Академический Документы
Профессиональный Документы
Культура Документы
by
Lianxiang Du
2001
Laboratory Investigations of Controlled Low-Strength Material
by
Dissertation
Presented to the Faculty of the Graduate School of
in Partial Fulfillment
of the Requirements
Doctor of Philosophy
Copyright 2001 by
Du, Lianxiang
________________________________________________________
UMI Microform 3031045
Copyright 2001 Bell & Howell Information and Learning Company.
All rights reserved. This microform edition is protected against
unauthorized copying under Title 17, United States Code.
____________________________________________________________
Committee:
David W. Fowler
David Trejo
Ellen Rathje
Alan F. Rauch
Dedication
people who encouraged and supported me during my study in China and in US.
This dissertation is especially dedicated to my dad, who always stayed late with
This report would not have been possible without Dr. Kevin J. Folliard.
Special gratitude goes to Dr. David Trejo at Texas A&M University, who
inspirations will benefit my whole life. Longhorns and Aggies can be not only
committee members, Dr. David Fowler, Dr. Ellen Rathje, and Dr. Alan Rauch.
Their guidance and technical support were very valuable to me. Special thanks
Lilma dos Santos Ribeiro and Jessica D. Reath in part of the research are
v
Sherian Williams, and many undergraduate students who helped me with the
laboratory work.
vi
Laboratory Investigations of Controlled Low-Strength Material
Publication No._____________
controlled low-strength material (CLSM) and the corresponding test methods for
foundry sand, and high-carbon fly ash was found to increase water demand to
achieve the desired flow in CLSM. Bottom ash tended to increase bleeding,
whereas foundry sand decreased bleeding tendencies. Two methods were used to
evaluate the setting and hardening of fresh CLSM mixtures, and their mechanisms
were evaluated. New methods were developed to measure the segregation and
vii
triaxial shear strength, and drying shrinkage were studied. Predictive models
capping materials, including neoprene pads, gypsum and sulfur, were studied. In
addition, factors affecting the strength gain of CLSM mixtures, including curing
using modified version of ASTM D 560. Higher air contents and compressive
freezing-and-thawing cycles.
Although CLSM was found less corrosive than soil, the potential
viii
Table of Contents
CHAPTER 1: INTRODUCTION 1
2.1 INTRODUCTION............................................................................................. 6
ix
2.3 MIXTURE PROPORTIONS........................................................................... 17
2.6 ECONOMICS.................................................................................................. 41
x
2.7.1.5 Utility Bedding ........................................................................ 51
2.7.1.6 Erosion Control ....................................................................... 57
2.7.1.7 Void Fill................................................................................... 57
2.7.1.8 Environmental Protection........................................................ 59
3.1 INTRODUCTION........................................................................................... 64
xi
3.5 CEMENT......................................................................................................... 71
4.1 INTRODUCTION........................................................................................... 82
xii
4.3.7 Excavation Study by CLSM Mixtures .............................................. 107
xiii
5.3.1.3 Effects of Drainage Conditions on Compressive Strength.... 154
5.3.1.4 Specimen Preparation before Testing.................................... 157
5.3.1.5 Alternative Capping Materials for Compression Testing...... 159
5.3.1.6 Effects of Load Rate on Compressive Strength .................... 168
5.3.1.7 Effect of Specimen Size ........................................................ 172
5.3.1.8 Effects of Different Testing Machine .................................... 173
5.3.1.9 Effects of Curing Temperature and Humidity....................... 174
5.3.2 Triaxial Shear Strength...................................................................... 182
5.3.3 Water Permeability............................................................................ 184
5.3.4 Drying Shrinkage ............................................................................... 186
5.3.5 California Bearing Ratio .................................................................... 187
5.3.6 Resilient Modulus.............................................................................. 188
5.3.7 Excavation Study............................................................................... 189
xiv
5.6 LEACHING AND ENVIRONMENTAL IMPACT..................................... 224
REFERENCES 244
VITA 255
xv
List of Tables
xvi
Table 5.10: Compressive strength at 91 days using different curing methods.... 149
Table 5.11: Compressive strength at 7 days under different drainage conditions .....
............................................................................................................................. 156
Table 5.12: Compressive strength at 28 days under different drainage conditions ..
............................................................................................................................. 157
Table 5.13: Influence of specimen air-drying on compressive strength (Mixture B-
2).............................................................................................................. 158
Table 5.14: Compressive strength results using different capping materials ..... 160
Table 5.15: Coefficients of variation for compressive strengths using different
capping materials ..................................................................................... 161
Table 5.16: Comparison of “a” values calculated by two methods.................... 165
Table 5.17: Effects of loading rates (wide range) on compressive strength ...... 169
Table 5.18: Effects of load rates (narrow range) on compressive strength........ 170
Table 5.19: Effects of loading rates (wide range) on deformation at peak load 170
Table 5.20: Effect of narrow range load rates (narrow range) on deformation at
peak load.................................................................................................. 171
Table 5.21: Effects of cylinder size on unconfined compressive strength......... 173
Table 5.22: Effects of different testing machines on compressive strength
(coefficient of variation is included in parentheses)................................ 174
Table 5.23: Compressive strength of specimens cured at different temperatures at
different ages ........................................................................................... 177
Table 5.24: Results of triaxial compression tests ............................................... 182
Table 5.25: Water permeability (hydraulic conductivity) of selected CLSM
mixtures ................................................................................................... 184
Table 5.26: Test results of water permeability after freezing-and-thawing cycles ...
............................................................................................................................. 185
Table 5.27: Drying shrinkage of selected CLSM mixtures ................................ 187
Table 5. 28: CBR values for six selected CLSM mixtures................................. 187
Table 5.29: Resilient modulus of six selected CLSM mixtures ......................... 188
Table 5.30: Results of excavation study............................................................. 190
Table 5.31: Comparison between field penetrometer values (field-cured samples)
and compressive strength values (field-cured cylinders) ........................ 191
Table 5.32: Results of excavatability study........................................................ 194
Table 5.33: Compressive and splitting tensile strengths at 7 and 28 days ......... 198
Table 5.34: Effects of temperature and drying conditions on splitting tensile
strength of CLSM.................................................................................... 200
Table 5.35: Freezing-and-thawing results for selected CLSM mixtures (after 7
days of curing) ......................................................................................... 202
Table 5.36: Freezing-and-thawing results for selected CLSM mixtures (after 28
days of curing) ......................................................................................... 202
Table 5.37: Freezing-and-thawing test results of D-series mixtures .................. 203
xvii
Table 5.38: Wetting-and-drying results for selected CLSM mixtures (after 7 days
of curing) ................................................................................................. 205
Table 5.39: Wetting and drying results for selected CLSM mixtures (after 28
days of curing) ......................................................................................... 206
Table 5.40: pH and resistivity values of CLSM mixtures.................................. 208
Table 5.41: Percent mass loss for samples embedded in CLSM only................ 213
Table 5.42: Statistical comparison of repeated CLSM mixtures (uncoupled
corrosion)................................................................................................. 216
Table 5.43: Statistical results of mass-loss test for uncoupled samples (CLSM
mixtures containing fly ash) .................................................................... 218
Table 5.44: Statistical results of mass-loss test for uncoupled samples (CLSM
mixtures containing no fly ash) ............................................................... 219
Table 5.45: Percent mass loss for coupled samples embedded in sand .............. 221
Table 5.46: Statistical comparison of repeated CLSM mixtures on corrosion of
coupled coupons embedded in sand ........................................................ 222
Table 5.47: Statistical results of mass-loss test for coupled samples (CLSM
mixtures containing fly ash) ...................................................................... 223
Table 5.48: Statistical results of mass-loss test for coupled samples (CLSM
mixtures without fly ash) ......................................................................... 224
Table 5.49: Analysis of heavy metal concentration of extract from raw materials
(mg/L) ...................................................................................................... 226
Table 5. 50: TCLP test results for Class C fly ash, Class F fly ash, and bottom ash
(mg/L) ...................................................................................................... 227
Table B-1: Needle penetration (NP) and soil penetrometer (SP) results............ 240
Table B-1: Needle penetrAtion (NP) and soil penetrometer (SP) results (cont’d)....
............................................................................................................................. 241
Table B-1: Needle penetration (NP) and soil penetrometer (SP) results (cont’d)242
Table B-2: Needle penetration (NP), soil penetrometer (SP) and vane shear tester
(VS) Results............................................................................................. 243
xviii
List of Figures
xix
Figure 5.4: Comparison between needle penetration (NP) and soil penetrometer
reading (SP) values and the back calculated resistance from soil
penetrometer (BSP). ................................................................................ 130
Figure 5.5: Setting time needle and the wedge at the tip of the needle (following
ASTM C 403).......................................................................................... 132
Figure 5.6: Comparison between soil penetrometer and vane shear values ........ 133
Figure 5.7: Effects of variables on bleeding of non-air-entrained CLSM mixtures.
“+” means increasing the bleeding, “-“ means decreasing, and “X” means
the effect is not significant compared with variations. ........................... 134
Figure 5.8: Correlation between bleeding (weight percent) and subsidence....... 137
Figure 5.9: Load-deformation response of Mixture #12 at 3, 7, and 28 days .... 140
Figure 5. 10: Comparison of measured and predicted strengths of air-entrained
CLSM mixtures at the ages of 7, 28, and 91 days. .................................. 142
Figure 5.11: Comparison of measured and predicted strengths of non-air-entrained
CLSM mixtures ....................................................................................... 144
Figure 5.12: Effects of curing conditions on compressive strength for Mixtures G-
7 and G-8 ................................................................................................. 150
Figure 5.13: Effects of curing conditions on compressive strength for Mixtures G-
9 and G-10 ............................................................................................... 151
Figure 5.14: Curing condition effects on compressive strength of Mixture G-3
and G-6 .................................................................................................... 152
Figure 5.15: Curing condition effects on compressive strength of Mixture G-4
and G-5 .................................................................................................... 153
Figure 5.16: Effects of capping methods on coefficient of variation (7-day
compressive strength) .............................................................................. 162
Figure 5.17: Effects of capping methods on 163
Figure 5.18: Effects of capping methods on coefficient of variation (91-day
compressive strength) .............................................................................. 163
Figure 5.19: Load-deflection curves for mixture E-1 (compression testing at 28
days) ........................................................................................................ 167
Figure 5.20: Compressive strength of mixture H-1 for different curing conditions .
............................................................................................................................. 178
Figure 5.21: Compressive strength of mixture H-2 for different curing conditions .
............................................................................................................................. 178
Figure 5.22: Compressive strength of mixture H-3 for different curing conditions .
............................................................................................................................. 179
Figure 5.23: Compressive strength of mixture H-4 for different curing conditions .
............................................................................................................................. 179
Figure 5.24: Compressive strength of mixture H-5 for different curing conditions .
............................................................................................................................. 180
Figure 5.25: Compressive strength of mixture H-6 for different curing conditions .
............................................................................................................................. 180
xx
Figure 5.26: Stress-strain curves of mixture I-1 and I-4 at 28 days (triaxial
compression testing with an effective confining pressure of 69 kPa)..... 183
Figure 5.27: Penetration resistance (ASTM C 403) vs. compressive strength at
different ages (all C-series mixtures) ...................................................... 192
Figure 5.28: Cracking of CLSM mixture C-2 during DCP penetration............. 195
Figure 5.29: Correlation between stiffness (from GeoGauge) and DCP index.. 196
Figure 5.30: Correlation between DCP Index and Removibility Modulus (RE)
calculated using 28-day compressive strength (lab-cured cylinders) ...... 197
Figure 5.31: Correlation between DCP Index and Removibility Modulus (RE)
calculated using 240-day compressive strength (field-cured cylinders) . 197
Figure 5.32: A cylinder from mixture E-1, before and after being tested for
splitting tensile strength........................................................................... 199
Figure 5.33: Specimens after freezing-and-thawing cycles (CLSM mixtures
containing foundry sand) ......................................................................... 204
Figure 5.34: Chloride profiles for select CLSM mixtures after 22 weeks of
exposure................................................................................................... 210
xxi
Chapter 1: Introduction
1.1 GENERAL
whose use has grown in recent years. CLSM, often referred to as flowable fill, is
and often times, fly ash. Other by-product materials, such as foundry sand and
applications, especially for backfill, utility bedding, void fill, and bridge
CLSM as a bedding material for pipe, electrical, and other types of utilities and
approaches, either as a subbase for the bridge approach slab or as backfill against
compacted fill in these applications. These benefits include reduced labor and
faster construction, and the ability to place material in confined spaces. The
1
excavation, if required. Another advantage of CLSM is that it often contains by-
product materials, such as fly ash and foundry sand, thereby reducing the demand
Despite these benefits and advantages over compacted fill, the use of
CLSM is not currently as widespread as its potential might predict. One reason
for this is that CLSM is somewhat a hybrid material; that is, it is a cementitious
material that behaves more like a compacted fill. As such, much of the
information and discussions on its uses and benefits have fallen outside the
that govern the use of CLSM. However, these specifications differ from state to
state, and moreover, a variety of different test methods are currently being used to
specifications and testing methods, has also hindered the proliferation of CLSM
applications.
widespread CLSM use. For instance, it is often observed in the field that
ages. This can be a significant problem that translates into added cost and labor.
Other technical issues deserving attention are the compatibility of CLSM with
different types of utilities and pipes, the potential leaching of constituent materials
2
and elements, and the durability of CLSM subjected to freezing-and-thawing
cycles.
technology,
mixtures and recommend test methods for laboratory and field testing,
strength of CLSM,
3
7. Study the effects of curing temperature, humidity, and other
CLSM. Typical applications of CLSM are also provided with case histories.
which included 38 mixtures in the Stage One study and 67 mixtures in Stage Two.
4
Three types of aggregates (concrete sand, foundry sand, and bottom ash), three
types of fly ash (Class C, Class F, and High carbon), and American Society for
Testing and Materials (ASTM) Type I cement were included in the study.
Relevant physical and chemical properties of these materials were also measured.
weight, air content, compressive strength, CBR, resilient modulus, traixial shear
5
Chapter 2: Literature Review
2.1 INTRODUCTION
used primarily as a backfill in lieu of compacted fill.1 Several terms are currently
used to describe this material, including flowable fill, unshrinkable fill, controlled
density fill, flowable mortar, plastic soil-cement, soil-cement slurry, K-Krete and
compressive strength of 2.1 MPa or less. Some researchers consider the range of
0.3 to 1.1 MPa as a good index of sufficient strength and easy future excavation.
CLSM is critical and this may determine the success of CLSM in practices such
as utility bedding.
practices for a long time. CLSM is relatively new and is different from
is not the case for CLSM. One of its earliest applications was carried out in 1964
the Canadian River Aqueduct Project, which runs from north of Amarillo to south
6
of Lubbock, Texas.2 The material used in that project was called plastic soil-
cement. The soil used consisted of local blow sand deposits. A new construction
procedure was introduced, and the cost of this project was estimated to be 40
percent less than using conventional backfilling techniques. The productivity was
to compacted granular fill utilizing fly ash and concrete batching techniques. This
new backfill material, called flowable fly ash, was used in several applications in
the late 1970s.3-5 This material was composed principally of fly ash and typically
River project, it was estimated that more than $1 million was saved by using this
new material.3 One impressive and exciting feature of this material was that it
remained cohesive when being placed. Another characteristic was the steep angle
patents for K-Krete were issued to “Brewer et al.”.6 A typical K-Krete mixture
cement, and up to 0.35 to 0.40 m3 of water per m3 of product. The four patents
included mixture design, backfill technique, pipe bedding, and dike construction.
These patents were sold to Contech, Inc., in Minneapolis, Minn., who ceded the
patent rights later to the National Ready Mix Concrete Association (NRMCA)
with the stipulation that those rights may not be used in a proprietary manner.6,7
7
Since then, ready-mixed concrete producers and contractors have used similar
conventional compacted fill, similar materials have been developed and used
throughout the United States and Canada. However, the lack of a centralized
Committee 229 was established in 1984 under the title “Controlled Low-Strength
report called “Controlled Low Strength Materials (CLSM),”1 which has been
referenced widely. In 1999, the revised edition was published with the same title.
Materials (flowable fill).”8 The articles in this book represented the state of art
and practice of CLSM in the field and in the research laboratory at that time.
waste and reduce the cost. Currently, there are five ASTM testing standards
available for CLSM. Specific information and description on these and other test
8
progress. For instance, The University of Louisville and Purdue University have
have developed powerful air-entraining agents to produce CLSM with high air
content. The air content of the air-modified CLSM can be as high as 15 to 35%,
Cooperative Highway Research Program (NCHRP) under Project 24-12 and 24-
12(1) (“Controlled Low-Strength Material for Backfill, Utility Bedding, Void Fill,
and Bridge Approaches”) has culminated in two reports. One report summarized
the state-of-the-art and current practice related to CLSM in 1999 and the other
2.2 MATERIALS
key role. The materials used in CLSM are thus of crucial importance for this
technology. One significant benefit of CLSM is its ability to use a wide range of
9
aggregates (e.g., foundry sand), chemical admixtures, and other by-product
used in CLSM, ASTM Type I is most commonly used. The prevailing criteria are
the local availability and cost of cement, and Type II or Type I/II cements may be
unlikely. Type III portland cement has been successfully used in CLSM to
similar proportions, CLSM mixtures with Class C fly ash demonstrate higher
compressive strength than those with Class F fly ash. One example is Flash Fill,
which contains predominantly Class C fly ash, concrete sand and water.27 No
compacting characteristics make trench road repairs faster, easier, and more
and replacement of weak subgrade beneath footings. Flash fill can be hand-
2.2.2 Aggregate
ingredient. Because of the low cement content of CLSM, the cost of aggregate
materials actually dominates the total cost. It is common that CLSM producers
use whatever suitable aggregate materials that are locally available in the
mixtures, such as concrete sand, foundry sand, crushed glass, by-product fines,
dominant source of CLSM). Sand that barely meets the ASTM C 33 requirements
CLSM have mostly evolved using only sand as aggregate. CLSM in the
Pacific Northwest has developed differently in that many mixtures use gravels up
to 25-mm top size.28 The reasons for the use of gravel center around economy
11
aggregate is used, the lowest void content in the combined aggregates will be
gravel may be expected to be similar to those with sand only. Subsidence can be
Foundry sand, a by-product of the metal casting industry, has been studied
and used successfully in CLSM, and its use has increased in recent years.10,30-32
Foundry sand is a viable candidate for use in CLSM because of its lower cost,
waste sand for every ton of metal castings produced and shipped.33 The most
commonly used waste foundry sand in CLSM is “green sand,” a term applied
when the original uniform sand is treated with a bonding agent (usually clay) to
optimize the efficiency of the sand in the molding process. After molding is
completed, the sand is discarded, often in landfills, and in 1994 typical costs were
$22/tonne to $44/tonne.8
which covers in detail the use of foundry sand (and fly ash) in CLSM and
12
provides guidelines for its proper usage.34 The Environmental Protection Agency
(EPA) has also recognized foundry sand, along with fly ash, as suitable materials
for CLSM.35
There has been a perception that using foundry sand in CLSM may have a
detrimental environmental impact because heavy metals may leach from the
foundry sand. Ferrous foundry sands are more commonly used in CLSM because
there are concerns about the contents of heavy metal in non-ferrous foundry sand.
The EPA does not recommend using non-ferrous foundry sand in CLSM because
of the concerns over the potential leaching of phenols and heavy metals, such as
cadmium, lead, copper, nickel, and zinc.35 Bhat et al. reported the results of their
screen out potentially hazardous waste foundry sands. The test was performed
both on expressed CLSM pore solution and raw foundry sand. No firm
entraining foam), high-fines limestone screenings have been used as aggregate for
CLSM.36,37 It was estimated that more than 81.6 million tonnes of quarry by-
products were generated in 1993.36 It is difficult to find viable uses for by-
product fines, which account for 15- to 20-percent of aggregate produced, because
most construction and highway material specifications limit the aggregate finer
13
showed that 14 to 30 percent of air worked well for CLSM mixtures containing
recommendations.
abundance of fines (passing 300-ìm sieve), which may be well suited for use in
CLSM.
Bottom ash and fly ash are both by-product materials of coal combustion.
the hot gases. These particles descend on hoppers or conveyors, at the bottom of
the furnace, in a solid or partially molten condition. The particles gradually cool
down to form bottom ash. Bottom ash particles are typically porous and angular
ponds. In this process, bottom ash is passed through a crusher to reduce the size
site. The typical range of particle sizes falls between 75 microns and 25
millimeters. Bottom ash can be used in compacted fill when combined with fly
14
2.2.2.5 Other Aggregate Materials
local bottle manufacturers has been crushed to pass a 12.5-mm (½ in) sieve and
that the crushed glass could be handled with bare hands. Phosphogypsum is a by-
product of the production of phosphoric acid and has been shown to be a viable
showed that the toxic contents of the individual materials and final design pass
mixtures were well below the EPA leachate standards. Ground granulated blast
furnace slag (GGBFS) and bentonite have also been used in CLSM.59
2.2.3 Admixtures
Fly ash is a by-product of coal combustion and has been used in a wide
classified into two classes based on the calcium content and pozzolanic properties
(ASTM C 618). Fly ash normally produced from lignite or subituminous coals is
coal is classified as Class F fly ash. It has only pozzolanic properties and has no
calcium oxide.
15
Fly ash is used mostly in portland cement concrete, but its use in CLSM
this unused fly ash does not meet the specifications for use in portland cement
LOI of fly ash used in concrete are typically limited to six percent for Class C and
Class F fly ash according to ASTM C 618, whereas certain bituminous fly ashes
contents increase water demand in concrete and can significantly raise chemical
Despite the limitations placed on fly ash for conventional concrete, it has been
demonstrated that CLSM can be successfully produced using a wide variety of fly
ash types and sources, including high-carbon fly ash that is not permitted in
concrete.
Because ASTM C 618 Class F and Class C fly ash are commonly
There are numerous benefits of using fly ash in CLSM, including improved
material cost.
16
2.2.3.2 Chemical Admixtures
CLSM, and set accelerators have been used to a lesser extent to increase the speed
CLSM is often produced with relatively high air contents (e.g. 20-35 percent),
either through the use of potent, liquid air-entraining agents or with foaming
bleeding, decreased water and/or cement content, improved frost resistance, and
lower material cost. Also, high air contents may be used to limit long-term
2.2.4 Water
general rule, any water that is suitable for concrete will work well for CLSM.
widely adopted for CLSM. A group of studies have focused on CLSM mixture
17
In general, there are four typical types of CLSM used in current practice,
II (high fly ash) Cement, fly ash Water, fly ash Fly ash
Type I CLSM was the earliest type that appeared in the construction field.
The mixture proportions tend to vary with different applications, especially the
cement content, which is the main factor affecting the strength of CLSM. Type II
CLSM mixtures usually have cement contents of 4 to 5 percent of the total solid
Type II CLSM could achieve compressive strength values as high as 690 kPa at
28 days. To produce Type III CLSM, potent chemical admixtures are required to
entrain large volume of air voids. Type III CLSM is efficient in reducing fill load
because a density as low as 673 kg/m3 can be reached by using foaming agents.47
Type IV CLSM is patented and little is known about its proportions, but the
18
mixtures are typically high in Class C fly ash content. Type I and III CLSM
8 MPa, the vast majority of applications are designed for future excavatability and
have strengths of less than 1 to 1.4 MPa. For these mixtures, low cement contents
are used (e.g., 30 to 60 kg/m3 ), with or without fly ash. Certain mixtures use very
high fly ash contents (Type III), where fly ash effectively serves both as a binder
and as an aggregate. Tables 2.2 and 2.3 show what the FHWA considers to be
typical mixture proportions for low and high fly ash contents. Mixtures that do
not contain fly ash are often produced with high air contents (20 to 35%), where
the high air contents provide certain similar benefits as fly ash, including
improved flow, reduced segregation and bleeding, and decreased cement content.
Table 2.2: Typical mixture proportions for low fly ash content CLSM 42
Material Range (kg/m3) Typical mixture proportions (kg/m3)
Fly ash (6-14%)* 119 - 297 178
Cement 30 - 119 59
Sand 1483 - 1780 1542
Water 198 - 494 297
Total: 2076
* Higher calcium fly ash is used in lower amounts than low calcium fly ash
19
Table 2.3: Typical Mixture Proportions for High Fly Ash Content CLSM 42
Material Range (kg/m3) Typical mixture proportions (kg/m3)
Fly ash (95%) 949 - 1542 1234
Cement 47 – 74 62
Water 222 - 371 247
Total: 1543
ACI Committee 229 opposes the use of plastic fines such as clay because
CPF. The term CDF often appears in literature. It is a term created by Brewer to
represent Controlled Density Fill, which indicates a uniform density from the
trench’s bottom to its top. The density of the CDF material can be adjusted by
concrete. Most CLSM is batched at ready-mix plants and mixed in truck mixers.
