Investigation of Grouted Dowel Connection For Precast Concrete Wa
Investigation of Grouted Dowel Connection For Precast Concrete Wa
Investigation of Grouted Dowel Connection For Precast Concrete Wa
Scholarship@Western
12-15-2016 12:00 AM
Part of the Civil Engineering Commons, Construction Engineering and Management Commons, and
the Structural Engineering Commons
Recommended Citation
Provost-Smith, Douglas J., "Investigation of Grouted Dowel Connection for Precast Concrete Wall
Construction" (2016). Electronic Thesis and Dissertation Repository. 4298.
https://ir.lib.uwo.ca/etd/4298
This Dissertation/Thesis is brought to you for free and open access by Scholarship@Western. It has been accepted
for inclusion in Electronic Thesis and Dissertation Repository by an authorized administrator of
Scholarship@Western. For more information, please contact wlswadmin@uwo.ca.
ABSTRACT
The grouted dowel connection is a simple and cost-effective connection used in many
precast concrete structural systems. The required dowel length is currently designed as a
regular bar in reinforced concrete, which underestimates the bond strength, thus resulting
in excessive connection lengths. Furthermore, precast wall construction continues
throughout cold weather, where in-situ heating of the grout used in the grouted dowel
connections is usually conducted for short periods of time. Hence, early-age exposure to
subfreezing conditions may affect the quality of the grout and subsequently the bond
strength of the connection, which can compromise structural integrity.
In this thesis, representative pullout tests were performed and their results were compared
with relevant data in the open literature in order to develop a reliable design equation for
predicting the required dowel development length. The equation was found to produce
results three times smaller than that determined by the ACI 318-14 code, while being
desirably 10% more conservative than equations proposed in previous research.
The effect of subfreezing exposure on the bond strength of the connection, along with the
mechanical properties, hydration process and pore size distribution of the grout were also
examined. Grout specimens were initially cured at ambient temperature (23 ± 1°C) for
one day and then placed inside an environmental chamber at -10°C. The compressive
strength of the grout was monitored at additional temperatures of 1°C and -20°C. It was
found that early-age subfreezing curing temperatures reduced the compressive strength of
the grout, leading to increased dowel embedment length to achieve bar fracture. The bond
strength of the connection remained proportional to the square root of compressive
strength, even when subsequent to early-age subfreezing exposure.
Keywords: Precast concrete, bond strength, connection, grout, duct, development length,
cold weather, curing, compressive strength, low temperature.
ii
CO-AUTHORSHIP STATEMENT
All experimental work, data analysis, and writing of the initial versions of this thesis were
performed by Douglas Provost-Smith. Mohamed Elsayed helped in conducting the
pullout tests. Dr. Moncef Nehdi developed the original concept of the project and created
the relationship with, and ensured the support from, the Industry Partner of this project,
who fabricated full-scale test specimens at their Precast Plant. He also provided guidance
and supervision, and helped in the development of the final versions of the publications.
Chapter 3 of this thesis was submitted for peer review and possible publication in the ACI
Structural Journal. Chapter 4 of this thesis was submitted for peer review and possible
publication to the Construction and Building Materials Journal. An abridged version of
Chapter 4 of this thesis was published as part of the 5th International Materials Specialty
Conference of the Canadian Society for Civil Engineering in London, Ontario, June
2016.
iii
ACKNOWLEDGMENTS
I would like to express my sincerest thanks to my supervisor Dr. Moncef L. Nehdi. His
guidance, advice, and encouragement have been incredibly valuable and instrumental to
the completion of my thesis.
Thank you to Mohamed (Gamal) Elsayed for his continued support. He provided
valuable discussions, and helped with experiments in the Structures Laboratory.
I would like to thank Wilber Logan for his assistance, suggestions, and valuable insight
throughout my experimental work.
I am grateful to Stubbe’s Precast for their strong support of this project. Manufacturing
the test specimens at their precast plant in Harley, Ontario was crucial for the progress of
these experiments.
Thank you to all the staff in the Department of Civil and Environmental Engineering for
the continued support throughout my period of study.
Finally, I would like to thank my Grandparents, who have always been my biggest
supporters. I would not be where I am today without you.
iv
TABLE OF CONTENTS
Abstract ............................................................................................................................... ii
Acknowledgments.............................................................................................................. iv
1 Introduction .................................................................................................................... 1
v
2.4.2 Effect on Bond Strength............................................................................ 19
3 Investigation of Grouted Dowel Connections for Precast Concrete Wall Panels ........ 23
vi
4.4.4 Thermogravimetric Analysis .................................................................... 57
Appendices ........................................................................................................................ 78
vii
LIST OF TABLES
Table 4.2: Grout tensile strength, Young’s modulus, and Poisson’s ratio ........................ 56
Table 4.3: Relative mass change (in %) and calculated content (in %) of the
corresponding hydration products..................................................................................... 59
Table 4.4: Average pore size and total intrusion volume from MIP test .......................... 60
Table A.1: Compressive strength data of concrete cured at ambient (23 ± 1°C) ............. 78
Table A.5: Tensile strength data of concrete cured at ambient (23 ± 1°C) ...................... 81
Table A.7: Young’s Modulus data for ambient (23 ± 1°C) curing ................................... 82
Table A.9: MIP data for grout cured for 1 day at ambient (23°C) ................................... 84
Table A.10: MIP data for grout cured for 7 days at ambient (23°C) ................................ 86
viii
Table A.11: MIP data for grout cured for 3 days at -10°C ............................................... 88
Table A.12: MIP data for grout cured for 7 days at -10°C ............................................... 90
ix
LIST OF FIGURES
Figure 2.1: Emulative precast wall connections using: a) grouted splice sleeve, and b)
grouted dowels; c) Section A-A. ......................................................................................... 7
Figure 2.2: Load transfer between deformed steel and concrete (ACI Committee 408,
2003). .................................................................................................................................. 8
Figure 3.1: Typical grouted dowel connection. .................Error! Bookmark not defined.
Figure 3.3: Specimen damage: (a) hairline cracking of concrete block; (b) grout conical
failure. ............................................................................................................................... 30
Figure 3.5: Tensile splitting failure of the concrete block for specimen 12CN-1. ........... 32
Figure 3.6: Reduction in strength due to absence of duct and/or eccentricity. ................. 33
Figure 3.7: Tensile splitting through grout cylinder of specimen 12CN-1. ...................... 33
Figure 3.8: Tensile splitting of surrounding concrete for specimen 12ED-1. .................. 34
Figure 3.9: Tensile splitting through grout and surrounding concrete (specimen 12EN-2).
........................................................................................................................................... 35
Figure 3.10: Bar stress-displacement plots for each embedment length. ........................ 37
Figure 3.11: Normalized bond stress-slip response for each embedment length. ............ 38
Figure 3.13: Normalized bar stress versus normalized embedment length. ..................... 41
x
Figure 3.14: Predicted versus experimental peak bar stress comparison.......................... 42
Figure 4.3: Internal grout temperature for varying subfreezing conditions. ..................... 53
Figure 4.4: Ratio of compressive strength for varying initial curing periods to the 7-day
strength of specimens initially cured for 1 day. ................................................................ 55
Figure 4.5: Ratio of 28-day to the 1-day tensile and compressive strengths for varying
curing temperatures. .......................................................................................................... 56
Figure 4.6: DTG curves for grout specimens at different ages and curing conditions. .... 58
Figure 4.7: TGA curves for grout specimens at different ages and curing conditions. .... 58
Figure 4.8: Incremental pore intrusion volume versus pore diameter for grout specimens
at different ages and curing conditions. ............................................................................ 60
Figure 4.9: Bar stress-displacement plots for each embedment length and curing regime.
........................................................................................................................................... 62
Figure 4.10: Failure modes of pullout specimens: (a) bar pullout; (b) bar fracture. ......... 63
Figure 4.11: Normalized bond stress versus normalized embedment length. .................. 64
Figure 4.12: Bond stress-slip response for each embedment length and curing regime
compared to established models. ...................................................................................... 66
xi
NOMENCLATURE
db = bar diameter
fn = confining pressure
ld = development length
l em b = embedment length
Rs = strength ratio
s = bar slip
δy = yielding displacement
xii
δu = ultimate displacement
μΔ = ductility ratio
xiii
1 INTRODUCTION
1.1 Background
Precast-concrete load-bearing wall panels have become a popular choice for low-,
medium-, and high-rise construction in North America. The repetitive nature of this
structural system allows for the mass production of high quality precast elements ready to
be quickly assembled on site. These panels are precast in an environmentally controlled
manufacturing facility, allowing for high quality control and constant production,
regardless of weather conditions. They also offer ease and speed of erection on site;
indeed, typical construction speeds of one floor per week and better are often achieved.
The reduction in site formwork and site labour, along with the speed of construction,
greatly reduce costs, often making it a more cost-effective alternative to traditional cast-
in-place construction.
An integral part of this structural system, which needs particular attention, is the
connections between individual panel units, since it directly affects the strength and
stability of the structure. The horizontal connection of precast concrete load-bearing wall
panels requires vertical continuity across joints. To achieve this, designers prefer to use
emulative connections rather than jointed connections as they perform similarly to a cast-
in-place structure, and can therefore be designed similarly. There are two common
methods of achieving emulative wall panel connections: mechanical splices, also referred
to as grouted splices; and the grouted dowel connection. There is much research
demonstrating the feasibility of grouted splices (Einea, Yamane, & Tadros, 1995;
Jansson, 2008; Ling, Ahmad, & Ibrahim, 2014). However, they generally have poor
construction tolerances and are costly due to the laborious fabrication required.
In the grouted dowel connection, a reinforcing bar protruding from the lower wall panel
is grouted into a corrugated steel duct cast into the upper wall panel, as shown in Fig. 1.1.
This connection is advantageous owing to its simplicity, favourable construction
tolerances, and cost-effectiveness. In this system, the grouted dowel carries the tensile
forces across the wall panel connection. Since ductility is required in this system, dowels
1
are required to undergo yielding, similar to regular reinforced concrete. Although there
are specific code requirements for the use of mechanical splices, there are none for the
use of the grouted dowel connection, and research on this connection is sparse.
Currently, this connection is designed using the development length equations for a
regular reinforcing bar in concrete (ACI Committee 318, 2014). However, this design
approach does not account for the additional confinement effect due to the presence of
the corrugated steel duct. Research has shown that when greater confinement is provided,
this method can lead to greatly overdesigned embedment lengths (Einea, Yehia, &
Tadros, 1999).
Previous research on the use of this connection for precast bridge bent caps demonstrated
that the ACI 318-05 Equation 12-1 produced development lengths three times longer than
necessary (Steuck, Eberhard, & Stanton, 2009). However, the latter research was
performed using larger diameter bars (32 mm to 57 mm) with larger duct diameter/bar
diameter ratios of 3.6, whereas a typical wall panel connection uses a 25 mm diameter
bar with a duct diameter/bar diameter ratio of 3.0. Steuck et al. (2009) research was also
conducted with 55 MPa grout, and may not be accurate for lower grout strengths.
2
1.2 Cold Weather Construction
Another major advantage of the precast wall structural system is that construction can
continue throughout adverse weather conditions, including cold weather. This is a
particular advantage since Canada has a cold climate that experiences subfreezing
temperatures over long periods of time, which can significantly slow or halt concrete
construction. Winter construction of cast-in-place concrete requires the heating of large
areas for extended periods of time, and typically involves the use of accelerators and/or
other admixtures to ensure the development of adequate compressive strength. However,
these methods can significantly increase energy consumption and cost, as well as make it
difficult to maintain consistent quality control.
Although the construction of precast concrete wall panels can continue throughout cold
weather conditions, the grouted dowel connection requires the placement of fresh grout.
For this purpose, the entire floor is blanketed and heated while the grout is mixed and
poured. After approximately one day, the heating is stopped and the connection is
exposed to subfreezing temperatures before the grout is fully cured. This can significantly
affect the quality of the grout and subsequently the bond strength of the connection,
which can therefore compromise structural integrity.
Research on the effects of cold weather curing on the bond strength is sparse. The author
could not access any studies in the open literature examining the effect of subfreezing
curing on the bond strength. Rather, several researchers focus on the effects which
subfreezing exposure had on the compressive strength of concrete since the latter is used
to calculate the development length required based on the ACI 318-14 equation
(25.4.2.3). However, Gardner and Poon (1976) investigated the effect of 2°C curing on
the bond strength of concrete and concluded that the bond strength remained proportional
to the square root of compressive strength, irrespective of the temperature or cement type.
Yet, there is a need to further investigate this effect, especially for the grouted precast
wall panel connection, where little is currently known.
3
1.3 Research Objectives
The primary objective of the research presented in this thesis is to acquire an advanced
understanding of the behaviour of grouted dowel connections used in precast concrete
wall panel construction. The specific research objectives are:
Chapter 2 provides a critical review of the mechanisms of bond, existing precast concrete
wall panel connections, and the effect of cold weather curing on compressive strength
and bond. Previous relevant investigations have been reviewed and discussed; with
particular on the effects of variable cold weather curing on the compressive strength of
concrete.
Chapter 3 examines the behaviour of the grouted dowel connection for use in precast wall
panels. The results are analyzed and compared to existing data to develop a rational
4
design equation for predicting the required development length of this connection. A
bond stress-slip model is also proposed and analysed in light of existing research.
Chapter 5 presents a summary and the main conclusions drawn from the research study,
along with recommendations for future research.
1.5 References
ACI Committee 318. (2014). Building Code Requirements for Structural Concrete (ACI
318-14) and Commentary (ACI 318R-14). American Concrete Institute.
Farmington Hills, MI.
Einea, A., Yamane, T., & Tadros, M. K. (1995). Grout-filled pipe splices for precast
concrete construction. PCI Journal, 40(1), 82–93.
Einea, A., Yehia, S., & Tadros, M. K. (1999). Lap splices in confined concrete. ACI
Structural Journal, 96(6), 947–955.
Gardner, N. J., & Poon, S. M. (1976). Time and Temperature Effects on Tensile , Bond ,
and Compressive Strengths. ACI Journal Proceedings, 73(7), 405–409.
Jansson, P. (2008). Evaluation of Grout-Filled Mechanical Splices for Precast Concrete
Construction. Michigan Department of Transportation. Lansing, Michigan.
Ling, J. H., Ahmad, A. B., & Ibrahim, I. S. (2014). Feasibility study of grouted splice
connector under tensile load. Construction and Building Materials, 50, 530–539.
Steuck, K. P., Eberhard, M. O., & Stanton, J. F. (2009). Anchorage of large-diameter
reinforcing bars in ducts. ACI Structural Journal, 106(4), 506–513.
5
2 LITERATURE REVIEW
2.1 Introduction
Research on precast concrete wall connections began in the late 1980s when the benefits
of precast concrete construction were beginning to be realized; the most prominent being
the precast seismic structural system (PRESSS) research program. It spanned over 10
years and concluded with the testing of a five-story precast building constructed to 60
percent scale and tested using pseudo-dynamic loading. The building consisted of four
different structural frame systems in one direction and a jointed structural wall system in
the orthogonal direction. It was found that the building performed very well in both
directions, with minimal damage to the shear wall despite seismic intensities reaching 50
percent above the design level (Priestley, Sritharan, Conley, & Pampanin, 1999). This
research demonstrated the ability of precast concrete structural systems to be used instead
of cast-in-place concrete, even under high levels of seismic excitation.
There are two types of precast concrete connections: jointed and emulative. Jointed
connections usually consist of bolted or welded steel plates, where the connection
stiffness differs from the wall stiffness, and therefore behaves differently from cast-in-
place joints. The use of these types of connections requires experimental evidence under
extreme loading conditions for special code acceptance. The research and labour required
to install these connections can make them very costly. Hence, most designers tend to
avoid them. Emulative connections are typically wet connections with rebar splicing. The
ACI/ASCE Joint Committee 550 on Precast Concrete Structures defines emulative
detailing as “the design of connection systems in a precast concrete structure so that its
structural performance is equivalent to that of a conventionally designed cast-in-place,
monolithic concrete structure” (ACI Committee 550, 2009). Since these connections
perform comparably to cast-in-place connections, they may be designed in a similar
manner; thus making such connections appealing to designers.
