7th CONFERENCE ON ASPHALT PAVEMENTS FOR SOUTHERN AFRICA

DESIGN RECOMMENDATIONS FOR HIGHLY FLEXIBLE SURFACINGS

K M Muthen
A O Bergh

CSIR Transportek
P O Box 395
Pretoria,0001
South Africa

Abstract

In South Africa, the successful use of surface seals has been widespread, even on
relatively heavily trafficked roads. In terms of traffic and maintenance costs, a significant
number of rural and urban unpaved roads may become candidates for surfacing in the near
future. This has fuelled the demand for the use of seals as a cost-effective upgrading option
for unpaved roads, an application which requires a highly flexible surfacing.

Surfacings on highly flexible light pavements can be expected to have reduced service lives
and high risks of failure because of the aggravating service conditions of high deflections
and possible marginal base quality. Although all thin surface dressings (seals) may be
regarded as flexible by comparison with asphalt, highly flexible surfacings refer to those
seals which offer performance characteristics superior to those of traditional surfacings
under conditions of abnormal pavement deflections. The design and maintenance
considerations of the surfacing is therefore a critical aspect in the overall design of such
pavement structures.

The design of highly flexible surfacings using the conventional seal design would be
inadequate, since the functions of highly flexible surfacings differ from those of
conventional seals. While the main functions of conventional seals is to provide a skid-
resistant surfacing and prevent the ingress of water, the waterproofing role of highly flexible
surfacings is usually critical in preventing pavement failure, while skid-resistance may be
of secondary consequence on low volume roads. Conventional methods of seal design
concentrate on optimising the skid resistance of the surfacing. In contrast, the prime
objective of a design procedure for highly flexible surfacings would be to optimise the
flexibility and fatigue resistance of the surfacing. Furthermore, conventional seal design
methods do not adequately take the enhanced properties of modified binders and the
characteristics of the base into account.

This paper reflects on recent research, conducted at CSIR Transportek on behalf of

SABITA, on the application of highly flexible surfacings. The investigation focussed on the
distress mechanisms, design recommendations and an economic analysis of highly flexible
surfacing options used in South Africa. In addition to binder considerations, other factors
which also play a role in the performance of such surfacings, such as the quality of the base
materials and the climatic conditions, were addressed.

Muthen KM et al -1- CAPSA’99

 Design recommendations for highly flexible surfacings

1. INTRODUCTION

Pavement design methods in use today are generally more geared for dense, heavy traffic
than for low trafficked roads, and are often very critical and demanding in terms of the
quality of materials. However, experience and research have indicated that, even at
relatively low traffic volumes, the upgrading of unpaved roads to sealed standards is cost
effective when compared with the cost of maintaining the unpaved gravel road, provided
that concessions with regard to Draft TRH4 (1996) and TRH14 (1987) pavement materials
specifications are made (Visser and van Niekerk, 1987; Paige-Green, 1994). As a result,
the use of surface treatments over light (marginal materials) pavement structures for the
upgrading of rural and urban unpaved roads has been widespread in South Africa. The
main reason for this is the substantial cost savings associated with light pavement
structures by comparison with the prohibitive costs of upgrading these roads to Draft TRH4
(1996) and TRH14 (1987) materials standards, which usually require crushed stone bases
and cemented layers. Light pavements have the structural capacity to carry high traffic,
provided that water can be prevented from entering and weakening the materials in the
supporting granular structure. In addition to providing adequate levels of service, the
sealing of unpaved roads has the advantages of :
C generating savings in construction cost due to use of local materials and natural
gravels;
C preventing the loss of gravel, resulting in elimination of the costs of replacing gravel;
C generating savings in vehicle operating costs due to improved riding quality;
C generating savings in maintenance costs to maintain an acceptable level of service
on gravel roads (dust and grader maintenance);
C the provision of access in all weather conditions, and
C providing a sound structure for future rehabilitation.

It is envisaged that there will be an increasing demand for the application of light pavement
technology for the upgrading of unpaved roads in developing communities in line with
social upliftment programmes in South Africa. This would typically be applicable to:
C Suburban streets and minor rural roads with traffic counts of less than 150 - 250 vpd
(Class III).
C Collectors and/or access roads to townships with traffic counts of 250 - 2000 vpd
(Class II).

