-
Research ArticleDelivering Sustainable Solutions through
Improved Mix andStructural Design Functions for Bitumen Stabilised
Materials
K. J. Jenkins , C. E. Rudman, and C. R. Bierman
Stellenbosch University, Stellenbosch, South Africa
Correspondence should be addressed to K. J. Jenkins;
[email protected]
Received 5 November 2019; Accepted 24 December 2019; Published
12 March 2020
Guest Editor: Alan Carter
Copyright © 2020 K. J. Jenkins et al. +is is an open access
article distributed under the Creative Commons Attribution
License,which permits unrestricted use, distribution, and
reproduction in any medium, provided the original work is properly
cited.
+e evolution of cold recycling using bitumen stabilisation
technology has been supported by progressive research initiatives
andbest practice guidelines. +e first generic guidelines for
bitumen stabilised materials (BSMs) were published only in 2002.
+eseguidelines provided a generic approach for the analysis of
foamed bitumen and bitumen emulsion technologies. From that
point,bitumen stabilisation became the common term for the
inclusion of either of the two bituminous binders. +e TG2 2nd
editionguideline of 2009 took a bold step recognising the shear
properties of the bitumen stabilised material (BSM) as the key
per-formance indicators. In addition, advancements in structural
design and application of BSMs provided practitioners with
robustguidelines. +e subsequent decade has provided an opportunity
to interrogate data from more than 300 BSM mix designs and 69LTPP
sections.+e data have led to research developments including
significant new performance properties of BSMs, refinedmixdesign
methods, and updated new pavement design methods. +is includes an
entire design process that has been updated with astreamlined mix
design procedure and a new frontier curve for the pavement number
design method, as well as a newmechanisticdesign function. It is
anticipated that the research findings and implementation of the
newly developed technology will lead toimproved application in BSM
technology.
1. Introduction
Since the 1950’s, there has been a strong emphasis on
findingefficient, cost-effective pavement rehabilitation
technologies.Part of this history included Prof. Csanyi [1]
experimentingwith hot bitumen and water in order to expand its
volumeand thus improve dispersion in marginal materials at am-bient
temperatures. Csanyi’s development of foamed bitu-men using steam
was taken further by Acott and Myburgh[2] and Ackeroyd and Hicks
[3, 4]. At that time, the impactthat foamed bitumen technology was
about to have onpavement rehabilitation and sustainable practice
was not yetapparent.
In the 1990’s, the patent rights that Mobil bought fromCsanyi
lapsed. In the interim, Mobil developed the foamingtechnology
further by applying accurate dosages of waterrather than steam.
Applications of foamed bitumen tostabilise crushed aggregates,
gravels, sands, and recycledasphalt provided good performance. At
the same time,
bitumen emulsion provided an alternate way to stabilise
baselayers. Both foam and emulsion technologies were
providingflexible and durable bases for a range of
rehabilitatedpavement structures using cold recycling. +e need
forrigorous mix design methods and structural design proce-dures
led South Africa to develop guidelines andmanuals forgranular
emulsion mixes (GEMS) [5], emulsion-treatedbases (ETBs) [6], and
foamed bitumen-treated materials [7].A generic term for the two
bitumen binders was createdunder the title of bitumen stabilised
materials (BSMs). +estructural design of BSM began with the Sabita
GEMS andETB Manuals, which incorporated mechanistic-empirical(ME)
design functions.
In 2002, the first Technical Guideline TG2 was pub-lished, for
foamed bitumen materials. In TG2, a mecha-nistic-empirical (ME)
structural design function wasprovided for pavements including cold
recycling. However,this function was based on only one data set and
was foundto be conservative. In 2009, the second edition of TG2
was
HindawiAdvances in Materials Science and EngineeringVolume 2020,
Article ID 7460174, 10
pageshttps://doi.org/10.1155/2020/7460174
mailto:[email protected]://orcid.org/0000-0001-7934-4313https://creativecommons.org/licenses/by/4.0/https://creativecommons.org/licenses/by/4.0/https://doi.org/10.1155/2020/7460174
-
released and included a new design method, the pavementnumber,
and the associated design equivalent materialslassification System.
+ere was no inclusion of MEfunction.
Ever increasing economic and environmental pressuresdrive the
development of more effective technologies for roadconstruction and
rehabilitation. By 2009, overwhelming evi-dence of the primary
distress mechanism for BSMs beingpermanent deformation and not
fatigue cracking was identi-fied. Although the principle of BSM
being a noncontinuouslyboundmaterial is well understood, the
prediction of long-termstiffness behaviour and permanent strain
development has notbeen developed to the same degree [8].