Typical emulative horizontal precast concrete wall connections consist of dry pack grout
and vertical continuity reinforcement. Although there are many different variations to
achieve continuity across the joint, three methods are commonly used: post-tensioning,
6
grouted splices sleeves (Fig. 2.1a), and grouted dowels (Fig. 2.1b). The first two will be
briefly reviewed, while the third is the focus of the present research. These connections
depend significantly on their bond, which is greatly increased by their high level of
confinement. Therefore, to fully understand these connections, a review of bond from
current research and relevant codes is first presented.
(a) (b)
(c)
Figure 2.1: Emulative precast wall connections using: a) grouted splice sleeve, and b)
grouted dowels; c) Section A-A.
2.2 Bond
The development length (ld) – the length required to achieve bar yield – depends on the
bond achieved between the bar and concrete, as well as the strength of the bar.
Considering the bond of deformed steel bars in concrete, the load is transferred from the
steel to the concrete through three mechanisms: chemical adhesion between the bar and
the concrete; frictional forces between surface roughness of the bar and the concrete; and
mechanical bearing of the deformed ribs against the concrete. This load transfer
mechanism is illustrated in Figure 2.2. Once initial slip occurs, the chemical adhesion is
lost and the load is transferred through friction and mechanical bearing. Furthermore,
7
since steel bars do not have significant roughness, mechanical bearing is the primary
mode for load transfer. According to ACI Committee 408R-03, “the forces on the bar are
balanced by compressive and shear stresses on the concrete contact surfaces, which are
resolved into tensile stresses that can result in cracking planes that are perpendicular and
parallel to the reinforcement” (ACI Committee 408, 2003).
Figure 2.2: Load transfer between deformed steel and concrete (ACI Committee 408,
2003).
Bond failure can occur through tensile splitting of the concrete, or a pullout failure which
results in shearing around the top surface of the ribs. The bond strength therefore depends
on the strength and volume of the concrete surrounding the bar (which resists the tensile
stresses), the surface deformations of the steel (which determine the concrete bearing
area), and the degree of confinement (which help prevent tensile splitting). Although the
two separate failure modes are known, bond failure can result from a combination of both
modes, making the bond behaviour of reinforcing bars in concrete very complex. This is
indicative of: the lack of mechanical design approach available for development lengths,
with the current models based on statistical regression analyses (ACI Committee 318,
2014; Orangun, Jirsa, & Breen, 1977), and the constant changes regarding development
lengths being made to ACI 318. Therefore, to accurately understand the bond behaviour
of specific connection details, representative testing must be done.
2.2.1 Confinement
The most common methods to increase confinement of reinforcement bonded in concrete
are through the use of transverse reinforcement (Soroushian, Choi, Park, & Aslani, 1991),
8
steel spirals (Hosseini, Rahman, Osman, Saim, & Adnan, 2015), steel pipes (Einea et al.,
1995), or increased concrete cover. These methods are provide passive confinement,
since they are reactionary in nature, and contribute through controlling propagation of
splitting cracks. Passive confinement allows the bar to fail in a shearing pullout failure
instead, which occurs at higher loads. When steel pipes have been used, their reactionary
confinement stresses have led to higher frictional resistance of the bar, resulting in much
shorter development lengths than in traditional reinforced concrete (Einea et al., 1995).
Where,
9
2.3 Precast Wall Connections
A review of the three most common emulative precast wall connections is presented with
a focus on the grouted dowel connection. To make comparisons between different bar
sizes easier, researchers normalize the embedment length in terms of the bar diameter
(db), and will be discussed similarly herein.
2.3.1 Post-Tensioning
The applicability of bonded post-tensioned shear walls has been previously demonstrated
(Hutchinson, Rizkalla, Lau, & Heuvel, 1991); nine different specimens were tested under
monotonic shear loading and equations to accurately predict their shear capacity were
developed. This research was continued by testing three different post-tensioning
configurations: bonded post-tensioned strands, bonded post-tensioned bars, and unbonded
post-tensioned bars (Soudki, Rizkalla, & Daikiw, 1995). All three configurations were
tested under reverse cyclic loading, and the two bonded connections were also tested
monotonically. It was found that the deformation capacities of all the connections were
very similar to mild steel connections; however, the energy dissipation capacities were
only about one-half compared to that of mild steel. Debonding was found to avoid
rupture of the post-tensioning steel, and more than double the deformation capacity
compared to the bonded connections. However this also resulted in a reduction in
strength and stiffness.
10
require the addition of bonded mild steel reinforcement across the horizontal joint; such
walls are called “hybrid walls”.
Hybrid walls have been shown to have greater energy dissipation capabilities compared
to that of regular un-bonded post-tensioned walls, increasing up to four times when
providing 0.75 times the amount of steel used in a comparable emulative wall (Y. C.
Kurama, 2002). The addition of mild steel also decreases the maximum lateral
displacement of the wall, and causes the response of the wall to decay faster, which
subsequently reduces the number of large displacement peaks (Holden, Restrepo, &
Mander, 2003; Y. C. Kurama, 2002; Restrepo & Rahman, 2007; Smith, Kurama, &
McGinnis, 2013). The bonding of the mild steel for hybrid walls is achieved similarly to
emulative walls, through either grouted splice sleeves, or grouted dowels (Restrepo &
Rahman, 2007). Thus, in the application of hybrid walls, significant consideration is
required in the secondary connection of the mild steel reinforcement.
The first use of non-proprietary pipe splices was proposed by Einea et al. (1995). They
tested four different splice sleeve configurations and reported that an embedment length
as short as seven times the bar diameter was capable of developing the bar. They reported
that regular steel pipe could generate a high level of confinement to the grout, preventing
tensile splitting failures and achieving high bond strength.
Further research by Einea et al. (1999) investigated a splice connection consisting of steel
spirals and multiple smaller lapping bars to splice the larger main bar. Different numbers
of lap bars were tested (one, two, and four), with two lap bars performing the best and
further tested. It was found that the steel spirals generated confinement similarly to
transverse reinforcement, and allowed the bars to be developed in just seven times the bar
11
diameter. An equation was derived to determine the required development length based
on a 95% confidence value. It was compared with the ACI 318-95 code equation and the
equation proposed by Darwin et al. (1996). The ACI and the Darwin et al. (1996)
equations were shown to overestimate the required development length by at least 76%,
and 28%, respectively.
Many other researchers have investigated the application of grouted splice sleeves with
the aim of creating simple, efficient, non-proprietary splice sleeves. Most splice sleeves
consist of a thick walled steel pipe, but utilize different configurations to enhance bond,
for instance via welded bars (Ling, Ahmad, Ibrahim, & Hamid, 2012), tapering the pipe
diameter, bolts (Sayadi, Rahman, Jumaat, Johnson Alengaram, & Ahmad, 2014), steel
rings, steel spirals (Hosseini et al., 2015), internal threading (Henin & Morcous, 2015),
grout keys, or combinations of the aforementioned (Ling et al., 2014). All these splice
sleeves generated high confinement stresses resulting in large bond stresses, which
allowed them to develop the bars in very short embedment lengths (6db to 8db). Although
these splices performed well, they require extensive fabrication (welding, threading, etc.)
and have very poor construction tolerances. For these reasons, many precast
manufacturers avoid them and prefer using grouted dowels.
12
available demonstrating the satisfactory use of this connection under full-scale seismic
testing. However, sparse research is available on the bond of this connection, specifically,
a reliable equation to accurately predict the required development length is yet to be
developed.
Raynor et al. (2002) elaborated upon previous research on the development of a precast
concrete framing system able to resist seismic loads and outperform conventional cast-in-
place moment frames, referred to as a hybrid frame (Priestley et al., 1999; Stone, Cheok,
& Stanton, 1995). The hybrid frame uses a combination of unbonded post-tensioned steel
and bonded mild steel, both of which pass through ducts in the beams and columns,
similar to the hybrid wall discussed earlier. The bonded steel bars are debonded in the
beam at a specified length to prevent premature fracture due to high strain concentration
at the beam-column interface. A finite element model was developed to generate a
reliable equation to predict the required unbonded length. Since the results were
proprietary to the company sponsoring the work, much data was withheld, and only the
results of the bond stress-slip model were published. This makes it difficult for
researchers to use/compare test results for future work. Furthermore, fiber-reinforced
grout was used in this study, which is not typical of wall panel connections; further
testing should be done with the desired grout type since it can play a significant role on
the bond behaviour of the connection.
The design of lightly reinforced precast concrete wall panels for use in low-rise buildings
subjected to high seismicity has previously been examined (Crisafulli, Restrepo, & Park,
2002). The wall-foundation connection consisted of a bed of dry-pack grout and two 16
mm diameter dowels with embedment lengths equal to 43db grouted into 50.8 mm (2 in.)
diameter corrugated ducts. This structural system was found to be suitable for use in low-
rise buildings with an abundance of wall panels subjected to high seismicity. The wall
panels can be lightly designed and experience minimal structural damage since the
plasticity concentrates at the connection region and is unable to spread through the wall
panel. Although the performance of the connection was deemed satisfactory, the
embedment length (43 db) was likely overdesigned; strain measurements taken along the
13
connecting bar found peak strains develop at the connection and decrease nearly linearly
over 20 db once the yield strength is reached.
The use of grouted dowel connections for a precast bridge bent cap system was
investigated by Steuck et al. (2009). A total of 14 pullout tests were conducted on bars
with varying sizes and embedment lengths, with or without the addition of polypropylene
fiber reinforcement. A non-linear finite element model was developed and used to
establish a design equation to predict the required development length for this
connection. It was found that this new equation requires a development length of 6.5 db,
while the ACI 318-05 equation 12-1 requires a development length of 20 db. The ACI
equation appears to be grossly conservative for this connection, requiring a development
length three times greater than what was found in this research. However, the test setup
was done to replicate a bridge bent cap system, using a much larger concrete block (914
mm diameter) than a wall connection would require. This extra concrete will affect the
connection since it provides greater confinement effects. Additionally, this research
focused on large diameter bars (db = 32 mm, 43 mm, and 57 mm) with a larger duct
diameter/bar diameter ratio (3.6). Typical wall panel connections use 25 mm diameter
bars with a duct diameter/bar diameter ratio of 3.0. For this reason, further testing is
required to replicate the connections intended use.
More recently, this connection method has been examined for use in column-to-
foundation connections subjected to high seismicity (Belleri & Riva, 2012; Popa,
Papurcu, Cotofana, & Pascu, 2015). Belleri and Riva (2012) tested and compared the
grouted dowel connection to cast-in-place and pocket-foundation connections. The
grouted dowel connection performed exceptionally well, with the damage localized to the
grout layer between the column and foundation, creating a simpler repair than that for
cast-in-place concrete or pocket foundations. High ductility of the connection related to
the confining effect of the corrugated sleeves was noticed; the high level of confinement
also prevented the reinforcement from buckling. Popa et al. (2015) compared grouted
dowel connections to cast-in-place concrete and had similar findings to that of Belleri and
Riva (2012), noting that the cast-in-place specimen dissipated more energy, but also had
more severe damage than that of the precast specimen. In both studies, it was concluded
14
that the grouted dowel connection is sufficient for use in a column-foundation connection
subjected to high seismicity. However, in both cases the connection length was designed
similarly to a reinforcing bar in concrete, likely overestimating the required embedment
length.
The reinforcing bars in the grouted dowel connection behave differently than bars cast
into regular reinforced concrete due to the confining effect of the duct (Belleri & Riva,
2012; Raynor, Dawn, & Stanton, 2002). Currently, the required development length is
calculated as a regular reinforcing bar in concrete. This greatly underestimates the bond
strength of this connection, since it does not accurately account for the confinement
effects generated by the corrugated duct. The peak bond stress is much higher than that of
reinforced concrete having comparable compressive strength, and the required
embedment length can be reduced below requirements of the current standards (Raynor et
al., 2002; Steuck et al., 2009). It has been reported that a well-confined connection can be
overestimated by as much as 97% by the ACI code equation (Einea et al., 1999);
therefore, further testing should be carried out to accurately predict the required
embedment length for this grouted dowel connection.
15
13%, 26%, and 70% of the 28 day strengths for cement Types II, I, and III, respectively.
However, when accelerating admixtures were used, the concrete performed much better,
gaining 56%, 66%, and 88%, of the 28 day strengths for cement Types II, I, and III,
respectively. Furthermore, after 28 days, when the -4°C concrete was exposed to moist
ambient conditions it reached or exceeded the long-term (3 months and 1 year) strength
of the corresponding control concrete continuously cured at ambient conditions.
To determine the feasibility of concrete construction in the Arctic Sea, Aitcin et al.
(1985) initially cured concrete at 4°C for 3 to 15 hours, and then placed it inside seawater
at a temperature of 0°C until testing. It was observed that as the initial curing period
increased, the compressive strength increased; and that the design strength of the concrete
could be achieved if it was initially cured for a minimum of 9 hours at 4°C before being
exposed to the 0°C seawater. The 28 and 56-day strengths were comparable to, and in
some cases exceeded, that of concrete cured under room temperature conditions. The
internal temperatures of the specimens were monitored, and it was found that due to the
small specimen size, the concrete had reached the seawater temperature very rapidly.
Although the temperature was cold, it remained above freezing, which could affect the
degree of hydration significantly.
Gardner and Poon (1976) tested the mechanical properties of Type I and Type III cement
concretes cured at low temperatures. The specimens were initially cured at 22°C for
either 1, 3, or 7 days, before curing continued at 22°C, 13°C, and 2°C. The results
showed that prolonged ambient curing increased compressive strength at early ages, but
had negligible effect after 7 days. This implies that relatively low temperature curing had
no detrimental effects on the long-term compressive strength.
Gardner et al. (1988) expanded on this work by curing concretes made with Type I and
Type III cements at varying w/c ratios (0.35, 0.45, and 0.55) in 0°C ocean water. Their
results indicated that the 0°C curing was only detrimental to the compressive strength at a
w/c ratio of 0.55. In fact, the Type III cement cured at 0°C outperformed the ambient
specimens after 7 days for a w/c ratio of 0.35, and after 14 days for a w/c ratio of 0.45. It
16
was concluded that the maturity method was not valid for concrete cured at cold
temperatures.
Later work by Gardner (1990) showed that curing concrete made with Type I and Type
I/fly ash cement concretes at 0°C had adverse effects on the compressive strength for w/c
ratios of 0.35 and 0.55. The rate of strength gain was significantly slowed down for both
concretes. This contradicts earlier work which determined that no adverse effects were
observed when curing at 2°C (Gardner & Poon, 1976) and 0°C (Gardner, Sau, & Cheung,
1988).
Marzouk and Houssein (1995) investigated the effect of early-age exposure to cold ocean
water on the mechanical properties of high-strength concrete containing silica fume and
fly ash. They initially cured the concrete at ambient conditions for either 1, 14, or 28 days
before exposing it to ocean water tanks at varying temperatures (20, 10, 0, -5, and -10°C).
The specimens initially cured for 14 and 28 days were negligibly affected by the
subfreezing exposure. The compressive strength, tensile strength, and modulus of rupture
of the concrete specimens cured for 1 day continued to increase after exposure to
subfreezing conditions, however, at a much slower rate than specimens cured at 20°C. At
7 days, the concretes cured at -5°C, and -10°C achieved 81% and 73% of that of the
control cured for 7-day at ambient conditions, respectively. While at 28 days, the
concretes cured at -5°C, and -10°C reached 78% and 69% of the strength of the control
concrete cured for 28 days at ambient conditions, respectively. This research
demonstrates that early-age subfreezing exposure had adverse effects on the compressive
strength of high-strength concrete, with lower subfreezing temperatures resulting in lower
compressive strength gain (Marzouk & Hussein, 1995). However, the subfreezing
exposure did not halt the strength gain, and still resulted in maximum strengths of 52.2
MPa, and 47.4 MPa, for -5°C, and -10°C, respectively. The maturity method was found
to reliably predict the compressive strength gain of most specimens. However, it was
found to be inaccurate for specimens exposed to subfreezing temperatures after 1 day of
initial curing.