The use of light pavements coupled with the promotion of ETB as a labour-intensive
stabilisation/modification option for marginal materials also has a significant impact on the
performance of the surfacing as these would result in more flexible pavements than
traditional cement or lime stabilised bases. A critical requirement for good performance
of these structures is a durable impermeable surfacing. Paradoxically, the risk of surfacing
failure under these conditions is high for the following reasons:
C lighter pavement structures are subject to large deflections which, depending on the
nature of the deflection bowl, can induce higher strains in the surfacing than
conventional pavements, resulting in premature cracking;
C the likely use of substandard pavement materials, the strength of which could be
highly moisture dependent;

Muthen KM et al -2- CAPSA’99

 Design recommendations for highly flexible surfacings

C the maintenance capability and control of vehicle overloading in rural areas may be
limited;
C increased rate of oxidation because oxidation occurs from the base (due to the
presence of air voids) as well as the atmosphere;
C low traffic volumes may contribute to the degradation of the surfacing due to a lack
of traffic moulding of the binder;
C drainage related problems which are prevalent in rural areas (surface and
subsurface), and
C rutting which results in stretching and thinning of the seal and consequent
weakening.

Consequently, in order to ensure good performance of light pavements, special

consideration should be given to the design, construction and maintenance of surfacings
placed over these structures.

The main thrust of this paper focuses on surface seals placed on low volume, light
pavement structures, which experience high deflections under standard axle loads. The
mechanisms governing the behaviour of highly flexible surfacings on these structures will
be discussed and the likely distresses identified from previous research and experience.
In particular, the impact of high deflections, marginal base materials and climate on the
performance of the surfacing will be considered. This will culminate in recommendations
to optimize the performance of thin surfacings on generally weak structures without the
need for expensive rehabilitation of the structural layers.

The results of an industry survey, conducted to identify any specialized highly flexible
surfacing products, as well as an economic analysis of the identified products are
discussed.

1.1 Definition of Highly Flexible Surfacings

A highly flexible surfacing can be described as a thin surface dressing (seal) which offers
performance characteristics superior to those of traditional surfacings under conditions of
abnormal pavement deflections (both elastic and non-recoverable). Their application is
therefore suitable for flexible, non-cemented pavement types with possible marginal
materials. Highly flexible surfacings are most likely to derive their benefits from modified
bituminous binders with their enhanced elasticity and enhanced resistance to ageing and
fatigue.

2 INDUSTRY SURVEY

An industry survey was conducted in order to identify suitable highly flexible surfacing
products available in the South African marketplace. This task was accomplished by
circulation of a questionnaire among the major members of the asphalt industry.

According to the responses received, a highly flexible surfacing should ideally have the
following properties in addition to the normal requirements of seal materials:
C the binder should be soft with high elastic recovery and reduced permanent

Muthen KM et al -3- CAPSA’99

 Design recommendations for highly flexible surfacings

deformation at high temperatures;

C high percentage of elastomer (SBR or SBS);
C high binder application rates; and
C elasticity and a high resistance to cracking at low temperatures.

2.1 Products Identified during the Industry Survey

All the highly flexible products identified were modified bituminous binders, containing
various polymers to enhance the flexibility and reduce the temperature susceptibility of
binders. These were:
C Rubspray Emulsion or Vialistic 65/5 Emulsion (Cationic spray grade 65 per cent
containing 5 per cent SBR latex)
C Vialistic 70/5 Emulsion (70 per cent Cationic spray grade containing 5 per cent
SBR latex)
C RMB3 or Viagrip 3 (Penetration Grade bitumen containing 3 per cent SBR)
C Elastomod 3 (Penetration Grade bitumen containing 3 per cent SBS)
C Elastomod 4 (Penetration Grade bitumen containing 4 per cent SBS)
C Rubberized Gravel Seal (Cutback, semi-priming SBS binder for gravel roads)
C RMB5 (Penetration Grade bitumen containing 5 per cent SBR)
C ARM-R-Shield Bitumen Rubber
C EVA Polymer-modified Bitumen

Table 2.1: Material Costs based on recent projects (courtesy of COLAS South
(Pty) Ltd.)
13.2/6.7 Double Seal using Vialastic 65/5 R5.55/m²
13.2/grit Double Seal using Vialastic 65/5 R4.58 /m²
13.2 Single Seal using Viagrip 3 (RMB3) R3.74 /m²
13.2 Single Seal using Elastomod 3 R4.10 /m²
Graded Seal using Rubberised Gravel Binder R4.24/m²
NB. Costs include binder, delivery, spraying and aggregate and exclude labour, equipment and VAT.

