Astm

Copyright by ASTM Int'l (all rights reserved); Fri Aug 23 10:02:14 EDT 2019
Downloaded/printed by
(UFCG) Universidade Federal de Campina Grande ((UFCG) Universidade Federal de Campina Grande) pursuant to License Agreement. No further reproductions authorized.
Selected Technical Papers STP1551
Pervious Concrete
Editors:
Heather J. Brown
Matthew Offenberg
ASTM International
100 Barr Harbor Drive
PO Box C700
West Conshohocken, PA 19438-2959
Printed in the U.S.A.
ASTM Stock #: STP1551
Library of Congress Cataloging-in-Publication Data
ISBN: 978-0-8031-7537-2
This publication has been registered with the Library of Congress.
Library of Congress Control Number: 2012043986
Copyright © 2012 ASTM INTERNATIONAL, West Conshohocken, PA. All rights reserved.
This material may not be reproduced or copied, in whole or in part, in any printed, mechanical,
electronic, film, or other distribution and storage media, without the written consent of the
publisher.
Photocopy Rights
Authorization to photocopy items for internal, personal, or educational classroom use, or the
internal, personal, or educational classroom use of specific clients, is granted by ASTM
International provided that the appropriate fee is paid to ASTM International, 100 Barr Harbor
Drive, P.O. Box C700, West Conshohocken, PA 19428-2959, Tel: 610-832-9634; online:
http://www.astm.org/copyright.
The Society is not responsible, as a body, for the statements and opinions expressed in this
publication. ASTM International does not endorse any products represented in this publication.
Peer Review Policy
Each paper published in this volume was evaluated by two peer reviewers and at least one editor.
The authors addressed all of the reviewers’ comments to the satisfaction of both the technical
editor(s) and the ASTM International Committee on Publications.
The quality of the papers in this publication reflects not only the obvious efforts of the authors
and the technical editor(s), but also the work of the peer reviewers. In keeping with long-standing
publication practices, ASTM International maintains the anonymity of the peer reviewers. The
ASTM International Committee on Publications acknowledges with appreciation their dedication
and contribution of time and effort on behalf of ASTM International.
Citation of Papers
When citing papers from this publication, the appropriate citation includes the paper authors,
“paper title”, J. ASTM Intl., volume and number, Paper doi, ASTM International, West
Conshohocken, PA, Paper, year listed in the footnote of the paper. A citation is provided as a
footnote on page one of each paper.
Printed in Bay Shore, NY

November, 2012
Foreword
THIS COMPILATION OF Selected Technical Papers, STP1551, on Pervious
Concrete, contains peer-reviewed papers that were presented at a symposium
held December 4, 2011 in Tampa, FL, USA. The symposium was sponsored by
ASTM International Committee C09 on Concrete and Concrete Aggregates and
C09.49 Pervious Concrete.
The Symposium Co-Chairpersons and STP Editors are Heather J. Brown,
MTSU/Concrete Industry Mgmt., Murfreesboro, TN, USA and Matthew Offenberg,
W. R. Grace, Canton, GA, USA.
Contents
Overview ............................................................ vii
Validation of the Performance of Pervious Concrete in a Field Application with Finite

Element Analysis
M. A. Alam, L. Haselbach, and W. F. Cofer .................................. 1
Performance of Pervious Portland Cement Concrete by Field and Laboratory Testing,
Including Void Structure, Unit Weight, Compressive and Flexural Strength
W. B. Denison, Jr. .................................................... 17
Impact of Pervious Concrete Porosity on Permeability by 3D Image Analysis
S. Meulenyzer, E. Stora, and F. Perez ...................................... 27
Variability of Fresh and Hardened Voids of Pervious Concrete
L. K. Crouch, J. P. Hendrix, A. Sparkman, and D. Badoe ........................ 52
The Development, Implementation and Use of ASTM C1701 Field Infiltration of In
Place Pervious Concrete
H. J. Brown and A. Sparkman ........................................... 69
Development of a New Test Method for Assessing the Potential Raveling Resistance
of Pervious Concrete
M. Offenberg ....................................................... 80
Potential Application of ASTM C1701 for Evaluating Surface Infiltration of Permeable
Interlocking Concrete Pavements
D. R. Smith, K. Earley, and J. M. Lia ....................................... 97
Overview
The symposium that was held discussed the importance of having testing stand-
ards for a segment of the industry that has been placing pervious concrete for
over 30 years in the United States. The community of industry professionals
now engaged in specifying, designing, testing and installing pervious concrete is
so large that appropriate standards are paramount for the industry to keep its
momentum. Currently, four standards are approved with several more concepts
being researched. These activities will impact the use/acceptance of pervious
concrete going forward and the confidence that the specifying community has in
the material.
This symposium provided a forum for presenting data collected on pervious
projects relating to fresh concrete properties, hardened properties, durability,
permeability and mix design alternatives. Topics that were presented contained,
but were not limited to:
• Surface durability,
• Use and intention of C1701 (ASTM Standard Test Method for Infiltration
of In Place Pervious Concrete) for field permeability,
• Density and voids of freshly delivered material,
• Use of admixtures to improve pervious attributes, and
• Correlations of field cores to C1688 (ASTM Standard Test Method for
Density and Void Content of Freshly Mixed Pervious Concrete).
vii
Pervious Concrete
STP 1551, 2012
Available online at www.astm.org
DOI:10.1520/STP104553
Md. Ashraful Alam,1 Liv Haselbach,2 and William F. Cofer2
Validation of the Performance of Pervious

Concrete in a Field Application with Finite
Element Analysis
REFERENCE: Alam, Md. Ashraful, Haselbach, Liv, and Cofer, William F.,
“Validation of the Performance of Pervious Concrete in a Field Application
with Finite Element Analysis,” Pervious Concrete on December 4, 2011 in
Tampa, FL; STP 1551, H. J. Brown and M. Offenberg, Editors, pp. 1–16,
doi:10.1520/STP104553, ASTM International, West Conshohocken, PA
2012.
ABSTRACT: Pervious concrete is a paving material that has a number of
stormwater and other environmental benefits. Most current applications of
pervious concrete are in residential streets, parking lots, driveways, and side-
walks, and it is being considered for shoulders and more high volume appli-
cations. Characterizations of stress distribution and deflection patterns in
pervious concrete systems may be useful parameters in the structural design
of these high volume uses. Pervious concrete panels with tandem axle dual
wheel loads were analyzed using finite element analysis. The wheel position
was considered in the corner, center, and edge of the pavement. The critical
stresses obtained from the analyses were compared against experimental
tensile and compressive strengths obtained from samples from a field appli-
cation for various pervious concrete layer thicknesses, and additional experi-
mental data. It was found that pervious concrete panels of sufficient
thickness have adequate strength to support the applied wheel loads. To
compare the long term performance when subjected to cyclic loading, the
critical tensile stresses for various pavement thicknesses were compared
with pavement condition index (PCI) rating data obtained from a field applica-
tion reflecting pavement performance of approximately 131,000 cycles of an
80 kN single-axle load. For this particular field application, it was found that,
for cyclic loading, the required thickness of the pervious concrete layer was
Manuscript received November 17, 2011; accepted for publication January 27, 2012; published
online April 2012.
1
Dept. of Civil and Environmental Engineering, Washington State Univ. P.O. Box 642910,
Pullman, WA 99164-2910 (Corresponding author), e-mail: ashraf@wsu.edu
2
Dept. of Civil and Environmental Engineering, Washington State Univ. P.O. Box 642910,
Pullman, WA 99164-2910.
Copyright V
C 2012 by ASTM International, 100 Barr Harbor Drive, PO Box C700, West Conshohocken, PA
19428-2959.
1
2 STP 1551 ON PERVIOUS CONCRETE
approximately 40 %–80 % higher compared with that for the static loading
condition.
KEYWORDS: finite element modeling, pervious concrete, field validation,
critical stress, PCI rating
Introduction
Pervious concrete pavement is regarded as a stormwater best management
practice and its use can limit water pollutants entering surface water, maintain
ground water levels, decrease road noise, and increase driver safety by reduc-
ing glaring and hydroplaning [1,2]. Its use may also aid in obtaining several
credit points in the United States Green Building Council (USGBC) Green
Building Rating System [3].
The use of pervious concrete in the construction industry is not new, being
first applied in residential walls in Europe in 1852 [2,4]. It has been used for
more than 20 years in the United States [5,6]. Pervious concrete has been typi-
cally used for residential streets, driveways, sidewalks, and parking lots, and it
is now being considered for additional applications such as highway shoulders
and high volume roadways. Some research has been performed to evaluate ma-
terial properties for flexural strength and to quantify the compressive stress-
strain relationship [7,8], aggregate size effects on strength [9], permeability
[10], freeze-thaw behavior [11], and porosity distribution [12]. There have also
been a few studies on pervious concrete pavement performance for compres-
sive behavior [8], visual inspection [13], and tire-pavement noise for overlays
[14].
Little information is available on finite element analysis (FEA) for struc-
tural performance evaluation of pervious concrete. FEA is a widely accepted
approach with respect to roads and pavement for identification of two-
dimensional and three-dimensional static and dynamic stress demand, deforma-
tions, pavement soil interactions, and various other pavement responses [15].
FEA is commonly used for rigid concrete pavement and flexible asphalt pave-
ment analyses and design, and recently it has been applied to porous asphalt
pavement. A three-dimensional micro-structural model has been used to repre-
sent pervious concrete with respect to percolation and transport characteristics.
This micro-structural model has been developed from three-dimensional digital
images through the identification of aggregate, cement paste, and voids in the
pervious concrete [16]. However, a model having such fine detail would be
impractical for the structural simulation of an entire pavement panel. Recently,
the finite element method has been applied to evaluate the unique performance
characteristics of pervious concrete in pavement by including its vertical poros-
ity distribution with three layers to account for the corresponding variation of
stiffness properties [17]. That finite element model was validated through a
convergence study and comparison with analytical theory (i.e., critical tensile
stress and deflection) for traditional concrete pavement for typical 80 kN single
ALAM ET AL., doi:10.1520/STP104553 3
axle dual wheel load applications, but it requires further validation with addi-
tional field and load applications.
The objectives of this study are as follows:
• to compare the critical stresses obtained from FEA for pervious concrete
with field application and laboratory material strength data,

• to relate the critical stresses and associated strengths with field applica-
tion performance data, and

• to compare the model with analytical theories for stresses and deflec-
tions for traditional concrete pavement for the field applied 200 kN tan-
dem axle dual wheel load. (The field application data were obtained
from a site in Oregon where a pervious concrete drive had been installed
to evaluate its structural performance under heavy loads for several
years [18].)
The comparisons will provide initial guidance for the structural design of
pervious concrete and its applicability for highway use. It should be noted that
fatigue and raveling in pervious concrete pavement is yet to be included in the
finite element model due to the current lack of theoretical model and experi-
mental data.
Experimental Data from Previous Studies

In previous studies, samples were taken from pervious concrete panels in the
main driveway at a concrete mixing plant in Oregon, which had been in use for
several years. The pavement condition index (PCI) of the panels, various mate-
rial characteristics taken from specimens when the panels were removed, and
material characteristics of laboratory prepared pervious concrete specimens
were analyzed in those studies [18,19]. In the driveway, there were a number
of panels with different pervious concrete pavement layer (top) thicknesses and
various other material properties. The subbase layer thicknesses at the site var-
ied from 100 mm to 250 mm, but there were no information relating the thick-
ness of the subbase layer to each panel. Compressive strength and flexural
strength data obtained for specimens extracted from the panels are listed in Ta-
ble 1. The panels on the egress side of the drive received full concrete truck
loads while the ingress side was for returning empty trucks, as noted with Yes
or No designations, respectively, for the Truck Loading Condition in Table 1.
Manifests were kept at the plant with respect to the loads and frequency of
these trucks and that information was used to estimate equivalent single axle
loads (ESALs) experienced by the panels over the period of study [18].
Finite Element Modeling of the Field Site

Pavement systems of different thicknesses were modeled using FEA to com-
pare predictions of critical flexural and compressive stress of the panels with
TABLE 1—Material properties for samples extracted from the Evolution Paving Site [18].
Depth of pervious Flexural Compressive

concrete layer, Truck loading Unit weight, strength, strength,
mm condition kg/m3 MPa MPa PCI rating
100 Yes 1858 – – 8
125 Yes 1792 1.42 15.58 77
125 Yes 1808 1.90 19.58 8
150 No 1920 1.79 19.86 8
175 Yes 1872 1.96) 20.27 86
188 No 1952 2.05 24.27 –
200 No 1888 2.27 19.99 –
200 Yes 1984 2.81 23.92 50
250 Yes 1778 – – 87
the experimental flexural strength and to relate the predicted demand to the
pavement conditions observed in a previous study [19]. Deflections in the
pavement were also summarized to compare with previous theory and tradi-
tional analysis of traditional concrete pavements. Because there was no panel
specific information for the subbase layer thickness, in the analyses it was mod-
eled for both 250 mm and 100 mm values.
The trucks used in the model have dual wheel tandem axles similar to those
traversing the drives at the site. The critical tensile stress typically occurs for
wheels at or near the edge of pavement, in the middle of two transverse joints,
while critical deflection in the pavement system usually occurs for wheel place-
ment at the corner of a pavement panel [20,21]. In this research, pervious con-
crete panels were modeled for both of these loading conditions and also for
wheel placement at the middle of the panel. However, it was reported that
wheel placement at the center of the pavement was not critical for deflections
nor for stresses for 80 kN single axle dual wheel load [17].
The pervious concrete panel was modeled as three layers; the top quarter,
the middle half, and the bottom quarter, with different moduli of elasticity for
each layer. Each of these layers is assumed to be perfectly bonded at its interfa-
ces to the other layers. This layered approach was done to more accurately rep-
resent the effect on material stiffness of the vertical porosity distributions that
are typically found in pervious concrete [17], which has been shown to be
directly related to its strength and stiffness [12]. Figure 1 depicts the three-
dimensional pervious concrete pavement system used in the FEA, modeled
with the ADINA software package [22].
Material Properties Considered in the FEA
Pervious Concrete—The moduli of elasticity for the pervious concrete

layers used in the finite element modeling were estimated from equations
ALAM ET AL., doi:10.1520/STP104553 5
FIG. 1—Three-dimensional representation of the pervious concrete system for

FE modeling.
derived in a previous study based on varying moduli, porosities, and compres-

sive strengths of cylinders tested on samples made in the laboratory, and on
flexural strength, compressive strength and porosity data from the pervious
concrete specimens extracted from the field placement [18]. In the first study,
cores extracted from the field had average values of unit weight of 1890 kg/m3,
flexural strength of 2.03 MPa, compressive strength of 21.3 MPa, and porosity
of 21 %. The relationships between flexural strength and compressive strength,
and between flexural strength and porosity were expressed in Eqs 1 and 2,
respectively [18]
qffiffiffiffi
MOR ¼ 0:44 fc0 (1)
pffiffiffi
MOR ¼ 7:62 P þ 5:52 (2)
where:
MOR ¼ flexural strength, MPa,
fc0 ¼ compressive strength, MPa, and
P ¼ porosity, %.
In that first study [18], the following relationships were derived as shown
in Eqs 3 and 4
fc0 ¼ 114P þ 43:8 (3)
pffiffiffiffi
E ¼ 50:6 106 w1:5
pc fc0 (4)
where:
fc0 ¼ compressive strength, MPa,
P ¼ porosity, %,
E ¼ modulus of elasticity, GPa, and
wpc ¼ unit weight of pervious concrete, kg/m3.
However, the porosities of these samples were higher than those in their
field study and the second study provided experimental moduli of elasticity
with porosities that were similar to this field study [23]. Thus, the modulus of
elasticity equations from this second study were substituted, resulting in Eq 5
0 1=2
E; s ¼ 42:2 106 w1:5
pc exp 2:2 ln fc ; s 3:6 (5)
where:
E; s ¼ modulus of elasticity, GPa,
wpc ¼ unit weight of pervious concrete, kg/m3, and
fc0 ; s ¼ compressive strength of pervious concrete at the site, MPa.
The compressive strength used in Eq 5 can be calculated from Eqs 1 and 2
for known porosity. However, pervious concrete placements typically have a
vertical porosity distribution due to surface compaction installation practices
with the lowest porosity on the top. A simplified set of porosity distribution
equations for the top quarter, the middle half, and the bottom quarter of the per-
vious concrete layer were developed in a previous study and applied herein
[12]. The higher porosity in the lower section is important for determining the
flexural strength limitations of a slab with an applied surface point load. These
can also be used at intermediate layers in FEA to more closely model internal
forces, stresses, and strains. The mean porosity was used for the middle, and
the porosity values for the top and bottom quarters were calculated from Eqs 6
and 7 and then applied to determine the varying moduli of elasticity within the
pervious concrete layer [12]
Ptop ¼ 1:07Pmean 7 (6)
Pbottom ¼ 0:93Pmean þ 7 (7)
where:
Pmean ¼ known mean porosity, %,
Ptop ¼ top quarter porosity, %, and
Pbottom ¼ bottom quarter porosity, %.
Thus, the moduli of elasticity of pervious concrete for the three different
layers in the finite element modeling were calculated for the porosity and unit
weight of the samples at the site from Eqs 1, 2, 5, 6 and 7.
ALAM ET AL., doi:10.1520/STP104553 7
Subbase and Subgrade—The modulus of elasticity and Poisson’s ratio of

the subbase were considered to be 0.14 GPa and 0.4, respectively, based on a
previous test for base materials under traffic loads [24]. The moduli of elastic-
ity for the subgrade vary for different types of soil underneath. For this study,
the modulus of elasticity and Poisson’s ratio for the subgrade were estimated to
be 0.035 GPa and 0.4, respectively, assuming soft clay soil [25].
Wheel Load
The test site was a service road into and out of a concrete mixing plant. The
entering trucks were typically empty (133 kN) and the trucks were loaded
(289 kN) when exiting the plant. The tandem axle dual wheel load was
assumed to be 200 kN. There were other small vehicles entering and exiting,
but their loading was considered to be insignificant. With two booster axles,
the trucks used in transporting the concrete had five axles. The booster axles
were used when the truck was full and typically raised when the truck was
empty [18]. Thus, the applied wheel load in the pavement was 200 kN that is
equivalent to 690 kPa tire pressure for tandem axle dual wheel.
Pavement Model and Meshing

The meshing pattern and wheel position for the three different loading condi-
tions are shown in Fig. 2. The mesh is refined around the loading zone to get
detailed stress results at those points of interest. The coarser meshes are in the
areas where the stress results have minimum significance in the design of pave-
ment. Because no dowel bars were used in the longitudinal and transverse
joints, each pervious concrete panel in the pavement system could be regarded
as an individual panel. Eight-node 3-D solid elements were used for modeling
the pervious concrete, subbase, and subgrade in the pavement system.
As a result of symmetry, only half of a panel was modeled for the edge
loading and center loading cases. The subgrade layer was modeled with a depth
of 2700 mm and it was also extended laterally to the same extent. This depth
and the side extensions were chosen because it was found through iterative
analyses that additional subgrade depths had little impact on the stresses in the
pavement panel. In a similar two-dimensional finite element pavement analysis
[26], it was recommended that the total depth of the pavement system be
2250 mm. The length and width of the pervious concrete panel used in the
analyses was 6.0 m and 4.5 m, respectively. For the outside edge of the pave-
ment, the subbase layer was extended 300 mm as is typically done in practice.
FEA Results and Discussion

Analyses were performed considering both minimum and maximum subbase
layer depths found in the field for all loading conditions. Maximum deflection,
FIG. 2—Mesh pattern and wheel load position for (a) edge loading; (b) corner
loading; and (c) center loading.
ALAM ET AL., doi:10.1520/STP104553 9
FIG. 3—Maximum (a) deflection; (b) critical tensile stress; and (c) critical
compressive stress for the three different loading types based on FE analyses.
critical tensile stress, and critical compressive stress in the panel for the three
different loading conditions in the FEA are portrayed in Fig. 3 for a 250 mm
depth subbase layer and a 200 mm depth of pervious concrete. Although it was
assumed to have a variation of stiffness through its thickness, pervious concrete
pavement was shown to behave in a manner similar to that of traditional con-
crete pavement. Stress from the wheel load positioned at the center of the panel
was shown to not be critical for the pavement design for the higher tire load
application. It should be noted that, in all the analyses, the weight of the pave-
ment itself was considered in addition to the wheel load.
In traditional concrete pavement design, compressive stress is not typically
the controlling factor. However, the compressive strength of pervious concrete
could vary greatly depending on the coarseness of the mixture, aggregate size
distribution, porosity, and additives [1,7,27,28]. Thus, on the basis of Fig. 3(c)
for corner loading, it is important to consider the compressive strength in the
design of pervious concrete pavement if low strength mixes are used for appli-
cations with high loads. However, because compressive strengths tend to be
correlated with flexural strengths in unreinforced concrete panels, lower
strength pervious concrete mixes would rarely be considered for heavy loading
applications.
The maximum deflections for variable thicknesses of the pavement for the
two specific depths of subbase are shown in Fig. 4 for the case of the wheel
placement at the corner of the pavement. The pervious concrete pavement at
the field site was not investigated for deflection and thus the deflection obtained
from FE analysis cannot be compared with experimental results. For both the
250 mm and 100 mm subbase layers, Fig. 4 shows that the deflection decreases
nonlinearly with increased pavement thickness. As expected, the effect is
FIG. 4—Maximum deflections for 250 mm subbase and 100 mm subbase with
corner loading as determined with FEA.
increasingly profound as the subbase thickness and, hence, its stiffness,

decreases.
Comparison Between Experimental Strength and Finite Element Stress Output

The critical flexural stress demand for various pervious concrete pavement
thicknesses as modeled by FEA are plotted in Fig. 5 for both subbase thick-
nesses. Experimentally determined material flexural strengths are also indi-
cated, where variation with depth for flexural strength may possibly be
connected to some additional compaction near the bottom with added mass on
top for thickened panels. When critical stresses from loading in a pavement
section reach these levels, then failure might occur. Points with demand above
the material’s flexural strength indicate the likelihood of failure. Note that both
the flexural strengths as measured for the materials, and the critical tensile
stresses as evaluated, were all in the lower quarter of the pervious concrete
layer. It could be surmised from Fig. 5 that, except for the 125 mm thick pan-
els, all of the panels in the pervious concrete drive have higher flexural strength
than the stress from loading for both subbase depth considerations.
The design of pavements also depends on cyclic loading, or load repeti-
tions. In traditional pavements, the number of load repetitions until failure can
be calculated from known flexural stresses and moduli following various em-
pirical formulae. In the case of pervious concrete, these empirical relationships
have not yet been established.
ALAM ET AL., doi:10.1520/STP104553 11
FIG. 5—Comparison of tensile strength from experimental investigation and

tensile stress from FEA.
Figure 6 depicts the material’s compressive strength from experiments and

the compressive stresses from FEA. The critical compressive stress demand is
significantly less than the field measured strength of the material for all the
pavement layer thicknesses analyzed and; thus, the likelihood of compressive
failure is small for these pavement thicknesses and associated subbase depths.
The variations in material strength related to thickness are assumed to be statis-
tically insignificant.
Comparison with Pavement Condition Index

