See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/354618904

Design and Construction of Bases and Subbases for Concrete Pavement

Performance

Conference Paper · September 2021

CITATIONS READS

0 2,361

3 authors, including:

Hesham Abdualla Thomas Yu

Iowa State University U.S. Department of Transportation
17 PUBLICATIONS 311 CITATIONS 22 PUBLICATIONS 519 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Hesham Abdualla on 16 September 2021.

The user has requested enhancement of the downloaded file.

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

Design and Construction of Bases and Subbases for Concrete Pavement

Performance
Shreenath Rao, Hesham Abdualla
Applied Research Associates, Inc.

Thomas Yu, P.E.

Federal Highway Administration

ABSTRACT

Effective base and subbase (referred to in this document as foundation) design and construction is
necessary to ensure good, long-term performance of pavements. However, structural models used in
mechanistic-empirical (ME) pavement design do not show significant impact of foundation layers on
pavement performance, especially for unbound layers beneath concrete pavements. This is in contrast
with experience of pavement engineers, who generally agree that the foundation layers perform
important functions, and thus contribute significantly to performance. The principal weakness of ME
pavement design models is that they are based only on pavement response, assuming the support
conditions are even and remain consistent throughout the life of the pavement. Seasonal changes in the
foundation layers are modeled, but only as uniform changes in layer moduli values. Such changes have
negligible effect on pavement response, especially for concrete pavements. Pavement foundations can
and do degrade unevenly over time, especially if the design is inadequate to resist deformation and
deterioration over time. The benefits provided by quality foundation layers are through the prevention
of adverse incidents that can lead to localized or progressive failures. The relatively simple, idealized
structural models used in ME designs are not meant to consider complex mechanisms involved in
failures resulting from foundation problems. Thus, qualitative understanding of the failure mechanism is
important to devise effective pavement layer designs that avoid foundation problems. Ideally, an
effective foundation should not degrade over time and should be designed against deformations,
decompaction, and infiltration of fines into the pavement layers, such that the engineering properties of
the foundation assumed in design are spatially consistent and retained throughout the life of the
pavement. This paper will demonstrate the effective uses of foundations for concrete pavements based
on the evaluation of a series of case studies encompassing a range of conditions wherein the foundation
layers contributed significantly (either positively or negatively) to overall pavement performance that is
not fully accounted for in ME design. Four of the ten case studies (Ontario, Michigan, Missouri, and
Virginia) developed under this study and focusing on drainage aspects of pavements are included in this
paper. The case studies are used in this study to develop guidelines for proper design and construction
of quality bases and subbases for concrete pavements.

1. BACKGROUND
Base course is a layer of the pavement structure immediately beneath the surface course. It typically
consists of high quality aggregate such as crushed gravel, crushed stone, or sand that provides a uniform

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 132

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

foundation support and an adequate working platform for construction equipment. Base may consist of
unbound materials, such as gravel or crushed stone, or stabilized materials, such as asphalt-, cement- or
lime-treated materials. The subbase course is typically a granular borrow that is placed between the
base and subgrade. It can be constructed as either a treated or untreated layer. Untreated or unbound
aggregate subbase layers are characterized in a manner similar to the subgrade in pavement design. The
material quality requirements of strength, plasticity, and gradation for subbase are not as strict as for a
base. The subbase course must be better quality than the soil subgrade, the subbase is often omitted if
soil subgrades are of high quality. Depending on site conditions, subgrade improvements may also be
performed. However, the role of different base and subbase layers and rationale for using different base
types and layering are not well documented. Many agencies specify typical base and subbase layers
based on historical performance and their own experiences. For example, it is unclear where and why a
treated base should be used, or why one type of treated base is preferred over another.

Pavement engineers generally agree that foundation layers perform important functions, including
providing uniform support, controlling pumping and erosion, protecting against frost heaving, and
reducing moisture related damage to paving materials. Based solely on structural analysis, the benefits
of unbound aggregate base cannot be demonstrated as the structural models used in mechanistic-
empirical (ME) pavement design do not show significant impact of foundation layers on pavement
performance. From structural perspective, the most efficient means of providing adequate structure is
by providing sufficient surface thickness, concrete or asphalt. However, experience shows that concrete
pavements placed directly on subgrade do not perform well in most areas, because of pumping and
migration of fines resulting in variability in foundation support. So, it should be clear that the foundation
layers perform a different function than the surface layer, which is the main structural component to
resist the applied loads. However, when evaluating the benefits of foundation layers, attempts are often
made to quantify benefits only in terms of structural effect.

