Upload
hacong
View
217
Download
1
Embed Size (px)
Citation preview
0
Performance of Concrete Rehabilitation using 1
Rapid Setting CalciumSulfoAluminate Cement 2
at the Seattle-Tacoma Airport 3
4
July 28, 2013 5
6
Corresponding Author 7 Chris Ramseyer 8
University of Oklahoma 9 Associate Professor, Director Fears Structural Engineering Lab 10
202 W. Boyd, room 334 11 Norman, Oklahoma 73019 12
United States 13 Phone: 405-406-2330 14
Fax: 405-325-4217 15 Email: [email protected] 16
17 18 19
Co-Author 20
Shideh Shadravan 21
Cornell University 22
325 Hollister Hall 23
Ithica, New York 14850 24
United States 25
Phone: 405-974-8066 26
Email: [email protected] 27 28 29
number of words in text = 4377 30 31
number of figures, tables, or photographs = 12 32 33
34
TRB 2014 Annual Meeting Original paper submittal - not revised by author
1
Abstract 1
2
This paper summarizes the results from an experimental study on the behavior of concrete 3
pavement in a controlled environment. The test program characterizes the volume change or 4
dimensional properties of selected concrete materials, evaluating their performance as real 5
pavements in that they are exposed to the controlled environment on the top surface and to 6
the ground moisture on the bottom surface. The concrete mix designs examined included 7
normal-strength Portland cement concrete (PCC), high-strength concrete (HSC), concrete 8
using shrinkage reducing admixtures (SRA), and concrete using Calcium sulfoaluminate 9
cement (CSA). The data includes standard concrete material characterization test results, joint 10
movements, internal relative humidity and temperature over the pavement depth. It was found 11
that the concrete using CSA was very stable with no long term shrinkage, cracking or 12
warping, while typical PCC and HSC continued to show crack growth at over 600 days of 13
age. The concrete using SRA had a minor decrease in shrinkage and crack growth at an early 14
age, and while this decrease extended for the length of the testing no further decrease in the 15
shrinkage growth or sectional stability was noted when compared to normal PCC at the end 16
of approximately two years. 17
18
19
TRB 2014 Annual Meeting Original paper submittal - not revised by author
2
1. Introduction 1
Drying shrinkage is one of the common causes of cracking in concrete slabs-on-ground 2
including pavements. As concrete dries, the water mix in concrete begins to evaporate from 3
the top surface of concrete. Therefore, a moisture difference is generated between the top and 4
bottom surface of the slabs. This causes a dimensional or “shrinkage” gradient to develop 5
into the slabs (Carlson, 1938; Tremper and Spellman, 1963; Ytterberg, 1987). Because the 6
concrete surface is restrained by many factors such as friction between bottom surface of slab 7
and subgrade, form work or surrounding objects and moisture difference through the depth of 8
the slab, it resists the effects of shrinkage. Therefore, an additional tension applies to the 9
concrete. As the tensile stress in the concrete exceeds the tensile strength of the concrete, 10
cracking and warping occurs (ACI, 2008). 11
12
Shrinkage is the most important factor influencing warping, curling and cracking of slabs or 13
pavements (Mailvaganam et al., 2002). In this research, warping is defined as a phenomenon 14
occurring due to uneven moisture conditions in the slab, whereas curling is defined as a 15
phenomenon occurring due to the uneven temperature through the depth of a slab. Warping 16
and curling have major effects on serviceability and performance of the concrete slabs-on-17
ground, necessitating their maintenance and repair. Drying shrinkage, a type of shrinkage in 18
concrete, is one of the major causes of cracks in concrete particularly in slabs-on-ground. 19
Shrinkage does not begin at the time of loading or drying; rather, it starts immediately after 20
cement and water come in contact during the concrete mix (Ramseyer, 1999; Holt, 2001). 21
Prediction of drying shrinkage in concrete is difficult because many affecting factors are 22
involved. According to many researchers (Tremper and Spellman, 1963; Hansen, 1987; Hart, 23
1928; Meininger, 1966; Neville, 2000; Powers, 1959; Washa, 1955), the amount of water per 24