17
Husem and Gozutok (2005) compared the effects of low temperature curing (10, 5, 0,
-5°C) on the compressive strength of ordinary and high-performance concrete. The high-
performance concrete incorporated fly ash and silica fume, and had lower w/c ratio than
that of ordinary concrete (0.30 compared to 0.50, respectively). Specimens were cured for
7 days at the specified temperature, and then re-exposed to ambient conditions until 28
days. The high-performance concrete performed much better than the ordinary concrete,
especially at lower temperatures; however, both concretes suffered significant reductions
in strength when cured below freezing. For the -5°C curing, the high-performance
concrete had a 63% reduction in strength at 7 days, while the ordinary concrete had a
79% reduction.
Nassif and Petrou (2013) found that concrete specimens cured at -5°C achieved 50% of
the 28-day strength of their counterpart concrete cured at 20°C. This strength gain was
largely attributed to the heat of hydration raising the internal temperature above freezing
for three days. However, in their study, relatively larger volumes of concrete were cast
(750x750x300 mm slabs) than what is typically used for the grouted dowel connections.
Thus, it is unlikely that the heat of hydration will significantly raise the internal
temperature of the grouted dowel connections above freezing for up to three days,
allowing further strength gain to occur.
It has been shown that concrete can gain compressive strength when exposed to
subfreezing temperatures at early ages; though the rate of strength gains and overall
strengths are typically much lower when compared to that at ambient curing conditions.
Yet, concrete can continue to hydrate once re-exposed to ambient conditions and can
reach, or even surpass, the long-term compressive strength of concrete continuously
cured at ambient conditions. However, from the literature presented herein, it is evident
that the rate of strength gain is very dependent on the concrete mix constituents and
proportions, with some researchers having contradictory results for similar mixes.
Therefore, it is important to accurately test the desired mix under appropriate conditions
that mimic the expected field situation.
18
2.4.2 Effect on Bond Strength
Data from several studies examining the effect of cold temperatures on concrete bond
behaviour has been compiled and examined (Huang, Chang, Shih, & Lee, 1989). The
studies encompassed a temperature range of -10°C to -70°C, and included both normal
and high-strength concrete. The authors normalized the bond stress, 𝑢𝑏 , and slip, 𝑠, to the
maximum bond stress, 𝑢𝑚𝑎𝑥 , and corresponding maximum slip, 𝑠𝑢𝑚𝑎𝑥 , and plotted
𝑢𝑏 /𝑢𝑚𝑎𝑥 versus 𝑠/𝑠𝑢𝑚𝑎𝑥 for six specimens with different compressive strengths. They
determined that the temperature had little influence on the bond behaviour, and that it is
mainly dependent on the concrete compressive strength. It was also found that the
relationship between bond strength and the square root of compressive strength of
concrete remained linear at subfreezing temperatures.
Another concern with the early-age exposure to subfreezing temperatures is that frost
damage due to trapped internal water may occur and affect the overall bond behaviour of
the connection. However, a study examining the effect of freeze-thaw damage on
concrete bond has shown that even in cases where freeze-thaw damage occurred, the
bond strength remained proportional to the square root of the compressive strength of
concrete (Shih, Lee, & Chang, 1988).
The bond behaviour of concrete at low temperatures remained essentially the same as at
normal temperatures, with the bond strength remaining proportional to the square root of
the compressive strength of concrete when exposed to subfreezing temperatures (Gardner
& Poon, 1976; Huang et al., 1989), even when frost damage occurred (Shih et al., 1988).
Since the bond strength has been shown to be approximately proportional to the square
root of the compressive strength of concrete, a reduction in the compressive strength of
the concrete due to cold weather curing will not affect the bond strength as greatly. For
example: if the concrete achieves only 80% of the desired compressive strength, it still
achieves ~89.4% of the desired bond strength. However, this relationship needs to be
investigated for this particular connection and specific grout mixture. From the above
discussion of relevant research in the open literature, it can be expected that the grouted
dowel connection will continue to gain some strength when exposed to subfreezing
19
temperatures at early-age. However, due to the complexity and variable nature of
concrete, dedicated testing and quantitative assessment is needed.
2.5 References
ACI Committee 318. (2014). Building Code Requirements for Structural Concrete (ACI
318-14) and Commentary (ACI 318R-14). American Concrete Institute.
Farmington Hills, MI.
ACI Committee 408. (2003). Bond and Development of Straight Reinforcing Bars in
Tension (ACI 408-03). American Concrete Institute. Farmington Hills, MI, USA.
ACI Committee 550. (2009). Guide to Emulating Cast-in-Place Detailing for Seismic
Design of Precast Concrete Structures (ACI 550-09). American Concrete
Institute. Farmington Hills, MI, USA.
Aitcin, P.-C., Cheung, M. S., & Shah, V. K. (1985). Strength Development of Concrete
Cured Under Arctic Sea Conditions. Temperature Effects on Concrete. ASTM
International.
Belleri, A., & Riva, P. (2012). Seismic performance and retrofit of precast concrete
grouted sleeve connections. PCI Journal, 57(1), 97–109.
CPCI. (2007). CPCI Design Manual. Canadian Precast/Prestressed Concrete Institute.
Ottawa, ON.
Crisafulli, F. J., Restrepo, J. I., & Park, R. (2002). Seismic Design of Lightly Reinforced
Precast Concrete Rectangular Wall Panels. PCI Journal, 47(4), 104–121.
Darwin, D., Zuo, J., Tholen, M. L., & Idun, E. K. (1996). Development length criteria for
conventional and high relative rib area reinforcing bars. ACI Structural Journal,
93(3), 347–359.
Einea, A., Yamane, T., & Tadros, M. K. (1995). Grout-filled pipe splices for precast
concrete construction. PCI Journal, 40(1), 82–93.
Einea, A., Yehia, S., & Tadros, M. K. (1999). Lap splices in confined concrete. ACI
Structural Journal, 96(6), 947–955.
Gardner, N. J. (1990). Effect of Temperature on the Early-Age Properties of Type I, Type
III, and Type I/ Fly Ash Concretes. ACI Materials Journal, 87(1), 68–78.
Gardner, N. J., & Poon, S. M. (1976). Time and Temperature Effects on Tensile Bond,
and Compressive Strengths. ACI Journal Proceedings, 73(7), 405–409.
Gardner, N. J., Sau, P. L., & Cheung, M. S. (1988). Strength development and durability
of concretes cast and cured at 0 C. ACI Materials Journal, 85(6), 529–536.
Henin, E., & Morcous, G. (2015). Non-proprietary bar splice sleeve for precast concrete
construction. Engineering Structures, 83, 154–162.
Holden, T., Restrepo, J., & Mander, J. B. (2003). Seismic Performance of Precast
Reinforced and Prestressed Concrete Walls. Journal of Structural Engineering,
129(3), 286–296.
20
Hosseini, S. J. A., Rahman, A. B. A., Osman, M. H., Saim, A., & Adnan, A. (2015).
Bond behavior of spirally confined splice of deformed bars in grout. Construction
and Building Materials, 80, 180–194.
Huang, X. P., Chang, K. C., Shih, T. S., & Lee, G. C. (1989). Analytical Models of Local
Concrete-Steel Bond at Low Temperature. Journal of Cold Regions Engineering,
3(4), 159–171.
Husem, M., & Gozutok, S. (2005). The effects of low temperature curing on the
compressive strength of ordinary and high performance concrete. Construction
and Building Materials, 19(1), 49–53.
Hutchinson, R. L., Rizkalla, S. H., Lau, M., & Heuvel, S. (1991). Horizontal Post-
Tensioned Connections for Precast Concrete Loadbearing Shear Wall Panels. PCI
Journal, 36(6), 64–76.
Jansson, P. (2008). Evaluation of Grout-Filled Mechanical Splices for Precast Concrete
Construction. Michigan Department of Transportation. Lansing, Michigan.
Klieger, P. (1958). Effect of mixing and curing temperature on concrete strength. ACI
Journal, 54(6), 1063–1081.
Kurama, Y. C. (2002). Hybrid post-tensioned precast concrete walls for use in seismic
regions. PCI Journal, 47(5), 36–59.
Kurama, Y., Pessiki, S., Sause, R., & Lu, L.-W. (1999). Seismic Behavior and Design of
Unbonded Post-Tensioned Precast Concrete Walls. PCI Journal, 44(3), 72–89.
Ling, J. H., Ahmad, A. B., & Ibrahim, I. S. (2014). Feasibility study of grouted splice
connector under tensile load. Construction and Building Materials, 50, 530–539.
Ling, J. H., Ahmad, A. B., Ibrahim, I. S., & Hamid, Z. A. (2012). Behaviour of grouted
pipe splice under incremental tensile load. Construction and Building Materials,
33, 90–98.
Marzouk, H., & Hussein, A. (1995). Effect of Curing Age on High-Strength Concrete at
Low Temperatures. Journal of Materials in Civil Engineering, 7(3), 161–167.
Moosavi, M., Jafari, A., & Khosravi, A. (2005). Bond of cement grouted reinforcing bars
under constant radial pressure. Cement and Concrete Composites, 27(1), 103–
109.
Nassif, A. Y., & Petrou, M. F. (2013). Influence of cold weather during casting and
curing on the stiffness and strength of concrete. Construction and Building
Materials, 44, 161–167.
Orangun, C. O., Jirsa, J. O., & Breen, J. E. (1977). A Reevaluation of Test Data on
Development Length and Splices. Journal of the American Concrete Institute,
74(3), 114–122.
Popa, V., Papurcu, A., Cotofana, D., & Pascu, R. (2015). Experimental testing on
emulative connections for precast columns using grouted corrugated steel sleeves.
Bulletin of Earthquake Engineering, 13(8), 2429–2447.
21
Priestley, M. J. N., Sritharan, S., Conley, J. R., & Pampanin, S. (1999). Preliminary
Results and Conclusions From the PRESSS Five-Story Precast Concrete Test
Building. PCI Journal, 44(6), 42–67.
Raynor, D. J., Dawn, E. L., & Stanton, J. F. (2002). Bond-Slip Response of Reinforcing
Bars Grouted in Ducts. ACI Structural Journal, 99(5), 568–576.
Restrepo, J. I., & Rahman, A. (2007). Seismic Performance of Self-Centering Structural
Walls Incorporating Energy Dissipators. Journal of Structural Engineering,
133(11), 1560–1570.
Sayadi, A. a., Rahman, A. B. A., Jumaat, M. Z. Bin, Johnson Alengaram, U., & Ahmad,
S. (2014). The relationship between interlocking mechanism and bond strength in
elastic and inelastic segment of splice sleeve. Construction and Building
Materials, 55, 227–237.
Shih, T. S., Lee, G. C., & Chang, K. C. (1988). Effect of Freezing Cycles on Bond
Strength of Concrete. Journal of Structural Engineering, 114(3), 717–726.
Smith, B. J., Kurama, Y. C., & McGinnis, M. J. (2013). Behavior of Precast Concrete
Shear Walls for Seismic Regions: Comparison of Hybrid and Emulative
Specimens. Journal of Structural Engineering, 139(11), 1917–1927.
Soroushian, P., Choi, K., Park, G., & Aslani, F. (1991). Bond of deformed bars to
concrete: effects of confinement and strength of concrete. ACI Materials Journal,
88(3), 227–232.
Soudki, K. A., Rizkalla, S. H., & Daikiw, R. W. (1995). Horizontal Connections for
Precast Concrete Shear Walls Subjected to Cyclic Deformations Part 2 :
Prestressed Connections. PCI Journal, 40(5), 82–96.
Steuck, K. P., Eberhard, M. O., & Stanton, J. F. (2009). Anchorage of large-diameter
reinforcing bars in ducts. ACI Structural Journal, 106(4), 506–513.
Stone, W. C., Cheok, G. S., & Stanton, J. F. (1995). Performance of hybrid moment-
resisting precast beam-column concrete connections subject to cyclic loading. ACI
Structural Journal, 92(2), 229–249.
Untrauer, R. E., & Henry, R. L. (1965). Influence of Normal Pressure on Bond Strength.
ACI Journal, 62(5), 577–586.
22
3 INVESTIGATION OF GROUTED DOWEL
CONNECTIONS FOR PRECAST CONCRETE WALL
PANELS1
3.1 Introduction
Precast concrete construction has become more popular since the 1980’s when the precast
seismic structural system (PRESSS) research program began. The conclusion of the
research showed that precast concrete structural systems can perform as well, and in
some cases better, than cast-in-place concrete (Priestley et al., 1999). Precast concrete
load bearing wall panels are commonly used in medium- and high-rise buildings due to
their cost-effectiveness, ease and speed of erection, and high quality control achieved at
the manufacturing plant. However, an integral part of this structural system that needs
particular attention is the connection since it directly affects the structure’s strength and
stability.
In the grouted dowel connection, a reinforcing bar protruding from the lower wall panel
is grouted into a corrugated duct cast into the upper wall panel as shown in Fig. 1.1 and
Fig. 2.1b. This connection is preferred due to its simplicity, favourable construction
tolerances, and cost effectiveness. In this system, the grouted dowel carries the tensile
forces across the wall panel connection. Since ductility is required in this system, these
1
A version of this chapter was submitted for publication as: Provost, D. J., Elsayed, M., & Nehdi, M. L.,
“Investigation of Grouted Dowel Connections for Precast Wall Construction.” Submitted to ACI Structural
Journal, November 2016.
23
dowels are required to undergo yielding, similar to regular reinforced concrete. Although
there are specific code requirements for the use of mechanical splices, there are none for
the use of reinforcing bars grouted into ducts, and research on this connection is sparse.
Raynor et al. (2002) tested the use of this connection type for a hybrid frame system.
They examined the bond-deformation relationship of the connection and concluded that it
behaved differently than bars in regular reinforced concrete. For instance, peak bond
stresses were found to be higher than those possible for comparable reinforced concrete,
therefore allowing for shorter embedment lengths than required by current codes. They
also observed different damage patterns compared to those in reinforced concrete. There
was no visible cracking in the grout cylinder; instead there was significant grout crushing
surrounding the reinforcing bar lugs. This behavior is different than that of bars in
reinforced concrete, which usually exhibit extensive tensile radial bond cracks (Goto,
1971). It was concluded that the lack of visible cracking along with the obvious local
grout crushing indicate that the duct provided sufficient confinement to prevent splitting
or cracking failure. Since the results were proprietary to the company sponsoring the
work, much data was withheld, including peak bond stresses and required development
lengths; only the connection’s bond-slip response was presented.
This connection was later examined for use in a precast bridge bent system (Steuck et al.,
2009). The authors conducted a total of 14 pullout tests with varying bar sizes,
embedment lengths, and the use of fiber reinforcement. Based on the experimental
results, the following design equation (Eq. 3.1) was proposed to calculate the required
development length.
fy d db
ld db duct (3.1)
10.8 f g' 2
Where,
24
f g' = grout compressive strength (MPa);
This equation was found to produce development lengths three times smaller than that
required by ACI 318-05 equation 12-1. However, this research considered only large
diameter bars (32 mm to 57 mm) with larger duct diameter/bar diameter ratios of 3.6,
whereas a typical wall panel connection usually uses a 25M bar with a duct diameter/bar
diameter ratio of 3.0. This design equation was also developed with only 55 MPa (8000
psi) strength grout, and may not be accurate with lower grout strengths.