Long-term pavement performance (LTPP) sections, i.e.,20 field
sites and 7 accelerated pavement testing (APT) siteswith 22
sections tested by the Heavy Vehicle Simulator (HVS)tester formed
the basis of the data for the 2009 guidelines.
Following this trend, the additional performance datarelated to
pavements with both BSM-foam and BSM-emulsion were expanded in the
10 years following 2009.+ispaper addresses the developments in mix
design andpavement design of BSM based on updated performancedata
from the LTPP and APT sections. In addition, a
newmechanistic-empirical design function has been developedin this
period.
+e noncontinuous binding behaviour of BSMs isexplained in the
Asphalt Academy’s Technical Guideline(TG2) [9]. +e failure
mechanism of BSMs, permanentdeformation, is also thoroughly
explained in the TG2. +edesign of BSMs should focus on permanent
deformation inthe layer. +e current design methods, specifically
thepavement number (PN) method, are based on experienceand tend to
be very conservative. +is paper aims to presenta design function
for BSMs which relates mechanical ma-terial properties and stress
conditions to pavement life basedon field observations.
2. Background
2.1.Defining aBSM. Firstly, it is important to provide a
briefdescription of BSM technology and what sets it apart fromgrave
emulsion and cold patching mixes. Bitumen stabili-sation is
typically used in the rehabilitation of existingpavement base
layers (granular, cemented, or asphalt) usingeither foamed bitumen
or bitumen emulsion. +e granularmaterial is treated with small
amounts of bitumen, i.e., lessthan 3%. A small amount of active
filler, typically 1%, is usedin themix to improve the bitumen
adhesion to the aggregate.+e resulting material is a
noncontinuously bound materialthat differs from hot mix asphalt
(HMA) and cement-sta-bilised material. Stabilisation with bitumen
significantlyincreases the cohesion of the aggregate particles with
little orno change in the angle of friction.
+e role of the BSM base in a pavement structure is toprotect the
underlying layers and subgrade from excessivestresses imposed by
traffic. BSMs, with resilient modulusvalues of typically 600 to
1400MPa, effectively achievedistribution the imposed loads. In this
way, the BSM layerprovides intermediate stiffness values between
the high
modulus of the asphalt layers above and granular supportbelow,
thus creating a balanced system to achieve the desireddesign
structural capacity.
2.2. Factors Influencing Permanent Deformation of BSMs.Permanent
deformation is the accumulation of shear de-formation caused by
repeated traffic loading. As BSMs fail inpermanent deformation, a
design function for this materialshould take the factors
influencing permanent deformationinto account. +e factors
influencing permanent deforma-tion of BSMs include
(1) Grading of the parent material(2) Physical properties of the
aggregate particles(3) Density achieved during compaction(4)
Moisture content (including equilibrium) and
moisture susceptibility(5) Number of load repetitions applied to
the material(6) Magnitude of the applied loads(7) Stress history of
the material(8) Lateral or confining pressure(9) +e amount of
active filler added to the mix
Of these variables, four significant factors were identifiedfor
incorporation into a mechanistic design function. +efour selected
factors are: retained cohesion, percentage ofmaximum dry density,
deviator stress ratio, and permanentstrain limit. +is was reduced
to three factors when thepermanent strain limit was included in a
reliability function.+ese factors incorporate an influence of the
minor factorsto some degree, for example, the grading of the
parentmaterial influences the achievable dry density, and the
de-viator stress ratio includes the effect of moisture in the
BSM.
3. Mix Design of BSM
3.1. Review of Mix Design. +e revision of the BSM mixdesign was
encouraged by the evolution of key performanceparameters and
appurtenant test methods. Key consider-ations in this review
included
Material variability: selection of appropriate condi-tioning
methods and reliable test methodsSpecifications: appropriate
guideline limits to providereliable performanceDuration:
streamlining of procedures to minimise themix design time
periodResource economy: minimisation of new testingequipment and
procedures to the essentialsPerformance related: evaluate material
properties witha reliable link to performance
To achieve the objective of meeting the road industry’sneeds,
each set of issues needed to be considered andaddressed.
Fortunately, a database with an excess of 300 mixdesigns using the
latest evaluation approach provided thenecessary platform for
optimising the mix design procedure.
2 Advances in Materials Science and Engineering
-
3.2. Optimisation of Mix Design. A two-part process is re-quired
to optimise a BSM mix. Each step requires testing ofnumerous
specimens to take account of themix compositionvariables and
moisture conditioning of the specimens. Tokeep the sample size of
the aggregate within manageableproportions, ITS tests are selected.