Because of the amount of fines included and the fact that CLSM does not usually
contain coarse aggregate (which would assist in breaking up cement clumps and
near-full loads in ready mixed trucks. To address this potential problem, certain
producers hold back part of the mixing water and/or liquid air-entraining agents
for on-the-job-site additions rather than at the plant, thereby reducing the mixture
volume in the truck on route to the site. The high fluidity of CLSM affects
Table 2.4 summarizes test methods and critical CLSM properties that
affect its performance in backfill, utility bedding, void fill, bridge approach, and
other applications.25 Currently, there are five ASTM test methods specifically
designed for CLSM. The table lists various test methods that are currently being
used to assess CLSM or tests that may have the potential to serve as a suitable
methods. CLSM is a hybrid material (falling between soil and concrete), and as
such, soil and concrete test methods may be adopted for use. Focus should be
placed on the most important CLSM properties, their effects on performance, and
21
2.5.1 Fresh CLSM Properties
2.5.1.1 Flowability
One of the most important attributes of CLSM is its capacity to easily flow
into confined areas and obtain required strength and/or density without the need
of CLSM significantly reduce labor and increase construction speed and safety.
quality-control parameter.
22
Table 2.4: Currently used or potential test methods for CLSM (continued)
CLSM Test methods
property
Hardening AASHTO T 197, “Time of Setting of Concrete Mixtures by Penetration Resistance”
time
Sampling ASTM D 5971, “Standard Practice for Sampling Freshly Mixed Controlled Low-
Strength Material”
AASHTO T 141, “Sampling Fresh Concrete”
Segregation ASTM C 940, “Standard Test Method for Expansion and Bleeding of Freshly Mixed
and bleeding Grouts for Pre- Placed Aggregate Concrete in the Laboratory”
Drying AASHTO T 160, “Shrinkage of Portland Cement Concrete”
shrinkage
Consolidation ASTM D 2435, “Standard Test Method for One-Dimensional Consolidation Properties
of Soil”
ASTM D 5084, “Measurement of Hydraulic Conductivity of Saturated Porous Materials
using a Flexible Wall Permeameter”
Permeability AASHTO T 277, “Rapid Determination of the Chloride Permeability of Concrete”
ASTM D 4525, “Standard Test Method for Permeability of Rocks by Flowing Air”
Freezing-and- AASHTO T 161, “Standard Test Method for Resistance of Concrete to Freezing and
thawing Thawing”
resistance ASTM D 560, “Standard Test Methods for Freezing and Thawing of Compacted Soil-
Cement Mixtures”
ASTM D 5334, “Standard Test Method for Determination of Thermal Conductivity of
Soil and Soft Rock by Thermal Needle Probe Procedure”
Thermal ASTM D 5335, “Standard Test Method for Linear Coefficient of Thermal Expansion of
properties Rock Using Bonded Electrical Resistance Strain Gages”
ASTM D 4612, “Standard Practice for Calculating Thermal Diffusivity of Rocks”
ASTM G 1, “Standard Practice for Preparing, Cleaning, and Evaluating Corrosion Test
Specimens”
Corrosion of
ASTM G 51, “Standard Test Method for Measuring pH of Soil for Use in Corrosion
metals in Testing”
CLSM Modified ASTM G 59, “Standard Practice for Conducting Potentiodynamic Polarization
Resistance Measurements”
Modified ASTM G 109, “Standard Test Method for Determining the Effects of Chemical
Admixtures on the Corrosion of Embedded Steel in Concrete Exposed to Chloride
Environments”
ASTM G 57, “Standard Test Method for Field Measurement of Soil Resistivity Using
the Wenner Four-Electrode Method”
Leaching ASTM D 4874, “Standard Test Method for Leaching Solid Material in a Column
Apparatus”
EPA SW-846, Method 1311, "Toxicity Characteristic Leaching Procedure (TCLP)"
Utility ASTM D 543, “Standard Practices for Evaluating the Resistance of Plastics to Chemical
degradation Reagents”
Controlled Low Strength Material,” has become the most common test to measure
CLSM flow since its adoption by ASTM. The test method uses a 75-mm × 150-
mm cylinder, open at both ends, which is filled with CLSM and lifted between 2
23
to 4 seconds allowing the CLSM to slump and to increase in diameter. The final
flowable mixture. Some researchers have found that this test method is not as
sensitive to consistency as the conventional slump test.49 Water and fine material
may run out of the cylinder, leaving a sand cylinder standing. When the slump
cone was used, the sand mix tended to flow with the water and fines.
Concrete”, has been used to a lesser extent to measure the flowability of CLSM.
ASTM C 939, “Standard Test Method for Flow of Grout for Preplaced-
flow cone. Several state DOTs have specified this test method for CLSM, and the
5 seconds.1
24
gradation and shape, air content, water content, fly ash type and quantity, and
flowability, and the lifting force has been found to be particularly important.51
proposed the concept of a “flow curves” to study possible factors affecting the
flow of CLSM.10 He found that when proper contents of fly ash were used,
adequate flow was obtained. But when too much fly ash was used, more water
was required. The study was limited mainly to CLSM using foundry sand as
aggregate. Gray used rheology to study the flow behavior and regarded the grout
flowable fly ash, but the high cost of this admixture usually limits its use in
CLSM. 54
CLSM mixtures, there is the potential for excessive segregation and bleeding,
especially for Type I and II CLSM mixtures. Because all components of CLSM
are in suspension when being placed, it is natural that segregation and bleeding
tend to occur when mixtures consolidate. When water is the main flow-enhancing
25
ingredient and there is insufficient surface area (fines), segregation and bleeding
are likely to occur. Generally, the use of fly ash and air-entraining agents reduce
the water demand, thus minimizing the potential for segregation and excessive
bleeding. Using ASTM Test Method C 940, Hoopes found that the 30% air-
modified mix had 0% bleed water, while the non-air-modified CLSM mixture
yielded 2.4% bleed water at the top of the sample.24 Little has been reported on
CLSM mixtures, even with high air content, are generally pumpable with
careful proportioning.
time required for CLSM to develop from the plastic state to a hardened state, with
can be as short as one hour, but it generally takes three to five hours under normal
conditions.55 There are many factors affecting the hardening of CLSM. These
factors include binder type and content, aggregate type, drainage conditions,
hardening. ASTM C 403, “Standard Test Method for Time of Setting of Concrete
26
pocket penetrometer has also been used to study the hardening of CLSM.
of this method.10 Resistance values from the soil pocket penetrometer generally
correlate well with those of ASTM C 403. Penetration values are often specified
2.5.1.4 Subsidence
air. CLSM mixtures with significant bleed water may have considerable
generally occurs during CLSM placement, up until the fresh material sets.
strength results from the cementitious (and pozzolanic) reaction of cement and fly
the single most important parameter that determines the unconfined compressive
insensitivity of strength to water cement ratio variations at high levels (>3).20 The
type and amount of fly ash (if used) also have important effects on the strength
CLSM mixtures.
strengths of 0.35 to 0.69 MPa (50 to 100 psi) have similar bearing capacities as
well-compacted granular soil. CLSM strength values are often specified for
Material (CLSM) Test Cylinders,” was adopted in 1995. It has gained acceptance
with state agencies and commercial testing laboratories. The relatively low
strength of CLSM may cause potential problems, such as damaging the test
cylinder when stripping of CLSM cylinders from plastic molds. Because large-
required low load ranges, the results often show significant variations.26 The
28
capacities of many load frames used to test CLSM specimens in research
laboratories are typically in the range of 1,300 kN to 2,220 kN. To fail a 150-mm
about 18 kN is required. This load is only about one percent of the load frame
most state DOTs used concrete machines in the testing of CLSM compression
specimens.25 Researchers using a loading rate of 0.24 MPa/sec (35 psi/sec) failed
CLSM specimens in a few seconds and could not satisfy the minimum
2.5.2.2 Permeability
the use of CLSM as utility bedding for gas pipelines. Difficulties have been
water permeability values of CLSM are in the range of 10-4 to 10-5 cm/sec and
may achieve 10-7 cm/sec in mixtures with high strength or high fine contents.1
CLSM mixtures with 30-percent and 21-percent air contents have yielded
29
is often adopted for CLSM mixtures. In experimental work of Bhat et al, a
backpressure of around 550 kPa was applied and maintained.10 In his study, the
“B” value was checked. The “B” value is the ratio of pore water pressure change
∆u
B= ( 2.1)
∆σcon
confining pressure
the study.10
When soil specimens are saturated, water is assumed to fill up the voids in
the soil skeleton and water becomes connected. Although the bulk modulus water
changes with temperature and confining pressure, it is much higher than that of
soil. If a soil sample is saturated and under undrained conditions, the change in
confining pressure will be carried mainly by the water rather than by the soil, as
demonstrated by high “B” values. For a CLSM samples, the skeleton has a
relatively higher bulk modulus, resulting from the bonding strength of the
Only under high pressure can water enter certain small sized voids in the
hydration products. In Bhat’s study, a backpressure around 550 kPa was used and
it was higher than the strength of certain CLSM mixtures.10 The confining
pressure was not provided in that study. The pore structure of CLSM specimens
30
can be damaged by such a high pressure. A hydraulic gradient of 10 can seldom
permeability values from different sources. CLSM mixtures may best be treated
as rock and the change of “B” value should be checked with the change in
Both the triaxial shear test and direct shear test are commonly performed in soils
laboratories, using ASTM D 3080 for the direct shear test and USACE EM 1110-
2-1906 for the consolidated-drained triaxial shear test. The applicability of these
There have been only a few published studies focusing on the triaxial
angles from triaxial compression tests range from 20 to 40 degrees. There are
even fewer published results of direct shear testing studies.16,24,60 The possible
31
to evaluate the strength of subbase and subgrade materials. Through years of
approximating the working stress, is repeated applied until certain criteria are met.
This test was initially used to evaluate pavement subgrades. The resilient
used. There have been very few studies on the resilient modulus of CLSM
mixtures.
CLSM may reduce its volume as it releases its free water and entrapped
to some extent by crushing at the particle contact points, and elastic distortions.
Similarly, the initial settlement of CLSM mixture within several hours after
32
CLSM volume due to applied load is similar to secondary consolidation of soil.
Because of the relatively strong and rigid skeleton of CLSM after hardening,
Soil”, can be used to measure the consolidation of CLSM mixtures. The test can
estimate both the rate and total amount of settlement of CLSM used in various
smaller.
High water-cement ratio and high water content are two factors known to
cause excessive drying shrinkage in concrete. Although CLSM has higher water
cement ratios and higher water contents than concrete, limited research on drying
shrinkage has shown that CLSM may not exhibit much more shrinkage than
0.02 to 0.05 percent, which is similar to concrete with low drying shrinkage.1 The
Gandham et al found that the maximum shrinkage and expansion values of CLSM
were generally less than the acceptable limit established for concrete.41 Lucht
also showed that shrinkage of CLSM was minimal when a shrinkage-ring method
33
was used.13 A shrinkage ring is often used to measure the cracking of concrete
cast around a steel ring. This approach represents 100% restraint and can be used
studs at both ends of the specimens and measuring the length change. Careful
measurements are required. CLSM specimens could be damaged when using this
approach because of the lower strengths of CLSM. Thus, this approach may not
be appropriate for CLSM. Lucht have also used the shrinkage ring method, which
Moisture content and dry density are two major factors affecting thermal
particle shape and size, gradation characteristics, organic content and specific
gravity. Low density, air-modified CLSM (Type III) is particularly suitable for
pipe backfill because of its enhanced insulating properties. The choice of thermal
34
Hot-Plate Apparatus”. Institute of Electrical and Electronics Engineers (IEEE)
Standard 442-1981 has also been used to measure thermal resistivity of CLSM.27
2.5.2.7 Excavatability
compressive strength of 350 kPa or less can be excavated manually and CLSM
with the range of 690 to 1400 kPa can be removed by mechanical equipment,
2
such as backhoes. Engineers in Hamilton County, Ohio, specify a Removability
thirty-day unconfined compressive strength and the dry unit weight. If the
calculated value of the Removability Modulus is less than 1.0, the specific
development of CLSM. When foundry sand was used in CLSM, the unconfined
cement content was high, the strength of CLSM more than doubled from the age
of four to thirteen weeks.64 . Fly ash types and contents also affect the ultimate
strength of CLSM.
very difficult to remove manually even when its strength is low. On the contrary,
Type III CLSM with strengths of 2.1 MPa can be removed with a backhoe.1
35
There is no existing laboratory test method to estimate the actual
construction materials. There have been several laboratory and field studies on
the freezing and thawing resistance of CLSM.4,11,17,57,65 Type II CLSM broke into
pieces about the size of a hand when it was placed in a zone of total water
saturation and subjected to severe winter freezing and temperatures below –18
°C.4 Without water saturation, Type II CLSM appears to perform well under
suitable backfill material, Gress proposed that the top 5 to 150 mm of CLSM
heaving of pavement and trench after 24 hours of CLSM placement.65 Type III
CLSM and Type II CLSM with a high cement content (10%) are recommended
Hoopes, Type I CLSM specimens failed in only five cycles during the testing
process of ASTM D 560.24 Type III CLSM mixtures with 21% and 30% air
contents performed satisfactorily when they were evaluated with ASTM D 560.
have been found to be severe for CLSM. The low strength and the high
36
concrete. The high permeability allows water to enter CLSM samples easily and
ice lenses to form inside the specimens. As a result, the material is damaged
internally. This phenomenon is very similar to the frost heave of soil. The
internal hydraulic pressure generated during water freezing also damages the
research community.45,65 This method is much less severe than those for concrete
2.5.2.9 Erosion
diameter hole, lengthwise through the center of the cylindrical test sample. The
water has a pressure of approximately 275 kPa and the test is continued for 28
days. Then, the total mass of eroded material is determined. Specimens with
37
Krell found that flowable fly ash resisted erosion in both its plastic and
hardened states.4 A tank was constructed with water being pumped continuously
across a sample. This test was developed to compare materials and determine
how they affect water quality. CLSM was shown to be superior to the other
backfill materials. This property eliminates the need for silt curtains, containment
CLSM mixture with unconfined compressive strength of 3,448 kPa (500 psi) was
used to fill the voids of riprap in a drainage ditch. The voids in front of and
behind the concrete floodwall along the Mississippi River were filled with CLSM
having a compressive strength of 690 kPa (100 psi). Similar material was also
foundry sand and fly ash, introduced into CLSM mixtures, some environmental
concerns have been raised. Heavy-metal ions may leach into the environment
more easily from CLSM than from concrete because of the higher permeability of
CLSM. There are two approaches to assess this problem. One approach is to test
the leachate of source materials and the other is to observe the long-term field
38
The Extraction Procedure (EP) toxicity method, Toxicity Characteristic
and ASTM method are well known laboratory test procedures to determine the
cadmium, chromium, lead, mercury, selenium, and silver. Bhat used “Bio-assay”
including organics.10
Bhat also measured the pH values and toxicity of expressed pore solutions
from both source materials and CLSM mixtures. Only one of the eleven mixtures
evaluated was found to be toxic.10 Naik et al. and Gandham et al. studied the pore
solution leachate characteristics of different mixtures using the TCLP test (EPA
SW-846, Method 1311).39,41 They found that the toxic contents of the mixtures
2.5.2.11 Corrosion
Although corrosion of metal pipes buried in CLSM has not yet surfaced as
appear as a problem in the future. Folliard et al. have discussed the potential
obtained from laboratory testing results. Abelleira et al. found that CLSM
were buried in CLSM mixtures and were then submerged under corrosive water
39
containing sulfate,, chloride and bicarbonate ions.60 Ramme et al. tested the
resistivity of CLSM mixtures and found that the corrosion potential of CLSM (fly
ash slurry) was significantly less than that of typical soils used for trench
backfill.61 Krell considered that the alkalinity of fly ash and cement could inhibit
the rusting of iron and steel.4 The pH of extruded pore solution from CLSM
“Standard Test Method for Preparing, Cleaning, and Evaluation Corrosion Test
Specimens”, ASTM G 51, “Standard Test Method for Measuring pH of Soil for
Use in Corrosion Testing”, and modified ASTM G 59, “Standard Practice for
decreased with respect to soil, there may be a concern with galvanic corrosion.
Galvanic cells may form when a pipe contacts both soil and CLSM, two different
environments. If CLSM is only used as utility bedding and the upper portion of
the backfill is conventional compacted fill, a large galvanic cell will be formed
with a small anode and a large cathode. Even though these pipes are buried in
CLSM, a galvanic cell will form between sections in CLSM and sections in the
surrounding soils.
40
2.5.2.12 Utility Degradation
2.6 ECONOMICS
A recent survey showed that most state DOTs considered the cost of
CLSM as the largest barrier for widespread use.25 Prices as high as $130 per
cubic meter were reported. When filling the same trench, CLSM was slightly
more expensive than conventional granular backfill.66 At the same time, because
reduce the cost below that of conventional backfill. Bhat et. al stated that savings
in cost could be achieved when waste foundry sand was used as fine aggregate in
CLSM mixtures.10
2.7.1.1 Introduction
placed and compacted for the purpose of raising the grade of a roadway (or
railway) above the level of the existing surrounding ground surface. A fill refers
41
to a volume of earthen material that is placed and compacted for the purpose of
defined. These definitions are used in the discussion to follow in this dissertation.
Bridge Replacement
the cases of converting deteriorated bridges. Those bridges are not demolished
but changed into culverts or other structures by using CLSM to fully fill the
Backfill
Backfill refers to the use of CLSM in filling the openings formed during
the installation or repair of utility. The material is only utilized to cover the utility
Structural Fill
42
Insulation and isolation fill refers to the use of CLSM as insulation and/or
isolation material to take the advantage of its appropriate thermal and acoustic
properties.
Utility Bedding
bedding belongs to the category of structural fill, but is listed separately here to
Erosion Control
Void Filling
Void filling refers to the use of CLSM in the practice of filling voids from
significant loads.
43
2.7.1.2 Backfills
compacted properly in the required layer thickness may not achieve the
uniformity of CLSM. Folliard et al. found that the production rates when placing
CLSM
A B C
The following are case studies where CLSM has been used as backfill:
the city of Peoria, Illinois, tested CLSM for backfilling utility trenches in 1988.1
As a result, the city changed its backfilling procedure to require the use of CLSM
44
Case history 2: At one location in Iowa, a laid-up limestone block
67
retaining wall bulged out, indicating the initial stage of failure. A new “H”-pile
wall with wood planks was built in front of the old wall and flowable mortar was
Use in Highway and Bridge Applications”, it was found that unshrinkable backfill
(25 kg cement per cubic meters, conventional concrete aggregates, 160 to 200 mm
slump, usually air entrained, maximum 28 days compressive strength of 0.4 MPa)
satisfied the technical requirements for utility cut restorations and this material
backfill.68
Thruway (I-95) between Pelham Parkway and Bartow Avenue, Albany, New
strength of 0.69 MPa) was specified for backfilling certain utilities, which was
placed above a soft organic silt containing a strata of highly organic peat and
plant, 122,570 m3 of flowable fly ash were used in trench fill and 84,267 m3 was
used in the power house area. The project estimates indicated that more than $1
45
Case history 6: In one tilt-up construction project in Denver, it was
estimated that approximately two days of construction time were saved by using
CLSM to fill voids under bridge pier footings in the Robert Street Bridge in St.
Approximately 126 m3 (165 yd 3 ) of K-Krete was placed under two piers at a cost
of $107,000.6
Case history 8: CLSM provided the solution when severe spring flooding
washed all of the fill from above and around a multiplate steel arch pipe in
Hutchinson County, South Dakota. The pipe was repaired in nine days at a cost
Case history 9: In Des Moines County, Iowa, CLSM was used to fill a
void along a wooden culvert, where access for conventional repairs was not
available. 6
underslab void for the US Navy at Rough & Ready Island, Stockton, California.
The total cost was less than 20 percent of the amount authorized by the owner for
facilities was backfilled with CLSM in downtown Milwaukee during the summer
of 1984.71
46
Case history 12: CLSM was used in backfill excavations to remove oil-
bridges. This process involved installing metal culverts, pipes, or reinforced box
culverts under the bridges. Soil was used as forms at both ends of the bridge.
The space between the bottom of the bridge and the top of the culvert was
Case history 13: Iowa DOT engineers developed flowable mortar, which
includes sand, portland cement, fly ash and water, to convert 10 bridges into
culverts. US Route 30 was kept open to normal traffic most of the time. The cost
Case history 14: Six narrow bridges were replaced in a 9-mile length of U.
these bridges were modified by using flowable mortar with concrete pipe culverts.
culvert. One was replaced with a new bridge. The installation of pipe culvert
with CLSM cost only one-third of a reinforced-concrete box culvert and one-
47
2.7.1.4 Structural Fill
structural fill, the compressive strength of CLSM can vary from 0.69 MPa to 8.3
MPa (100 psi to 1200 psi) depending upon the application. The structural
coefficient of CLSM layer is estimated to range from 0.16 to 0.28 for compressive
strengths from 2.8 MPa to 8.3 MPa (400 psi to 1,200 psi).1
the embankment,
approach construction because use of CLSM may eliminate causes No. 3 and No.
4. Possibly, if CLSM is used for the bridge approach embankment, cause No. 2
can be avoided.
Case history 15: Near Boone, Iowa, 2,145 cubic meters of CLSM were
used to provide proper bearing capacity for the footing of a grain elevator that was
48
Case history 16: In June 1979, a small peninsula about 48.8 meters long
and 9.1 meters wide was constructed of flowable fly ash at Edison's Monroe
Power Plant.3
Case history 17: CLSM was used to fill between a spread footing and the
also used for updating the manufacturing plant without interrupting the
manufacturing process. The mix was line pumped over 305 m during some
placements.74
CLSM was used to stabilize the soil surrounding the sheet-piled shaft that would
be used to launch a micro tunnel boring machine. The use of this fly-ash-based
CLSM greatly improved the stability of the soils and increased the safety of the
shaft during the launch. The use of CLSM provided a saving of more than
Case history 19: In July 1995, a contractor placed more than 306 cubic
meters (400 cubic yards) of CLSM as backfill for the abutments of a bridge
located along Colorado State Highway 135 near Crested Butte, Colorado.
Case history 20: In the Muroran Bay Bridge project, a man-made island of
soft seabed. From October 1988, 53,600 m3 of slurry, containing 46.1 x 106 kg of
fly ash, 22.2 x 106 kg of volcanic ash, and 2.6 x 106 kg of cement, was filled into a
49
steel-pile cofferdam during a three-month period. Observations included 1)
reduced lateral earth pressure on the cofferdam walls, resulting in smaller bending
moments and lateral displacements; and 2) higher strength of the island for
supplied from the hydration of the slurry kept the fill temperature above 10 ºC and
Case history 21: Pockets of soft material of an underlying soil layer were
excavated and then backfilled with flowable fly ash (Class F ash). The next
morning, the contractor continued his work on building the six-story apartment
building.4
Case history 23: The Oklahoma DOT and Oklahoma State University
research program involved construction of three new bridges on U.S. 177 north of
Stillwater, Oklahoma. They found that the use of CLSM as construction material
reduce the potential for “the bump at the end of the bridge”.
Case history 24: Flowable fly ash was used to save Detroit's Northeast
78
Raw Water Tunnel, Detroit, Michigan, 1985-6. CLSM was used to support the
50
2.7.1.5 Utility Bedding
utilities. Improper bedding preparation can decrease the desired pipe carrying
capacity or cause excessive backfill settlement. There are several methods to test
the inherent strength of pipe sections. These methods include the two-edge
bearing test, the three-edge bearing test, the sand bearing test, and the Minnesota
bearing test. The three-edge bearing method is the simplest to perform. The ratio
of the strength of a pipe under any stated condition of loading, whether in the field
or in the laboratory, to its strength by the three-edge bearing test is called the load
factor or strength ratio for the stated condition and is designated as Lf. There are
Impermissible bedding (Class D), Figure 2.2, is the case where little or no
effort is taken to shape the foundation to fit the lower part of the conduit exterior
or to refill all spaces under and around the conduit with granular materials, at least
partially compacted.
51
Figure 2.2: Class D: impermissible beddings. load factor = 1.1. (From Soil
Engineering, 4th Edition, M. G. Spangler and R. L. Handy, Harper &
Row, Publisher, New York, 1982)
Ordinary bedding (Class C), shown in Figure 2.3, is the case where the
the lower part of the conduit exterior with reasonable closeness for a width of at
least 50% of the conduit breadth. The remainder of the conduit is surrounded to a
height of at least 0.15 m above its top by granular materials that are shovel-placed
and shovel-tamped to completely fill all spaces under and adjacent to the conduit.
All this work must be done under the general direction of a competent engineer.