For most cast-in-place reinforced concrete members, a tensile splitting failure is the most
common bond failure mode exhibited. Current design codes require the development
length of this connection to be designed as a regular bar in normal reinforced concrete
(ACI Committee 318, 2014). Since this design equation is based upon test results for bar
splices in concrete, which typically fail through tensile splitting, it predicts much lower
bond stresses and consequently much longer embedment lengths than required for
reinforcing bars grouted into ducts. Researchers have demonstrated that this can lead to
greatly overdesigned embedment lengths when this failure mode is suppressed or
delayed, allowing much larger bond stresses to be generated (Einea et al., 1999).
It has been shown that the grouted dowel connection is quite different than regular bar in
reinforced concrete. However, since limited research is available on this connection, it is
not well known which parameters will influence the connections behavior. The studies
that have so far examined the bond behavior of this connection for uses in hybrid frames
(Raynor et al., 2002) and precast bridge bent cap systems (Steuck et al., 2009) have used
larger concrete covers (127 mm [5 in.], and 356 mm [14 in.], respectively) than that
typically encountered in precast concrete wall panels (64 mm [2.5 in.]). Furthermore,
neither of these studies examined the effect of eccentric bar placement within the duct, or
how the absence of the duct affects the behavior of the connection. Therefore, it is
25
important to replicate field conditions when testing grouted dowel connections in order to
accurately estimate its strength for each particular application.
Unless otherwise stated, the test bars were grouted concentrically within the duct and
extended 25 mm (1 in.) above the concrete block to measure slip during testing. To
examine the effect of eccentric bar placement within the connection, test bars were placed
directly against the edge of the duct and grouted. For specimens with no ducts present, a
76 mm (3 in.) diameter cylindrical void cast in the concrete block was grouted. Similar to
field conditions, the grout was poured from the passive end of the connection. In typical
field conditions, an additional 76 mm (3 in.) is grouted above the bar at the top end of the
26
duct to avoid the bleeding distance of the grout (Crisafulli et al., 2002). However, during
grouting, care was taken to ensure no bleeding, and this distance was removed to allow
easier measurements of bar slip at the top (passive) end.
In field construction, precast concrete wall panels are typically braced for one week.
Hence, to mimic field conditions, the test specimens were tested at 7 days. Embedment
lengths of 4, 6, 8, and 12 bar diameters were tested using a 25M bar to examine the bond
behavior envelope (elastic and plastic behavior). A much longer embedment length of 36
bar diameters was also tested since this is a typical connection length used in practical
applications. For each embedment length, two identical specimens were tested to
examine the repeatability of results. The specimen notation is as follows (lAB-n), to
indicate the embedment length (l = 4, 6, 8, 12, or 36 bar diameters), bar placement (A = C
for concentric or E for eccentric), presence of the duct (B = D for duct or N for no duct),
and n represents the specimen number.
The concrete blocks were cast using self-consolidating concrete having an average
compressive strength of 50.6 MPa (7340 psi). Commercially available non-shrink grout
was used as per industrial practice, mixing one 25 kg bag with 3.75 L of water to achieve
adequate fluid consistency. The average 7-day compressive strength of the grout was 38.4
MPa (5570 psi), measured on 3 replicate 76 x 152 mm (3 x 6 in.) cylinders. The duct
27
could not be accurately tested because disassembling it would result in plastic
deformations, therefore altering its mechanical properties. The test bars used were 25M
deformed steel with a specified yield strength of 400 MPa (58 ksi).
The pullout test setup is illustrated in Fig. 3.1. The test specimens were placed on a 216 x
216 mm (8.5 x 8.5 in.) steel bearing plate atop the active end of an open loop Tinius
Olsen testing machine. The bearing plate had a square internal void that was used to
distribute the load to a 25 mm wide outer edge of the concrete block so as to avoid
inducing compression stresses in the bonded region, which is known to affect the bond
behavior of the bar (ACI Committee 408, 2003). One strain based linear variable
displacement transducer (LVDT) was placed on the unloaded end of the bar to measure
slip. Another LVDT was attached to the testing machine crosshead, measuring its relative
movement, which represents the bar elongation. The specimen was tested at a loading
rate of 0.5 kN/s.
fs
Rs (3.2)
fy
A strength ratio exceeding 1.25 was reached with an embedment length of 8 bar
diameters. At an embedment length of 12 bar diameters (305 mm [12 in.]) – the shortest
allowable embedment length according to ACI 318-14 – the bar approached its ultimate
stress, reaching an average of 679 MPa (98.5 ksi). The degree of bar yielding appears to
have an effect on the bond stress; at an embedment length of 12 bar diameters, the bar
had undergone most yielding, approaching fracture, yet had the lowest recorded bond
stress. Steuck et al. (2009) also observed a similar trend and attributed it to the inelastic
elongation of the bar which reduced the effective cross-sectional area, partially
28
disengaging the bar from the grout, and therefore reducing the bond capacity. However,
further testing should be carried out to accurately quantify this effect.
29
(a) (b)
Figure 3.2: Specimen damage: a) hairline cracking of concrete block; b) grout conical
failure.
Conical grout failures at the free surface were also observed for all specimens (Fig. 3.2b).
This failure can be attributed to Goto cracks, which form cones at the bar lugs near the
free ends of concrete for reinforcing bars projecting from concrete and placed in tension
(Goto, 1971). The presence of this failure at an overdesigned embedment length of 36 bar
diameters confirms that it is based on the boundary conditions, and not due to other bond
parameters. Goto (1971) originally observed such cracks to have angles of = 45 to 80
degrees to the bar axis, with most being approximately = 60 degrees. This failure mode
was also observed by Steuck et al. (2009), who recorded cracks producing angles of =
45 to 60 degrees from the bar axis. In the present study, the cone varied between = 30
to 60 degrees, and in some specimens, the cone followed the spiral lug as shown in Fig.
3.2b. It is important to note that some applications of this connection require the bar to be
debonded at the free surface to prevent premature fracture (Raynor et al., 2002); in such a
case the conical failure described above will not be present.
Specimens with an embedment length of 36 bar diameters all experienced failures due to
bar fracture. Bar fracture is the preferred failure mode since it allows the bar to fully
develop, therefore offering a higher tensile capacity of the connection; and causing larger
30
bar displacement due to post-yielding bar elongation, ensuring more ductile behavior of
the connection. However, for the purpose of accurately quantifying the bond strength of
this connection, a pullout due to bond failure is desirable, which is the rationale for
exploring shorter embedment lengths in the present study.
The low duct stiffness was also observed in preliminary testing, where the present authors
tested the grouted duct without surrounding concrete. The duct split open through tensile
splitting expansion of the grout (Fig. 3.3) at around 100 kN, which is a significantly
lower force than that required to fully develop a 25 mm diameter reinforcing bar. This
31
was also observed by Ling et al. (2014) who tested a similar grouted duct for use as a
splice sleeve, failing at very low load.
Even though the duct has insufficient strength to develop the bar alone, its use is
paramount for the connection to function properly. The primary role of the duct is to
provide adequate bond between the surrounding precast concrete and the freshly placed
grout. However, the duct does also help to confine the surrounding concrete, preventing it
from a tensile splitting failure. In the absence of the corrugated duct, two additional
failure modes can occur: the entire grout sleeve can pull out from the surrounding
concrete; or the concrete can experience a tensile splitting failure. Figure 3.4 displays the
failure mode of specimen 12CN-2, where the grout sleeve began to slip before the
surrounding concrete failed through tensile splitting. These failure modes are conflict
with findings of researchers who previously investigated bars grouted into concrete
(Darwin & Zavaregh, 1996). This is likely due to the difference in the size of the grouted
cavity; since for a 25 mm diameter bar, Darwin and Zavaregh (1996) used 32 mm (1.25
in.) holes, while the current research used 76 mm (3 in.) holes.
Figure 3.4: Tensile splitting failure of the concrete block for specimen 12CN-2.
The reduction in strength due to the absence of the corrugated steel duct can be observed
in Fig. 3.5. This chart displays the ratios of failure loads for specimens with no duct
and/or cast with eccentricity to the average failure load of control specimens 12CD-1 and
12CD-2. When the duct was absent from specimens grouted concentrically (specimens
32
12CN-1 and 12CN-2), there was a 30% reduction in the connection’s strength. This is
due to the lack of confinement, therefore causing the surrounding concrete block to fail
through tensile splitting. For instance, in specimen 12CN-1, the tensile cracking split
through the grout cylinder as shown in Fig. 3.6.
1.00
0.90
0.80
0.70
P / P12CD-avg
0.60
0.50
0.40
0.30
0.20
0.10
0.00
No Duct Eccentric Eccentric with No
Duct
33
3.4.3 Effect of Eccentricity
Specimens grouted eccentrically within ducts experienced a pullout failure as described
earlier. However, these specimens achieved lower bond stresses when compared to
concentrically grouted control specimens. This was apparently due to the extensive
tensile cracking of the surrounding concrete block (Fig. 3.7). Unlike specimen 12CN-1 in
which cracks occurred in the surrounding concrete and grout (Fig. 3.6), the tensile
cracking in specimens 12ED-1 and 12ED-2 only occurred in the surrounding concrete,
but not in the duct or grout. This suppression of tensile cracking is attributed to the
confining effect of the duct. This tensile cracking reduced the stiffness and subsequently
the confinement effect of the surrounding concrete. As discussed earlier, the large bond
stress achieved by this connection is dependent upon the high confinement effect.
Therefore, this reduction in confinement led to a reduction in strength by about 10% as
shown in Fig. 3.5. When the bars are aligned eccentrically against the duct wall, the
connection generates greater concentrated tensile stresses in the surrounding concrete,
leading to the extensive tensile cracking displayed in Fig. 3.7.
Eccentric bar placement had a relatively small effect on longer connection lengths in
which specimens experienced bar fracture. When specimens with an embedment length
of 36 bar diameters were tested, eccentric specimens still failed through bar fracture,
though about 4% reduction in strength was recorded. It is inconclusive whether the
34
eccentric bar placement affects the connection in the absence of the duct; specimen
12EN-1 (eccentric with no duct) performed similarly to both specimens 12CN-1 and
12CN-2 (no duct but concentric). Yet, specimen 12EN-2 had a reduction in strength of
16% due to eccentricity, and an overall reduction of 42% due to a combined effect of the
absence of duct and eccentricity. Specimen 12EN-2 also had a very brittle failure through
tensile splitting of both the grout and surrounding concrete (Fig. 3.8).
Figure 3.8: Tensile splitting of grout and surrounding concrete for specimen 12EN-2.
35
P 103
ueff (3.3a)
.db .leff
Where,
d db
leff lemb duct (3.3b)
2 tan
With a constant duct diameter of 3db, and assuming an average angle of 45 degrees,
l eff l em b d b (3.3c)
Where,
P = load (kN);
36
The bar stress-displacement behavior corresponding to each dowel embedment length
experiencing pullout failure is displayed in Fig. 3.9. The recorded displacement is the
relative movement of the testing machine’s crossheads; this therefore accounts for both
slip and bar elongation. For all specimens, the behavior mimicked that of a regular steel
bar until the peak load was reached, then a steep drop in load was observed, followed by
a more gradual decrease in load until the test was stopped. The criterion for stopping the
test was once a slip of ~8mm was reached since it envelopes the entire bond-slip
response.
Displacement (in.)
0 0.5 1 1.5 2 2.5 3
800 140
700 120
Bar Stress (MPa)
600
Previous research has reported that the average bond stress is directly proportional to the
square root of the concrete compressive strength (Untrauer & Henry, 1965). Therefore, to
accurately analyze and compare the bond strengths and bond-slip behavior, the effective
bond strength, ueff, was normalized with the square root of the grout compressive
strength, √fg’. The normalized bond stress-slip behavior is similar for all embedment
lengths, with small differences in peak bond stress and corresponding slip (Fig. 3.10).
However, this is rather common in pullout tests, due to the variable nature of concrete
materials (Eligehausen, Popov, & Bertero, 1983).
37
Slip (in.)
0 0.1 0.2 0.3 0.4 0.5
3.5 45
4CD-2 40
3
6CD-2 35
Figure 3.10: Normalized bond stress-slip response for each embedment length.
The normalized bond stress-slip relationship was defined by fitting curves to the
experimental data for an embedment length of 8 bar diameters, which represents the
average response for all tested specimens. The corresponding equations are:
Where,
s = slip (mm).
The normalized bond stress-slip model developed herein had most variability with
specimen 4CD-2 with a mean squared error of 0.12. However, comparing model
38
predictions to experimental results of specimen 8CD-2, which represents the average
response for all tested specimens, it had a mean squared error of 0.014. The normalized
bond stress-slip model is shown in Fig. 3.11 and compared to previously established
bond models by Raynor et al. (2002) and Steuck et al. (2009) for bars grouted in ducts,
and by Eligehausen et al. (1983) for regular reinforced concrete. The differences are
apparent, with each bond model representing bars grouted in ducts predicting much larger
bond stresses than the model for reinforced concrete. This is due to the aforementioned
confinement effect of the duct.
The model developed by Raynor et al. (2002) predicts a much higher bond stress than the
other models. This may be attributed to differences in testing; in the present study and in
Steuck et al. (2009), the load was distributed to the outer edge of the concrete to prevent
inducing additional compressive stresses affecting the bond behavior. Conversely, in
Raynor et al. (2002), the concrete block was actually bolted to the test frame with steel
bearing plates, which may have induced confinement stresses increasing the bond
capacity. However, the differences in bond stress between the current model, and the
model proposed by Steuck et al. (2009) are small, and within the scatter of the test results
as can be observed in Fig. 3.12.
s/db
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
6 70
Steuck et al.
5 60
Raynor et al.
ueff/√f g ' (√MPa)
4
Eligehausen et al.
40
3
30
2
20
1 10
0 0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
s/db
39
3.4.5 Analysis of Experimental Results
The tensile bar stress versus embedment length results from the tests conducted in the
present study were plotted and compared with the results from Steuck et al. (2009).
Initially the results from Steuck et al. (2009) appeared to have higher stresses for similar
embedment lengths. However, once the bar stress was normalized with the square root of
the compressive strength of grout, the difference disappeared (Fig. 3.12). This suggests
that the duct diameter/bar diameter ratio only affects the effective length of the bar, and
not the overall strength.
It can be observed in Fig. 3.12 that the normalized tensile stress varies approximately
linearly with the normalized embedment length. This indicates that the maximum bond
stress was approximately constant in all tests, with the exception of one specimen that
fractured in the testing by Steuck (2009). A design equation (Eq. 3.5) was derived by
fitting a lower bound 99% confidence curve to the data.
fy d db
ld d b duct (3.5)
9.6 f g' 2
Where,
40
ld/db
0 5 10 15
120 1600
Present Study
Steuck 1400
100
The proposed design equation shares the same material variables as Eq. 3.1 by Steuck et
al. (2009), while accounting for lower strength grouts, resulting in desirably more
conservative design. As can be observed in Fig. 3.12, the design equation crosses the
embedment length axis at approximately 1.0; this represents the conical failure of the
grout which as mentioned, does not contribute to the bond strength. This is accounted for
in the design equation through the second term, and should be ignored when debonding
of the bar is present. This design equation is applicable for duct diameter/bar diameter
ratios of 3.0-3.6 using non-shrink grouts only; the use of different grout types has not
been investigated herein, and thus needs to be validated for specific use in this particular
connection. The data from Steuck et al. (2009) had a maximum grout compressive
strength of 70 MPa. Therefore, the proposed design equation should be limited to 70 MPa
until such data on using higher strength grouts becomes available.