unit volume of concrete or the ratio of water to cement has a significant effect on the 25
shrinkage process of concrete. Also, the amount of cement per unit volume is one of the most 26
important factors contributing to shrinkage (Mailvaganam et al., 2002; Troxell et al., 1968; 27
Nagataki, 1970). Thus, concrete shrinkage increases with using more water and cement 28
(paste) in concrete. The moisture content of the sub-base has also an important effect on the 29
drying shrinkage and curling moment of the concrete (Nagataki, 1970; Suprenant, 2002). As 30
the shrinkage gradient rises the curling moment of concrete increases (Carlson, 1938; Keeton, 31
1979). Also, temperature and relative humidity at the surface of the slab, along with moisture 32
content in the sub-base, greatly affects the amount of curling at slab edges (Mailvaganam et 33
al., 2002; Suprenant, 2002). 34
TRB 2014 Annual Meeting Original paper submittal - not revised by author
3
1
As described in the preceding paragraphs, there have been numerous analytical and 2
experimental investigations to characterize drying shrinkage as a material property. There, 3
however, are not enough data to evaluate the strains and stresses within concrete subjected to 4
moisture gradients and restrained shrinkage. Given this gap, this research provides an 5
opportunity to improve our understanding of warping and our ability to predict its effects. 6
The general objectives of this research are: 7
8
(1) To provide reliable warping data in a controlled environment for typical Portland cement 9
concrete, high-strength concrete, Portland cement concrete with shrinkage reducing 10
admixture, and Portland cement concrete with calcium sulfoaluminate cement (also known as 11
“shrinkage compensating concrete”); 12
(2) To investigate a representative selection of shrinkage magnitude for concrete; 13
(3) To develop a better understanding of ASTM C 157 and C 878 test methods (ASTM, 14
2008) versus the “shrinkage from time zero” test method; and 15
(4) To characterize behavior of slabs-on-ground for the selected materials where shrinkage 16
and warping are a concern. 17
18
This paper summarizes the research for seven large-scale test specimens of slabs-on-ground 19
with the six selected concrete mixtures monitored for approximately 2 years. This research 20
follows the test protocol of Bissonnette et al. (2007) for slabs-on-ground with more realistic 21
conditions. The main variables of the investigation are the concrete composition (Portland 22
cement concrete, high-strength concrete, concrete with shrinkage reducing admixtures, and 23
concrete with calcium sulfoaluminate cement). The items monitored for the investigation 24
include visual observation of cracks; surface strains obtained using Demec target strain 25
gauges; interior relative humidity and interior temperature in 12.7 mm increments through the 26
depth of a slab; compressive strength; unrestrained and restrained expansion; and shrinkage 27
from time zero test results. This was done in order to obtain as much information as possible 28
for a better understanding of warping and shrinkage, to enhance the prediction of its effects, 29
and to compare the dimensional stability sensitivity of various materials. 30
31
32
33
34
TRB 2014 Annual Meeting Original paper submittal - not revised by author
4
2. Experimental program 1
2.1 Concrete mixtures and materials 2
Six concrete mixtures were selected for the investigation: normal-strength Portland Cement 3
Concrete (PCC), High-Strength Concrete (HSC), normal concrete mixes with two types of 4
Calcium SulfoAluminate cement [Komp I and Rapid Setting CSA Cement] (CSA and 5
RSCC), and normal concrete mixes with two types of common Shrinkage Reducing 6
Admixtures [Eclipse and Tetraguard] (SRA #1 and SRA #2). The only difference between 7
PCC and concrete mixes with SRAs is adding SRAs to the same PCC mixes. The concrete 8
mixes used in this research are summarized in Table 1. The mixtures (CSA and RSCC) using 9
two types of calcium sulfoaluminate cement, Komp I(CSA), a so-called shrinkage 10
compensating concrete and Rapid setting CSA cement (RSCC) which were made in the 11
mixer at the laboratory. The rest of the concrete mixtures (PCC and HSC) using normal-12