3. DETERIORATION MECHANISMS

Thin pavements and pavements constructed of weak materials are expected to have high
deflections. This is because deflection is directly related to the elastic modulus of the
underlying pavement layers which, in turn, is proportional to properties such as CBR and
in situ density. Furthermore, the seasonal variation in the moisture content reduces the
shear strength of the base and underlying layers, resulting in higher deflections during the
wet seasons. These high deflections result in reduced surfacing and consequently in
reduced pavement lives as shown in Figure 3.1, for non-modified binders. Jooste (1997),
on evaluation of the performance models used to generate Figure 3.1, found that a
calibration factor of approximately 0.8 was applicable for seal lives in the Gauteng
province.

Muthen KM et al -4- CAPSA’99

 Design recommendations for highly flexible surfacings

Figure 3.1: Effect of surface deflection on seal life.

(Paterson, 1987)

The predominant form of distress affecting surfacings on highly flexible pavements is crack
formation. This may occur as a result of fatigue or brittle failure of the binder. Permanent
deformation of the surfacing at high temperatures is also possible. On marginal bases,
other problems such as bleeding, poor adhesion, punching, ravelling and potholes are
common.

3.1 Cracking of the Surfacing

The main component affecting the ability of the surfacing to withstand cracking is the
binder. The rheology of a bituminous binder allows it to deform and relieve the stresses in
the surfacing caused by deflections of the pavement. However, the maximum strain that a
particular binder can withstand is dependent on a variety of factors such as:
C chemical composition;
C operating temperature, and
C degradation (hardening) resulting from the loss of volatile solvents and oxidation.

The binder should be capable of conforming to the deflections in the coldest conditions
expected otherwise cracking will result. Once cracks have occurred, ingress of water will
usually result in rapid degradation of the surfacing and pavement.

Chemical modification can be used to improve the flexibility, durability and temperature
susceptibility of bitumen binders. Special polymer additives and crumb rubber are being
used for this purpose by binder producers in South Africa. Modification of bitumen with a
variety of polymers and rubber is known to result in surfacings with the following properties:
C increased flexibility enabling them to withstand larger strains both as a result of
active cracks as well as of flexible substrata;
C reduced incidence of bleeding at high temperatures, implying that relatively higher

Muthen KM et al -5- CAPSA’99

 Design recommendations for highly flexible surfacings

spray rates can be used;

C improved durability and extended service life by slowing oxidation and hardening;
C flexibility at low temperatures;
C improved shear resistance and aggregate retention properties, and
C improved adhesion to aggregate.

The long-term performance of the surfacing is affected by the durability of the binder with
respect to ageing and hardening. From the Australian experience (Dickinson, 1982), the
main factor which determined the seal life was the rate of hardening of the binder. This
hardening occurs as a result of the loss of volatile solvents and, more significantly, reaction
of the binder with oxygen. Higher temperatures also increase the rate of oxidation. It was
also noted that a limiting value for the viscosity existed, above which distress did not occur,
with the limit varying according to climatic regions (Oliver, 1984). In South Africa, Paige-
Green (1994) observed that ‘the hardening of bitumen with time resulted in a decreased
seal flexibility with concomitant cracking under high deflections’. Generally, the rate of
ageing in surface treatments is high because the surface area of binder exposed to the
atmosphere is large, while the diffusion path for oxygen is small. Seals placed directly on
gravel roads are also exposed to oxidation from below, due to the presence of oxygen in
the voids of the gravel layer. A means of increasing the service life of seals is to increase
the binder content, thus effectively reducing the oxidation rate. However this translates
directly into higher construction costs, and to the need for countermeasures to reduce
bleeding.

It is common practice in South Africa to increase binder spray rates when surfacings are
constructed on poor quality base materials, which tends to aggravate the bleeding problem.
However, Paige-Green (1994) noted that the most successful surfacings investigated had
high binder contents which, in his opinion, helped to maintain the integrity of the surfacing.
He also contended that the associated bleeding is of lesser consequence on low volume
roads than the presence of cracks.

