur- to truck, mix delivery delays or queues, equipment malfunction,
face near the newly placed concrete where they remained undisturbed for 7 days. and weather to name a few [13]. It is important to note that each
During the placement of pervious concrete for Projects A and C, slabs of pervi- infiltration test is a measure of the infiltration of the pavement sur-
ous concrete were also cast to determine if slab specimens provided a more consis-
tent test specimen compared to the cylinders. The slabs were prepared by placing
face in one localized spot having a diameter of 300 mm. However,
wooden forms on top of the stone base material so the tops of the forms were flush in the case of these three pavements, the number of locations hav-
with the finished pavement grade. Concrete was discharged into the forms as the ing low infiltration values was limited and they were surrounded
pavement was placed. After finishing with the roller screed and cross-roller, the by areas having adequate infiltration where the water would drain
forms containing the concrete were removed from the fresh concrete pavement
to and infiltrate through the surface and into the underlying infil-
and covered with plastic and left to cure for 7 days. All of the slabs had the same
thickness as the pavement (150 mm) and three different sizes of slabs were cast tration bed.
(300  300 mm, 450  450 mm, and 600  600 mm). One slab of each size was cast Fig. 3 shows the results of the properties of the cores obtained
for each of Projects A and C. from each project. These results indicate that there was no statis-
tically significant difference between the in situ properties of the
2.2.3. Test procedures (cylinders, slabs, cores)
different pavements. This can be attributed to the fact that the
Seven days after placement of each pavement, the plastic was removed and same construction crew placed the three pavements and the same
infiltration tests were conducted on the pavement and cores were obtained for
lab testing. Each pavement was divided into sublots ranging in length from 6 to
12 m. One infiltration test was conducted in accordance with ASTM C1701 at a ran- 8000
dom location within each sublot [11]. Following the infiltration testing, 5–6
150 mm diameter cores were taken from locations in the pavement that covered 7000
Infiltration Rate (cm/hr)

the range of infiltration rates. All of the specimens (cores, cylinders, and slabs) were
then transported to the laboratory for further testing.
6000
The scope of testing on the specimens for this study focused on the volumetric 5000
and hydrologic related properties of the pervious concrete specimens. The test plan
is summarized in Fig. 1. After the molds were stripped from the cylinders, each cyl- 4000
inder was cut into four equal slices. This was done to evaluate the porosity distri- 3277
bution along the vertical axis of the cylinders. The porosity (accessible air void) 3000
and density of each slice was measured using the vacuum sealing method in accor- 2286
2000 1830
dance with ASTM D7063 [11,12].
The slabs that were cast during the placement of Projects A and C were removed 1000
from the molds and the infiltration rate was tested at the center of each slab using
ASTM C1701 [11]. To mimic the field conditions, the slabs were placed on top of 0
150 mm of crushed stone based having the same gradation as the pavement for A B C
conducting the infiltration rate of each slab. Following the infiltration tests, cores Project
were taken from each slab and tested in the same manner as the cylinders discussed
previously (porosity and density). This testing was completed to compare the prop- Fig. 2. Infiltration rates of pavements prior to opening to traffic. Error bars
erties of the slabs to the cylinders as well as to attempt to correlate the porosity of represent one standard deviation.
the slabs to the infiltration rate.

Project A Project B Project C

Cylinders Slabs Pavement Cylinders Pavement Cylinders Slabs Pavement

Porosity & Porosity & Porosity &

Infiltration Infiltration Infiltration Infiltration Infiltration
Density Density Density

Cores Cores Cores Cores Cores

Porosity & Porosity & Porosity & Porosity & Porosity &
Density Density Density Density Density

Fig. 1. Flow chart of experimental approach.