52
H > 7.3 m-13 mm per 305 mm of H
Figure 2.3: Class C ordinary beddings. load factor = 1.5 (From Soil
Engineering, 4th Edition, M. G. Spangler and R. L. Handy, Harper
& Row, Publisher, New York, 1982)
First class bedding (Class B), shown in Figure 2.4, is the case where the
is carefully pre-shaped by means of a template to fit the lower part of the conduit
exterior for a width of at least 60% of the conduit breadth. The remainder of the
granular materials that are carefully placed to completely fill all spaces under and
adjacent to the conduit. The granular materials are thoroughly tamped on each
side and under the conduit as far as practicable in layers not exceeding 0.15 m in
thickness.
53
Figure 2.4: Class B first class beddings. load factor for ditch conduits =1.9 (From
Soil Engineering, 4th Edition, M. G. Spangler and R. L. Handy,
Harper & Row, Publisher, New York, 1982)
Concrete-cradle bedding (Class A), Figure 2.5, is the case where the lower
thickness under the lowest part of the conduit and extending upward on each side
of the conduit for a distance not greater than one-fourth the outside diameter.
When CLSM is used for utility bedding, there is no or only minimal preparation
flowability and strength characteristics of CLSM may make this material a cost-
concrete pipe along the entire Canadian River Aqueduct Project, which stretches
515 km (322 miles) from Amarillo to Lubbock, Texas. It was estimated that soil-
54
cement slurry reduced bedding costs by 40 percent. Production increased from
122 linear meters to 305 linear meters of pipe placed per shift.2
Figure 2.5: Class A: concrete-cradle bedding. load factor 2.25 to 3.4. (From
Soil Engineering, 4th Edition, M. G. Spangler and R. L. Handy,
Harper & Row, Publisher, New York, 1982)
installed a pipe arch culvert (381 cm × 241 cm, 2745 cm long with only 840 mm
of CLSM top cover) in Paulding County, 1989 and a long span arch culvert (10.5
m long) in Cuyahoga County in 1990. He found that: 1): the minimum wall
thickness required for the structural capacity of flexible culverts, pipe arches and
long span arches, could be reduced when using CLSM-CDF as the backfill
55
development; 3) more triaxial shear information is needed for wider CLSM-CDF
strength ranges, (0.35 to 10.4 MPa (50 to 1500 psi)) to help develop more
conventionally placed backfill exerts less vertical and horizontal load on a culvert
than does a CLSM-CDF backfill, but after CLSM-CDF's cohesion buildup the
reverse is true.20
downtown Seattle.74
Airport, CLSM was used to backfill a 1.5 m diameter rigid concrete pipe in a
narrow 3.1 m trench, where the material was placed up to the pipe springline. The
culvert pipe has been inspected three times since installation and it is performing
well.80
Case history 29: CLSM was specified as bedding material for a 32,000-m
between 1991 and 1993). The flowable backfill was designed to completely
surround the pipe and extend a minimum of 152 mm above the top of the pipe.69
Case history 30: A roadway over a myriad of petroleum pipes was needed
Toledo, Ohio. A rectangular area was excavated and backfilled with a monolithic
mass of CLSM to totally envelope the pipes. After a three-day curing period, a
56
100-mm granular layer was placed over the area as a wearing course for the
roadway. 6
abutment movement, bridge approach settlement, and other damage. The use of
erosion control mat to avoid further erosion of the Iowa River bank.6
Case history 32: Riprap was put into the erosion cavity and the spaces
between the rocks were filled with flowable mortar in Iowa counties.7
inside and/or between objects. In practice, CLSM is used widely to fill tunnel
Case history 33: CLSM was used to fill an abandoned sewer and a
deserted tunnel that passed under the Menomonee River in downtown Milwaukee.
The material was reported to flow 72 meters and 92 meters, respectively. Only
four hours were needed to fill an exploratory shaft 36.6 m deep, 3.7 m in diameter
with a 9.2 m long branch tunnel. A total of 602 cubic meters of CLSM were
cast.1
57
Case history 34: To verify the practicality of the hydraulic backfilling
slurry into an abandoned mine was completed at the Fairfax mine in Preston
practical solution to ash disposal with collateral benefits in preventing Acid Mine
Case history 35: CLSM was chosen to fill the an abondoned pipeline
Case history 36: Flowable fly ash was used to fill abandoned gasoline
Case history 37: 1.CLSM was used to fill the a void under a 30.5-meter
Iowa at Iowa City, old stream tunnels were filled with CLSM. 3. Many
Case history 38: 1. Wisconsin Electric filled two obsolete steam service
diameter by 88.5 m long (1.8 x 88m) and the other was a 1.5 m high by 1.2 m
Case history 39: In LaSalle, Illinois, 306 m3 of CLSM was pumped to fill
58
bus station in a tunnel was filled with CLSM. CLSM has also been used to fill
Case history 40: CLSM was used to fill old basements full of loose
Case history 41: Iowa DOT filled two 7.6-m3 and one 3.8-m3 abandoned
fuel tanks with approximately 20 m3 of CLSM near Ames in 1987. The project
was completed at a cost of $1,140. The cost of removing the tanks was estimated
at approximately $8,000.6
Case history 42: With CLSM, the City of Des Moines repaired the grade
under the street pavement and filled an undermined culvert for Pope John Paul's
visit.7
2.8.1 Introduction
Although CLSM has gained popularity in recent years, there are several
obstacles for which some engineers are reluctant to specify it for construction
engineers are not familiar with this material. Education is extremely important
The following are some important issues that deserve additional emphasis
To date, there are only five ASTM test methods and no AASHTO method
Material (CLSM)”;
60
Obviously, there is a lack of standard test methods for CLSM. Only a few
CLSM properties are routinely measured by state DOTs and commercial testing
laboratories, and even these properties are measured with different test methods.25
used in different locations throughout the United States. These materials include
fly ash of different carbon contents, bottom ash, crushed concrete, high fines, and
CLSM users about how to best use and specify these materials, or how to address
specific issues that may arise from their inclusion in CLSM.25 Procedures or test
gain for CLSM because the majority of CLSM applications may eventually
61
diggibility of CLSM mixtures placed in field. Density and aggregate type also
CLSM containing fly ash demonstrates more long-term strength gain because of
the pozzolanic and/or cementitious reaction of fly ash. Because CLSM is usually
made with locally available materials and the properties of these materials may
construction material. There have not yet been significant durability problems
identified in field applications.25 But this may be due to the relatively short
period of time that CLSM has been used. Issues of particular interest include
Fly ash, foundry sand, and other by-product materials are often used in
CLSM. These materials may contain heavy metals and other potentially harmful
materials that may leach into the surrounding environments. The leaching
62
develop suitable test approaches and to provide specifications to warrant the safe
internal friction angle, consolidation and settlement, and bedding factors for
practice.
Soil test methods that may be applied to CLSM should be simple methods,
63
Chapter 3: Materials and Mixture Proportions
3.1 INTRODUCTION
portland cement, and water. Chemical admixtures are sometimes used to obtain
CLSM strength.
product materials. Foundry sand and bottom ash are two of the most common by-
product materials used in CLSM, and both were included in this study. A natural
sand, typically used in concrete, was also used throughout the investigation.
64
3.2.2 Aggregate Gradation
with ASTM C 136-96a, “Standard Test Method for Sieve Analysis of Fine and
bottom ash, only materials passing the 6.35 mm sieve were used in this study.
The gradations of the three fine aggregates are shown in Table 3.1 and Figure 3.1.
For comparison, the upper and lower limits set by ASTM C 33, “Standard
Specification for Concrete Aggregates,” are also shown in the graph. Although
the river sand met the specification of ASTM C 33-93, it approached the coarse
limit of the gradation band. The bottom ash was found to be slightly coarser and
65
100
90
80
70
Percent passing (%)
60
50
40 River sand
30 Bottom ash
Foundry sand
20 ASTM C33 (lower limit)
ASTM C33 (upper limit)
10
0
0.1 1 10
Sieve opening size (mm)
The absorption capacity and specific gravity of each aggregate were found
in accordance with ASTM C 128-97, “Standard Test Method for Specific Gravity
gravity and absorption capacity of each aggregate were determined. The results
are shown in Table 3.2. The Fineness Modulus (FM) of the river sand was
66
concrete technology to assess the fineness of an aggregate. The cumulative
percentages retained sieves: No. 100 (150 ìm), No. 50 (300 ìm), No. 30 (600
ìm), No. 16 (1.18 mm), No. 8 (2.36 mm), No. 4 (4.75 mm), and larger –
increasing in the ratio of 2 to 1, are calculated and divided by 100. The value
obtain is called FM. The absorption capacities of foundry sand and bottom ash
(SSD) state. The following equation was used to perform moisture correction:
Wbatch=WSSD(1+p)/(1+a) (3.1)
analytical laboratory and the results are shown in Table 3.3. The Material Safety
Data Sheet (MSDS) from the producer indicated that this material might comprise
The CLSM mixtures in this study included three types of fly ash, ASTM C
618 Class F, Class C and high-carbon fly ashes. Fly ash with less than 70% SiO 2
and Al2 O3 , resulting in 15-35%CaO is classified as Class C fly ash. On the other
68
hand, fly ash with more than 70% SiO 2 and Al2 O3 , and lower CaO contents, is
Specification for Coal Fly Ash and Raw or Calcined Natural Pozzolan for Use as
six percent for Class C and Class F fly ashes. The LOI of each fly ash was
Sampling and Testing Fly ash or Natural Pozzolans for Use as a Mineral
reflect higher carbon contents. The high-carbon fly ash used in this study would
not typically be allowed for use in conventional concrete. High-carbon fly ash is
concrete, but high-carbon fly ash has been allowed in CLSM mixtures by state
DOTs according a survey. 25 The LOI results of the fly ashes are shown in Table
3.2.
Chemical analyses of the four fly ash used in this study were performed by
a commercial analytical laboratories, and the results are shown in Table 3.4.
Class F fly ash had high proportions of silica and alumina (a total of 85%) where
Class C had only 55% of these oxide. As a result, the Class C fly ash had a much
higher calcium oxide content. Because of the presence of free lime, some Class C
fly ashes rapidly set and harden when being mixed with water, and are considered
to be true hydraulic cements. The high-carbon fly ash had a higher calcium oxide
69
content than the Class F fly ash, but this by itself does not necessarily indicate that
high-carbon fly ash has higher reactivity than the Class F fly ash. The fly ash
used in water permeability and triaxial shear tests was also analyzed.
compounds. The average specific gravity of the agent is 1.0 and the average total
solids are 95%. This admixture is capable of entraining stable air contents
70
3.4.2 Accelerating Agent
was utilized in this study for a small number of mixtures. This agent belongs to
ASTM C 494 Type C, and the average specific gravity is 1.34. It has a solid
contents of 41.5%.
3.5 CEMENT
ASTM Type I portland cement was used in this study. ASTM Type II
cement was used only in the study of water permeability and triaxail shear tests.
71
3.7 CLSM MIXTURE PROPORTIONS
This section summarizes the mixture proportions used in this study. There
were two distinct stages of mixing for this study. In the first stage, a total of 38
mixtures were cast and tested. In the second stage, additional mixtures were cast
properties (and corresponding test methods) were prioritized and classified into
the following three groups, depending on the research objectives of this study:
The most common types of CLSM materials and mixtures were selected
for the initial laboratory study. For each of these mixtures, the types and amounts
of cement, fly ash, and aggregates were selected prior to mixing, and the water
content of each mixture was then adjusted to achieve a flow of about 200 mm, as
and the range of materials chosen for the initial 38 mixtures included:
72
CLSM (with fine aggregates)
Type I portland cement: 1 type, 2 levels - 30 kg/m3 , 60 kg/m3
Fly ash: 3 types, 3 levels - 0 kg/m3 , 180 kg/m3 , 360 kg/m3
Fine aggregate: 2 types, 1 level - 1500 kg/m3
Air content: 3 levels - Entrapped air only, 15-20% air, 25-30% air
(Air-entraining agents will not be used for CLSM containing fly
ash)
defined, and specifying the desired dosage range for the key components, a
statistical program (ECHIP) was used to generate portions for a test matrix of the
trials. In other words, rather than producing CLSM with every possible
test matrix was produced that could be used to predict test results across the entire
compare the results of one test to another or to determine the effects of individual
or combined variables on test results. The program is also designed to assess the
test matrix. The repeatability of a test method greatly affects its acceptance as a
73
Although an air-entraining dosage 13 times the manufacture’s
recommended dosage was used in the CLSM mixtures containing foundry sand,
less than 3% of air was introduced in these mixtures. This may be attributed to
the presence of coals and organic materials in the foundry sand. Thus, foundry
sand was not included in air-entrained CLSM mixtures. For the air-entrained
mixtures, a full test matrix was used. Seven CLSM mixtures other than the above
mentioned mixtures were chosen for the stage 1 study. As such, a total of 38
were studied in further detail. For each aspect studied, specially selected CLSM
mixture series were generated using the statistical software, one for non-air-
entrained CLSM (with fly ash), and one for air-entrained CLSM (without fly ash).
The non-air-entrained mixtures are shown in Table 3.6. Mixture proportions for
numbering, two new mixtures, mixture #20 and #21, were substituted for the
74
Table 3.6: Non-air entrained CLSM mixture proportions
Mixture* Type I cement Fly ash type Fly ash content Fine aggregate
content type**
(kg/m3 ) (kg/m3 )
1 30 Class C 180 River sand
2 60 Class C 180 River sand
1r 30 Class C 180 River sand
9 60 Class F 360 Foundry sand
15 30 Class C 360 Foundry sand
4 30 Class F 360 River sand
5 60 Class F 180 Bottom ash
3 60 Class C 360 Bottom ash
8 60 High carbon 180 Foundry sand
10 30 High carbon 180 Bottom ash
12 30 Class C 360 Bottom ash
6 30 High carbon 360 River sand
3r 60 Class C 360 Bottom ash
4r 30 Class F 360 River sand
7 30 Class F 180 Foundry sand
11 60 High carbon 360 Bottom ash
14 60 Class F 360 River sand
13 60 Class C 360 Foundry sand
5r 60 Class F 180 Bottom ash
2r 60 Class C 180 River sand
* ECHIP randomizes order of mixtures and provides for duplicates.
** Fine aggregate content was held constant at 1500 kg/m3.
Because of the modifications to the initially planned matrix for the air-
entrained mixtures, the statistical program for designing the air-entrained mixtures
was not used. Mixtures covering all of the selected variables were cast and
evaluated. That is, two cement contents (30 kg/m3 and 60 kg/m3 ), two target air
contents (15-20% and 25-30%), and two aggregate types (concrete sand and
bottom ash) were used in all combinations to create a total of eight mixtures.
From these eight mixtures, three were selected for replicate mixtures, bringing the
75
Table 3.7: Air-entrained CLSM mixture proportions
Mix # Cement content Air content Fine aggregate
(kg/m3 ) type**
23 60 25-30% Bottom Ash
17 30 15-20% Bottom Ash
19 30 25-30% Bottom Ash
18 60 15-20% Concrete Sand
16 30 15-20% Concrete Sand
22 60 25-30% Concrete Sand
20* 60 15-20% Bottom Ash
22 60 25-30% Concrete Sand
21* 30 25-30% Concrete Sand
16 30 15-20% Concrete Sand
20* 60 15-20% Bottom Ash
* These mixes were substituted for the originally proposed mixtures because of difficulty in
entraining air in mixtures containing foundry sand. The originally proposed mixtures
containing foundry sand were still cast, but without entrained air. Table 3.8 shows these
mixtures (Mix # 29 and #30).
** Fine aggregate content was held constant at 1500 kg/m3.
specific mixture types that may be of interest. The mixtures represent typical
CLSM paste mixtures (i.e., 5% cement, 95% fly ash) and also include mixtures
Mixtures #24 and #25 were CLSM mixtures without fine aggregates. The
effect of acceleration admixtures was studied on mixtures #26, #27 and #28.
Because of the high fine content of foundry sand, it was possible to design CLSM
mixtures without fly ash or high air contents. Two mixtures, #29 and #30, were
76
Table 3.8: Selected CLSM mixtures included in Stage One
Type I cement Fly ash
Mixture content Fly ash content Air content Fine aggregate
# (kg/m3 ) type (kg/m3 ) type*
24 60 Class F 1200 Entrapped air only None
25 60 High 1200 Entrapped air only None
carbon
26** 60 None 0 25-30% River sand
27** 60 Class F 1200 Entrapped air only None
28** 60 Class F 180 Entrapped air only River sand
29 60 None 0 Entrapped air only Foundry sand
30 30 None 0 Entrapped air only Foundry sand
* For mixtures 26, 28, 29, and 30, the fine aggregate content was held constant at 1500 kg/m3
** Mixtures contain accelerating admixture
After casting and testing the initially proposed mixtures (as summarized in
Table 3.6 through 3.8), additional mixtures were cast to further investigate or
refine selected test methods or to study selected CLSM properties in more detail.
The mixtures were based in most cases on previously cast mixtures (from the
original 38). However, there were several variations used to evaluate certain
properties or tests.
commonly measured (and often the only hardened property measured), particular
emphasis was placed on developing a refined test that is more reliable and
reproducible. Issues such as load rate, curing condition, temperature effects, and
77
Nine sets of additional mixtures were cast and will be referred to
shown below.
Mixture Number of
Series Description Mixtures
A Effects of load rate on compressive strength (Table 3.9) 7
B Effects of curing and air-drying on compressive strength (Table 3.10) 2
C Long-term strength gain and excavatability (Table 3.11) 9
D Freezing and thawing resistance (Table 3.12) 11
E Alternative capping materials for compression cylinders (Table 3.13) 8
F Effects of drainage on compression cylinders (Table 3.14) 8
G Effects of storage conditions on compression cylinders (Table 3.15) 10
H Effects of temperature and humidity on compressive strength (Table 3.16) 6
I Permeability and triaxial shear strength (Table 3.17) 6
Tables 3.9 through 3.17 show the mixture proportions for the additional
aggregate content was held constant at 1500 kg/m3 . These tables also contain
78
Table 3.9: CLSM mixture proportions for load-rate study
Mixture Cement Fly ash Fly ash Fine Water Flow Air Unit
3
content type content aggregate type (kg/m ) (mm) content weight
3 3
(kg/m ) (kg/m ) (%) (kg/m3 )
A-1 60 None None River sand 156 175 25.0 1739
A-2 60 Class F 360 Foundry sand 520 200 1.9 1755
A-3 60 Class F 1200 None 486 213 1.6 1620
A-4 30 Class C 180 River sand 265 330 1.7 2161
A-5 30 Class C 180 River sand 213 216 Not 2226
measured
A-6 60 Class F 1200 None 501 216 Not 1635
measured
A-7 60 None None River sand 156 165 24.5 1740
79
Table 3.12: CLSM mixtures used for freezing and thawing study
Mixture Cement Sand Fly ash Water Flow Air content Unit weight
(kg/m3 ) (1500 kg/m3 ) (180kg/m3 ) (kg/m3 ) (mm) (%) (kg/m3 )
D-1 30 River None 119 180 27% 1630
D-2 30 River Class F 205 200 Entrapped 2196
D-3 30 River High-carbon 256 229 Entrapped 2078
D-4 30 River Class C 200 216 Entrapped 1980
D-7 30 Foundry Class F 425 238 Entrapped 1835
D-6 30 Foundry High-carbon 481 229 Entrapped 1757
D-5 30 Foundry Class C 399 200 Entrapped 1800
D-8 30 Bottom ash Class F 357 200 Entrapped 1870
D-9 30 Bottom ash High-carbon 407 200 Entrapped 1733
D-10 30 Bottom ash Class C 282 200 Entrapped 1896
D-11 45 River None 96 152 30% 1569
80
Table 3.15: CLSM mixture proportions for cylinder storage-condition study
Mixture Cement Sand Fly ash Fly ash Aggregate Water Flow Air content
(kg/m3 ) (kg/m3 ) (kg/m3 ) type type (kg/m3 ) (mm) (%)
G-1 60 - 1140 Class F - 485 200 Entrapped
G-2 0 2000 275 Class C River sand 252 200 Entrapped
G-3 30 1500 - - River sand 112 187 29
G-4 15 1500 180 Class F River sand 177 200 Entrapped
G-5 30 1500 180 Class F River sand 175 200 Entrapped
G-6 45 1500 - - River sand 103 190 30
G-7 30 1500 180 Class F Foundry sand 349 216 Entrapped
G-8 30 1500 180 Class C Foundry sand 352 190 Entrapped
G-9 30 1500 180 Class F Bottom ash 424 140 Entrapped
G-10 30 1500 180 Class C Bottom ash 367 152 Entrapped
Table 3.17: CLSM mixture proportions for triaxial shear strength and water
permeability
Mixture Cement Fly ash Sand Water Flow Air Fresh unit Dry unit Moisture
(kg/m3 ) (kg/m3 (kg/m3 (kg/m3 ) (mm) content weight weight content
) ) (%) (kg/m3 ) (kg/m3 ) (%)
81
Chapter 4: Test Methods and Procedures
4.1 INTRODUCTION
Table 4.2 provides a summary of all the tests ultimately performed in this study,
that were later added to the program (using the mixtures in Tables 3.9 to 3.17).
This section provides information on the specific test methods used in this study,
into three categories based on the properties that they are intended to measure:
strength of CLSM.
82
Table 4.1: Summary of initial test program for this study
CLSM Test Specimen Test ages
property method(s) details
IMPORTANT PROPERTIES - TESTING FOR ALL MIXTURES
Flow ASTM D 6103 75 mm x 150 mm After mixing
Setting/Hardening Accordingly (function of early
ASTM C 403 150 mm x 150 mm
time stiffness)
Compressive 75 mm x 150 mm & 3 cylinders each @ 3, 7, 28,
ASTM D 4832
strength 150 mm x 300 mm and 91 days
CLSM -sand ASTM G 1 75 mm x 150 mm 182 days
comparative study
ASTM G 1 &
Galvanic cells 100 mm x 200 mm 182 days and monthly
ASTM G 109
50 mm diameter x 100
pH ASTM G 51 7, 28, and 182 days
mm height
100 mm x 150 mm x
Resistivity ASTM G 57 28 and 182 days
50 mm prism
POTENTIALLY IMPORTANT PROPERTIES - TESTING FOR LIMITED MIXTURES
Segregation and No standard, see test Every 15 min. for first hour,
150 mm x 150 mm
bleeding procedure then hourly
Subsidence No standard 610 mm x 100 mm After hardening & 48 hours
Triaxial shear USACE EM 1110- 73 mm diameter, 150 3 samples each at 7, 28, and
strength 2-1906 mm height 91 days
California Bearing 152 mm diameter, 178
AASHTO T 193 1 sample at 28 days
Ratio (CBR) mm height
73 mm diameter, 150
Resilient modulus AASHTO T 292 1 sample at 28 days
mm height
73 mm diameter, 150 2 samples each at 28 and 91
Water permeability ASTM D 5084
mm height days
87.5 mm × 26.3 mm × 1 sample each daily for 7 days
Drying shrinkage No standard
1000 mm angle & weekly
450 mm x 450mm x
Excavatability No standard 91 days (and later ages)
300 mm
Chloride diffusion ASTM C 1152 150 mm x 300 mm 30 and 91 days
Freezing and 102 mm diameter, 125
ASTM D 560 28 days
thawing mm height
EPA SW-846,
Leaching/environ-
Method 1311
mental impact
(TCLP)
LESS IMPORTANT PROPERTIES - NO TESTING
136 mm diameter, 20-
Direct shear strength ASTM D 3080 None
24 mm height
Thermal ASTM D 5334 76 mm x 203 mm None
conductivity
Air/gas permeability ASTM D 4525 73 mm x 150 mm None
Consolidation ASTM D 2435 63 mm x 200 mm None
83
Table 4.2: Overview of all properties and tests studied
Initial study (see Table
CLSM property Additional studies
4.1)
Flow ASTM D 6103
Soil pocket penetrometer
Setting/Hardening time ASTM C 403
Pocket torvane
Sample size effect
Small vs large machine
Loading rate
Effect of drying sample for 4
to 8 hours before testing
Compressive strength ASTM D 4832 Alternative capping materials
Effects of drainage
Effects of curing methods
Effects of curing temperature
and relative humidity
CLSM-sand comparative study ASTM G 1
ASTM G 1 & ASTM G
Galvanic cells
109
pH ASTM G 51
Resistivity ASTM G 57
Segregation and Bleeding ASTM C 940
Subsidence No Standard
ASACE EM 1110-2-
Triaxial shear strength
1906
California Bearing Ratio (CBR) AASHTO T 193
Resilient modulus AASHTO T 292
Water permeability ASTM D 5084
Drying shrinkage No standard
Excavatability No standard Splitting tensile strength
Chloride diffusion ASTM C 1152
Freezing and thawing ASTM D 560 Effects on permeability
Direct shear strength None
Thermal conductivity None
Air/gas permeability None
Consolidation None
Chemical and toxicity
Leaching/environmental impact None analyses of constituent
materials
84
4.2 FRESH CLSM TEST METHODS
4.2.1 Mixing Procedure and Measurement of Flow, Air Content, and Unit
Weight
For each of the 38 initial mixtures studied trial mixing was performed to
determine the approximate water demand needed for the target 200 to 250 mm
flow. To measure flow in trial mixtures, as well as actual test mixtures, ASTM D
6103-97 was followed. This method uses a 75-mm x 150-mm plastic cylinder,
with the bottom removed. The cylinder is raised, and the flow diameter is
After determining the target water content, the actual mixtures were cast
and test samples were prepared. For all of the 38 mixtures, the constituent
materials were batched using a Sartoris 38-kg capacity scale (0.5-gram accuracy).