strength and high-strength concrete were provided by a local ready mix concrete company in 13
Oklahoma. 14
15
Table 1 Concrete mixes for slab specimens. 16
Materials
(kg/m3)
SRA #1 SRA #2 PCC HSC CSA-a CSA-b RSCC
Komp I - - - - 71.2 71.2 -
Portland Cement 211.1 210.5 210.5 322 219.4 219.4 -
Fly Ash 52.2 52.2 52.2 106.7 - - -
Rapid setting CSA Cement - - - - - - 390.2
Citric Acid - - - - - - 2.97
Course Aggregate 57 1097.1 1097.1 1097.1 1097.1 1037.8 1037.8 1050.8
Sand 867.6 867.6 867.6 709.2 779.8 779.8 775.1
Water 157.7 157.7 157.7 156.6 160.1 161.3 172
MR (Polyheed 1020 (L/m3)) - - - - 2.48 2.5 2.04
MR (Pozzolith 80 (L/m3)) 0.5 0.54 0.54 1.12 - - -
Eclipse (L/m3) 4.95 - - - - - -
Tetraguard (L/m3) - 4.95 - - - - -
Water-to-cement ratio 0.023 0.023 0.023 0.014 0.021 0.021 0.017
28 day compressive strength
(MPa) 26.2 23.8 21.7 36.2 41 39 69
17
18
19
20
21
22
23
TRB 2014 Annual Meeting Original paper submittal - not revised by author
5
2.2 Characterization tests for materials 1
A number of characterization tests, as listed in Table 2, were developed in the investigation to 2
typify the materials. The characterization tests followed ASTM standard tests such as 3
compressive strength test (ASTM C 39), length change test (ASTM C 157), and restrained 4
expansion test (ASTM C 878). Therefore, cylinders and prism test specimens were filled with 5
the concrete from each batch. The ASTM test measurements began at the first few hours after 6
placing concrete into the form according to the test method description. Also, the shrinkage 7
from time zero method was used for the purpose of characterizing the materials. Test 8
specimens were made 75 x 75 x 325 mm. A 25 mm plastic block was placed at one end of the 9
specimen. Therefore, concrete was free to move due to shrinkage at this end (free end). A 10
steel plate (form) on the other end of specimen restrained the concrete in shrinkage and 11
expansion provided fixed end (Figure 1). Two bolts were screwed at the center of the ends. 12
The bolts were 279 mm apart and were used as the original length of specimens for 13
calculating the strain 14
Table 2 Material characterization tests 15
Test description Lab RH & T
60% RH 21 °C 30% RH 21 °C
Compressive
strength (ASTM C
39)
+
–
Length change
(ASTM C 157) + +
Restrained
expansion (ASTM C
878)
+
–
shrinkage from time
zero test method
+
–
16
(Figure 1). A dial gauge was installed at the free end (the end with the plastic block) of each 17
specimen and the length change was recorded at the time “zero” which was immediately after 18
placing concrete into the form (Figure 1) and were continued for 110 days. The dial gauge 19
was connected to the bar and bar went through the hole and touched the back of the nut 20
(bolt). The specimens were placed in the chamber at 21°C and 60% relative humidity and 21
were wet cured by placing wet sponges on top of specimens for 7 days. Then the specimens 22
were demolded from the sides (not the ends). The concrete mixes used for the shrinkage from 23
time zero method were similar to the large-scale slab concrete used in this study with the 24
minor exception being due to the type of water reducer used on 4 of the samples. 25
TRB 2014 Annual Meeting Original paper submittal - not revised by author
6
1
Figure 1 Shrinkage from time zero test mold and shrinkage from time zero test specimen. 2
2.3 Slab experiments 3
Seven slab test specimens of 75 mm x 900 mm x 6000 mm were cast on 100 mm moist 4
compacted sand on the ground (Figure 2). Two slabs (CSA-a and CSA-b) produced with 5
calcium sulfoaluminate cement of Komp I and one slab per other combination of concrete 6
mixes (PCC, HSC, RSCC, SRA #1 and SRA #2) was cast. The slab specimens were located in 7
the test facility with a controlled environmental chamber. The laboratory relative humidity 8
and temperature are generally 60% and 21 ºC. The relative humidity was dropped to 30% for 9
the second year of the investigation as part of the experiment. The purpose of reducing 10
ambient relative humidity was to increase the moisture gradient into the slab due to a lower 11
relative humidity at the top surface of the slab. Slabs were cast in specially designed pits, 12