Distress surveys were performed for the field installation at the field site and
the associated Pavement Condition Indices (PCI) were calculated in previous
studies for the panels after five to six years of use [18,19]. In this previous
study, the ESAL factor for the concrete trucks was found to be 2.1, which indi-
cates that the pavement at the site was subjected to approximately 131,000
ESALs from the fully loaded trucks. Distress refers to surface raveling, crack-
ing, and any other type of visible damage on the surface of the pavement and
the PCI is the numerical value that represents the existing surface condition.
The PCI rating ranges between 0–100, with 0 representing the worst condition
of the pavement and 100 the best possible condition. ASTM has different ver-
sions for PCI rating. The PCI values reported in the aforementioned study [18]
are listed in Table 1 and were based on the 2007 PCI rating, which is similar to
the 2009 update [29,30]. There are a number of distress types listed in the PCI
standard for concrete pavements, but not all of these distresses were observed
FIG. 6—Comparison of material compressive strengths from experimental

investigation and compressive stresses from FE analysis.
in the pervious concrete pavement at the site. Also, due to the uniqueness of
pervious concrete pavements, many of the distress types were ignored. For
example, the presence of polished aggregate in the pervious concrete pavement
surface was ignored as a distress type, because pervious concrete has a very
thin layer of binder on the aggregate making polished aggregate a frequent
occurrence. The most notable distress types observed in the pavement at the
site were linear cracking, scaling, shrinkage cracks, corner breaks, and divided
slabs in the outside edge of the pavement.
The PCI ratings for different thicknesses of the pervious concrete panels at
the site in the exiting lanes (most heavily loaded) are plotted in Fig. 7. The PCI
points have been connected for the lowest values for each thickness, indicative
of worst case pavement performance. This worst case performance line might
be a first estimate of minimum performance for cyclic loading. It should be
emphasized that there were several panels of lower thicknesses which showed
very little distress despite the many years of testing, and these are also indi-
cated on Fig. 7 above the minimum performance line.
Discussion
The depth corresponding to the intersecting point between an imaginary
strength curve and the stress curve in Fig. 5 may be defined as the critical thick-
ness for static loading. From Fig. 5, for 100 mm thickness in the subbase layer,
ALAM ET AL., doi:10.1520/STP104553 13
FIG. 7—Minimum performance line and additional thickness values with

respect to PCI ratings.
the required thickness of the pervious concrete layer is 143 mm, and for
250 mm thickness in the subbase layer, the thickness of the pervious concrete
layer is 134 mm. The PCI rating curve provides a measure of the ability of the
pavement to withstand long-term traffic loads because the PCI rating represents
62,400 cycles of full truck loads or 131,000 ESALs. Thus, the depths corre-
sponding to different PCI ratings on the minimum performance line in Fig. 7
may be defined as the depths required for long-term cyclic loading for these
pavement conditions. According to the ASTM standard [30], a PCI rating
between 55 and 70 indicates fair pavement condition, while ratings between 70
and 85 are satisfactory and those between 85 and 100 are good. From Fig. 7,
for a fair, satisfactory, or good condition pavement, the minimum required
thickness in the pervious concrete layer is 211 mm, 231 mm, and 251 mm,
respectively. As can be seen from Figs. 5 and 7, increased thickness of the
pavement layer leads to increased longevity with respect to static and cyclic
loading, respectively. For the minimum depth subbase (100 mm) layer, the per-
centage increases required for thickness of the pervious concrete layer for
cyclic loading compared to the thickness required for static loading are 47.5 %,
61.5 %, and 75.5 % for fair, satisfactory, and good performance pavement,
respectively.
Conclusion
Pervious concrete pavement was analyzed for static loading with finite element
methods that included vertical porosity variations in addition to other material
characteristics. The analyses were performed for four different thicknesses in
the pervious concrete layer and two different subbase depths. Deflections and
stresses were obtained from the analysis output and compared with laboratory
and field data, which included tensile strengths, compressive strengths, and
PCI ratings. A summary of the findings includes the following:
• Pervious concrete pavement behavior follows that of classical analytical
theory for traditional concrete pavement, with maximum deflections for

wheel positions at the corner of the pavement and maximum tensile
stresses for edge loading, even with a vertical porosity distribution and
associated stiffness.
• A comparison of stress demand with material strength data obtained
from experiments with respect to static loading, and also PCI ratings af-
ter long-term loading, both indicate that pervious concrete of sufficient
thickness, and with adequate subbase, can be an alternative material in
the design of highway pavements.
• Preliminary results indicate that, in order to calculate the required thick-
ness for cyclic loading, the pavement thickness required for static load-
ing might need to be multiplied by a factor of safety of two (2).
It should be noted that each panel considered in the FEA was an isolated
one to represent the field conditions. However, for some field applications, the
bottom portions of the pervious concrete panel may remain connected in the
longitudinal direction of the pavement. This should be analyzed in future stud-
ies as should other wheel types. Different combinations of material properties
and different layer thicknesses; in addition to variability in subgrade compac-
tion, should also be analyzed to prepare further guidance for the structural
design of highways with pervious concrete. Finally, although the PCI rating
seems to indicate that pervious concrete pavement is able to withstand many
cycles of loading, further research on material performance for cyclic loading
is required to quantify its fatigue properties.
Acknowledgments
The writers gratefully acknowledge the financial support from Transportation
Northwest. The writers also appreciate the assistance from Evolution Paving,
and Will Goede, and Michelle Boyer in evaluating the performance of pervious
concrete.
References
[1] Tennis, P., Leming, M., and Akers, D., Pervious Concrete Pavements,
Portland Cement Association, Skokie, IL, 2004.
[2] ACI Committee 522, “Pervious concrete,” ACI 522R-06, American Con-
crete Institute, Farmington, MI, 2006.
ALAM ET AL., doi:10.1520/STP104553 15
[3] Haselbach, L., The Engineering Guide to LEED-New Construction,

McGraw-Hill, New York, 2008.
[4] Saber, A., “Pervious Concrete Pavements for Parking Lots and Low Vol-
ume Roads,” ASCE J. La Sect., Vol. 17, No. 2, 2009, pp. 15–17.
[5] Huffman, D., “Green Transportation Infrastructure,” Statement to the
Committee on Science and Technology and Subcommittee on Technol-
ogy and Innovation, United States House of Representatives, May 10,
2007, Congr. Rec., Vol. 110-27, pp. 43–51.
[6] Henderson, V., Tighe, S., and Norris, J., “Pervious Concrete Pavement:
An Integrated Laboratory and Field Study,” Transp. Res. Rec., Vol. 2113,
2009, pp. 13–21.
[7] Yang J., and Jiang, G., “Experimental Study on Properties of Pervious
Concrete Pavement Materials,” Cem. Concr. Res., Vol. 33, 2003, pp.
381–386.
[8] Deo O., and Neithalath, N., “Compressive Behavior of Pervious Con-
cretes and a Quantification of the Influence of Random Pore Structures
Features,” Mater. Sci. Eng., Vol. 528, 2010, pp. 402–412.
[9] Crouch, L. K., Pitt, J., and Hewitt, R., “Aggregate Effects on Pervious
Portland Cement Concrete Static Modulus of Elasticity,” J. Mater. Civ.
Eng., Vol. 19, No. 7, 2007, pp. 561–568.
[10] Neithalath, N., Sumanasooriya, M. S., and Deo, O., “Characterizing Pore
Volume, Sizes, and Connectivity in Pervious Concretes for Permeability
Prediction,” Mater. Charact., Vol. 61, 2010, pp. 802–813.
[11] Guthine, W. S., DeMille, C. B., and Eggett, D. L., “Effects of Soil Clog-
ging and Water Saturation on Freeze–Thaw Durability of Pervious Con-
crete,” Transp. Res. Rec., Vol. 2164, 2010, pp. 89–97.
[12] Haselbach, L. M. and Freeman, R. M., “Vertical Porosity Distributions in
Pervious Concrete Pavement,” ACI Mater. J., Vol. 103, No. 6, 2006, pp.
452–458.
[13] Delatte, N., Mrkajic, A., and Miller, D. I., “Field and Laboratory Evalua-
tion of Pervious Concrete Pavements,” Transp. Res. Rec., Vol. 2113,
2009, pp. 132–139.
[14] Schaefer, V. R., Kevern, J. T., Izevbekhai, B., Wang, K., Cutler, H. E.,
and Wiegand, P., “Construction and Performance of Pervious Concrete
Overlay at Minnesota Rd. Research Project,” Transp. Res. Rec., Vol.
2164, 2010, pp. 82–88.
[15] Mackerle, J., “Finite Element and Boundary Element Analysis of Bridges,
Roads and Pavements—A Bibliography (1994–1997),” Finite Elem. Anal.
Design, Vol. 29, 1998, pp. 65–73.
[16] Bentz, D. P., “Virtual Pervious Concrete: Microstructure, Percolation,
and Permeability,” ACI Mater. J., Vol. 105, No. 3, 2008, pp. 297–301.
[17] Alam, A., Haselbach L., and Cofer, W., “Three-Dimensional Finite Ele-
ment Modeling and Analysis of Pervious Concrete Pavement: Simplified
Vertical Porosity Distribution Approach,” Finite Elem. Anal. Design

(submitted).
[18] Goede, W. G., 2009, “Pervious Concrete: Investigation into Structural
Performance and Evaluation of the Applicability of Existing Thickness
Design Methods,” M.S. Thesis, Washington State University, Pullman,
WA.
[19] Goede W., and Haselbach, L., “Investigation into the Structural Perform-
ance of Pervious Concrete,” J. Transp. Eng., (in press).
[20] Huang, Y. H., Pavement Analysis and Design, Prentice Hall, Upper Sad-
dle River, NJ, 2004.
[21] Portland Cement Association, Thickness Design of Concrete Highway
and Street Pavement, Portland Cement Association, Skokie, IL, 1984.
[22] ADINA, ADINA—Finite Element Analysis Software, Version 8.7.2,
ADINA, Watertown, MA, 2010.
[23] Ghafoori N., and Dutta, S., “Pavement Thickness Design for No-Fines
Concrete Parking Lots,” J. Transp. Eng., Vol. 121, No. 6, 1995, pp.
476–484.
[24] Adu-Osei, A., Little, D. N., and Lytton, R. L., “Structural Characteristics
of Unbound Aggregate Bases to Meet AASHTO 2002 Design Require-
ments, Interim report,” Report No. ICAR/502-1, Aggregates Foundation
for Technology, Research, and Education, Arlington, VA, 2001.
[25] Coduto, D. P., Foundation Design Principles and Practices, Prentice-
Hall, Englewood Cliffs, NJ, 1994.
[26] Cho, Y., McCullough, B. F., and Weissmann, J., “Considerations on
Finite-Element Method Application in Pavement Structural Analysis,”
Transp. Res. Rec., Vol. 1539, 1996, pp. 96–101.
[27] Onstenk, E., Aguado, A., Eickschen, E., and Josa, A., “Laboratory Study
of Porous Concrete for Its Use as Top-Layer of Concrete Pavements,”
Fifth International Conference on Concrete Pavement Design and Reha-
bilitation, Vol. 2, Purdue University and School of Civil Engineering, 550
Stadium Mall Drive, West Lafayette, IN, 1993, pp. 125–139.
[28] Vassilikou, F., Kringos, N., Kotsovos, M., and Scarpas, A., “Application
of Pervious Concrete for Sustainable Pavements: A Micro-Mechanical
Investigation,” Highways; Pavements; Materials (CD-ROM), Transporta-
tion Research Board of the National Academies, Washington D.C., 2011.
[29] ASTM D 6433-07, 2007, “Standard Practice for Roads and Parking Lots
Pavement Condition Index Surveys,” Annual Book of ASTM Standards,
Vol. 04.03, ASTM International, West Conshohocken, PA.
[30] ASTM D 6433-09, 2009, “Standard Practice for Roads and Parking Lots
Pavement Condition Index Surveys,” Annual Book of ASTM Standards,
Pervious Concrete
STP 1551, 2012
DOI:10.1520/STP104561
William B. Denison, Jr.1
Performance of Pervious Portland Cement

Concrete by Field and Laboratory Testing,
Including Void Structure, Unit Weight,
Compressive and Flexural Strength
REFERENCE: Denison, William B., Jr., “Performance of Pervious Portland

Cement Concrete by Field and Laboratory Testing, Including Void Structure,
Unit Weight, Compressive and Flexural Strength,” Pervious Concrete on De-
cember 4, 2011 in Tampa, FL; STP 1551, H. J. Brown and M. Offenberg, Edi-
tors, pp. 17–26, doi:10.1520/STP104561, ASTM International, West
Conshohocken, PA 2012.
ABSTRACT: Titan Virginia Ready Mix (TVRM) sponsored a study involving
field and laboratory testing on numerous placement segments of the 4700
yd3 (8 acres) of pervious pavement supplied to the Prime Outlet Mall project
in Williamsburg, Virginia, 2008. The intent of the study was to gather a statis-
tically sufficient quantity of test data of plastic and hardened properties using
ASTM test standards from a project large in square footage to provide such
data. The TVRM approach involved representative sampling, plastic and
hardened density testing, void % determination, thickness measurements,
and compressive and flexural strength testing. The hardened test specimens
were obtained by coring and sawing from precast panels cast onsite during
pavement production. The test panels measuring 3 ft 3 ft 6 in. in depth
had supporting bottoms and was cast from composite samples obtained at
the project site. The test panels were cast and cured in the same manner as
the pervious pavement, one test panel for every 150 to 200 yd3. The test pan-
els were cured a minimum of 7 days to promote the necessary aggregate/
cement paste bond development as is referenced in ACI 522.1-08 [ACI 522-
1, “Specification for Pervious Concrete Pavement,” American Concrete Insti-
tute, Farmington Hills, MI, 2008]. Once the curing was complete the pervious
test panels were cushioned, and transported to the TVRM laboratory for fur-
ther curing and hardened strength and density testing. The TVRM objective
Manuscript received November 19, 2012; accepted for publication March 26, 2012; published
online October 2012.
1
Titan Virginia Ready Mix, LLC, 2125 Kimball Terrace, Norfolk, Virginia 23504,
e-mail: wdenison@titanamerica.com
Copyright V
19428-2959.
17
was to compile enough raw data with consistent test results to help designers
and specifiers in designing for future pervious projects.
KEYWORDS: pervious concrete, precast panels, curing, plastic and hard-
ened properties, ASTM
Introduction
Developments of testing standards for pervious concrete have long been desira-
ble in our industry. The Titan Virginia Ready Mix approach was influenced by
my involvement with the Shot Crete nozzelman certification program in ACI
506.3R-91 (Guide to Shotcreting) in 1997. Shotcreting was used widely by the
Dept. of Defense in the port of Hampton Roads, Virginia for dry dock wall
repairs. There was a high demand for certified Shot Crete nozzelman because
of the high volume of shotcreting work. Nozzelman were required to complete
thickness and sloughing test in 2 ft 2 ft 4 in. deep test panels. The test pan-
els simulated actual shooting direction and conditions. They were shaved and
cured with polyethylene sheeting immediately after shooting and intentionally
field cured so that strength development would more typically reflect field
performance. Cores and sawed beams were obtained from each test panel in
accordance with ASTM C42-04 [4] (Test Method for Obtaining and Testing
Drilled Cores and Sawed Beams of Concrete) at 3 days and were subjected to
moist curing in a fog room at 73.4 F 6 3 F and 100 % humidity until tested at
28 days.
Our methological approach is a modified version of the Shot Crete technol-
ogy from 1997. Similarities were incorporated into a testing program to evalu-
ate both plastic and hardened properties of pervious concrete.
Testing Program
Evaluation of Plastic Properties of Pervious Concrete
Sampling—ASTM C172-04 (Standard Practice for Sampling Freshly

Mixed Concrete), Paragraph 5.2.2, Sampling from a Paving Mixer.
Density (Unit Weight) of Concrete, lb/ft3—ASTM C29-03 [2] (Standard

Test Method for Bulk Density (“Unit Weight”) and Voids in Aggregate), Para-
graph 11, Jigging. ASTM C138-01 [6] (Standard Test Method for Density
(“Unit Weight”), Yield, and Air Content by the Volumetric Method):
D ¼ ðMc Mm Þ=Vm (1)
where
D ¼ density (unit weight) of concrete, lb/ft3,
Mc ¼ mass of measure filled with concrete, lb,
DENISON, doi:10.1520/STP104561 19
Mm ¼ mass of measure, lb, and

Vm ¼ volume of measure, ft3.
Air Void Percentage, %—ASTM C138-01 Standard Test Method for Den-
sity (Unit Weight), Yield, and Air Content (Gravimetric) of Concrete:
A ¼ ½ðT DÞ=T 100 (2)
where
A ¼ air content (percentage of voids) in the concrete, %,
T ¼ theoretical density of the concrete computed on an air free basis, lb/ft3,
and
D ¼ density (unit weight) of concrete, lb/ft3.
Evaluation of Hardened Properties of Pervious Concrete
Hardened Density, lb/ft3—ASTM C42-04 [4] (Test Method for Obtaining

and Testing Drilled Cores and Sawed Beams of Concrete) Paragraph 7.5,
Density.
Thickness, in.—ASTM C174-06 [7] (Test Method for Measuring Thickness

of Concrete Elements Using Drilled Concrete Cores).
Compressive Strength, psi—ASTM C39-05 [3] (Compressive Strength of

Cylindrical Concrete Specimens).
Flexural Strength, psi—ASTM C78-02 [5] (Flexural Strength of Concrete

Using Simple Beam with Third Point Loading). ASTM C42-04 (Test Method
for Obtaining and Testing Drilled Cores and Sawed Beams of Concrete),
Appendix XI—Sawed Beams for Flexural Testing.
Pervious Mix Proportioning

When proportioning a pervious mix design, it should be the primary goal of the
producer to incorporate their raw materials into a mixture that will meet the
void requirements when tested. The density and void content results will heav-
ily depend on the individual constituent built in the pervious design; cement
content, coarse aggregate content, water content, and maximum aggregate size.
The lack of fines gives pervious the void structure and permeability.
The TVRM pervious mix design utilized on the Prime Outlet Mall was
designed with a total void volume of 18.5 %, with enough water to produce a
pervious mixture with wet-metallic sheen. This allowed placement without
TABLE 1—TVRM pervious mix design proportions.
Material Plastic Mix Proportions (lb/yd3)a

Roanoke Cement I/II 600
3/8-in. Coarse aggregate (angular stone), SSD 2723
Water 174.9 (21 gal)
Entrained air voids, % 5.0
Entrapped air voids, % 13.5
Admixtures Type A, water-reducing
Hydration stabilizer
Air entraining agent
Design density, lb/ft3 Theoretical density (unit weight) ¼ 129.6 lb/ft3
Theoretical density (air free unit weight) ¼ 159.0 lb/ft3
a
Designed in accordance with the provisions set forth in ACI 522R-10.
causing the paste to flow from the aggregate. The actual weights are given in
Table 1.
Field Samples
Field sampling was performed in accordance with ASTM C172-04, Paragraph
5.2.2, Sampling from Paving Mixers. After discharge of the pervious concrete,
composite samples were obtained from at least five portions of the pile. The
first and final composite samples were obtained within 15 min as required in
ASTM C172-04. The composite samples were transported and remixed at the
field testing location and protected from evaporation during testing. Density
(unit weight) testing in accordance with ASTM C29-05 and ASTM C138-01
commenced within 5 min after obtaining the final portion of the composite
sample. The density testing was performed with a 25-ft3 measure, because the
results have been more repeatable and reproducible than with the 0.50 ft3 or
larger measures. Air void percentages were then calculated in the field by
gravimetric method per ASTM C138-01, Paragraph 7.5, Percentage of Air
Voids. The test results and averages are listed in Table 2.
During field testing, casting of the test panels occurred parallel to the pave-
ment. This was to assure that the pervious concrete was in place within a 20
min time exposure limitation as specified in ACI 522.1 [1]. Curing commenced
immediately after casting with 6 mil polyethylene sheeting.
Test Panels
Test panels measuring 3 ft 3 ft 6 in. in depth were cast with bottoms for
every 150 to 200 yd3 except for the first six placements, where test panels were
cast for each placement. The test panels simulated actual placement and con-
solidation methods as used in the pervious pavement. The test panels were cast
DENISON, doi:10.1520/STP104561 21
TABLE 2—Field and laboratory test results (plastic and hardened densities to air void percentages)
of TVRM’s pervious concrete.
Total Plastic Plastic Void Hardened Hardened Void

Panel Placement Volume Density Percentages Density Percentages
Number Date (yd3) (lb/ft3) (%) (lb/ft3) (%)
1 1/10/2008 84 125.1 21.3 130.2 18.1
2 1/11/2008 87 126.1 20.7 129.6 18.5
3 1/12/2008 90 127.0 20.1 128.5 19.2
4 1/14/2008 101 130.0 18.2 131.6 17.2
5 1/15/2008 102 128.1 19.4 130.5 17.9
6 1/16/2008 91 127.4 19.8 130.0 18.2
7 1/17/2008 49 127.6 19.9 128.9 18.9
1/18/2008 98 126.9 20.2 – –
8 2/16/2008 91 130.0 18.2 131.5 17.3
9 2/19/2008 91 128.1 19.4 129.8 18.4
2/20/2008 91 126.4 20.5 – –
10 2/21/2008 83 126.4 20.5 128.4 19.2
2/23/2008 91 125.8 20.9 – –
11 2/25/2008 84 124.1 21.9 126.0 20.8
2/26/2008 109 124.7 21.6 – –
12 2/27/2008 89 125.0 21.4 127.5 19.8
2/28/2008 90 126.4 20.5 – –
13 2/29/2008 98 125.6 21.0 126.0 20.8
3/1/2008 80 125.8 20.9 – –
14 3/3/2008 63 125.6 21.0 127.8 19.6
3/4/2008 112 125.3 21.2 – –
15 3/5/2008 107 126.1 20.7 130.0 18.2
3/6/2008 56 125.5 21.1 – –
16 3/10/2008 81 125.8 20.9 127.5 19.8
3/11/2008 42 125.3 21.2 – –
17 3/18/2008 110 125.5 21.1 128.7 19.1
3/20/2008 106 124.3 21.8 – –
18 3/21/2008 124 126.0 20.9 128.6 19.2
3/24/2008 101 125.3 20.8 – –
19 3/25/2008 211 126.4 20.5 131.5 17.3
20 3/26/2008 152 129.1 18.8 131.5 17.3
21 3/27/2008 224 130.0 18.2 131.0 17.6
22 3/28/2008 147 128.1 19.4 129.9 18.3
23 3/29/2008 163 126.9 20.2 128.6 19.2
24 3/31/2008 245 129.5 18.6 132.0 17.0
25 4/1/2008 320 127.4 19.9 129.4 18.6
26 4/2/2008 275 127.6 19.8 129.8 18.4
27 4/3/2008 120 127.0 20.1 130.4 18.0
4/4/2008 91 125.8 20.9 – –
TABLE 2—Continued
Total Plastic Plastic Void Hardened Hardened Void

Panel Placement Volume Density Percentages Density Percentages
Number Date (yd3) (lb/ft3) (%) (lb/ft3) (%)
28 4/7/2008 124 124.8 21.5 129.8 18.4
4/8/2008 5 No test – – –
29 5/13/2008 28 131.0 17.6 132.5 16.7
Average 126.7 20.30 129.6 18.52
next to the in place pervious pavement, where curing commenced immediately

after casting with polyethylene sheeting. The test panels were left undisturbed
for 7 days in the field to simulate the strength development more typical in the
field. On the seventh day of curing, the test panels were cushioned and trans-
ported to the TVRM laboratory in Norfolk, Virginia for final curing and further
hardened testing.
Laboratory Testing
Procedures Performed at the TVRM Laboratory, Norfolk, VA

Three cores were obtained from each test panel as they arrived in the laboratory
upon completion of the 7-day field curing. The cored test specimens were
obtained in accordance with ASTM C42-04 for the purpose of thickness meas-
urements. The cored specimens were measured for thickness in accordance
with ASTM C174-06 (Test Method for Measuring Thickness of Drilled Cores).
After thickness measurements, the cores were trimmed, weighed, and measured
for density (unit weight) in accordance with ASTM C42-04, Paragraph 7.5,
Density. The density test results and averages are listed in Table 2 (see Fig.
1(a) and 1(b)) and thickness measurements in Table 3 (see Fig. 1).
The test panels were then moist cured in a fog room at 73.5 F 6 3.5 F and
100 % humidity for the remaining curing period. Once curing was terminated,
which included initial and final moist curing time, the test panels were removed
from the fog room and covered with wet burlene to prevent excessive surface
drying. Three additional cores and two sawed beams were obtained in accord-
ance with ASTM C42-04 (Test Method for Obtaining and Testing Drilled Cores
and Sawed Beams of Concrete). The three cored test specimens were visually
examined, trimmed and capped in accordance with ASTM C617-03 [8] (Test
Method for Capping Cylindrical Concrete Specimens), submerged in a lime satu-
rated bath at 73.5 F 6 3.5 F along with the two sawed beams for 24 h prior to
testing. The three cores were tested for compressive strength in accordance with
ASTM C39 at 28 days and the two sawed beams were tested for flexural strength
DENISON, doi:10.1520/STP104561 23
TABLE 3—Laboratory test results (relationship between compressive, flexural strength, and thick-
ness) of TVRM’s pervious concrete.
Panel Placement Thickness Average Compressive Average Flexural