For the smaller mix volumes, a 0.028-m3 drum mixer was used. For the larger
mix volumes (needed for measuring additional properties on the six selected
For non-air-entrained mixtures, the dry materials (e.g., sand, fly ash, and
cement) were first mixed with approximately half of the expected mixing water
(based on trial mixing) for three minutes, followed by a two-minute rest period.
After the rest period, the remainder of the batched water was added, followed by
measurements were taken. In most cases, as a result of the trial mixing, the target
flow of 200 to 250 mm was obtained. If the flow was less than desired, small
85
amounts of water were added, followed by an additional minute of mixing, to
obtain the target flow. Rarely was the flow greater than the desired range. Care
slump) was first obtained in the mixer, followed by the addition of the AEA. This
process was necessary because of the potency of the AEA. If the AEA was added
to an already fluid mixture, the flow would far exceed the desired range, and the
6103, and the air content was measured using slightly modified ASTM C 231
(pressure method). If the flow and air content met the desired ranges, test samples
were then cast. If the air content was found to be lower than desired, additional
AEA was added (within the specified dosage of the admixture). It can often be
challenging to obtain both the desired flow and air content. When CLSM is
relatively dry, adding AEA will increase both air content and flow. Conversely,
adding water to a mixture that already has the desired air content but has a lower
than desired flow may actually decrease the air content (the flow may actually
gained in trial mixing allowed converging on the target air content and flow in the
test mixtures much more efficiently. After obtaining the desired flow and air
content, the unit weight was measured using ASTM C 231. Upon successfully
meeting the target air content and flow, various test samples were cast and tested.
86
4.2.2 Setting and Hardening
The setting time of CLSM was measured using two methods, needle
typical soil penetrometer is shown in Figure 4.1. Fresh CLSM was placed in two
standard containers (150 mm x 150 mm), one container for each test method.
Before each measurement, the bleed water was removed and weighed. The
cumulative weight of the bleed water removed throughout the test was used to
calculate the sample bleeding (bleed water mass divided by total sample mass).
It should be noted that the depth of penetration for the needle penetrometer
was approximately 25 mm, compared with only 6.4 mm for the soil penetrometer.
the influence of bleed water and obtain a more accurate estimate of early
stiffness/strength gain.
87
Figure 4.1: A typical soil pocket penetrometer
(VS) tester, as shown in Figure 4.2, was used. This device is often used by
in the field. It was utilized in this study to assess the setting/hardening of fresh
88
Figure 4.2: Pocket vane shear tester used in this program
4.2.3 Segregation
was used for this purpose. Each cylindrical section had a diameter of 100 mm and
mm. After the samples had set, steel separating plates were inserted at the
junctions between the cylinder sections, thus yielding three separate samples
(upper, middle, and lower). Each sample was then wet sieved, using the #4, #8,
#16, #30, #50, #100 and #200 sieves. Each portion retained on these sieves was
89
then dried in the oven at 110 ºC for 24 hours and weighed. Material passing the
#200 sieve was not collected. Using the resultant gradation from each of the three
as typically used for FM. As is the case with the normal treatment of the FM, two
different gradations can yield the same FM. As such, the overall gradation was
4.2.4 Subsidence
The term “subsidence” used here is defined as the settlement of the top
aligned and firmly bonded together using high-strength tape. The bottoms of the
upper two molds were removed, allowing for the casting of a 600 mm high CLSM
specimen for subsidence measurement. Extra tall specimens were cast to ensure
that sufficient subsidence would occur and that the reduction in the column height
could be more readily measured. The choice of specimen height was rather
arbitrary. Because the maximum pressure increases with the height of fresh
CLSM mixture and higher pressure will result in denser material, it is anticipated
that the measured subsidence (in percentage) will increase with specimen height.
A 600-mm height was chosen to represent a typical layer thickness. After casting
the specimen, the top edge of the specimen was leveled off and periodic
probe, which was lowered from the top of the mold onto the center of the
90
subsided surface. Bleed water usually evaporated before the first measurement.
Measurements were conducted until subsidence ceased, usually within two hours
of casting the specimens. Because the friction from the mold wall restrains the
settlement of CLSM, the results from laboratory tests may under estimate the real
developing a test method with improved accuracy and reliability. This section
mixtures previously shown in Tables 3.9 through 3.16. The results of the initial
compression study, as well as the findings of the various follow-up studies, can be
CLSM.
is very important to carefully handle test cylinders, especially when stripping the
91
cylinders from the molds. Although steel split molds are a good option for
casting CLSM specimens, because of the large number of specimens cast in this
program, a lower cost method was used. Before casting CLSM cylinders for
compression testing, the plastic cylinder molds were cut lengthwise in half till the
bottom. Electrical tape was then used to bind the mold back to its original shape.
This approach only has negligible impact on the shape and size of CLSM
specimens. Figure 4.3 shows a typical plastic mold in this study. After mixing
the CLSM, the cylinders were filled and tapped lightly on the sides to remove
large entrapped air voids. Plastic lids were then placed firmly on the cylinders,
which was maintained at 100% relative humidity (RH) and 23 ºC. After curing in
the molds for seven days, the cylinders were then stripped by simply removing the
electrical tape and removing the CLSM specimens from the cylinders.
Conventional stripping tools were not used because of possible damage to the
specimens. The cylinders were then kept outside of their molds in the fog room
until testing. Some CLSM mixtures tend to leach and soften slightly upon long-
Moist curing was selected for this study so that test results can be
compared from one laboratory to another, even though CLSM is rarely, if ever,
moist-cured in the field. The same argument can also be made for concrete
testing. That is, concrete is rarely moist cured for more than seven days (if at all)
92
in field applications, but standard curing in a fog room provides a benchmark for
Figure 4.3: Plastic mold were pre-cut and taped before casting specimens
to the method for this project, as described later. Most of the compression tests
control (100 kN Instron machine), but some testing was also performed on a
larger capacity machine with load control (1780 kN Tinius Olson machine) to
evaluate the effects of machine capacity. When using the smaller machine,
displacement control was used. Additional testing was also performed to examine
93
the effects of cylinder size (75 x 100 mm, 100 x 200 mm, and 150 x 300 mm),
and, for this study, a constant apparent strain rate, 2,493 microstrain/min, was
used. For this, the crosshead displacement was set at 0.38 mm/min for the 150-
mm-high specimen, 0.51 mm/min for the 200-mm-high specimen, and 0.76
mm/min for the 300-mm-high specimen. The objective was to produce failure in
about the same amount of time for each cylinder size for a given mixture. A
floating, spherical head was used to minimize eccentricities while loading. The
Figure 4.4.
94
For the larger capacity compression machine, load-controlled testing was
employed, as is the case for concrete testing. The typical load rates used for
concrete, 138 to 345 kPa/sec, would fail most CLSM specimens in a matter of
seconds. Thus, a lower load rate of 6.9 kPa/sec was selected. This lower load
rate was possible on the machine for this study but may not be available for many
Only the peak load was obtained from the large capacity machine, and no
deformation data was obtained, which again is typical of concrete testing. Sulfur
capping was used for almost all of the cylinders, except for some weaker mixtures
at early ages, where it was not possible to use sulfur caps. For these mixtures,
neoprene pads were used for testing at seven days. As previously mentioned,
including cylinder size, machine capacity, capping method, load rate, and curing.
method states only to “Apply the load at a constant rate such that the cylinder will
fail in not less than 2 min.” Because of the vagueness in defining the load rate,
compressive strength.
Table 3.9 shows the mixture proportions used for the load rate study.
These mixtures (a total of seven) were selected based on previous testing of the 38
95
mixtures in this project and were chosen to provide a wide range of materials and
proportions. Using these seven mixtures, the effects of displacement rate (cross-
From each of the mixtures in Table 3.9 (A1-A7), standard cylinders (75
mm x 150 mm) were cast and moist-cured until the age of testing. The effects of
loading rate were first assessed for mixtures A-1 to A-4 with a wide range of load
rates (0.13 mm/min, 0.38 mm/min, and 0.89 mm/min). After performing these
tests, a narrower range of load rates (0.25 mm/min, 0.38 mm/min, 0.51 mm/min,
and 0.64 mm/min) were evaluated using mixtures A5-A7. Later in this chapter,
the terms “wide range” and “narrow range” refer to the aforementioned
investigations of load rates. The aim was to determine a suitable load rate range
relatively short time. The latter concern was due to the fact that several mixtures
from early research took a relatively long time (i.e., greater than 10 to 15 minutes)
to fail in compression under displacement control, which would not be ideal for a
the curing conditions and the treatment of cylinders before testing. According to
ASTM D 4832, curing of CLSM cylinders is performed with the specimens in the
molds (in the fog room) until the time of testing. This is different than the
96
normal concrete approach to stripping the cylinders from their molds after about
the first day of curing and then curing them in the fog room. ASTM D 4832 also
specifies a drying time of 4 to 8 hours for test cylinders after the moist-curing
period and before they are tested in compression. Concrete cylinders, on the other
hand, are specified to remain moist until the time of testing, with no required
drying time. Research was conducted to investigate the effects of cylinder storage
The mixtures shown in Table 3.10 were used to study the effects of drying
time (0.5, 2, 4, and 8 hours) and cylinder curing (in mold, out of mold, and in
After curing according to the above regimes, the test cylinders were then
0.38 mm/min. The cylinders were kept moist until the time of testing, i.e., no
additional mixtures were subsequently cast to study these issues in more detail.
To study variations in curing regimes, the mixtures in Table 3.15 were cast and
test cylinders were prepared. This study aimed to identify possible differences in
97
summarized in Table 4.3. Note that for curing condition D, the cylinders were
placed outside the laboratory and were exposed to the high summer temperature
and dry atmosphere of Austin, Texas. All cylinders were capped with sulfur
Table 4.3: Four curing conditions used in the study (using G-series mixtures)
Curing Curing regime
condition
A (normal) Keep sample in mold with cap on, for seven days in fog room. Then
strip cylinder and keep cylinders in fog room until time of testing
B (mold) Keep sample in mold, with cap on, for seven days in fog room. Then
remove cap and keep cylinder in mold in fog room until time of testing
C (cap) Keep sample in mold, with cap on, in the fog room until time of testing
D (out) Keep sample in mold, with cap off, outdoors until time of testing
certain fly ashes are used. Because CLSM is used in many different environments
in practice, the same mixture proportions could exhibit different strength values.
mixtures. This study was a follow-up to earlier testing that suggested that
Three curing temperatures (10 ºC, 21 ºC, and 38 ºC) and six CLSM
mixtures (H-1 to H-6 in Table 3.16) were selected to study the strength gain of
CLSM across a range of practical construction conditions. CLSM was cast into
standard cylinder molds (75 mm x 150 mm) and moved to the appropriate
98
temperature-controlled chamber until the date of testing. The cylinders were
stored in two different manners. Half the cylinders from each mixture were
stripped after three days and returned to the same chamber until the time of testing
designated later in this report as “dry” curing. Temperature and humidity were
monitored throughout the test. The other half of the specimens from a given
mixture were kept inside the molds with the caps firmly placed on top until the
day of testing (designated as “wet” curing). These cylinders were placed directly
After compression testing, the moisture contents of the specimens were measured
to assess the effects of curing conditions on the moisture content (or evaporable
is used as a geotechnical material and placed without concern for curing. During
exposed to air and in contact with the surrounding soil and/or structures.
Different environments may significantly affect the final strength of CLSM as the
water cement ratio may be affected by the seepage of water into surrounding
99
The effects of seepage and evaporation were investigated in a study using
the mixtures detailed in Table 3.14 (F-1 to F-8). This study also investigated the
effects of temperature on strength gain, using the fog room as a control and
later in this report, the findings of this temperature effects study were quite
molds buried in loose sand. Figure 4.5 shows the test set-up. To simulate the
condition of no water loss, CLSM mixtures were cast in plastic molds without
holes and tight lids were put on (condition “Cap”) immediately after casting. To
simulate conditions that only surface water evaporation is possible, mixtures were
cast in plastic molds without holes and lids (condition “No cap”). To simulate
moderate water seepage, mixtures were cast into molds with seven uniformly
lid was put on the cylinders. To simulate severe drainage condition, mixtures
were cast into molds with holes not only on the bottom but also on the side
(condition “Side holes”). There were 36 holes on the wall and seven holes on the
bottom per mold. All holes were of 3.6 mm diameter. No lid was put on the
100
Loose
Mold sand
450 mm 300 mm
Compacted
450 mm sand
450 mm
repeatable compressive strength data. However, for early age samples and/or for
particularly low strength cylinders, it may not be possible to cap cylinders with
sulfur because of the risk of damage to the sample. In these instances, neoprene
pads were used. In addition, some limited comparative studies were performed
neoprene pads. However, because only limited testing was performed, it was
decided to significantly expand the scope of the original work to investigate not
only sulfur and neoprene, but also high-strength gypsum capping compound, for a
101
are the potential health concerns over the fumes generated from sulfur capping
stations and the length of time needed to cap cylinders with sulfur.
The use of neoprene pads has gained popularity in recent years for
sulfur, neoprene pads are easier and faster to use than having to cap with sulfur. It
is well known that higher strength concrete requires higher neoprene durometer
values, and vice-versa. Thus, for CLSM, it is expected that softer neoprene pads
(much softer than those used for concrete) may be needed. Another possible
sometimes used for concrete. Because of the high flowability of the gypsum
paste, it could be helpful in capping CLSM cylinders that are prone to suffering
edge damage (e.g., weak cylinders). The set-up of gypsum capping is shown in
Figure 4.6. In addition, gypsum is non-hazardous, without the odor and health
alternative capping materials was launched. Included in this study were sulfur
caps, gypsum (or hydrostone) caps, and neoprene pads with durometer values of
20, 40, 50, 60, and 70. Eight mixtures (E-series) were used in this study, as
102
Figure 4.6: Method of capping cylinders with gypsum compound
followed for testing six CLSM mixtures, as I-series shown in Table 3.17. The
stripped after seven days. The testing was performed under consolidated and
drained conditions. The pore water pressure was maintained at 34.5 kPa, and the
confining pressures were 69.0 kPa, 103.5 kPa, and 172.5 kPa, respectively.
Theses values resulted in effective confining stress of 34.5, 69.0, and 138.0 kPa,
respectively. The loading rate was 0.38 mm/min, the same loading rate used for
most of the unconfined compression tests. The tests were terminated when the
residual strength was reached or the stress-strain curve became essentially flat.
The approach of stress paths was used to interpret the data. By curve fitting, the
effective internal friction angle, φ´, and effective cohesion intercept, c´, were
found.
103
4.3.3 Water Permeability of CLSM
using ASTM D 5084 -90, “Standard Test Method for Measurement of Hydraulic
The cylinders were cured in the fog room for 28 days. A back pressure of 69 kPa
was applied and maintained until no additional water entered the sample
Because the samples were moist-cured prior to testing, the samples were already
relatively incompressible, thus, the requirement that B ≥ 0.95 was lifted. Rather,
the change of the “B” value with respect to the change of confining pressure was
study the freezing and thawing effects on permeability. The CLSM cylinders
were first moisture cured for 28 days and then were exposed to 12 freezing-and-
thawing cycles. Because CLSM specimens without air entrainment usually break
down in the testing process, polyvinyl hear-shrink plastic film was used to hold
the specimens together. The specimens were 100 mm × 125 mm. Porous stones
were wrapped on both ends of the specimens. Figure 4.7 shows the specimen set-
104
Figure 4.7: Specimen set-up for testing of permeability of CLSM mixture
after freezing-and-thawing cycles
Germany for self-leveling floor screeds was modified and used in this study.
CLSM was cast into an 87.5 mm × 26.3 mm × 1000 mm steel “C” channel. This
angle had one fixed end plate with an anchor and one movable end plate with an
anchor. Prior to casting, wax paper was placed on the inside of the angle to
reduce friction. CLSM was then placed in the fabricated forms. The amount of
105
an accuracy of 0.01 mm, which measured the displacement of the movable end
plate. Shrinkage measurements were taken daily for the first week and once a
week thereafter.
Method of Test for The California Bearing Ratio,” for testing CLSM in this study.
The only difference was that the material (CLSM) was cast or poured into the
molds, without the need for compaction, as is needed for testing soils. After
seven days of curing, the collar was removed, and the surface was trimmed level
using a straight edge. The specimens were tested at the age of 28 days. Six
testing, it was found that the deviator stresses listed in Table 4 of AASHTO T
292-91 were not high enough to introduce sufficient deformations. Thus, the
University. Load conditioning of 41 kPa was used for the 1000 repetitions.
106
4.3.7 Excavation Study of CLSM Mixtures
The excavatability of CLSM was assessed for six of the original 38 CLSM
mixtures included in this study to gain an “order of magnitude feel” for the
relative ease of excavating various CLSM mixtures. CLSM was cast into 450-
gained strength in the first few hours. Walkability is the time at which an average
person can walk on the material. Long-term excavatability was assessed at an age
of approximately nine months using typical hand tools, including a shovel and a
pick for six selected mixtures. The compressive strength of lab-cured cylinders
GeoGauge®.
term strength gain and excavatability was conducted. Nine CLSM mixtures (C-1
to C-9 in Table 3.11) were included in the study. A field penetrometer (field
version of ASTM C 403) was used to evaluate the strength gain of CLSM
mixtures in the boxes. The dynamic cone penetrometer (DCP) was also utilized
situ CBR and to estimate the strength of soil as a function of depth. The testing
107
consists of dropping a hammer (8 kg in mass) from a height of 575 mm, which
forces a steel rod with a conical head into the CLSM or soil. The penetration
depth per blow is then recorded. The corresponding DCP index value was used to
estimate a soil strength value. This value is often correlated with CBR.
Equation 1.a.
W 1 .5 × 104 × C 0 .5 (4.1)
RE =
106
Where, W – In situ unit weight (pcf)
W 1. 5 × 0.619 × C 0 .5 (4.2)
RE =
106
Where, W – In situ unit weight (kg/m3 )
Engineers in Hamilton County, Ohio and the city of Cincinnati have found
future excavatability. The same approach was used to calculate RE values for the
C-series mixtures and comparing the results to other direct or indirect indices of
108
In preliminary investigations, it was found that the splitting tensile
cap the cylinders), and the test results may be helpful in predicting excavatability.
testing may be quite similar to the stress conditions associated with digging the
Table 3.13) were tested for tensile strength, and comparisons were made with
thaw resistance of soil-cement mixtures, was used with one minor modification.
The modification was that the application of wire scratch brush on thawed
specimens was not required because of the low strength of CLSM. One freeze-
thaw cycle involved cycling between -18 °C (a freezer) and 23 °C (the fog room).
earlier time. Mass loss was used as the index of damage. For the initial study,
three cylinders (100 mm x 125 mm) were moist-cured for seven days and three
109
should be noted that tests typically used for concrete, such as ASTM C 666, were
found to be too severe in preliminary trials, and the modified soil cement method
mixtures was performed. After 7 and 28 days of moist curing, the wetting-and
drying cycles were initiated. The specimens were soaked in water at 23 ºC for 24
hours and then dried for 24 hours at room temperature (relative humidity
approximately 50%). The mass of CLSM specimens was recorded before and
after soaking. The testing was stopped after 12 cycles. Six selected mixtures (#4,
result in the need for costly repairs. Concerns have been expressed regarding the
corrosion performance of ductile iron pipe in soil and CLSM. Limited work has
been performed on corrosion of ductile iron pipe embedded in CLSM and work is
can influence the corrosion performance of utilities embedded in both soils and
110
• Acidity or alkalinity (pH of the solution at the pipe interface)
• moisture
program, uniform corrosion will be assumed such that mass loss can be measured.
CLSM and soil at different parts of one cross-section or at different sections along
the pipe. For the first scenario, when a small portion of pipe contacts native soil
(forming an anode of a small area) and the large part contacts CLSM (forming a
cathode of a large area), the worst case of galvanic corrosion is set up. Mass loss
testing was performed in both uncoupled and coupled conditions to assess the
CLSM and evaluated for mass loss in laboratory conditions. Analyses of the
111
corrosion activity of coupons embedded in CLSM mixtures can be compared with
For comparative testing, Ottawa sand was selected as a control material due to its
minimal variance in gradation and composition. However, any soil type could be
used in such a test. The experimental test set-up is shown in Figure 4.8. All of
the initial 38 CLSM mixtures were evaluated in this study. Specimens were cast
and moist cured for 28 days, then immersed in a 3.5% sodium chloride (NaCl)
measurement of average corrosion rate during the exposure period and thus can be
75 x 150mm
cylinder
3.5% NaCl
solution
Pipe
coupon 100 mm
CLSM
or sand
112
4.5.2.2 Coupled Condition (Galvanic)
Figure 4.9 shows the laboratory test set-up for evaluating samples that
may experience galvanic coupling. Samples were moist cured for 28 days before
being exposed to a 3.5% NaCl solution and being electrically connected by a 10-
Ω resistor. Two coupons with essentially equal areas were used in this set-up.
All of the 38 original CLSM mixtures were evaluated in this study. Preparing,
1–90.
10 Ω
resistor
100 x 200 mm
cylinder
3.5% NaCl solution
113
ASTM G 51-95, “Standard Test Method for Measuring pH of Soil for Corrosion
Testing.” The pH values were recorded when the data were stabilized, usually
two or three minutes after inserting the meter into the solution.
This test followed the California DOT Test 532, “Method for Estimating
composed of plastic plates for the frame and steel plates as electrodes for the
current. The box size measures 100 mm ×150 mm × 50 mm. The resistance
Steel Screws
Plates
114
4.5.5 Chloride Diffusion
For underground iron pipes, the rate of corrosion can be accelerated by the
a function of depth can be determined using specific ion probes to determine how
Five CLSM mixtures (#4, #6, #22r, #24, and #26) were evaluated for
chloride ingress rate, as shown in Figure 4.11. From each mixture, two CLSM
cylinders (150 mm x 300 mm) were ponded for 22 weeks with a 3.5% NaCl
solution. Samples were then tested for chloride contents as a function of cylinder
depth.
Chloride
Solution
150 x 300 mm
CLSM cylinder
115
4.6 LEACHING AND ENVIRONMENTAL IMPACT
regarding the use of coal combustion products in construction have rekindled the
coal combustion products and other waste materials. Although the EPA did not
ban or curtail the use of coal combustion products, as originally expected, the
topic will likely resurface in the future. Because the use of by-product and waste
potential toxicity of materials used in this study and to develop a protocol for
and recycled materials included in this project, including fly ash (class F, Class C,
and high-carbon), bottom ash, and foundry sand. The chemical analysis was
atomic absorption. Information such as the calcium oxide (CaO) content of fly
ashes can be very helpful in assessing the potential reactivity and effects on long-
heavy metals are more of a concern than organics in fly ash, so the limited
laboratory study only tested for heavy metals. Various techniques are available
116
for the measurement of organics in constituent materials and CLSM mixtures, but
For each of the by-products included in this study (three fly ashes, one
bottom ash, and one foundry sand), the total heavy metal concentration was first
determined in accordance with EPA Method 610, where nitric acid and hydrogen
peroxide were used to digest the materials. The eight elements analyzed included
Because this testing determines the total amount of heavy metals, and not the
leachable amount, the extraction values are often assumed to be 20 times that of
the TCLP limits. If any of the by-product materials yielded values in excess of
the toxicity limits (20 times the TCLP limit), TCLP was then conducted to assess
the type and amount of heavy metals that are actually leachable from the
materials. Method 40CFR 261.24 was used to extract the samples for TCLP
testing, and the same eight heavy metals as previously listed were measured. The
concentrations of the extracts were compared to the TCLP limits which follow:
Testing the constituent materials directly, rather than testing the materials
117
toxicity screening. If the constituent materials test below the TCLP toxicity limits
under this worst-case scenario, the materials can be classified as non-toxic, and no
additional tests need to be performed. None of the materials tested in this study
exceeded the TCLP limits, and thus tests on actual CLSM specimens were not
deemed necessary. However, if any of the constituent materials had exceeded the
TCLP limits when tested by themselves, CLSM mixtures would have been cast
for subsequent leaching tests. The overall approach to the toxicity and leaching
118
STEP 1:
Chemical analysis of all raw materials
STEP 2:
Heavy metal test of all raw materials
using EPA 610
Yes
Less than
TCLP limits ?