truly on-grade. The idea was that while the base is kept moist by ground water, the top of the 13
slabs are exposed to the low ambient relative humidity environment in the lab; this increases 14
the moisture gradient in the slab and increases warping. 15
16
Figure 2 (a) Slab on 102 mm moist-compacted sand on grade and (b) 225 mm of concrete 17
slab edge restrained by trusses 18
TRB 2014 Annual Meeting Original paper submittal - not revised by author
7
1
The slabs were fully restrained longitudinally by casting the test slabs around a transverse 2
steel truss that was attached to the edges and the test pits. The steel trusses shown in Figure 2 3
transfer all the loads to two D25 reinforcing bars (diameter = 25 mm), cast previously in the 4
existing floor along each edge of the test specimen. The test slab ends were thickened to 230 5
mm to accommodate this detail (Figure 2). A minimum longitudinal reinforcing steel for 6
temperature ( s = 0.0015) was used for all of the slabs. Four 6.35 mm diameter longitudinal 7
reinforcing bars were continuous through the joints. Upon casting the slabs, contraction joints 8
were saw cut 25 mm deep at 1500 mm from each end to provide a 3000 mm long central test 9
section. All the slab specimens except one (RSCC) were cured with wet burlap and a plastic 10
sheet for 7 days. The slab (RSCC) made with Rapid setting CSA cement was cured with 11
spraying water for 4 ~ 5 hours after casting. 12
13
Slab monitoring began after 7 days of moist curing. The monitoring consisted of the 14
following: (1) visual observation for surface cracking; (2) surface strain and joint opening 15
measurement in the longitudinal direction using Demec target strain gauges with 178 mm 16
gauge length; (3) internal relative humidity at 12.7 mm increments through the depth of the 17
slab (ASTM F 2170); (4) internal temperature at 12.7 mm increments through the depth of 18
the slab; and (5) ambient relative humidity and temperature. The slab monitoring represented 19
the behavior of slab exposed to the low relative humidity at the top surface and high relative 20
humidity at the bottom surface of the slab. Figure 3(a) shows the location of Demec target 21
strain gauges at mid-span at 1500 mm from the ends and the location of the RH (Relative 22
Humidity) meter at 375 mm from mid-span of the slab. Figure 3(b) shows a profile of the slab 23
deformation due to warping. It displays the end restraints and warping at joint opening. 24
25
26
27
28
29
TRB 2014 Annual Meeting Original paper submittal - not revised by author
8
1
(a) Top view of slab specimen 2
3
(b) Profile of slab deformation due to warping 4
Figure 3 Top view and profile of slabs-on-ground specimen (Note: 1 in. = 25 mm; 1 ft = 300 5
mm) 6
7
3. Test results and discussion 8
3.1 Characterization test results 9
The ASTM C 39 compressive strength results were recorded for one year, the ASTM C 157 10
length change results were recorded for approximately 2 years (Figure 4(a)), and the ASTM 11
C 878 restrained expansion results for CSA were recorded during the time the concrete 12
expanded due to the test method description. The ASTM C 157 method is a standard test 13
method for length change of hardened hydraulic-cement mortar and concrete, and the ASTM 14
C 878 method is a standard test method for restrained expansion of shrinkage compensating 15
concrete (CSA). Upon curing the specimens into the limestone and water (saturated water 16
with lime) for 7 days, measurement of length variations was initiated. From Figure 4(a), the 17
comparison between PCC and SRA #1 results represents that SRA #1 had a minimal impact 18
on shrinkage at short term and almost no impact at long term. Therefore, it can be concluded 19
that the shrinkage reducing admixture did not provide a noticeable improvement on reducing 20
over all shrinkage in normal concrete. The HSC shrank in a greater rate in comparison to the 21
other concrete mixes used in this research. The difference between HSC and PCC was due to 22
many issues that affect the shrinkage in HPC, like all types of concrete shrinkage (e.g., 23
TRB 2014 Annual Meeting Original paper submittal - not revised by author
9
autogenous shrinkage, porous system, capillary size, paste quantity, etc.) but the increased 1