41
connection involving various uncertain parameters. A comparison of the predicted bar
stress from Eq. 3.5 versus experimental peak bar stress for the present data, and data
from Steuck et al. (2009) is displayed in Fig. 3.13, where anything under the line results
in a conservative design. It can be observed that the proposed equation does not
overestimate any of the experimental data; rather it conservatively under predicts the bar
stress. While this requires slightly longer development lengths, it allows for a desirable
conservative level of safety necessary in design.
f s , Experimental (ksi)
0 20 40 60 80 100 120
800 120
700 Present Study 100
f s , Predicted (MPa)
f s , Predicted (ksi)
600 Steuck
80
500
400 60
300
40
200
20
100
0 0
0 100 200 300 400 500 600 700 800
f s , Experimental (MPa)
Assuming a specified steel yield strength of 420 MPa, grout compressive strength of 35
MPa, and a duct diameter of 3db, the proposed design equation (Eq. 3.5) results in an
embedment length of 8.4db. The current ACI 318-14 equation (25.4.2.3) greatly
overestimates this length, resulting in a development length 3.08 times larger than
required. The current ACI code does not account for the higher confinement generated by
the presence of the duct. This was also observed by Einea et al. (1999) who used spirals
to generate greater confinement effects than in regular concrete; they found that the ACI
equation overestimated the required length by almost double.
Since Eq. 3.1 developed by Steuck et al. (2009) predicts higher bar stresses, it tends to
underestimate the required length. For the same properties mentioned above, Eq. 3.1 only
42
provides 90% of the calculated required length. The equation proposed in the present
study is 10% more conservative than Eq. 3.1 for all grout strengths. An accurate
relationship between cyclic and monotonic load could not be found for this connection.
However, Steuck et al. (2009) suggested a conservative factor of 1.5 for determining the
seismic development lengths. Therefore, the development length of this equation for use
in seismic applications would approximately be 12.6db.
1. This grouted dowel connection can be used with a clear concrete cover of as low
as 64 mm (2.5 in.) for 50 MPa (7300 psi) concrete.
2. The steel duct creates a higher confinement effect than present in regular
reinforced concrete, allowing for a shear pullout failure rather than a tensile
splitting failure.
3. This failure mode occurs at significantly higher bond stresses, allowing for a
reduction in required embedment length.
43
4. The absence of the duct can reduce the strength of this connection by about 30%.
6. The current ACI equation (25.4.2.3) does not account for the greater confinement
effect of the duct, therefore greatly overestimates the required development
length.
7. The equation proposed in the present study requires development lengths of 8.4db
for yield, and 13.1db for bar fracture, assuming that a 35 MPa grout is used.
9. The equation proposed in the present study is applicable for non-shrink grouts
having a compressive strength up to 70 MPa. Using other grout types has not been
investigated herein, and thus needs to be validated for specific use in this
particular connection. The use of higher strength grouts (>70 MPa) has also not
been explored, and therefore the proposed design equation should be limited to 70
MPa until such data on using higher strength grouts becomes available.
3.8 References
ACI Committee 318. (2014). Building Code Requirements for Structural Concrete (ACI
318-14) and Commentary (ACI 318R-14). ACI 318-14. Farmington Hills, MI.
ACI Committee 408. (2003). Bond and Development of Straight Reinforcing Bars in
Tension Reported by ACI Committee 408. ACI 408-03. Farmington Hills, MI,
USA.
Crisafulli, F. J., Restrepo, J. I., & Park, R. (2002). Seismic Design of Lightly Reinforced
Precast Concrete Rectangular Wall Panels. PCI Journal, 47(4), 104–121.
44
Darwin, D., & Zavaregh, S. S. (1996). Bond Strength of Grouted Reinforcing Bars. ACI
Structural Journal, 93(4), 486–495.
Einea, A., Yamane, T., & Tadros, M. K. (1995). Grout-filled pipe splices for precast
concrete construction. PCI Journal, 40(1), 82–93.
Einea, A., Yehia, S., & Tadros, M. K. (1999). Lap splices in confined concrete. ACI
Structural Journal, 96(6), 947–955.
Eligehausen, R., Popov, E. P., & Bertero, V. V. (1983). Local bond stress-slip
relationships of deformed bars under generalized excitations. Report No.
UCB/EERC-83/23. Earthquake Engineering Research Center, Universisty of
California-Berkeley.
Goto, Y. (1971). Cracks Formed in Concrete Around Deformed Tension Bars. ACI
Journal Proceedings, 68(4), 244–251.
Jansson, P. (2008). Evaluation of Grout-Filled Mechanical Splices for Precast Concrete
Construction. Michigan Department of Transportation. Lansing, Michigan.
Ling, J. H., Ahmad, A. B., & Ibrahim, I. S. (2014). Feasibility study of grouted splice
connector under tensile load. Construction and Building Materials, 50, 530–539.
Ling, J. H., Ahmad, A. B., Ibrahim, I. S., & Hamid, Z. A. (2012). Behaviour of grouted
pipe splice under incremental tensile load. Construction and Building Materials,
33, 90–98.
Priestley, M. J. N., Sritharan, S., Conley, J. R., & Pampanin, S. (1999). Preliminary
Results and Conclusions From the PRESSS Five-Story Precast Concrete Test
Building. PCI Journal, 44(6), 42–67.
Raynor, D. J., Dawn, E. L., & Stanton, J. F. (2002). Bond-Slip Response of Reinforcing
Bars Grouted in Ducts. ACI Structural Journal, 99(5), 568–576.
Steuck, K. P., Eberhard, M. O., & Stanton, J. F. (2009). Anchorage of large-diameter
reinforcing bars in ducts. ACI Structural Journal, 106(4), 506–513.
Untrauer, R. E., & Henry, R. L. (1965). Influence of Normal Pressure on Bond Strength.
ACI Journal, 62(5), 577–586.
45
4 EARLY-AGE EXPOSURE OF GROUTED PRECAST
CONCRETE WALL CONNECTIONS TO SUBFREEZING
CURING TEMPERATURES1
4.1 Introduction
In cold climates, subfreezing temperatures can prevail for long periods of time, which can
significantly slow or halt concrete construction. The ACI Committee 306R-10 defines
cold weather as a period of three or more successive days when the average daily air
temperature drops below 4°C and does not exceed 10°C for more than one-half of any
24-h period (ACI Committee 306, 2010). Concrete must reach a compressive strength of
3.5 MPa before exposure to subfreezing temperatures; failure to do so will result in
significant reduction of both the strength and stiffness of the concrete. Winter
construction of cast-in-place concrete requires the heating of large areas for extended
periods of time, and typically includes the use of accelerating admixtures to ensure the
development of adequate compressive strength. However, such methods can significantly
increase cost, while making it difficult to maintain superior quality control.
1
Parts of this chapter were published or submitted as: Provost, D. J., Elsayed, M., & Nehdi, M. L., (2016)
“Investigation of Grouted Precast Concrete Wall Connections at Subfreezing Conditions,” Proceedings of
the CSCE 5th International Materials Specialty Conference, London, ON, MAT-719, and Provost, D. J.,
Elsayed, M., & Nehdi, M. L., “Early-age Exposure of Grouted Precast Concrete Wall Connections to
Subfreezing Curing Temperatures,” submitted to Construction and Building Materials Journal, November
2016.
46
The authors could not retrieve any studies in the open literature examining the effect of
subfreezing curing conditions on the bond strength of grouted dowel connections.
However, Gardner and Poon (1976) investigated the effect of 2°C curing on the bond
strength of concrete, and concluded that the bond strength was affected proportionally to
the square root of the compressive strength, irrespective of the temperature or cement
type. Most existing research work has focused on the effect subfreezing exposure on the
compressive strength, since the latter is used in the ACI (2014) equation (25.4.2.3) to
determine the bar development length.
Results of several studies generally indicate that the compressive strength gain of
concrete was adversely affected by early-age exposure to subfreezing temperatures. For
instance, Klieger (1958) found that curing concrete at -4°C resulted in significantly lower
28-day compressive strengths by 13% and 70%, for concretes made with Type I and
Type III cements, respectively. Gardner (1990) showed that curing concrete at 0°C had
adverse effects on the compressive strength of concrete, contradicting earlier work which
determined that no adverse effects were observed when cured at 2°C (Gardner & Poon,
1976) and 0°C (Gardner et al., 1988). Furthermore, the magnitude of strength loss of
concrete cured in cold weather varied significantly between studies. In two separate
investigations, high-strength concrete incorporating silica fume and fly ash was cured in -
5°C ocean water, resulting in 7-day strengths of 37% (Husem & Gozutok, 2005), and
81% (Marzouk & Houssein, 1990). Such inconsistent results demonstrate the importance
of testing the actual grout mixture to be used in precast wall panel construction, while
accurately replicating the subfreezing conditions to be experienced in the field.
47
structures including beam-column joints in hybrid frames (Raynor et al., 2002) and
bridge bent caps (Steuck et al., 2009), the research conducted herein on the effect of
subfreezing exposure on grouted dowel connections could be extended to the winter
construction of various precast structures such us buildings and bridges.
The main temperatures examined in this study were -10°C and ambient (T = 23 ± 1°C).
The mechanical properties of the grout including compressive strength, tensile strength,
young’s modulus, and bond strength were evaluated. Thermogravimetric analysis (TGA)
was used to elucidate the effect of subfreezing curing on the grout’s hydration products.
Mercury intrusion porosimetry (MIP) tests were also performed to examine how the
subfreezing curing affected the porosity and pore size distribution of the grout.
48
compressive strength result is the average value obtained on three identical specimens.
The compressive strength gain of the grout was monitored at ambient (T = 23 ± 1oC), and
at temperatures of 1°C, -10°C, and -20°C. The initial ambient curing time before
exposure to subfreezing conditions was varied at 0, 8, and 12 hours. The grout was tested
at ages of 1, 3, 5, and 7 days, for all curing regimes, and 28 days for curing at -10°C and
ambient conditions.
The tensile strength of the grout was determined according to ASTM C496 (Standard
Test Method for Splitting Tensile Strength of Cylindrical Concrete Specimens) using the
average value of three identical 75 x 150 mm (3 x 6 in.) cylinders at each curing regime.
The tensile strength gain of the grout was assessed at 1, 3, 7 days and at 28 days for
curing at -10°C and ambient temperature.
The modulus of elasticity and Poisson’s ratio of the grout were determined according to
ASTM C469 (Test Method for Static Modulus of Elasticity and Poisson's Ratio of
Concrete in Compression) using the average of two 100 x 200 mm (4 x 8 in.) identical
cylinders. The modulus of elasticity was measured at 7 and 28 days for specimens cured
at -10°C and ambient temperature.
Moreover, mercury intrusion porosimetry (MIP) tests were performed to determine the
effect of early-age subfreezing curing on the porosity and pore size distribution of the
49
grout. Fragments were taken from the internal sections of crushed grout cylinders used
for testing the compressive strength of the grout. Care was taken to ensure that these
fragments were not fractured, and multiple tests were run to confirm the repeatability of
the results. The specimens were immediately submersed in isopropanol to stop hydration
reactions, and subsequently dried in a desiccator until a constant mass was reached. The
tests were run using a Micrometrics AutoPore IV 9500 Series porosimeter having a range
of pressures from 0 to 414 MPa (60,000 psi).
Furthermore, pullout test specimens were designed to represent typical grouted dowel
connections used in full-scale precast concrete wall panel construction. Each pullout test
specimen used an unreinforced precast concrete block with a thin-walled corrugated steel
duct concentrically cast into the block. The concrete blocks were cast using self-
consolidating concrete with an average compressive strength of 50.6 MPa (7340 psi), and
had a cross-sectional area of 203 x 203 mm (8 x 8 in.) and a height of 406 mm (16 in.).
The corrugated steel ducts had an internal diameter of 76 mm (3 in.) and a thickness of
0.36 mm (0.014 in.). The mechanical properties of the ducts could not be accurately
tested since disassembling it would result in plastic deformations, therefore altering its
engineering properties.
Deformed 25M steel bars were placed concentrically within the duct and grouted from
the passive (top) end of the connection. The bars extended 25 mm (1 in.) above the
concrete block to measure slip during testing. The top and bottom of the test bars were
wrapped with plastic to de-bond sections of the bar so that the embedment length, ld,
would lie in the middle of the specimen. This de-bonding was done to avoid compression
stresses induced from the pullout testing, which is known to affect the bond behavior of
the bar (ACI Committee 408, 2003). The test reinforcing bars had a specified yield
strength of 400 MPa.
The pullout test specimen and setup are illustrated in Fig. 4.1. After 7 days of curing, the
specimens were placed on a 216 x 216 mm (8.5 x 8.5 in.) steel bearing plate atop the
active pulling end of an open loop Tinius Olsen testing machine with a maximum
capacity of 530 kN (119 kips). A 216 x 216 mm (8.5 x 8.5 in.) steel bearing plate with a
50
square internal void was used to distribute the load to a 25 mm wide outer edge of the
concrete block to further help avoid compression stresses induced from the pullout test,
as previously mentioned. One strain based linear variable displacement transducer
(LVDT) with a 25 mm gauge length was placed on the unloaded end of the bar to
measure slip. A second LVDT was used to measure the testing machine crosshead
movement, which represents the bar elongation. The specimen was loaded monotonically
in tension at a rate of 0.5 kN/s. The test was ended once either the test bar fractured, or
the LVDT measuring slip reached its gauge length.
51
control specimens cured at ambient temperature. There appeared to be no significant
difference between the results of grout specimens cured at -10°C or -20°C, with both
curing conditions yielding about 84% of that of the control specimens. When the grout
specimens were exposed to ambient temperatures subsequent to sub-freezing curing for
an additional 28 days, it began to gain strength once more, achieving 97% of that of the
control. The compressive strength development for all tested temperature curing regimes
was fitted using regression analysis as plotted in Fig. 4.2. It was found that the strength
development of the grout at all temperatures followed a logarithmic function; fg’= a + b ln
(t); where, fg’ = compressive strength of the grout in MPa; t = time in days, and a, b =
regression constants.
The internal temperature of the grout specimens cured at subfreezing conditions are
presented in Fig. 4.3. The figure was truncated at 3 days since no change in temperature
was observed up until testing at day 7. All temperature recordings were taken on 75 x 150
mm (3 x 6 in.) cylinders unless otherwise stated. Due to the small specimen size, the
internal temperature of the grout reached the external subfreezing temperature very
rapidly; this has also been reported in previous studies (Aitcin, Cheung, & Shah, 1985).
Some researchers argue that the heat of hydration causes the internal temperature of the
concrete to remain above freezing to allow hydration to proceed (Nassif & Petrou, 2013).
However, as observed in Fig. 4.3, such an effect was not observed herein. The relatively
larger pullout specimen took 2.5 hours longer to reach freezing temperature compared to
the smaller cylindrical specimens due to an insulation effect of the concrete surrounding
52
the grout. Yet, this time difference is unlikely to lead to significant additional strength
gain.
45
40
35
30 -10°C Pullout Specimen
25 -10°C
20 -10°C (12 h)
Temperature ( C)
15 -10°C (8 h)
10
-20°C
5
0
-5
-10
-15
-20
-25
0 0.5 1 1.5 2 2.5 3
Age (Days)
Grout specimens that were allowed to cure at ambient conditions for 8 hours before
exposure to subfreezing temperatures achieved 7-day compressive strength of 3.5 MPa.
53
The specimens were tested at 8 hours before exposure to subfreezing temperature inside
the environmental chamber and achieved zero compressive strength. Examining the
internal temperatures from Fig. 4.3, it can be observed that the specimen initially cured
for 8 hours followed a similar trend to that of specimens initially cured for 1 day at
ambient conditions. At 8 hours when it was placed inside the environmental chamber, its
internal temperature was about to begin increasing due to the exothermic cement
hydration reactions. The initial and final setting times of the grout are 4 hours and 6.5
hours, respectively. The lack of internal temperature increase, along with the zero
strength achieved at 8 hours, were indications that the specimen just achieved final set,
and that cement hydration reactions which produce mechanical strength were just
beginning.