No
STEP 3:
TCLP test (Method 40CFR 261.24)
on potentially toxic raw materials
from Step 2
No
STEP 4:
TCLP test (Method 40 CFR 261.24)
on CLSM mixture containing
potentially toxic raw materials from
Step 3
Yes
Less than
TCLP limits ?
No
119
Chapter 5: Results and Discussions
5.1 INTRODUCTION
In this section, results from the testing program described in Chapter 4 are
provided. Specifically, the results of tests on the fresh and hardened properties,
are presented. Detailed discussions on the test results are included, along with
test methods for field applications and on better understanding factors affecting the
short- and long-term behavior of CLSM, with specific focus on factors influencing
strength.
An important aspect of this study was the assessment of the fresh or plastic
and fresh CLSM properties. Table 5.1 summarizes some of the important
parameters, including water demand (to obtain the target flow), air content, flow,
unit weight, and bleeding (%) for the initial 38 mixtures. Note that mixtures ending
in “r”, such as mixture 2r, denote mixtures that were repeated for statistical
120
Table 5.1: CLSM mixture proportions and fresh properties
Mixture Cement Fly ash type Fly ash Fine aggregate Water flow Total Air Fresh unit
content content type demand bleeding content weight
(kg/m3) (kg/m3) (kg/m3) (cm) (%) (%) (kg/m3)
1 30 Class C 180 Concrete sand 211 20 NA 0.9 1965
2 60 Class C 180 Concrete sand 206 20 2.45 0.95 2108
1r 30 Class C 180 Concrete sand 206 21 2.08 0.9 1974
15 30 Class C 360 Foundry sand 486 20 0.13 2.75 1741
3 60 Class C 360 Bottom ash 577 18 4.32 1.65 1754
8 60 High carbon 180 Foundry sand 532 24 1.04 3.3 1647
10 30 High carbon 180 Bottom ash 628 14 4.81 2.0 1681
9 60 Class F 360 Foundry sand 520 20 0.54 2.5 1684
5 60 Class F 180 Bottom ash 600 18 5.84 2.5 1739
12 30 Class C 360 Bottom ash 572 22 3.64 2.7 1774
4 30 Class F 360 Concrete sand 220 20 0.39 2.2 2199
7 30 Class F 180 Foundry sand 501 20 0.57 2.1 1817
3r 60 Class C 360 Bottom ash 541 20 2.58 2.1 1997
4r 30 Class F 360 Concrete sand 220 22 2.92 1.8 2211
24 60 Class F 1200 None 486 24 2.25 2.8 1635
23 60 None 0 Bottom ash 454 14 1.30 28.5 1382
18 60 None 0 Concrete sand 200 22 0.70 16.5 1836
14 60 Class F 360 Concrete sand 216 22 1.00 1.3 2174
2r 60 Class C 180 Concrete sand 206 25 0.21 0.5 2291
29 60 Foundry sand 0 None 373 23 0.28 2.6 1812
30 30 Foundry sand 0 None 414 20 0.40 2.0 1789
17 30 None 0 Bottom ash 582 13 4.35 20.0 1447
11 60 High carbon 360 Bottom ash 573 23 6.42 1.7 1743
6 30 High carbon 360 Concrete sand 315 20 2.26 1.3 2103
16 30 None 0 Concrete sand 295 20 2.33 16.0 1922
21 30 None 0 Concrete sand 170 18 0.62 25.5 1789
22 60 None 0 Concrete sand 131 20 0.05 26.5 1748
22r 60 None 0 Concrete sand 136 18 0.43 25.5 1802
5r 60 Class F 180 Bottom ash 600 16 7.20 1.4 1887
26 60 None 0 Concrete sand 136 17 0 25.5 1802
16r 30 None 0 Concrete sand 295 19 2.35 15.5 1874
13 60 Class C 360 Foundry sand 499 20 0 1.8 1902
25 60 High carbon 1200 None 853 24 7.38 1.3 1322
19 30 None 0 Bottom ash 492 13 1.08 25.0 1385
20 60 None 0 Bottom ash 525 13 3.41 18.5 1485
27 60 Class F 1200 None 486 23 1.28 0.7 1638
20r 60 None 0 Bottom ash 525 13 1.44 15.5 1511
28 60 Class F 180 Concrete sand 220 20 1.33 1.4 2182
121
purposes. The results of the mixtures are presented in the order in which they were
mixed and are not sorted by mixture type. The fresh properties of the A- to I-series
The water demand to obtain the target flow was found to be one of the most
carbon fly ash, bottom ash, and foundry sand, were found to particularly increase
the water demand of CLSM. By increasing the water content of CLSM, most of
ASTM D 6103 was found to be a suitable method for measuring the flow of
CLSM. For mixtures with low density (high air content), it was found that the 200
mm flow was sometimes difficult to obtain, and for these mixtures, a flow of 175
mm was found to be satisfactory. The bottom ash used in this study lacked
sufficient fines for good workability and was prone to bleeding. However, when
used with high contents of fly ash, the mixtures were more workable and exhibited
reduced bleeding.
The effects of material source and quantity on the water demand of CLSM
mixtures. Figure 5.1 shows the results. This graph illustrates the statistically
significant variables that affect water demand. In this graph, the effect is the
difference between the specified variable level and the mean water demand of
446.3 kg/m3 . The reference variable levels are concrete (river) sand, Class C fly
ash, 180 kg/m3 fly ash content, and 30 kg/m3 cement. The effect of reference level
is the opposite of the sum of remaining levels. The plus sign (+) in the key after the
122
variable description indicates that the difference was positive and dash (-) indicates
that the difference was negative. The negative sign means that the variable level
will reduce that amount of water from the mean. The effect of a reference level is
the opposite of the combination effects of all other variable levels. From the figure
it can be seen that the fine aggregate source was the most significant factor
affecting the water demand of mixtures. The use high-carbon fly ash also increased
the water demand. There is no significant difference between the use of the Class
C and Class F fly ash. The effect of high content of fly ash (360 kg/m3 ) was
negligible.
10
0
0 20 40 60 80 100 120 140 160
Effect on water demand (kg/m3 )
123
Table 5.2 shows the results from the analysis of variance (ANOVA)
or more populations or treatment. The Mean Squares column shows the summary
statistic for the source and the degree of freedom (DF) column shows the number
of model terms summarized by the mean square. The P column gives the
significance. Values less than 0.05 are usually considered statistically significant
and thus affecting the water demand. Thus, this table identifies the significant
variables as fly ash type, fine aggregate source, and the interactions between
presented in Chapter 4. This approach was found to be easy and fast. The results
are shown in Table 5.1. The range of entrapped air contents was similar to those of
containing foundry sand. This is most likely attributed to the presence of clay, coal
fines and other organic materials in the foundry sand. It is believed that the
124
foundry sand prevented the encapsulation or formation of entrapped tiny air voids,
thus making the AEA none-effective. Air-entraining agent dosages were increased
The results of unit weights of fresh CLSM mixtures are shown in Table
5.1. The unit weights of CLSM mixtures in this study ranged from 1322 to 2291
kg/m3 . The unit weights of CLSM mixtures with entrapped air were only slightly
lower than ordinary Portland cement concrete because of the higher water contents
densities as a result of the higher water demand and the lower specific gravity of fly
ash.
5.2.4 Setting/Hardening
construction demands (i.e., timing between lifts or early opening to traffic). Test
methods are needed to more easily assess the setting of CLSM, both in the
laboratory and in the field. This section discusses some of the preliminary findings
• Needle penetrometer
125
• Soil pocket penetrometer
mixtures, it is not possible to assign initial and final CLSM setting times as is
(ASTM C 403) was correlated with soil penetrometer values for all mixtures, and
hardening time related to “walkability time” was investigated for select mixtures.
Appendix B shows the raw data from the soil penetrometer (SP) and needle
penetrometer (NP) tests for the Stage One 38 mixtures, as well as the eight
mixtures from the I-series. The walkability time was assessed by preparing large
“CLSM boxes” that were walked on at various ages. It was found that the soil
penetrometer values were in the range of 4.3 kPa to 7.4 kPa (average of 6.1 kPa)
when CLSM mixtures were able to support the weight of an average person with an
approximate 6 mm indentation.
Thus, it is often not feasible to compare the setting time of CLSM mixtures to each
other at predefined time increments, but rather, the timing of measurements was a
subject to bleed water effects. Despite these differences (25 mm versus 6.4 mm) in
penetration depth, there was a limited correlation between soil penetrometer and
126
needle penetrometer values for the 38 mixtures, as shown in Figure 5.2 with a R2
(the square of the correlation coefficient) of approximately 0.75 for all the CLSM
0.8
0.7
0.6 R2 =0.75
0.5
0.4
0.3
0.2
0.1
0
0 2 4 6 8 10 12 14
Figure 5.2: Correlation between ASTM C 403 and soil penetrometer values
One potential problem with using the soil pocket penetrometer is the effects
(and fines) is common in CLSM construction and can make surface measurements
CLSM and being able to quantify the penetration resistance of the top layer of
Figure 5.3, to minimize the effects of bleeding on field measurements. When using
127
the pokcet penetrometer, the stand was first placed on the mixture surface. The top
of the penetrometer was then inserted into the hole on the top the stand, where its
dead weight pushed the tip of the penetrometer through the soft fines and bleed
water and came to a rest on the dense surface that marked the “true” top. The 6.4
mm penetration depth was then marked on the tip, based on the measurement taken
from the top surface of the stand. Finally, the needle was pushed down and the
reading was recorded. This modified stand was used for the E-series mixtures.
penetrometer gives the bearing pressure. This test is based on Terzaghi’s bearing
capacity equation for circular footings, and the soil friction angle is assumed to be
The unconfined compressive strength is two times the cohesion, according to the
theoretical total stress failure envelope. For the pocket soil penetrometer, the
failed at this depth. According to Bhat, Bucchi suggested this method for the rapid
assumes that only cohesion contributes to the bearing strength, the following
128
1.3 × 5.7q u
qf = 1.3 Nc Cu = = 3.71 qu (5. 1)
2
multiplying the reading from the soil penetrometer by 3.71, the penetration
resistance (qf) can be back calculated. This is used in Figure 5.4 to compare the
penetration tests.
129
3.00
SP
BSP
2.50
Series3
back calculated soil penetrometer resistance
a
Soil penetrometer reading (SP) and
2.00
(BSP) (MPa)
1.50
1.00
0.50
0.00
From Figure 5.4, it was observed that at the early stage of setting, the soil
penetrometer may yield higher resistance values than those measured by needle
penetration. This may be because the soil penetrometer is more sensitive at low
hydration proceeds and CLSM gains stiffness, the needle penetrometer clearly
exhibited higher values than those of the soil penetrometer. When using the needle
penetrometer for a penetration depth of 25.4 mm, it was observed that that the load
per unit area increased to a peak and then dropped off considerably. This
phenomenon was not observed for the soil penetrometer, which only requires a
130
penetration depth of 6.4 mm. Because CLSM mixtures are often plastic for several
hours after mixing, the soil penetrometer may not generate a bearing failure.
Figure 5.5 shows a needle (ASTM C 403) after the penetration testing.
Note the small wedge at the tip of the needle. This confirmed that a general failure
deformations, thus affecting the bearing capacity. This also offers a possible
The pocket VS tester was used to test the shear strength of CLSM mixtures
at early ages. There was some limited correlation between the VS values and the
soil penetrometer values, as shown in Figure 5.6 (R2 =0.45). The VS values
some assumptions on failure plane, whereas soil penetrometer values are measured
indirectly.
131
Wedge
Figure 5.5: Setting time needle and the wedge at the tip of the needle
132
0.50
0.40
R2 = 0.45
0.30
0.20
0.10
0.00
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Figure 5.6: Comparison between soil penetrometer and vane shear values
Bleeding was found to be common for CLSM mixtures. Even for CLSM
mixtures with high air content, a very small amount of bleeding water was
observed on the mixture surface. Flash fill (mixture C-2) showed little bleeding
because of its fast setting nature. The bottom ash used in this study lacked
sufficient fines for good workability and was prone to bleeding. However, when
used with high contents of fly ash, the mixtures were workable with reduced
bleeding.
contents were also analyzed using the statistical software program, ECHIP. The
results from the analysis are shown in Figure 5.7. The mean bleeding is 2.52%.
The foundry sand used in this study reduced the bleeding significantly, while the
133
bottom ash significantly increased the bleeding. The fly ash type had a minimal
10
Bottom ash (+)
XXX Class F FA
XX 60 kg cement
XX 60 kg cement * Class F
XX 60 kg cement * 360 kg FA
0
0 1 2 3 4
Bleeding (% of weight)
the results from the test shown in Table 5.3 were obtained. These data represent
the gradation of aggregates sieved from different depths within a CLSM cylinder.
calculate the FM for sand used in concrete. Based on the data shown in Table 5.3,
134
very little segregation occurred in any of the mixtures. However, the test method
with the intention of correlating the laboratory results with actual field
performance.
135
5.2.7 Subsidence
The results of subsidence testing for six CLSM mixtures are shown in Table
5.4. All of the CLSM mixtures exhibited appreciable subsidence, with the
exception of mixture #23, which had relatively poor flowability. With the
exception of mixture #23, there was a limited correlation between subsidence and
136
3
2.5
1.5
0.5
0
0 0.5 1 1.5 2 2.5
Bleeding ( weight %)
findings from the initial mixtures (38 in all), in which several aspects of
compression testing were examined, including the effects of cylinder size, capping
material, and load rate. Based on these original mixtures, some useful predictive
models were developed to predict the strength gain (short- and long-term) of
CLSM. The original study led to several follow-up studies, each of which focused
137
in more detail on issues involving testing parameters. Detailed investigations on
load rate, curing and conditioning of cylinders, effects of drainage on strength, and
the use of alternative capping materials are included in following sections. The
findings of the initial, broad study and the later, detailed studies could be used to
strength of CLSM.
at 3, 7, 28, and 91 days are shown in Table 5.5. The results were found to be
repeatable, with quite low values of coefficient of variation. The data shown in this
most mixtures show a drastic change in the load-deflection curve as the curing time
is increased. Figure 5.9 illustrates this behavior for Mixture #10, which was typical
of most CLSM mixtures. At early ages, CLSM acts more like a soil, with more
ductile behavior. But as time progresses, CLSM begins to act more like concrete,
were studied in the initial investigation, some of which were later addressed in
more detailed research. Table 5.6 shows a comparison between sulfur-capped and
neoprene-capped cylinders for the selected six CLSM mixtures. In general, sulfur-
capped cylinders yielded higher strengths than cylinders using neoprene pads. The
138
Table 5.5: Unconfined compressive strength of original 38 CLSM mixtures
3-day f’c C.O.V. 7-day f’c C.O.V. 28-day C.O.V. 91-day C.O.V.
Mixture
(MPa) (%) (MPa) (%) f’c (MPa) (%) f’c (MPa) (%)
1 0.12 8.17 0.21 6.81 1.09 4.91 1.87 2.8
2 0.29 2.42 1.76 10.11 3.69 4.03 6.26 13.5
1r 0.13 14.36 0.24 1.18 1.35 7.69 2.34 0.3
15 0.07 8.22 0.11 5.66 0.18 3.97 0.25 1.4
3 0.33 10.50 0.57 2.42 1.36 6.83 2.02 2.9
8 0.09 9.75 0.11 8.27 0.25 4.69 0.33 10.6
10 0.12 2.27 0.16 15.12 0.22 7.93 0.26 1.3
9 0.09 9.66 0.13 12.40 0.22 4.99 0.25 2.9
5 0.14 13.70 0.18 8.15 0.46 16.11 0.57 5.3
12 0.30 16.17 0.27 6.16 0.57 4.70 0.86 3.4
4 0.34 4.94 0.48 6.18 0.79 11.04 1.08 13.6
7 0.09 3.22 0.11 3.40 0.12 9.70 0.16 5.8
3r 0.46 13.08 0.58 4.36 1.49 5.78 1.97 8.3
4r 0.41 13.61 0.57 5.80 0.94 3.92 1.03 6.9
24 0.34 4.77 0.22 1.40 0.44 0.13 0.58 4.6
23* - - 0.04 6.42 0.14 9.52 0.18 7.6
18* - - 0.33 6.75 0.70 1.08 0.79 4.0
14 0.58 6.61 1.07 13.46 2.15 8.14 3.49 16.8
2r 0.42 9.30 1.58 2.86 4.90 2.44 6.87 0.3
29 0.18 0.57 0.31 0.22 0.63 1.9 0.98 6.1
30 0.09 7.08 0.14 3.10 0.26 8.55 0.28 2.8
17* - - 0.01 31.79 0.07 18.90 0.13 16.8
11 0.33 1.66 0.42 4.29 0.75 3.48 0.94 4.7
6 0.40 10.20 0.47 0.74 0.83 4.88 1.09 4.7
16* - - 0.06 11.90 0.13 12.03 0.16 8.5
21* - - 0.09 10.59 0.16 11.84 0.18 11.7
22* - - 0.43 8.97 0.73 4.33 1.01 4.8
22r 0.32 4.19 0.50 9.68 0.96 17.53 0.93 7.0
5r 0.17 12.50 0.28 10.33 0.55 10.08 0.78 16.3
26 0.43 7.08 0.76 8.23 1.14 15.75 1.53 2.6
16r* - - 0.07 9.74 0.15 23.57 0.17 8.3
13 0.28 0.45 0.35 3.33 0.74 3.2 1.12 4.4
25 0.17 4.44 0.30 5.72 0.40 30.90 0.50 9.0
19* - - 0.02 0.71 0.06 45.02 0.06 13.2
20* - - 0.04 36.42 0.21 1.01 0.29 26.0
27 0.22 4.61 0.29 1.72 0.36 3.83 0.55 6.6
20r* - - 0.04 49.90 0.15 32.77 0.24 10.4
28 0.28 3.02 0.47 0.86 0.70 1.95 0.94 0.2
* Mixtures were too weak to be tested at three days.
139
3000
28 days
2500
2000
7 days
Load (N)
1500
1000
3 days
500
0
0 1 2 3 4 5
Deformation (mm)
neoprene pads used in this study had a durometer value of 50 (an index of
140
pads with varying durometer values, was subsequently performed, as described
Efforts were made in this study to develop predictive models for the
considered. No single model was found to work well for the entire range of
materials and mixture proportions; however, predictive models for subsets of the
mixtures were found to be quite accurate. For instance, separate models were
developed for air-entrained (both for “high” and “moderate” air contents), and non-
air-entrained CLSM.
the mixtures containing bottom ash as the fine aggregate, because of the
aforementioned problems with bleeding, were not included in the model. By trial-
f c' = a ⋅ e b ( w / c ) ( 5 .2 )
a = 0 . 3074 ⋅ ln( t ) + 0 . 22
b = 0 .0086 ⋅ ln( t ) − 0 . 272
Where, f c' is the compressive strength in MPa, w/c is the water-cement ratio, and t
mixtures are plotted in Figure 5.10. There was very good correlation, with a R2
value of 0.97. This formula was also found to be effective in predicting long-term
141
strength gain (i.e., beyond 91 days). For example, cylinders from mixture 22r were
tested for compressive strength after 256 days, with an average strength of 1.0
1.4
1.2
R2 =0.97
1
Predicted strength (MPa)
0.8
0.6
0.4
0.2
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Measured strength (MPa)
mixtures. An initial approach, similar to that used by Bhat10 , was used in which the
water-cement ratio was the only variable used to predict the compressive strength.
This approach yielded an R2 value of 0.8. To improve this accuracy, a formula was
developed that included the water-cement ratio, aggregate type, fly ash type, and
approach was to assign numerical values to the non-numerical variables used in the
142
formula. Through a trial and error process, the following constants (k) were
selected for the materials used in this investigation. Concrete or river sand (kriver
sand) was assigned a value of 1.0, foundry sand (kfoundry sand) a value of 0.2, bottom
ash (kbottom ash ) a value of 1.0, Class C fly ash (kC ash ) a value of 2.2, Class F fly ash
(kF ash ) a value of 1.0, and high-carbon fly ash (kHC ash ) a value of 0.75. The
equation for predicting the compressive strength, S(t), is shown below, and a
5.11.
S(t) = b 0 (t) ⋅ (k agg.type ) b 1 (t) ⋅ (k fly ash type ) b 2 (t) ⋅ (w/c) b 3 (t) ⋅ (k fly ash content ) b 4 (t) (5. 2)
where,
b3 (t)= -0.0034•t-1.03
Where, S(t) is the compressive strength in MPa and t is the age in days.
For field applications, the empirical equations presented in this report can
design specification. It should be noted that the derived equations have limitations.
As is the case for experimental studies, the validity of the equations is only for the
143
specific materials used in this study. It is recommended that those interested in
generate a database that can serve as the basis for predictive models.
6 R2 =0.97
5
0
0 1 2 3 4 5 6 7
for the specific materials used in this study, a reasonable correlation was obtained
between the equations and test data from a different series of mixtures using
completely different materials sources. Table 5.7 shows the mixture proportions
for this study, along with a comparison between the measured and predicted 7- and
144
28-day compressive strength values. Although there was a reasonable correlation
between predicted and actual strengths, this correlation could have been improved
Mixture* Cement Fly ash Concrete Water Air Flow Fresh unit
(kg/m3 ) (kg/m3 ) sand (kg/m3 ) content (mm) weight
(kg/m3 ) (%) (kg/m3 )
I-1 30 180 1500 283 1.30 200 2036
I-2 60 180 1500 327 1.25 210 2077
I-3 120 180 1500 335 0.65 220 2087
Mixture* 7 days (MPa) 28 days (MPa)
Measured Predicted Measured Predicted
I-1 0.202 0.314 0.368 0.704
I-2 0.572 0.536 1.062 1.277
I-3 1.832 1.081 2.728 2.696
* Mixture codes are different from those used in present study. These mixtures from a previous study contained high carbon
fly ash in combination with three dosages of Type I portland cement
properties was obtained from this study. Most of the significant effects were
tended to increase water demand (such as foundry sand and high-carbon fly ash)
generally yielded lower compressive strengths than materials with lower inherent
water demands. The chemical reactivity of fly ashes was found to be critical, as the
strength of CLSM containing Class C fly ash was higher than similar mixtures
containing Class F or high-carbon fly ash. Class C fly ashes, by definition, have
higher calcium (CaO) contents than Class F fly ashes (and the high-carbon fly ash
145
used in this study) and are more effective in increasing strength, even as early as
seven days.
smaller investigation on the effects of cylinder curing (i.e., in mold, out of mold, in
lime water) is described first (using the B-series mixtures), followed by a more
comprehensive study on cylinder curing (using the G-series mixtures). The overall
objective of these studies was to determine the most efficient and accurate
Table 5.8 shows the effects of curing method on mixture B-1. The findings
show that the compressive strength of CLSM can vary widely, depending on
whether the cylinders are cured in or out of their molds, or whether the cylinders
are cured out of their molds in lime-saturated water. This initial study, using only
strength was achieved when the cylinders were left continuously in their molds,
suggesting that cylinders stored out of their molds or in lime water may be subject
the potential effects of specimen leaching and loss of mass, led to further
in this section.
146
Table 5.8: Effects of curing methods on compressive strength (Mixture B-1)
Curing 7-day strength C.O.V. 28-day strength C.O.V. 91-day strength C.O.V.
method* (kPa) (%) (kPa) (%) (kPa) (%)
Cylinders kept
in molds, 313.6 2.9 1360.9 2.0 3811.5 9.4
stored in fog
room
Cylinders were
stripped at 3 362.2 1.2 1394.7 3.8 2396.5 4.4
days and cured
in fog room
Cylinders were
stripped at 7 352.6 4.6 1157.2 4.8 2670.8 3.8
days and cured
in lime-
saturated water
* Refers to curing regimes previously described in Chapter 4. All curing was performed at 23 °C.
more comprehensive investigation was performed using ten CLSM mixtures (G-
used to better understand the potential effects on strength values for different curing
• Curing Condition A (“normal”) – Cylinder kept in mold (with cap on) for
seven days in fog room. Cylinder then stripped and kept in fog room until time
of testing. This is referred to as “normal” because most cylinders tested in this
project followed this procedure.
• Curing Condition B (“mold”) – Cylinder kept in mold (with cap on) for seven
days in fog room. Cap is then removed and cylinder is kept in fog room until
time of testing.
• Curing Condition C (“cap”) – Cylinder kept in mold (with cap on) until time of
testing.
• Curing Condition D (“out”) – Cylinder kept in mold (with cap off) outdoors
until time of testing.
147
After subjecting cylinders from each of the G-series mixtures to the curing
regimes just described, all cylinders were capped with sulfur capping compound
and tested at a loading rate of 0.38 mm/min. The compressive strengths measured
at 28 and 91 days are shown in Tables 5.9 and 5.10, respectively. The remainder of
this section discusses these results and also describes some of the nuances observed
when testing different materials and mixture proportions. Some interesting
observations were made that illustrate that the strength of CLSM is significantly
affected by variations in curing regime, and further, that these variations are a
function of specific materials and mixture proportions. Some of the more
interesting results are illustrated in Figures 5.12 to 5.15, as described later in this
section.