paste volume in HSC is generally referenced as a leading contributor to the increase in 2
shrinkage (Mailvaganam et al., 2002; Troxell et al., 1968; Nagataki, 1970). 3
4
5
Figure 4 (a) Unrestrained expansion ASTM C 157 versus time (PCC, HSC and SRA #1) and 6
(b) restrained expansion ASTM C 878 versus time (CSA) 7
8
On the other hand, the CSA expanded more than 4 times greater than the other concrete 9
mixtures during the first 7 days of curing (Figure 4(b)). It can be expected that the large 10
expansion of CSA is able to offset the restrained shrinkage caused by drying shrinkage of 11
concrete at the long term and that the possibility of warping and cracking caused by drying 12
shrinkage is reduced. Note that shrinkage compensating cement concrete (CSA) was used for 13
two of the concrete specimens (CSA-a and CSA-b), and both concrete mixes exhibited similar 14
characteristics. Therefore, all the results showing CSA are the average of the data from the 15
two concrete specimens using CSA. It is noted that the ASTM C 878 method provides data 16
only for expansion of CSA; therefore, the shrinkage from time zero tests were additionally 17
conducted for an equal basis comparison as described in the following paragraphs. 18
19
The shrinkage from time zero test results were recorded for 28 days and 110 days. Comparing 20
the results from the shrinkage from time zero method to the ASTM C 157 unrestrained 21
0 50 100 150 200 250 300 350 400 450 500 550 600
Day
-0.05
-0.04
-0.03
-0.02
-0.01
0
0.01
0.02
0.03
Str
ain
(%
)
SRA #1
HSCPCC
0 50 100 150 200 250 300 350 400 450 500 550 600
Day
-0.15
-0.1
-0.05
0
0.05
0.1
Str
ain
(%
)
CSA (Avg. of CSA-a & CSA-b)
SRA #1
TRB 2014 Annual Meeting Original paper submittal - not revised by author
10
expansion results shows that the trend after 7 days was the same for both testing methods 1
(Figure 5). The concrete specimens for the shrinkage from time zero method did not show 2
expansion in the first 7 days of curing because the expansion was restraind due to the steel 3
plate at the fixed end of the specimens (Figure 3), while the expansion was unrestrained for 4
the ASTM C 157 method.. It is noted that the shrinkage from time zero method was found 5
easier to perform in comparison with the ASTM C 157 and C 878 methods. Figure 5 is a 6
representative result comparing two testing methods for PCC and SRA #1. 7
8
9
Figure 5 Shrinkage from time zero test results for SRA #1 in comparison with those obtained 10
from ASTM C 157 11
12
The Shrinkage from time zero test results are also compared with those of the ASTM C 878 13
method for CSA in Figure 6. The shrinkage from time zero test results show the same general 14
trend when compared to that of the ASTM C 878 method. Both testing methods showed 15
expansion for the first 7 days of curing; however, the amount of expansion was different due 16
to the smaller stiffness of the rod used to restrain expansion of concrete for the ASTM C 878 17
18
Figure 6 Shrinkage from time zero test results for CSA in comparison with those obtained from ASTM C 878 19
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
Day
-0.04
-0.03
-0.02
-0.01
0
0.01
0.02
0.03
0.04
Str
ain
(%
)
SRA #1 (Shrinkage from Time Zero)
SRA #1 (ASTM C 157)
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28
Day
0
0.02
0.04
0.06
0.08
0.1
Str
ain
(%
)
CSA (Shrinkage from Time Zero)
Avg. of CSA-a & CSA-b (ASTM C 878)
TRB 2014 Annual Meeting Original paper submittal - not revised by author
11
method, compared with the stiffness of the steel frame used to restrain expansion of concrete 1
for the shrinkage from time zero method. Overall, there was little shrinkage in CSA for about 2
110 days of curing after the initial expansion, which verified the effectiveness of shrinkage 3
compensating concrete in reducing the drying shrinkage. 4
5
3.2 Slab monitoring results 6
3.2.1 Surface measurements and joint opening 7
The average joint opening data for the six concrete mixtures of seven slabs are represented in 8
Figure 7, with complete surface measurements for all the slabs provided in Figure 3. The 9
average of the east and west opening strains were used. The shrinkage was measured from the 10
point of initial set and continued for 600 days. This measurement is not exactly the same as 11