When the grout was allowed to cure for 12 hours at ambient temperature before exposure
to subfreezing temperatures, it achieved a 7-day compressive strength of 11.4 MPa. At 12
hours, the grout had achieved a compressive strength of 4.9 MPa before being placed
inside the environmental chamber at -10°C. Specimens initially cured for 12 hours at
ambient temperature performed better than those initially cured for only 8 hours because
they experienced significant cement hydration reactions before subfreezing exposure.
This can be observed in Fig. 4.3, where the specimens initially cured for 12 hours
experienced a steady increase in internal temperature, approaching the peak temperature
of the specimens cured for 1 day, before being placed inside the environmental chamber.
Again, this increase in internal temperature is attributed to the progress of exothermic
cement hydration reactions.
The ratio of compressive strength at varying initial ambient curing periods to the 7-day
strength of identical specimens initially cured for 1 day is displayed in Fig. 4.4. It can be
observed that the compressive strength continued to increase up to 7 days for all curing
conditions, with the specimens initially cured for 12 hours greatly outperforming those
cured for only 8 hours. Yet, both types of specimens incurred large reduction in
compressive strength of 65% and 89%, for 12 hours and 8 hours initial curing,
respectively, when compared to specimens initially cured at ambient temperature for 1
day. This emphasizes the importance of the initial curing period before early-age
54
exposure of the grout to subfreezing temperatures, in agreement with previous research
(Aitcin et al., 1985; Gardner & Poon, 1976).
0.4
8h 12h
0.35
0.3
0.25
f g '/f g '24 h 0.2
12 hours
0.15
0.1
0.05
0
1 3 5 7
Age (Days)
Figure 4.4: Ratio of compressive strength for varying initial curing periods to the 7-day
strength of specimens initially cured for 1 day.
The ratio of the 28-day to the 1-day tensile and compressive strengths for specimens
cured at ambient and -10°C curing is displayed in Fig. 4.5. It can be observed that the
subfreezing curing was more detrimental to the tensile strength than to the compressive
strength. Specimens cured at ambient temperatures experienced a gain of 105% in
compressive strength, and 107% in tensile strength, between 1 and 28 days, respectively.
Yet, specimens cured at -10°C experienced a gain of only 59% in compressive strength
and 22% in tensile strength between 1 and 28 days, respectively. This contradicts the
results of previous studies that concluded that exposure of concrete to subfreezing
resulted in lower percentage of compressive strength gain than that of tensile strength
gain (Marzouk & Hussein, 1995).
55
Table 4.2: Grout tensile strength, Young’s modulus, and Poisson’s ratio
1 3.02 - -
Ambient 3 3.29 - -
(23°C) 7 4.49 20712 0.2285
28 6.25 22713 0.2346
1 3.01 - -
3 3.18 - -
-10°C
7 3.22 19971 0.2315
28 3.68 20563 0.2322
2.5
Tensile Strength Compressive Strength
Strength Ratio of 28 Day/1 Day
1.5
0.5
0
Ambient -10°C
Figure 4.5: Ratio of 28-day to the 1-day tensile and compressive strengths for varying
curing temperatures.
56
increased by 5% from 7 to 28 days, despite subfreezing exposure, yet at a much slower
rate compared to that at ambient curing.
57
Dehydration of calcium hydroxide (mass loss = 24.34%): Ca(OH)2 → CaO + H2O (4.1)
De-carbonation of calcium carbonate (mass loss = 44.0%): CaCO3 → CaO + CO2 (4.2)
2.00E-02
0.00E+00
-4.00E-02
-6.00E-02
1 Day
-8.00E-02 3 Day -10°C
7 Day -10°C
-1.00E-01
7 Day Ambient
-1.20E-01
0 200 400 600 800 1000
Temperature ( C)
Figure 4.6: DTG curves for grout specimens at different ages and curing conditions.
102
1 Day
100 3 Day -10°C
7 Day -10°C
Relative Mass (%)
98 7 Day Ambient
96
94
92
90
0 200 400 600 800 1000
Temperature ( C)
Figure 4.7: TGA curves for grout specimens at different ages and curing conditions.
From the TGA results in Table 4.3, it can be observed that there was notable increase in
portlandite between 1 and 7 days for grout specimens cured 1 day at ambient temperature
and then at -10°C up to 7 days. This is apparently due to the progress of hydration
reactions under subfreezing conditions, since the amount of CSH/ettringite also continued
to grow, accompanied by increased mechanical properties. However, the control
58
specimen cured for 7 days at ambient temperature had a much smaller increase in
portlandite, and a much larger increase in CSH/ettringite. Hence, it appears that this
relatively large increase in portlandite under subfreezing conditions is indicative of the
pozzolanic reactions of fly ash incorporated in the grout being significantly hampered
compared to the main cement hydration process under subfreezing temperatures, as
previously hypothesized by other researchers (Marzouk & Hussein, 1995). This is also
supported by previous researchers who have shown that the portlandite content decreased
with hydration time for concretes incorporating pozzolans (Payá, Monzó, Borrachero, &
Velázquez, 2003; Pinto et al., 2007; Soriano et al., 2013; Trník et al., 2016; Vedalakshmi
et al., 2003).
Table 4.3: Relative mass change (in %) and calculated content (in %) of the
corresponding hydration products
59
Table 4.4: Average pore size and total intrusion volume from MIP test
Temperature Age (Days) Average Pore Diameter (μm) Total Intrusion Volume (mL/g)
0.012
Incremental Pore Volume (mL/g)
1 Day
0.01 3 Day -10°C
7 Day -10°C
0.008 7 Day Ambient
0.006
0.004
0.002
0
100 10 1 0.1 0.01 0.001
Pore Size Diameter (μm)
Figure 4.8: Incremental pore intrusion volume versus pore diameter for grout specimens
at different ages and curing conditions.
Comparing the incremental pore intrusion volume versus pore diameter for grout
specimens cured at ambient conditions for 1 and 7 days (Fig. 4.8), it can be observed that
the specimens cured for 7-days at ambient temperature had a decrease in the larger pore
sizes (>1μm), as well as a decrease in the smaller pore sizes (<0.07 μm) compared to the
specimens cured similarly for only one day. The majority of pores ranged from 1 to 0.02
μm in diameter. The peak was also much smaller for the 7-day specimens, resulting in
lower total intrusion volume as expected. For both specimens cured for 3 and 7 days at
-10°C, there was a minimal pore refinement effect taking place. Aside from a decrease in
pore sizes larger than 10 μm, the pore size distribution has changed negligibly. This
further supports the argument that pozzolanic reactions are significantly slowed or halted
at subfreezing temperatures.
60
4.4.6 Bond Behaviour of Grouted Dowel
Test results of 10 pullout specimens are summarized in Table 4.5. The specimen notation
is as follows (lT-n), where l indicates the embedment length (l = 6, 12, or 16 bar
diameters), T represents the curing temperature (T = A for ambient, or 10 for -10°C), and
n represents the specimen number. Specimens cured at -10°C, were initially maintained at
ambient for one day. The yielding displacement, δy (mm), is taken as the point on the bar
stress-displacement plot where the specimen’s stiffness degraded suddenly after the
elastic response. The ultimate displacement, δu (mm), and maximum slip, smax (mm),
correspond to the point where the maximum load, P (kN), occurs. The maximum bond
stress, ub (MPa), is calculated assuming an average stress distribution along the bar (Eq.
4.3).
P 10 3
ub (4.3)
.d b .l d
The bar stress-displacement behavior of each specimen type is displayed in Fig. 4.9. The
recorded displacement is the relative movement of the testing machine’s crossheads; this
therefore accounts for both slip and bar elongation. For all specimens, the behavior
mimicked that of a regular steel bar until the peak load was reached. For specimens
experiencing a pullout failure, a steep drop in load then occurred until a small plateau was
reached due to the residual bond stress, then the test was stopped.
61
600 16-10-1
12-A-2
500 12-10-1
6-A-1
300
200
100
0
0 20 40 60 80 100 120 140 160
Displacement (mm)
Figure 4.9: Bar stress-displacement plots for each embedment length and curing regime.
As previously discussed in Chapter 3, mechanical splices, which are also commonly used
as precast concrete wall panel connections, are required by current design codes to
develop 125% of the bars yield strength to account for material over-strength and ensure
that the bar yields (ACI Committee 318, 2014). Therefore, the same strength ratio, Rs,
(Eq. 3.2) was used herein.
It has been established that precast concrete connections in low to moderate seismic
regions require a ductility ratio of 4.0 (Soudki, Rizkalla, & Leblanc, 1995). This ratio was
therefore used herein since there are minimal code requirements for proving the adequacy
of these connections. The ductility ratio, μΔ, is the ratio between the ultimate
displacement and the yielding displacement (Eq. 4.4).
δu
μΔ (4.4)
δy
62
4.4.6.1 Failure Modes
There were only two failure modes exhibited by all pullout specimens, pullout failure
consisting of bar-grout bond failure (Fig. 4.10a), and bar fracture (Fig. 4.10b). In
practical applications, bar fracture is the preferred failure mode since it allows the bar to
fully develop, therefore offering higher tensile capacity for the connection. It also
provides desired superior ductility through the bar post-yielding elongation. No cracking
of the concrete block was observed for all tested specimens. Also, since the bars were
wrapped, no conical failure of the grout at the active end was experienced, as observed in
Chapter 3, and reported by previous researchers investigating similar connections (Steuck
et al., 2009).
Figure 4.10: Failure modes of pullout specimens: (a) bar pullout; (b) bar fracture.
63
At first, the bond strength appeared to be noticeably affected by the -10°C curing
temperature for specimens with embedment lengths of 6 bar diameters. However,
previous research has reported that the average bond stress is directly proportional to the
square root of the concrete compressive strength (Untrauer & Henry, 1965), even under
low temperature curing (Gardner & Poon, 1976). Therefore, to accurately analyze the
effect of subfreezing exposure on bond strength, the peak bond strength, ub, was
normalized with the square root of the grout compressive strength, √fg’. The normalized
bond strength was plotted versus the corresponding normalized length for all specimens
as displayed in Fig. 4.11.
4
Ambient
3.5 -10C
u b /√f g' = 0.55 +16.7*d b /ld
3
2.5
ub/√f g '
1.5
0.5
0
0 5 10 15 20
l d/db
It can be observed in Fig. 4.11 that there is very little difference in the normalized bond
strength between specimens cured at ambient temperature or at -10°C. Indeed, the
reduction in average bond strength for specimens having an embedment length of 6 bar
diameters cured at ambient temperature versus -10°C was 8.3%, and the reduction in the
square root of compressive strength was 8.1%. This indicates that the bond strength
remained directly proportional to the square root of the compressive strength of the grout
as previously established, regardless of the early age exposure of specimens to
subfreezing. The peak bond strength, ub , normalized with the square root of the grout
compressive strength, √fg’ , for the various curing conditions was fit using Eq. 4.5 below,
with a coefficient of correlation of 0.987 as shown in Fig. 4.11.
64
ub
0.55 16.7 * db ld (4.5)
f g'
Where,
ub
= normalized bond strength (√MPa);
f g'
This fit followed a similar behaviour to the relationship proposed by Orangun et al.
(1977), which provides the basis for the current ACI 318-14 equation (25.4.2.3).
However, this trend does not adequately represent the behavior of specimens in the
elastic range when compared to results from Chapter 3 and previous research (Steuck et
al., 2009), and is therefore only applicable for specimens that have undergone extensive
yielding. For practical purposes, it is desirable to know how the bar stress and required
development length are affected by the change in compressive strength due to exposure
to subfreezing temperature. Since the change in bond strength is directly proportional to
the change in the square root of the grout’s compressive strength, and knowing that
through equilibrium,
f s db
ub (4.6)
4l d
It can be concluded that the bar stress will also be affected proportionally, and the
required development length is inversely proportional to the change in the square root of
the grout’s compressive strength. Therefore, the development length will require an
increase by the same magnitude that the square root of grout compressive strength is
decreased. For example, a decrease in compressive strength from 40 MPa to 30 MPa due
to early-age subfreezing exposure would require a development length 1.13 times longer.
65
4.4.6.3 Bond Stress-Slip Response
All pullout test specimens with an embedment length of 6 bar diameters had a similar
bond stress-slip response. Specimens 12-10-1 and 12-10-2, which have embedment
length of 12 bar diameter, behaved differently due to the much lower peak bond stress
reached. However, they have similar characteristic regions to that of the pullout
specimens with embedment length of 6 bar diameters, but smaller in scale due to their
lower peak bond stresses. The bond stress-slip response of specimens 6-A-2, 6-10-1, and
12-10-2 are shown in Fig. 4.12, along with previously established models for the grouted
dowel connection.
s/db
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
6 70
6-A-2
5 6-10-1 60
12-10-2
ueff/√f g ' (√MPa)
Proposed Eq 3.4 50
Figure 4.12: Bond stress-slip response for each embedment length and curing regime
compared to established models.
The bond stress-slip response of all pullout specimens is characterized by four regions: i)
a curvilinear ascending branch from the beginning of the test until the peak bond stress
was reached; ii) a region of approximately constant stress at the peak bond stress; iii) an
approximately linear region of sudden decrease until the residual bond stress; and iv) a
linear region with a more gradual decrease in bond stress until the test was stopped.
Specimens with a longer embedment length (12 bar diameters) had a much stiffer
ascending branch than that of specimens with an embedment length of 6 bar diameters,
experiencing minimal slip until the peak bond stress was reached. This may be due to the
extensive yielding undergone by specimens 12-10-1 and 12-10-2, leading to significant
66
reduction in the bar-cross section, which disengaged the bar from the grout, causing
sudden failure and resulting in little slip beforehand.
It can be observed in Fig. 4.12 that the specimens having an embedment length of 6 bar
diameters followed similar behavior to the model previously developed in Chapter 3,
except with larger peak bond stresses. Such peak bond stresses are comparable to that
predicted by the Steuck et al. (2009) model, and are likely due to differences in test setup.
It can be observed that the only difference in bond-slip behavior between specimens
cured at ambient and -10°C is the third region where the bond stress suddenly decreased
to the residual bond stress. However, it is difficult to determine whether this is substantial
since such small differences have been reportedly due to the variable nature of concrete
(Eligehausen et al., 1983).
2. The initial curing time at ambient temperature (23 ± 1oC) was crucial for the
grout’s compressive strength development. Reducing this initial curing time of 24
hours by half to 12 hours reduced the compressive strength by 65%.
3. Curing the grout at -10°C was more detrimental to the tensile strength gain than to
the compressive strength gain. When cured at -10°C, the increase in strength gain
between 1 and 28 days was 59% for compressive strength, yet only 22% for
tensile strength.
67
4. The development of the modulus of elasticity was greatly slowed between 7 and
28 days subsequent to early-age exposure of the grout to subfreezing conditions.
6. Curing the grout at -10°C resulted in greater total mercury intrusion volume
compared to that for specimens cured at ambient temperatures, indicating higher
total porosity.
7. The bond strength of the grouted dowel connection remained proportional to the
square root of the grout compressive strength even after early-age subfreezing
exposure. Furthermore, curing the grout at -10°C resulted in an increase in the
dowel embedment length from 12 to 16 bar diameters to achieve bar fracture.
4.6 References
ACI Committee 306. (2010). Guide to Cold Weather Concreting (ACI 306R-10).
American Concrete Institute. Farmington Hills, MI, USA.
ACI Committee 318. (2014). Building Code Requirements for Structural Concrete (ACI
318-14) and Commentary (ACI 318R-14). American Concrete Institute.
Farmington Hills, MI.
ACI Committee 408. (2003). Bond and Development of Straight Reinforcing Bars in
Tension (ACI 408-03). American Concrete Institute. Farmington Hills, MI, USA.
68
Aitcin, P.-C., Cheung, M. S., & Shah, V. K. (1985). Strength Development of Concrete
Cured Under Arctic Sea Conditions. Temperature Effects on Concrete. ASTM
International.