Each specimen is 152mmdiameter and 95mm high. +e increase from the
100mmdiameter specimens in the previous decade is to reduce
thecoefficient of variation of the tests, dictated by the ratio
ofaggregate size to specimen diameter.
Firstly, the selection of active filler must be undertaken.+is
is achieved by using three variables of active filler: 1%lime, 1%
cement or no active filler, and two variables formoisture
conditioning: dry or wet. A standard bitumencontent is added to
each specimen. +ree repeat tests areconducted for each set of
variables.
Secondly, the selection of optimum bitumen content isundertaken.
+is includes the selected active filler, fourvariables of bitumen
content, and two variables for moistureconditioning. +e trend in
ITS results, including the threerepeat tests, is plotted as an
example in Figure 1.
Taking account of material variability of the repeat tests,the
ITS limits from Table 1 can be used to select the designbinder
content on Figure 2.
+e flowchart for BSM mix design is captured in a se-quence
provided in Figure 2. After ITS testing, there is onefinal step,
i.e., triaxial testing. +e challenge was to developreliable but
relatively simple and cost-effective equipmentfor triaxial testing.
+is included
Vibratory hammer compaction method: simulates fieldcompaction in
producing BSM specimens in the lab-oratory. Both ITS and triaxial
specimens are producedthis way.Triaxial cell: comprises an
inflatable tube in a confiningcylinder for testing 150mm diameter×
300mm highspecimens. It enables a standardised testing procedure.+e
shear parameters that are determined serve as a BSMclassification
tool in accordance with Table 2, as well asinput into performance
models for structural design.
For the first time, the shear strength properties of BSMtake
cognisance of the RA content for the classification.Research shows
that the addition of high RA content gen-erally results in an
increase in the cohesion and a slightreduction in the friction
angle. At the same time, a higher RAcontent invariably leads to
improved moisture resistance forthe BSM. +is is captured in an
increased retained cohesionvalue measure for triaxial specimens
that have been con-ditioned under water before testing.
+e new compaction methods and test protocols arecurrently being
tailored into SANS norms, although in theinterim, they will be
included in the revised TG2.
4. Development of a BSM Transfer Function
4.1. Architecture of Transfer Function. +e failure mecha-nism
for BSMs, permanent deformation (or rutting), issimilar to that of
granular materials. +erefore, the transferfunction for BSMs is
based on the design function for
waterbound macadam shown in equation (1) [10]. +ewaterbound
macadam transfer function calculates thebearing capacity in terms
of the number of standard axleload repetitions (N) that can be
sustained before a certainlevel of plastic strain is induced in the
layer:
logN � 1.891 + 0.075(RD) − 0.009(S) + 0.028(PS)
− 1.643(SR),(1)
where N� number of standard axles the layer can sustainbefore
reaching the deformation limit, RD� relative density
Added bitumen (%) foam or emulsion
Indi
rect
tens
ile st
reng
th IT
S (k
Pa)
o
1.8 2.0 2.2 2.4
100
150
200
250
300
XXX
X
XX
XXX
XXX
BSM1Min ITSWET
BSM1Min ITSDRY
oo
ooo
ooo
ooo
Figure 1: Selection of bitumen content from ITSDRY and
ITSWET.
Table 1: Indirect tensile strength limits for
classification.
ClassITS limits
ITSDRY (kPa) ITSWET (kPa)BSM1 >225 >125BSM2 >175
>100
Pavement evaluationand sampling
Compliance testing Determine blendrequirements
Determination of optimumbitumen content
ITS testing
Determination of the shearproperties
triaxial testing
Is the materialsuitable?
Active filler selectionlime vs. cem vs. none
ITS testing
No
Yes
Figure 2: Flowchart for BSM mix design steps.
Advances in Materials Science and Engineering 3
-
(%), S� saturation (%), PS� plastic strain limit as a
per-centage of the layer thickness (%), and SR� stress ratio
(—)
+e resistance to permanent deformation of unboundgranular
materials under repeated loading can be improvedby increasing the
density of the material [11]. +e durabilityand performance of a BSM
mix depends on its level ofcompaction [12]. Moisture damage
contributes significantlyto the deterioration of pavement
materials, including BSMs.A reduction in shear strength through
moisture ingress, i.e.,a higher degree of saturation (S), results
in an acceleratedrate of permanent deformation [13].
Finally, the rate of permanent deformation accelerateswith an
increase in deviator stress and decreasing confiningpressure [14].
