By
Kiplagat Chelelgo
Requirement for the Award of the Degree of Master of Science in Civil Engineering
August, 2009
AUTHORS DECLARATION
The research detailed in this report was conducted in the summer of 2009 at the
Engineering Centre. I hereby declare that it is my own work and has not been
presented in any form to any other institution other than the University of
organisations.
i
ABSTRACT
depending on the rate at which curing proceeds. The challenge presented by its
regimes should reliably simulate on-site trends without excessively aging the
This project set out to investigate the likely effects of pavement temperature
the incorporated bituminous binder. To achieve this, a predictive model was used
to estimate high and low pavement temperature peaks in tropical and temperate
conditions. This resulted in two cyclic curing protocols that were used alongside
artificially aging Hot Mixed Asphalt and mixing it with a Cationic Emulsion. Asphalt
cores were moulded from the mix and subjected to the three curing regimes. A
series of tests were conducted on the cores to track curing progression and the
under-estimation of long term strength and severe aging of the binder as attested
ii
results, a study incorporating more variables that interactively influence the
iii
TABLE OF CONTENTS
SECTION Page
Authors Declaration..i
Abstract .ii
Abbreviations ..ix
Symbols xi
Acknowledgements xii
CHAPTER1. INTRODUCTION.1
iv
2.3.1 Stiffness .14
2.4.1 Penetration.15
3.1 Equipments.17
3.2 Materials20
3.2.1 Aggregates20
v
3.6.3 Softening Point of bitumen37
5.1 Conclusions46
5.2 Recommendations.47
LIST OF REFERENCES.48
BIBLIOGRAPHY.53
APPENDICES 54
A Schedule of tests55
B Results.56
C Journal paper.57
vi
LIST OF TABLES
vii
LIST OF FIGURES
viii
ABBREVIATIONS
HR Hot Recycling
CR Cold Recycling
ix
ITST Indirect Tensile Stiffness Test
x
SYMBOLS
Bulk density ( )
Dry density ( )
xi
ACKNOWLEDGEMENTS
Nottingham for the financial support offered through the Developing Solutions
for Africa scholarship scheme. It is through the fund that I enrolled in the Msc.
Many thanks also go to my supervisor Assoc. Prof. Tony Parry for his guidance
and invaluable technical advice throughout the research period. The able
Mr. Oke Oluwaseyi deserves recognition for the role he played at the problem
definition and concept development stages. The many hours we spent together in
the laboratories are not taken for granted. Together with him are the members of
I would also like to thank the members of NTEC laboratory led by Richard
Blakemore, Lawrence Pont, Nancy Hodge, Richard Meehan and Dave Edwards for
all the knowledge that they freely shared. Their trainings on the use of
I am indebted to my friends and family members for the emotional and financial
support provided. My honourable friend Mr. Joshua Boiwo deserves a lot of credit
Last but not least is my dearest friend, Angela. Thanks for urging me on.
xii
CHAPTER1. INTRODUCTION
Flexible and rigid pavements are designed to function effectively under traffic and
environmental loads for the duration specified as the design life and are
this period. Conventionally, the pavement would be overlaid with a Hot Mixed
Asphalt or ripped and replaced with a new layer of asphalt altogether, in a bid to
lowering these costs are highly desired. The current trend is to increase re-use of
materials from existing pavements and to minimize use of materials from raw
recycling performs much better from the economic point of view and savings of
up to 30% have been reported in some projects [1] . Materials from the two major
asphalt pavements dominate due to their more popular use in paved areas.
rehabilitation methods started on a small scale in the early 1930s and gained
much of its popularity in the 1970s due the energy crisis that hit the globe in
1
United States and South Africa. Laboratory models and field trial sections were
binder and thus allow for uniform coating of both fine and coarse aggregates.
Hot-Recycling (HR) was the first form of asphalt pavement recycling and was
based on precisely the same technology as HMA, only that reclaimed aggregates
coated with aged bitumen were used in place of virgin aggregates. In that case,
can be done with either softer grades of bitumen, bitumen emulsions, foamed
bitumen or bitumen cut-backs. The cut-backs have since been relegated to small
scale patching works due to their poor Health Safety and Environment (HSE)
record [3].
Cold emulsions were the first on the scene with the first anionic bitumen
emulsions being patented in 1922 by Hugh Mackay, an English chemist, under the
commercial name Cold Spray that was later changed to Cold Asphalt (COLAS)
in the 1950s by Professor Ladis Csanyi of Iowa State College [4]. The two
emulsifying agents while foaming of bitumen involves injecting cold water into hot
bitumen to form vapour that is entrapped in bubbles of bitumen. They share the
2
need for curing or setting for the mix to attain the ultimate materials
properties.
factors that interactively determine the curing rate. On-site curing may take
material properties for use in pavement structural design. The laboratory curing
regime so selected should closely correlate with the expected in-situ conditions
This project set out to investigate the likely effects of alternate heating and
cooling on predicted asphalt and binder properties of a wearing course cold mix.
The current protocols are based on steady state temperature curing regimes
which may not in essence capture the temperature dynamics taking place in the
This study was motivated by the need to reliably promote use of cold emulsion
asphalts and recyclable pavement materials as the cost effective and environment
friendly road construction options for both the developing and the developed
3
yet a consensus on the laboratory curing modes for use, though one element runs
through all the protocols that have so far been used by researchers. They are all
steady state temperature curing regimes. The study experimented with cyclic
the tropics and the temperate regions. For purposes of comparison, one steady
state protocol in popular use was tested alongside those under study. To achieve
the goals of the study, the laboratory and research work focused on the following
points:
4
CHAPTER2. LITERATURE REVIEW
constituent materials without input of heat. It can be carried out on site as Cold
In-Place Recycling (CIPR) or off-site as Cold Central Plant Recycling (CCPR) [5].