those from the ASTM C 157 and C 878 methods, because the slab bottom is exposed to the 12
moisture in the soil and the slab top is exposed to the relatively low humidity in a controlled 13
environment, while for the ASTM C 157 and C 878 methods, all sides of the specimens are 14
exposed to the controlled relative humidity. The relative humidity of the lab (air above the 15
top of the slab) was 60% for the first year and then was reduced to 30%. 16
17
Figure 7 Width of joint opening versus time 18
19
Figure 7 shows the average width of the joint opening versus time and illustrates that the use 20
of SRA only reduced shrinkage in the first week. After the first week the slope of the line 21
representing the SRA data matches the slope of the line for PCC. This once again indicates 22
that the use of SRA reduced shrinkage at short term but did not impact the shrinkage at long 23
term. Figure 7 also shows that HSC had the largest expansion/crack at the joints and CSA had 24
the smallest expansion/crack at the joints. When comparing the strain of control joints for 25
CSA with SRAs, it is found that shrinkage compensating concrete (CSA) had a positive effect 26
on reducing shrinkage of concrete at both short and long terms compared with all concretes 27
0 50 100 150 200 250 300 350 400 450 500 550 600
Day
0
1
2
3
4
5
6
7
8
Wid
th o
f Jo
int
Op
en
ing
(m
m)
0
1
2
3
4
5
6
7
8
Wid
th o
f Join
t O
penin
g (
mm
)
HSCPCC
SRA #2
SRA #1RSCC
CSA (Avg. of CSA-a & CSA-b)
TRB 2014 Annual Meeting Original paper submittal - not revised by author
12
including SRAs. Conversely, joint opening expansion continued over the long term (over 600 1
days) with PCC, HSC and SRAs. The comparison between HSC and PCC shows that the 2
cracks of HSC were wider than those of PCC, and HSC also shrank quicker than normal 3
concrete (PCC). The HSC shrank at a faster rate than other mixes during the first few weeks 4
(early age) after curing. Based on the discussion in this and the preceding paragraphs, it is 5
concluded that the use of CSA reduces warping in comparison to the other types of concrete 6
used in this research. 7
8
Table 3 summarizes the joint expansion (crack growth) for the mixes after 574 days. It can be 9
seen that, the slab using SRAs shows join expansion 2.1 to 2.4 times greater than CSA, PCC’s 10
joint expansion was 3.7 times greater than CSA, and HSC’s joint expansion was 4.6 times 11
greater than CSA. Therefore, it can be concluded that CSA is very stable with less crack and 12
warping at long term compared to other mixes in this research. 13
14
Table 3 Joint expansion at 574 days, 15
Mix Joint expansion (mm) Compared to CSA
CSA 1.6 -
SRA #1 3.4 231%
SRA #1 3.9 244%
PCC 5.9 367%
HSC 7.4 459%
16
17
3.2.2 Comparing slab monitoring with ASTM C 157 18
Figure 8 shows the slab behavior in comparison with the ASTM C 157 method for SRA #1, 19
PCC and HSC. A significant difference was observed between the slab behavior and the 20
ASTM C 157 results. Again, this might be due to the different conditions of humidity on each 21
face of the slab. In light of that, it appears that the ASTM C 157 method does not provide 22
accurate results for predicting slab-on-ground behavior. Conversely, the current slab-on-23
ground experiment reflects the realistic exposure conditions and actual dimensional and 24
material properties of concrete. 25
TRB 2014 Annual Meeting Original paper submittal - not revised by author
13
1
2
3
Figure 8 Strains at mid-span for SRA #1, PCC and HSC in comparison with those obtained 4
from ASTM C 157 5
6
3.2.3 Interior relative humidity and temperature 7
This subsection presents general results for internal temperature and relative humidity of 8
slabs-on-ground, measured after 1 year. Figure 9(a) shows that there were no temperature 9
changes through the depth of the slabs which means the ambient temperature remained 10
constant, whereas Figure 9(b) presents that the RH of the slabs within the top 12.5 mm of 11
concrete slab was higher than the ambient relatively humidity in the air. From the internal 12
relative humidity of slabs (Figure 9(b)), it is found that HSC had the highest moisture 13
gradient and CSA had the least moisture gradient. The PCC, SRA #1, SRA #2 and RSCC 14