The largest increase in strength was observed for mixtures G-2, G-8, and G-
10, which contained Class C fly ash, where the high temperatures apparently
helped to activate the fly ash. Mixtures G-2, a typical “flash fill” mixture with 275
kg/m3 of Class C fly ash and no portland cement, showed a 40 percent increase in
compared directly to Class F fly ash for mixtures containing foundry sand (Figure
5.12) and bottom ash (Figure 5.13). Mixtures containing Class F or Class C fly
ashes exhibited higher strengths for cylinders cured outdoors in a hot climate, but
the differences were more profound for Class C fly ash. In fact, the difference
149
between fog-room cured and outdoor-cured cylinders were as high as 250 percent
for the mixtures containing Class C fly ash. Thus, using laboratory-cured cylinders
to assess the field performance of CLSM containing fly ash (especially high-
calcium fly ash) in hot environments must be done with caution. The effects of
fly ash are used, and when the fly ash to cement ratio is high. Another impact of
higher temperature and outdoor curing (with the caps removed from the cylinders)
is the reduction in water content caused by evaporation, which lowers the effective
water-binder ratio.
For high air content mixtures, G-3 and G-6, the effects of curing methods
varied with cement content (and strength levels), as shown in Figure 5.14. Mixture
G-3 contained 30 kg/m3 of cement and exhibited relatively low strengths. In fact,
150
mixture G-3 suffered such a reduction in strength when stored outdoors (compared
days, cylinders could be tested, but the resultant strength was significantly lower
than fog-room cured cylinders. Mixture G-6, which contained 45 kg/m3 of cement,
than the lower cement content mixture (G-3). For mixture G-6, there was a
findings suggest that high-air content CLSM mixtures (without fly ash) benefit
151
Figure 5.14: Curing condition effects on compressive strength of Mixture G-3
and G-6
Figure 5.15 for mixtures containing low (15 kg/m3 ) and intermediate (30 kg/m3 )
cement contents and Class F fly ash. Mixture G-4, the low cement mixture, yielded
quite low strength values, and curing in the mold showed negative effects on the
5, with the intermediate cement content, the strengths of outside cured specimens
showed 1.4 times those of specimens cured in the fog room. This further confirms
that both material type and mixture proportions affect the compressive strength of
152
standard test method is essential in generating reproducible test results that can be
temperature and cylinder storage led to a final investigation on the effects of curing
153
5.3.1.3 Effects of Drainage Conditions on Compressive Strength
CLSM in native soil, and various drainage conditions were assessed. This section
presents the results of compression testing for cylinders stored in the following
manners:
• “Normal” – standard fog room curing (the same curing regime used throughout
most of this study)
• “No Cap” – cylinders were placed in the curing box with the cylinder caps
removed to simulate surface evaporation.
• “Cap” – cylinders were placed in the curing box with the cylinder caps left
firmly on top to simulate a situation without water loss.
• “Bottom holes” - cylinders were placed in the curing box with the cylinder caps
removed and with holes drilled in the bottom of the cylinders to simulate both
surface evaporation and bottom drainage.
• “Side holes” - cylinders were placed in curing box with the cylinder caps
removed and with holes drilled in the bottom and sides of the cylinders to
severe drainage (from top, bottom, and sides).
154
same time-temperature history as the other cylinders stored in the curing box, but
the cylinders were not subjected to drainage or evaporation.
The compressive strength test results at 7 and 28 days are shown in Table
5.11 and Table 5.12, respectively. Note that mixture F-2, which was a Class C fly
ash mixture without cement, was not tested at 28 days, due to the rapid setting time
for the four curing methods that utilized the curing box, suggesting that the
based on this laboratory simulation study. If differences were observed for a given
mixture, the difference was more apparent at 7 days than at 28 days. Once again,
the effects of temperature were evident as cylinders cured in the plywood box set-
cylinders cured in the fog room. This was particularly evident for mixtures
containing fly ash. The only fly ash mixture that did not show higher strengths in
the curing box, compared to the fog room, was mixture F-2, which was a so-called
“flash fill” mixture. This mixture set extremely fast, with little or no water
available for seepage or evaporation. Also, because the mixture set so fast and
hydrated so rapidly within the first few hours, the effects of long-term temperature
was that the effects of curing temperature on strength may be quite complex
because it is not just a function of constituent material type but also of material
155
quantity. For instance, when low cement contents (15 kg/m3 ) were used in
combination with fly ash, such as in mixtures F-4, F-6, and F-8, there were only
the cement content was increased to 30 kg/m3 (with the fly ash content remaining
the case of mixtures F-5 and F-7. Thus, one must consider the combined effects of
material type and quantity when assessing the long-term strength gain of CLSM,
156
Table 5.12: Compressive strength at 28 days under different drainage
conditions
Mixture Normal No cap Cap Bottom holes Side holes
Average C.O.V. Average C.O.V. Average C.O.V. Average C.O.V. Average C.O.V.
(%) (%) (%) (%) (%)
F-1 532.3 16.3 615.1 14.4 511.5 29.7 554.2 19.4 658.7 9.0
F-2* - - - - - - - - - -
F-3 144.1 14.2 138.4 14.2 132.0 17.7 153.6 25.6 157.3 30.0
F-4 291.9 5.7 309.0 7.5 253.1 16.3 250.7 6.8 283.6 10.2
F-5 842.1 9.0 1497.2 3.1 1329.3 3.6 1576.2 5.0 1586.0 3.6
F-6 255.1 2.7 334.3 18.6 297.4 14.1 394.8 8.6 412.3 9.2
F-7 779.5 14.1 1215.4 8.1 1412.6 2.2 1217.2 4.0 1278.3 1.7
F-8 493.1 6.9 499.8 5.7 462.5 2.4 528.7 3.8 560.3 4.0
* Cylinders from mixture F-2 were not tested at 28 days.
in many CLSM mixtures and suggests that drainage and evaporation may not be as
CLSM placed in actual applications to gain insight into real drainage conditions, as
conditioning (i.e., drying specimens after curing but before testing) is described.
to air-dry for 4 to 8 hours before capping. It was observed that this long waiting
period may be inconvenient and impractical in a testing laboratory and may also
lead to variations in test results. To investigate this, cylinders from mixture B-2
157
were moist-cured for various ages, removed from the fog room, then subjected to
The results of testing after 7, 28, and 91 days of curing are shown are shown
in Table 5.13. Although there were some variations in strength results at each age
when the specimens were allowed to dry for different periods of time, the
variations were not appreciable in most cases. This suggests that it is not necessary
to air dry specimens for 4 to 8 hours before testing, as is specified in the ASTM
expected that more accurate results may be obtained, and laboratory productivity
158
5.3.1.5 Alternative Capping Materials for Compression Testing
included in this investigation were neoprene pads with Shore “A” durometer values
of 20, 40, 50, 60, and 70, as well as sulfur-capping compound and gypsum (or
“hydrostone”). After curing the CLSM cylinders for various ages (7, 28, and 91
days), the cylinders were capped and tested using a similar approach which
followed throughout most of this project. A load rate of 0.38 mm/min was used for
all cylinders, and load and deformation values were recorded by a data acquisition
system.
method of using this capping material. When using the gypsum, a water-gypsum
ratio of 0.3 was used, and a split 75-mm x 150-mm plastic mold was used as a
“dam” to retain the fluid gypsum mixture on top of the cylinder end. Sealing
grease was used to prevent the gypsum mixture from leaking through the gap
between the cylinder surface and plastic mold. Because of the self-leveling
property of gypsum mixtures, cylinders with damaged ends could be easily capped.
The long setting time of gypsum mixtures, up to 40 minutes, made the capping
process time consuming, and thus, gypsum was only used for testing at 28 days.
Table 5.14 shows the compressive strength values measured for the
different mixtures and capping materials at 7, 28, and 91 days. Table 5.15 shows
the various capping methods. The sulfur capping method provided the highest
159
strength values for all the eight mixtures. Also, in general, sulfur capping
generated the lowest variations compared to the other capping methods. Note that
Mixture E-2, a “flash fill” mixture, set so quickly that it was not possible to
produce enough cylinders for testing at all ages. Thus, testing was only performed
at an age of 7 days.
160
Table 5.15: Coefficients of variation for compressive strengths using different
capping materials
Mixture E-1 E-2 E-3 E-4 E-5 E-6 E-7 E-8
Sulfur 7 day 1.6 5.8 5.0 7.1 15.7 17.6 8.4 6.9
28 day 16.3 14.2 5.7 20.7 13.3 4.0 10.3
91 day 3.6 15.3 10.6 10.5 6.9 5.3 8.1
Gypsum 28 day 18.0 7.7 6.1 5.5 5.3 7.8 14.9
D70 7 day 5.3 12.4 20.3 5.9 3.7 12.0 9.7 9.5
28 day 45.9 19.3 22.0 44.5 16.1 14.9 20.8
91 day 22.8 6.8 24.9 8.3 3.0 3.7
D60 7 day 18.0 6.5 15.5 7.2 17.3 12.0 13.2 22.1
28 day 12.9 2.9 2.5 4.0 23.9 7.6 22.0
91 day 30.4 12.9 16.4 12.5 6.6 21.4
D50 7 day 17.5 5.7 28.7 15.9 7.4 0.7 4.6 30.0
28 day 18.1 15.5 5.0 20.9 15.1 6.4 6.5
91 day 17.0 22.2 12.6 4.2 4.0 11.5
D40 7 day 17.9 27.6 23.3 21.4 10.1 13.0 6.8 15.2
28 day 15.4 14.0 9.5 20.5 19.5 10.8 11.2
91 day 12.7 2.7 4.6 27.8 6.6 14.6 9.4
D20 7 day 16.5 1.9 20.5 15.8 10.3 7.3 9.6 16.6
28 day 7.1 1.1 14.5 10.1 19.2 6.8 5.9
91 day 6.1 7.4 22.2 9.1 6.0 23.5
161
Figure 5.16: Effects of capping methods on coefficient of variation (7-day
compressive strength)
162
Figure 5.17: Effects of capping methods on coefficient of variation (28-day
compressive strength)
163
The process for the qualification of unbonded capping systems described in
ASTM C 1231 was followed in analyzing the test data in this study. Under close
examination, the relationship expressed as equation (3) in the test method was
deviation of the difference should be based on xp and 0.98 xs. When testing
concrete according to ASTM C 1231, the strength from different capping methods
is assumed to be essentially equal, and the error using the proposed process can
typically be neglected. However, it has been shown that different capping materials
yield different strength values for CLSM and as such, the validity of the specified
ASTM C 1231 analysis process was examined in more detail. The theoretically
correct process should calculate the difference in strength for each pair of cylinders
d i = x pi − ax si (5. 3)
Where:
than “a” percent of strength measured with sulfur capping at a 95% confidence.
The initial estimate of a0 was selected to be the value obtained through the
ASTM C 1231 process satisfying equation (3) in ASTM standard. Adhering to the
164
process outlined in ASTM C 1231-99 (using the proposed equation 5.4 above, a
repeating until converging on an accurate value. Usually, only one or two cycles
were needed for convergence. Table 5.16 illustrates the qualification of unbonded
caps using the unmodified ASTM C 1231 process and the proposed, modified
method. For each durometer hardness and gypsum cement, more than 40 pairs of
specimens were included in the statistical analysis. The differences between the
two processes were found to be within six percent of each other, a relatively small
difference. However, it was believed that it was worthy of consideration, given the
165
ASTM D 4832 states that the alternative capping systems are acceptable
when the average strength obtained is not less than 80% of the average strength of
criterion, the use of gypsum and neoprene pads with durometer values of 20 and 50
could be qualified when using the ASTM C 1231 process. By using the modified
method (as shown in Table 5.16), gypsum and neoprene pads with durometer
each of the compression tests. Rather than providing all of this data in this
dissertation, some representative results are presented next. Figure 5.19 shows a
between loading platens, the compression of neoprene pads is also included the
deformation value. As a result, higher deformation values and longer loading times
were recorded in the testing. Thus, care should be taken in performing tests and
analyzing data when different capping materials are used because both the
measured deflection and testing time are directly influenced. Unlike concrete
difficult to measure deflections or strains from the cylinder surface because of the
inherent low strength of CLSM. For CLSM, the most relevant data can be obtained
from using rigid caps (gypsum or sulfur) because the measured deflections (based
material.
166
4
D60
3.5
3 D20
D20
2.5 D40
Load (kN)
D70 D50
2
Sulfur D40 D60
1.5
D70
1 Gypsum Gypsum
D50
0.5 Sulfur
0
0 100 200 300 400 500 600 700
Displacement (mm)
In summary, the main findings from this capping material study include the
following:
167
results. However, when applying the acceptance criteria described
regarding the loading rate, additional emphasis was placed on assessing the effects
of loading rate on the compressive strength of CLSM in this study. This section
describes the results of two studies, the first of which focused on a relatively wide
range of loading rates (0.085 to 0.58 %/min), and the second of which covered a
machine, often used to test soils, was used for this investigation. The results of
these two studies, coupled with the findings from the initial investigation, could be
used to recommend loading rates for a refined method to measure the unconfined
testing. The testing of CLSM cylinders should be accurate. If strength values are
found to be influenced by loading rate, then a load rate range must be defined and
required for accurate testing. The loading rate also determines the length of time
needed to test a given cylinder. This required length of time must be sufficient to
ensure accuracy (i.e., not a sudden cylinder failure) and to complete a given test in
test cylinders may be critical, especially if many tests are performed in a single day.
168
The compressive strength results for the two loading rate range studied,
referred to as “wide range” and “narrow range” studies, shown in Tables 5.17 and
5.18. The cylinder measures 75 mm x 150 mm were used in this study. The full
load-deflection curve was also obtained for each cylinder tested, and the
deformation at peak load for each test specimen was recorded, as shown in Tables
5.19 and 5.20. The deformation at peak load was used to determine if changes in
affected both the compressive strength of CLSM and the deformation at peak load.
169
Table 5.18: Effects of load rates (narrow range) on compressive strength
Age Mixture 0.16 %/min 0.25 %/min 0.33 %/min 0.42 %/min
(days) Average C.O.V. Average C.O.V. Average C.O.V. Average C.O.V.
(kPa) (%) (kPa) (%) (kPa) (%) (kPa) (%)
A-5 533.4 9.1 532.6 4.6 530.8 11.3 565.7 7.3
7 A-6 279.1 4.3 291.3 6.4 288.8 1.9 310.1 2.3
A-7 193.5 8.0 190.5 3.1 188.0 2.2 220.1 8.4
A-5 1050.1 7.1 1090.5 5.0 1048.3 4.1 1066.7 5.0
28 A-6 752.0 9.0 745.3 5.3 760.4 12.1 779.3 10.9
A-7 1039.1 5.4 1022.0 3.5 1085.3 3.6 1078.8 4.0
A-5 1339.5 3.4 1614.4 7.6 1529.3 8.4 1590.2 4.5
91 A-6 829.2 9.0 990.3 5.3 1079.7 12.1 1016.5 10.9
A-7 1871.9 5.4 1784.3 3.5 1863.7 3.6 1983.8 4.0
Table 5.19: Effects of loading rates (wide range) on deformation at peak load
0.085 %/min 0.25 %/min 0.58 %/min
Age Mixture Deformation C.O.V. Deformation C.O.V. Deformation C.O.V.
(days) (mm) (%) (mm) (%) (mm) (%)
A-1 0.711 23.9 0.673 2.7 0.803 20.2
7 A-2 4.252 16.3 3.459 15.9 4.252 2.3
A-3 0.960 23.5 0.892 28.6 0.765 13.4
A-4 1.521 13.4 1.547 19.8 1.402 5.1
A-1 0.572 13.5 0.566 11.8 0.566 6.4
28 A-2 1.466 13.3 1.801 17.2 1.537 11.8
A-3 0.665 2.3 0.861 12.3 0.848 5.1
A-4 0.648 1.4 0.683 20.9 0.691 14.4
A-1 0.716 0.7 0.597 1.7 0.638 1.9
91 A-2 1.920 0.4 2.042 2.7 1.880 3.4
A-3 0.909 1.8 0.925 1.3 0.851 2.6
A-4 0.777 1.7 0.843 4.7 1.031 3.9
170
Table 5.20: Effect of narrow range load rates (narrow range) on deformation
at peak load
0.16 %/min 0.25 %/min 0.33 %/min 0.42 %/min
Age Mixture Deformation C.O.V. Deformation C.O.V. Deformation C.O.V. Deformation C.O.V.(
(days) (mm) (%) (mm) (%) (mm) (%) (mm) %)
A-5 0.577 15.8 0.526 14.5 0.480 34.1 0.434 31.0
7 A-6 0.648 11.6 0.620 12.2 0.627 9.0 0.599 8.1
A-7 1.229 17.9 1.224 31.1 1.166 7.6 1.077 2.9
A-5 0.610 16.0 0.726 20.3 0.650 9.7 0.528 11.6
28 A-6 0.711 7.2 0.640 18.7 0.681 3.0 0.759 5.1
A-7 0.838 12.3 0.861 12.2 0.780 12.8 0.828 23.4
A-5 0.493 4.8 0.732 12.9 0.759 3.2 0.602 6.9
91 A-6 0.866 20.5 0.676 6.7 1.036 17.7 1.214 13.5
A-7 0.775 21.5 0.640 9.6 0.813 20.4 0.808 24.1
depending upon the type of mixture and the age of testing. Interestingly, some
mixtures (i.e., mixture A-2) were relatively insensitive to load rate, whereas others
were quite sensitive. There was no consistent trend among the mixtures, which
load rate. However, there were some general trends that were insightful, as
described next.
For testing at 7 and 28 days, the range of loading rates from 0.16 to 0.42
%/min was generally found to generate consistent results. The effects of loading
rate on strength were more significant at 91 days than at 7 or 28 days. Mixture A-2
(containing foundry sand) was an interesting case where the compressive strength
varied very little, even across a wide range of loading rates. This is advantageous
containing foundry sand were found to be large, and testing in some cases took
more than 10 minutes per sample when using a loading rate of 0.25 %/min. Higher
171
loading rates could thus be used for these types of mixtures to speed up the testing
strength with loading rate were to some extent mixture or material specific, a load
rate between 0.and 0.42 %/min may be selected for improved compressive strength
testing.
were profound, where at early ages (7 days), the deformations were somewhat
greater than those at later ages (28 and 91 days), as the specimens were more
peak load did not vary as much as the actual compressive strength values. This
suggests that the mode of failure was not greatly affected by loading rate
modifications. Thus, the range of loading rates between 0.16 to 0.42 % per minute
can be recommended for the compression test method, based mainly on selecting a
range that yielded accurate and repeatable strength values in a reasonable amount
of time.
no discernible size effects were observed. It should be noted that all cylinders were
tested at the same effective strain rate (0.25 %/min), using the same deflection-
172
controlled machine. There is no general trend found for the effects of specimen
size.
laboratories use larger capacity concrete compression machines under load rate
control. To assess the relative difference in strength values, a limited study was
capacity) under deflection control with a larger capacity machine (1780 kN Tinius
Olson) under load control. For instance, for specimens from mixture #22 at the age
of 7 days, the small machine yielded a strength of 0.50 MPa and a COV of 1.4%,
and the large machine yielded a compressive strength of 0.24 MPa and a COV of
values and significantly higher variations than those obtained with the smaller
machine. Thus, caution should be taken when using a large capacity machine
173
(Tinius Olson, 1780 kN capacity) under load control. The results from this
meet the-time-to-failure limits described in ASTM D 4832 (not less than two
minutes), a load rate of 6.9 kPa/sec was used for the large machine. In general, the
results obtained from the larger compression machines lack the accuracy in the
that suggested that temperature plays a major role in CLSM strength development.
Three curing temperatures (10 ºC, 21 ºC, and 38 ºC) and six CLSM mixtures (H-1
to H-6 in Table 3.16) were selected to study the strength gain of CLSM across a
range of practical construction conditions. After casting the specimens into plastic
174
molds, the cylinders were immediately transported to the appropriate temperature-
controlled chambers. After three days of storage in the chambers, half the cylinders
from each mixture were stripped or removed from the molds and returned to the
same chamber until the time of testing. This curing regime is referred to as “dry”
in subsequent discussions. The other half of the specimens from a given mixture
were kept inside the molds with the caps firmly on the top of the mold until the day
of testing (designated as “wet” curing). These cylinders were placed directly next
The results of compression testing at 7, 28, and 91 days are shown in Table
5.23. Also included in Table 5.23 are the moisture contents of specimens measured
the moisture content (or evaporable water content) and strength of CLSM. The test
results at 7, 28, and 91 days are given in Table 5.23 and illustrated in Figures 5.20
through 5.25.
conclusion that temperature and storage conditions play a major role in the
clear trends in behavior across all material types and mixture proportions, but
175
strength when cured at 38 ºC, whereas mixture H-2, containing Class F fly
ash, showed a 40% increase in strength for the same conditions. The CaO
content of fly ash is generally the most important variable affecting
compressive strength and should be taken into account when using CLSM,
especially when temperature extremes may be encountered.
176
Table 5.23: Compressive strength of specimens cured at different
temperatures at different ages
177
Figure 5.20: Compressive strength of mixture H-1 for different curing
conditions
178
Figure 5.22: Compressive strength of mixture H-3 for different curing
conditions
179
Figure 5.24: Compressive strength of mixture H-5 for different curing
conditions
180
• Drying (from an age of three to seven days) generally increased the 7-day
strength of CLSM, compared to keeping the cylinders in molds
continuously for 7 days. For 91-day testing, the opposite effect was
generally observed. That is, cylinders continuously left in sealed plastic
molds exhibited higher strengths than companion cylinders that were
allowed to dry after the first three days. The results at an intermediate age
of 28 days were not as well defined, and general trends were not evident.
• The measurement of the moisture content of cylinders at the time of testing
may provide useful information on its performance, especially when
subjected to drying conditions. In some cases, the reduction in moisture
content for cylinders cured outside of their molds at higher temperatures
was accompanied by microcracking and subsequent strength reductions.
• The standard curing methods might severely under-estimate CLSM strength
development in the field, especially in humid environments. On the
contrary, for the case of mixture H-4, the dry specimens at 21 ºC only
developed 25% of compressive strength estimated in the lab. When
choosing a CLSM mixture, its proportions should be carefully studied in
relation to its application environments, such as temperature and loss of
moisture.
• In general, mixtures without fly ash were less sensitive to curing
temperature and drying conditions. The predictive models described earlier
in this chapter were generally effective in predicting CLSM without fly ash,
but significant differences were evident when applying the predictive
models to fly ash mixtures. Thus, particular attention should be paid to
assessing the long-term strength gain of CLSM containing fly ash,
especially when high-calcium ashes are used in warmer climates.
• Field experience is essential in understanding the performance of local
materials in specific applications and climates. The experience gained is
181
particularly important in assessing long-term strength gain and its effects on
excavatability.
• An assessment of the on-site strength of CLSM should take into account
laboratory-obtained test results, but it should also take into account climatic
conditions. An understanding of material reactivity is helpful in
extrapolating laboratory results to field performance.
Using the I-series CLSM mixtures, the triaxial shear strength of several
CLSM mixtures was measured. The results, shown in Table 5.24, confirmed the
chemical bonding and internal frictional resistance. For the mixtures investigated
in the present study, the behavior was found to be a function of specific material
and mixture proportions, and the effects changed with increased curing time.
mixtures were calculated. For mixtures I-1, I-2, and I-3, the internal friction angles
and cohesion both increased with time (between 7 and 28 days). Their friction
angles at 28 days were in the range of a very dense granular soil, and the mixtures
behaved like dense sand, with lower residual strengths than ultimate strengths. For
internal friction angle, whereas for mixture I-5, an increase in cohesion was the
dominant factor. These mixtures (I-4 and I-5) had high air contents and exhibited
behavior similar to loose or uncompacted sands. It was interesting to note that for
mixture I-6, the friction angle decreased and cohesion greatly increased with time.
182
Figure 5.26 illustrates some of the more profound differences in CLSM behavior,
contrasting the stress-strain behavior of mixtures I-1 and I-4 samples at 28 days.