Number Date Measurements (in.) Strength, psi Strength, psi
1 1/10/2008 6.06 2380a 420b
2 1/11/2008 6.00 2780a 425b
3 1/12/2008 6.11 2690a 435b
4 1/14/2008 6.16 2850a 450b
5 1/15/2008 6.04 2720a 415b
6 1/16/2008 5.97 2760a 435b
7 1/17/2008 6.06 2650a 425b
8 2/16/2008 6.18 2900a 440b
9 2/19/2008 5.97 2820a 425b
10 2/21/2008 6.12 2750a 420b
11 2/25/2008 6.18 2450a 410b
12 2/27/2008 5.97 2580a 430b
13 2/29/2008 5.97 2690a 425b
14 3/3/2008 6.00 2870a 445b
15 3/5/2008 6.12 2790a 430b
16 3/10/2008 6.06 2640a 420b
17 3/18/2008 5.97 2490a 415b
18 3/21/2008 6.09 2660a 425b
19 3/25/2008 6.03 2860a 435b
20 3/26/2008 6.07 2870a 425b
21 3/27/2008 6.02 2790a 430b
22 3/28/2008 6.03 2690a 425b
23 3/29/2008 6.07 2910a 445b
24 3/31/2008 6.12 2980a 475b
25 4/1/2008 6.07 2580a 425b
26 4/2/2008 5.97 2640a 435b
27 4/3/2008 6.00 2840a 445b
28 4/7/2008 6.07 2710a 435b
29 5/13/2008 6.03 2890a 460b
Average 6.05 2730 435
a
Mean of three cores.
b
Mean of two beams.
in accordance with ASTM C78 at 28 days. The compressive and flexural strength
test results and averages are listed in Table 3 (see Fig. 1(c)).
Analysis of Results
The Titan Virginia Ready Mix testing approach used a unique and closely con-
trolled test panel program, along with plastic and hardened density testing and
FIG. 1—(a) This figure represents two graphs from Table 2 data. (b) This fig-
ure shows plastic versus hardened void percentages from Table 2. (c) This fig-
ure graphs flexural versus compressive strength from Table 3 data.
compressive and flexural strength testing. The TVRM testing methology

yielded consistent test values correlating to the pervious concrete and pave-
ment. The consistent data will help both the designer and specifiers to incorpo-
rate pervious pavements in areas that previously had been ignored or bypassed
because of unreliable or inaccurate test results. The test results in Tables 2
and 3 offer a satisfactory look into test data that correlates the pervious design
with actual plastic and hardened properties.
Both the plastic and hardened density is within the design tolerance of
65 pcf, and their corresponding air void percentage is within the allowable tol-
erance range of 65 % of air void design. The compressive and flexural strength
testing is still in court at this present time, but with our method of testing could
DENISON, doi:10.1520/STP104561 25
shed some light into consistent compressive and flexural strength numbers
required by civil engineers for pavement design.
Conclusion
There are many problems surrounding the measurement of quality in pervious
concrete pavements. The problems exist because the material properties and
their sensitivity to variations, which occur in consolidating and curing. The
placing, consolidating, and curing of pervious concrete is more sensitive than
conventional impervious concrete. The standard practices and test methods
typically utilized in the field, such as slump, air content, and fabrication of test
specimens are not applicable. If the resulting field and laboratory test methods
used by TVRM can be correlated to the desirable properties of both the mate-
rial and pavement, it would seem reasonable to use these test methods and
results to test for acceptance.
Acknowledgments
The author sincerely appreciates the technical assistance provided by Robert
Justice, Quality Assurance Technician for Titan Virginia Ready Mix, LLC
SOVA. The author also gratefully acknowledges the management support of
Dan Osborne, General Manager of Titan Virginia Ready Mix, LLC. Technical
support from Larry Necessary, Technical Service Manager, Roanoke Cement,
and Technical support from Scott Manning, TVRM Corporate Engineering.
Special thanks go out to Chris Cartwright, Department Head of Civil Engineer-
ing Technologies at Tidewater Community College for additional lab facilities
as needed. Disclaimer: The opinions, findings, and conclusions expressed here
are those of the author and not necessarily those of Titan Virginia Ready Mix,
LLC.
References
[1] ACI 522-1, “Specification for Pervious Concrete Pavement,” American

Concrete Institute, Farmington Hills, MI, 2008.
[2] ASTM C29-03, 2007, “Standard Test Method for Bulk Density (“Unit
Weight”) and Voids in Aggregate,” Annual Book of ASTM Standards,.
[3] ASTM C39/C39M-05, 2007, “Standard Test Method for Compressive
Strength of Cylindrical Concrete Specimens,” Annual Book of ASTM
Standards, Vol. 04.02, ASTM International, West Conshohocken, PA.
[4] ASTM C42/C42-04, 2007, “Standard Test Method for Obtaining
and Testing Drilled Cores and Sawed Beams of Concrete,” Annual Book
of ASTM Standards, Vol. 04.02, ASTM International, West Consho-

hocken, PA.
[5] ASTM C78-02, 2007, “Flexural Strength of Concrete (Using the Simple
Beam with Third Point Loading),” Annual Book of ASTM Standards, Vol.
04.02, ASTM International, West Conshohocken, PA.
[6] ASTM C138-01, 2007, “Standard Test Method for Density (“Unit
Weight”), Yield, and Air Content by Volumetric Method,” Annual Book
hocken, PA.
[7] ASTM C174-06, 2007, “Standard Test Method for Measuring Thickness
of Concrete Elements Using Drilled Concrete Cores,” Annual Book of
ASTM Standards, Vol. 04.02, ASTM International, West Conshohocken,
PA.
[8] ASTM C617-03, 2007, “Standard Test Method for Capping Cylindrical
Concrete Specimens,” Annual Book of ASTM Standards, Vol. 04.02,
ASTM International, West Conshohocken, PA.
Pervious Concrete
STP 1551, 2012
DOI:10.1520/STP104562
Samuel Meulenyzer,1 Eric Stora,1 and Fabien Perez1
Impact of Pervious Concrete Porosity

on Permeability by 3D Image Analysis
REFERENCE: Meulenyzer, Samuel, Stora, Eric, and Perez, Fabien, “Impact

of Pervious Concrete Porosity on Permeability by 3D Image Analysis,” Pervi-
ous Concrete on December 4, 2011 in Tampa, FL; STP 1551, H. J. Brown
and M. Offenberg, Editors, pp. 27–51, doi:10.1520/STP104562, ASTM Inter-
national, West Conshohocken, PA 2012.
ABSTRACT: Pervious concrete is an effective solution to manage storm water
runoff because of its ability to allow permeation of huge quantities of water.
The objective of the present study is to define relationships between formula-
tion parameters (like aggregate sizes and paste quantities) and final properties
of pervious concrete (permeability, mechanical resistance, and porosity)
through 3D images obtained by micro-tomography. Through an original experi-
mental program, we attempt to determine the permeability of different pervious
concretes directly from microstructural parameters extracted from 3D images.
We exploit mathematical morphology tools, such as two-point correlation func-
tions to access specific surface area, porosity value, and granulometric distri-
bution of porosity. The permeability of pervious concrete is finally estimated by
solving the Stokes equation on the 3D pore network numerically with finite ele-
ments. Permeability values obtained from 2D images and 3D acquisitions with
water permeability measured in laboratory are then compared.
KEYWORDS: pervious concrete, x-ray tomography, 3D image analysis,
permeability, pore structure
Introduction
Pervious concretes have been shown a renewed interest in recent years and are
recognized as a sustainable urban drainage system (SUDS). To work effi-
ciently, pervious concretes require two important properties: high permeability
and sufficient strength to withstand light traffic.
Manuscript received November 20, 2011; accepted for publication February 29, 2012; published
1
Lafarge Centre de Recherche, Saint-Quentin-Fallavier, 38291 France.
Copyright V
19428-2959.
27
However, it has generally been observed in the literature (e.g., Ref 23) that
permeability tends to decrease and compressive strength tends to increase as
the amount of voids in the pervious material decreases. The objective of this
study is thus to optimize the pore structure to get a satisfying balance between
permeability and compressive strength.
To reach this goal, both physical measurements (porosity, permeability,
compressive strength) and advanced image analysis techniques based on
acquisitions of 3D microtomography have been employed. 2D image analysis
has been widely used in the pore structure characterization of pervious con-
cretes [20], whereas the use of 3D microtomography has been more limited.
This tool is, however, well suited for characterizing the pore structure of pervi-
ous concretes and its connectivity, because the sizes of the voids range from
tens of lm to a few mm, and is presently used to link the different permeability
measured experimentally with the observations of the pore structure.
Materials and Methods
Materials
Numerous formulations have been tested but the present study focuses on eight
formulas with different aggregate sizes, volume of paste, quantity of fillers,
and water-to-binder (w/b) ratios. Crushed aggregates have been used. Table 1
below show a summary of the different pervious concretes mixed designs men-
tioned in this paper related to their different components.
Methods
Porosity Measurement—Porosity is a parameter of primary importance

when one wants to connect the drainage properties of a material to its intrinsic
characteristics such as mechanical performance and workability. The level of
mechanical performance of these materials, however, could be improved if it
TABLE 1—Sample designation extracted from the experimental program.
Specimen Range of Paste Range of Aggregate Paste

Designation Volume (L/m3) Size (mm) Modified
A/G1 140–160 8–12.5 No
A/G6 140–160 8–12.5 No
A/G8 140–160 8–12.5 Yes
A/G14 197–217 8–12.5 Yes
A/G21 83–103 8–12.5 Yes
B/G1 140–160 1–6.3 No
B/G6 140–160 1–6.3 Yes
B/G8 197–217 1–6.3 Yes
MEULENYZER ET AL., doi:10.1520/STP104562 29
was possible to reduce the amount of voids. It is indeed well known now that
the void ratio is related to the compressive strength of pervious concrete (ACI
522R-06 [1]), as shown in Fig. 1 below (cross-dotted curve).
It is well known that the permeability of a porous media is closely linked
to its porosity. As illustrated in Fig. 1, permeability decreases with porosity,
which increases the mechanical strength. The aim of ability to match the net-
work of voids of a pervious concrete would be to keep the same level of perme-
ability for a lower total porosity and thus better mechanical performance. This
is illustrated by the dotted line. A second perspective would be to optimize the
amount of voids available to flow to the total amount of voids.
The measurement techniques of porosity developed for standard concretes
are difficult to apply to pervious concretes because of the large amount of voids
contained [16,21,24,25]. Montes et al. [16] proposed a simple method suited
for pervious concretes and called “water displacement method.” This technique
is similar to the one described by ASTM standard D6857 for porous asphalt. A
cylindrical sample is tightly sealed in a plastic film, stretched, and placed in an
adapted container filled by a volume of water V0. Once the sealed sample is
immersed in water, a new volume V1 of water is measured. The film is then
taken off from the sample and the latter is immersed again in the water-filled
container V0. The container is then placed in a vacuum pump system and is
pressurized to increase the level of saturation of the permeable pores. The
pump, capable of evacuating a sealed and enclosed chamber to a <7 mbar vac-
uum, is activated for at least 15 min to remove the air bubbles from the pervi-
ous concrete cylinders. A volume V2 is thus obtained and the material porosity
is finally computed with the following formula
FIG. 1—Scheme illustrated the relationships between permeability, compres-

sive strength, and porosity for a pervious concrete (from authors).

V1 V2
U ¼ 100 (1)
V1 V0
where U is the material porosity (%) and V0 is recalled to be the volume of the
empty system (predetermined, constant).
Permeability Measurement—The vertical permeability of pervious con-

crete is measured by means of a constant-head permeabilimeter (see Fig. 2).
This system is designed so that a column of water remains constant over the
sample that is permanently saturated during the test. A high-input flow of at
least 10 L/min is required to compensate the water percolating through the cy-
lindrical sample of 80-mm diameter and 110-mm height, and to keep the water
height constant. The principle of the technique consists in measuring the differ-
ence between the flow of the water coming in the column and the one of the
water evacuated by the overflow to obtain the output flow of water passing
through the sample. The vertical permeability Kv of the pervious concrete can
be computed using Darcy’s formula
4L
Kv ¼ ðQ1 Q2 Þ (2)
p h D2
where Kv is the vertical permeability (mm/s), Q1 is the input flow (m3/s), and
Q2 is the water flow from the overflow of the permeabilimeter (m3/s), L is the
sample height (mm), h is the height of the water column imposed (mm), D is
the diameter of the sample exposed to water (m).
3DXRCT and 3D Image Analysis Tools—Tomography is a technique to

reconstruct a volume from a set of images taken from different angles,
FIG. 2—Picture of the constant-head permeabilimeter.

representing the change of the signal from the sample. Depending on the nature
of the signal, there are several types, such as electrical, magnetic, gamma, or
x-ray tomography. In all these cases, the principle is the same. Initially, acquisi-
tions consist on collecting the signal and, in a second step, rebuilding the volume
using different reconstruction algorithms. The result is a 3D image in which it is
possible to access the sections individually. 3D x-ray computed tomography
(3DXRCT) was invented by G. Hounsfield, who made the first CT scanner, and
A. Cornack, who, independently, developed a mathematical method of recon-
struction. The principle of acquisition is shown schematically in Fig. 3.
In this case, the signal resulting from interactions between the sample and
an x-ray beam follows the Beer-Lambert law
I ¼ I eðl lÞ (3)
This law delivers the intensity I of light radiation according to the intensity I of
the illumination source, the linear attenuation coefficient l and the path through
the sample. Then the volume is reconstructed by an algebraic reconstruction
technique, which consists of an inversion of the Radon transform [17]. The result
is a 3D image of the linear attenuation coefficient of the sample. This coefficient
is proportional to atomic number Z and the signal intensity is characteristic of
the chemical composition. The specificity of 3DXRCT its low resolution, which
can vary from a few tenths of a lm to a few hundred lm, depending on the de-
vice used. It is, therefore, particularly well suited to the study of materials: by
FIG. 3—Illustration of principles for 3D x-ray computed tomography.

delivering low-resolution three-dimensional information of their composition, it

is a non-destructive powerful method for the study of their microstructure.
Image-analysis techniques are now widely used in network analysis of
porous materials including building materials, whatever their amount of porosity
(usually expressed in %). The pervious concrete specimens were made in a mixer
in which the different components are mixed (aggregates, paste, admixtures, addi-
tions). Cylindrical containers of size 150 mm high and 80 mm in diameter are
used to consolidate the structure of concrete during drying. The choice of this size
is very important because it ensures that the size of the representative elementary
volume (REV) is correct for all the characterizations made thereafter.
We usually consider the elementary representative volume reached when
the size of the coarse aggregate is about 10 times smaller than the size of the
specimen. A diameter of 8 cm is very close to this value. The maximum size of
the largest aggregate is indeed around 1 cm for the largest aggregate of cut.
Aggregates used for the purpose, have maximum sizes around 5 mm. The
specimens were then demolded and scanned using a laboratory-based tomo-
graph. Because of their height of 14 cm, they are scanned in two parts (top and
bottom) and then gathered into one volume. The sensor variant used allows
viewing area of 1900 1900 pixels for the total field of view. The field of
reconstruction was set to 1000 pixels. Volumes have a final size of approxi-
mately 1000 1000 2000 voxels. The height can be between 1500 and 2000
sections in the foreground as one chooses. Indeed, the distribution of
FIG. 4—The different steps applied to 3D pervious concrete images. Tomo-

graphic images are provided in the Phoenix V-tome-X apparatus, which will be
able to support a big size of samples and give, in our case, a voxel size of 140 lm.
aggregates is not perfectly planar near the top and bottom of the specimen
because no settlement is carried out, as not to influence the compactness of the
pervious concrete. 3D images are visualized with Avizo Software [3].
Considering all these acquisitions settings, the voxel size is equal to
140 lm. Because the specimens are cylindrical, square-shaped cropping is per-
formed to take care of possible side effects. The images in grayscale are then
thresholded which is a simple operation of separating a biphasic medium.
From the extraction of the pore network, it is possible to estimate the degree of
porosity throughout the total volume or ad hoc basis which allows for high-
lighting the porosity gradients all the way up. The diagram in Fig. 4 summa-
rizes this path that provides access to the network of pores.
3D Median Filter—The median filter is a nonlinear filter. It is an alterna-

tive filter than anisotropic for example [22]. It gives the central pixel the me-
dian value of pixels of the window. It allows the elimination of isolated groups
of pixels having a size less than half a window. It is known to be effective in
treating impulse noise, while ensuring a certain conservation of the contours.
The diagram below (see Fig. 5) shows the principle of the median filter.
FIG. 5—Illustration of median filter principle.

FIG. 6—Stack fraction of porosity for the three pervious concrete formulae
made with coarse aggregates.
Porosity Gradient—Analysis of porosity is a first easy tool to use on 2D

and 3D images obtained from the XRCT. The evaluation of the porosity of per-
vious concrete is generally a good indication of their performance. It is also an
important data to be integrated in models for predicting performance. There
are several techniques for determining the level of porosity of a material. Imag-
ing techniques are generally very accurate and fast, as long as the images are
available in sufficient numbers and with resolutions of the measured quantity.
An experimental design is built to characterize the percentage of porosity using
the measurement of surface area fractions of each section. These sections have
a size of 500 500 pixels, that means 7 cm2. Then the measurement is per-
formed automatically on each cut that allow us to represent the density value
as a function of height. The graph below (see Fig. 6) shows the porosity frac-
tions for each of the three samples made with the coarse aggregate.
It is generally found on the appearance very characteristic of void ratio in
the height of pervious concrete and directly related to their implementation.
Indeed, the first top half of the specimen shows that the level of porosity
increases gradually before stabilizing in the second part of the height. The total
porosity is calculated by averaging all the fractions measured.
3D Granulometry—Granulometry is a powerful tool which can be applied

to a porous network. Performing granulometry on the pore phase of the micro-
structure provides a description of the average pore size, and the distribution of
pore sizes. Similarly, it can be applied to the matrix phase to describe the thick-
ness of the matrix material between the pores.
A further subject for investigation is the characteristic size of each phase.
This is not trivial, particularly in 3D, as objects may be connected together.
Therefore, a population of small objects may be linked to form a single large
volume. Granulometry is the technique used to calculate a representative size
for such connected objects.
Granulometry is carried out by performing a series of 3D erosion and dila-
tion operations on the phase of interest (fluid or pore media). At each step of
the series, the phase is eroded by a certain radius, and then re-dilated. If a fea-
ture has a characteristic size smaller than the erosion radius, it will disappear
during erosion, and not be re-dilated. Larger features will be re-dilated to their
original size after the erosion step. The radius of erosion/dilation is increased
until all features in the volume are removed. Thus, a histogram can be pro-
duced showing the proportion of voxels of the phase that are found in features
of a particular size. The histogram represents the distribution of feature sizes in
the sample, and this data can be used to compute mean and standard deviation
values to describe the sample.
All of these operations have been implemented as plugins. To aid the
understanding of the effect of using different structural elements, the plugins
include a feature to allow each voxel of the phase studied to be coloured
according to the size of the feature to which it belongs. For each type of
structural element attempts have been made to reduce the computation time
by applying logical simplifications to the processes performed. Another plu-
gin has been developed which further increases the speed of calculation by
scaling the studied volume to 50 % and 25 % of the original size. The granul-
ometry of small features is calculated from the full size, unscaled volume.
Granulometry of larger features is carried out on the scaled blocks, greatly
reducing the processing time. This plugin uses either octahedrons or cubes as
the structural element, although in principle the same concept could be
applied to spheres.
Tortuosity from Images—Tortuosity describes how direct a path can be

found through a percolated pore network. Different definitions of tortuosity
exist; the tortuosity defined by Jaouen et al. [12] is used in this paper. It has
been calculated from tomography data as the microstructure is segmented into
two phases: the pores and the matrix.
The pore voxels on the midplane of the volume are labelled with distance
1. Adjacent pore voxels are coloured 2, then 3, and so on, until no connected
pore voxels remain unlabelled. The average label colour is then calculated as a
function of linear distance normal to the midplane. The average label colour
gives the distance though the pore network, which is a tortuous network will be
greater than the linear distance. The average colour is plotted as a function of
FIG. 7—Calculation of tortuosity.
the linear distance, and a linear gradient taken as a measure of the tortuosity.
For a non-tortuous system; for example a linear pipe, the distance through the
pore network will equal the linear distance, and the gradient will be unity. In a
tortuous system the gradient will be greater than one. Using this method the
tortuosity can be calculated for any direction in the volume (usually for the or-
thogonal x, y, and z directions) to investigate anisotropies in the structure. The
propagation of labels can be performed considering either 6, 18, or 26 neigh-
bours around each voxel. Figure 7 demonstrates the calculation and shows a
microstructure with the pores labelled as described.
A second parameter, the extent of the connected phase, is also calculated.
This is the distance through the microstructure that is connected to the mid-
plane of the volume. In a fully percolated system, the extent of the porosity
will be the same as the dimensions of the block. In a less percolated system,
the network connected to the pores located on the midplane may extend only
part of the way through the volume.
Tortuosity from dc3d Software—The tortuosity has been measured in

another way using formation factors denoted F. The software dc3d developed
at NIST [8] allows for computing the formation factors from 3D images. The
3D image obtained by micro-tomography is converted into a conductor net-
work using the process explained in Ref 8. The conductor network is composed
of two types of conductances: the pores have a conductivity of r0 and the solid
part has a conductance of 0. A conjugate gradient relaxation algorithm is used
to solve this electrical problem and provides the voltage at every node of the
conductor network submitted to a difference of potential. The equivalent con-
ductance r of the network can then be deduced and the formation factor F
linked to the material tortuosity is obtained as
r0 U
F¼ ¼ (4)
r s
For computational reasons, the simulations were performed on images of

300 300 300 ¼ 27 106 voxels. Table 5 presents the tortuosities computed
with this method.
Description of the Two-Point Correlation (TPC) Functions—Beyond the

empirical characterization of pore and throat sizes, the pore geometry can be
characterized in a rigorous way mathematically using correlation functions,
which can be measured using image analysis. It is possible to use the curves
of covariance to determine statistically some indicators of the microstructure
of a material with two phases, as the value of porosity, pore size characteris-
tics, and specific surface of pervious concrete samples. We can also use
these two-point correlation functions (also called covariance function) to cal-
culate the intrinsic properties of materials, such as electrical conductivity
and elastic modulus [9], and the homogeneity of phases [7]. This notion was
introduced by Matheron [14] and widely detailed by Berryman et al. [5].
The two-point correlation functions S2(r) are obtained through lines of
lengths r and successively cutting the biphasic phase usually consists of the
solid phase (1) and the liquid phase (2). The next step is to count the intercep-
tion fractions of these segments through the middle of the material and to rep-
resent the distribution by the following function [6]

1 X 2r
pl
S2 ðrÞ ¼ S2 r; (5)
2r þ 1 l¼0 4r
In isotropic materials, S2(r) reaches a maximum value when r ¼ 0. This value

corresponds to the porosity of the medium. For very large values of r, the cova-
riogram tends to an asymptote, the value of which is the square of the porosity.
Between these two characteristic values, the shape of the curve reflects the ho-
mogeneity of the medium studied. The Fig. 8 below shows an example of cova-
riograms obtained for pervious formulae with coarse aggregates.
These curves are obtained from processing two-dimensional images
selected from the stack of images obtained by micro x-ray tomography.
FIG. 8—Example of TPC curve.
The 3D images without any edge effects were further processed using an
image processing and analysis software (ImageJ [11]) to treat the 1000
2D slices obtained. The gray-scale images were resized into images of
300 300 pixels, and thresholded to obtain binary images (showing the pore
and solid phases) by analyzing the grey-level histogram. From the 1000
300 pixel 300 pixel square images, 40 2D images were extracted and used
to obtain the pore structure features by means of FORTRAN programs devel-
oped at NIST [19]. These programs developed at NIST by Bentz have been
used in other studies [15] and is presently employed to compute the specific
surface on 2D slices extracted from the 3D microtomographic acquisitions.
The characteristic pore sizes for pervious concrete mixtures were extracted
using two-point correlation (TPC) functions on the two-dimensional images
of pervious concrete cross sections [18]. The TPC function provides informa-
tion about area fraction of pores, characteristic pore sizes, and the specific sur-
face area of the pores. The TPC function for a two-phase material can be
obtained by randomly throwing line segments of length r, with a specific orien-
tation, into the structure and counting the fraction of times the end points
of the line lie in the phase of interest [26]. Figure 14 shows a typical TPC func-
tion [S2(r)] for a pervious concrete mixture. The correlation length (lTPC) is
defined as
lTPC
dTPC ¼ (6)
1/
where lTPC is determined as shown in Fig. 14. The specific surface area of
the pores (sp), defined as the total pore interface area for a unit volume of
the material, can also be extracted from the slope of the function S2(r) at
r ¼ 0 [5,26]

@S2 ðrÞ
sp ¼ 4 (7)
@r r!0
Modelling Tools—The application of the Katz-Thompson (KT) law [13]

was initially developed for computing the permeability of rocks. Its applicabil-
ity to cement-based materials is a debated subject. Halamickova et al. [10]
have studied the possibility of using the KT law for water permeability and
chloride ion coefficients. Their results show that this law is applicable if the po-
rosity is very connected. The KT law may be written as follows
dc2
Kw ¼ C (8)
F
where C is a parameter, dc is the critical pore diameter, and F ¼ r=r ¼ U=s
is the formation factor, where s is tortuosity. This Katz-Thompson law is
compared with the water permeability measured previously on pervious
concretes.
3D Modelling of Permeability Tensor—The permeability of a porous mate-

rial is an intrinsic value describing its ability to pass a fluid. This is a value that
describes the interactions at the macroscopic level. It is possible to calculate
the permeability by simulating the flow through the Stokes equations applied
to the geometry of the pore space at the microscopic scale. The calculation of a
macroscopic value from terms that take into account the microscopic factors
involved to effect a change of scale. It was achieved by the method of making
local volume averaging introduced by Whitaker [27].
The principle of local volume averaging applied to the permeability is
described below. The principle of the local volume averaging is to consider
that an equation describing a phenomenon of a particular space can be aver-
aged. This provides the applicable laws in each point of the volume. In our
case, it is theoretically possible to model single-phase transport phenomena
through a pore network from the equations governing the transport across the
pore. However, applied across a material, such an approach will not lead to a
certain outcome, given the often complex nature of the porous medium. The
method of taking average density considered in this case it is indeed more
effective to express the problem in terms of averaged equations from informa-
tion across the pore and describing the phenomenon in all the material. For
single-phase transport in porous media, the flow is described at the macro-
scopic scale by Darcy’s law
K
q¼ rP (9)
l
This law expresses that the flow q is proportional to the pressure gradient. It
has been shown that the coefficient of proportionality has two terms. The first
depends on the fluid; it is expressed by the viscosity l, the second is a charac-
teristic of the porous medium; it is the permeability K.
At the microscopic level and under certain conditions, a flow can be
expressed by the Stokes equations:
In the phase b
rpb þ qb g þ lb r2 vb ¼ 0 (10)
r vb ¼ 0 (11)
At the interface br: vb ¼ 0 (12)
The Navier-Stokes equations describe the general behavior of a fluid from the
conservation of momentum. We can deduce Eq 9 by making assumptions.
Indeed, it is assumed to consider only laminar flow where the viscosity effects
related to predominate over inertial effects.
It is also assumed that the density is constant in the local form of the con-
servative principle of mass conservation. This equation is, therefore, valid for
an incompressible fluid. The boundary condition of Eq 11 corresponds to the
condition of fluid adhesion near the wall. This is one of two conditions on the
interaction of the fluid with the wall that is possible to solve the Stokes equa-
tion. In the case of a fluid satisfying these conditions, Darcy’s law is valid at
the macroscopic scale, that is to say, anywhere across the pore. It is then possi-
ble to consider that it actually corresponds to a local average of Stokes equa-
tions. That is what was demonstrate by Whitaker [27], obtaining the tensor
form of Darcy’s law
K
vb ¼ r Pb (13)
l
where K is the permeability tensor.