On site recycling eliminates the need for haulage, despite introducing the element
of variability in the mix. Cold In-Place Recycling requires use of specialized set of
materials.
Weight Deflection (FWD) tests for foundation analysis [6]. Once it has been
ascertained that the foundation is sufficient and drainage has been given due
itself.
Once the analysis process has been completed and the pavement section
from the section by either coring or by milling. The material thus collected is
residual binder content, penetration of the recovered binder and the total
moisture content in the material. Material properties will inform the choice and
5
amount of the binder, pre-wetting water and the compaction effort required to
pavement recycling exercise. This is attributed to the fact that reduction in its
bitumen emulsion could be anywhere between 25% and 60% [3]. To achieve the
ultimate properties of the base bitumen, the emulsion has to break to release
the water and allow bitumen globules to coalesce. The rate of emulsion breaking
Emulsion breaking and mix curing occur concurrently and act to get rid of the
effective moisture in the mix so as to allow for direct contact between the
aggregates and the binder. Curing happening within the first few days in the life
of the mix is more dependent on stabilizing agent than environmental factors and
is responsible for the initial cohesion of asphalt that gives it strength to support
light construction traffic. The final stages that carry on to the attainment of
Equilibrium Moisture Content (EMC) are dependent on both the stabilizing agent
conditions is responsible for unique regional curing rates [7]. Humidity, wind
speed, drainage conditions and temperature dominate the curing process and the
6
2.2 Temperature Dynamics in Asphalt Pavements
This project makes use of pavement temperature simulation models that have
hitherto found use in selection of appropriate performance grade bitumen fit for
such as solar radiation, wind velocity and ambient air temperature to simulate
and characterize pavement temperatures over time and depth. From analysis of
Minnesota Road Weather Information System (RWIS) sites over a six-year period
between 1999 and 2005, the trio concluded that the maximum daytime pavement
This report concerns itself with Cold-Emulsion Reclaimed Asphalt Pavement whose
on-site temperature trends are expected to mirror those of Hot Mixed Asphalt as
model that incorporated the latitude of the location concerned and the specific
time of the year. Latitude and the location are two parameters that were intended
at catering for the variation in solar radiation which is associated with the Earths
revolution round the sun. Solaimanian & Bolzan [11] and Yavuzturk et al [8]
7
confirmed that incident solar radiation and ambient temperatures are the two
knowledge, they developed the relationships below to define the two temperature
(2.2, a)
(2.2, b)
= Latitude (Degrees)
and El-Korchi [12] at the time of this project was employed in computing
this project. The model allows for computation of temperature at the surface and
8
(2.2, c)
(2.2, d)
laying and compaction. Heating reduces the viscosity of the aged binder and thus
enables it to homogeneously blend with the fresh added binder. The resulting
effective binder uniformly coats the aggregates to produce a stiff asphalt mix
capable of supporting traffic loads shortly after construction. Such a mix can be
Strength of a cold mix on the other hand evolves over time and is dependent on
the rate of emulsion breaking, mix curing and fluxing. The challenge presented by
the curing phenomenon and scepticism associated with lack of reliable data and a
definitive design procedure have confined the use of cold asphalt mixtures to
9
to estimate service life of cold mixes. All protocols encountered in the course of
this study use steady state temperatures and a few are cited below.
field life of the mix i.e. up to one year in the field [13]
and three years in the field for the temperate regions under low to
3. 14 days at 18C and Relative Humidity 50% to simulate short term curing
on site conditions under which the asphalt mix will serve. The idea of diurnal and
sought to investigate temperature cycles in one tropical city and one city in the
temperate region. The curing protocol proposed by the British Standards Institute
Nairobi in Kenya was chosen to represent the tropical conditions since its climatic
region covers bulk of the country and represents the most populated and
productive area of the country. Kenya has two rainy seasons, one between March
and May and the other between November and December. The remaining part of
the year is dry although there are no significant variations in the average monthly
temperatures. The driest months lie between June and October and the
10
conditions will be quite favourable for laying of a cold asphalt mix. The average
number of sunshine hours as computed from the table below for the duration
Average
Discomfort Relative
Sunlight Temperature Average
from heat Humidity Wet Days
Month (hours) Precipitation
& (+0.25 mm)
Average Record (mm)
humidity am pm
Min Max Min Max
Jan 9 12 25 8 29 Moderate 74 44 38 5
Feb 9 13 26 9 31 Moderate 74 40 64 6
June 6 12 21 7 27 - 89 60 46 9
July 4 11 21 6 26 - 86 58 15 6
Aug 4 11 21 7 27 - 86 56 23 7
Sept 6 11 24 5 28 - 82 45 31 6
Oct 7 13 24 7 30 - 82 43 53 8
Nov 7 13 23 6 28 - 86 53 109 15
Dec 8 13 23 8 28 - 81 53 86 11
11
London in the UK was chosen to represent the temperate conditions with the
average number of sunshine hours in a typical summer day as computed for the
Discomfort Relative
Average Temperature Average
from heat Humidity Wet Days
Month Sunlight Precipitation
Average Record & (+0.25 mm)
(hours) am pm (mm)
Min Max Min Max humidity
Jan 1 2 6 -10 14 - 86 77 54 15
Feb 2 2 7 -9 16 - 85 72 40 13
March 4 3 10 -8 21 - 81 64 37 11
April 5 6 13 -2 26 - 71 56 37 12
May 6 8 17 -1 30 - 70 57 46 12
June 7 12 20 5 33 - 70 58 45 11
July 6 14 22 7 34 - 71 59 57 12
Aug 6 13 21 6 38 - 76 62 59 11
Sept 5 11 19 3 30 - 80 65 49 13
Oct 3 8 14 -4 26 - 85 70 57 13