600
500
I-4
400
I-1
300
200
100
0
0 2 4 6 8 10
Strain (%)
Figure 5.26: Stress-strain curves of mixture I-1 and I-4 at 28 days (triaxial
compression testing with an effective confining pressure of 69
kPa)
183
5.3.3 Water Permeability
The water permeability (or hydraulic conductivity) for six CLSM mixtures
(I-series) are shown in Table 5.25. These permeability values (measured after 28
days of moist-curing) were in the range of silts. The results show that water-
cement ratio affected the coefficient of permeability greatly. Generally, the lower
the w-c ratio, the lower the permeability. Interestingly, the high air content of
mixture I-5 did not increase the permeability significantly, indicating that the
entrained air bubbles were not well connected. The testing performed on these six
mixtures indicates that CLSM can be easily assessed for water permeability using
184
results are shown in Table 5.26 and compared with permeability values of CLSM
increased permeability and others showing decreased permeability. This was most
likely due to the test set-up, which was designed to keep the samples intact, thus
intact (and confined), it may not have allowed for an accurate estimate of in-situ
permeability. Because the samples were confined, the expansion due to freezing
and thawing may have actually compacted the samples, resulting in an apparent
185
reduction in permeability. More work is needed to elucidate the effects of freezing-
shrinkage of CLSM, and only limited emphasis was placed in this project on the
topic. A method developed in Germany for flooring applications was used in this
study, with the results shown in Table 5.27. The temperature during the testing
period was 20 o C, and the relative humidity was approximately 60 to 65%. For
mixtures without air entrainment, most of the shrinkage apparently occurred during
the first day. Much of this may be attributable to early bleeding and subsidence,
which may result in a net volume decrease of the solid mixture. This method is
made to modify or optimize the method for CLSM. It may be possible that CLSM
does not exhibit enough strength (or stiffness) to effectively cause meaningful
of CLSM and to assess and recommend suitable test methods. Because the topic of
drying shrinkage was not identified as a critical issue for this study, it is not
186
Table 5.27: Drying shrinkage of selected CLSM mixtures
-6
Shrinkage strain (x 10 )
Time Mix #4 Mix #24 Mix #23 Mix #6 Mix #22r Mix #26
1 day 2260 2830 80 1440 90 10
2 days 2280 2850 80 1450 90 30
3 days 2280 2860 80 1450 90 50
4 days 2280 2860 80 1450 100 70
5 days 2300 2860 80 1460 100 100
6 days 2310 2880 80 1480 110 130
7 days 2330 2880 80 1500 120 160
2 weeks 2390 2930 160 1540 150 200
3 weeks 2410 2960 180 1590 150 190
4 weeks 2410 2980 180 1529 160 210
5 weeks 2410 2980 180 1600 160 220
6 weeks 2420 2960 180 1610 160 220
7 weeks 2410 2960 180 1600
testing was performed on assessing CBR values of CLSM. Standard crushed rock
has a CBR value of 100. The results of testing performed on six CLSM mixtures
are shown in Table 5.28. Most of the mixtures exhibited quite high CBR values,
with the exception of some fly ash mixtures and high-air mixtures. Based on the
values obtained, it can be concluded that the specific mixtures investigated would
function as suitable base or subbase material. More importantly, the results and
experience confirm that it is feasible to determine CBR values for CLSM using
187
5.3.6 Resilient Modulus
The resilient moduli of the six selected CLSM mixtures were measured and
the results are shown in Table 5.29. These values were magnitude higher than
typical soils. Based on the values obtained, it can be concluded that the specific
resilient modulus values for CLSM using equipment commonly used to evaluate
188
5.3.7 Excavation Study
This section summarizes the results of tests that are directly or indirectly
related to the excavatability of CLSM. Included are the initial findings from tests
conducted on selected CLSM mixtures (six from the original mixture series) and
indices. The use of the splitting tensile test as a potential index of excavatability
was investigated, and as such, the tensile results are presented in this section.
assessed for six of the original 38 mixtures by casting CLSM into 450-mm x 450-
assessed using a soil penetrometer, and these values were correlated with the
“walkability” or the time at which an average person can walk on the material with
a slight impression. It was found that the soil penetrometer values in the range of
4.32 kPa to 7.35 kPa correlated with the initial “walkability.” Long-term
excavatability was assessed for the six CLSM mixtures at an age of approximately
nine months using typical hand tools, including a shovel and a pick. Prior to
assessing the excavatability, the “stiffness” of the samples was measured using the
Gauge). For example, the lab-cured compressive strength of mixture #23 was quite
low, but the field-cured excavation box was not excavatable. It has been shown in
189
previous testing that laboratory-cured cylinders may not be accurate indicators of
temperatures in the field (as was the case for these samples). Also, the results
excavatability. Another example of lack of correlation was the fact that mixture
#24 had a higher “stiffness” than mixture #22r, yet it was much easier to excavate.
excavatability, including the assessment of other test methods and indirect indices,
as described next.
was performed. A wide range of CLSM mixtures (C-series) was included in the
190
• Dynamic cone penetrometer (DCP)
• CBR (estimated from DCP)
• Stiffness gauge (GeoGauge)
• Removability modulus (RE)
• Splitting tensile strength
191
0.6
0.5
R2 = 0.67
Compressive strength
0.4
(MPa)
0.3
0.2
0.1
0
0 2 4 6 8 10 12 14 16
Needle penetration (ASTM C 403) (MPa)
was a general trend that mixtures with higher compressive strength exhibited higher
penetration values, the correlation was not very strong. A possible explanation for
the lack of fit between the two data sets is that the data include all mixtures from
the C-series, even though some mixture types show different inherent relationships
between strength and penetration resistance. For example, paste mixtures exhibit
resistance can be developed for specific materials and commonly used mixture
192
proportions. However, care should be taken in trying to extrapolate trends
developed for one set of common mixtures (i.e., paste mixtures) to other mixture
Removability Modulus (RE) values. The table also shows the compressive strength
for lab-cured cylinders (at 28 days) and field-cured cylinders, which were cured
adjacent to the excavatability boxes and tested at the time of excavation (240 days).
The densities of the lab-cured and field-cured cylinders were measured before
testing them in compression, and these values were used in RE calculations. The
The DCP index values, which indicate the penetration depth per blow, were
measured for each of the excavation boxes. The minimum value for a recordable
found to decrease until the specimens ultimately suffered large cracks, as shown in
Figure 5.28. After the large cracks appeared, the DCP values progressively
increased. Thus, the lowest index value was taken for each mixture and used in
Table 5.32 because it represented the most difficult portion to excavate, thus
193
Table 5.32: Results of excavatability study
Mix 28-day Density 240-day RE using RE using DCP CBR Stiffness Relative
# compressive (kg/m3 ) compressive 28-day 240-day index (%) (using ease of
strength strength lab-cured field- (mm) GeoGauge) excavation
(lab-cured) (field-cured) cylinders cured (MN/m) (with
(kPa) (kPa) cylinders shovel)*
C-1 533 1512 1134 0.84 1.22 6.4 37 19.37 3
C-2 807 1946 493 1.51 1.18 6.5 36 40.56 7
C-3 144 1724 63 0.53 0.35 29.0 7 19.03 1
C-4 292 1858 447 0.85 1.05 5.2 46 28.97 6
C-5 1027 2094 3290 1.90 3.39 0.5 100 30.87 9
C-6 332 2252 362 1.20 1.25 5.0 48 18.54 8
C-7 1192 2143 1397 2.12 2.29 1.1 100 11.96 8
C-8 658 2171 1615 1.60 2.51 0.6 100 23.89 10
C-9 417 1660 199 0.85 0.59 10.0 22 23.03 4
*Note: The ease of excavation wa s compared by assigning different values to each mixture, where 1 is easiest and 10 is most
difficult.
The GeoGauge was used to assess the relative “stiffness” of the CLSM
specimens. As CLSM mixtures were quite strong (relative to soil), a thin layer of
wet fine sand was placed on the surface prior to testing, as per the
mixture. The variations were quite high for the device, with coefficients of
values and DCP values is shown in Figure 5.29. There was no clear trend between
stiffness values and DCP, nor was there a clear trend between stiffness values and
actual excavatability (by shovel). In general, the GeoGauge was not found to be an
The correlation between the DCP index and the two sets of calculated RE
values (based on 28-day lab-cured cylinders and 240-day field-cured cylinders) are
shown in Figure 5.30 and 5.31. The RE values calculated using field-cured
194
cylinders, tested at the time of excavation, showed a better correlation with DCP
index than RE values based on 28-day lab-cured cylinders. The RE values based
on 28-day lab-cured cylinders did, however, identify all the mixtures that were not
excavatable (i.e., their RE values were greater than 1). However, using the 28-day
data is quite conservative and may not be suitable for all CLSM mixtures.
195
45
40
35
30
25
20
15
10
5
0
0 5 10 15 20 25 30 35
Figure 5.29: Correlation between stiffness (from GeoGauge) and DCP index
corresponding California Bearing Ratio (CBR) values were obtained from the DCP
index. However, according to the manual, the highest possible CBR value (100%)
calculated was reached for some of the CLSM mixtures, making it difficult to use
the value quantitatively. Thus, the DCP index, as previously described, was used
with an RE value of 1.0. This suggests that the DCP may be an effective, user-
further investigation in field testing, where excavatability may not be assessed only
196
using hand tools, but also with typical, commercial excavation equipment (i.e.,
backhoe).
35
30
25
y = 5.2798x-2.557
R2 = 0.76
20
15
10
0
0.00 0.50 1.00 1.50 2.00 2.50
Figure 5.30: Correlation between DCP Index and Removibility Modulus (RE)
calculated using 28-day compressive strength (lab-cured
cylinders)
40
35
30
25 y = 5.4769x-1.8217
20 R 2 = 0.93
15
10
5
0
0.00 1.00 2.00 3.00 4.00
Figure 5.31: Correlation between DCP Index and Removibility Modulus (RE)
calculated using 240-day compressive strength (field-cured
cylinders)
197
Another parameter that potentially might be used as an index for
found that tensile strength may, in fact, be more suitable than compressive strength
the C-series mixtures, some tests were performed on other mixture series. The
results are provided in this section because of the potential of applying tensile data
to excavatability predictions.
shown in Table 5.33. A split specimen of mixture E-1 is shown in Figure 5.32. For
the E-series CLSM mixtures, the splitting tensile strength to compressive strength
ratio ranged from 9 to 17 percent, which is higher than those typically observed for
conventional concrete. Unlike concrete, this ratio did not substantially decrease
198
Figure 5.32: A cylinder from mixture E-1, before and after being tested for
splitting tensile strength
Additional splitting tensile tests were performed using the H-series mixtures
to assess the effects of drying on tensile strength and the tensile to compressive
strength ratio. This testing was initiated because drying generally has a more
profound effect on tensile strength than compressive strength, at least ni the case of
microcracks. The results, shown in Table 5.34, confirmed that drying had a similar
The effects of temperature and drying on the tensile strength of CLSM, especially
research. It is expected that further research will help to classify some of these
effects and to determine the potential for using splitting tensile strength (by itself or
199
Table 5.34: Effects of temperature and drying conditions on splitting tensile
strength of CLSM
Mixture H-1 H-2
Condition Average C.O.V. ft /fc Average C.O.V. ft /fc
(kPa) (%) (%) (kPa) (%) (%)
10 ºC, dry 16.7 23.1 6.5 6.0 20.2 5.1
10 ºC, wet 25.6 19.3 7.9
21 ºC, dry 23.8 8.6 4.9 8.8 39.6 6.1
21 ºC, wet 89.0 27.9 14.1
38 ºC, dry 55.0 18.8 7.3 18.9 11.4 7.2
38 ºC, wet 74.5 6.1 8.1 53.4 15.5 8.4
Mixture H-3 H-4
Condition Average C.O.V. ft /fc Average C.O.V. ft /fc
(kPa) (%) (%) (kPa) (%) (%)
10 ºC, dry 100.4 14.1 7.2 65.5 33.0 7.3
10 ºC, wet 166.3 32.1 12.2 95.8 101.4 5.7
21 ºC, dry 49.2 8.5 5.8 32.2 5.0 7.0
21 ºC, wet 114.5 15.5 12.4 455.9 36.0 12.2
38 ºC, dry 87.8 7.5 8.4 158.6 6.3 7.7
38 ºC, wet 525.4 30.9 13.5 1791.7 12.5 15.6
Mixture H-5 H-6
Condition Average C.O.V. ft /fc Average C.O.V. ft /fc
(kPa) (%) (%) (kPa) (%) (%)
10 ºC, dry 28.8 12.3 6.8 56.1 67.5 8.4
10 ºC, wet 133.0 7.5 9.8 84.5 22.8 9.2
21 ºC, dry 25.6 16.6 7.7 64.1 5.3 17.2
21 ºC, wet 127.8 8.6 8.5 142.8 7.3 15.4
38 ºC, dry 106.8 9.1 8.0 114.1 16.6 14.6
38 ºC, wet 198.2 6.0 7.5 125.3 12.5 12.6
200
5.4 DURABILITY OF CLSM
thawing resistance of CLSM. The first study involved testing six CLSM mixtures
from the original mixture series using a modified version of ASTM D 560, as
described in Chapter 4. This method was originally developed for the assessment
of soil-cement. The second study used the same method to assess a wider range of
5.35 and 5.36 for samples cured for 7 days and 28 days, respectively. According to
the “soil-cement laboratory handbook”, the maximum allowable weight loss is 14%
criterion, most CLSM mixtures with high air content behaved well, especially after
material. There has been very little field evidence suggesting that this is a problem
with CLSM, but research was needed, as a minimum, to assess the laboratory
201
Table 5.35: Freezing-and-thawing results for selected CLSM mixtures (after 7
days of curing)
Mixture # 4 24 23 6 22r 26
Initial weight (g) 2241 1639 1441 2085 1861 1868
Initial moisture content (%) 10.86 37.78 30.37 14.79 10.38 9.17
1 cycle (% of initial weight retained) 100.1 101.2 102.1 100.1 100.5 100.4
2 cycles (% of initial weight retained) 100.3 102.3 100.6 99.9 100.4 100.2
3 cycles (% of initial weight retained) 100.9 103.3 103.0 97.8 100.2 100.5
4 cycles (% of initial weight retained) 101.2 103.2 100.9 95.4 100.3 100.3
5 cycles (% of initial weight retained) 101.3 103.0 103.3 92.6 100.1 100.4
6 cycles (% of initial weight retained) 101.3 99.2 101.3 91.3 100.1 100.5
7 cycles (% of initial weight retained) 97.7 Damaged Damaged 88.6 100.2 100.5
8 cycles (% of initial weight retained) 92.4 88.7 100.2 100.4
9 cycles (% of initial weight retained) 87.0 88.3 100.2 100.6
10 cycles (% of initial weight retained) 80.2 88.2 100.1 100.4
11 cycles (% of initial weight retained) 76.7 87.7 100.1 100.4
12 cycles (% of initial weight retained) 73.7 87.2 100.0 100.4
Moisture content at end of test (%) 13.35 15.92 11.16 6.97
Mass loss (%) 27.9 13.7 0.7 -2.5
202
The freezing-and-thawing test results of the D-series CLSM mixtures are
shown in Table 5.37. It is interesting to note that D-1 mixture (high-air mixture
with 30 kg/m3 of cement) did not pass the twelve cycles, whereas a similar mixture
with 45 kg/m3 of cement (D-11) did pass the entire 12 cycles, suggesting that both
Mixture D-10 survived all 12 cycles, most likely due to its higher strength
(contributed from the Class C fly ash). The remaining mixtures (D-2 through D-9)
did not survive all 12 cycles. Mixtures containing foundry sand (D5 to D7) were
damaged after only one or two cycles. Figure 5.33 shows damaged CLSM samples
203
Figure 5.33: Specimens after freezing-and-thawing cycles (CLSM mixtures
containing foundry sand)
cycles at the ages of 7 and 28 days respectively. In this study, the samples were
dried at room temperature, approximately 20 °C. Table 5.38 and 5.39 show the
results. In these tables, from 1 to 11 cycles, the weights of dry samples before
being submerged in water were expressed as a percentage of the weights before the
204
wetting-and-drying cycles. After 12 cycles, the weight loss was expressed in terms
#23, cured for 7 and 28 days, performed satisfactorily throughout the wetting-and-
drying cycles. Mixtures 22r and 26 gained weight when exposed to wetting-and-
drying cycles at 7 days. This was probably due to the continuing hydration of
portland cement. The mass loss of mixture #23 was significantly reduced when it
was tested after 28 days of curing when compared to 7 days of curing. This
damage.
205
Table 5.39: Wetting and drying results for selected CLSM mixtures (after 28
days of curing)
Mixture # 4 24 23 6 22r 26
Initial weight (g) 2261.1 1646.9 1472.3 2134 1790.1 1860.5
Initial moisture content (%) 0.1072 0.3709 0.2875 0.1538 0.09788 0.08642
1 cycle (% of initial weight retained) 96.4 97.4 97.4 99.8 99.0 98.6
2 cycles (% of initial weight retained) 95.4 94.8 99.5 88.6 96.5 96.7
3 cycles (% of initial weight retained) 95.5 95.3 100.0 96.6 97.9 98.0
4 cycles (% of initial weight retained) 95.2 92.3 95.7 95.3 97.5 97.3
5 cycles (% of initial weight retained) 97.6 36.0 99.5 96.4 95.8 95.9
6 cycles (% of initial weight retained) 96.2 75.7 79.2 96.5 95.3 95.7
7 cycles (% of initial weight retained) 94.0 93.4 98.2 95.3 98.3 98.2
8 cycles (% of initial weight retained) 97.6 92.1 96.5 95.6 97.8 97.7
9 cycles (% of initial weight retained) 91.2 93.2 98.5 96.2 97.7 97.8
10 cycles (% of initial weight retained) 96.9 93.1 97.8 96.1 97.2 97.0
11 cycles (% of initial weight retained) 94.8 92.7 94.9 95.8 97.3 97.9
12 cycles (Mass loss, %) 0.2 3.6 5.8 0.0 0.0 0.0
Moisture content at end of test (%) 7.6 33.2 33.8 11.0 7.1 6.1
5.5.1 Introduction
values of CLSM mixtures were measured. The chloride penetration was illustrated
in Figure 5.34. Uncoupled and coupled ductile-iron specimens were exposed to the
3.5% NaCl solution after 28 days of curing. The mass loss of steel specimens was
measured after being exposed to the NaCl solution. Because in cases where the
dissolved oxygen in the water entrapped in the soil, the composition of carbon steel
206
has little effect on corrosion resistance in a given soil. Thus, the composition of the
extruded solutions were found to vary with time, with no general trend. A
mixtures whose pore solutions had pH values less than 12.5 likely had a lower
influencing the pH value of the pore solution include the amount of calcium
Class C fly ash, alkalis from raw materials, the consumption of calcium hydroxide
The most corrosive soils are those that contain large concentrations of
soluble salts. Because of the presence of salts, such soils have relatively high
207
CLSM mixtures were measured using the approach specified in Chapter 4. The
182-day results are shown in Table 5.40. The moisture content of a soil also affects
its resistivity. In this aspect, the saturated condition of CLSM mixtures, as required
by the test method in this study, represents the worst-case scenario when no
208
But during testing, chloride ions were present and, as such, it may be more
same solution as that used for mass loss testing. However, resistivity alone is not a
For underground iron pipes, the rate of corrosion can be accelerated by the
aggressive ions at the pipe location. This study evaluated five CLSM mixtures for
chloride ingress rate. CLSM cylinders (150 mm x 300 mm) were ponded for 22
weeks with a 3.5% chloride solution. Samples were then tested for chloride
contents as a function of cylinder depth. Figure 5.34 shows results from the testing
program. These results indicate that chloride ions easily migrate through CLSM,
209
Chloride Profiles for CLSM Mixtures Ponded with 3.5% Chloride Solution
0
50
Depth from top of cylinder (mm)
Mix 4
Mix 6
100 Mix 24
Mix 22R
Mix 26
150
200
250
300
0 10 20 30 40 50 60 70
3
Total chloride content (kg/m )
Figure 5.34: Chloride profiles for select CLSM mixtures after 22 weeks of
exposure.
Visual observations during the testing program indicated that the ponded
solution was migrating through the CLSM at relatively high rates. It was observed
that approximately 650 mL was permeating through the sample from mixture #23
of solution through the sample, ponding was stopped and mixture #23 was not
It was observed that all other samples exhibited wetting of the cylinder
sides as a result of ponding the chloride solution on the top of the samples. It is
likely that permeation (i.e. flow as a function of a pressure head) of the chloride
210
solution into the CLSM samples was the predominate mode of chloride ingress.
conditions. The test set-up was explained in Chapter 4. As is the case for most
laboratory corrosion tests, the results should be interpreted with caution. This is
especially true for evaluating corrosion of metals in soils, where there are inherent
may also influence the corrosion characteristics of steel pipelines. Thus, the results
should also be evaluated to assess corrosion rates and provide guidelines for future
applications. The drawback of field testing is the length of time required for
consuming, field tests and accelerated, but preliminary, laboratory tests. Thus, the
ASTM G 51, “Standard Practice for Preparing, Cleaning, and Evaluating Corrosion
211
embedded in CLSM only are shown in Table 5.41. Significant variations in mass
loss existed between different mixtures, repeated mixtures, and samples within one
mixture.
Several factors might have contributed to the variations in test results. First,
the variations were possibly due to the inherent nature of CLSM. Because of the
lack of coarse aggregates, it is difficult to break cement lumps and disperse the
specimens, such as 75-mm x 150-mm cylinders for compression, were tested. For
corrosion coupons, due to their small size and corresponding small surface areas,
slight changes of local CLSM conditions may significantly affect the corrosion
results. Local areas containing higher cement and fly ash contents may show
higher local pH values and suffer less corrosion. Also, the properties of the
transition zone between CLSM and coupons may affect the corrosion behavior.
transition zone exists and may play an important role. For example, the existence
of entrapped air may affect the corrosion behavior by increasing the available
chloride ions and dissolved oxygen. Another potential factor leading to variations
is the actual measurement itself. Because of the very small values of percentage of
mass loss, a very small error may significantly influence the results. Although steel
specimen surfaces are prepared to meet certain requirements, variations are still
expected. The different surface conditions of steel specimens could also affect the
212
Table 5.41: Percent mass loss for samples embedded in CLSM only
Sample ID Specimen No. Average
1 2 3
1 0.05 0.05 0.05
1R 0.04 0.05 0.05 0.04
2 0.02 0.02 0.05 0.03
2R 0.03 0.01 0.02 0.02
3 0.01 0.01 0.02 0.01
3R 0.01 0.02 0.06 0.03
4 0.03 0.02 0.02 0.02
4R 0.01 0.01 0.01 0.01
5 0.03 0.05 0.05 0.04
5R 0.04 0.03 0.02 0.03
6 0.01 0.02 0.03 0.02
7 0.03 0.02 0.02 0.02
8 0.02 0.00 0.01
9 0.03 0.02 0.02 0.03
11 0.05 0.01 0.01 0.02
12 0.03 0.04 0.04
13 0.02 0.02 0.13 0.06
14 0.02 0.04 0.02 0.02
15 0.02 0.03 0.02 0.02
16 0.19 0.22 0.15 0.19
16R 0.10 0.12 0.09 0.10
17 0.28 0.72 0.50
18 0.13 0.37 0.25
19 0.62 0.35 0.48 0.48
29 0.03 0.03 0.02 0.02
30 0.03 0.01 0.31 0.12
22 0.37 1.02 0.17 0.52
22R 0.08 0.09 0.11 0.09
23 2.10 5.80 4.60 4.17
24 0.04 1.30 0.67
25 0.03 0.03 0.94 0.33
26 0.11 0.07 0.09 0.09
27 0.06 0.12 0.09
28
20 0.03 0.17 0.03 0.08
21 5.85 1.95 2.85 3.55
213
Because of the variations within a given mixture, the results from a repeated
mixture may be quite different or close to the original mixture. This is because
means from two different populations are compared, which are random variables.
denoted as µ1 and µ2 , the null hypothesis for a test on two independent means
would be:
H0 : µ1 = µ2
HA: µ1 ≠ µ2
Step 2: Select the appropriate model. For the case of two independent
samples, the hypotheses of step 1 can be tested using the following test statistic:
X1 − X 2
t= (5. 6)
1 1
s ( + ) 0. 5
n1 n 2
in which X1 and X 2 are the means of the samples from populations 1 and 2,
respectively; n1 and n2 are the samples sizes for the samples from populations 1 and
degrees of freedom of (n1 +n2 -2); and s is the square root of the pooled variance
given by
(n1 − 1) s12 + ( n 2 − 1) s 22
s2 = (5. 7)
n1 + n2 − 2
214
where s12 and s22 are the variances of the samples from populations 1 and 2,
respectively. This test statistic assumes that the variances of the two populations
Step 4: Compute estimate of test statistic. Samples are drawn from the
two populations and the sample means and variances are drawn. In this study,
three samples are drawn per population. The statistic is then calculated to test the
function of the degrees of freedom (n1 + n2 –2), the level of significance, and the
statement of the alternative hypothesis. In this case, the region of rejection is:
Step 6: Select the appropriate hypothesis. If the sample value lies in the
shown in Table 5.42. It is estimated that four out of seven repeated mixtures did
not yield the same corrosion mass losses as the original mixtures. There were two
possible causes for the differences between repeated and control mixtures. One
cause was that the mass loss was very sensitive to small changes of CLSM
ingredient materials or/and mixing conditions, which is different for every mixture.