ASTM C39. (2016). Standard Test Method for Compressive Strength of Cylindrical
Concrete Specimens. American Society for Testing and Materials. West
Conshohocken, PA. doi:10.1520/C0039
ASTM C469. (2014). Standard Test Method for Static Modulus of Elasticity and
Poisson’s Ratio of Concrete in Compression. American Society for Testing and
Materials. West Conshohocken, PA. doi:10.1520/C0469
ASTM C496. (2011). Standard Test Method for Splitting Tensile Strength of Cylindrical
Concrete Specimens. American Society for Testing and Materials. West
Conshohocken, PA. doi:10.1520/C0496
Bhatty, J. (1991). A review of the application of thermal analysis to cement-admixture
systems. Thermochimica Acta, 189, 313–350. doi:10.1016/0040-6031(91)87128-J
Cole, W. F., & Kroone, B. (1960). Carbon Dioxide in Hydrated Portland Cement.
Journal of the American Concrete Institute, 56(6), 1275–1296.
Dweck, J., Buchler, P. M., Coelho, A. C. V., & Cartledge, F. K. (2000). Hydration of a
Portland cement blended with calcium carbonate. Thermochimica Acta, 346(1-2),
105–113. doi:10.1016/S0040-6031(99)00369-X
Eligehausen, R., Popov, E. P., & Bertero, V. V. (1983). Local bond stress-slip
relationships of deformed bars under generalized excitations. Report No.
UCB/EERC-83/23. Earthquake Engineering Research Center, Universisty of
California-Berkeley.
Gardner, N. J. (1990). Effect of Temperature on the Early-Age Properties of Type I, Type
III, and Type I/ Fly Ash Concretes. ACI Materials Journal, 87(1), 68–78.
Gardner, N. J., & Poon, S. M. (1976). Time and Temperature Effects on Tensile, Bond,
and Compressive Strengths. ACI Journal Proceedings, 73(7), 405–409.
Gardner, N. J., Sau, P. L., & Cheung, M. S. (1988). Strength development and durability
of concretes cast and cured at 0 C. ACI Materials Journal, 85(6), 529–536.
doi:10.14359/2289
Husem, M., & Gozutok, S. (2005). The effects of low temperature curing on the
compressive strength of ordinary and high performance concrete. Construction
and Building Materials, 19(1), 49–53. doi:10.1016/j.conbuildmat.2004.04.033
Klieger, P. (1958). Effect of mixing and curing temperature on concrete strength. ACI
Journal, 54(6), 1063–1081.
Marzouk, H., & Houssein, A. (1990). Properties of high-strength concrete at low
temperatures. ACI Materials Journal, 87(2), 167–171.
Marzouk, H., & Hussein, A. (1995). Effect of Curing Age on High-Strength Concrete at
Low Temperatures. Journal of Materials in Civil Engineering, 7(3), 161–167.
doi:10.1061/(ASCE)0899-1561(1995)7:3(161)
69
Nassif, A. Y., & Petrou, M. F. (2013). Influence of cold weather during casting and
curing on the stiffness and strength of concrete. Construction and Building
Materials, 44, 161–167. doi:http://dx.doi.org/10.1016/j.conbuildmat.2013.03.016
Orangun, C.O., Jirsa, J. O., & Breen, J. E. (1977). A reevaluation of test data on
development length splices. Journal of the American Concrete Institute, 74(3),
114-122. doi:10.14359/10993
Payá, J., Monzó, J., Borrachero, M., & Velázquez, S. (2003). Evaluation of the
pozzolanic activity of fluid catalytic cracking catalyst residue (FC3R).
Thermogravimetric analysis studies on FC3R-Portland cement pastes. Cement
and Concrete Research, 33(4), 603–609. doi:10.1016/S0008-8846(03)00014-0
Pinto, C. A., Büchler, P. M., & Dweck, J. (2007). Pozzolanic properties of a residual FCC
catalyst during the early stages of cement hydration: Evaluation by thermal
analysis. Journal of Thermal Analysis and Calorimetry, 87(3), 715–720.
doi:10.1007/s10973-006-7772-2
Raynor, D. J., Dawn, E. L., & Stanton, J. F. (2002). Bond-Slip Response of Reinforcing
Bars Grouted in Ducts. ACI Structural Journal, 99(5), 568–576.
Šauman, Z. (1971). Carbonization of porous concrete and its main binding components.
Cement and Concrete Research, 1(6), 645–662.
Sha, W., O’Neill, E. a., & Guo, Z. (1999). Differential scanning calorimetry study of
ordinary Portland cement. Cement and Concrete Research, 29(9), 1487–1489.
doi:10.1016/S0008-8846(99)00128-3
Shih, T. S., Lee, G. C., & Chang, K. C. (1988). Effect of Freezing Cycles on Bond
Strength of Concrete. Journal of Structural Engineering, 114(3), 717–726.
doi:10.1061/(ASCE)0733-9445(1988)114:3(717)
Soriano, L., Monzó, J., Bonilla, M., Tashima, M. M., Payá, J., & Borrachero, M. V.
(2013). Effect of pozzolans on the hydration process of Portland cement cured at
low temperatures. Cement and Concrete Composites, 42, 41–48.
doi:10.1016/j.cemconcomp.2013.05.007
Soudki, K. A., Rizkalla, S. H., & Leblanc, B. (1995). Horizontal Connections for Precast
Concrete Shear Walls Subjected to Cyclic Deformations Part 1: Mild Steel
Connections. PCI Journal, 40(4), 78–96.
Steuck, K. P., Eberhard, M. O., & Stanton, J. F. (2009). Anchorage of large-diameter
reinforcing bars in ducts. ACI Structural Journal, 106(4), 506–513.
Trník, A., Scheinherrová, L., Kulovaná, T., & Černý, R. (2016). Simultaneous
Differential Scanning Calorimetry and Thermogravimetric Analysis of Portland
Cement as a Function of Age. International Journal of Thermophysics, 37(1), 12.
doi:10.1007/s10765-015-2028-7
Untrauer, R. E., & Henry, R. L. (1965). Influence of Normal Pressure on Bond Strength.
ACI Journal, 62(5), 577–586.
Vedalakshmi, R., Raj, A. S., Srinivasan, S., & Babu, K. G. (2003). Quantification of
hydrated cement products of blended cements in low and medium strength
70
concrete using TG and DTA technique. Thermochimica Acta, 407(1-2), 49–60.
doi:10.1016/S0040-6031(03)00286-7
71
5 SUMMARY AND CONCLUSIONS
5.1 Summary
Precast concrete load-bearing wall panels have become a popular choice for low-,
medium-, and high-rise construction in North America. An integral part of this structural
system is the horizontal connections between wall panels, since it directly affects the
strength and stability of the structure. A very common connection method is the grouted
dowel connection, where a reinforcing bar protruding from the lower wall panel is
grouted into a corrugated steel duct cast into the upper wall panel. Despite the common
use of this connection in practice, there are no pertinent specific code requirements that
guide this use, and related research is sparse. Furthermore, this construction proceeds
throughout cold weather conditions, with the connection area typically heated for one day
then exposed to subfreezing temperatures, before the grout is fully cured. The effects of
exposure to early-age subfreezing temperatures on the bond behaviour of this connection
are still not well understood. Thus, the focus of this research was to fill this knowledge
gap by exploring the bond behaviour of this connection for use in precast wall panel
construction, and the effects of exposure to early-age subfreezing temperatures on the
connection’s bond strength.
Chapter 2 presented a literature review of the mechanisms of bond, as well as the few
relevant studies that have examined this connection. It was shown that this connection is
currently designed as a regular reinforcing bar in concrete (ACI Committee 318, 2014),
which does not account for the additional confinement provided by the presence of the
corrugated steel duct. It was also demonstrated that when greater confinement is present,
it can lead to greatly overdesigned embedment lengths (Einea et al., 1999). The only
available study on the bond strength of this connection examined its use for precast
bridge bent caps using larger diameter bars (32M to 57M), with larger duct diameter/bar
diameter ratios of 3.6 (Steuck et al., 2009), whereas a typical wall panel connection
generally uses a 25M bar with a duct diameter/bar diameter ratio of 3.0. This
demonstrates the lack of information available on this connection, specifically for its use
in precast wall panels. Furthermore, there is a lack of research in the open literature
examining the effects of subfreezing curing on bond strength. Yet, one study
72
investigating the effect of 2°C curing on concrete concluded that the bond strength
remained proportional to the square root of compressive strength, irrespective of the
temperature or cement type (Gardner & Poon, 1976). It was concluded from this
literature review that the corrugated steel duct present in this connection generates greater
confinement than what is accounted for in the current design codes, allowing for a
reduction in development length compared to a regular bar in traditional reinforced
concrete. Yet, the limited research available requires further investigation into this
connection for its use in precast wall panels, along with thorough examination of how the
exposure to early-age subfreezing temperatures affects the bond strength of the
connection.
73
exposure to subfreezing temperature on the bond strength of the connection were
quantified, and recommendations for cold weather construction were provided.
5.2 Conclusions
The major conclusions of this research are as follows:
1. The steel duct creates a higher confinement effect than present in regular
reinforced concrete, allowing for a shear pullout failure rather than a tensile
splitting failure. This failure mode occurs at significantly higher bond stresses,
allowing for a reduction in required embedment length. The absence of the
corrugated steel duct results in a tensile splitting failure, and can reduce the
strength of this connection by about 30%.
2. Eccentric bar placement generates tensile stress concentrations in the surrounding
concrete, which can result in tensile splitting of the concrete not observed in
specimens with concentric bar placement. This results in about 10% strength
reduction for embedment lengths of 12 bar diameters, and 4% for embedment
lengths of 36 bar diameters.
3. The current ACI 318-14 equation (25.4.2.3) does not account for the greater
confinement effect of the duct. It therefore greatly overestimates the required
development length, resulting in values 3.08 times larger than required.
4. The equation proposed by Steuck et al. (2009) is not sufficiently conservative,
which is a disadvantage for designing the grouted dowel connection since safety
is necessary considering the high variability of concrete materials. The equation
proposed in the present study is 10% on the conservative side, which is a
desirable feature in design applications.
5. The equation proposed in the present study is applicable for non-shrink grouts
having a compressive strength up to 70 MPa. Using other grout types has not been
investigated herein, and thus needs to be validated for specific use in this
particular connection. The use of higher strength grouts (>70 MPa) has also not
been explored, and therefore the proposed design equation should be limited to 70
MPa until such data on using higher strength grouts becomes available.
74
6. Lower curing temperature resulted in lower compressive strength of the grout, as
expected. However, there was no significant difference in compressive strength
for grouts cured at -10°C and -20°C. The initial curing time at ambient
temperature (23 ± 1°C) was crucial for the grout’s compressive strength
development. Reducing this initial curing time by half reduced the compressive
strength by 65%.
7. Curing the grout at -10°C was more detrimental to the tensile strength gain that to
the compressive strength gain. When cured at -10°C, the increase in strength gain
between 1 and 28 days was 59% for compressive strength, yet only 22% for
tensile strength.
8. The development of the modulus of elasticity was greatly slowed between 7 and
28 days subsequent to the early-age exposure of the grout to subfreezing
conditions.
9. Pozzolanic reactions appeared to be greatly slowed/halted compared to the basic
cement hydration reactions when the grout was exposed to early-age subfreezing
temperatures.
10. Curing the grout at -10°C resulted in greater total mercury intrusion volume
compared to that for specimens cured at ambient temperatures, indicating higher
total porosity.
11. The bond strength of the grouted dowel connection remained proportional to the
square root of the grout compressive strength even after early-age subfreezing
exposure. Furthermore, curing the grout at -10°C resulted in an increase in the
embedment length from 12 to 16 bar diameters to achieve bar fracture.
12. The bond stress-slip response of the grouted dowel connection remained
unchanged for specimens exposed to early-age subfreezing conditions.
13. The mechanical properties and their associated strength development under
subfreezing conditions found herein are only applicable to the specific grout
tested. As discussed in Chapter Two, different cementitious proportions will result
in different final strengths, and rate of strength gain under subfreezing conditions.
Therefore, it is important to accurately test the desired grout for use in subfreezing
conditions.
75
5.3 Recommendations for Future Work
Recommendations for future work are as follows:
5.4 References
ACI Committee 318. (2014). Building Code Requirements for Structural Concrete (ACI
318-14) and Commentary (ACI 318R-14). American Concrete Institute.
Farmington Hills, MI.
Einea, A., Yehia, S., & Tadros, M. K. (1999). Lap splices in confined concrete. ACI
Structural Journal, 96(6), 947–955.
Gardner, N. J., & Poon, S. M. (1976). Time and Temperature Effects on Tensile Bond,
and Compressive Strengths. ACI Journal Proceedings, 73(7), 405–409.
76
Steuck, K. P., Eberhard, M. O., & Stanton, J. F. (2009). Anchorage of large-diameter
reinforcing bars in ducts. ACI Structural Journal, 106(4), 506–513.