Nov 2 5 10 -5 19 - 85 78 64 15
Dec 1 4 7 -7 15 - 87 81 48 15
12
Two temperature peaks for use in the laboratory to simulate cyclic curing were
computed using equations 2.2(c), 2.2 (d) and the average maximum and
minimum temperatures for the two cities under study, as obtained from BBC
weather records [18,19]. The upper peaks were assumed to occur during the day
when the maximum average air temperatures are recorded while lower peaks
were assumed to occur during the night when the average minimum
which is 50mm thick and can reliably be assumed to represent the average
pavement conditions. The duration of sunshine hours were used in the laboratory
to cure samples at the upper peaks, i.e. 44 C and 37C in the table below. The
lower peaks were 34C and 29C for the tropical and temperate conditions
respectively
Nairobi 1.27 11.6 22.2 36.3 97.3 46.9 116.4 92.9 33.8 111.1 43.9
London 51.50 11.8 20.0 31.6 88.8 39.8 103.6 84.8 29.3 98.9 37.2
13
2.3 Fundamental Properties of Asphalt Mixtures
an engineer to specify the thickness of a pavement layer and predict its service
life with some degree of certainty. Historically, this has been done empirically
using recipes specified by rule of thumb [21]. Recipes are calibrated to suit local
conditions and materials and may therefore not be quite useful in developing
guidelines fit for universal comparison of design methods. There has been
and strains induced by traffic and environmental conditions . The two pavement
cracking and rutting due to permanent deformation. Stiffness and fatigue can be
performance.
2.3.1 Stiffness
Stiffness refers to the relationship between the stress applied to a material and
a visco-elastic material whose elastic modulus varies with both temperature and
asphalt mix, the term stiffness modulus has been adopted [22].
either uniaxial test, indirect tensile test or the beam tests [22]. Laboratory
14
was done in accordance with BS DD 213: 1993 [28], which specifies the method
of performing the Indirect Tensile Stiffness test in the Nottingham Asphalt Tester.
strength of the material. Traffic and thermal loads in an asphalt pavement induce
alternate stretching and relaxation in the binder matrix which eventually leads to
Determination of fatigue life of an asphalt mix in the laboratory can be done using
simple flexure, uniaxial test or the indirect tensile test. The latter is preferred due
specification- BS DD ABF: 2003 and used in this exercise. The aim of the test is
To compare the effects of the curing regimes on the aging properties of the
binders. Bitumen was recovered before and after curing of the cylinders and the
2.4.1 Penetration
penetration test is mostly used for classifying bitumen into grades but it can give
good indication of the aging phenomenon since its value is largely dependent on
Softening Point is a laboratory test used to determine the transit point between
is by Ring and Ball test in accordance with BS 2000, part 58 [31]. Aged bitumen
is expected to have higher softening point than fresh bitumen since it will have
Saturates and Asphaltenes. Their relative presence depend on the source of crude
oil from which the bitumen was extracted and the age of the particular bitumen.
Asphaltenes are polar components with the highest molecular weight and function
Asphaltenes are insoluble in n-heptane and can therefore be sieved out from the
bitumen sample.
16
CHAPTER3. LABORATORY WORK
3.1 Equipment
The Nottingham Asphalt Tester (NAT) [22] is the main piece of equipment that
was used to carry out the core tests in this project. Other equipments that came
Sieve Shaker, Hobart Mixer, Shear Gyratory Compactor (SGC) and Conditioning
Cabinets.
proceeded, use was made of a suite of tests available in the Nottingham Asphalt
unit [22].
diameter via the loading platens. The sinusoidal load pulses applied produce
mounted diametrically opposite each other, in the case of ITSM testing, or above
the loading platens, in the case of ITFT tests. All operations are controlled by
specialist computer software that keeps a data log of average transient loads and
the peak lateral or longitudinal deformations. The three tests i.e. ITFT, ITSM and
RLAT can all be done in the same NAT machine by simply changing the mounting
while ITFT is destructive. To prevent damage to the LVDTs during ITFT testing,
17
they are mounted above the upper loading platen where they measure vertical
deformation at failure.
Pneumatic
Load Actuator
Load Cell
18
LVDTs
standard 600 kPa while being given 30 gyrations per minute. The available
software allows for setting the number of gyrations or the desired density as the
target in the gyration process [21]. The software automatically plots density of
the sample against the number of gyrations as compaction proceeds. Once the
19
Loading Ram
Upper Platen
Steel Mould
Rotating Base
3.2 Materials
This research utilized 60 asphalt cores of dimensions 100mm diameter and 50mm
mixed asphalt of granite origin, dust and a mineral filler. The focus was on cold
emulsion reclaimed asphalt mix fit for use in the surface course and thus,
3.2.1 Aggregates
The aggregates used in this research were derived from artificially aged Dense
Bitumen Macadam (DBM) obtained from Cliffe Hill Quarry in Leicester. The
residual binder in the DBM after being kept at ambient temperatures for close to
ranged between 20dmm and 21dmm. This material came in lumps and had to be
20
reheated to 160C before being laid into slabs of 305mm by 305mm by 50mm
thick and allowed to cool in the laboratory for a day or two. The penetration of
binder recovered from the material after this treatment had dropped to 14dmm
and the asphalt was considered to have been sufficiently aged to qualify as
To generate Reclaimed Asphalt Pavement aggregates, the slabs were broken into
small blocks with the aid of a pneumatic drill before being passed through a jaw
Sieve analysis was performed on a number of RAP, Filler and Dust samples and
results averaged to get their overall gradations. It was noticed that RAP gradation
fell below the ORN 19 & 31[25,26] lower bound, an indication that it was deficient
in fines.