0 50 100 150 200 250 300 350 400 450 500 550 600
Day
-0.05
-0.04
-0.03
-0.02
-0.01
0
0.01
0.02
Str
ain
(%
)
Slab at Mid-Span
ASTM C 157
(a) SRA #1
0 50 100 150 200 250 300 350 400 450 500 550 600
Day
-0.05
-0.04
-0.03
-0.02
-0.01
0
0.01
0.02
Str
ain
(%
)
Slab at Mid-Span
ASTM C 157
(b) PCC
0 50 100 150 200 250 300 350 400 450 500 550 600
Day
-0.05
-0.04
-0.03
-0.02
-0.01
0
0.01
0.02
Str
ain
(%
)
Slab at Mid-Span
ASTM C 157
(c) HSC
TRB 2014 Annual Meeting Original paper submittal - not revised by author
14
moisture gradients were between CSA and HSC. There were almost no changes in the 1
moisture content at 63.5 mm depth in the slab. Most of the moisture content changes occurred 2
between 25 and 63.5 mm depth from the top surface of the slab. Similar trends were observed 3
after reducing the ambient RH (in the air) to 30% from 60%. The measured interior relative 4
humidity (RH) of the slabs-on-ground is directly related to warping. The materials RH 5
gradient is related to a strain gradient due to the materials tendency to swell of shrink due to 6
moisture. In this case CSA has less of a RH gradient which indicates it has less of a strain 7
gradient and as such it is less sensitive to warping than typical Portland cement concrete or 8
high-strength concrete, Shadravan (2011) reports the complete results for the slabs’ interior 9
temperature and relative humidity, as well as ambient temperature and relative humidity in 10
the air. 11
12
Figure 9 Interior slab temperature and relative humidity as a function of depth from slab top 13
surface, measured after 1 year of curing 14
15
The discussion in the preceding paragraph brings up two possible hypotheses. The first 16
possible hypothesis is that the porosity of the HPC is low since it is hard for moisture to get 17
in to the concrete from the bottom surface of the slab and it is much easier to dry it out from 18
the top. Thus, the least porous material has the lowest average relative humidity, The second 19
possible hypothesis is that HPC is the most porous material that has the lowest relative 20
humidity at the top surface (i.e., the concrete is drying further into its core). These two 21
hypotheses, however, would require further research to validate and may be influenced by 22
other factors such as the amount of paste in each mix. 23
24
25
26
67 68 69 70 71 72 73 74
Temperature (oF)
50 60 70 80 90 100
Relative Humidity (%)
-80
-70
-60
-50
-40
-30
-20
-10
0
Depth
(m
m)
-80
-70
-60
-50
-40
-30
-20
-10
0
Depth
(m
m)
(a) Interior Slab Temperature
RSCC
CSA
(Avg. of
CSA-a & CSA-b)
SRA #1, SRA #2, HSC and PCC (b) Relative Humidity
RSCCHSC
SRA #1 SRA #2
CSA (Avg.)PCC
[Air: 60% RH]
[Soil: 100% RH]
TRB 2014 Annual Meeting Original paper submittal - not revised by author
15
4. Conclusions 1
2
The main objective of this research was to provide reliable warping data based on the internal 3
strain measurements of pavement, ASTM C 157, ASTM C 878, and shrinkage from time zero 4
drying shrinkage measurements for the selected concrete mixtures in a controlled 5
environment. The main conclusions of the investigation are summarized as follows: 6
7
(1) Typical Portland cement concrete, high-strength concrete and concrete with SRAs 8
continued to exhibit increased crack growth at approximately 2 years. 9
10
(2) The shrinkage from time zero test method gets the same results, measures shrinkage more 11
accurately, and is easier to perform when compared to ASTM C 157 and C 878 method. 12
13
However, the shrinkage from time zero tests provide less accurate results for the prediction of 14
slabs-on-ground behavior when compared to the large-scale tests of actual slab-on-ground 15
exposed to realistic conditions. 16
17
(3) The ASTM C 157 method appears inadequate to assess slab shrinkage behavior because 18
the ASTM C 157 method and the actual slab-on-ground test method show significant 19
differences in the results. 20
21
(4) Based on the measured interior relative humidity of the slabs-on-ground, shrinkage 22
compensating concrete or concrete using calcium sulfoaluminate cement (CSA) was 23
inherently less sensitive to warping than typical Portland cement concrete or high-strength 24
concrete. 25
26