Ultimately, the benefits provided by the foundation layers can be related to the structural and
functional performance; however, the benefits are more in the way of preventing bad things from
happening that can lead to localized or progressive failures and increased roughness. The relatively
simple, idealized structural models used in ME designs are not meant to consider complex mechanisms
involved in failures resulting from subgrade and foundation problems. This is not to say that the current
ME design models are deficient; it would not be practical nor necessary for design procedures to model
complex failure mechanisms associated with foundation problems. For the purposes of pavement
design, qualitative understanding of the failure mechanism and protecting against those failures is
sufficient to devise effective pavement layer designs that avoid foundation problems. A pavement
foundation that does not degrade over time does not need to be replaced. A permanent foundation has
obvious advantages in environmental impact, and it could also have significant economic advantages. In
congested areas, not having to replace the foundation could be highly advantageous in expediting
pavement rehabilitations and reconstructions.

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 133

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

2. FUNCTION OF PAVEMENT FOUNDATION IN RIGID PAVEMENTS

Pavement foundation in rigid pavements has less appreciable impact on the structural capacity and the
primary function of the foundation layers is providing a uniform support for the concrete slabs. A
uniform and sound quality support layer enhances the rigid pavement performance more than a
stronger and non-uniform support (ACPA 2007, Hein et al. 2017). The functions of the rigid pavement
foundation are as follows:

 Provide a uniform support to the PCC layer with adequate stiffness.

 Offer a stable construction platform.
 Prevent loss of slab support due to erosion and pumping.
 Provide resistance against frost-heave and soil expansion.
 Separate the subgrade from the main structural component.
 Improve drainage and prevent moisture-related damage.
 Provide a gradual vertical transition in layer moduli (stiffness) from the slab to the subgrade.

If the primary functions of the rigid pavement foundation are not adequately considered during the
design process or it is not properly constructed, the pavement system may not achieve the desired
performance. Moreover, misuse of a foundation layer beneath a rigid pavement may lead to premature
deficiencies. For instance, the base and subbase type and thickness should be selected based on specific
site conditions. More often than not, the base and subbase type is selected based on a variety of factors
such as the agency policy, cost and availability of materials, and past experience. Given these conditions,
the base type and thickness should still be selected to meet the needs (e.g., drainage, protect against
frost heaving, protect against swelling and unstable soils) of the project site.

3. QUALITATIVE DESCRIPTION OF PAVEMENT FOUNDATION

3.1 How Rigid Pavements Work

PCC slabs have elastic modulus that is an order of magnitude higher than asphalt concrete. The typical
flexural strength is about 700 psi and the modulus of elasticity is about 5 million psi. Therefore, unlike
flexible pavement structures, which transfers the wheel load gradually to the layers underneath (see
Figure 1), the traffic load applied to rigid pavement structures is primarily distributed by the concrete
slabs over a wider area before it is transmitted to the layers beneath the slabs (Hein et al. 2017) . As
such, the pavement responses induced in the layer below the concrete slabs, including the stresses (i.e.,
pressure) and strains as well as the deflections, are relatively smaller.

Previous studies showed that the load-induced compressive stress on top of subgrade in rigid
pavements is substantially lower than its bearing strength. As an example, 12,000 lb tire load with 100
psi contact stress applied on a typical rigid pavement structure induces compressive stress of about 7 psi
for the corner loading. In this case, the induced stress on top of subgrade drops to as low as 3 psi for
interior loading. Such observations confirm that concrete pavements obtain the desired structural
capacity from concrete slabs and therefore, the uniformity and stability of support layers in rigid
pavements is a more important than their stiffness and strength (ACPA 2007).

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 134

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

WheelWheel
LoadLoad WheelLoad
Wheel Load

Flexible Pavement Rigid Pavement

Flexible Pavement
Base
Rigid Pavement
Base
Base Base

Subgrade Subgrade
Subgrade Subgrade

Figure 1. Load Distribution in Flexible Pavement Figure 2. Load Distribution in Rigid

Pavement

3.2 Failure Mechanisms

The purpose of a uniform support for concrete pavement is to ensure that pavement will attain its
service life and uniformly distribute loads over the foundation throughout the entire service life. A
uniform support can be achieved by reducing the effect of three key factors: frost heave, pumping of
fine-grained soils, and volume changes of the soil. Other factors responsible for non-uniform support
include variability in compaction, in cut/fill and transitions, and ineffective drainage system. Table 1
summarizes the causes and effects of non-uniform support on the performance of concrete pavement
and the recommended practicable solutions to eliminate such problem (Hein et al. 2017, ACPA 2007,
ACPA 1995, Christopher et al. 2006, Snethen et al. 1977).