+e stress ratio (SR) used in this equation is afunction of the load
intensity, the shear properties of thematerials, as well as the
overall pavement structure.
4.2. Stellenbosch BSM Transfer Function. Using the archi-tecture
of design function for granular type behaviour, aBSM function can
be developed.+e new transfer function isbased on 14 different roads
and 22 different analysis sectionsas follows:
logN � A − B(DSR)3 + C Pmod · RetC( + D, (2)
where DSR� deviator stress ratio as a fraction (—),Pmod �maximum
dry density as a percentage of modifiedAASHTO density (%), RetC�
retained cohesion (%),A� constant based on design reliability, and
B, C, andD� constants from data correlation (Table 3).
+e DSR term describing the effect of deviator stress onpavement
performance yielded a power three function,underlining the
importance and sensitivity of this term. +eDSR is evaluated at a
depth of 25% of the BSM layer thickness:
DSR �σ1 − σ3σ1,f − σ3
�σ1 − σ3σ1,f − σ3
, (3a)
σ1,f �(1 + sinϕ) · σ3 + 2 · C · cos ϕ
(1 − sinϕ), (3b)
where DSR� deviator stress ratio as a fraction, σ1
�majorprincipal stress (kPa), σ3 �minor principal stress,
i.e.,confining pressure (kPa), σ1,f �major principal stress
atfailure (kPa), C� cohesion of the BSM default value (kPa),F�
friction angle of the BSM default value (°), andF� friction angle
of the BSM default value (°).
4.3. Calibration of BSMDesign Function. +e purpose of
thiscalibration is to determine values for the constants in
equation (2) and to best describe the relative influence ofeach
of the input variables on BSMs. +e transfer functionrelates the
number of standard axles the pavement canaccommodate with the
remaining amount of permanentstrain in the BSM layer, i.e., a
limiting value set in by designcriteria. +is enables the designer
to specify the amount ofpermanent deformation that may occur before
the BSM isdeemed to have failed, linked to the design
reliability.
Figure 3 shows the comparison of the accumulation ofplastic
strain due to repeated loading (N actual) compared to theprediction
of remaining life, using the new transfer function (NTF). +e
calibration process aims to reduce the differencebetween the
estimated life and the actual life of BSMs.
4.4. Long-Term Pavement Performance Data. +e transferfunction
can be calibrated to perfectly describe a specific caseif the
information is sufficient. However, this will not beuseful as it
will only describe the life of a specific BSM on alocal scale,
limiting the relevance for design. By investigatinga number of
pavement structures, with data available atmultiple points in time,
the transfer function can be cali-brated to describe the life of
BSMs on a more extensive scale.
Data such as the material properties and subsequentlong-term
performance of pavement structures were gath-ered and analysed in
the permanent deformation model. Foran overall representation of
the properties and performanceof BSMs, results of fourteen
long-term pavement perfor-mance (LTPP) studies were interrogated.
+is was part of along-term pavement performance study, by SANRAL.
Eachof these pavements either had a BSM 1 or BSM 2 base layerwith a
minimum thickness of 100mm. In total, 69 LTPPsections have played a
role in data provisions and shapingthe design functions.
+e data available for these pavements included densi-ties,
moisture contents, layer thicknesses, material types, andclasses as
well as traffic data. +e data also included FWDand rut depth
measurements at different stages in their fieldlife. +e information
used during the analysis of thesepavements is discussed in detail
by Long and Jooste [15]. Asummary of the information available for
these pavements ispresented in Table 4.
Table 2: Shear parameter limits for triaxial tests.
Class RA (%)Triaxial
Cohesion (kPa) Friction angle (°) Retained cohesion (%)
BSM 1
-
Areas of insufficient data required assumptions to bemade based
on the performance and conditions of thesepavements. Where cohesion
(C), friction angle (ϕ), andretained cohesion values were not
available, realistic andrepresentative default values were used.
Reasonable andconservative values were applied for each pavement,
basedon the minimum design values specified in the TG2 [9].
+e data obtained from these pavements were used forthe
calibration of the transfer function. +e structures ofeach of these
pavements was modelled in Rubicon Toolboxto determine the critical
(highest) DSR value when subjectedto an 80 kN standard axle load
(E80). +e software requiredthe layer thicknesses and material
properties for each of theLTPP pavements to calculate the DSR.
4.5. Permanent Strain Development with Repeated Loading.+e new
BSM transfer function uses permanent strain as thelimiting factor
when determining the life of a BSM. During
the analysis of the pavements used to calibrate the
transferfunction, rutting measurements were taken at
differentpoints in time. +e rutting measurements reflect the
per-manent deformation of the pavement structure as a whole.