ORN
ORN
19&31
Maximum 19&31
Size RAP 5mm Dust Mineral Filler Lower
Density Upper
(mm) (% Passing) (% Passing) (% Passing) Bound
(% Passing) Bound
for
for DBM
DBM
21
Figure 3.2.1: Gradation of Individual Aggregates
polythene sample bags in readiness for blending. There were five bags in total
containing aggregates of sizes 20mm, 14mm, 10mm, 6.35mm, 3.35mm and dust
respectively.
A cationic bitumen emulsion containing 60% bitumen and 40% water was used as
the binder in preparation of the cold mix. The bitumen emulsion of Venezuela
origin was supplied by Nynas Asphalts, UK. At the time of supply, the
manufacturer reported the penetration of the emulsion as being 48dmm and its
these values
Asphalt mix design is the process of determining the properties of the constituent
asphalt binder, amount of pre-wetting water and the compaction effort required
to achieved the required field density [23]. Since the mix being prepared was
Reclaimed Asphalt Pavement, Dust and Filler were graded separately and blended
falling within the envelope defined by the lower and upper bounds of the
Overseas Road Notes No. 19 and 31, as shown in the table below [25,26].
Several proportions were tried with the aim of approaching the maximum density
(3.3.1, a)
The maximum aggregate size is one sieve size larger than the first sieve to
retain more than 10% aggregates, which is 28mm according to this gradation.
Size ranges from 75 microns to 28mm. The ratios shown in the table are those
23
of RAP, Dust and Filler respectively, which were blended using the simple
(3.3.1, b)
sieve
individual aggregates
combination
The approximate total binder content requirement for the mix was estimated at
6.46%, using the empirical formula below developed by the Asphalt Institute
[21]. This was based on the combined aggregates gradation presented in table
3.3.1 above.
(3.3.2)
aggregates
25
= Percentage of mineral aggregates passing sieve size 2.36mm
The binder content of the Reclaimed Asphalt Pavement which constitutes 65% of
the mix had previously been determined as 4.25%. Taking this into consideration,
the total binder requirement of the mix was calculated as 3.7%. Since a cationic
emulsion containing 60% base bitumen was to be used, 6.16% was needed to
Maximum Dry Density (MDD) of the blended aggregates was performed using
ensuring that some were above and others below the estimated optimum
100mm diameter steel moulds and compacted in the gyratory compactor. The
LVDTs installed in the gyratory compactor accurately measured the heights of the
with the mass and diameter of the specimens to compute the bulk density. The
computed bulk density was input into the equation below to arrive at the dry
26
Where, = Bulk density ( )
= Dry density ( )
= Moisture content
Moisture
Dry Density (kg/m)
Content (%)
The data series of dry densities were plotted against the moisture contents and a
smooth curve fitted through the points. The peak of the curve represents the
MDD (2,177 ) and the corresponding moisture content is the OMC (6.55%).
27
Based on the emulsion content of 6% determined in section 3.3.2, an asphalt mix
table below. Cylinders prepared from the mix were cured for three days at 40 in
controlled at 20 .
The cylinders prepared by varying amounts of pre-wetting water were tested for
strength in the Indirect Tensile Strength test (ITS) and the results tabulated in
table 3.3.3 (b) as dry strengths (ITSD). To simulate moisture damage in the
worse case scenario, the cylinders were kept overnight in a water bath at
25 and re-tested for strength. The strength values were tabulated in the table
below as wet strengths (ITSW). The Indirect Tensile Strength ratio (ITSR),
computed as a ratio of the wet strength to the dry strength, was plotted
alongside the dry and wet strength values to yield the graphs in Figure 3.3.3 (b)
below.
Average
28
Figure 3.3.3 (a): Determination of Pre-wetting water (%)
It is evident from the plots in figure 3.3.3 (a) that the dry strength drops as the
total fluid content increases. The wet strength, on the other hand, rises slightly to
dry aggregates.
The pre-wetting water content of and the emulsion content of were used
to fabricate three cylinders for determination of the target bulk density. The
gyration of ; ram pressure of 600 and 200 gyrations. The average of the
three cylinders gave a reference bulk density of 2,458 . 95% of this value,
29
3.4 Preparation of gyratory specimens
Cylindrical specimens for laboratory curing and testing were fabricated based on
and the optimum pre-wetting water content. The target dimensions of the
laboratory cylinders suitable for both ITSM and ITFT tests were 100mm diameter
and 45mm height. From the estimate of the target density, this translates to
into metallic tins that were conditioned overnight at 35 to drive out any
moisture in the aggregates. This material was sufficient to produce four asphalt
cylinders.