This tensor can be expressed from the local field tensor resulting from the
close problem. It can express the perturbation of the local field velocities (mi-
croscopic) from the macroscopic pressure gradient. The permeability tensor
thus characterized the flow conditions within the material for the pressure gra-
dients applied along the three principal directions of space (see Fig. 9).
These equations were integrated in a Fortran code that allow the calcula-
tion of this permeability tensor from a 3D image of a porous media. It is also
possible to get a transcription of fields vectors of this tensor for each voxel that
FIG. 9—Schematic description of the permeability tensor K for all space

directions.
allow us the visualization of “strength” lines based on the calculated vectors,

as illustrated in the picture below (Fig. 10):
These figures clearly shown that porosity network is very well connected.
We found that more of 99 % of pores are percolated.
More details about this method and some examples of application should
be found for the interested reader in Refs 2 and 4.
Results and Discussion

Figure 11 below shows that there is a good agreement between the water acces-
sible porosity obtained experimentally and the one measured on the 3D micro-
tomographic images. Note that the water porosity is higher than the one
measured by 3D image analysis because of water in small pores that cannot be
detected in tomography.
The relationship between compressive strength, permeability and global po-
rosity has been investigated on a large panel of pervious concretes (see Fig. 12).
Figure 12 shows that for the formulae tested there is no simple relation
between porosity and permeability as claimed by Neithalath et al. [18]. It
should be noted that the permeability can vary significantly for systems with
porosites of approximately 30 %. Based on the figures above, the eight formu-
lations detailed in Table 4 have been chosen and studied in more detail. The
Fig. 13 clearly shows that there is no direct correlation between the permeabil-
ity and the total porosity measured on 3D images. Therefore, other morphologi-
cal indicators of pore structure, such as specific surface area, pore size
distribution, and critical pore diameter have been investigated.
FIG. 10—3D pictures of strength lines for permeability calculated from

3DXRCT images. Example given for sample A/G1. Left: flux lines through the
pore network. Right: flux lines superposed to the porosity network (in white).
Some interesting points have been chosen from series A formulations con-
taining coarse aggregate (A/G1, A/G6, and A/G8). These samples all have a
porosity of 30 % and compressive strength around 10 MPa. However, there is a
significant difference in permeability, as shown in Table 2.
FIG. 11—Relation between the water porosity measured the water displace-
ment method and the one obtained from the 3D images.
FIG. 12—Variations of permeability (triangles) and compressive strength (dia-

monds) as functions of the porosities obtained from the 3D images for pervious
concretes composed of (a) fine aggregates, and (b) coarse aggregates.
For similar reasons, B/G1, B/G6, and B/G8 pervious concrete samples have
been chosen for their physical characteristics, as shown in Table 3 below.
3D characterizations described above were performed on these interested
points. The pore size distributions were performed for the eight formulae of
Table 4. The results are presented in Fig. 14.
FIG. 13—Variations of permeability as functions of the porosities obtained from

the 3D images on the eight formulae of pervious concretes detailed in Table 4.
TABLE 2—Physical values for some coarse aggregates pervious.
Coarse Aggregates Kv (cm/s) Compressive Strengh (MPa) XRCT Porosity (%)

A/G1 3.06 10 29.9
A/G6 4.4 9.7 29.6
A/G8 5.49 10.6 30.3
TABLE 3—Physical values for some fine aggregates pervious.
Fine Aggregates Kv (cm/s) Compressive Strength (MPa) XRCT Porosity (%)

B/G1 1.99 10.5 33.1
B/G6 2.05 8 36
B/G8 1.46 12.3 27.2
TABLE 4—Some details of the eight formulae chosen from a large panel of pervious concrete and
their physical properties.
Specimen Paste Kv Cs Voids from Voids from Specific Surface TPC Diameter
Designation Volume (cm/s) (MPa) XRCT (%) TPC Area (m2/m3) (mm)
A/G1 Small 3.06 10 29.94 29.58 145.38 1.05
A/G6 Small 4.40 9.7 29.58 29.67 149.49 1.03
A/G8 Small 5.49 10.6 30.28 30.45 144.41 1.17
A/G14 High 1.99 10.5 33.12 31.27 240.71 0.75
A/G21 Very low 2.05 8 31.10 31.26 241.84 0.75
B/G1 Small 1.46 12.30 23.60 20.59 117.57 0.38
B/G6 Small 1.24 13.6 22.84 21.69 79.93 0.67
B/G8 High 3.69 5.9 35.34 33.59 185.53 1.26
FIG. 14—3D volume cumulated distribution of pore network for six on the
small and coarse aggregate pervious concrete formulae.
These curves are interesting because they show a difference in the average
pore size for the A and B series. B-series samples (containing fine aggregates)
have a smaller mean value of pore size compared to A-series samples (contain-
ing coarse aggregates).
Tortuosities are obtained from the calculation described in the section,
“Tortuosity from dc3d Software,” and compared with the values of tortuosity
obtained through the diffusion model described in the section, “Description of
the Two-point Correlation (TPC) Functions.” Results are shown in Table 5.
The influence of the volume of paste on the pore structure and permeability
has been investigated in Fig. 15. The large ball size denotes the formulae with
a high volume of paste.
TABLE 5—Tortuosities for A and B series from 3D images and from diffusion modeling.
Tortuosity from 3D Images Tortuosity from Conductivity Model

B/G1 1.002 1.010
B/G6 1.002 –
B/G8 1.004 –
A/G1 1.003 1.011
A/G6 1.007 –
A/G8 1.006 –
FIG. 15—Variations of permeabilities as functions of (a) specific surfaces, and

(b) the critical pore diameters obtained from the 3D images on the formulae of
pervious concretes with different volumes of paste.
Figure 15 shows the specific surface areas (sp) of the pervious concrete
mixtures in m2/m3 of pervious concretes. 1 m3 of pervious concrete thus devel-
ops a significant pore surface 80 to 240 m2 depending on the formulation
tested. As can be seen in Fig. 15, sp and dTPC are found to decrease as the size
of aggregates diminishes and the volume of paste increases. The TPC diame-
ters are generally higher than in Neithalath et al. [18], but it is logical because
the permeability measured on these formulae are higher, too.
FIG. 16—Relationship between specific surface area and permeability for fine
and coarse selected samples.
Figure 16 below illustrates the variations of specific surfaces, critical pore

diameters and permeability of pervious concretes with a same volume of paste.
A clear difference can be observed between the formulation containing coarse
and finer aggregates. Permeability of pervious concretes with coarse aggregates
and the same volume of paste can still vary significantly, even if their specific
surfaces and critical diameter are similar. Further investigations are required to
understand these variations.
It also appears that the permeability is closely linked to the critical pore di-
ameter. Therefore, the models proposed in literature to link critical pore diame-
ter and permeability, such as Katz-Thomson, has been applied on the pervious
concretes tested.
In the present study, two different kinds of models are applied to pervious
concretes. The first one is the application of the Katz-Thompson law. If we
consider the tortuosity calculated from 3D images equal to 1 (see Table 5), the
correlation between the permeability measured and that predicted by Katz-
Thompson law is good, as shown in Fig. 17.
The 3D permeability tensor described in the section, “3DXRCT and 3D
Image Analysis Tools,” on the formulation with fine aggregates has also been
estimated (see Fig. 18). The results below are normalized with respect to the
permeability of the B/G1 formula, because the value of the permeability tensor
in the axis of gravity (z axis) estimated by Permea3D does not directly corre-
spond to the absolute values that can be measured physically. Note that the
other values of the tensor are non-zero, which is logical because the flow, even
FIG. 17—Relationship between the permeability measured experimentally and

the one predicted with Katz-Thompson law.
if it is mainly in this direction is also spread on the sides. It is, therefore, diffi-
cult to compare the measured and computed absolute values. But the same
trends as the one obtained experimentally are observed in the following figure
for B-series pervious concretes.
FIG. 18—Relationship between the permeability measured and the predicted

by Permea3D.
Conclusions
A panel of techniques has been employed to characterize the pore structure and
the permeability of different formulations of pervious concretes. The main con-
clusions of this study are as follows:
(1) There is no simple relation between total porosity and permeability as
also already shown by Neithalath et al. [18]. Therefore, more advanced
characterizations of the pore structure with image analysis have been
carried out.
(2) Image analysis techniques provide important information on the micro-
structure of pervious concretes, such as:
– paste thickness,
– total porosity and pore size distribution,
– porosity gradients, and
– morphological parameters of the pore structure (tortuosity, specific
surface, and critical pore diameter). These microstructural parame-
ters are important data for the mechanical and physical properties of
pervious concretes.
(3) The aggregate size and volume of paste are found to have a strong
influence on pore parameters, such as critical pore diameter. The per-
meability is directly impacted by the changes of these pore structure
parameters.
(4) Modelling tools based on image analysis are useful to understand the
link between pore structure and permeability of this material. The
Katz-Thompson model works well on the formulae tested but its valid-
ity should be tested on other formulations. The 3D numerical tool Per-
mea3D is also a promising tool.
(5) The present approach based on physical experiments, 3D image analy-
sis, and modeling helps in the formulation of pervious concretes (vol-
ume of paste and aggregate sizes) to get good compromise between
permeability and compressive strength.
Acknowledgments
The writers would like to acknowledge Dominique Bernard from ICMCB, Pes-
sac (France) and Dale P. Bentz from NIST (USA) for helping us in this work.
The writers also would like to acknowledge all people from Lafarge Centre de
Recherche for their precious help.
References
[1] ACI Committee 522, “Pervious Concrete,” 522R-06, American Concrete

Institute, Farmington Hills, MI.
[2] Anguy, Y., Bernard, D., and Ehrlich, R., “The Local Change of Scale
Method for Modeling Flow in Natural Porous Media (I): Numerical
Tools,” Adv. Water Res., Vol. 17, 1994, pp. 337–351.
[3] Avizo software, http://www.vsg3d.com/avizo/.
[4] Bernard, D., “Using the Volume Averaging Technique to Perform the
First Change of Scale for Natural Random Porous Media,” Advanced
Methods for Groundwater Pollution Control, G. Gambolati and G. Verri,
Eds., Springer-Verlag, New York, 1995, pp. 9–24,
[5] Berryman, J. G. and Blair, S. C., “Kozeny–Carman Relations and Image
Processing Methods for Estimating Darcy’s Constant,” J. Appl. Phys.,
Vol. 62, 1987, pp. 2221–2228.
[6] Berryman, J. G. and Blair, S. C., “Use of Digital Image Analysis to
Estimate Fluid Permeability of Porous Materials: Application of Two-
Point Correlation Functions,” J. Appl. Phys., Vol. 60, 1986, pp.
1930–1938.
[7] Coster, M. and Chermant, J.-L., “Image Analysis and Mathematical Mor-
phology for Civil Engineering Materials,” Cement Concr. Comp., Vol.
21, 1999, pp. 403–412.
[8] Garboczi, E. J. and Bentz, D. P., “Computer Simulation of the Diffusivity
of Cement-Based Materials, J. Mater. Sci., Vol. 27, No. 8, 1992, pp.
2083–2092.
[9] Garboczi, E. J., Bentz, D. P., and Martys, N. S., “Digital Images and
Computer Modelling,” Experimental Methods in the Physical Sciences,
Vol. 35, Chap. 1, Academic Press, San Diego, 1999, pp. 1–41.
[10] Halamickova, P., Detwiler, R. J., Bentz, D. P., and Garboczi, E. J., “Water
Permeability and Chloride Ion Diffusion in Portland Cement Paste Mor-
tars: Relationship to Sand Content and Critical Pore Diameter,” Cem.
Concr. Res., Vol. 25, No. 4, 1995, pp. 790–802.
[11] ImageJ, “Image Processing and Analysis in Java,” http://imagej.nih.gov/
ij/download.html.
[12] Jaouen, L., Olny, X., and Sgard, F., Microstructure et propriétés des
matériaux, Marne-la-Vallée, 2005.
[13] Katz, A. J. and Thompson, A., “Quantitative Prediction of Permeability in
Porous Rocks,” Phys. Rev. B, Vol. 34, 1986, pp. 8179–8181.
[14] Matheron, G., Random Sets and Integral Geometry, John Wiley & Sons,
New York, 1975.
[15] Marolf, A., Neithalath, N., Sell, E., Wegner, K., Weiss, J., and Olek, J.,
“Influence of Aggregate Size and Gradation on Acoustic Absorption of
Enhanced Porosity Concrete,” ACI Mater. J., Vol. 101, No. 1, 2004, pp.
82–91.
[16] Montes, F., Valavala, S., and Haselbach, L. M., “A New Test Method for
Porosity Measurements of Portland Cement Pervious Concrete,” J. ASTM
Int., Vol. 2, No. 1, 2005.
[17] Natterer, F., The Mathematics of Computerized Tomography, SIAM: So-

ciety for Industrial and Applied Mathematics, Philadelphia, PA, 2001.
[18] Neithalath, N., Weiss, J., and Olek, J., “Characterizing Enhanced Porosity
Concrete Using Electrical Impedance to Predict Acoustic and Hydraulic
Performance,” Cement Concr. Res., Vol. 36, No. 11, 2006, pp.
2074–2085.
[19] Neithalath, N., Marolf, A., Weiss, J., and Olek J., “Modeling the Influence
of Pore Structure on the Acoustic Absorption of Enhanced Porosity Con-
crete,” J. Adv. Concr. Technol., Vol. 3, No. 1, 2005, pp. 29–40.
[20] Neithalath, N., Sumanasooriya, M. S., and Deo, O., “Characterizing
Pore Volume, Sizes, and Connectivity in Pervious Concretes towards Per-
meability Prediction,” Mater. Charact., Vol. 61, No. 8, 2010, pp.
802–813.
[21] NF EN12697-19 þ A1, 2007, “Mélanges bitumineux—Méthodes d’essai
pour mélange hydrocarboné à chaud—Partie 19,” Perméabilité des
éprouvettes.
[22] Perona, P. and Malik, J., “Scale-Space and Edge Detection Using Aniso-
tropic Diffusion,” IEEE Trans. Pattern Anal. Machine Intell., Vol. 12,
No. 7, 1990.
[23] Schaefer, V. R., Wang, K., Suleiman, M. T., and Kevern, J. T., “Mix
Design Development for Pervious Concrete in Cold Weather Climates,”
Report No. 2006-1, National Concrete Pavement Technology Center, Feb
2006.
[24] ASTM D6752-01, “Standard Test Method for Bulk Specific Gravity and
Density of Compacted Bituminous Mixtures Using Automatic Vacuum
Sealing Method,” Annual Book of ASTM Standards, Vol. 04.03, ASTM
International, West Conshohocken, PA.
[25] Sumanasooriya, M. S. and Neithalath, N. “Stereology and Morphology
Based Pore Structure Descriptors of Enhanced Porosity (Pervious) Con-
cretes,” ACI Mater. J., Vol. 106, 2009, pp. 429–438.
[26] Torquato, S., Random Heterogeneous Materials—Microstructure and
Macroscopic Properties, Springer Science and Business Media, New
York, 2002.
[27] Whitaker, S., “The Method of Volume Averaging,” Theory and Applica-
tion of Transport in Porous Media, Vol. 13, Kluwer Academic Publishers,
New York, 1999.
Pervious Concrete
STP 1551, 2012
DOI:10.1520/STP104501
L. K. Crouch,1 John P. Hendrix,2 Alan Sparkman,3

and Daniel Badoe4
Variability of Fresh and Hardened Voids

REFERENCE: Crouch, L. K., Hendrix, John P., Sparkman, Alan, and Badoe,
Daniel, “Variability of Fresh and Hardened Voids of Pervious Concrete,” Per-
vious Concrete on December 4, 2011 in Tampa, FL; STP 1551, H. J. Brown
and M. Offenberg, Editors, pp. 52–68, doi:10.1520/STP104501, ASTM Inter-
national, West Conshohocken, PA 2012.
ABSTRACT: ASTM C1688-08 currently contains only a single-operator
standard deviation for the density criterion for determining allowable range.
The adequacy of this criterion was put to the test in recent research under-
taken for the Tennessee Concrete Association (TCA). Twenty replications
of three pervious mixtures were used. The control mixture had 356-kg/m3
(600-lb/yd3) of cementing materials with a 0.3 w/cm ratio, 1533-kg/m3 (2584-
lb/yd3) of No. 89 limestone coarse aggregate and no fine aggregate. The sec-
ond mixture had 3.5 % river sand replacement of the coarse aggregate (by
total aggregate volume) and a 0.31 w/cm ratio. The third mixture had 7 %
river sand in the combined aggregate gradation and a w/cm ratio of 0.32.
Coarse aggregate content was reduced to accommodate the changes in the
second and third mixtures. Twenty replications of each TCA mixture yielded
coefficients of variation of ASTM C1688 density values of 0.3, 0.5, and 0.7
%, respectively for the control mixture, the 3.5 % sand mixture, and the 7 %
sand mixture. All three mixtures met current allowable ASTM C1688 range
criteria based on ASTM C670-03 multipliers. However, the variability in den-
sity increased substantially as sand and w/cm ratios were increased, possibly
indicating that the variability criterion for density may need to be revised.
Manuscript received October 31, 2011; accepted for publication February 29, 2012; published
1
Ph.D., P.E., Professor, Dept. of Civil and Environmental Engineering, Tennessee Technological
Univ., Cookeville, TN 38505, e-mail: lcrouch@tntech.edu
2
P.E., McClone Construction Company, Sterling, VA 20166.
3
CAE, CCPf, LEED AP, Tennessee Concrete Association, Nashville, TN 37203.
4
Ph.D., Professor, Dept. of Civil and Environmental Engineering, Tennessee Technological
Univ., Cookeville, TN 38505.
Copyright V
19428-2959.
52
CROUCH ET AL., doi:10.1520/STP104501 53
Other findings include a good correlation between C1688 voids and pervious
properties such as effective (water accessible) voids (ASTM D7063-05) and
compressive strength (ASTM C39-06) indicating that C1688 voids may be
useful as a mix design and adjustment tool. Twenty replications of each TCA
mixture yielded coefficients of variation of ASTM C1688 void values of 1.1,
2.5, and 3.6 %, respectively, for the control mixture, 3.5 % sand mixture, and
7 % sand mixture. Information on compressive strength and effective void
variability is also provided.
KEYWORDS: density, fresh concrete, pervious concrete, proctor hammer,
void content
Introduction and Research Significance

ASTM C1688-08 [1] is a relatively new test method with tremendous potential.
First, the results of the test method may be used to characterize and limit
batch-to-batch variability of pervious concrete as the ASTM sub-committee
that developed the method intended. Secondly, and possibly more importantly,
ASTM C1688 voids may prove extremely valuable as a mixture design and
adjustment tool. Historically, in Tennessee, pervious concrete mixture design
was accomplished by field trial and error modified with experience. The field
placements, subsequent coring and testing proved expensive and time consum-
ing. Research at Tennessee Technological Univ. (TTU) sponsored by the Ten-
nessee Concrete Association (TCA) indicates that pervious concrete
compressive strength and hardened voids may be strongly related to ASTM
C1688 voids of the plastic pervious concrete. Figures 1 and 2 show the com-
bined results of the research undertaken by TTU researchers Medley [2] and
Hendrix [3] for pervious concrete mixtures containing ASTM C33/C33M-08
[4] No. 8, No. 89, and similar coarse aggregates. Each figure contains about 85
data points. The coefficients of determination, both greater than 0.83, indicate
the possibility of a strong relationship between hardened properties and ASTM
C1688 voids. The purpose of this paper is to provide further information on
ASTM C1688 variability, especially variability of voids which is not currently
addressed in the test method.
Literature Review
Little literature is available on ASTM C1688 use; therefore, the literature
review will be concerned with the need to improve pervious PCC compressive
strength and mixture design alterations to achieve that goal. ASTM C1688 may
play a vital role in pervious PCC compressive strength improvement.
Compressive strengths have been reported to range from 2.8 to 27.8-MPa
(400 to 4000-psi) [5]. However, typical compressive strengths ranging from
13.8 to 17.2-MPa (2000 to 2500-psi) allow the concrete to be used for
FIG. 1—Pervious concrete hardened voids versus ASTM C1688 plastic voids.
FIG. 2—Pervious concrete core compressive strength versus ASTM C1688

plastic voids.
vehicle traffic [6]. There is currently no standard method for compressive