Apercentage of the total permanent deformation was used toobtain
the permanent deformation within the BSM baselayers.
+e accumulated permanent strain in the BSM layers ofthe LTPP
pavements varies for each pavement. +is poses achallenge for the
calibration of the transfer function.Pavements that showed little
deformation would reflect earlylife permanent deformation
behaviour, while pavementsthat showed more significant deformation
would reflect thelong-term permanent deformation behaviour. +is
onlyallows the rate of permanent deformation accumulation
fordifferent pavements to be compared and not the
absolutevalues.
In order to obtain values for the actual traffic at higherstrain
levels, the Huurman model for permanent straindevelopment was
implemented for each pavement. +eHuurman model for PS development
was used to determinethe number of load repetitions at each of
these PS levels up tothe final level of 10%.+e Huurman model for PS
predictionis as follows [16]:
εp � A ·N
1000
B
, (4)
where N� number of load repetitions, εp � plastic
strainaccumulation (%), and A and B� regression constants.
4.6. Calibration of the New BSM Transfer Function. +e
dataobtained from the fourteen LTPP pavements were used forthe
calibration of the transfer function. +e transfer func-tion’s
prediction of the number of load repetitions (N TF) toreach a
specific level of permanent strain was compared tothe estimated
traffic (N estimate) that the various pavementsaccommodated to
reach that strain level.
Figure 4 illustrates a comparison of the transfer func-tion’s
predictions compared to the estimated traffic over arange of
permanent strain values. An estimation was con-sidered good if it
predicted a value close to the estimatedvalue. By using a large
number of data points, the slope oflinear trendline through the
data was used to determine thefunction’s accuracy. A slope greater
than one indicates thatthe function overestimates the life of the
layer, whereas aslope smaller than one would indicate
underestimation.
+e transfer function was calibrated using linear re-gression to
determine the values of A, B, C, and D thatyielded a slope of one,
while minimizing the dispersion ofdata points. +e comparison
between the calibrated transferfunction and the estimated traffic
is shown in Figure 5. +ecalibrated transfer function is shown in
equation (5). In thisfigure, the slope of the trendline (the dotted
line) correlatesvery well with the one-to-one relationship.
+e trendline for the calibrated transfer function wasy � 1.0006x
+ 1.0572.+e slope of the trendline was deemedto be a good
representation of the estimated life. However,the intersect point
(1.0572) indicates that the transfer
0
2
4
6
8
10
12
14
PS (%
)
0 2 4 6 8 10 12N actual
N actualN TF
Minimizedifference bycalibration
Figure 3: Transfer function prediction compared to actual
loadrepetitions [10].
Table 4: LTTP pavement information [10].
Road BSM constructionyearBSM thickness
(mm)
Standard axlesaccommodatedto date (MESA)
MR27 1988 100 5MR504(A) 1995 175 1.6
MR504(B) 1995 175 1.6
N1-1 1984 100 17N1-13 1980 150 14N1-14 1980 150 14N2-16 1980 140
3.4N2-20 2000 180 2.4N4-1 1997 170 5.6N4-5X 1996 150 6.7N11-8 2004
280 1.1N12-19(3) 1974 100 18
N12-19(4) 1974 135 18
P243-1 2000 250 0.48
Advances in Materials Science and Engineering 5
-
function may overestimate the life of the material at
lowremaining strain values.+is is due to the logarithmic natureof
the function, which produces positive results even whenthe
remaining strain is very small (or zero). +erefore, it
isrecommended that a limit to the minimum remaining strainis set in
place for design purposes.
Permanent strain development obtained from trafficestimation was
compared to that of the transfer function foreach of the LTTP
pavement structures. Figure 6 shows thisrelationship between
predicted and actual life for a section ofthe MR502. It highlights
significant differences in correla-tion at low strain levels. +e
slope of transfer function’spredicted vs. actual life at higher
strain levels is close toparity. +is indicates that the transfer
function accuratelypredicts the long-term permanent strain
development of theBSM.
4.7. SafetyAdjustments forDesign. +enew transfer functionfor
BSMs is calibrated to best describe the observations madefrom the
LTPP data.+is function only aimed to describe theobserved trends. A
design function, however, requires dif-ferent levels of reliability
to be explicitly built into the
function. Based on South African pavement design, fourcategories
and appurtenant reliability have been incorpo-rated into the
transfer function, i.e., category A (95% reli-ability), category B
(90% reliability), category C (80%reliability), and category D
(50%).