4. DESIGN OF BASES AND SUBBASES FOR CONCRETE PAVEMENT

The selection of base and subbase type for a given a project should be based on (1) the function of
base/subbase layer with the pavement structure, (2) improve the short and long-term performance, (3)
cost-effective approach, and (4) local experience (Hall et al. 2005). Modulus of subgrade reaction
(known as k-value) is typically used to quantify the stiffness (strength) of rigid pavement support.
Composite k-value is a representative of pavement foundation stiffness consisting base and subbase.
The k-value is determined by plate load test in accordance with AASHTO T122 and ASTM D1196. The
stiffness of pavement support may increase by placing subbase and base layer on top of subgrade.
However, increasing the support strength (or stiffness) to reduce the PCC thickness, to expedite the
construction process, or as a surrogate for improving the durability of the base is not recommended.
Increasing k-value within the typical range does not substantially affect the required thickness of
concrete slab (ACPA 2007).

Aggregate base and subbase with 15% or more fines (i.e., passing the sieve No. 200) are highly prone to
pumping. The use of non-erodible or treated base and subbase materials can control and prevent
pumping. The requirements in AASHTO M155 entitled “Standard Specification for Granular Material to
Control Pumping under Concrete Pavement” should be followed when the unbound granular materials
is to be used (AASHTO 2004). In general, the higher the application of heavy truck traffic, materials with
lower fine content and lower plasticity should be selected.

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 135

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

Table 1. Failure Mechanisms and Recommended Solutions for Obtaining a Uniform Support

ITEMS CAUSE EFFECT SOLUTION

Pumping  Base erodibility  Faulting  Use less erodible
 Fine grained  Fatigue base materials
subgrade cracking  Limit the amount of
 Present of free  Corner fines to 10% or less for
water cracking aggregate base course
 Water table  Lane shoulder  Use a well-design
 Heavy truck drop off drainable system
 Loss of  Use a separation
support layer when using unstabilized
granular layer
 Use dowel bars
Frost heave  Frost susceptible  Random  Use non-frost-
soil cracking susceptible materials
 Source of water  Blocked  Cover frost-
 Freezing drainage susceptible soil with sufficient
temperature penetrating  Reduction in thickness of non-frost-
the soil bearing capacity susceptible material
 Loss of
support
Expansive soil  Degree of  Random  Consider use of
moisture changes within cracking stabilization and membranes
the soil  Longitudinal  Remove and replace
cracks near the small areas of swelling soils
pavement’s edge  Covering the soils
 Significant with a sufficient depth
surface roughness  Identify optimum
 Loss of moisture content
support
Longitudinal  Variability of soil  Increase long-  Use quality materials
or transverse along the project term roughness for fill sections to ensure
Cut/fill  Settlements, long-term performance
which causes random  In cut-sections,
cracking drainage system may be
 Loss of needed if ground water table
support is high

Stiffer bases are not necessarily better support under rigid pavements as they fail to conform to the
shape of the curled PCC slabs and may lead to loss of support, higher curling stresses, and subsequent
cracking. It should be noted that providing thicker concrete slab, higher concrete strength, the use of
dowel bars and widened slabs are more economical to substantially reduce the cracking potential in
concrete slabs and pumping of materials. A stiff support has potential to cause cracking because of the
higher environmentally-induced stresses in the slabs. This can be detrimental for relatively young

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 136

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

concrete slabs leading to development of random cracks. It is recommended that the compressive
strength of cement treated bases and lean concrete basses should range from 300 to 800 psi and 750 to
1,200 psi, respectively (Hein et al. 2017).

Stabilized bases including cement treated bases and lean concrete bases have potential to expand and
contract due to moisture and temperature variations. These movements can sometimes induce stresses
greater than the strength of freshly placed surface PCC (when the strength in the freshly placed PCC is
low as it is hydrating and gaining strength), thus increasing the potential for early-age cracking in the
PCC layer. In addition, rough slab-base interfaces increases frictional forces at the interface due to the
excessive axial restraint to volumetric shrinkage and to thermal expansion and contraction (Hall et al.
2005). To mitigate this potential risk, it is common practice to have a debonding separator layer (like a
plastic sheet) between the cementitious stabilized base and the PCC layer. However, an unbonded base
contributes less to the long-term fatigue performance of concrete pavement as compared to a fully
bonded base, and this may need to be considered in the pavement design process, for example, by
increasing the thickness of the PCC layer. This is less of an issue with dense asphalt-treated bases which
are sufficiently flexible and do not expand and contract due to thermal effects to the same extent as
cementitious stabilized bases.