strength for pervious concrete, often ASTM C39 is used even though repeat-
ability of the test method for pervious concrete has not been verified. There
are several factors that influence compressive strength, such as effective
(water accessible) voids, w/cm ratio, aggregate size, and the level of com-
paction [7]. The need for higher compressive strength in pervious PCC has
been of concern. Because of the relatively low compressive strength, pervi-
ous concrete has been designated for lower traffic load applications, such as
sidewalks, driveways, recreation trails, and parking lots [8]. According to
research completed at Tsinghua Univ., the weakest link of pervious PCC is
strength of the cement binder. This research tested two methods of improv-
ing the compressive strength. The first method was to increase the volume
of paste and cement binder area. Smaller aggregate was used to increase the
surface area of the paste, which, in turn, increased the compressive strength.
The second method used in the research was to increase the actual strength
of the cement paste. This was accomplished by adding fine supplementary
cementitious materials to the paste content [9]. In agreement, research
conducted at Iowa State Univ. had similar compressive strength improve-
ments from adding a small amount of fine aggregate to the pervious PCC
matrix [10].
The void content of a pervious PCC slab is typically more important to its
performance than its compressive strength [11]. Pervious PCC void content is
not only caused by the scarcity of fine aggregate contained in the mixture, but
compactive effort, w/cm ratio, cementitious material content, and aggregate
gradation [7]. These voids can range from 15 to 30 % of the entire concrete
structure giving the pavement its porosity to allow rainwater to infiltrate into
the soil below [12]. The size of these voids can range anywhere from 2.0 to
8.1-mm (0.08 to 0.32-in.) [7]. Keeping the void content between 20 and 25 %
has been shown to consistently yield compressive strengths in the range of 13.8
to 17.2-MPa (2000 to 2500-psi) [6]. A decrease in the effective air voids will
also have a decreasing effect on the permeability, but will result in an increase
in compressive strength [13].
There are two measurements of voids that can be taken from pervious con-
crete. Using ASTM C1688, the density and void content of fresh pervious con-
crete can be taken. This void content of the fresh pervious concrete is related
to, but not indicative of, the void content of the in-situ pervious PCC [1]. The
other method for measuring voids was explored by researchers at TTU using
an adaptation of ASTM D7063-05. This method used a modification of an
ASTM specification for determining the effective porosity and effective air
voids of compacted bituminous pavements to determine the effective air void
content of pervious PCC [11]. Research has shown that smoother, rounder
gravel-type aggregates will produce fewer effective voids than the more angu-
lar limestone aggregates [14].
Aggregates
Typical coarse aggregates meet the requirements of ASTM C33/C33M-08 No.
7, No. 8, No. 67, No. 89, or a similar aggregate [7]. The aggregates for pervious
concrete are typically smaller than that of typical PCC. This gives the mix
additional workability, while also producing a smoother riding surface [11].
Even though the compressive strengths of pervious PCC are typically low com-
pared to that of conventional PCC, an increase in compressive strength can be
seen when the aggregate size is reduced [9]. This is because of the amount of
contact surface area between more aggregates [9]. In addition, the shape of the
aggregate has been shown to have an effect on the compressive strength.
Round aggregates, such as river gravel, have shown an increased compressive
strength greater than that of angular limestone aggregates with similar grada-
tion [15]. The amount of coarse aggregate used in typical pervious mixes
ranges from 1424 to 1543-kg/m3 (2400 to 2600-lb/yd3) [16]. Although pervious
concrete is also known as “no-fines” concrete, a small amount of fine aggregate
(sand) has been added with success. The addition of sand increases compres-
sive strength; however, void content and permeability are decreased because of
the sand filling a portion of the previously open voids [10]. The addition of
sand will not only enhance the compressive strength of the pervious PCC, but
it will also increase the freeze–thaw resistance [5].
Cementing Materials
Type I portland cement along with supplementary cementitious materials
(SCM) complying with ASTM C150/C150M, C595/C595M, or C1157/
C1157M are used to form the paste for pervious PCC [7]. Typical pervious
PCC mixes consist of approximately 267 to 415-kg/m3 (450 to 700-lb/yd3) of
cement products [5]. SCMs are used in over 60 % of the ready mixed concrete
today. Fly ash is the most commonly used of the SCMs. It is used in nearly half
of the ready-mixed concrete mixes [17]. Fly ash is a by-product of coal com-
bustion for electricity generating plants. Using fly ash as a partial replacement
(15 to 25 %) for portland cement reduces the energy needed to produce the
cement, and reduces the amount of CO2 produced by the portland cement pro-
duction process [17,18]. Not only is the use of fly ash good for the environment
in the way of energy consumption, it is also beneficial to the concrete mix
itself. When fly ash is added to a concrete mix partially replacing portland
cement, it improves workability of the mix. It acts as a water reducing agent
and makes the mix flow better [18]. The cement paste binder layer between the
aggregates is thin; therefore, this results in the weak point of the hardened con-
crete where it ultimately fails under compression [9]. The smaller sized aggre-
gate increases the paste binder surface area, which, in turn, strengthens the
concrete [9].
Chemical Admixtures
Pervious PCC mix is a harsh mixture that has a small placement window with-
out the addition of chemical admixtures. This makes placing pervious concrete
a labor-intensive task. From the discharge of the mix from the concrete truck,
to the compactive effort and final strike-off, the placement of pervious PCC
without chemical admixtures can be an aggravating process [19]. With the low
w/cm ratio, this creates a scenario where the placement of pervious concrete is
extremely difficult. The addition of a mid-range water reducer (MRWR) cre-
ates a more fluid product without having an excess amount of water in the mix
[19]. Too much water will result in the sealing of the open voids and loss of
permeability.
The higher portland cement level along with the lower w/cm ratio of pervi-
ous concrete allows for faster setting of the cement paste. With the faster set-
ting time, placement again becomes an issue. A cement hydration stabilizer
can be added to prolong the plastic state of the pervious PCC mix, gaining
more time to place the mix [19]. With the addition of a hydration stabilizing
chemical admixtures, the rather short discharge time of 1 h may be extended to
1.5 h or more for a concrete truck delivering pervious PCC [15].
The lack of, or small amount of, fine aggregate in pervious concrete has
been documented as a factor contributing to the harshness of the mix. A
viscosity-modifying admixture (VMA) can be added to give the mixture more
flow. The VMA will assist in the discharge of the mix from the concrete truck
[19]. In addition, using a VMA in concrete will help reduce aggregate
segregation, and control paste drain down. Paste drain down is an event that
occurs when the cement paste is too fluid, and, therefore, settles to the bottom
of the pervious concrete, sealing it, and not allowing any fluid to percolate
though the pervious section.
For conventional PCC, the lower the w/cm ratio, the higher the compres-
sive strength is. This relationship does not directly apply to pervious PCC.
According to McCain and Dewoolker, higher w/cm ratios corresponded with
an increase in compressive strength because of lower voids [20]. The typical
range for the w/cm ratio is 0.26 to 0.45 [7]. However, with the proper admix-
ture combination in the mixture, w/cm ratios are generally in the range of 0.27
to 0.30 [15].
Compaction
The compaction that is applied to pervious concrete is the key to all of the
hardened properties. If the in-place concrete is compacted too much, the effec-
tive voids will decrease. This will begin to close off some, and eventually all,
of the interconnected voids restricting water flow rendering it impervious.
However, compressive strength of over compacted pervious PCC is extremely
high. Not enough compaction will result in an increase of effective voids and
will lower compressive strength [13]. Research completed by Meininger,

looked at different levels of compaction effort, and how it affects compressive
strength, unit weight, and air void content, concluding that some compaction is
desired [13]. However, compactive effort of test specimens needs to resemble
the compaction that is applied to the pervious PCC slab in the field. Literature
by Mahboub et al. [16], concludes that the conventional method of preparing
concrete cylinders by rodding is not an accurate representation of the compac-
tion that is applied in the field. A closer demonstration of the compactive effort
in the field is by pneumatically pressing the pervious PCC cylinders at
0.07-MPa (10-psi) under lab conditions [5]. The durability of pervious PCC is
also affected by compaction. Compaction is required to seat the aggregates into
the surface of a slab, and to reduce raveling [13].
Materials
The PC, obtained from a local supplier, used in the research met ASTM
C150-06 Type I requirements [21]. Class F fly ash that met ASTM C618-08a
was used as 20 % of the total cementitious material in all mixtures in this study
[22]. The coarse aggregate for this research was sold by a local supplier as a
9.5-mm (3/8-in.) crushed limestone aggregate. Standard sieve analyses in
accordance with ASTM C117-04 and ASTM C136-06 were performed on the
coarse aggregate [23,24]. The results are shown in Table 1. The coarse aggre-
gate saturated surface dry (SSD) bulk specific gravity (BSG) and absorption,
determined as per ASTM C127-06, were 2.67 and 1.5 %, respectively [25].
From previous research, the BSG SSD and absorption of the Ohio River Sand
was 2.607 and 0.9 %, respectively. In addition, the gradation of the fine aggre-
gate is also shown in Table 1. A combination of chemical admixtures, as pre-
scribed by TCA for pervious PCC, was used in each batch. Finally, all pervious
concrete mixtures were made with tap water.
TABLE 1—Aggregates.
Project Coarse ASTM No. 89 Project River ASTM C33 Fine

Sieve Aggregate Stone Sand Aggregate
Size (% Finer by Mass) (% Finer by Mass) (% Finer by Mass) (% Finer by Mass)
12.5 mm 100 100 – –
9.5 mm 99 90–100 100 100
4.75 mm 33 20–55 96 95–100
2.36 mm 5 5–30 88 80–100
1.18 mm 3 0–10 79 50–85
600 lm – – 55 25–60
300 lm 3 0–5 11 5–30
150 lm – – 1 0–10
75 lm – – 0.5 –
TABLE 2—Mixtures.
Control 3.5% Sand 7% Sand

Component Mixture Mixture Mixture
Type I Portland Cement, kg/m3 (lb/yd3) 285 (480) 285 (480) 285 (480)
Class F fly ash, kg/m3 (lb/yd3) 71 (120) 71 (120) 71 (120)
No. 89 limestone SSD, kg/m3 (lb/ yd3) 1533 (2584) 1473 (2482) 1413 (2382)
River sand SSD, kg/m3 (lb/ yd3) 0 52 (88) 103 (174)
Water, kg/m3 (lb/CY) 107 (180) 110 (185) 112 (189)
Hydration stabilizer, mL/100 kg (oz/wt.3) 261 (4) 261 (4) 261 (4)
Mid-range water reducer, mL/100 kg (oz/wt.3) 326 (5) 326 (5) 326 (5)
Viscosity modifier, mL/100 kg (oz/wt.3) 130 (2) 130 (2) 130 (2)
Pervious PCC Mix Designs

The three pervious PCC mixture designs used in this research are shown in
Table 2. The control mixture had 356-kg/m3 (600-lb/yd3) of cementing materi-
als with a 0.3 w/cm ratio and no fine aggregate. The second mixture had 3.5 %
river sand replacement of the coarse aggregate (by total aggregate volume) and
a 0.31 w/cm ratio. The third mixture had 7 % river sand in the combined
aggregate gradation and a w/cm ratio of 0.32. Coarse aggregate content was
reduced to accommodate the changes in the second and third mixtures. The ini-
tial mixture design was based on previous experience and the subsequent mix-
ture designs on the desire to reduce ASTM C1688 void contents. Table 3
shows mixture parameters such as w/cm ratio and % fine aggregate for each
mixture. The differences in the mixtures were in fine aggregate and water con-
tent. Hereafter, mixtures will be referred to as either the control (no fine aggre-
gate) or by their % fine aggregate.
Procedure
Casting and Plastic Properties

The batch size of the 60 pervious PCC batches for this research project was
0.017-m3 (0.60-ft3) mixed in a 0.028-m3 (1-ft3) capacity electric shear mixer.
TABLE 3—Mixture parameters.
Control 3.5% Sand 7% Sand

Quantity/Ratio/Percentage Mixture Mixture Mixture
Cementing materials content, kg/m3 (lb/yd3) 356 (600) 356 (600) 356 (600)
Water-cementing-materials-ratio 0.30 0.31 0.32
Percent fine aggregate by total aggregate volume 0 3.5 7
Percent fly ash by weight of total cementing materials 20 20 20
The mixing process complied with ASTM C192/C192M-07 [26]. Immediately

after the mixing of each batch was completed, ASTM C1688 was conducted on
the fresh pervious PCC to determine the density and percentage of voids. The
pervious concrete was then placed uniformly into a 152-mm 152-mm 610-
mm (6-in. 6-in. 24-in.) oiled wooden mold. Concrete from the ASTM
C1688 test was reused in the beam mold. The fresh pervious concrete was
allowed to overflow the mold. A 152-mm 305-mm (6-in. 12-in.) concrete
rolling pin was then used to lightly compact the concrete into the mold. The
only weight used to compact the concrete was the weight of the approximately
9-kg (20-lb) rolling pin. Two passes were made with the rolling pin. This type
of compaction was used to simulate light compaction of pervious concrete in
the field placement operations common in Tennessee. The excess concrete
above the top of the wooden mold was removed and discarded by passing a
steel straight edge across the length of the mold.
Curing and Hardened Properties

Once the final strike-off had been completed, the fresh pervious PCC beam
was covered with a plastic bag and left to cure for 2464 h, in compliance with
ASTM C192. After the initial 24h curing period, the pervious PCC beam was
removed from the wooden molds, and marked with spray paint. The beam was
then transferred to a lime-water immersion tank. The pervious beam remained
submerged in the tank for at least 7 days. At that time, the beam was removed
from the tank. Four cores were taken from the beam in accordance with ASTM
C42-06 [27]. Cores were not cleaned or flushed to remove cuttings because
observations did not indicate a need for cleaning. Two of the cores
were marked and returned to the lime-water tank where they remained until the
28-day compressive strength test. The remaining two cores were marked and
placed in an industrial oven with a temperature of 110 6 3 C (230 6 5 F).
These cores remained in the oven between 7 and 14 days when hardened prop-
erty void tests were performed.
The test method that was used to measure the effective air void content
of the hardened pervious concrete is a hot-mix asphalt (HMA) test, ASTM
D7063-05 [28]. Approximately seven days after the two cores were placed in
the oven, the cores were removed and allowed to cool for several hours prior to
commencing the test. The oven-drying time period of 7 days was to ensure that
the cores were completely dry. It was determined through research at TTU that
an effective way to measure the voids present in the pervious PCC sample was
to use the Instroteck Corelok System [11]. The resulting effective voids of the
two cores were averaged and reported as the effective void content of the pervi-
ous batch.
The compressive strength test method was performed on two approxi-
mately 100-mm (4-in.) pervious cores from each beam. After being drilled
from the beam, these two cores remained in the lime-water tank until 28 days
from the original casting date. Prior to testing, the cores were capped with sul-
fur in accordance with ASTM C617-06 [29]. The actual compression test was
performed in compliance with ASTM C39-06 [30]. Because the length (l) to di-
ameter (d) ratio was less than 1.8, an l/d correction factor was used in calculat-
ing the compressive strength. This correction factor was different for almost
every core and was determined from ASTM C39.
Results
The results of the plastic and hardened property tests are shown in Tables 4–6
for the control, 3.5 %, and 7 % sand mixtures, respectively. Table 7 shows the
results of several two-tailed, paired t-tests at the 90 % confidence level to deter-
mine if the results of the plastic and hardened properties differed significantly
for each of the three mixture designs. Using a t-critical value of 1.729, the
results of all plastic and hardened tests were found to differ significantly for all
three mixtures.
TABLE 4—Control mixture results.
ASTM C1688 Average C39 Core

Density (Unit Weight), ASTM C1688 ASTM D7063 Compressive Strength,
Batch kg/m3 (lb/ft3) Void Content (%) Effective Air Voids (%) MPa (psi)
1 2009 (125.4) 19.5 32.3 14.3 (2070)
2 2016 (125.8) 19.2 33.6 15.0 (2170)
3 2005 (125.2) 19.6 32.1 11.8 (1710)
4 2004 (125.1) 19.7 32.9 12.1 (1760)
5 2008 (125.4) 19.5 33.3 12.8 (1860)
6 2011 (125.5) 19.4 32.0 13.2 (1920)
7 2012 (125.6) 19.4 32.1 15.4 (2230)
8 2008 (125.3) 19.5 31.1 14.8 (2150)
9 2012 (125.6) 19.4 31.7 15.5 (2250)
10 2013 (125.7) 19.3 32.1 15.5 (2250)
11 2015 (125.8) 19.2 32.0 15.0 (2180)
12 2021 (126.2) 19.0 31.5 14.2 (2060)
13 2012 (126.6) 19.3 31.4 15.3 (2220)
14 2022 (126.2) 18.9 31.3 14.8 (2150)
15 2012 (125.6) 19.4 32.1 14.1 (2050)
16 2018 (126.0) 19.1 32.6 14.8 (2150)
17 2005 (125.2) 19.6 31.9 12.9 (1870)
18 2015 (125.8) 19.2 32.2 14.6 (2120)
19 2005 (125.1) 19.6 31.8 15.0 (2170)
20 2007 (125.3) 19.6 32.3 15.0 (2180)
TABLE 5—3.5% Sand mixture results.
ASTM D7063 Average C39 Core

ASTM C1688 Density ASTM C1688 Effective Compressive Strength,
Batch (Unit Weight), kg/m3 (lb/ft3) Void Content (%) Air Voids (%) MPa (psi)
1 2069 (129.2) 16.8 30.3 16.1 (2330)
2 2055 (128.3) 17.4 30.7 16.4 (2380)
3 2055 (128.3) 17.4 31.3 17.2 (2500)
4 2055 (128.3) 17.4 31.5 16.1 (2340)
5 2065 (128.9) 17.0 31.2 16.0 (2320)
6 2065 (128.9) 17.0 30.9 15.7 (2280)
7 2054 (128.2) 17.4 31.0 16.2 (2230)
8 2035 (127.0) 18.2 31.8 16.3 (2350)
9 2032 (126.8) 18.3 31.4 15.7 (2270)
10 2046 (127.7) 17.8 31.4 16.7 (2420)
11 2051 (128.0) 17.5 31.2 16.7 (2420)
12 2044 (127.6) 17.8 32.2 15.9 (2310)
13 2061 (128.7) 17.1 31.1 16.6 (2410)
14 2052 (128.1) 17.5 32.1 16.1 (2330)
15 2052 (128.1) 17.5 31.1 16.1 (2330)
16 2045 (127.6) 17.8 31.3 15.8 (2290)
17 2044 (127.6) 17.8 31.8 15.7 (2280)
18 2046 (127.7) 18.4 31.7 15.9 (2310)
19 2050 (128.0) 17.6 31.6 16.1 (2340)
20 2040 (127.3) 18.0 31.8 16.4 (2380)
Analysis of Results
The results of the statistical analysis of plastic and hardened property tests are
shown in Table 8. The ASTM C1688 allowable range for 10 samples was
computed using single-operator standard deviation, 22-kg/m3 (1.4-ft3), and the
largest multiplier, 4.5 for 10 test samples, was used to compute the maximum
acceptable range from ASTM C670-03 Table 1 [31]. Throughout the paper, the
term allowable range has been used to indicate the range that should be
expected to be exceeded only once in 20 times. The actual range for each mix-
ture was within the allowable range limit. However, Table 9 shows that the
mixtures containing fine aggregate had actual ranges much higher than that of
the control mixture and used much more of the allowable range. In the past,
TCA has used fine aggregate contents by volume of total aggregate as high as
10 %. If time and materials had been available, it would have been interesting
to see if a 10 % sand mixture could meet current maximum allowable range
criteria. Perhaps the ASTM subcommittee should conduct such an experiment
and consider revising the maximum allowable range criteria.
The mean values, actual ranges, standard deviations and coefficients of vari-
ation for ASTM C1688 voids for all three mixtures are also shown in Table 8.
TABLE 6—7% Sand mixture results.
ASTM C1688 Density Average C39 Core

(Unit Weight), ASTM C1688 ASTM D7063 Effective Compressive Strength,
Batch kg/m3 (lb/ft3) Void Content (%) Air Voids (%) MPa (psi)
1 2084 (130.1) 16.5 29.9 18.3 (2660)
2 2091 (130.5) 16.2 29.6 18.8 (2720)
3 2089 (130.4) 16.3 29.0 17.7 (2560)
4 2105 (131.4) 15.6 29.3 18.0 (2610)
5 2115 (132.0) 15.2 29.0 17.4 (2530)
6 2115 (132.0) 15.2 30.4 18.3 (2660)
7 2118 (132.2) 15.1 28.1 17.9 (2590)
8 2112 (131.8) 15.4 28.7 18.5 (2680)
9 2123 (132.5) 14.9 28.7 18.5 (2680)
10 2111 (131.8) 15.4 28.1 18.4 (2670)
11 2107 (131.5) 15.5 27.9 17.8 (2580)
12 2129 (132.9) 14.6 28.0 18.8 (2720)
13 2121 (132.4) 15.0 27.9 17.7 (2570)
14 2117 (132.1) 15.2 28.1 17.8 (2580)
15 2093 (130.6) 16.1 28.2 18.8 (2720)
16 2101 (131.1) 15.8 28.2 18.8 (2720)
17 2125 (132.7) 14.8 29.0 17.7 (2570)
18 2090 (130.4) 16.2 29.3 18.4 (2670)
19 2122 (132.5) 14.9 29.3 18.4 (2670)
20 2118 (132.2) 15.1 30.0 17.2 (2490)
Table 9 shows that the mixtures containing fine aggregate had actual void ranges
much higher than that of the control mixture, similar to the behavior of ASTM
C1688 density ranges. The authors assume that because the coefficients of varia-
tion are all below 4 %, data precision was good. The mean void percentages dif-
fer between ASTM C1688 and ASTM D7063 test methods because of the
different compactive efforts applied. A much greater compactive effort was used
in ASTM C1688. The cores that were used in the ASTM D7063 standard test
TABLE 7—Paired t test results.
Mixture 3.5% Sand 7% Sand

Control ASTM C1688 density 15.26 ASTM C1688 density 29.85
ASTM C1688 voids 15.62 ASTM C1688 voids 29.63
ASTM D7063 voids 3.53 ASTM D7063 voids 18.89
ASTM C39 strength 6.91 ASTM C39 strength 15.57
3.5% Sand ASTM C1688 density 13.49
ASTM C1688 voids 12.44
ASTM D7063 voids 11.28
ASTM C39 strength 12.15
TABLE 8—Statistical analysis results.
Test Parameter Control 3.5% Sand 7% Sand

ASTM C1688 Mean value 2011.5 (125.56) 2050.8 (128.01) 2109.2 (131.66)
density Actual range 18 (1.1) 37 (2.4) 45 (2.8)
(unit weight), ASTM single 22 (1.4) 22 (1.4) 22 (1.4)
kg/m3 (lb/ft3) operator
standard deviation
Allowable range 99 (6.3) 99 (6.3) 99 (6.3)
Within allowable Yes Yes Yes
Standard deviation 5.35 (0.33) 9.76 (0.61) 13.79 (0.86)
Coefficient of 0.3 0.5 0.7
variation (%)
ASTM C1688 Mean value 19.37 17.59 15.45
void content (%) Actual range 0.8 1.6 1.7
Standard deviation 0.22 0.44 0.56
variation (%)
ASTM D7063 Mean value 32.12 31.37 28.84
effective air Actual range 2.5 1.9 2.5
Voids (%) ASTM single 0.35% 0.35% 0.35%
operator
standard deviationa
Allowable rangea 1.575% 1.575% 1.575%
Within allowable No No No
Standard deviation 0.63 0.47 0.77
variation (%)
Average core Mean value 14.33 (2076) 16.19 (2348) 18.15 (2633)
compressive Actual range 3.7 (540) 1.5 (220) 1.6 (230)
strength, ASTM single 3.2% 3.2% 3.2%
MPa (psi) operator
coefficient of
variationb
Allowable rangeb 14.4% or 2.06 (299) 14.4% or 2.33 (338) 14.4% or 2.61 (379)
Within allowable No Yes Yes
Standard deviation 1.13 (164) 0.40 (58) 0.48 (70)
Coefficient of 7.9 2. 5 2.7
variation (%)
a
For hot-mix asphalt samples.
b
For higher strength cores of normal concrete.
method were compacted in pervious beam mold using the weight of the approxi-
mate 9-kg (20-lb) concrete rolling pin. This was a much less vigorous type of
compaction, which can be seen in the void percentages. However, the difference
in compaction is the only difference between ASTM C1688 voids and in-place
TABLE 9—Relative comparison of statistical analysis parameters.
Parameter Control 3.5% Sand 7% Sand

ASTM C1688 density (percent of control range) – 206 250
ASTM C1688 density (percent of allowable range) 18 37 45
ASTM C1688 density (percent of control standard deviation) – 182 258
ASTM C1688 density (percent of control COV) – 178 241
ASTM C1688 voids (percent of control range) – 200 213
ASTM C1688 voids (percent of control standard deviation) – 200 255
ASTM C1688 voids (percent of control COV) – 221 322
pervious PCC voids. Therefore, it stands to reason that there should be a strong
relationship.
The mean values, actual ranges, standard deviations, and coefficients of
variation for ASTM D7063 effective air voids and ASTM C39 compressive
strengths are also shown in Table 8. Most of the results, with the exception
of the control mixture compressive strengths, have coefficients of variation
less than 3 % and, therefore, are presumed to be good results. Allowable
range information is provided but is not particularly useful because in one
case the allowable range is for hot-mix asphalt samples and in the other is
for stronger normal PCC. The information is provided in the hope that it will
be useful in later development of allowable criteria for hardened pervious
PCC properties.
Table 10 shows relative changes in mean plastic and hardened properties
as a percent of the mean results of the control mixture from Table 8. Tables 8
and 10 confirm literature review findings that an increase in compressive
strength can usually be obtained by adding small amounts of fine aggregate.
The fine aggregate is able to add strength to the concrete by increasing the area
for the cement paste to adhere. As stated earlier, there are two ways to increase
compressive strength in pervious concrete. The first way was to increase the
cement paste binder area, and the other was to increase the actual strength of
the paste. However, care should always be exercised in using fine aggregate
substitution to increase compressive strength.
TABLE 10—Relative changes in mean plastic and hardened properties.
Parameter 3.5% Sand 7% Sand

Percent of control ASTM C1688 mean density 102.0 104.9
Percent of control ASTM C1688 mean voids 90.8 79.8
Percent of control ASTM D7063 mean voids 97.7 89.8
Percent of control ASTM C39 mean compressive strength 113.0 126.7
Conclusions
On the basis of the laboratory data and the preliminary analysis done, the fol-
lowing conclusions can be drawn:
1. ASTM D1688 voids appear to be strongly related to ASTM D7063
effective voids and compressive strength of pervious PCC cores.
2. Increasing fine aggregate and water content can increase the average
compressive strength of pervious PCC cores significantly.
3. Increasing fine aggregate and water content can decrease the ASTM
D7063 effective voids of pervious PCC cores significantly.
4. Increasing fine aggregate and water content can increase the variability
of both ASTMC 1688 density and voids substantially. A reconsideration
of the criteria for density may be needed.
Acknowledgments
The writers gratefully acknowledge the support of the Tennessee Concrete
Association and the Tennessee Technological University Department of Civil
and Environmental Engineering. In addition, the writers thank Jeff Holmes
and Perry Melton for their patience and skill in fabrication, maintenance, and
repair of the equipment. The writers appreciate the laboratory help provided
by Allen Browning, Aaron Crowley, Lindsay Bryant, Sarah Dillon, and Mar-
tin Medley.
References
[1] ASTM C1688-08, 2009, “Standard Test Method for Density and Void Con-
tent of Freshly Mixed Pervious Concrete,” Annual Book of ASTM Stand-
ards, Vol. 04.02, ASTM International, West Conshohocken, PA, p. 844.
[2] Medley, M. L., II, 2010, “Pervious Concrete Mixture Design by the Unit
Weight Method,” M.Sc. thesis, Tennessee Technological Univ., Cooke-
ville, TN 38505.
[3] Hendrix, J. P., 2011, “Utilizing TCA Mix Design and Adjustment Method
to Improve Pervious Concrete,” M.Sc. thesis, Tennessee Technological
Univ. Cookeville, TN 38505.
[4] ASTM C33/C33M-08, 2009, “Standard Specification for Concrete
Aggregates,” Annual Book of ASTM Standards, Vol. 04.02, ASTM Inter-
national, West Conshohocken , PA, pp. 1–12.
[5] Obla, K., “Pervious Concrete for Sustainable Development,” Proceedings
of the First International Conference in Washington, D.C. for NRMCA,
Rec. Adv. Concr. Technol. 2007.
[6] Kuennen, T., “Voids Add Value to Pervious Concrete,” Better Roads,
Vol. 73, No. 8, 2003, pp. 22–29.
[7] ACI 522R-10, Report on Pervious Concrete, American Concrete Institute,