The dry aggregates were then transferred into the mixers bowl where 1.5%
(56g) of water by mass of dry aggregates was added. This was followed by 1
aggregates was added into the moist aggregates and mixing continued for
a further 1 minute. The resulting mix was slightly segregated but there was
30
Stainless
Steel Bowl
Flat beater
Figure 3.4 (a): 20 Quarts Hobart Mixer Figure 3.4 (b): Flat type Agitator
The resultant mix was weighed into a steel mould assembly that was placed in
the gyratory compactor and compacted to the target density. The number of
gyrations to achieve the target density varied from cylinder to cylinder and the
reason for this variation was not immediately understood. Attempts were made to
keep the mould in the oven for a few minutes before testing in a bid to control
the mix temperature but that did not change the trend. It was suspected that the
process of weighing the mixtures into the mould might have introduced some
was not easy to control segregation in every cylinder since batching was done
hours to allow them gain sufficient shear strength before being transferred into
necessary for some cylinders to be extruded after 18 hours and kept in sealed
from the weight loss in the cylinders revealed similarity in the rates of curing of
31
the cylinders extruded after 24 hours and those extruded after 18 hours and kept
in polythene bags. The latter method was therefore adopted for all the
Two thermostatically controlled conditioning cabinets had been set aside for the
table 2.2.2 (c) were rounded off to whole figures for ease of handling. For the
tropics, 34 and 44 were used as the two peaks, while 29 and 37 were
used for the temperate conditions. Temperature in the cabinet representing the
temperature of 35 for the 14 day curing duration selected for all the cabinets
The tests described hereunder were specifically chosen for their ability to
of the mix and that of the binder. The rate of moisture loss was tracked to give
content at the time of moulding. Indirect Tensile Stiffness tests were conducted
Despite the ITSM test being non-destructive, it was decided to use cylinders not
previously tested for stiffness at each interval to eliminate the possibility of any
damages. The pulse shapes for specimens tested after one day of curing indicated
some failures in the specimens and thus justifying the decision. Samples cured
32
for 10 and 12 days were taken back to the conditioning cabinets after stiffness
determination to provide sufficient number of samples for use in the ITFT test
which was scheduled to be done on cores cured for 14 days. The standard code
followed for the test required a minimum of 10 cylinders per curing regime which
To study the effects of the curing regimes on the bitumen properties, bitumen
was extracted from cylinders after 24 hours of cure at ambient temperature and
The Indirect Tensile Stiffness modulus test was performed in the laboratory in
of the code. The diameter of each specimen was averaged from values taken at
three equally spaced points using external callipers in the manner shown in figure
3.6.1 (a). A steel rule was used to compute the average height by taking two
readings from each of the perpendicular directions as shown in figure 3.6.1 (b).
Steel Rule
External
Callipers
Straightedge
Steel plate
Figure 3.6.1 (a): Specimen diameter Figure 3.6.1 (b): Specimen height
33
The Nottingham Asphalt Tester configured as shown in figure 3.3.1 (a) was used
in this test. The cylindrical specimen of average diameter 100 and 50 thick
was correctly positioned between the upper and the lower platens of the NAT. The
LVDTs were secured centrally about the axis of the specimen and opposite each
other with the help of the alignment jig. Much care was taken to ensure that the
LVDT probes did not get into voids in the specimen. The target parameters were
input into the controlling software and the specimen given five conditioning
was recommended for the 100 diameter specimens. After the five
conditioning pulses, the LVDTs were checked and readjusted as indicated on the
computer screen by the software. The load actuator of the Nottingham Asphalt
A further five load pulses were applied and the stiffness from the five pulses
averaged to give the stiffness of the cylinder. It was recommended that the test
be repeated each time by rotating the specimen through 90 10. The mean value
34
of the stiffness in the two directions was taken as the stiffness of the cylinder.
(3.6.1)
Where,
t = Specimen thickness
= Poissons ratio
= Stiffness modulus
The above equation has been programmed into the software controlling the test
and a print out containing all the required values is generated at the end of the
test. A Poissons ratio of 0.35 is recommended for asphalt when tests are
undertaken at 20
The Indirect Tensile Fatigue Test was performed in the laboratory in accordance
with BS DD ABF: 2003 [29]. According to the draft code, the specimen to be
12 hours. For a mix with aggregate size 20mm or below, 10 specimens are
recommended for testing at each stress level to give a good spread of data. The
target stress level for the first specimen should either be 600 or 500 ,
subsequent stresses will depend on the number of cycles to failure of the first
sample.
35
A controlled strain failure criterion was chosen for the specimens in this research
because the criterion represents failure mechanism in thin layers such as the
controlled stress version of the Indirect Tensile Stiffness Test (ITST) using the
stress level to be adopted in ITFT testing. The configuration of the NAT is then
switched to accommodate the ITFT frame and the specimen centrally positioned
between the upper and the lower loading strips. The flat faces of the specimen
should be perpendicular to the bottom loading strip. Once the LVDTs have been
correctly positioned above the upper loading strip, the specimen is loaded by
horizontal strain against the number of cycles to failure on logarithmic scales. The
tensile strain (microstrains) at the centre of the specimen are calculated as:
; (3.6.2)
Where
36
3.6.3 Bitumen Penetration Test
in BS 2000-49 [30]. Bitumen extracted from cylinders before and after curing
25 for between 1hr and 1.5 hrs prior to the test. A needle with a 100g weight is
then brought to touch the surface of bitumen at right angles and allowed to fall
through it for 5 seconds. The procedure is repeated three times and the depths in
deci-millimetres are averaged to get the penetration value for the bitumen.