+e calibration process was repeated by only adjustingthe value
of constant A, which adjusts the prediction offunctions without
altering the relationship between thefunction’s variables.
Reliability in the transfer function wasmeasured as the percentage
of data predicted by the transferfunction that had a smaller value
than the observed data.Figure 7 illustrates the principles used to
calibrate thetransfer function to different levels of reliability.
It is shownthat the transfer function produced lower life
predictionsthan was estimated for 90% of the data points.
+e new design function’s prediction of the LTPP lifecompared to
the estimated life is shown in Figure 8. +eslope of the transfer
functions predictions compared to theestimated life reduces with
the increase in reliability. +ereduction in slope indicates that
the transfer function un-derestimates the life of the material by
higher margins. +isunderestimation increases the probability of a
designedmaterial to achieve the calculated design life.
0
2
4
6
8
10
12
14
16N
TF
Before calibrationAfter calibration
1
1
Calibration
0 2 4 6 8 10 12N estimate
Figure 4: Measure of transfer function’s accuracy [10].
05
1015202530354045
0 10 20 30 40
N T
F
N actual
Predicted vs. actual life1:1 relationship
Figure 5: Predictions of the calibrated transfer function
comparedto actual traffic [10].
0.01.02.03.04.05.06.07.08.09.0
10.0
0.0 2.0 4.0 6.0 8.0
PS (B
SM) (
%)
N actual (MESAs)
Huurman modelTransfer function
Figure 6: Transfer function’s permanent strain prediction
forMR504 (A) [10].
1
10
4.3
N (M
ESA
)
PS (%)
N actual90%
90% of transferfunction
Figure 7: Reducing the transfer function’s predictions to
increasereliability [10].
6 Advances in Materials Science and Engineering
-
+e new calibrated BSM design function is shown inequation (5).+e
design function can be used to estimate thelife of BSMs for design
purposes. However, this function’suse should be compliant with
rules outlined in TG2 [17]where allocated materials stiffness and
analysis positions inthe BSM layer are outlined.
logN � A − 57.286(DSR)3 + 0.0009159 Pmod · RetC(
+ 0.86753.(5)
5. Design Method Comparison
As part of the validation process, the new function wascompared
to the older design methods for BSMs. In Bier-man’s research [10],
five different pavement structures wereinitially investigated, and
the results were compared. +eseanalyses were carried out before a
structural pavementdesign procedure was developed. +e methods
included aheuristic design pavement number (PN) method and
themechanistic-empirical (ME) method. Unfortunately, theanalysis
was carried out before the PN method, which re-quired updating with
new data and recalibration. Never-theless, reasonable pavement life
comparisons weredeveloped.
+e Stellenbosch BSM design function, however, could beapplied
using the updated TG2 [17], thus providing morerealistic insights.
+ese results are presented in this paper.Focus was placed on a
comparison with the old and new PNdesignmethods [7, 18], and the
lives of ten different pavementstructures were compared. Only two
of these structures arehighlighted in this paper (see Table 5).
+ere are fundamental differences between the heu-ristic PN
design method and the ME method. +e PNprovides an estimate of
pavement life of the entirepavement structure based on performance
of the LTPPpavements with BSM layers. +e architects of the
PNmethods used ME analysis to develop the design model, soa design
only needs to input the pavement materials’classifications and
structure. +e Stellenbosch BSM designfunction interrogates the
structural mechanics of eachlayer to determine the “weakest link”
and thus determinethe design life.
+e comparison between the design life of the selectedpavements
using the different design models is provided inFigures 9(a) and
9(b). +e variables included in the lifecalculation include
Design models: PN method old-2009 and new-2019 aswell as
Stellenbosch BSM ME Design functionReliability levels: 95%
(category A) and 90% (category B).
0.0
5.0
10.0
15.0
20.0
25.0
30.0
35.0
40.0
0 10 20 30 40
N T
F
N actual
50%85%
90%95%
Figure 8: Underestimation of life at different levels of
reliability [10].
Table 5: Pavement structures for design comparison [10].