Farmington Hills, MI, March 2010.
[8] Ghafoori, N. and Dutta, S., “Laboratory Investigation of Compacted No-
Fines Concrete for Paving Materials.” J. Mater. Civ. Eng.. Vol. 7, No. 3,
1995, pp. 183–191.
[9] Yang J. and Jiang, G. “Experimental Study on Properties of Pervious
Concrete Pavement Materials.” Cem. Concr. Res. Vol. 33, 2003. pp.
381–386.
[10] Kevern, J., Wang, K., Suleiman, M., and Schaefer, V., “Mix Design
Development for Pervious Concrete in Cold Weather Climates,” Proceed-
ings of the 2005 Mid-Continent Transportation Research Symposium,
Aug 2005, Ames, IA.
[11] Crouch, L. K., Cates, M., Dotson, J., Honeycutt, K., and Badoe, D. A.,
“Measuring the Effective Air Void Content of Portland Cement
Pervious Pavements,” Cement, Concrete, Aggregates, Vol. 25, No. 1,
June 2003.
[12] Huffman, D., “Understanding Pervious Concrete,” The Construction
Specifier, Construction Specifications Institute, Dec 2005.
[13] Meininger, R., “No-Fines Pervious Concrete for Paving,” Concrete Int..
Vol 10, No. 8, 1998, pp 20–27.
[14] Crouch, L. K., Smith, N., Walker, A., Dunn, T., and Sparkman, A.,
“Pervious PCC Compressive Strength in the Laboratory and the Field:
The Effects of Aggregate Properties and Compactive Effort,” NRMCA
Concrete Technology Forum: Focus on Pervious Concrete, May 2006.
[15] Tennis, P., Lemming, M., and Akers, D., “Pervious Concrete Pavements,”
EB302.02, Portland Cement Association, Skokie, IL, and National Ready
Mixed Concrete Association, Silver Spring, MD, 2004, 36 pp.
[16] Mahboub, K. C., Canler, J., Rathbone, R., Robl, T., and Davis, B.,
“Pervious Concrete: Compaction and Aggregate Gradation.” ACI Mater. J.,
2009, pp. 523–528.
[17] “Green in Practice 107 – Supplementary Cementitious Materials (SCMs),”
Technical Brief, Portland Cement Association, http://www.concretethinker.
com/technicalbrief/Supplementary-Cementitious-Materials.aspx.
[18] “Building Materials: Better Performance without Higher Cost,” Head-
waters Resources, http://flyash.com/flyashenvironment.asp.
[19] Bury, M., Mawby, C., and Fisher, D., “Making Pervious Concrete Place-
ment Easy Using a Novel Admixture System,” 2006 Concrete Technology
Forum, National Ready Mixed Concrete Association, 2006.
[20] McCain G. and Dewoolkar, M., “Porous Concrete Pavements: Mechani-
cal and Hydraulic Properties,” TRB 2010 Annual Meeting.
[21] ASTM C150-09, 2010, “Standard Specification for Portland Cement,” Annual
Book of ASTM Standards, Vol. 04.01, ASTM International, West Consho-
hocken, PA, p. 155.
[22] ASTM C618-08a, 2009, “Standard Specification for Coal Ash and Raw or
Calcined Natural Pozzolans for Use in Concrete,” Annual Book of ASTM
Standards, Vol. 04.02, ASTM International, West Conshohocken, PA, p.
331.
[23] ASTM C117-04, 2009, “Standard Test Method for Materials Finer than
75 lm (No. 200) Sieve in Mineral Aggregates by Washing,” Annual Book
hocken, PA, p. 65.
[24] ASTM C136-06, 2009, “Standard Test Method for Sieve Analysis of Fine
and Coarse Aggregates,” Annual Book of ASTM Standards, Vol. 04.02,
ASTM International, West Conshohocken , PA, p. 93.
[25] ASTM C127-07, 2009, “Standard Test Method for Density, Relative Den-
sity (Specific Gravity), and Absorption of Coarse Aggregate,” Annual
hocken, PA, p. 76.
[26] ASTM C192/C192M-07, 2009, “Standard Practice for Making and Curing
Concrete Test Specimens in the Laboratory,” Annual Book of ASTM
Standards, Vol. 04.02, ASTM International, West Conshohocken, PA,
p. 136.
[27] ASTM C42-04, 2009, “Standard Test Method for Obtaining and Testing
Drilled Cores and Sawed Beams of Concrete,” Annual Book of ASTM
Standards, Vol. 04.02, ASTM International, West Conshohocken, PA,
p. 32.
[28] ASTM D7063-05, 2010, “Standard Test Method for Effective Porosity
and Effective Air Voids of Compacted Bituminous Paving Mixture
Samples,” Annual Book of ASTM Standards, Vol. 04.03, ASTM Interna-
tional, West Conshohocken, PA, p. 918.
[29] ASTM C617-09, 2009, “Standard Practice for Capping Cylindrical Con-
crete Specimens,” Annual Book of ASTM Standards, Vol. 04.02, ASTM
International, West Conshohocken, PA, p. 326.
[30] ASTM C39-05, 2009, “Standard Test Method for Compressive Strength
of Cylindrical Concrete Specimens,” Annual Book of ASTM Standards,
Vol. 04.02, ASTM International, West Conshohocken, PA, p. 23.
[31] ASTM C670-03, 2009, “Standard Practice for Preparing Precision and
Bias Statements for Test Methods for Construction Materials,” Annual
hocken, PA, p. 355.
Pervious Concrete
STP 1551, 2012
DOI:10.1520/STP20120019
Heather J. Brown1 and Alan Sparkman2
The Development, Implementation,

and Use of ASTM C1701 Field Infiltration
of In Place Pervious Concrete
REFERENCE: Brown, Heather J. and Sparkman, Alan, “The Development,

Implementation, and Use of ASTM C1701 Field Infiltration of In Place Pervi-
ous Concrete,” Pervious Concrete on December 4, 2011 in Tampa, FL; STP
1551, H. J. Brown and M. Offenberg, Editors, pp. 69–79, doi:10.1520/
STP20120019, ASTM International, West Conshohocken, PA 2012.
ABSTRACT: In place pervious concrete begins its design life with an initial
infiltration rate that is in part determined by the mixture proportions and com-
pactive effort. This means that both the producer and the contractor influence
the final infiltration rate of the pervious concrete. It was this fact that led to
the development of a field based ASTM method to determine what pervious
concrete infiltration rate was being constructed. Other researchers have
investigated field infiltration to determine pollutant removal, hydrological per-
formance, clogging potential, and water quality. It was necessary to have a
test that could monitor a pervious concrete pavement over time in relation to
maintenance, because so many pervious projects had already been placed
across the United States. ASTM C1701 was adopted in June 2009, and engi-
neers are beginning to understand the usefulness of such a test method.
This paper focuses on the methodology, showcasing projects that have used
the test method and correlating existing pervious placements with current
infiltration data to determine a best practices design infiltration rate. Addition-
ally, an investigation into possible maintenance techniques is discussed for
situations in which a clogged placement is encountered.
KEYWORDS: pervious concrete, infiltration rate, permeability, porosity
Manuscript received March 1, 2012; accepted for publication September 7, 2012; published
1
Chair and Associate Professor, Concrete Industry Management, Middle Tennessee State Univ.,
P.O. Box 24, Murfreesboro, TN 37132, e-mail: heather.brown@mtsu.edu
2
Executive Director, Tennessee Concrete Association, Nashville, TN 37203,
e-mail: asparkman@tnconcrete.org
Copyright V
19428-2959.
69
Introduction
Pervious concrete became well known through many on-site demonstrations

and video clips of spectators watching water disappear into surface voids.
Since pervious concrete’s first unveiling, in Florida in the 1970s, many “lunch-
n-learns” have culminated with the visual impact of the pervious concrete
infiltration rate. It was after a surge of pervious concrete placement took place
that design and construction practices became more formal with the intro-
duction of ACI 522.1-08 [1], the Portland Cement Association’s Pervious
Concrete Pavements manual [2] and Hydrologic Design of Pervious Concrete
[3], and the formation of ASTM Subcommittee C09.49. Each of these placed
importance on quantifying infiltration as it relates to overall system perform-
ance and stormwater management. As important as it is to determine initial
infiltration, it is equally important to understand system exfiltration and storage
capacity for a long-lasting design.
As producers began to optimize mixtures based on strength, unit
weight, and porosity using existing ASTM standards, it became clear that
balancing strength and porosity in the field was critical for pervious con-
crete success. Field testing after contractor placement was not formalized
for pervious concrete. Instead, laboratories began using known methods
such as the falling head [4] and constant head permeability [5] test methods
to ensure that a certain void content was producing adequate infiltration
while achieving strength [6]. The infiltration rates were much greater than
anticipated, and thus no maximum or minimum infiltration rates were seen
in many pervious concrete designs. Testing for permeability using these
two methods was expensive and time consuming. The Florida Concrete &
Products Association designed a laboratory method that required wrapping
a cylinder with duct tape and creating a small lip to hold a 1/2 in. head of
water to determine permeability; this method yielded a rough estimate suffi-
cient for verification [7]. Because cylinder preparation was (and to date is
not) an ASTM standard, there was little confidence that any laboratory per-
meability method could correlate well to the material delivered to the site.
Additionally, infiltration is greatly influenced by contractor and equipment
placement methods. Early research noted that certain pervious concrete
mixtures worked best with low compactive effort equipment (e.g., roller
screed), whereas other mixtures could be compacted with a high density
paver if properly proportioned [8]. This compactive effort influence led
the ASTM C09.49 Committee to focus on a field test for infiltration to
accurately assess design storm capacities. Preliminary laboratory and field
testing began to reveal values of pervious concrete infiltration of 150 in./h
or higher [1]. Little research had attempted to determine a minimum or
maximum acceptable value; instead the focus was on strength, density, and
porosity.
BROWN AND SPARKMAN, doi:10.1520/STP20120019 71
ASTM C1701 Method Development
During the ASTM Committee method development, a review of the literature

revealed that several test methods were being used across the country. Early
permeability testing was developed into a version of the now adopted ASTM
C1701 employing a watering can, a stopwatch, a calculator, and a measuring
tape [9]. The advantages of this method were that it utilized off-the-shelf
equipment, was easy to perform, and indicated surface clogging potential.
Devices such as a double-ring infiltrometer [10] that rests on top of the con-
crete were considered because of their presence on job sites for determining
soil infiltration [11]. It was difficult to maintain a head in the outer ring with
these devices, as the infiltration rates were high in pervious concrete. This was
viewed as a disadvantage despite the very accurate measurements they pro-
duced with vertical surface infiltration.
The literature also revealed that an embedded single-ring infiltrometer
that could be installed either post- or pre-construction was being used to
determine the infiltration rates of pavement systems [11–13]. The option of
post-construction installation was important because an owner might
want multiple measurements taken of a larger site. This method was not
pursued by the ASTM C09.49 Committee because it was destructive to the
pavement and because of the expense of coring a section and filling in areas
after installation of the ring. The main obstacle the committee faced was
choosing a method that combined all the advantages and simulated
actual rain conditions with the movement of water through the system. It
was perceived that this could not be done with a surface test, but the com-
mittee strived to keep the test simple and flexible with good operator
repeatability.
The ASTM D3385-03 theory, along with the promoted single-ring method,
led the committee to utilize a rigid surface ring of known diameter along with
a known volume of water. A timed event was used for the calculation to get an
infiltration rate that would exceed the design storm rate in inches per hour as
commonly used in civil site designs. To address concerns about the lateral
movement of water from a surface-only test method, several committee mem-
bers ran repeatability tests at their locations with varied heads of water. It was
determined that a head of 0.5 in. was easy to maintain by the operator and did
not promote excessive lateral movement of water, which would give a false
sense of infiltration of the system. Lateral movement of water is present during
testing whether the surface is clogged or not. A visual lateral movement at the
surface can be a helpful tool in determining whether possible sedimentation is
occurring at the surface. Research has shown that the type of clogging in a
pavement is directly related to the type of sediment present at the site [14].
This will dictate the type of remediation measures that must be performed in
order to clean the pavement.
FIG. 1—Ring dimensions.
Procedure
ASTM C1701, “Test Method for Determining Permeability of Field Placed
Pervious Concrete Pavements,” was published in late 2009, and this section is
meant to summarize the method for further discussion [15]. A rigid ring made
of plastic or metal should be chosen following the dimensions in Fig. 1. A con-
tainer to hold water and a stopwatch will be needed for the testing. Prewetting
is done to adjust for the capillary action of the system. Two “runs” are made
subsequently with a chosen volume of water (1 or 5 gal, depending upon the
time of the prewet). A 1 gal test is used for slower draining pavements, and a
5 gal test is used for faster draining pavements, so that the operator can main-
tain the 0.5 in. head of water easily throughout the timing portion of the
procedure.
Precision and Bias Testing

In order for ASTM C1701 to be published, a precision and bias study
must be conducted. A single operator study was completed on a hybrid
180 000 ft2 commuter lot at Middle Tennessee State University (MTSU),
and there are plans for an inter-laboratory precision statement to be com-
pleted by 2014. Figure 2 illustrates the lot and the three test locations.
Tables 1 and 2 show a sample of the data collected at one test location and
statistics for the three locations from the study, respectively. A measure of
repeatability was obtained by making two replicate measurements at three
locations on a newly placed pervious concrete pavement. The replicate
measurements were repeated daily from an age of 1 day to an age of
10 days. The single-operator coefficient of variation of the infiltration rate
at one test location was found to be 4.7%. This value has been published
in the method for a single operator. The temperature did vary on each
test day, and this has been shown to cause fluctuations in the infiltration
results.
FIG. 2—MTSU commuter lot for single operator precision.
TABLE 1—MTSU commuter lot on days 1–3: data at one test location.
Day Runs after Prewet Time, s Volume, gal Diameter Infiltration Rate, in./h
1 Run 1 25.6 5.0 11.75 1495.7
Run 2 22.6 5.0 11.75 1695.3
2 Run 1 23.8 5.0 11.75 1612.7
Run 2 25.0 5.0 11.75 1532.8
3 Run 1 20.3 5.0 11.75 1893.7
Run 2 21.4 5.0 11.75 1794.5
TABLE 2—Repeatability statistics for three test locations on the MTSU commuter lot.
Average Repeatability Standard Deviation Repeatability Limit
Material X sr R
Location 1 1600.9 141.6 396.5
Location 2 2022.1 59.3 166.1
Location 3 279.0 18.8 52.5
Case Studies
ASTM C1701 is used for determining the initial field infiltration of a pervious
concrete placement. This value is to be not accepted or rejected by a design
engineer but simply noted by the owner as a starting point for infiltration.
Research has shown that up to 2000 in./h can be achieved while still optimizing
density, porosity, and strength [16]. The American Concrete Institute and the
Portland Cement Association reference a minimum of 150 in./h as a design
limit, but this has been far exceeded in designs in the most recent years. This
wide range of field results has not been formally introduced to the industry as
any form of acceptance. This method can assist in that development by track-
ing initial infiltration and the potential decline of that value over time. If a
parking area shows signs of sedimentation, then an ASTM C1701 test can pin-
point the problem areas, and they can be remedied, thus avoiding an excessive
loss of permeability. At the same time, ASTM C1701 can address more serious
issues of clogged areas and aid in determining whether the maintenance prac-
tices being used are effective. The following case studies represent a range of
issues that utilized this method for determining success or failure of the pro-
posed solution.
McCabe Park
The McCabe Park pervious lot was constructed in 2010 and had received heavy
sedimentation from an unseeded hillside onto an isolated section of parking
stalls. The Tennessee Concrete Association was asked to perform maintenance
on the lot after the hillside had been dressed appropriately. The data in Table 3
TABLE 3—McCabe Park before and after permeability measurements using ASTM C1701.
McCabe Pervious Lot, Ball Park Side Date July 21, 2011
Right-side Parking Stall Water, gal Time of Test, s Inches per Hour
Initial condition 1.0 1500.0 4.9
After cleaning 1.0 39.2 185.5
24-hour testing 1.0 35.0 207.8
Right Side, Lower Middle
24-hour testing 1.0 273.0 26.6
1.0 155.0 46.9
Left Side, Upper End
24-hour testing 5.0 135.0 269.4
FIG. 3—Clogged parking stall.
represent the initial permeability of the clogged lot and the final permeability
after cleaning was performed. This site utilized the Ditch Witch FX-30 (Figs. 3
and 4) with a special vacuum head attachment for pervious concrete. It is
shown in each case that the area was extremely clogged and that thorough
cleaning led to definite improvement of the permeability.
FIG. 4—Cleaning with Ditch Witch FX-30.

Big Box Store Study

This case study showcases how new pervious concrete construction can be
perceived as failing when the overall infiltration of the lot is more than
adequate to handle an isolated clogged area. Traditional surface parking lots
are designed with gradual slopes directed toward inlet structures for storm-
water management. This particular project was originally designed for a tra-
ditional asphalt surface and then was flipped to pervious concrete after the
project was bid on and started. The grading of the lot had been done, so the
sloped design and the introduction of underground utilities did not change
midstream in the project. A hybrid design of asphalt drive lanes and pervious
concrete in the back half of the stall was installed. Over the course of com-
pleting the project and getting the final contract closed out, rain events
occurred that caused massive movement of water across isolated stalls near
inlet structures. Suspended solids in the stormwater were forced across two
stalls on a routine basis during all rain events. It did not take long for pond-
ing to be evident on these two stalls. An ASTM C1701 test was performed,
and each stall showed 0 in./h infiltration.
These clogged areas raised red flags suggesting that possibly the pervious
concrete was not installed correctly. The data in Table 4 show that despite the
two clogged parking stalls next to the traditional inlet structure, the rest of the
lot was performing well above minimum suggested design infiltration rates.
The stalls directly next to the clogged stalls were tested to prove that the over-
all performance of the lot was intact and the isolated areas were going to con-
tinue to be a maintenance issue throughout the life of the parking lot. This
finding was important, as it notified the owners that if left unattended, the
clogged area would increase over time, causing a larger problem.
TABLE 4—ASTM C1701 infiltration values for parking stalls next to clogged stalls.
Location Stall 4, South Side
Test 1 Water, lb Time, s Inches per Hour

Prewet 8.0 30.0 255.6
1 40.0 129.0 297.3
2 40.0 110.0 348.6
Location Stall 7, South Side
Test 2 Water, lb Time, s Inches per Hour
Prewet 8.0 18.0 426.1

1 40.0 90.0 426.1
2 40.0 85.0 451.1
TABLE 5—MTSU commuter lot: Comparison of ASTM C1701 values over time.
Location 2008 Infiltration Rate, in./h 2011 Infiltration Rate, in./h

1 1709 1189
2 1863 910
3 286 11
MTSU Commuter Lot

The MTSU Commuter Lot was used for the initial repeatability study in 2008. It
was constructed in 2006 and has received moderate traffic to date. A follow-up
ASTM C1701 test was run in November 2011 and showed an interesting trend
that the authors are beginning to see in certain pervious designs. The same three
spots that were tested for repeatability in 2008 were tested in 2011 because they
had been permanently marked (Table 5). Locations 1 and 2 were approximately
50 % less permeable after nearly 4 years of service. It should be noted that with
no maintenance, this is a remarkable permeability rate. Location 3 was nearly
completely clogged with only 11 in./h of infiltration. It is the opinion of the
authors that screenings from asphalt drive lanes are a source of heavy sedimenta-
tion over time. Several projects that utilize the hybrid design of asphalt drive
lanes with pervious concrete stalls are introducing a source of fines that become
lodged in the top layer of pervious concrete. A wearing away of the asphalt sur-
face course creates this situation, and because the lanes are crowned to drain to-
ward the pervious concrete, it is a condition created by the design. This
particular area also has the lowest elevation of the entire lot, so it receives addi-
tional surface runoff in large rain events with heavy sedimentation. It is noted
that the 2008 value is only 286 in./h after 2 years of service; an initial infiltration
of greater than 500 in./h would be more appropriate. The recommendation for
this lot is that it be routinely street-swept and vacuumed annually.
Conclusions
ASTM C1701 can be utilized in the early quantification of pavement perme-
ability performance and then subsequently used as a measure for possible
remediation. Mixture optimization can be achieved by comparing permeability
with both the void percentage and the fresh density of the mixture. Aged pervi-
ous pavements can utilize the method to determine trouble areas that need
localized cleaning. As the method is specified for determining new pavement
permeability, the data captured will allow the industry to better understand
long-term pavement performance in terms of clogging and maintenance. The
method is not intended to be used as an acceptance criterion for the pavement,
as the range of permeability in successful projects runs from 150 in./h to over
2000 in./h.
References
[1] ACI 522.1-08, 2008, “Specification for Pervious Concrete,” American

Concrete Institute, Detroit, MI.
[2] Tennis, P. D., Leming, M. L., and Akers, D. J., Pervious Concrete Pave-
ments, Portland Cement Association, Skokie, IL, 2004.
[3] Leming, M. L., Malcom, H. R., and Tennis, P. D., Hydrologic Design
of Pervious Concrete, EB303, Portland Cement Association, Skokie, IL,
and National Ready Mixed Concrete Association, Silver Spring, MD,
2007.
[4] ASTM D2434-00, 2000, “Standard Test Method for Permeability of
Granular Soils (Constant Head),” Annual Book of ASTM Standards,
[5] ASTM D5084-90, 1990, “Standard Test Method for Measurement of Hy-
draulic Conductivity of Saturated Porous Materials Using a Flexible Wall
Permeameter,” Annual Book of ASTM Standards, Vol. 4.08, ASTM
[6] Crouch, L. K., Smith, N., Walker, A. C., Dunn, T. R., and Sparkman, A.,
Determining Pervious PCC Permeability With a Simple Triaxial Flexible-
Wall Constant Head Permeameter, Tennessee Concrete Association,
Nashville, TN, 2006.
[7] Florida Concrete & Products Association, Inc. (FCPA), Portland Cement
Pervious Pavement Manual, FCPA, Orlando, FL, 1990.
[8] Crouch, L. K., Smith, N., Walker, A. C., Dunn, T. R., and Sparkman, A.,
Pervious PCC Compressive Strength in the Laboratory and the Field:
The Effects of Aggregate Properties and Compactive Effort, Tennessee
Concrete Association, Nashville, TN, 2006.
[9] Youngs, A., “Pervious Concrete: It’s for Real,” Pervious Concrete and
Parking Area Design Workshop, Omaha, NE, 2005.
[10] ASTM D3385-09, 2009, “Standard Test Method for Infiltration Rate of
Soils in Field Using Double-Ring Infiltrometer,” Annual Book of
ASTM Standards, Volume 4.08, ASTM International, West Consho-
hocken, PA.
[11] Bean, E. Z., Hunt, W. F., Bidelspach, D. A., and Smith, J. T., Study on the
Surface Infiltration Rate of Permeable Pavements, Interlocking Concrete
Pavers Institute, Washington, D.C., 2004.
[12] Chopra, M., Wanielista, M., Spence, J., Ballock, C., and Offenberg, M.,
Hydraulic Performance of Pervious Concrete Pavements, Stormwater
Management Academy, University of Central Florida, Orlando, FL,
2006.
[13] Haselbach, L., Valavala, S., and Montes, F., “Permeability Predictions for
Sand-Clogged Portland Cement Pervious Concrete Pavement Systems,”
J. Environ. Manage., Vol. 81, No. 1, 2006, pp. 42–49.
[14] Mata, L. A., Sedimentation of Pervious Concrete Pavement Systems,

North Carolina State University, Raleigh, NC, 2008.
[15] ASTM C1701-09, 2009, “Test Method for Determining Permeability of
Field Placed Pervious Concrete Pavements,” Annual Book of ASTM
[16] Kevern, J. and Montgomery, J., “Hitting the Mark,” Concr. Int., Vol. 32,
No. 3, 2010, pp. 45–48.
Pervious Concrete
STP 1551, 2012
DOI:10.1520/STP104555
Matthew Offenberg1
Development of a New Test Method for

Assessing the Potential Raveling Resistance
REFERENCE: Offenberg, Matthew, “Development of a New Test Method for