The other was due to the small sample sizes. As already pointed out, the within-
batch variations were found to be significant. Thus, larger sample sizes (> 3) are
215
required to represent the whole population in corrosion test. Further study is
needed to determine which factor is more dominant or whether both are equally
important.
The mass loss of samples in different CLSM mixtures was found to vary
significantly. Many factors may affect the corrosion behavior of metal in different
entrained air contents from 15 to 30% showed much higher corrosion rates than
216
Constituent materials of CLSM mixtures may also influence the corrosion
Models (GLM) Procedure was used because of its ability to handle unbalanced
utilized to evaluate the significance of factors. A p-value was calculated for each
F-statistic and the factor was determined significant if the p-value was less than a
0.05 significance level. This procedure can also cope with cases of missing data.
As mentioned above, the use of fly ash may significantly change the corrosion
separately for mixtures with and without fly ash. Factors analyzed include cement
content (two levels), fly ash type (three levels), fly ash content (two levels),
aggregate type (three levels), and water-cement ratio (continuous variable). The
Results in Table 5.43 indicate that there was no single factor significantly
affecting the corrosion of samples in CLSM mixtures with fly ash. However, the
interaction between cement content and fly ash type did seem to play an important
role. For instance, samples in mixture #11 (60 kg/m3 cement and 360 kg/m3 high-
carbon fly ash) corroded less than those in mixture #1 (30 kg/m3 cement and 180
kg/m3 Class C fly ash). Thus, when choosing CLSM mixture proportions,
precautions should be taken to evaluate combinations of cement content and fly ash
type.
217
For CLSM mixtures with entrained air (without fly ash), air content and
aggregate type were found by one test to significantly affect the corrosion rate of
embedded samples. It is worth noting that air content may influence corrosion
rates. One test in Table 5.44 indicates that the air content could affect corrosion
rate. More study is required to further assess these factors. The use of different
aggregates (concrete sand and bottom ash) may also affect the ingress of chloride
ions by permeability.
Table 5.43: Statistical results of mass-loss test for uncoupled samples (CLSM
mixtures containing fly ash)
Source DF Type I SS Mean Square F Value Pr > F
Cement content 1 0.00021188 0.00021188 0.62 0.4335
Fly ash type 2 0.00109877 0.00054939 1.62 0.2091
Fly ash content 1 0.00045676 0.00045676 1.35 0.2519
Aggregate 2 0.00134288 0.00067144 1.98 0.1497
Water-cement ratio 1 0.00003464 0.00003464 0.10 0.7508
Cement * fly ash type 2 0.00224639 0.00112319 3.31 0.0453
Source DF Type II SS Mean Square F Value Pr > F
Cement content 1 0.00022344 0.00022344 0.66 0.4213
Fly ash type 2 0.00115231 0.00057616 1.70 0.1942
Fly ash content 1 0.00022754 0.00022754 0.67 0.4171
Aggregate 2 0.0097992 0.00048996 1.44 0.2464
Water-cement ratio 1 0.00000012 0.00000012 0.00 0.9852
Cement * fly ash type 2 0.00224639 0.00112319 3.31 0.0453
Source DF Type III SS Mean Square F Value Pr > F
Cement content 1 0.00001570 0.00001570 0.05 0.8306
Fly ash type 2 0.00141477 0.00070739 2.08 0.1358
Fly ash content 1 0.00022754 0.00022754 0.67 0.4171
Aggregate 2 0.0097992 0.00048996 1.44 0.2464
Water-cement ratio 1 0.00000012 0.00000012 0.00 0.9852
Cement * fly ash type 2 0.00224639 0.00112319 3.31 0.0453
Source DF Type IV SS Mean Square F Value Pr > F
Cement content 1 0.00001570 0.00001570 0.05 0.8306
Fly ash type 2 0.00141477 0.00070739 2.08 0.1358
Fly ash content 1 0.00022754 0.00022754 0.67 0.4171
Aggregate 2 0.0097992 0.00048996 1.44 0.2464
Water-cement ratio 1 0.00000012 0.00000012 0.00 0.9852
Cement * fly ash type 2 0.00224639 0.00112319 3.31 0.0453
Note that “*” means interaction.
218
Table 5.44: Statistical results of mass-loss test for uncoupled samples (CLSM
mixtures containing no fly ash)
Source DF Type I SS Mean Square F Value Pr > F
Cement content 1 0.17285714 0.17285714 0.10 0.7584
Air content 1 20.88580011 20.88580011 11.72 0.0024
Aggregate 1 18.73837547 18.73837547 10.52 0.0037
Water-cement ratio 1 0.01619013 0.01619013 0.01 0.9249
Cement * aggregate 1 0.71627884 0.71627884 0.40 0.5326
Source DF Type II SS Mean Square F Value Pr > F
Cement content 1 0.09839648 0.09839648 0.06 0.8164
Air content 6.74734560 6.74734560 3.79 0.0645
Aggregate 1 1.75261123 1.75261123 0.98 0.3321
Water-cement ratio 1 0.60023984 0.60023984 0.34 0.5675
Cement * aggregate 1 0.71627884 0.71627884 0.40 0.5326
Source DF Type III SS Mean Square F Value Pr > F
Cement content 1 0.27679692 0.27679692 0.16 0.6973
Air content 1 6.74734560 6.74734560 3.79 0.0645
Aggregate 1 0.00236750 0.00236750 0.00 0.9713
Water-cement ratio 1 0.60023984 0.60023984 0.34 0.5675
Cement * aggregate 1 0.71627884 0.71627884 0.40 0.5326
Source DF Type IV SS Mean Square F Value Pr > F
Cement content 1 0.27679692 0.27679692 0.16 0.6973
Air content 1 6.74734560 6.74734560 3.79 0.0645
Aggregate 1 0.00236750 0.00236750 0.00 0.9713
Water-cement ratio 1 0.60023984 0.60023984 0.34 0.5675
Cement * aggregate 1 0.71627884 0.71627884 0.40 0.5326
Comparing mixture 22r and 26, 24 and 27 suggests that the addition of an
accelerating admixture did not change the corrosion performance of pipe coupons
in CLSM. Because only two pairs of mixtures were tested, this finding is only
preliminary.
Table 5.45. In this testing, a galvanic cell was formed as a result of pipe coupons
219
being embedded in different environments, which were electrically connected to
complete the circuit for corrosion current. A 10-Ω resister was used to obtain
repeatability of CLSM mixtures with regard to corrosion. The results are shown in
Table 5.46. Five of six mixtures were found to be statistically repeatable. The
repeatability of mass loss of coupled samples was much higher than those of
functioned as a cathode and the coupon in sand as anode. The lower variation was
evident for the within-batch and between-batch variations of results provided here.
A t-test indicated that the means of mass loss for coupled samples
embedded in sand were not significantly different for CLSM mixtures with and
without fly ash content. This was probably because current flow resulting from the
analyzed separately for CLSM with and without fly ash. The results are shown in
220
Table 5.45: Percent mass loss for coupled samples embedded in sand
Sample ID Specimen No. Average
1 2 3
1R -- -- 1.68 1.68
2 2.42 1.48 1.80 1.90
2R 2.48 1.52 1.80 1.93
3 3.70 -- 5.70 4.70
3R 1.90 1.25 1.55 1.57
4 2.35 1.45 1.85 1.88
4R 3.50 2.20 1.50 2.40
5 1.50 2.65 2.45 2.20
5R 2.65 2.00 3.58 2.74
6 1.75 2.85 4.65 3.08
7 4.60 5.95 1.50 4.02
8 2.80 -- 3.10 2.95
9 1.42 0.18 -- 0.80
11 2.70 7.60 1.60 3.97
12 1.97 1.78 2.35 2.03
13 1.25 1.75 0.10 1.03
14 2.25 1.84 1.60 1.90
15 4.20 3.60 4.00 3.93
16 2.80 8.30 12.00 7.70
16R 3.00 6.25 1.30 3.52
17 3.90 3.50 4.95 4.12
18 2.80 2.95 1.25 2.33
19 5.60 4.70 -- 5.15
29 2.10 1.60 3.44 2.38
30 1.80 1.44 1.61 1.62
22 2.90 2.70 4.25 3.28
22R 1.20 2.35 2.30 1.95
23 2.80 2.60 -- 2.70
24 11.60 2.80 5.30 6.57
25 4.00 4.65 2.00 3.55
26 3.25 2.55 2.55 2.78
27 6.00 2.65 -- 4.33
28 5.50 5.40 0.40 3.77
20 2.90 4.00 5.80 4.23
21 2.30 1.58 1.30 1.73
221
Table 5.46: Statistical comparison of repeated CLSM mixtures on corrosion
of coupled coupons embedded in sand
Population Population Estimated Critical value Accepted
# #r X1 X2 S2 statistic (t) (t), hypothesis
α=0.05
2 2r 1.90 1.93 0.24 -0.084 2.776 Null
3 3r 4.70 1.57 0.74 3.998 3.182 Alternative
4 4r 1.88 2.40 0.62 -0.806 2.776 Null
5 5r 2.20 2.74 0.50 -0.937 2.776 Null
16 16r 7.70 3.52 13.88 1.375 2.776 Null
22 22r 3.28 1.95 0.57 2.169 2.776 Null
For CLSM mixtures with fly ash, all four statistical tests indicated that
Three out of four tests showed that cement contents and aggregate type were also
important factors.
For CLSM mixtures without fly ash, no single factor or interaction was
the case of uncoupled corrosion, air content did not play a key role here. Only one
test indicated that cement content affects the corrosion behavior. Further study is
222
Table 5.47: Statistical results of mass-loss test for coupled samples (CLSM
mixtures containing fly ash)
Source DF Type I SS Mean Square F Value Pr > F
Cement content 1 1.83194087 1.83194087 0.11 0.7402
Fly ash type 2 22.93176577 11.46588288 0.70 0.5036
Fly ash content 1 23.10524882 23.10524882 1.41 0.2429
Aggregate 2 62.42531783 31.21265891 1.90 0.1633
Water-cement ratio 1 99.80592246 99.80592246 6.08 0.0183
Source DF Type II SS Mean Square F Value Pr > F
Cement content 1 84.9477888 84.9477888 5.17 0.0287
Fly ash type 2 0.6473494 0.3236747 0.02 0.9805
Fly ash content 1 21.3668191 21.3668191 1.30 0.2611
Aggregate 2 162.0849627 81.0424813 4.94 0.0124
Water-cement ratio 1 99.80592246 99.80592246 6.08 0.0183
Source DF Type III SS Mean Square F Value Pr > F
Cement content 1 84.9477888 84.9477888 5.17 0.0287
Fly ash type 2 0.6473494 0.3236747 0.02 0.9805
Fly ash content 1 21.3668191 21.3668191 1.30 0.2611
Aggregate 2 162.0849627 81.0424813 4.94 0.0124
Water-cement ratio 1 99.80592246 99.80592246 6.08 0.0183
Source DF Type IV SS Mean Square F Value Pr > F
Cement content 1 84.9477888 84.9477888 5.17 0.0287
Fly ash type 2 0.6473494 0.3236747 0.02 0.9805
Fly ash content 1 21.3668191 21.3668191 1.30 0.2611
Aggregate 2 162.0849627 81.0424813 4.94 0.0124
Water-cement ratio 1 99.80592246 99.80592246 6.08 0.0183
#29 and #30 showed that higher cement content dramatically increased the
corrosion rate of the anode in this study. There was not enough information to
environments, where one environment is CLSM and the other is sand, indicates that
placing pipe in these different environments while coupled, could lead to increased
corrosion.
223
Table 5.48: Statistical results of mass-loss test for coupled samples (CLSM
mixtures without fly ash)
224
toxicity of the by-product materials used in this study, including bottom ash,
foundry sand, and fly ash (Class C, Class F, and high carbon).
The chemical analysis, heavy metal test, and TCLP test on raw materials
dissertation, a systematic approach (shown earlier in Figure 4.13) was taken to first
study, then to quantify the total heavy metals present in these materials, and lastly
to measure the heavy metals that actually are able to leach from these materials.
The chemical compositions of the fly ashes and foundry sand were already reported
in Table 3.3 and Table 3.4. This information is particularly important for fly ashes,
where the chemical composition (especially CaO and SiO 2 content) can be used to
the raw by-product materials used in this study. These results represent the total
concentration of the eight key heavy metals. A “rule of thumb” that some
times the standard TCLP limits.26 According to this guideline, bottom ash, Class C
fly ash, and Class F fly ash exceeded the “rule of thumb” value for arsenic. Thus,
additional testing was performed (using the TCLP method) to determine the actual
amount of heavy metals that are available to leach from these materials. Because
the foundry sand and high-carbon fly ash did not have significant amounts of total
225
heavy metals, the materials were classified as non-toxic, and no subsequent
The TCLP results for Class C fly ash, Class F fly ash, and bottom ash (raw
materials) are shown in Table 5.50. The concentration of heavy metals that leached
from each material was well below the EPA-recommended TCLP limits, and as
such, the materials were classified as non-toxic and suitable for use in CLSM. If
any of the by-product materials had exhibited significant leaching of heavy metals
(above the TCLP limits), the last step would have been to assess the actual leaching
of heavy metals from CLSM containing the material(s). This systematic approach
can be followed for any material being considered for use in CLSM. Although all
the materials used in this study were deemed non-toxic, it may be possible that
certain materials considered for a given CLSM application may be more of an
environmental concern.
226
Table 5. 50: TCLP test results for Class C fly ash, Class F fly ash, and bottom
ash (mg/L)
Element TCLP limits Bottom ash Class C fly ash Class F fly ash
Arsenic 5.0 0.12 0.074 0.37
Barium 100.0 3.61 0.30 0.17
Cadmium 1.0 0.001 0.004 0.024
Chromium 5.0 0.01 0.29 0.11
Lead 5.0 <0.01 <0.01 <0.01
Mercury 0.2 <0.01 <0.01 0.11
Selenium 1.0 <0.01 0.37 0.02
Silver 5.0 <0.01 <0.01 <0.01
227
Chapter 6: Conclusions and Future Work
6.1 CONCLUSIONS
CLSM mixtures. A wide range of CLSM mixtures were included in this research
materials, such as fly ash, bottom ash, and foundry sand, were used. Suitable
existing test methods were examined and modified in this study to compensate for
the lack of CLSM test methods. Existing geotechnical test methods, such as
triaxial shear and resilient modulus, were included in the research to assess their
1. Source of aggregates i.e. foundry sand, bottom ash, and river sand is
2. The foundry sand and bottom ash used in this study led to difficulties
good approach to measure the air contents and unit weights of fresh
CLSM mixtures;
228
4. Both the needle (ASTM C 403) and pocket soil penetrometer can be
are different;
5. The initial and final setting concepts, used in concrete technology, may
6. For CLSM mixtures with high air content (without fly ash), the water-
only the water-cement ratio, but also the fly ash type, fly ash content,
9. Testing machines with smaller, more accurate load cells yield more
229
10. The air drying of CLSM specimens for 4 to 8 hours, as specified by
testing;
12. Curing temperature and humidity can significantly affect the strength
curing conditions;
13. The inclusion of Class C fly ash in CLSM mixtures typically yields
230
that appropriate conditions should be clearly specified for the mixtures
sulfur. However, the capping procedure may take more than half an
16. Neoprene pads with hardness durometer less than 50 were found
the in-situ strength of CLSM mixtures. The DCP index correlates well
21. Measuring the California Bearing Ratio (CBR) and Resilient Modulus
231
22. The water permeability of CLSM mixtures were found to be in the
range of silts;
check the saturation of samples, the requirement that B > 0.95 can be
25. Beyond the first day, the drying shrinkage of CLSM mixtures may be
26. For CLSM specimens cured in the fog room in this study, the pH
27. The resistivity of CLSM mixtures after 182 days ranged from 1781 to
232
30. For the uncoupled conditions, no single factor was found to
uncoupled condition;
32. For the coupled conditions, corrosion rates of specimens in sand with
CLSM containing fly ash were affected by the water-cement ratio and
if cured only for seven days, provided that the samples are allowed to
air dry;
233
1. Evaluate the feasibility of ASTM D 6023-96, “Standard Test Method
humidity;
234
9. More severe wetting-and-drying tests should be performed on CLSM
235
Appendix A: Referenced Standard Test Methods
ASTM C 128-1997, “Standard Test Method for Specific Gravity and Absorption
of Fine Aggregate”
ASTM C 136-1996a, “Standard Test Method for Sieve Analysis of Fine and
Coarse Aggregates”
ASTM C 231-1997, “Standard Test Method for Air Concrete of Freshly Mixed
Concrete by the Pressure Method”
ASTM C 311-1998b, “Standard Test Method for Sampling and Testing Fly Ash
or Natural Pozzolans for Use as a Mineral Admixture in Portland-Cement
Concrete”
ASTM C 618-1998, “Standard Specification for Coal Fly Ash and Raw or
Calcined Natural Pozzolan for Use as a Mineral Admixture in Concrete”
236
ASTM C 939-1997, “Flow of Grout for Preplaced-Aggregate Concrete (Flow
Cone Method)”
ASTM D 559-1996, “Standard Test Methods for Wetting and Drying Compacted
Soil-Cement Mixtures”
237
ASTM D 6023-1996, “Standard Test Method for Unit Weight, Yield, and Air
Content (Gravimetric) of Controlled Low Strength Material”
ASTM D 6024-1996, “Test Method for Ball Drop on Controlled Low Strength
Material to Determine Suitability for Load Application”
ASTM G 51-1999, “Standard Test Method for Measuring pH of Soil for Use in
Corrosion Testing”
AASHTO T 24-1997 “Obtaining and Testing Drilled Cores and Sawed Beams of
Concrete”
238
AASHTO T 161-2000, “Standard Test Method for Resistance of Concrete to
Freezing and Thawing”
239
Appendix B: Setting/Hardening Test Results
Table B-1: Needle penetration (NP) and soil penetrometer (SP) results
240
Table B-1: Needle penetrAtion (NP) and soil penetrometer (SP) results
(cont’d)
241
Table B-1: Needle penetration (NP) and soil penetrometer (SP) results
(cont’d)
242
Table B-2: Needle penetration (NP), soil penetrometer (SP) and vane shear
tester (VS) Results
243
References
3. Funston, J. J., Krell, W. C. and Zimmer, F. V., “Flowable Fly Ash, A New
1234, 1989.
5. Krell, W. C., “Flowable Fly Ash,” Concrete International, Vol. 15, No. 7,
1989.
(Flowable Fill), Howard, A. K. and Hitch, J. L., eds, STP 1331, 1998.
10. Bhat, S. T., Use of Coal Combustion Residues and Foundry Sands in
244
11. Burns, F., Flowable Fly Ash Backfill for Buried Pipelines, M. S. Thesis,
12. Hagerty, J. L., Investigation and Testing of Controlled Low Strength Material
Louisville, 1989.
13. Lucht, D. A., Thermal Performance of Flowable Fill Mixtures for Horizontal
17. Nantung, T. E., Design Criteria for Controlled Low Strength Materials, Ph. D.
20. Brewer, W. E., “The Design and Construction of Small Span Bridges and
245
21. Lundvall, J. F., “Mitigation of Roadway Settlement above Buried Culvert and
Transportation, 1997.
23. AASHTO, “A Look at What Some States Are Doing,” AASHTO Quarterly,
25. Folliard, K. J.,Trejo, D., Du, L., and Sabol, S. A., “Controlled Low-Strength
Material for Backfill, Utility Bedding, Void Fill, and Bridge Approaches,”
26. Folliard, K. J.,Trejo, D., Du, L., and Sabol, S. A., “Controlled Low-Strength
Material for Backfill, Utility Bedding, Void Fill, and Bridge Approaches,”
28. Fox, T. A., “Use of Coarse Aggregate in Controlled Low Strength Materials,”
246
29. Bhat, S. T. and Lovell, C. W., “Design of Flowable Fill: Waste Foundry Sand
30. Bhat, S. T. and Lovell, C. W., “Mix Design for Flowable Fill,” Transportation
31. Bhat, S. T. and Lovell, C. W., “Flowable Fill Using Waste Sand: a Substitute
for Compacted or Stabilized Soil,” Testing Soil Mixed with Waste or Recycled
32. Stern, K., “The Use of Spent Foundry Sand in Flowable Fill in Ohio,”
35. Environmental Protection Agency (EPA), “Back Document for Proposed CPG
36. Crouch, L. K., Gamble, R. Brogdon, J. F. and Tucker, C. J., “Use of High-
37. Crouch, L. K. and Gamble, R., “Putting More Fines in Flowable Fill,” Rock
247
38. Karim, A. K., Salgado, R. and Lovell, C. W., “Compaction of Fly and Bottom
39. Naik, T. R., Kraus, R. N., Sturzl, R. F. and Ramme, B. W., “Design and
Testing Controlled Low Strength Materials (CLSM) Using Coal Ash,” Testing
Soil Mixed with Waste or Recycled Materials, ASTM STP 1275, West
1996.
42. Federal Highway Administration (FHWA), “Fly Ash Facts for Highway
43. McGrath, T. J. and Hoopes, R. J., “Bedding Factors and E Values for Buried
44. Nmai, C. K., McNeal, F. and Martin, D., “New Foaming Agent for CLSM
45. Janardhanam, R., Burns, F. and Peindl, R. D., “Mix Design for Flowable Fly-
3,1992.
248
46. Ayers, M. E., Wong, S. Z. and Zaman, M., “Optimization of Flowable Fill
47. Fesenmaier, T., “CLSM for New England Thruway,” Concrete International,
51. Dandria, G. G., Frost, D. J., Ashmawy, A. and Patterson, K. R., “Potential
Factors Affecting Flow consistency Test Method for Controlled Low Strength
52. Gray, D. D., “Filling Abandoned Mines with Fluidized Bed Combustion Ash
249
54. Janardhanam, R., Burns, F. and Peindl, R. D., “Mix Design for Flowable Fly-
2, 1992.
1991.
56. McLaren, R. J. and Balsamo, N. J, Fly Ash Design Manual for Road and Site
Institute, 1986.
58. Ayers, M. E., Wong, S. Z. and Zaman, W., “Optimization of Flowable Fill
59. Dolen, T. P., “Properties of Low-Strength Concrete for Meeks Cabin Dam
60. Abelleria, A., Berke, N. S. and Pickering, D. G., “Corrosion Activity of Steel
Fill), ASTM STP 1331, West Conshohocken, PA, American Society for
250
61. Ramme, B. W., Naik, T. R. and Kolbeck, H. J., “Use of Fly Ash Slurry for
62. Landwermeyer, J. S. and Rice, E. K., “Comparing Quick Set and Regular
Properties for Regular and Quick-Set Flowable Fill,” The Design and
STP 1331, West Conshohocken, PA, American Society for Testing Materials,
1998.
65. Gress, D., “The Effect of Freeze-Thaw and Frost Heaving on Flowable Fill,”
67. Buss, W. E., “Iowa Flowable Mortar Saves Bridges and Culverts,”
251
68. Fougere, K. A., “Evaluation of Flowable Fill for Use in Highway and Bridge
Maine, 1994.
69. Hook, W. and Clem D. A., “Innovative Uses of Controlled Low Strength
70. Mason, T. F., “Use of Controlled Density Fill to fill Underslab Void,” The
Fill), ASTM STP 1331, West Conshohocken, PA, American Society for
71. Naik, T. R., Barun R., Ramme, B. W. and Lolbeck, H. J., “Filling Abandoned
72. Walker, M. P., and Ash, J. P., “Flowable Fill Backfill for Use in Sequential
252
Low-Strength Materials (Flowable Fill), ASTM STP 1331, West
75. Green, B. H., Staheli, K., Bennett, D. and Walley, D. M., “Fly-Ash-Based
76. Kawasaki, H., Horiuchi, S., Akatsuka, M. and Sano, S., “Fly-Ash Slurry
Embankment to Minimize the Bump at the End of the Bridge,” The Design
ASTM STP 1331, West Conshohocken, PA, American Society for Testing
Materials, 1998.
78. Swaffar, K. M. and Price, H. R., “Tunnel Saved by Fly Ash,” Civil
79. Spangler, M. G. and Handy, R. L., Soil Engineering, Fourth Edition, Harper &
80. Hegarty, J. R. and Eaton, S. J., “Flowable Fill Promotes Trench Safety and
253
Fill), ASTM STP 1331, West Conshohocken, PA, American Society for
82. Lea’s Chemistry of Cement and Concrete, Hewlett, P. C., Editor, Fourth
254
Vita
August 25, 1971, the son of Wuyin Du and Yueying Yuan. After completing his
the Zhao Zhiying Award. In 1993, he obtained the degree of Bachelor of Science.
his master’s degree in science with the thesis titled, “Experimental Study on
received the prestigious Bloc Fellowship and entered The University of Delaware
614200
255