 B.J. Putman, A.I. Neptune / Construction and Building Materials 25 (2011) 3480–3485 3483

100
(a) BC C A AB C
100
2000 2000

80 80
1800 1800
1770
1733

Density (kg/m3)

Porosity (%)
1704
Density (kg/m3)

Porosity (%)
60
60 1600
1600 Density Density
Porosity Porosity 40
40 1400
1400
28.2 29.9
27.0 1200 20
1200 20
A AB B B AB
1000 0
1000 0
A B C
Project Consolidation Method

Fig. 3. Properties of cores from Projects A, B, and C. Error bars represent one
standard deviation. (b) AB C BC AB AB A BC
100
2000

80
50 1800

Porosity (%)
Density (kg/m3)
60
40 1600 Density
Porosity
Porosity (%)

40
30 1400

20 1200 20

y = -0.05x + 106 A A AB AB B B AB
10 R² = 0.86 1000 0

0
1400 1600 1800 2000 Consolidation Method
Density (kg/m3)

Fig. 4. Correlation between pavement density and porosity.

(c) A A A A A A A A
100
2000

80
1800

Porosity (%)
Density (kg/m3)

2000 50 60
1600
Density
40 Porosity 40
1500 1400
Density (kg/m3)

Porosity (%)

30 1200 20

1000 A A A A A A A A
20 1000 0
Density = -0.08x + 1813
500
R² = 0.72
10 Consolidation Method
Porosity = 0.004x + 24.1
R² = 0.69 Fig. 6. Porosity and density of cylinders cast from (a) Project A; (b) Project B; and
0 0 (c) Project C. Error bars represent one standard deviation. Treatments with one or
0 1000 2000 3000 4000 more common letter on the plot are statistically similar (a = 0.05).

Pavement Infiltration Rate (cm/hr)

Fig. 5. Relationship of pavement infiltration rate to pavement density and porosity. increase in density. When combined with the infiltration testing
data, Fig. 5 shows the correlation between pavement infiltration
rate and both density and porosity of the in-place pavement. These
mix design was used from the same supplier for the three projects. relationships were developed using the infiltration data from all
This further demonstrates that this particular contractor has a pro- three projects and the properties of the cores taken from the loca-
cess to place pervious concrete that is consistent and reproducible. tions where the infiltration was measured. There is a definite cor-
Fig. 4 shows the correlation between density and porosity of the relation between both pavement density and porosity with
pavement cores taken from Projects A, B, and C. The correlation is pavement infiltration. As the pavement density decreases, the
quite strong (R2 = 0.86) with the porosity decreasing with an porosity increases, which leads to increased infiltration rate. While
3484 B.J. Putman, A.I. Neptune / Construction and Building Materials 25 (2011) 3480–3485

Fig. 7. Photos of core slices for rodded specimens (R-2-10); Proctor hammer specimens (P-2-10); and drop-compacted specimens (D-2-10).

8000
(a) B AB A
100
Infiltration Rate (cm/hr)

7000 2000
6774
6000
Slabs 80
5000 Pavement 1800
1734
Density (kg/m3)

Porosity (%)
4000 1671 60
1650
3000 3073 1600
2891 Density
2000 Porosity 40
1344 1400
1000 31.2 30.2
26.1
0 20
A C 1200
Project
B A A
Fig. 8. Infiltration rate of slabs cast from (a) Project A and (b) Project C. Error bars 1000 0
represent one standard deviation. 300 300 450 450 600 600
Slab Dimensions (mm)
2
this trend was expected, the relatively high R values for field data
(0.72 and 0.69 for density and porosity, respectively) indicate (b) A B AB
100

favorable correlations. 2000

Density
Porosity 80
3.2. Properties of cylinders 1800
Density (kg/m3)