Assessing the Potential Raveling Resistance of Pervious Concrete,” Pervious
Concrete on December 4, 2011 in Tampa, FL; STP 1551, H. J. Brown and
M. Offenberg, Editors, pp. 80–96, doi:10.1520/STP104555, ASTM Interna-
tional, West Conshohocken, PA 2012.
ABSTRACT: One of the key concerns with pervious concrete is the material’s
surface durability, specifically, its resistance to raveling. As the market for
pervious concrete grew, this was one of the hurdles in the way of broader
adoption of the technology. This paper documents the process of developing
a test method to determine the potential raveling resistance of a pervious
concrete mixture. The process included a study that used lab cast cylinders
to compare the raveling resistance potential of pervious concrete mixtures
containing different aggregates, varying cement contents, and basic chemi-
cal admixtures. A refined version of the test method was developed after an
unsuccessful ASTM round robin evaluation. The results from this new
method will provide the industry with beginning correlations between basic
mix ingredients and the surface durability of a finished pervious concrete
pavement.
KEYWORDS: pervious concrete, raveling, test
Introduction
In the United States, pervious concrete use is growing rapidly in the paving
and stormwater markets. This technology is not new; it has been in use for over
30 years in pavement and for almost 100 years in structural materials.
Manuscript received November 17, 2011; accepted for publication April 23, 2012; published
1
Technical Service Manager, Civil Engineering, W. R. Grace & Company, 6606 Marshall
Boulevard, Lithonia, GA 30058.
Copyright V
19428-2959.
80
OFFENBERG, doi:10.1520/STP104555 81
However, as the adoption of this technology has accelerated, the engineering

community has requested more assurance that the material will be durable
through the pavement’s entire design life. Unfortunately, the standard test
methods for the quality control of plain concrete are not applicable to pervious
concrete, so new test methods must be developed.
The most common cause of failure in pervious concrete is surface dete-
rioration. The key concern, then, is the material’s surface durability, specifi-
cally, its resistance to raveling. Even though the technology has advanced in
mature market areas so as to reduce raveling in light duty applications, this
is still the primary hurdle preventing broader acceptance in new markets
and heavier traffic loads. This project began as a study of lab cast cylinders
intended to compare the potential raveling resistance of pervious concrete
mixtures using different aggregates, varying cement contents, and basic
chemical admixtures. After its initial introduction to the industry in 2008, it
was brought before ASTM for standardization. This paper describes the
process to date.
The Problem
The U.S. Federal Highway Administration uses the following definition of
“raveling” [1]:
“Raveling—the wearing away of the pavement surface caused by the
dislodging of aggregate particles.”
Technically speaking, a pervious concrete pavement can be thought of
as a prismatic mass of aggregate, with each particle having a coating of ce-
mentitious paste that bonds it to adjacent coated aggregate particles. As a
paste bridge forms between the particles’ coatings, the paste should be fluid
enough that there can be no determination of where the coating of one parti-
cle begins and that of the other ends; the paste should be cohesive enough
to form a smooth shoulder (Fig. 1). The coating of paste on each particle
should be thick enough to give the concrete structural strength, yet thin
enough to allow the voids in the mass to remain open, even under compac-
tion. In order for a pavement to ravel, then, excluding aggregate failure,
there must be sufficient stress imparted on the coated aggregate to cause the
shoulder (paste bridge) to fracture completely with each adjacent aggregate
particle. Sufficient stress could be caused by a large load, a weak paste,
or a small paste bridge (the area of contact between adjacent aggregate
particles).
Fundamentally, the paste strength is one of the key components in raveling
performance. Compressive strength testing can give some indication of the rel-
ative performance in terms of raveling resistance of pervious concrete mixtures
utilizing coarse aggregate from a given aggregate source but with varying paste
FIG. 1—Close-up image of an ideal shoulder (paste bridge) on a cut and pol-
ished section of pervious concrete with 3/8 in. nominal maximum sized coarse
aggregate.
contents. Deo and Neithalath [2] reported higher compressive strengths in mix-
tures with incrementally higher paste contents and correspondingly lower void
contents. A similar increase in raveling resistance could also be expected for
this series of mixtures as the paste content increased. This team also reported
that larger aggregates cause greater variability in compressive strength,
whereas smaller aggregate sizes generally result in a more homogeneous
material structure as compared to specimens with larger aggregate, resulting
in more repeatable stress–strain responses in laboratory replicates. The
increasing degree of heterogeneity with increasing aggregate size was
demonstrated in both visual and mechanical measurements of the pervious
concrete mixtures studied. Similar performance could be expected in terms of
raveling resistance: mixtures with larger aggregates would tend to show
more variability in the laboratory and the field than mixtures with smaller
aggregates.
Raveling is traditionally thought of as occurring due to issues with batch-
ing, handling, or curing. Unfortunately, pervious concrete pavements can ravel
even if batched with sufficient mix water, handled properly, and properly
cured. Therefore, a test is needed that can identify the mixture proportions that
enhance surface durability through raveling resistance. This test will help con-
crete producers use the best possible ingredients, help contractors understand
the best possible construction techniques, and help researchers pave the way to
advanced pervious concrete applications.
Test Concepts
In working to solve this problem, the team worked through several criteria that
the test should fit.
The test should use existing, readily available equipment.
The specimens should be easy to fabricate.
The test specimens should be as light as is practical.
The test should be able to be performed in the lab or the field.
Lab cast specimens should be representative of field conditions.
For wide adoption, the test should be quick and easy to run.
The test should be repeatable.
Several existing tests were considered.
ASTM C944 (“Standard Test Method for Abrasion Resistance of Concrete
or Mortar Surfaces by the Rotating-Cutter Method”) [3] is a standard test for
concrete abrasion resistance and was ruled out because it did not test a repre-
sentative sample area and is not readily available.
ASTM C779 (“Standard Test Method for Abrasion Resistance of Horizon-
tal Concrete Surfaces”) [4] is another test for concrete abrasion resistance and
was not chosen because the specimen would be too heavy.
ASTM E303 (“Standard Test Method for Measuring Surface Frictional
Properties Using the British Pendulum Tester”) [5] is the British pendulum test
for skid resistance. The team liked the concept of a swinging pendulum as a
tool that could be used to dislodge aggregate, but they decided this pendulum
was not heavy enough to be efficient.
ASTM C39 (“Standard Test Method for Compressive Strength of Cylindri-
cal Concrete Specimens”) [6] is the standard test for the compressive strength
of concrete. Many of the factors that impact raveling resistance probably also
impact compressive strength, but the results of this test are impacted by more
than the surface properties alone.
ASTM C131 (“Standard Test Method for Resistance to Degradation of
Small-size Coarse Aggregate by Abrasion and Impact in the Los Angeles
Machine”) [7] is a test for measuring the abrasion resistance of aggregates,
but the apparatus may be used to test raveling resistance if the test is not
too aggressive. The team decided this was the best apparatus to study.
Because the test will not measure actual field raveling, the properties
described have been termed “raveling potential” and “potential resistance to
raveling.”
Developing the Test

In ASTM C131 [7], a sample of aggregate is placed in a shelved steel chamber
along with a charge of steel spheres, and the chamber is rotated for a set num-
ber of revolutions. During the development of this test for pervious concrete,
all options were open for consideration and study. The team considered speci-
men size and geometry, the number of specimens to be put in the machine, the
number of revolutions for a test run, and the number of spheres to be used. The
test specimens were cast from a widely used pervious concrete mix in an
attempt to provide some basis for comparison to field performance. The raw
materials in the mixture included cement, a soft limestone coarse aggregate,
and a lignin-based water reducer.
For the first round of evaluation, a single 4 in. long, 4 in. diameter cylinder
was placed in the machine without any spheres. The specimen was weighed
before testing, and then a final mass was recorded after 50, 100, and 500 revo-
lutions of the machine. Any pieces of the original sample that were retained on
a 1 in. sieve were considered part of the final mass. The mass lost was calcu-
lated as a percentage of the original specimen mass. An identical specimen was
tested with a full complement of 12 spheres. A summary of the results can be
found in Table 1.
Visual observation of the specimens after each sequence of revolutions
showed that at 50 revolutions, there was little abrasion of the soft limestone
coarse aggregate. After 100 and 500 revolutions, the aggregate had been
abraded significantly. The intent of the test is not to test aggregate durability,
but to test the ability of the cement paste to hold on to the aggregate. Thus, it
was decided to set the test initially at 50 revolutions.
Further, based on these results, it appeared that the spheres had no impact
on the mix performance in this test. In ASTM C131, the spheres are included
in order to grind up the individual aggregate particles. Because the mechanical
TABLE 1—Data from first initial evaluation.
Revolutions Weight, g Individual Loss, g Individual Loss, % Total Loss, g Total Loss, %
0 Spheres
0 1413.8
50 1337.6 76.2 5.4 76.2 5.4
100 1301.2 36.4 2.6 112.6 8.0
500 1083.1 218.1 15.4 330.7 23.4
12 Spheres
0 1329.8
50 1264.8 65 4.9 65 4.9
100 1219.3 45.5 3.4 110.5 8.3
500 931.8 287.5 21.6 398 29.9
attrition of the aggregate particles that had raveled and broken loose was not of
interest in this test, the spheres were not needed. Therefore, spheres were elimi-
nated from further testing as an unnecessary element.
The next round of testing focused on varying the cylinder size and number.
Single and multiple 4 in. diameter, 8 in. tall cylinders were tested. The larger
cylinders broke in half in some tests (Fig. 2), skewing the results with larger
surface areas to ravel. Because of the variation in the mix proportions, this
round of testing was able to demonstrate a wide range of potential surface dura-
bilities, suggesting that the test might be able to detect resistance to raveling
potential spanning a broad scale. In addition, the team was able to confirm that
the aggregate was not being abraded.
Based on these rounds of testing, the team proposed the following guide-
lines for a test:
• Samples would be cylinders 4 in. long and 4 in. in diameter.
• Specimens would be cast to the mix design void content.
• Cylinders would cure for seven days in a covered, sealed cylinder mold
at laboratory temperature.
• Immediately after stripping, the cylinder would be weighed and
measured.
• The cylinder would be subject to 50 revolutions in the ASTM C131
chamber without any spheres.
FIG. 2—Final mass of 4 in. diameter, 8 in. tall cylinders tested as a pair. Note
the broken cylinder.
• The final weight would be calculated.

• The mass lost would be calculated as a percentage of the original speci-
men mass.
Expanded Testing Program

In order to test the effectiveness of the proposed procedure, a number of mix-
tures were tested with several variables. Research [8] has demonstrated that
field cast pervious concrete typically has a lower void content near the top
surface and a higher void content near the bottom. In an effort to reduce the
variability of the test, an attempt was made to cast specimens at the desired
void content with minimal variation in the vertical porosity distribution.
Unfortunately, ASTM did not yet have a standardized process for casting per-
vious concrete specimens. To achieve this, the mass necessary to achieve that
void content was weighed and placed into a standard 4 in. 8 in. single-use
cylinder mold. The cylinder was dropped from a 1 in. height ten times and then
compacted manually, using a gyratory motion, with a plastic cylindrical tool 3
in. in diameter and 6 in. in height until the cylinder was 4 in. tall. The gyratory
motion was a manual attempt to mimic the compactive effort mechanically
induced by a gyratory compactor as described in ASTM D6925 [9]. A plastic
bag was immediately placed over the cylinder, which was allowed to cure
uninterrupted for seven days at constant laboratory temperature.
Testing was conducted with a number of variables, including aggregate
type, void content, fly ash addition, and water/cement ratio (Figs. 3–5). Each
series of tests demonstrated the anticipated range of results. In most series, as
the void content increased, the potential for raveling resistance decreased. In
FIG. 3—Results of preliminary experiments at 50 revolutions to show proof of

concept.
FIG. 4—Detail of low mass loss results.
the series demonstrating water/cement ratios from 0.25 to 0.34, the mass loss
did not change.
After reviewing the results from these series, it appears that the test method
works effectively for demonstrating a significant span of raveling resistance
potential.
The method, as performed on a cast cylinder, will yield different results
than field installations, as contractors’ methods [10] play an important role in
the final void content and surface durability.
ASTM Round Robin

The proposed test method was presented to ASTM Subcommittee C09.49 at
the December 2008 meeting. Following the presentation, a task group was
FIG. 5—Effect of water/cement ratio (w/c) on potential raveling resistance at

50 revolutions in proof of concept testing.
TABLE 2—Data including coefficient of variation of mass loss from each laboratory from the initial round robin.
Mix 1 Mix 2 Mix 3
Designed Void Average Mass Mass Loss Designed Void Average Mass Mass Loss Designed Void Average Mass Mass Loss
Lab Content, % Loss, % COV, % Content, % Loss, % COV, % Content, % Loss, % COV, %
1 20 5.06 15 25 10.59 40 30 29.57 59
2 20 8.77 20 25 12.54 12 30 12.43 14
3 20 7.05 20 25 12.79 32 30 11.44 12
4 15 8.31 23 22.5 10.63 16 30 15.05 26
5 20 8.92 9 25 9.16 4 30 10.09 9
6 20 10.81 37 25 15.62 31 30 12.73 52
7 20 8.41 16 25 8.43 25 30 9.92 19
8 15 6.88 26 22.5 7.95 29 30 13.21 16
9 20 8.69 52 25 14.07 19 30 46.97 28
10 20 12.80 29 25 14.73 18 30 65.08 32
11 20 9.42 48 25 19.62 15 30 14.25 53
12 20 4.49 27 25 12.36 13 30 18.28 38
13 18 6.64 13 20.5 7.65 13 23.9 10.94 12
Average 8.2 25.7 12.0 20.4 20.8 28.5
TABLE 3—Data from the initial round robin comparing variability in mass loss to variability in sample height.
Mix 1 Mix 2 Mix 3
Mass Loss Average Sample Sample Height Mass Loss Average Sample Sample Height Mass Loss Average Sample Sample Height
Lab COV, % Height, mm COV, % COV, % Height, mm COV, % COV, % Height, mm COV, %
1 14.8 101.5 1.9 39.6 107.0 0.5 59.4 107.1 0.6
2 19.8 103.7 0.0 11.9 103.3 0.1 13.9 101.4 0.0
3 20.3 101.3 0.0 31.8 101.2 0.0 11.7 101.2 0.0
4 23.3 108.6 0.5 16.2 104.4 1.4 26.2 103.1 2.0
5 8.7 101.2 0.1 3.9 101.5 0.1 9.3 101.4 0.1
6 37.1 111.9 2.7 30.5 113.5 2.3 51.9 103.5 2.1
7 15.9 101.3 0.1 24.9 101.3 0.1 18.9 101.3 0.1
8 25.5 106.6 2.8 28.6 103.7 3.1 16.4 103.5 1.0
9 52.2 109.6 2.1 18.6 110.1 1.2 28.2 105.0 1.3
10 28.5 110.7 3.9 18.0 108.3 1.2 32.2 105.0 1.3
11 47.9 112.2 2.4 14.8 111.6 0.9 52.6 105.0 3.1
12 26.8 101.0 3.5 13.0 105.7 1.9 37.8 105.8 1.8
13 12.8 106.4 0.6 13.2 105.9 1.1 11.5 105.7 1.2
Average 25.7 105.8 1.7 20.4 106.0 1.1 28.5 103.8 1.1
OFFENBERG, doi:10.1520/STP104555
89
FIG. 6—Impact of variability in sample height on variability in mass loss for

crushed and rounded aggregates.
formed to optimize the test procedure and verify it through a round robin evalu-
ation. The task group made one major change to the procedure: to compact the
specimen with a drop hammer. Teams compacted specimens separately with a
Marshall Hammer (ASTM D6926) [11] and a Proctor Hammer (ASTM D698)
[12] in order to evaluate the performance of each. Each participating lab cast
six specimens on each of three mixtures. Mixtures 1, 2, and 3 were representa-
tive of low, medium, and high void contents, respectively.
Table 2 shows a summary of the results from the first round on specimens
compacted with the Marshall Hammer. The full data set can be downloaded
from the WK23367 collaboration area at www.ASTM.org. The task group
agreed that with the coefficient of variability (COV) of the mass loss averaging
almost 25 %, the variability of the test was not adequate for standardization.
However, the task group did reach a consensus on the fact that the Marshall
Hammer was the most effective tool for compacting specimens. In addition to
casting nicely shaped cylinders, it allowed the operator to count the number of
blows that it took to compact the specimen to the 4 in. (100 mm) final height,
thus providing an indication of the workability of the mixture.
Although it was hypothesized that variability in specimen fabrication
would correlate well with variability in mass loss, such was not the case. Spe-
cifically, the height and mass of the specimens were of concern. After review-
ing the data from the first round, it was found that there was little correlation of
the COV of mass loss to the COV of specimen height (Table 3, Fig. 6) and the
COV of specimen mass (Table 4, Fig. 7). There was almost no variability in
specimen diameter, even among the cylinders cast in plastic single-use molds.
TABLE 4—Data from the initial round robin comparing variability in mass loss to variability in sample mass.
Mix 1 Mix 2 Mix 3
Mass Loss Average Sample Sample Mass Mass Loss Average Sample Sample Mass Mass Loss Average Sample Sample Mass
Lab COV, % Mass, g COV, % COV, % Mass, g COV, % COV, % Mass, g COV, %
1 14.8 1673 0.2 39.6 1584 0.3 59.4 1498 0.2
2 19.8 1575 0.7 11.9 1527 1.2 13.9 1429 1.2
3 20.3 1604 0.8 31.8 1496 0.7 11.7 1429 0.4
4 23.3 1706 0.5 16.2 1555 0.4 26.2 1402 0.9
5 8.7 1635 0.8 3.9 1577 0.1 9.3 1467 0.2
6 37.1 1488 0.3 30.5 1396 0.2 51.9 1312 0.7
7 15.9 1672 0.4 24.9 1576 0.2 18.9 1472 0.3
8 25.5 1713 0.8 28.6 1548 0.8 16.4 1408 0.7
9 52.2 1728 0.4 18.6 1650 0.3 28.2 1571 0.0
10 28.5 1732 0.0 18.0 1648 0.2 32.2 1571 0.1
11 47.9 1492 0.3 14.8 1400 0.2 52.6 1307 0.4
12 26.8 1671 0.3 13.0 1581 0.3 37.8 1501 0.2
13 12.8 1650 0.3 13.2 1613 0.8 11.5 1586 0.6
Average 25.7 1641 0.4 20.4 1550 0.4 28.5 1458 0.45
OFFENBERG, doi:10.1520/STP104555
91
FIG. 7—Impact of variability in sample mass on variability in mass loss for

crushed and rounded aggregates.
This was surprising, as some distortion of the cylinder mold was expected with
the drop hammer. The highest COV reported for the diameter was 0.68 %,
representing a standard deviation of 0.70 mm.
Further review of the data showed a clear difference in the variability of
rounded and crushed coarse aggregates. Crushed coarse aggregates tend to be
angular in shape as a result of the mechanical fracturing process, whereas grav-
els tend to be rounded from years of slow erosion in a waterway. The variabili-
ty in mass loss was higher for crushed materials, as was the variability in
sample height, but no direct correlation was found between these two. Further
research will be necessary in order to determine whether this relationship is
broadly observed or was an artifact of the limited data set.
Before proceeding, it was clear that the task group needed to modify the
procedure in order to reduce the variability of the mass loss. After reviewing
the procedure from the first round robin, the task group attempted several varia-
tions of the test, including testing single specimens at 50, 75, 100, 125, 150,
200, 300, and 500 revolutions and testing specimens in sets of three at 150,
500, 1000, and 1500 revolutions. After the data had been reviewed, it was
found that the most practical and most efficient protocol with the lowest rea-
sonable coefficient of variation (averaging less than 6 %) was to test cylinders
in sets of three at 500 revolutions with no steel spheres.
When examining the specimens and the remaining broken down mass after
this procedure, it appears that most of the mass loss comes from the impact of the
specimen after it has been dropped by the shelf within the rotating drum.
Therefore, aggregate particles that are well bonded tend not to break loose upon
impact. Conversely, aggregate particles that are loosely bonded tend to separate
easily. The variables that contribute to the robustness of a bond include the paste
strength, the paste thickness, points of contact with other aggregate particles, and
the diameter of the cement paste bridge at those points of contact.
Originally, 500 revolutions abraded the soft Florida limestone aggregate,
so this level of assessment was discounted as being too aggressive. However,
when the testing procedure was refined with more common, harder aggregates,
the increased revolution count did more to break loose aggregates than to cause
abrasion. Thus, the task group agreed to proceed with 500 revolutions as the
standard for the test.
In samples that showed high mass loss, there might have been some cush-
ioning of the drop impact of each cylinder by the powdery or fine-grained ma-
terial that had broken loose in previous revolutions. If so, this would tend to
reduce the mass loss of those poorer performing mixtures. Thus, results might
be skewed toward better performance. However, considering the range of mass
losses calculated (19.2 % to 94.6 %), the test demonstrates a broad range of
potential durability.
After the second round robin evaluation of this modified procedure
(Table 5), the coefficient of variation, averaging less than 4 %, was found to be
acceptable to the task group and subcommittee. The full data set from the second
round robin also can be downloaded from the WK23367 collaboration area at
TABLE 5—Data including coefficient of variation of mass loss from each laboratory from the second
round robin.
18 % Voids 20 % Voids 22 % Voids
Average Mass Mass Loss Average Mass Mass Loss Average Mass Mass Loss
Lab Loss, % COV, % Loss, % COV, % Loss, % COV, %
1 41.3 1.9 46.4 0.6 48.0 2.8
2 40.2 1.1 40.4 31 50.3 2.0
3 48.0 1.1 53.4 3.2 51.7 5.0
4 50.5 2.5 54.9 0.8 52.6 1.9
5 37.7 3.5 52.4 1.0 92.7 1.0
6 37.6 1.2 52.4 2.9 94.6 0.5
7 32.8 2.3 32.6 4.2 36.2 2.6
8 29.9 0.6 35.9 2.0 36.5 2.3
9 31.2 4.0 27.5 19 36.4 3.5
10 32.1 7.0 31.5 5.3 39.2 3.1
11 37.8 1.3 39.0 2.0 41.9 1.7
12 37.2 1.4 39.2 1.1 40.5 0.8
13 22.9 1.5 19.6 8.0 25.8 11
14 22.0 5.0 19.2 3.9 29.0 10
Average 35.8 2.5 38.9 6.0 48.2 3.4
TABLE 6—Sample data from one lab for one mix tested in second round robin.
Initial Weight, g Final Weight, g Mass Loss, %

Sample 1 5045 3400 32.6
Sample 2 5046 3351 33.6
Sample 3 5042 3424 32.1
Average 32.8
Standard Deviation 0.76
COV 2.32
www.ASTM.org. For this round of testing, each laboratory compacted three sets
of three cylinders for mixtures proportioned at 18 %, 20 %, and 22 % voids. An
example of the data submitted by a lab for a mix is shown in Table 6.
For this second assessment, teams were also asked to report the number of
blows each mixture required for compaction. This index might give an indica-
tion of the workability of the mixture. The data showed no correlation between
workability and raveling resistance; however, none was expected based on the
testing protocol. With more study, this might turn out to be a useful quality
control tool in the future, or a tool to help concrete producers optimize mixture
proportions for workability.
In the current version, the test method is intended only to help a concrete
producer or researcher select the best possible mixture proportions based on
potential raveling resistance [13] or to compare the potential raveling resist-
ance of different pervious concrete mixture proportions. Additionally, it is
intended for use only with coarse aggregates with a nominal maximum size
smaller than 1 in., as larger aggregate sizes have not yet been evaluated. Until
there are enough data to develop correlations between test results and field per-
formance, meaningful limits of acceptable results cannot be established, so this
test should not be used as part of a project specification.
The registered work item has worked its way through ASTM Subcommit-
tee C09.49 as of this writing and is proceeding through the fourth concurrent
main committee ballot.
Conclusion
The test method described effectively demonstrates a wide range of potential
raveling resistance for pervious concrete mixtures. The specimens are easy to
fabricate and easy to handle due to their small size. Because it recommends
that one count the number of blows required in order to compact the specimen,
the procedure provides an indication of the workability of the fresh concrete.
The short curing time mimics field-curing conditions and allows the test to pro-
duce results quickly. Based on the simple procedures for casting a specimen,
the test can be performed on laboratory-mixed or field-batched concrete.
There are some shortcomings to this test method, though. Specimens that
are cast with a paste phase that is too fluid, and thus are sealed on the top or
bottom, should not be tested. Additionally, no cores were tested in this project;
in order to establish good correlations to field performance, cores or other
methods of evaluating field raveling resistance will need to be evaluated.
Finally, this test method has come under criticism for not testing any real prop-
erties of pervious concrete. This could be debated at length; however, the data
in this study and others seem to correlate with field experience with pervious
concrete. Until the test method is in place and data can be gathered on the
impact the standard might have on the success or failure of projects, final con-
clusions cannot be drawn about the efficacy of the test.
References
[1] “Glossary - Distress Identification Manual for The LTPP (Fourth Revised
Edition),” June 2003 - FHWA-RD-03-031, http://www.tfhrc.gov/pavement/
ltpp/reports/03031/glossary.htm (Last accessed February 1, 2008).
[2] Deo, O. and Neithalath, N., “Compressive Behavior of Pervious Con-
cretes and a Quantification of the Influence of Random Pore Structure
Features,” Mater. Sci. Eng. A, Vol. 528, 2010, pp. 402–412.
[3] ASTM C944, 2008, “Standard Test Method for Abrasion Resistance of
Concrete or Mortar Surfaces by the Rotating-Cutter Method,” Annual
hocken, PA.
[4] ASTM C779, 2010, “Standard Test Method for Abrasion Resistance of
Horizontal Concrete Surfaces,” Annual Book of ASTM Standards, Vol.
[5] ASTM E303, 2008, “Standard Test Method for Measuring Surface Fric-
tional Properties Using the British Pendulum Tester,” Annual Book of
ASTM Standards, Vol. 04.03, ASTM International, West Conshohocken,
PA.
[6] ASTM C39, 2012, “Standard Test Method for Compressive Strength of
Cylindrical Concrete Specimens,” Annual Book of ASTM Standards, Vol.
[7] ASTM C131, 2006, “Standard Test Method for Resistance to Degradation
of Small-size Coarse Aggregate by Abrasion and Impact in the Los
Angeles Machine,” Annual Book of ASTM Standards, Vol. 04.02, ASTM
[8] Haselbach, L. M. and Freeman, R. M., “Vertical Porosity Distributions in
Pervious Concrete Pavement,” ACI Mater. J., Vol. 103, No. 6, 2006, pp.
452–458.
[9] ASTM D6925, 2009, “Standard Test Method for Preparation and Deter-
mination of the Relative Density of Hot Mix Asphalt (HMA) Specimens
by Means of the Superpave Gyratory Compactor,” Annual Book of ASTM