Softening point of bitumen is determined in the laboratory using the Ring and Ball
apparatus in accordance with BS 2000-58 [31]. In this test, two steel balls are
bath. The temperature of the water bath is raised in steps of 5 till the bitumen
samples deform to let the balls drop to the bottom of a steel plate placed 25mm
below the rings. The temperature at which this occurs is recorded as the
(EVT) of the bitumen, i.e. the temperature at which its penetration is 800
37
CHAPTER4. DATA PRESENTATION AND ANALYSIS
The tests described in the previous chapter were specifically selected to reliably
order to effectively collect all the relevant data and to conduct all the tests in the
correct order and timings. It was not always possible to do the tests at the
intended time due to tight schedules in the use of the laboratory equipments and
flexibility had to be built in to allow for postponement of some tests. ITSM tests
were mostly affected since some of the cylinders had to be kept in the cold room
for up three days awaiting availability of testing equipment. The help of the
A measure of the moisture loss by samples under the three temperature regimes
was made to provide information regarding the curing progress. It was decided to
use the residual moisture content as a percentage of the total mix to ease
comparison of the three data sets. Pre-wetting water in the mix at the time of
moulding constituted 1.5% by mass of dry aggregates. The emulsion on the other
hand was made up of 40% water. This together with the pre-wetting water
brought the amount of water in the mix to 3.64% of the total mass. Curiously,
the three curing regimes displayed closely similar trends in water loss. A plot of
the percentage water in the mix against period in time indicated an approximate
loss of 10% after curing in the mould for 24 hours, 80% after curing in the
conditioning cabinets for 1 day and 90% lost after curing in the conditioning
38
cabinets for 2 days. This seems to suggest that temperature plays a minor role in
34C & 44C Cure 29C & 37C Cure 35C Cure
15 0.07 13 -0.11
15 0.13
39
Figure 4.2: Cold-Emulsion RAP Moisture Loss
In Table 4.2 and Figure 4.2 above, day 1 represents the 24 hour period within
which the cylinders were cured in the moulds before being transferred to the
conditioning cabinets. As can be seen from the graphs, moisture loss stabilized at
around the sixth day though there was a peculiar rise in moisture content at
around that period for the three regimes. This is the time during which the
samples had been kept in the cold room for three days prior to ITSM testing.
Some moisture was noticed to have collected within the storage bags and it is
It can also be noticed that at some instances, more than 100% of moisture had
been lost by the samples. That obviously cannot happen in the normal
circumstances. There are two possible explanations to this occurrence and the
theres a possibility that the aggregates used contained some moisture that was
not accounted for in moisture calculations. Most specimens also lost some
40
fragments during handling and curing due to their fragility giving the false notion
that they had lost more water than they actually contained. The other possibility,
which is most likely, is the loss of volatile components of the bitumen. The latter
could only be confirmed after the other laboratory tests had been done.
Stiffness modulus of cylinders from the three curing protocols was determined at
six time intervals. The cyclic temperatures started on the lower peaks and the
effect was reflected in the stiffness modulus determined after a day of curing in
the ovens. That would have been equivalent to on-site laying of cold-emulsion
asphalt in the evening when temperatures are low. Strength development in the
peak.
41
Figure 4.3: Strength Evolution in Cold-Emulsion RAP
The drop in stiffness of cores witnessed on the 10 th and 14th days was a break
from the general trend. Since stiffness determined at each interval was done on
fresh set of cylinders, the batching process was suspected to have introduced
variability in the mix. The other phenomenon that was not closely monitored is
cylinders into the ovens. Going by the results of stiffness determined on the first
day, this could as well be the reason we have lower values on the tenth day.
Wide variations in material properties are common in cold mixes but a general
trend can be drawn from the results plotted in Figure: 4.3 above. It can be
the long term strength. The tropical conditions show a faster rate of strength
42
to support low to medium traffic [24], which in this can be achieved from as early
Fatigue characteristics of materials cured under the three different regimes were
compared by plotting the maximum horizontal tensile strains against the number
lines. Fatigue life of each mixture was obtained from its linear regression model
and 37 exhibited the highest level of fatigue resistance while those cured at
and thus ending up being brittle. In terms of fatigue, the standard cure of asphalt
emulsion at 35 slightly underestimates the fatigue life of the mix laid in the
temperate environment.
43
Figure 4.4: Comparison of Fatigue Life
Bitumen was recovered from the cores before and after being subjected to the
three curing conditions. Table 4.5 below presents the characterization of the
44
bitumen caused by the curing regimes. The results indicate that curing at the
the other two regimes. The increase in asphaltene contents after curing is partly
due to aging and partly due to activation of aged bitumen in the RAP.
Asphaltene
Penetration Softening
Contents
( ) Point ( )
(%)
35oC 21 64 16.3
After Curing
29oC &
32 59.2 16.2
37oC
34oC &
22 63 16.1
44oC
45
CHAPTER5. CONCLUSION AND RECOMMENDATIONS
5.1 Conclusion
The study of the fundamental properties of the materials cured under three curing
protocols, two cyclic and one steady state, revealed a potentially useful
results from the study form a basis for a detailed study into the precise behaviour
of cold asphalts under different site conditions. Based on the results of the
46
5.2 Recommendations
the results. The study makes the following recommendations for future research:
samples
asphalt pavements.
47
REFERENCES
[1] Santucci, L.E. and Hayashida, M.T., Testing, Design and Evaluation of Cold
Pavements for Southern Africa, Cape Town, South Africa. March, 1984.
[2] Van Wijk, A.J., Wood, L.E. and Kercher, K.J., Construction of Cold Recycled
Asphalt Pavements for Southern Africa. Cape Town, South Africa. March, 1984.
10,2006.
Stellenbosch,2008.
[7] Jenkins, K.J., Mix Design Considerations for Cold and Half- Warm Bituminous
SA,2000.