To provide drainable base layers, permeable granular or stabilized bases with drainage system or free
draining delighted bases can be used. Permeable granular layers should only be used where there is
potential for moisture damage to pavement on roadways with medium to heavy truck traffic, and
should be properly designed and constructed. However, the owner agency should have a commitment
to regular inspection and routine maintenance of the edge-drains or the exposed (daylighted) area of
the aggregate drainage layer. An open-graded base needs a suitable separator layer beneath it to
prevent subgrade fines from migrating up into and clogging the base. This may be an appropriately
graded untreated aggregate subbase, an appropriate geotextile fabric, or a layer of subgrade soil treated
with sufficient lime or cement to achieve good long-term stability and resist erosion. Stabilized open-
graded drainage layers have very little aggregate passing the No. 200 sieve. Asphalt cement contents
typically range between 1.6 and 1.8 percent by mass of aggregates. Cement treated open graded
drainage layers are typically produced with a water to cement ratio of 0.37 and a cement content of 185
to 220 lbs/yd3 (Hein et al. 2017). Permeable bases must be constructed strong enough to resist
construction traffic and paving machine without deformation (Hall et al. 2005). The recommend
permeability values are ranging from 500 to 800 ft/day with taking in consideration the stability of the
bases (Hein et al. 2017).

The use of “daylighted” base course that is exposed to the open along the edge of the pavement is
recommended to drain water infiltrating from the surface into the base layers, particularly in situations
where moisture conditions are not extremely severe. Daylighting allows water to slowly drain out of the
pavement structure without the use of edge drains. Daylighted bases are well suited for roadways with
flat grades (1% or less) and shallow ditches, where it is difficult to outlet drainage pipes at an adequate
height above the ditch. However, it requires careful construction and periodic maintenance to keep the
exposed edge clear of soil, vegetation, and debris, and prevent clogging. Typical maintenance activities
include weeding and manual removal of debris. The bottom of the exposed edge of the daylighted base

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 137

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

should be at least 6 in. above the 10-yearstorm flow line of the ditch to prevent water from backing up
into the daylighted base during or after a heavy rainfall. Daylighting the base layers is more “forgiving”
than using edge drains. With edge drains, there is the potential for trapping water within the pavement
layers causing a “bathtub” effect and resulting in significantly greater damage, if they get clogged from
not being regularly maintained or from improper installation. However, when properly maintained, edge
drains are effective and drain water efficiently out of the pavement system, particularly in areas with
high water tables and cut sections.

5. CASE STUDIES FOR EFFECTS OF BASES AND SUBBASES ON PAVEMENT PERFORMANCE

5.1 U.S. 460 Bypass, Appomattox County, VA
The project is located on the U.S. 460 bypass in the northern part of Appomattox County, Virginia. The
doweled jointed plain concrete pavement (JPCP) section, an approximate 2.8-mile long section, of the
U.S 460 bypass exhibited premature failure at several location approximately 5 years (i.e., 1998) after
paving. The Virginia DOT engineers and researchers conducted field and laboratory investigations to
identify the causes of premature failures and evaluate the condition of the pavement section (Hossain
and Elfino 2005, Elfino and Hossain 2007). The project is located in a wet-freeze climate and the average
daily traffic (ADT) in 2003 was 13,000 with 10% truck traffic.

5.1.1 Design and Construction

The U.S. 460 bypass was designed to carry an equivalent of 8 million single axle loads (ESAL) with an
expected design life of 30 years. The following design was used for this section:

 9.0 in. doweled JPCP slab with 15 foot spacing.

 4.0 in. cement-stabilized open-graded drainage layer (OGDL).
 6.0 in. cement-treated soil, using 10% hydraulic cement by volume.
 9.0 to 6.0 in. variable depth jointed concrete undoweled tied shoulder.
 4.0 in. aggregate base materials for shoulder (VDOT Type 1, Size 21A).
 Pavement edge drain UD-4 in accordance with the standard pavement edge drainage and outlet
pipes.

The subgrade soil was classified as A-7-5 red clay and silt with a CBR of 9.

5.1.2 Performance
A visual survey was conducted to evaluate the cause of premature pavement failure. The survey results
showed that about 24% of the eastbound slabs were distressed, compared with 12% of the westbound
slabs. The pavement exhibited mid-slab cracks, broken joint seals, lane-shoulder drop-off and pumping,
and joint faulting. Field and laboratory investigations were performed to assess the causes of pavement
distresses. The overall observation of the laboratory and field investigation can be summarized as
follows:

 A majority of the drainage layer was clogged and filled with red soil (see Figure 3a).
 Cracks propagate through the drainage layer in the mid-slab crack core sample.
 Water trapped underneath the slab was observed under damaged slabs during coring.
 The OGDL was not extended over the edge drainage in some areas (see Figure 3b).