77
APPENDICES
Table A.1: Compressive strength data of grout cured at ambient (23 ± 1°C)
Age Compressive Strength Standard
Sample Average
(Days) (MPa) Deviation
1 18.14
1 2 20.29 19.18 1.08
3 19.1
1 34.14
3 2 33.16 33.75 0.52
3 33.96
1 36.58
5 2 36.09 36.64 0.59
3 37.26
1 39.02
7 2 37.75 38.43 0.64
3 38.52
78
Table A.3: Compressive strength data of grout cured at -10°C
Age Compressive Strength Standard
Sample Average
(Days) (MPa) Deviation
1 20.97
1 2 19.22 20.55 1.18
3 21.46
1 27.99
3 2 25.84 26.72 1.13
3 26.34
1 33.55
5 2 31.21 32.32 1.18
3 32.19
1 31.7
7 2 33.16 32.51 0.74
3 32.68
Table A.4: Compressive strength data of grout cured at -10°C (12 hours initial curing)
Age Compressive Strength Standard
Sample Average
(Days) (MPa) Deviation
1 4.73
1/2
2 4.92 4.88 0.13
(12 Hours)
3 4.99
1 9.62
3 2 10.14 9.85 0.27
3 9.79
1 10.77
5 2 11.22 11.02 0.23
3 11.07
1 11.68
7 2 11.07 11.41 0.31
3 11.49
79
Table A.5: Compressive strength data of grout cured at -10°C (8 hours initial curing)
Age Compressive Strength Standard
Sample Average
(Days) (MPa) Deviation
1 0
1/3
2 0 0.00 0.00
(8 Hours)
3 0
1 2.07
3 2 1.87 1.95 0.11
3 1.91
1 2.97
5 2 3.51 3.27 0.28
3 3.34
1 3.81
7 2 3.23 3.54 0.29
3 3.57
80
Table A.7: Tensile strength data of grout cured at ambient (23 ± 1°C)
Age Standard
Sample Tensile Strength (MPa) Average
(Days) Deviation
1 2.95
1 2 3.20 3.02 0.16
3 2.90
1 3.57
3 2 3.18 3.29 0.24
3 3.12
1 4.56
7 2 4.48 4.49 0.07
3 4.42
1 6.41
28 2 6.12 6.25 0.15
3 6.23
81
Table A.9: Young’s Modulus data of grout cured at ambient (23 ± 1°C)
Young's Poisson's
Load Stress Lateral Reading Vertical Reading Vertical
Age Lateral Strain Modulus, Ratio,
(lbs) (MPa) (mm) (mm) Strain
E (MPa) γ
5000 2.74 0.003 1.47638E-05 0.029 0.0001087
10000 5.49 0.008 3.93701E-05 0.066 0.0002474
15000 8.23 0.013 6.39764E-05 0.097 0.0003636
20000 10.97 0.02 9.84252E-05 0.132 0.0004948
25000 13.72 0.027 0.000132874 0.168 0.0006297
7 Day 20712 0.2285
5000 2.74 0.004 1.9685E-05 0.041 0.0001537
10000 5.49 0.009 4.42913E-05 0.076 0.0002849
15000 8.23 0.015 7.38189E-05 0.113 0.0004235
20000 10.97 0.022 0.000108268 0.149 0.0005585
25000 13.72 0.029 0.000142717 0.186 0.0006972
5000 2.74 0.005 2.46063E-05 0.027 0.0001012
10000 5.49 0.01 4.92126E-05 0.055 0.0002061
15000 8.23 0.015 7.38189E-05 0.085 0.0003186
20000 10.97 0.022 0.000108268 0.117 0.0004385
28 25000 13.72 0.028 0.000137795 0.149 0.0005585
22712.5 0.2346
Day 5000 2.74 0.004 1.9685E-05 0.032 0.0001199
10000 5.49 0.01 4.92126E-05 0.067 0.0002511
15000 8.23 0.016 7.87402E-05 0.101 0.0003786
20000 10.97 0.021 0.000103346 0.135 0.0005060
25000 13.72 0.027 0.000132874 0.168 0.0006297
82
Table A.10: Young’s Modulus data of grout cured at -10°C
Young's Poisson's
Stress Lateral Reading Vertical Reading Vertical
Age Load (lbs) Lateral Strain Modulus, Ratio,
(MPa) (mm) (mm) Strain
E (MPa) γ
5000 2.74 0.004 1.9685E-05 0.037 0.0001387
10000 5.49 0.01 4.92126E-05 0.073 0.0002736
15000 8.23 0.016 7.87402E-05 0.11 0.0004123
20000 10.97 0.023 0.000113189 0.15 0.0005622
7 Day 19971.2 0.23145
5000 2.74 0.008 3.93701E-05 0.035 0.0001312
10000 5.49 0.015 7.38189E-05 0.072 0.0002699
15000 8.23 0.021 0.000103346 0.105 0.0003936
20000 10.97 0.028 0.000137795 0.143 0.0005360
5000 2.74 0.006 2.95276E-05 0.033 0.0001237
10000 5.49 0.012 5.90551E-05 0.068 0.0002549
15000 8.23 0.018 8.85827E-05 0.102 0.0003823
28 20000 10.97 0.024 0.00011811 0.138 0.0005172
20563 0.2322
Day 5000 2.74 0.007 3.44488E-05 0.035 0.0001312
10000 5.49 0.014 6.88976E-05 0.07 0.0002624
15000 8.23 0.02 9.84252E-05 0.106 0.0003973
20000 10.97 0.027 0.000132874 0.144 0.0005397
83
Table A.11: MIP data for grout cured for 1 day at ambient (23°C)
Pore Diameter (μm) Incremental Pore Volume (mL/g) Cumulative Intrusion (mL/g)
107.8672 0 0
71.3185 0.0005 0.0005
53.727 0.0009 0.0014
38.99 0.001 0.0024
35.7431 0.0006 0.003
28.5721 0.0024 0.0054
25.2089 0.0007 0.0061
22.5513 0.0013 0.0074
20.3938 0.0016 0.009
16.4683 0.0038 0.0128
13.3582 0.0026 0.0154
10.6897 0.0026 0.018
8.5479 0.0038 0.0218
8.033 0.0002 0.022
5.1879 0.0009 0.0229
4.5679 0.0003 0.0232
3.8121 0.0011 0.0243
2.9695 0.0018 0.0261
2.4356 0.0014 0.0275
1.9115 0.0014 0.0289
1.5587 0.0009 0.0298
1.2382 0.0008 0.0306
0.9796 0.001 0.0316
0.7971 0.0008 0.0324
0.6534 0.0008 0.0332
0.5118 0.001 0.0342
0.4119 0.0008 0.035
0.3347 0.0007 0.0357
0.2675 0.0007 0.0364
0.216 0.0008 0.0372
0.178 0.001 0.0382
0.1425 0.0013 0.0395
0.1125 0.0017 0.0412
0.0909 0.0025 0.0437
0.0737 0.005 0.0487
0.0594 0.0084 0.0571
0.0478 0.01 0.0671
0.0476 0.0001 0.0672
0.0382 0.0091 0.0763
84
0.031 0.0084 0.0847
0.0249 0.0075 0.0922
0.0202 0.0052 0.0974
0.0162 0.0042 0.1016
0.0144 0.0019 0.1035
0.013 0.0014 0.1049
0.0107 0.0021 0.107
0.0085 0.0018 0.1088
0.0071 0.0004 0.1092
0.0061 0.0008 0.11
0.0053 0.0005 0.1105
0.0047 0.0004 0.1109
0.0043 0.0004 0.1113
0.0039 0.0006 0.1119
0.0036 0.001 0.1129
0.0046 0.0011 0.114
0.006 0.0008 0.1148
0.0078 0.0005 0.1153
0.0102 0.0003 0.1156
0.0133 0.0002 0.1158
0.0172 0 0.1158
0.0222 -0.0003 0.1155
0.0292 -0.0006 0.1149
0.0373 -0.0009 0.114
0.0495 -0.0018 0.1122
0.0646 -0.0025 0.1097
0.0819 -0.0031 0.1066
0.1065 -0.0044 0.1022
0.142 -0.0059 0.0963
0.1776 -0.0048 0.0915
0.2366 -0.0057 0.0858
0.304 -0.0038 0.082
0.4256 -0.0033 0.0787
0.5306 -0.0013 0.0774
0.7067 -0.0011 0.0763
0.8857 -0.0006 0.0757
1.1066 -0.0006 0.0751
1.4695 -0.0006 0.0745
1.908 -0.0005 0.074
2.4232 -0.0005 0.0735
3.1829 -0.0005 0.073
4.0175 -0.0005 0.0725
6.6126 -0.0009 0.0716
85
Table A.12: MIP data for grout cured for 7 days at ambient (23°C)
107.4817 0 0
71.3757 0.0003 0.0003
53.6886 0.0001 0.0004
38.971 0.0002 0.0006
35.718 0.0001 0.0007
28.5655 0.0003 0.001
25.1992 0.0001 0.0011
22.5461 0.0002 0.0013
20.397 0.0001 0.0014
16.465 0.0001 0.0015
13.3595 0.0003 0.0018
10.689 0.0004 0.0022
8.2248 0 0.0022
5.0469 0.0002 0.0024
4.5629 0 0.0024
3.7947 0.0004 0.0028
2.9529 0.0007 0.0035
2.4316 0.0013 0.0048
1.8934 0.0014 0.0062
1.5506 0.0013 0.0075
1.2327 0.0028 0.0103
0.9899 0.0029 0.0132
0.7956 0.0039 0.0171
0.6518 0.005 0.0221
0.5107 0.0048 0.0269
0.4126 0.0033 0.0302
0.335 0.0037 0.0339
0.2676 0.0032 0.0371
0.2159 0.002 0.0391
0.1777 0.0018 0.0409
0.1425 0.0025 0.0434
0.1125 0.0053 0.0487
0.0909 0.007 0.0557
0.0737 0.006 0.0617
0.0594 0.0053 0.067
0.0476 0.005 0.072
0.0382 0.0041 0.0761
0.031 0.0038 0.0799
0.0249 0.0037 0.0836
86
0.0202 0.0026 0.0862
0.0162 0.002 0.0882
0.0144 0.0007 0.0889
0.013 0.0005 0.0894
0.0107 0.001 0.0904
0.0085 0.0011 0.0915
0.0071 0.0011 0.0926
0.0061 0.001 0.0936
0.0053 0.0016 0.0952
0.0047 0.0012 0.0964
0.0043 0.0019 0.0983
0.0039 0.0016 0.0999
0.0036 0.0008 0.1007
0.0046 0.0018 0.1025
0.006 0.0007 0.1032
0.0078 0.0004 0.1036
0.0102 0.0002 0.1038
0.0133 0 0.1038
0.0172 -0.0002 0.1036
0.0222 -0.0004 0.1032
0.0292 -0.0006 0.1026
0.0373 -0.0007 0.1019
0.0495 -0.001 0.1009
0.0646 -0.0013 0.0996
0.082 -0.0016 0.098
0.1066 -0.0024 0.0956
0.1421 -0.0034 0.0922
0.1775 -0.0032 0.089
0.2367 -0.0042 0.0848
0.3041 -0.0035 0.0813
0.425 -0.0039 0.0774
0.5307 -0.0021 0.0753
0.7049 -0.0021 0.0732
0.8803 -0.0013 0.0719
1.1097 -0.0012 0.0707
1.4412 -0.0012 0.0695
1.9168 -0.001 0.0685
2.4626 -0.0006 0.0679
3.1403 -0.0005 0.0674
4.0598 -0.0004 0.067
6.3971 -0.0005 0.0665
87
Table A.13: MIP data for grout cured for 3 days at -10°C
Pore Diameter (μm) Incremental Pore Volume (mL/g) Cumulative Intrusion (mL/g)
107.6613 0 0
71.2716 0.0003 0.0003
53.6674 0.0002 0.0005
38.9513 0.0006 0.0011
35.711 0.0001 0.0012
28.5585 0.0002 0.0014
25.1923 0.0001 0.0015
22.5451 0.0003 0.0018
20.3963 0.0004 0.0022
16.4676 0.0006 0.0028
13.3528 0.0004 0.0032
10.6853 0.0023 0.0055
8.5446 0.0012 0.0067
8.124 0.0001 0.0068
5.016 0.0004 0.0072
4.6132 0.0001 0.0073
3.7609 0.0005 0.0078
2.9664 0.0015 0.0093
2.4542 0.0013 0.0106
1.9058 0.0013 0.0119
1.5538 0.0013 0.0132
1.2316 0.0015 0.0147
0.9816 0.0015 0.0162
0.7996 0.001 0.0172
0.6512 0.0011 0.0183
0.5116 0.0017 0.02
0.4122 0.002 0.022
0.3351 0.002 0.024
0.2672 0.0017 0.0257
0.2158 0.0012 0.0269
0.1782 0.0011 0.028
0.1424 0.0014 0.0294
0.1125 0.0017 0.0311
0.091 0.0019 0.033
0.0737 0.0037 0.0367
0.0594 0.0078 0.0445
0.0475 0.0094 0.0539
0.0382 0.0071 0.061
0.031 0.0076 0.0686
0.0249 0.0061 0.0747
88
0.0202 0.0046 0.0793
0.0162 0.0041 0.0834
0.0144 0.0019 0.0853
0.013 0.0016 0.0869
0.0107 0.0027 0.0896
0.0085 0.0025 0.0921
0.0071 0.0015 0.0936
0.0061 0.0012 0.0948
0.0053 0.0013 0.0961
0.0047 0.0005 0.0966
0.0043 0.0012 0.0978
0.0039 0.0012 0.099
0.0036 0.001 0.1
0.0046 0.0011 0.1011
0.006 0.0008 0.1019
0.0078 0.0003 0.1022
0.0102 0.0002 0.1024
0.0133 0 0.1024
0.0172 -0.0004 0.102
0.0222 -0.0008 0.1012
0.0292 -0.0014 0.0998
0.0373 -0.0017 0.0981
0.0495 -0.0027 0.0954
0.0646 -0.0032 0.0922
0.082 -0.0033 0.0889
0.1066 -0.0042 0.0847
0.1421 -0.005 0.0797
0.1778 -0.0039 0.0758
0.2367 -0.0046 0.0712
0.3045 -0.0034 0.0678
0.4254 -0.0032 0.0646
0.53 -0.0015 0.0631
0.7052 -0.0014 0.0617
0.8789 -0.0009 0.0608
1.1063 -0.0008 0.06
1.4388 -0.0009 0.0591
1.8743 -0.0008 0.0583
2.4182 -0.0006 0.0577
3.1602 -0.0006 0.0571
3.9805 -0.0005 0.0566
6.5197 -0.0008 0.0558
89
Table A.14: MIP data for grout cured for 7 days at -10°C
Pore Diameter (μm) Incremental Pore Volume (mL/g) Cumulative Intrusion (mL/g)
107.8672 0 0
71.3185 0.0003 0.0003
53.727 0.0002 0.0005
38.99 0.0003 0.0008
35.7431 0 0.0008
28.5721 0.0002 0.001
25.2089 0.0001 0.0011
22.5513 0.0001 0.0012
20.3938 0.0001 0.0013
16.4683 0.0006 0.0019
13.3582 0.0009 0.0028
10.6897 0.0012 0.004
8.5479 0.0031 0.0071
7.9605 0.0001 0.0072
5.1595 0.0016 0.0088
4.5484 0.0016 0.0104
3.7995 0.0017 0.0121
2.9623 0.0023 0.0144
2.431 0.0016 0.016
1.9089 0.002 0.018
1.5571 0.0013 0.0193
1.2373 0.001 0.0203
0.979 0.0012 0.0215
0.7967 0.0008 0.0223
0.6531 0.001 0.0233
0.5116 0.0017 0.025
0.4118 0.0012 0.0262
0.3346 0.0014 0.0276
0.2674 0.0013 0.0289
0.2159 0.0012 0.0301
0.178 0.0012 0.0313
0.1425 0.0015 0.0328
0.1125 0.0015 0.0343
0.0909 0.0017 0.036
0.0737 0.003 0.039
0.0594 0.0071 0.0461
0.0478 0.0092 0.0553
0.0476 0.0001 0.0554
0.0382 0.0084 0.0638
0.031 0.0078 0.0716
90
0.0249 0.0068 0.0784
0.0202 0.0052 0.0836
0.0162 0.0044 0.088
0.0144 0.002 0.09
0.013 0.0016 0.0916
0.0107 0.0026 0.0942
0.0085 0.0023 0.0965
0.0071 0.0014 0.0979
0.0061 0.0011 0.099
0.0053 0.001 0.1
0.0047 0.001 0.101
0.0043 0.0012 0.1022
0.0039 0.0013 0.1035
0.0036 0.0016 0.1051
0.0046 0.0012 0.1063
0.006 0.0004 0.1067
0.0078 0.0002 0.1069
0.0102 0 0.1069
0.0133 -0.0002 0.1067
0.0172 -0.0005 0.1062
0.0222 -0.0007 0.1055
0.0292 -0.0011 0.1044
0.0373 -0.0015 0.1029
0.0495 -0.0024 0.1005
0.0646 -0.003 0.0975
0.0819 -0.0034 0.0941
0.1065 -0.0046 0.0895
0.142 -0.0058 0.0837
0.1776 -0.0046 0.0791
0.2366 -0.0049 0.0742
0.304 -0.0032 0.071
0.4256 -0.0027 0.0683
0.5305 -0.0011 0.0672
0.7066 -0.001 0.0662
0.8856 -0.0007 0.0655
1.1064 -0.0006 0.0649
1.469 -0.0007 0.0642
1.9072 -0.0006 0.0636
2.4219 -0.0005 0.0631
3.1804 -0.0006 0.0625
4.0134 -0.0005 0.062
6.6016 -0.0008 0.0612
91
CURRICULUM VITAE
Post-secondary Education
Publications:
Provost, D. J., Elsayed, M., & Nehdi, M. L., “Investigation of Grouted Dowel
Connections for Precast Wall Construction.” Submitted to ACI Structural Journal.
Provost, D. J., Elsayed, M., & Nehdi, M. L., “Early-age Exposure of Grouted Precast
Wall Connections to Subfreezing Curing Temperatures.” Submitted to Construction and
Building Materials.
Provost, D. J., Elsayed, M., & Nehdi, M. L., (2016) “Investigation of Grouted Precast
Concrete Wall Connections at Subfreezing Conditions.” Proceedings of the CSCE 5th
International Materials Specialty Conference, London, ON. MAT-719.
Elsayed, M., Provost, D. J., and Nehdi, M. L. (2016) “Bond Behaviour of Grouted
Connections Under Monotonic Tensile Loads.” Proceedings of the CSCE 5th
International Materials Specialty Conference, London, ON. MAT-723.
92