48
[9] Herb, W., Marasteanu, M., and Stefan, H.G., Simulation and Characterization
[10] Diefenderfer, B.K, Al-Qadi, L.I, Reubush, D.S, Freeman, T.E., Development
[11] Solaimanian, M. and Bolzan, P., Analysis of the Integrated Model of Climatic
[12] Mallick, B and El-Korchi, T., Pavement Engineering: Principles and Practice.
[13] Ruckel, P.J, Acott, S.M and Bowering, R.H Foamed-Asphalt Paving Mixtures:
[14] BS 434-2:2006 Bitumen Road Emulsions - Part2: Code of Practice for the
Use of Cationic Bituminous Emulsions on Roads and Other Paved Areas. British
[15] Serfass, J.P., Poirier, J.E., Henrat, J.P and Carbonneau, X., Influence of
Curing on Cold Mix Mechanical Performance. Materials and Structures, Vol. 37,
[16] Brennan, M.J., Sheahan, J., Gilbin, P.M. and Kavanagh, A., Laboratory
49
[17] Kekwick, S.V Best Practice: Bitumen-Emulsion and Foamed Bitumen
Pretoria,2005.
[18] BBC Weather Centre (2009), Average Conditions for Nairobi, Kenya
Available at:
http://www.bbc.co.uk/weather/world/city_guides/results.shtml?tt=TT0003
00
[19] BBC Weather Centre (2009), Average Conditions for London, United
Kingdom
Available at:
http://www.bbc.co.uk/weather/world/city_guides/results.shtml?tt=TT0037
90
[20] Fatani, M., Al-Abdul Wahhab, H.I, Balghunaim, F.A., Bubshait, A. and Al-
Pavements in Saudi Arabia". Final Report submitted to King Abulaziz City for
[21] Needham, D., Developments in Bitumen Emulsion Mixtures for Roads. PhD
50
[23] Jitareekul, P. (2009), An Investigation into Cold In-Place Recycling of
[24] Thanaya, N.A., Evaluating and Improving the Performance of Cold Asphalt
Hot Mix Asphalt in Tropical and Sub-Tropical Countries. Overseas Road Note 19.
TRL, UK,2002.
Standards Institute,2003
[30] BS 2000-49: Methods of Test for Petroleum and Its Products- Bitumen and
Institute, 2000.
51
[31] BS 2000-58:, Methods of Test for Petroleum and Its Products-: Bitumen
[32] Assignment: The World (2003), Finding Places with Longitude and Latitude
52
BIBLIOGRAPHY
[1] Maccarone, S., Holleran, G. & Ky, A., Cold Asphalt Systems as an Alternative
[2] Erlan, O.L., Stubstad, R. and Briggs, R., Temperature Predictions and
[3] Engelbrecht, J.C., Roberts, F.L, and Kennedy, T.W Cold Recycled Mixtures,
Changes Occurring During the Oxidative Aging of Two Selected Asphalts from
[5] Khalid, H.A. and Monney, K.O., Moisture Damage Potential of Cold Asphalt.
53
APPENDICES
54
APPENDIX A: SCHEDULE OF TESTS
55
APPENDIX B: RESULTS
56
APPENDIX C: JOURNAL PAPER
57
EFFECTS OF DIURNAL TEMPERATURE DYNAMICS ON
CURING OF COLD-EMULSION RECLAIMED
ASPHALT PAVEMENTS
Kiplagat CHELELGO, School of Civil Engineering,
ABSTRACT
Accelerated laboratory curing regimes come in handy in design and
specification of Cold-Emulsion Reclaimed Asphalt Pavements. The
emphasis is in the choice of a laboratory curing protocol that adequately
simulates expected curing trends in the field. Popular protocols employ
steady state curing temperatures to predict long term behaviour of Cold-
Emulsion Reclaimed Asphalt Pavements. Two cyclic temperatures were
studied alongside one popular steady state temperature in an attempt to
investigate the likely effects of dynamic pavement temperatures on curing
progression and properties of bituminous binders. Analysis of stiffness and
fatigue properties of asphalt cores cured under the three regimes revealed
that steady state curing overestimates the early-life strength of mixes
while underestimating the long term strength. Properties of bitumen
recovered after curing under the three regimes pointed to a severity in
aging of asphalt cores cured under constant temperatures over time.
INTRODUCTION
Pavements recycling fall into the categories of Hot Recycling (HR) and Cold
Recycling (CR). Hot recycling uses the same technology as Hot Mixed Asphalt and
therefore has the disadvantage of huge energy consumption and is associated
with high gaseous emissions. Cold recycling utilizes softer grades of bitumen,
bitumen cut-backs or bitumen emulsions. The emulsions have gained favour
owing to their good environmental record.
58
Curing is the phenomenon that controls the rate of strength gain in Cold-Emulsion
Reclaimed Asphalts. The role of curing is twofold. One is to rid the mix of
moisture so as to allow for direct contact between the aggregates and the binder
and secondly is, to activate the aged binder in Reclaimed Asphalt Pavement by
the process of fluxing. This is whereby new binder penetrates aged binder to
render in active in contributing to the binding effect.
The project set out to study the effects of diurnal pavement temperature cycles
on the rate of curing, strength development and fatigue properties of Cold-
Emulsion Reclaimed Asphalt Pavement. The aim was to assess the suitability of
the existing protocols to mimic on site conditions without exaggerating the aging
process in bitumen.