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 138

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

(a) (b)
Figure 3. Slab Replacement (2005): (a) discontinuity of open-graded drainage layer over the edge drain
and (b) clogged drainage layer

5.1.3 Lesson Learned/Summary

 A poor drainage system and increased truck traffic can significantly affect pavement
performance.
 If the OGDL is not continued to the edge drain, trapped water in the drainage layer will seep
vertically and cause an increase in base/subbase and subgrade moisture.
 Water abrades the soil cement base/subbase under repeated heavy loads leading to localized
loss of support, disintegration, resulting in pavement distresses, including structural and
durability Figure 4.

Figure 4. Pavement distresse (2018): (a) mid slab cracking (structural distress) and (b) joint deterioration
(durability distress).

5.2 U.S. 63, Callaway County, MO

The Missouri River flooding significantly damaged pavements, culverts, bridges, etc. in Jefferson City,
Missouri, resulting in the closure of roadways, delay in traffic, and economic losses to the city. The
roadway was completely washed out due to the flooding in 1993. The project is located on southbound
US-63 in Callaway County, MO, just across the Missouri River from Jefferson City, MO. The original
pavement design of southbound US 63 sections consisted of 9 inches of joint reinforced concrete
pavement (JRCP) with 61 foot joint spacing on 4 inches of dense graded crushed rock base. The Missouri
Department of Transportation (MoDOT) researcher and engineers conducted a comprehensive study to
further enhance the pavement design of US 63 sections to resist such environment. The major finding of
the study led to developing a new standard specification provision of a thick daylighted rock base, which

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 139

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

has capabilities to drain water from the pavement structure and to improve the load-bearing capacity of
pavement structure.

5.2.1 Design and Construction

The new design of US 63 section consisted of 12 inches of doweled joint plain concrete pavement (JPCP)
with 15 foot joint spacing over 24 inches daylighted rock base and was constructed in 1994. This was the
first implementation of daylighted rock base in Missouri. The 24 inches base was selected to increase
the structural capacity as well as improve drainage during heavy rain or flood periods. The grade was
raised about 6 feet because of the flooding damage. The daylighted rock base was placed on the top of
the subgrade. A cross slope gradient from the median to the outside fill slope was provided on the top
surface of the subgrade to remove water effectively from the pavement structure, prior to the
placement of the 24 inch rock fill base. The subgrade soils in this area consisted of A-6 and A-7-6 soils.

5.2.2 Performance
A visual survey of US-63 sections conducted in 2016 and revealed that all sections performed in
excellent condition and there were no signs of cracking or faulting (see Figure 5a). In early 2018, a
second survey was conducted indicated excellent conditions, which could be attributed to the
effectiveness of the daylighted rock base (see Figure 5b). The pavement section was constructed in
October of 1994 and has performed extremely well with minimal cracking, faulting, and roughness.
Minimal maintenance has been done on the section since construction and all joints look exceptional.
The section experienced another flood in 1995 and maintained in good condition. The success of this
pavement was attributed to the daylighted 2-foot rock base and its superior drainage capabilities. After
24 years of relatively heavy traffic, US 63 section is still in perfect structural condition and no repairs
were performed. IRI measurements were taken from 2007 to 2017 for the project and the data shows
the consistency of the roughness of the pavement over 10 years of initial life.

(a) (b)
Figure 5. Pavement Performance of US-63, Callaway County, MO: (a) performance in 2016 and (b)
performance in 2018

5.2.3 Lesson learns

 The stability and drainability of base material is essential to enhance the performance of
pavement during heavy rain or flooding incidents.

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 140

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

 The original base, 4 inches dense graded crushed rock, was filled with sand and relatively
undrainable. Concrete pavement constructed over a dense-graded rock base has a high risk of
being damaged during flooding incidents due to ineffective base drainability.
 The use of a thick daylighted rock base significantly improved the long term pavement
performance. The 24 inch daylighted rock base was effective for removing water from the
pavement structure, which enhanced the JPCP performance and eliminated the moisture-
related damages.