Porosity (%)
The porosity and density results of the tests on the cylinders 60
cast from each project are included in Fig. 6. The objective of this
1600 1597
1539
portion of the study was to determine which cylinder consolida- 1506
38.9 40
tion technique(s) produced cylinders having similar properties as 1400 37.1 38.1
the in-place pavements. For each graph in this figure, the average
porosity and density of the cores extracted from the respective
1200 20
pavement are included (dashed lines) along with the standard
deviation (dotted lines) as a means for comparison with the
A A A
in situ pavement properties. 1000 0
A statistical analysis (Fisher’s test for least significant differ- 300 300 450 450 600 600
ence) was conducted on the data to determine which cylinder con- Slab Dimensions (mm)
solidation techniques produced cylinders having statistically
similar properties (porosity and density) at a 95% level of signifi- Fig. 9. Porosity and density of cylinders cast from (a) Project A and (b) Project C.
Error bars represent one standard deviation. Treatments with one or more common
cance (a = 0.05). For the cylinders prepared from Project A, the
letter on the plot are statistically similar (a = 0.05).
P-2-10, R-2-10, and D-2-10 consolidation methods produced cylin-
ders having similar porosity and density values. These methods
were also closest to the average core (in-place pavement) porosity and R-2-25 were closest to the average pavement density. Finally,
and density. Of the cylinders prepared from Project B, those consol- for Project C, all of the consolidation techniques resulted in poros-
idated using P-2-10, R-2-5, R-2-10, and D-2-10 had statistically ity and density values that were statistically similar to each other.
similar porosity values. However, the P-2-10 and D-2-10 methods However, P-2-5, P-2-10, R-2-5, and D-2-10 were closest to the
produced cylinders having porosity values closest to the mean mean core density.
pavement porosity. In terms of density, methods P-2-10, R-2-15, Another important aspect of this study to consider is the
R-2-25, and R-3-25, were statistically similar, but P-2-10, R-2-15, repeatability, or variability, of the results. To this end, it is evident
 B.J. Putman, A.I. Neptune / Construction and Building Materials 25 (2011) 3480–3485 3485

from the error bars in Fig. 6 that rodding resulted in the greatest the cylinders consolidated using the Proctor hammer and drop-
variability. This is most likely due to the presence of holes left by ping. A major cause of the increased variability was the creation
the rod during the rodding procedure. Due to the nature of pervi- of rod holes in the cylinders.
ous concrete having very low slump, the rod holes do not get com-  Consolidation by means of the standard Proctor hammer pro-
pletely filled in when the rod is removed as seen in Fig. 7. These rod duced cylinders having porosity and density closest to that of
holes affect the porosity and density of a specimen and the occur- the pavement in all three projects. However, because five blows
rence varied based on the mix properties (slump), number of rep- of the hammer was only used on one project, it cannot be con-
etitions, and number of layers. The cylinders prepared using the cluded whether five blows or 10 blows of the Proctor hammer
Proctor hammer and dropping methods were more consistent replicate the in situ pavement properties better.
(lower standard deviation) and obviously did not have any rod  While the cylinders had porosity values that were close to the
holes (Fig. 7). pavement porosity, the density of the cylinders were generally
greater than the pavement. However, the slabs, had porosity
3.3. Properties of slabs and density values that were relatively close to those of the
pavement. Based on these limited results, the 600  600 mm
The infiltration rates of the pervious concrete slabs cast from slabs having the same thickness as the pavement (150 mm)
Projects A and C are illustrated in Fig. 8. The average infiltration had properties that were most similar to the in-place pavement
rate of the pavement at the same location from where the slabs with the exception of infiltration rate. Additionally, the slabs
were cast is also included in each plot for comparison purposes. removed from the freshly placed pavement can also be mea-
After completing the infiltration testing, cores were cut from sured for in-place fresh concrete density, which would be
each slab and tested for porosity and density. These results are in- advantageous over the existing fresh pervious concrete unit
cluded in Fig. 9. As with Fig. 6, the average in-place pavement weight test procedure (ASTM C1688) that may not represent
porosity and density value are included in the graphs (dashed the in-place properties because of the variability in placement
lines) along with the standard deviation (dotted lines) for compar- procedures for different contractors.
ison purposes.
As with the cylinders, a statistical analysis was conducted to
determine statistically significant differences between the differ- Acknowledgements
ent sized slabs (a = 0.05). For Project A, the 450 mm and 600 mm
slabs had statistically similar porosity and density values and the The authors wish to acknowledge the City of Aiken for the
density and porosity of the 600 mm slab was closest to the in-place financial support of this research. Additionally, the cooperation of
pavement. For Project C, the porosity of all of the slabs were statis- TT&S Contracting, the contractor who constructed the pavements
tically similar and the 300 mm and 600 mm slabs had similar den- studied in this project was instrumental in this project.
sity. Additionally, the 300 mm slab had a porosity and density that
was most similar to the in-place pavement for Project C. References