EXPERIMENTATION
Curing Protocols
(1)
(2)
Climatic data obtained from BBC weather were used to compute the average
minimum and maximum air temperatures [3,4]. The resulting values of the
59
upper and lower temperature peaks for the tropical and temperate regions were
44C & 34C and 37C & 29C respectively. The protocol proposed by asphalt
institute i.e., 14 days at 35C [5] was adopted as the steady state curing
temperature. Three thermostatically controlled conditioning cabinets were used to
simulate the curing protocols in the laboratory.
Specimen Preparation
Hot Mixed Asphalt artificially cured in the laboratory was graded and mixed with a
cationic emulsion to give a mix that was used to produce 18 cylindrical cores for
each curing protocol. The average dimensions of the cylinders used were 100mm
diameter and 50mm height. The cylinders were cured at room temperature for 24
hours before being transferred to the respective conditioning cabinets. Weight
loss was tracked as a means of evaluation curing progression.
Laboratory Tests
Cylinders cured for 10 and 12 days were cured for further four and two days
respectively. This was aimed at providing additional samples for determination of
fatigue life which was targeted at samples cured for 14. 10 samples conditioned
overnight were subjected load pulses ranging from 600 to 100 . Fatigue
characteristics were determined by plotting maximum tensile horizontal strain
versus life to failure on logarithmic scales.
At the end of 14 days of curing, bitumen was extracted from the cores and
characterised to give Penetration, Softening Point temperature and percentage of
Asphaltenes in recovered bitumen.
60
RESULTS AND ANALYSIS
Moisture loss with time was reckoned from the residual moisture as a percentage
of the total mix. Curiously, the three curing regimes displayed closely similar
trends in water loss. As can be seen in figure 2 below, the three protocols
followed more less the same trend in moisture loss. 10% of the total water
content was lost in the 24 hour period of curing in the mould and another 70%
was lost after curing for one day in the conditioning cabinets. 90% of the total
moisture content had been lost on the second day suggesting that temperature
may not be the key player in the evaporation mechanism. Equilibrium Moisture
Content seemed to have been arrived at after 6days of curing
Stiffness modulus of cylinders from the three curing protocols was determined at
six time intervals. The cyclic temperatures started on the lower peaks and the
effect was reflected in the stiffness modulus determined after a day of curing in
the ovens. That would have been equivalent to on-site laying of cold-emulsion
asphalt in the evening when temperatures are low. Strength development in the
cylinders cured at 35 rose gradually in an almost linear manner while those
under cyclic temperatures rose gradually with a decreasing gradient towards a
peak.
61
Figure3: Strength Evolution in Cold-Emulsion RAP
The drop in stiffness of cores witnessed on the 10th and 14th days was a break
from the general trend. Since stiffness determined at each interval was done on
fresh set of cylinders, the batching process was suspected to have introduced
variability in the mix. The other phenomenon that was not closely monitored is
the effect on stiffness of the temperature peak at the time of introduction of
cylinders into the ovens. Going by the results of stiffness determined on the first
day, this could as well be the reason we have lower values on the tenth day.
Wide variations in material properties are common in cold mixes but a general
trend can be drawn from the results plotted in Figure: 4.3 above. It can be
deduced that laboratory curing of cold-emulsion asphalt at 35 closely predicts
the intermediate strength, overestimates the early strength and underestimate
the long term strength. The tropical conditions show a faster rate of strength
development which may generally be underestimated if use is made of the
existing protocol of curing at 35 . A stiffness modulus of 2000 is sufficient
to support low to medium traffic [24], which in this can be achieved from as early
as 12days in the tropical conditions.
Fatigue characteristics of materials cured under the three different regimes were
compared by plotting the maximum horizontal tensile strains against the number
of cycles to failure and generating fatigue relationships by use of power trend
lines. Fatigue life of each mixture was obtained from its linear regression model
by assuming logarithmic linearity of fatigue life. Samples cured cyclically at 29
and 37 exhibited the highest level of fatigue resistance while those cured at
34 an 44 had the lowest life to failure.
Bitumen was recovered from the cores before and after being subjected to the
three curing conditions. Table 4.5 below presents the characterization of the
binders in terms of Penetration, Softening Point and Asphaltene Content
percentage. These three parameters were used to assess severity of aging of
bitumen caused by the curing regimes. The results indicate that curing at the
steady temperature of 35 results in the highest degree of aging as compared to
the other two regimes. The increase in asphaltene contents after curing is partly
63
due to aging and partly due to activation of aged bitumen in Reclaimed Asphalt
Pavement.
Asphaltene
Penetration Softening
Contents
( ) Point ( )
(%)
35oC 21 64 16.3
After Curing
29oC &
32 59.2 16.2
37oC
34oC &
22 63 16.1
44oC
The study of the fundamental properties of the materials cured under three curing
protocols, two cyclic and one steady state, revealed a potentially useful
correlation between laboratory curing temperatures and on-site curing. The
results from the study form a basis for a detailed study into the precise behaviour
of cold asphalts under different site conditions. Based on the results of the
laboratory tests, the study makes the following tentative conclusions.
5.2 Recommendations
64
REFERENCES
[2] Mallick, B and El-Korchi, T., Pavement Engineering: Principles and Practice,
[3] BBC Weather Centre (2009), Average Conditions for Nairobi, Kenya
Available at:
http://www.bbc.co.uk/weather/world/city_guides/results.shtml?tt=TT0003
00
[4] BBC Weather Centre (2009), Average Conditions for London, United Kingdom
Available at:
http://www.bbc.co.uk/weather/world/city_guides/results.shtml?tt=TT0037
90
[5] BS 434-2:2006 Bitumen Road Emulsions - Part2: Code of Practice for the
Use of Cationic Bituminous Emulsions on Roads and Other Paved Areas. British
65