5.3 U.S. 23, Monroe County, MI

In 1992, Michigan DOT constructed an Aggregate Test Road on southbound US-23 with the main
purpose of studying the effect of frost susceptible coarse aggregate on concrete durability. The project
starts just north of the US-23 and US-223 interchange and ends at the border line between Michigan
and Ohio. The test road was constructed with concrete mixtures including five different coarse
aggregates (Groups A through E) of varying degrees of freeze-thaw properties. The coarse aggregate
type for group A was 6AA crushed limestone, group B was 6AA blast furnace slag, group C was 6A
naturel gravel, group D was crushed limestone from different quarry, and group E was natural gravel. All
other factors of the concrete mix design were kept the same. The average annual daily traffic (AADT)
was about 20,000 with 18% of commercial (Hansen et al. 2007, Quiroga 1992).

5.3.1 Design and Construction

The original pavement was removed to the existing sand subbase. The new pavement structure
consisted of a 10.5 inches jointed reinforced concrete pavement (JRCP) with 27 ft. joint spacing, on a 4
inches asphalt treated permeable base (ATPB) on a 3 inches gravel separator layer. Half of each of the
five test sections was built on the original poorly-draining subbase while the other half constructed on a
well-draining permeable sand subbase to evaluate the effect of subbase layer on concrete performance.
The existing subbase material exhibited a much finer mix compared to the new subbase which greatly
affects the drainage properties of the materials. The existing subbase was considered impermeable, and
the new subbase is very drainable. The other half was constructed on a well-draining, special select
subbase which showed extremely high drainability values ranging from 198 ft/day to 288 ft/day, well
exceeding the specification requirement of 7.7 ft/day. The subgrade soil below the pavement structure
consist of a wet clay. During reconstruction, subgrade undercuts were performed at locations with
unstable grade, followed by the installation of a 4.0 inch underdrain and backfill.

5.3.2 Performance
The main purpose of the test road was to study the effect of freeze-thaw on pavement performance. All
JRCP sections did not exhibit any distresses related to freeze-thaw problems such as joint deterioration
or D-cracking. The ATPB was a major factor in preventing D-cracking along with a good air-void system
of the concrete. Mid-panel deflections were measured and there were smaller deflections underneath
the well-draining subbase compared to the existing poor subbase. Dowel bar looseness was also
observed which contributed to higher deflections and poor load transfer. After 23 years, all section
performed well with the exception of section B. Section B (i.e., aggregate type was blast furnace slag)
exhibited significant full lane width mid-panel cracks in about 75 percent of the truck-lane panels which
was followed by crack spalling. Full depth repairs were made after 19 years of service. It was observed

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 141

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

from coring that some minor deterioration of the ATPB at the crack edge occurred in isolated case which
caused some erosion issues in section B. All joints did not exhibit pumping and joint faulting was less
than 0.04 inches. Overall, the project performed very well regarding freeze-thaw resistance, durability,
drainage, and distresses. The freeze-thaw performance was attributed to the well-draining ATPB layer
which prevents water from accumulating at the bottom of the PCC layer.

5.3.3 Lessons learned

 A well-draining base/subbase structure improved pavement performance as well as freeze-thaw
resistance.
 Higher mid-panel deflections were observed for the “existing” poor-draining subbase compared
to the well-draining subbases.
 The ATPB has the potential to drain water from pavement, prevent pavement becomes
saturated, which eliminating the effect of freeze-thaw damage and moisture related distresses.

5.4 Ontario, Canada

The Ontario Ministry of Transportation (MTO) is responsible for the management of 10,300 miles of
paved roads and rigid pavements are about 6% of the total. The Ontario MTO conducted several
forensic investigations and gathered the information from other highway agencies in North American to
develop specification for three types of open graded drainage layers (OGDL). They categorized the OGDL
into three types; (1) untreated, (2) asphalt cement treated, and (3) Portland cement treated. Beginning
in the early 1980s, the MTO constructed a series of test sections to monitor the performance of the
drainage system and pavement. The key design considerations for the OGDL layers include:

 Permeability of the OGDL (ensure water movement away from the travel lanes).
 Stability/strength (to allow proper placement and compaction as well as support for the
pavement surface).
 Collector system (ensure water entering the pavement will be moved away from the travel lanes
and ensure long-term performance of the system – not clog).
 OGDL protection (ensure the OGDL and drainage system are not clogged by fine aggregate and
soil particles reducing the system permeability).

Based on the key findings of the OGDL research investigation, the MTO developed new specifications
requiring that a 4 inch layer of OGDL be placed beneath the concrete slab in all new rigid pavement
designs (Marks et al. 1992, Hajek et al. 1992, Bradbury and Kazmierowski 1993, and Kazmierowski et al.
1999). The gradation of the OGDL consists of coarse aggregates retained on the No. 4 sieve size. As
untreated aggregates were not considered to be stable enough to support construction traffic without
distortion, the OGDL is treated with 1.8 percent asphalt cement. In addition, the longitudinal drainage
system was modified to be integral with the OGDL to ensure that water entering the system exits the
pavement as soon as possible. The OGDL should be extended 3 feet past the edge of the concrete
pavement or paved shoulder, if present.