[10] Offenberg, M., “Producing Pervious Pavements,” Concr. Int., Vol. 27,
No. 3, 2005, pp 50–54.
[11] ASTM D6926, 2010, “Standard Practice for Preparation of Bituminous
Specimens Using Marshall Apparatus,” Annual Book of ASTM Standards,
[12] ASTM D698, 2012, “Standard Test Methods for Laboratory Compaction
Characteristics of Soil Using Standard Effort (12 400 ft-lbf/ft3
(600 kN-m/m3)),” Annual Book of ASTM Standards, Vol. 04.08, ASTM
[13] Offenberg, M., Frew, J., Seaman, S., Rafalko, L., and Lee, R.,
“Laboratory Evaluation of Coal Combustion By-products on Raveling
Potential of Pervious Concrete,” Proceedings of the 2011 International
Concrete Sustainability Conference, August 8–10, 2011, Cambridge MA,
National Ready Mixed Concrete Association, Silver Spring, MD.
Pervious Concrete
STP 1551, 2012
DOI:10.1520/STP104560
David R. Smith,1 Kevin Earley,2 and Justin M. Lia3
Potential Application of ASTM C1701

for Evaluating Surface Infiltration
of Permeable Interlocking Concrete
Pavements
REFERENCE: Smith, David R., Earley, Kevin, and Lia, Justin M., “Potential
Application of ASTM C1701 for Evaluating Surface Infiltration of Permeable
Interlocking Concrete Pavements,” Pervious Concrete on December 4, 2011
in Tampa, FL; STP 1551, H. J. Brown and M. Offenberg, Editors, pp. 97–105,
doi:10.1520/STP104560, ASTM International, West Conshohocken, PA
2012.
ABSTRACT: As a sister sustainable pavement to pervious concrete, perme-
able interlocking concrete pavement (PICP) has seen increased use for
stormwater management and low impact development. Surface infiltration is
a key performance indicator for both pavement types. This paper provides a
brief background of the development of test methods for measuring the sur-
face infiltration of permeable pavements. Among these test methods is the
single ring infiltrometer method described in ASTM C1701, which was devel-
oped to test the surface infiltration of PICP, concrete grid pavements, and
pervious concrete. Research literature references on surface infiltration test-
ing at sites in Long Island, NY, confirm that ASTM C1701 is suitable for
measuring the surface infiltration rate of PICP. The post-construction pave-
ment surface infiltration results there demonstrated an average rate of
1.4 103 m/s (200 in./h) or greater. Test results are also referenced from
U.S. Environmental Protection Agency surface infiltration testing that used a
modified version of ASTM C1701 at a permeable pavement research facility
in Edison, NJ, consisting of PICP, pervious concrete, and porous asphalt.
Manuscript received November 18, 2011; accepted for publication March 30, 2012; published
1
Technical Director, Interlocking Concrete Pavement Institute, 13921 Park Center Rd., Suite
270, Herndon, VA 20171, e-mail: dsmith@icpi.org
2
LEED Green Associate, M.S., Engineering Geology, Director of Commercial Sales, Nicolock
Paving Stones, 3025 Fairhill Dr., Collegeville, PA 19426, e-mail: kearley@nicolock.com
3
P.E., LEED AP BC þ D, President of 4Site Engineering, PLLC, 58 Janet St., Port Jefferson
Station, NY 11776, e-mail: jlia@4siteli.com
Copyright V
19428-2959.
97
Modifications to ASTM C1701 are proposed that include the use of (1) mod-
eling clay to seal the ring to the pavement in hot weather and (2) graduated
bucket(s) to determine the mass of infiltrated water. Graduations can elimi-
nate the use of a scale on the test site to determine the infiltrated water’s
mass. In addition, changes to ASTM C1701 are proposed to include PICP,
concrete grid pavements, and porous asphalt for comparative purposes.
KEYWORDS: permeable interlocking concrete pavement, permeable pave-
ment, pervious pavement, surface infiltration testing
Background
There are several ways to assess the infiltration of rainfall and runoff into per-
meable pavements. In ascending cost order, they include (1) visual inspection
during or immediately after a rainstorm for ponding; (2) measuring the surface
infiltration in sampled small areas; (3) generating synthetic rainfall and runoff
for a distinct rain event, which often involves a rain simulator; and (4) continu-
ous monitoring of rainfall and surface runoff, usually over a period of years.
The pervious concrete pavement industry elected to use method 2, i.e., sam-
pling and testing the surface infiltration rate of small areas across a larger pave-
ment for acceptance testing of newly constructed pavements and to assess
in-service surface infiltration. This is likely due to the speed and economy of
conducting tests in this manner.
PICP Surface Infiltration Testing with ASTM C1701

The method for the sampling and testing of small areas is similar to that
described in ASTM D3385 for measuring soil infiltration [1]. This test method
is intended for use on soils with infiltration rates of between 104 m/s (14 in./h)
and 108 m/s (0.001 in./h). In 2004, ASTM D3385 was employed by Bean et al.
[2] in an evaluation of the surface infiltration of permeable interlocking concrete
pavement (PICP), concrete grid, and pervious concrete pavement sites in NC,
VA, MD, and DE. They moved to a single ring infiltrometer because of the high
surface infiltration rates of the permeable pavements tested and the volume of
water required in order to maintain hydraulic heads within the rings.
Bean et al. tested 14 PICP sites using a single ring device sealed with
plumber’s putty to the pavement surface and measured values between
1.1 102 m/s (1575 in./h) and 4.4 106 m/s (0.625 in./h) on PICP. The low-
est values were to the result of clogging from fines. Values between 1.9 102
m/s (2756 in./h) and 3 105 m/s (4.3 in./h) were measured on pervious con-
crete, again with the lowest values being due to clogging from fines.
Bean et al.’s use of a single ring device is similar to that recommended by
ASTM C1701 [3], which involves a single ring infiltrometer sealed to the pave-
ment surface. Developed after Bean et al.’s research, this test method consists
of a 300 mm (12 in.) diameter ring, plumber’s putty used to seal the ring to the
SMITH ET AL., doi:10.1520/STP104560 99
pavement surface, water (typically provided in 19 l or 5 gal buckets), and a

stopwatch (found on most cell phones). The test method is recommended for
acceptance testing and in-service performance of PICP by the Interlocking
Concrete Pavement Institute [4]. A minimum infiltration rate acceptance for
new construction of 7 104 m/s (100 in./h) is recommended. The same rate is
recommended for acceptance testing of pervious concrete pavement in a New
York State Department of Transportation specification [5] and a draft specifica-
tion by the California Department of Transportation.
Bean et al. used ASTM C1701 on 14 PICP sites. In order to further demon-
strate the utility of this test method for PICP, results from two sites from 4Site
Engineering obtained in 2010 and 2011 are provided [6,7]. Both sites are in
Lindenhurst, NY, with the first being a 880 m2 (8800 ft2) parking lot at a
building materials supply store. The paving units were 120 mm 225
mm 80 mm (4.75 in 9 in 3.125 in.) thick with 13 mm (0.5 in.) joints
filled with a stone gradation conforming to ASTM No. 9 stone per ASTM
D448 [8]. According to The Aggregate Handbook [9], this material has a hy-
draulic conductivity of at least 3.5 103 m/s (500 in./h).
The paving units and jointing material were installed over 40 mm (1.5 in.)
thick ASTM No. 8 bedding stone. This was placed over a 100 mm (4 in.) thick
ASTM No. 57 base and a 675 mm (27 in.) thick subbase of ASTM No. 3 stone.
This was constructed in lieu of a drywell drainage system. The design storm
required by the local municipality is equivalent to a storage volume of 40 mm
(1.5 in.) in a 24 h period or a 90 % rainfall event as defined by the New York
State Stormwater Management Design Manual. Figure 1 illustrates the test site
with the ASTM C1701 test apparatus.
FIG. 1—ASTM C1701 test apparatus at a building supply store parking lot.
TABLE 1—PICP test results for a building supply parking lot in Lindenhurst, NY, in meters per sec-
ond (inches per hour).
Date Test Location 1 Test Location 2 Test Location 3

3 3
Oct. 21, 2010 1.5 10 (206) 2.1 10 (302) 2.5 103 (358)
1.3 103 (182) 2.5 103 (350) 2.2 103 (313)
Apr. 4, 2011 2.3 103 (321) 2.4 103 (341) 2.4 103 (336)
2.4 103 (340) 2.3 103 (327) 2.5 103 (354)
Nov. 2, 2011 1.8 103 (250) 1.6 103 (229) 1.7 103 (242)
1.7 103 (243) 1.6 103 (232) 1.7 103 (239)
Both test locations in Lindenhurst used plumber’s putty to seal the metal
ring against the paving. At both sites, the jointing stones were removed
between the joints under the ring and filled with plumbers putty to further
direct the water downward. Removal of the jointing stones can be done with a
putty knife and a screwdriver. A key consideration in placing the ring is to
frame the paver joint pattern within the ring. This framed area should represent
the percentage of open area in the overall surface to best characterize surface
infiltration. In addition to characterizing the overall permeable jointing pattern,
this location can reduce the time and effort required in order to remove jointing
stones and fill the joints with plumber’s putty.
After pre-wetting, the ASTM C1701 test method was conducted three
times over the first 11 months of service, which resulted in an average infiltra-
tion rate of 2.025 103 m/s, or 287 in./h. Table 1 provides the test data for
three test locations.
Another test using ASTM C1701 was conducted at the 600 m2 (6000
2
ft ) PICP parking lot of a public library in Lindenhurst, NY. The paving units
were 200 mm 200 mm 80 mm (8 in. 8 in. 3.125 in.) thick with 13 mm
(0.5 in.) joints filled with ASTM No. 8 stone. The paving units and jointing ma-
terial were installed over 40 mm (1.5 in.) thick ASTM No. 8 bedding stone.
This was placed over a 100 mm (4 in.) thick ASTM No. 57 base and a 150 mm
(6 in.) thick subbase of ASTM No. 3 stone over a permeable soil subgrade.
Like at the site mentioned above, this PICP was constructed in lieu of a drywell
drainage system and according to the same New York State design criteria.
Figure 2 illustrates the test site with the ASTM C1701 test apparatus.
After pre-wetting, the ASTM C1701 test was conducted in April 2011 dur-
ing the first months of service and resulted in an average infiltration rate of
3.8 103 m/s, or 538 in./h. Table 2 provides the test data.
U.S. EPA Experience

In the fall of 2009, the U.S. Environmental Protection Agency (EPA) opened a
110-car parking lot for staff at their Edison, NJ, National Risk Management
FIG. 2—ASTM C1701 test apparatus at a public library parking lot.
Laboratories. The parking lot was designed with the aim of evaluating the per-
formance of PICP, pervious concrete, and porous asphalt. The parking facility
is illustrated in Fig. 3; the three permeable pavement surfaces were each
approximately 530 m2 (5300 ft2). The research objectives and parameters for
this multi-year monitoring project are shown in Table 3. One objective listed is
to measure the surface infiltration in order to assess maintenance cleaning
methods. In order to gain a better understanding of the surface cleaning
required, one half of the parking lot is vacuum swept twice a year, and the
other half is not. Surface infiltration measurements are measured monthly in
both areas with a modified version of ASTM C1701 so as to characterize the
clogging potential of each surface and when cleaning might be required.
Instead of plumber’s putty, to create a seal between the ring and the pave-
ment surface, neoprene is applied to the ring and the ring is pressed into the
pavement surface by plastic buckets weighted with stones. This apparatus
increases setup, measurements and clean-up time. Figure 4 illustrates the appa-
ratus. All other aspects of the test method appear to be similar to those
described in ASTM C1701 [10].
Borst et al. reported infiltration rates for the initial months of the parking
lot’s surface [10]. There was no difference between maintained and
TABLE 2—PICP test results for a public library parking lot in Lindenhurst, NY, in meters per second
(inches per hour).
Date Test Location 1 Test Location 2 Test Location 3

Apr. 4, 2011 3.4 103 (481) 3.5 103 (492) 4.6 103 (656)
3.4 103 (487) 3 103 (422) 4.8 103 (687)
FIG. 3—Pervious concrete (light colored parking area on left), PICP (center),
and porous asphalt (right) at a test facility at the U.S. EPA laboratories in Edi-
son, NJ.
unmaintained area surface infiltration rates, likely due to the newness of the
surfaces. The unweighted mean for pervious concrete was 1.1 102 m/s
(1574 in./h); PICP was measured at 6.6 103 m/s (944 in./h), and porous
asphalt at 5.6 104 m/s (79 in./h). The report does not mention whether the
stones were removed from between the pavers in the PICP and filled with neo-
prene or plumber’s putty.
Proposed Changes to ASTM C1701 Testing Procedures

Jennifer Drake, a doctoral student at the University of Guelph, is conducting a
series of surface infiltration tests on PICP and pervious concrete sites in On-
tario, Canada, using ASTM C1701, as well as double ring test methods. She is
also investigating water volume and pollutant removal from these pavements.
Her observations regarding both pavement types are that they require regular
vacuum sweeping in order to maintain their surface infiltration rates. She noted
at an October 2011 presentation in Ontario that in hot weather, non-oil base
TABLE 3—U.S. EPA research objectives and parameters measured at the Edison, NJ, permeable
pavements research facility.
Monitoring Objective Parameters Measured

Hydrologic performance Volume, exfiltration rate
Water quality performance Soils, indicator organisms, metals, nutrients, organic compounds
Urban heat island effects Net radiation, infrared radiation, temperature
Maintenance effects Surface infiltration rate, visual assessment
Use Car counter, visual assessment
Infiltrating water parameters Water depth, redox, pH, conductivity, chloride
FIG. 4—Modified ASTM C1701 using neoprene seal on the porous pavement
at the U.S. EPA laboratories in Edison, NJ [10].
modeling clay works better than plumber’s putty, as the latter material becomes
viscous, stringy, and difficult to handle [11]. The modeling clay molds quickly
and creates a seal between the ring and the paving.
ASTM C1701 currently requires that the mass of infiltrated water be deter-
mined so that the value can be entered into a formula used to calculate the sur-
face infiltration rate. This can mean that one has to bring a scale to the site in
order to weigh the before and after mass of the water usually dispensed from
buckets. A bucket or other suitable container(s) with graduations related to the
mass of water could obviate the need for a scale on the site while still providing
the mass of water dispensed during the test.
Conclusions
The most common permeable pavement surfaces are pervious concrete, PICP,
and porous asphalt. There are millions of square meters of each in service.
ASTM C1701 is an inexpensive and rapid test method for measuring surface
infiltration by simulating a small hydraulic head on the surface test area like
those generated by intense rain storms and contributing runoff. Data and expe-
rience from Bean et al., Lia, and Drake confirm that ASTM C1701 is suitable
for testing the surface infiltration of PICP, and Bean extends its use to the suc-
cessful testing of concrete grid pavements [2,6,7,11]. Borst et al. report using a
modified version of ASTM C1701 to test PICP, pervious concrete, and porous
asphalt as part of a nationally visible evaluation of these pavements [10].
In order for contractors, stormwater agencies, and project owners to better
understand the performance and maintenance needs of all permeable
pavements, ASTM C1701 should be revised to address PICP, concrete grid

pavements, and porous asphalt testing. The literature demonstrates that perme-
able pavements can experience a reduction in their surface infiltration rates,
especially if they are not maintained with vacuum sweeping [12,13]. Broaden-
ing the application of ASTM C1701 to characterize the surface infiltration of
more than just pervious concrete enables an objective comparison of the per-
formance of other pavement systems designed to reduce stormwater runoff.
The scope of ASTM Committee C09 on Concrete and Concrete Aggregates
is “[t]he assembling and study of data pertaining to the properties of hydraulic-
cement concrete and its constituent materials, including the study of the effect
of characteristics of materials and mixtures on the properties of concrete; and
(2) the development of standards for concrete and for the constituent materials
of concrete (except cement), as well as for certain related materials, such as
materials used in curing.” This scope appears to enable the inclusion of the test-
ing of PICP and concrete grid pavements in ASTM C1701, especially given that
ASTM C1701 evolved from tests on PICP and grids, as well as from pervious
concrete. Although including porous asphalt surface testing in ASTM C1701
might not fall exactly within the scope of Committee C09, it could help with
comparisons of performance among the most common permeable pavements.
With this in mind, the proposed changes to ASTM C1701 include the
following:
• Change the title to “Standard Test Method for Infiltration Rate of In
Place Pervious Pavements” and expand the scope to include PICP, con-
crete grid pavements, and porous asphalt.
• As a substitute for plumber’s putty, allow the optional use of (non-oil
base) modeling clay when testing in hot temperatures with material

specifications.
• Allow the optional use of graduated containers to determine the water
volume and mass for dispensing water. This could eliminate the use of a
scale on the test site.
• Include test procedures for PICP, concrete grid pavements, and porous
asphalt, including a description of centering the ring over concrete

paver/grid patterns and joints.
• Include precision and bias statements for the above.
References
[1] ASTM D3385, 2003, “Standard Test Method for Infiltration Rate of Soils
in Field Using Double-Ring Infiltrometer,” Annual Book of ASTM Stand-
ards, Vol. 04.02, ASTM International, West Conshohocken, PA.
[2] Bean, E. Z., Hunt, W. F., and Bidelspach, D. A., “Field Survey of Permea-
ble Pavement Surface Infiltration Rates,” J. Irrig. Drain. Eng., Vol. 137,
2007, pp. 249–255.
[3] ASTM C1701, “Standard Test Method for Infiltration Rate of In Place
Pervious Concrete,” Annual Book of ASTM Standards, ASTM Interna-
tional, West Conshohocken, PA.
[4] Smith, D. R., Permeable Interlocking Concrete Pavement—Design Speci-
fications Construction Maintenance, 4th Edition, Interlocking Concrete
Pavement Institute, Herndon, VA, 2011, pp. 73–74.
[5] New York State Department of Transportation, 2011, “Item
502.010700OD—Pervious Portland Cement Concrete,” Albany, NY.
[6] Lia, J. M., “Post Construction Surface Testing of the Nicolock Eco-Ridge
Permeable Interlocking Concrete Paver Installation (PICP) at the Century
Building Materials Facility Hamlet of Lindenhurst, Town of Babylon Suf-
folk County, NY,” Report to 4Site Engineering, Port Jefferson Station,
NY, October 2010, April and November 2011.
[7] Lia, J. M., “Post Construction Surface Testing of the Nicolock SF-Rima
Permeable Interlocking Concrete Paver Installation (PICP) at the Linden-
hurst Public Library,” Report to 4Site Engineering, Port Jefferson Station,
NY, April 2011.
[8] ASTM D448, 2006, “Standard Classification for Sizes of Aggregate for
Road and Bridge Construction,” Annual Book of ASTM Standards, Vol.
04.03, ASTM International, West Conshohocken, PA, pp. 62–64.
[9] National Stone Association (NSA), The Aggregate Handbook, NSA,
Washington, D.C., 1991.
[10] Borst, M., Rowe, A. A., Stander, E. K., and O’Connor, T. P., 2010, Sur-
face Infiltration Rates of Permeable Surfaces: Six Month Update (Nov
2009 through April 2010), Water Supply and Water Resources Division,
National Risk Management Research Laboratory, U. S. Environmental
Protection Agency, Edison, NJ.
[11] Drake, J., “Permeable Pavement Monitoring at the Kortright Centre for
Conservation & Maintenance Testing on Established Permeable
Pavements,” Presentation to the Toronto and Region Conservation
Authority, October 2011, Toronto, ON, Canada.
[12] Chopra, M. B., Stuart, E., and Wanielista, M. P., “Pervious Pavement Sys-
tems in Florida—Research Results,” Proceedings of Low Impact Develop-
ment 2010: Redefining Water in the City, American Society of Civil
Engineers, Reston, VA, pp. 193–206.
[13] Vancura, M., and Khazanovich, L., Performance Evaluation of In-Service
Pervious Concrete Pavements in Cold Weather, Department of Civil En-
gineering, University of Minnesota, Minneapolis, Minnesota, 2010.

Astm

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

Astm

Încărcat de

Drepturi de autor:

Formate disponibile

Copyright by ASTM Int'l (all rights reserved); Fri Aug 23 10:02:14 EDT 2019

Printed in the U.S.A.

ASTM Stock #: STP1551

Printed in Bay Shore, NY

Validation of the Performance of Pervious Concrete in a Field Application with Finite

Md. Ashraful Alam,1 Liv Haselbach,2 and William F. Cofer2

Validation of the Performance of Pervious

with field application and laboratory material strength data,

tion performance data, and

Experimental Data from Previous Studies

Finite Element Modeling of the Field Site

Depth of pervious Flexural Compressive

Material Properties Considered in the FEA

Pervious Concrete—The moduli of elasticity for the pervious concrete

FIG. 1—Three-dimensional representation of the pervious concrete system for

derived in a previous study based on varying moduli, porosities, and compres-

fc0 ¼ 114P þ 43:8 (3)

Ptop ¼ 1:07Pmean 7 (6)

Pbottom ¼ 0:93Pmean þ 7 (7)

Subbase and Subgrade—The modulus of elasticity and Poisson’s ratio of

Pavement Model and Meshing

FEA Results and Discussion

increasingly profound as the subbase thickness and, hence, its stiffness,

Comparison Between Experimental Strength and Finite Element Stress Output

FIG. 5—Comparison of tensile strength from experimental investigation and

Figure 6 depicts the material’s compressive strength from experiments and

Comparison with Pavement Condition Index

FIG. 6—Comparison of material compressive strengths from experimental

FIG. 7—Minimum performance line and additional thickness values with

theory for traditional concrete pavement, with maximum deflections for

[3] Haselbach, L., The Engineering Guide to LEED-New Construction,

Vertical Porosity Distribution Approach,” Finite Elem. Anal. Design

William B. Denison, Jr.1

Performance of Pervious Portland Cement

REFERENCE: Denison, William B., Jr., “Performance of Pervious Portland

Evaluation of Plastic Properties of Pervious Concrete

Sampling—ASTM C172-04 (Standard Practice for Sampling Freshly

Density (Unit Weight) of Concrete, lb/ft3—ASTM C29-03 [2] (Standard

Mm ¼ mass of measure, lb, and

A ¼ ½ðT DÞ=T 100 (2)

Evaluation of Hardened Properties of Pervious Concrete

Hardened Density, lb/ft3—ASTM C42-04 [4] (Test Method for Obtaining

Thickness, in.—ASTM C174-06 [7] (Test Method for Measuring Thickness

Compressive Strength, psi—ASTM C39-05 [3] (Compressive Strength of

Flexural Strength, psi—ASTM C78-02 [5] (Flexural Strength of Concrete

Pervious Mix Proportioning

TABLE 1—TVRM pervious mix design proportions.

Material Plastic Mix Proportions (lb/yd3)a

Total Plastic Plastic Void Hardened Hardened Void

Total Plastic Plastic Void Hardened Hardened Void

next to the in place pervious pavement, where curing commenced immediately

Procedures Performed at the TVRM Laboratory, Norfolk, VA

Panel Placement Thickness Average Compressive Average Flexural

compressive and flexural strength testing. The TVRM testing methology

[1] ACI 522-1, “Specification for Pervious Concrete Pavement,” American

of ASTM Standards, Vol. 04.02, ASTM International, West Consho-

Samuel Meulenyzer,1 Eric Stora,1 and Fabien Perez1

Impact of Pervious Concrete Porosity

REFERENCE: Meulenyzer, Samuel, Stora, Eric, and Perez, Fabien, “Impact

Materials and Methods

Porosity Measurement—Porosity is a parameter of primary importance

TABLE 1—Sample designation extracted from the experimental program.

Specimen Range of Paste Range of Aggregate Paste

A ¼ ½ðT DÞ=T 100 (2)