Pavement Pavement layer Material class +ickness (mm) Stiffness
(MPa) BSM properties
1
Surfacing HMA 35 3500 C (kPa) 250Base BSM 1 250 720 φ (°) 40
Sub-base G5 250 240 RetC (%) 75Selected subgrade G8 180 120 Pmod
(%) 100
Subgrade G8 N/A 90
2
Surfacing HMA 20 2500 C (kPa) 225Base BSM 2 280 500 φ (°) 39
Sub-base G6 250 200 RetC (%) 70Selected subgrade G8 150 100 Pmod
(%) 98
Subgrade G8 N/A 90
Advances in Materials Science and Engineering 7
-
Layer ME analysis of BSM: single layer vs. sublayer (atleast
100mm)
BSM Classification: BSM1 vs. BSM2
Default values for BSM properties based on the TG2[17]
guidelines
+e pavement life determined using the PN methodcompares well
with the new Stellenbosch ME BSM designfunction. Sensitivity to
design reliability is evident. Sub-layering of the BSM layer and
using an analysis position at1/4 depth of the layer provide a more
realistic design life. Itshould be noted that sublayering provides
a more
35mm HMA
250mm BSM1
250mm G5
180mm G7
Infinite G8
3500 MPa
720 MPa
240 MPa
120 MPa
90 MPa
0
5
10
15
20
25
Des
ign
life (
MES
As)
Moderate climate and category (on bar)
PN (old)PN (new)PN (new)
Stell (single)Stell (sub)Stell (sub)
A A B A A B
(a)
B B B+ B B B+
20mm HMA
280mm BSM2
250mm G6
150mm G8
Infinite G8
2500 MPa
500 MPa
200 MPa
100 MPa
90 MPa
01234567
Des
ign
life (
MES
As)
Moderate climate and category (on bar)
PN (old)PN (new)PN (new)
Stell (single)Stell (single)Stell (sub)
(b)
Figure 9: (a) Design life of pavement 1 with BSM1 base vs.
design method. (b) Design life of pavement 2 with BSM2 base vs.
design method.
Option 4Option 3
0
100 000
200 000
300 000
400 000
500 000
600 000
700 000
800 000
Constr cost LCC (disc = 6%) Energy cost LCC + energy cost
Cos
t per
opt
ion
(US$
/km
)
Option 1Option 2
Figure 10: Combined life cycle cost and energy consumption
[19].
8 Advances in Materials Science and Engineering
-
representative density profile and resilient modulus profilefor
a thick layer of material that is stress-dependent.
6. Economics
Jenkins and Collings [19] compared four different
rehabil-itation strategies in terms of life-cycle costing and
energyconsumption, based on actual projects. +e four interven-tions
were
(1) Patch and overlay(2) Mill and replace(3) Cement stabilise
the existing base + surfacing and
overlay(4) Recycle using bitumen stabilisation technology.
It is obvious in Figure 10 that initial construction costsalone
are inadequate for selection of rehabilitation alter-natives. +ey
can provide skewed and unrealistic rehabili-tation selection, which
will lead to unnecessary wastage ofresources. Cheapest is the
dearest.
+e whole-of-life analysis using PWOC (present worthof costs)
provides more realistic financing requirements forpavement upkeep
over the entire analysis period. +ecombined energy and life cycle
costs provide the most in-formative insight into sustainable
solutions. For this par-ticular project, bitumen stabilisation
technology scores asthe most cost-effective and sustainable
solution forrehabilitation.
In this case, energy is used as a surrogate for emissionstoo;
however, life cycle analysis (LCA) approach would bethe ultimate
way to analyse different rehabilitationinterventions.
7. Conclusions
+e past decade has provided the opportunity to gatherinvaluable
performance data of BSMs from several hundredmix designs and 69
LTPP sections. In turn, these data haveprovided the opportunity to
upgrade the design and ap-plication and bitumen stabilised
materials (BSMs). Some ofthe highlights of areas where BSMs are
being taken forwardby many strategic developments include
Guidelines for a refined mix design method for
BSMsincluding:
Vibratory hammer compaction of ITS and triaxialspecimensPhased
analysis to determine active filler (type andcontent) and binder
content for optimal performanceTriaxial testing of the proposed mix
to evaluate shearproperties at equilibrium and wet conditioning to
beused in material classification and provide inputs intostructural
designCertification of all test procedures using SANSstandards
An updated material classification system that usessignificantly
more materials data to provide a morerobust method for evaluation
and design
An upgrade pavement number design system thatremoved biases and
provides for a greater range oftypes of pavement structures and
more accurate designoutcomes based on extended dataA new
mechanistic-empirical design system for BSMsbased on extensive
analysis of LTPP sections andperformance parameters of the
BSMEconomic and environmental (energy consumption)analyses provide
important insight into the sustain-ability of rehabilitation
options and highlights theadvantages of cold recycling
technology
+e improved understanding of key performance pa-rameters of BSMs
and implementation of these findings inthe application of the
technology will offer more effectiveand reliable solutions in
pavement rehabilitation.