5.4.1 Design and Construction

This section demonstrates the design and construction of highway using three types of OGDL basses.
The highway 115 is located near the city of Pererborough. A section of Highway 115 Pererborough with
a total length of 10.20 mile was built in 1991 with three different OGDL to evaluate the performance of

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 142

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

each type. Section one (0.6 mile long) consisted of 8 inches JPCP, on 4 inches of untreated OGDL with
increase in percent of passing No. 4 to increase layer stability, on 4 inches of aggregate base, over 12
inches of aggregate subbase. Section two was similar to section one, but with 4 inches of cement
treated base (200 lb/yd3) instead of untreated OGDL. Section three was similar to section one, but with
asphalt cement treated base (1.8 percent) instead of untreated OGDL. The purpose of 4.0 aggregate
base was used to act as the filter layer between OGDL and the subgrade. The longitudinal subdrain was
placed under the shoulder, 2 feet away from the lane edge. A 4 inches diameter outlets was placed at
330 feet intervals to roadway ditches. The cement treated OGDL was placed with the concrete slipform
that was used to place the concrete and there was no issues during the placement reported. Minor
damage of the surface of the cement treated OGDL was observed during the placement of concrete
pavement. The cement treated OGDL was cured by water “sprinkling” every 2 hours for 8 hours. The
asphalt treated OGDL was placed using hot mix asphalt paver with no placement issues. The untreated
OGDL was placed using trucks and a grader to achieve the design profile.

5.4.2 Performance
Laboratory testing was conducted to evaluate the permeability of the three OGDL. The results indicated
that all three types of OGDL met the initial permeability and stability requirements. The untreated OGDL
was able to carry construction traffic without any significant damage. The FWD testing was conducted in
1992-1993 indicted that the deflection of cement treated OGDL was 17 percent less than asphalt treated
OGDL and about 28 percent less than untreated OGDL. In general, the performance of the Highway 115
pavement has been excellent. There were several issues related to late season construction that were
documented on this contract including; cold weather concrete delivery, subgrade instability in cut to fill
transition, and premature cracking due to late saw cutting of transverse joints.

In 2005, a pavement evaluation was undertaking to identify and prioritize concrete pavement
restoration requirements for the pavements. The evaluation included a detailed pavement surface
condition survey, Falling Weight Deflectometer (FWD) testing, subgrade and pavement layer materials
testing, MIT Scan to check dowel bar alignment, and Ground Penetrating Radar (GPR) testing and test
pits at the side of the roadway to verify the operation of the drainage system. The results of the
pavement investigation completed in 2005 revealed 0.5 percent cracked slabs (2 slabs) in the eastbound
direction of the highway and 2.4 percent (50 slabs) in the westbound direction at an age of 13 years. By
this time, the pavement had carried approximately 4.67 million equivalent single axle loads (ESALs).
These slabs were replaced in a 2006 construction contact. The majority of the slab replacements were at
cut to fill transition areas. In addition, the concrete pavement was diamond ground for
smoothness/friction in 2011 and then grooved in 2014. As of 2017, the pavement had carried
approximately 13.3 million ESALs.

5.4.3 Lessons learned

 Open graded drainage layers and their drainage system should be protected from the intrusion
of fines. The movement of fine soil particles such as silty clay by pumping action of repetitive
axle loading can lead to early pavement failures.
 OGDL layers should be separated from the subgrade through the use of a granular layer. The use
of granular layers was found to be more effective than using a geotextile.

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 143

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

 The continuity of the OGDL and subdrain system to remove water from the pavement is critical
to prevent water from being trapped within the pavement structure.
 The OGDL should not be left uncovered for long periods or over the winter.
 Asphalt treated OGDL can more easily be completed when the layer has been allowed to cool
below a temperature of 150oF.

6. SUMMARY AND CONCLUSION

The goal of this paper was to identify and document the useful information related to the effect of the
pavement foundation on the performance of concrete pavements. The function of pavement foundation
include prevention of pumping, protection against frost action, drainage, prevention of volume change
of the subgrade, increased structural capacity, and a stable construction platform. The primary function
of base is to prevent pumping, and therefore it must be free draining or highly resistant to erosion.
Various case studies of the effect of pavement foundation on concrete performance were summarized
based on data/field investigations. These case studies show the effect of base/subbase in terms of
increasing or decreasing the overall pavement performance. The case studies included in this paper
documents the substantial effect of drainage on the structural and functional performance of concrete
pavements. Poorly designed or constructed drainage systems have a detrimental effect of pavement
performance, while removal of water through well-designed and well-constructed drainage system is
crucial for long term pavement performance in areas where the potential for moisture damage is high.