Light applications of diluted emulsions are used as a routine maintenance measure to

enrich the aged binder and to restore some flexibility to the surfacing, but the benefits are
limited for periods up to two years.

The cyclical deformation in flexible pavements results in cyclical flexural stresses in the
surfacing. Like most materials, bituminous binders experience fatigue. The adhesion
between the binder and the aggregate is likely to be affected by fatigue. In the case of
pavements which have cracks in the base course, these cracks will quickly reflect through
surfacings with low fatigue strength. Findings by Coetser (1994) suggest that, based on the
CMS test, binders with high stiffness (high E modulus) have a lower fatigue life than binders
with low stiffness, which implies that as the binder ages and hardens, its fatigue resistance
also decreases. Bitumen-rubber seals have proven track records at retarding propagation
of cracks from the underlying layers. Although the mechanisms involved in crack
propagation differ from those of elastic deformation, intuitively, the benefits of seals with
modified binders are applicable to highly flexible pavements.

Muthen KM et al -6- CAPSA’99

 Design recommendations for highly flexible surfacings

3.2 Other Distresses

Paige-Green (1994) highlighted several surfacing distress modes encountered during a
study on marginal base materials:
C poor adhesion between seal and base resulting from degradation of the base
materials at the interface;
C surface distress associated with large stones near the surface of the base. This
resulted from poor adhesion of the seal to the large stones and low densities in
these areas, and
C embedment of the seal aggregates into soft bases (punching), resulting in bleeding.

The often critical moisture-strength relationship of marginal materials results in significant

weakening of the pavement when water enters it. Heavy Vehicle Simulator (HVS) tests
described by Visser et al (1983) on two light pavement sections in the Transvaal showed
that, in terms of onset of cracking, the life of a newly resealed section was two to three
times greater than that of a section with an aged surfacing. Under wet conditions, the
weakened state of pavement increased the rate of deterioration to such an extent that it
was suggested that even light vehicles contributed to the distress. The effect of moisture
is particularly severe in the outer wheel paths, where moisture is usually free to enter the
pavement from the unsealed shoulders.

3.3 Analysis of Database

The intention of this part of the study was to identify surface distress and to confirm the
detrimental effect of high deflections and marginal base materials on the surfacing. A
database of results pertaining to light pavement structures, obtained during the study by
Paige-Green (1994), was consulted in order to evaluate the performance of various seal
types constructed on these structures. The study by Paige-Green (1994) focussed mainly
on the use of marginal base course materials, but an evaluation of the surfacing condition
was performed at each of the sites. Because of the limited information on the seal, only the
trends with respect to seal performance could be obtained. No attempt was made to model
the data.

The performance of the seals is based on an index calculated from the observed condition
with regard to cracking, texture, bleeding, binder condition and riding quality (QI). The visual
ratings were converted to a performance index, after weighting of the individual visual
ratings according to significance.

A variety of seal types was evaluated. Generally a sand seal was the most frequently used
initial surfacing type. Single seals were the most frequently used reseal type.

3.3.1 Base Strength

The average penetration rate of the DCP through the base layer was found to be the most
convenient measure of the base strength. Obviously, a higher rate of penetration is
indicative of a weaker base (either as a result of poor quality materials, poor compaction
or moisture damage), which will experience greater deflections. The performance indices
for various seals are plotted against the DCP penetration rate in Figure 3.2. The diagram
indicates that the performance of all seal types decreases with a decrease in base

Muthen KM et al -7- CAPSA’99

 Design recommendations for highly flexible surfacings

strength. It can also be seen that double seals are the most resilient on a weak base while
sand seals are the most sensitive. This is probably related to the thickness of the seal with
thicker seals having increased resistance to damage. Sand seals seem to perform well on
bases with a DCP penetration rate of less than about 4.5 mm/blow while single and double
seals were observed to perform well on bases with a DCP penetration rate of less than 7
mm/blow.

7.5

6.5
Performance Index

5.5

4.5

4
0 2 4 6 8 10 12
DCP Penetration Rate (mm/blow)

single sand double

Figure 3.2: Seal Performance on bases of varying strength.