[1] Tennis PD, Leming ML, Akers DJ. Pervious concrete pavements. Technical
4. Conclusions report, EB302.02. Portland Cement Association, Skokie, Illinois, and National
Ready Mixed Concrete Association, Silver Spring, Maryland; 2004.
Based on the results of this limited study that was developed to [2] Schaefer VR, Wang K, Suleiman MT, Kevern J. Mix design development for
pervious concrete in cold climates. Technical report. National Concrete
potentially determine the most appropriate method(s) that can be Pavement Technology Center, Iowa State Univ., Ames, Iowa; 2006.
used to prepare pervious concrete test specimens that most closely [3] Beeldens A, Van Gemert D, Caestecker C. Porous concrete: laboratory versus
resemble the properties of in-place pervious concrete pavements, field experience. In: Proc 9th int symp concrete roads, Istanbul, Turkey; 2003.
[4] Bury MA, Mawby CA, Fisher D. Making pervious concrete placement easy:
the following conclusions can be made.
using a novel admixture system. Concrete in Focus Fall; 2006. p. 55–9.
[5] Neithalath N, Weiss J, Olek J. Characterizing enhanced porosity concrete using
 The infiltration rate of pervious concrete pavement can be electrical impedance to predict acoustic and hydraulic performance. Cem
highly variable for a number of reasons including mix variabil- Concr Res 2006;36:2074–85.
[6] Haselbach L. Pervious concrete and mitigation of the urban heat island effect.
ity, mix delivery timing (delays and queues), placement prac- In: Proc 88th annual meeting transportation res board; 2009.
tices, and weather. However, if the areas of low (to no) [7] Kevern JT, Schaefer VR, Wang K. Temperature behavior of a pervious concrete
infiltration are localized and surrounded by pavement having system. In: Proc 88th annual meeting transportation res board; 2009.
[8] Offenburg M. Producing pervious pavements. Concrete International; March
infiltration, then stormwater will be able to infiltrate the pave- 2005.
ment into the underlying stone infiltration bed. [9] Rizvi R, Tighe SL, Henderson V, Norris J. Laboratory sample preparation
 There is a strong correlation between the density and porosity techniques for pervious concrete. In: Proc 88th annual meeting transportation
res board; 2009.
of pervious concrete mixtures. Additionally, there is a good cor- [10] Mahboub KC, Canler J, Rathbone R, Robl T, Davis B. Pervious concrete:
relation between pavement infiltration rate and both density compaction and aggregate gradation. ACI Mater J 2009;106:523–8.
and porosity. This indicates that the control of the pavement [11] ASTM. Annual book of ASTM standards. ASTM International; 2008.
[12] Neptune AI. Investigation of the effects of aggregate properties and gradation
density, and in turn porosity, is important during placement on pervious concrete mixtures. MS Thesis, Clemson University, Clemson, SC;
to ensure desired infiltration performance and pavement 2008.
functionality from a stormwater management perspective. [13] Putman BJ. Field performance of porous pavements in South Carolina. In: Proc
of the 2010 South Carolina water resources conference; 2010.
 Consolidation of pervious concrete cylinders by rodding
produced cylinders that had a greater degree of variability than