7. REFERENCES
AASHTO M155-04. (2004). Standard Specification for Granular Material to Control Pumping under
Concrete Pavement. American Association of State Highway and Transportation Officials,
Washington, DC.

AASHTO M147-17. (2017). Standard Specification for Materials for Aggregate and Soil-Aggregate
Subbase, Base, and Surface Courses. American Association of State Highway and Transportation
Officials, Washington, DC.

AASHTO T222-81. (2017). Standard Method of Test for Nonrepetitive Static Plate Load Test of Soils and
Flexible Pavement Components for Use in Evaluation and Design of Airport and Highway
Pavements. American Association of State Highway and Transportation Officials, Washington,
DC.

ACPA. (1995). Subgrades and Subbases for Concrete Pavements. Engineering Bulletin TB011P. American
Concrete Pavement Association.

ACPA. (2007). Subgrades and Subbases for Concrete Pavements. Engineering Bulletin EB204P. American
Concrete Pavement Association.

ASTM D1196/ D1196M-12. (2016). Standard Test Method for Nonrepetitive Static Plate Load Tests of
Soils and Flexible Pavement Components, for Use in Evaluation and Design of Airport and
Highway Pavements. Book of Standards, 04.03, ASTM International, West Conshohocken, PA.

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 144

 GAP 2019 Proceedings | www.GAP2019.com | 4 – 7 November 2017 | Colorado Springs, CO

Bradbury, A., and Kazmierowski, T.J. (1993). Field Evaluation of Various Types of Open Graded Drainage
Layers, Transportation Association of Canada Annual Conference, Ottawa, Ontario.

Christopher, B. R., Schwartz, C., and Boudreau, R. (2006). Geotechnical Aspects of Pavements, FHWA
NHI-14-014. Federal Highway Administration, Washington, DC.

Elfino, M. K., and Hossain, M. S. (2007). “Subsurface Drainage and Premature Distresses in Concrete
Pavement: A Case Study in Virginia.” Transportation Research Record, No. 2004(1), pp. 141-149.

Hajek, J., Kazmierowski, T.J., Sturm, H., Bathurst, R.J., and Raymond, G.P. (1992). Field Performance of
Open Graded Drainage Layers, Transportation Research Board Annual Meeting, Washington,
D.C.

Hein, D.K., Rao, S., Tayabji, S., and Lee, H. (2017). Bases and Subbases for Concrete Pavements, Report
No. FHWA-HIF-16-005, Federal Highway Administration, Washington, DC.

Hall, J. W., Mallela, J., and Smith, K. (2005). Stabilized and Drainable Base for Rigid Pavement – A Design
and Construction Guide. Report No. IPRF-01-G-002-021(G). Innovative Pavement Research
Foundation, Airport Concrete Pavement Technology Program, Federal Aviation Administration.

Hossain, M. S, and Elfino, M. K. (2005). Forensic Investigation of Concrete Pavement: U.S. 460,
Appomattox Bypass. VTRC 06-R9. Virginia Transportation Research Council, Charlottesville,
Virginia.

Hansen, W., Stanton, J.F., and Bennett, A. (2017). US-23 Aggregate Test Road Long-Term Performance
Evaluation. Report No. SPR-1652. Michigan Department of Transportation. Lansing, MI.

Kazmierowski, T.J., Marks, P. and Anderson, P., (1999). Development of Cement-Treated Open Graded
Drainage Layer Placement Techniques in Ontario, Transportation Research Record 1673, Paper
No. 99-0410, Washington, D.C., 1999.

Marks, P., Hajek, J, Sturm, H and Kazmierowski, T.J. (1992). Ontario’s Experience with Pavement
Drainage Layers, Transportation Association of Canada Annual Conference, Québec City,
Québec.

Snethen, D. R., Johnson, L. D., Patrick, D. M. (1977). An Evaluation of Expedient Methodology for
Identification of Potentially Expansive Soils, Report No. FHWA-RD-77-94. Federal Highway
Administration, Washington, DC.

Quiroga, H. (1992). Aggregate Durability and the Performance of PCC Pavements. Materials and
Technology Engineering and Science (MATES). No. 70. Michigan Department of Transportation. Lansing,
MI.

Published by Minerva TRI |Austin, Texas, USA | www.minervatri.com 145

View publication stats