Data Availability
+e research data that have been used in this publicationemanates
from research at Stellenbosch University and isclearly referenced.
Additional data can be obtained from theaccredited publications in
the reference list.
Conflicts of Interest
+e authors declare that they have no conflicts of interest.
Acknowledgments
+e authors acknowledge the financial support provided byRoadmac
Surfacing Cape toMr Carl Bierman in carrying outthis research
project.
References
[1] L. H. Csanyi, Foamed Asphalt in Bituminous Paving
Mixes,Highway Research Board Bulletin, Washington, DC,
USA,1957.
[2] S. M. Acott and P. A. Myburgh, “Design and performancestudy
of sand bases treated with foamed asphalt,” Trans-portation
Research Record, vol. 898, pp. 290–296, 1982.
[3] F. M. L. Akeroyd and B. J. Hicks, Foamed Bitumen
RoadRecycling, Alad Limited, London, UK, 1988.
[4] F. M. L. Akeroyd, “Advances in foamed bitumen technology,”in
Proceedings of the 5th Conference on Asphalt Pavements forSouthern
Africa, Mbabane, Swaziland, June 1989.
[5] SABITA, GEMS—:e Design and Use of Granular EmulsionMixes,
Southern African Bitumen Association, Cape Town,South Africa,
1993.
[6] SABITA, ETB—:eDesign and Use of Emulsion Treated
Bases,Southern African Bitumen Association, Cape Town, SouthAfrica,
1999.
[7] SABITA, :e Design and Use of Foamed Bitumen
TreatedMaterials, Southern African Bitumen Association, CapeTown,
South Africa, 2002.
[8] D. Collings and K. Jenkins, “+e long-term behaviour
ofbitumen stabilised materials,” in Proceedings of the
10thConference on Asphalt Pavements for Southern Africa,
Dra-kensberg, South Africa, 2011.
[9] Asphalt Academy, Technical Guideline TG2: Bitumen
Stabi-lised Materials: A Guideline for the Design and Use of
Bitumen
Advances in Materials Science and Engineering 9
-
Emulsion and Foamed Bitumen Stabilised Materials,
AsphaltAcademy, Cape Town,South Africa, 2009.
[10] C. R. Bierman, A Design Function for Bitumen
StabilisedMaterial Performance Based on Laboratory and Field
Evalu-ation: MEng (Research), Stellenbosch University,
Stellenbosch,South Africa, 2017.
[11] A. A. Van Niekerk, “Mechanical behavior and performance
ofgranular bases and sub-bases in pavements,” Doctoral
degree,Technical University Delft, Delft, Netherlands, 2002.
[12] K. J. Jenkins, “Mix design considerations for cold and
half-warm bituminous mixes with emphasis on foamed
bitumen,”Doctoral degree, University of Stellenbosch,
Stellenbosch,South Africa, 2000.
[13] F. Lekarp, U. Isacsson, and A. Dawson, “State of the art II
:permanent strain response of unbound aggregates,” Journal
ofTransportation Engineering, vol. 126, no. 1, 2000.
[14] S. Werkmeister, Permanent Deformation Behaviour of Un-bound
Granular Materials in Pavement Constructions, Tech-nischen
Universitat Dresden, Dresden, Germany, 2003.
[15] F. Long and F. Jooste, Summary of LTTP Emulsion andFoamed
Bitumen Treated Sections, Technical Memorandum,Sabita, Cape Town,
South Africa, 2007.
[16] M. Huurman, “Permanent deformation in concrete
blockpavements,” Doctoral degree, Technical University Delft,Delft,
Netherlands, 1997.
[17] Asphalt Academy, Technical Guideline TG2: Bitumen
Stabi-lised Materials: A Guideline for the Design and Construction
ofBitumen Emulsion and Foamed Bitumen Stabilised Materials,Asphalt
Academy, Pretoria, South Africa, 3rd edition, 2020.
[18] Asphalt Academy, Technical Guideline TG2: Bitumen
Stabi-lised Materials: A Guideline for the Design and Construction
ofBitumen Emulsion and Foamed Bitumen Stabilised Materials,Asphalt
Academy, Pretoria, South Africa, 2nd edition, 2009.
[19] K. J. Jenkins and D. C. Collings, “Combining LCC and
energyconsumption for enhancing decision making regarding
re-habilitation options,” in Proceedings of the 10th Conference
onAsphalt Pavements for Southern Africa, Drakensberg, SouthAfrica,
2011.
10 Advances in Materials Science and Engineering