(5) Shrinkage compensating concrete or concrete using calcium sulfoaluminate cement was 27
extremely stable, with little or no long-term shrinkage, cracking or warping. This stability 28
was noted at both early age and at approximately 2 years. 29
30
(6) Shrinkage reducing admixtures decreased shrinkage and crack growth a small amount at 31
an early age but did not impact the rate of shrinkage growth or sectional stability after that 32
period of time. 33
34
TRB 2014 Annual Meeting Original paper submittal - not revised by author
16
1
Acknowledgements 2
The research lab construction and the materials used for the investigation has been financially 3
supported by CTS Cement Manufacturing Corporation. The views expressed are those of 4
authors, and do not necessarily represent those of the sponsors. 5
6
References 7
Carlson RW (1938) Drying Shrinkage of Concrete as Affected by Many Factors. Proceedings 8
of the American Society for Testing and Materials, V. 38, Part II, ASTM, West 9
Conshohocken, Pa. 10
Tremper B and Spellman DL (1963) Shrinkage of Concrete-Comparison of Laboratory and 11
Field Performance. High-way Research Record No. 3, Highway Research Board, pp. 30–12
61. 13
Ytterberg RF (1987) Shrinkage and Curling of Slabs on Grade: Part II-Warping and Curling. 14
Concrete International, V. 9, pp. 54–61. 15
ACI Committee 224 (2008) Control of Cracking in Concrete Structure (ACI 224R-01; 16
Reapproved 2008). American Concrete Institute, Farmington Hills, Mich., 46 pp. 17
Mailvaganam NP, Wellington L, Repette WL and Taylor DA (2002) Curling in Concrete 18
Floors-Causes and Repair. The Indian Concrete Journal, Vol. 76, No. 6, pp. 341–348. 19
Ramseyer CCE Investigation of Very Early Strength Concrete with Low Shrinkage 20
Properties. M.S. thesis, School of Civil Engineering and Environmental Science, The 21
University of Oklahoma, Norman, Okla. 22
Holt EE (2001) Early Age Autogenous Shrinkage of Concrete. VTT Technical Research 23
Center of Finland, ESPOO, 184 pp. 24
Hansen W (1987) Drying Shrinkage Mechanisms in Portland Cement Paste. Journal of the 25
American Ceramic Society, V. 70, No. 5, pp. 323–328. 26
Hart WE (1928) The Technique Involved in Laying a Good Concrete Floor. Engineering and 27
Contracting, V. 67, No. 8, pp. 393–394. 28
Meininger RC (1966) Drying Shrinkage of Concrete. Engineering Report No. RD3 (A 29
Summary of Joint Research Laboratory Series J–135, J–145, 173, and 143), National 30
Ready Mixed Concrete Association, Silver Spring, Md., 22 pp. 31
Neville AM (2000) Properties of Concrete. Fourth Edition, Pearson Education Limited, 32
England, 844 pp. 33
TRB 2014 Annual Meeting Original paper submittal - not revised by author
17
Powers TC (1959) Causes and Control of Volume Change. Journal, PCA Research and 1
Development Laboratories, V. 1, No. 1, pp. 29–39. 2
Washa GW (1955) Volume Changes and Creep. Significance of Tests and Properties of 3
Concrete and Concrete Aggregates, STP–169, ASTM, pp. 115–128. 4
Troxell GE, Davis HE and Kelly JW (1968) Composition and Properties of Concrete. 2nd 5
Edition, McGrawHill Book Co., New York, N.Y., 529 pp. 6
Nagataki S (1970) Shrinkage Restraints in Concrete Pavements. Journal of the Structural 7
Division, Proceedings of the American Society of Civil Engineers, Vol. 96, No. 7, pp. 8
1333–1358. 9
Suprenant BA (2002) Why Slabs Curl, Part I: A Look at the Curling Mechanism and Effect 10
of Moisture and Shrinkage Gradients on the Amount of Curling. Concrete International, 11
Vol. 24, No. 4, pp. 56–61. 12
Keeton JR (1979) Shrinkage Compensating Cement for Airport Pavement. Report No. FAA–13
RD–79–11, Phase 2, Federal Aviation Administration, Washington D.C. 14
ASTM V.04.09 (2008). American Society for Testing and Materials Annual Book of ASTM 15
Standards. American Society for Testing and Materials, Pa. 16
Bissonnette B, Attiogbe EK, Miltenberger MA and Fortin C (2007) Drying Shrinkage, 17
Curling, and Joint Opening of Slab on Ground. ACI Materials Journal, V. 104, No. 3, pp. 18
259–267. 19
Shadravan S (2011) Dimensional Stability of Concrete Slabs on Grade. Ph.D. thesis, School 20
of Civil Engineering and Environmental Science, The University of Oklahoma, Norman, 21
OK, 310 pp. 22
23
TRB 2014 Annual Meeting Original paper submittal - not revised by author