The life expectancy of seals was found to be longer on stronger bases. Figure 3.3 shows
the observed lives (before being resealed) of sand seals on bases of various strengths.
This chart clearly indicates that the life of sand seals is reduced dramatically when they are
placed on highly flexible bases.

An assessment of the surface texture depth was carried out by calculating a texture depth
index based on the visual assessment results. This gave an indication of which surfaces
were prone to bleeding or fatting up. The texture depth index was plotted against the
corresponding base DCP penetration rate value to obtain Figure 3.4. These results
indicate that there is an increased tendency for the surfacing to bleed when the base is
weak (DCP > 5 mm/blow). This may be attributed to the punching of the seal aggregates
into the soft base.

Muthen KM et al -8- CAPSA’99

 Design recommendations for highly flexible surfacings

12

11
10

8
Life (years)

0
0 2 4 6 8 10 12
DCP Penetration Rate (mm/blow)

Figure 3.3 : Observed lives of sand seals.

10

8
Texture Depth Index

0
0 2 4 6 8 10 12
DCP Penetration Rate (mm/blow)

single sand double

Figure 3.4: Influence of soft bases on seal texture depth.

These results confirm the detrimental effect of weak bases on the life and performance of
a traditional surfacing, which is most likely due to abnormal deflections.

3.3.2 Climate

Muthen KM et al -9- CAPSA’99

 Design recommendations for highly flexible surfacings

Generally seals evaluated in this report showed decreased performance in regions with
high rainfall (Figure 3.5). Analysis of the data suggests the following reasons for this:
C bases in wet climates are weaker than those in the dry climatic regions, resulting in
high deflections, especially when wet;
C the quality of the base materials in wet areas, as measured by the fineness product
(PI x % passing 0,425mm sieve), proved to be poorer than that of base materials
in drier areas. The performance of seals shows remarkable dependence on the
quality of the base materials (Figure 3.6), and
C the seasonal wetting and drying of the base materials results in a higher rate of
degradation in wet regions than in dry regions.

The implications of these results are that the performance of marginal materials is far more
critical in wet areas or where moisture conditions vary greatly between seasons.

6.5

6
Performance Index

5.5

4.5

4
500 700 900
Annual rainfall (mm)

Figure 3.5: Performance of single seals in various rainfall regions.

The model for selecting marginal materials, developed in Australia by McLennan (1986),
is shown in Figure 3.7. A suitable pavement material is selected by considering the traffic
and environment (rainfall). In terms of this model a suitable material should give adequate
long term performance if the CBR is greater than 60, deflections are less than 1.5 mm for
low traffic volumes and in-situ moisture conditions do not vary significantly.

Muthen KM et al -10- CAPSA’99

 Design recommendations for highly flexible surfacings

7.5

6.5
Performance Index

5.5

4.5

4
0 200 400 600 800 1000 1200
PI x % Passing 0,425mm sieve

single sand double

Figure 3.6: Seal performance related to material quality of the base

(fineness product).

In light of the distress mechanisms identified above, it can be appreciated that the selection
of an appropriate binder is critical in order to ensure the good performance of the
surfacing. For highly flexible pavements, the binder should ideally possess the following
properties:

Figure 3.7: Australian Model for the selection of marginal materials (after
McLennan, 1986)

Muthen KM et al -11- CAPSA’99

 Design recommendations for highly flexible surfacings

C low stiffness at low temperatures so as to avoid brittle failure;

C high stiffness at high temperatures to avoid plastic deformation;
C ability to withstand and recover from high deformations without cracking; and
C high resistance to fatigue and ageing.
The above requirements are ideally provided by modified bituminous binders.

3.4 Comparison of Binders

The flexibility of binders changes according to their composition and operating temperature
and ageing characteristics. Experience in South Africa (Bergh, 1997) indicates that:
C Penetration grade bitumens remain flexible for between 2 to 3 years. They become
brittle at temperatures below 5°C;
C Modified binders retain flexibility for between 5 to 7 years depending on the type of
elastomer (SBS modified binders are likely to offer similar performance to bitumen-
rubber, but insufficient South African field performance data is available to
substantiate this); and
C Bitumen rubber remains flexible for up to 15 years and at temperatures down to -
5°C.

The flexibility, durability and waterproofing ability of the seal is highly dependent on the
amount of binder present. Assuming that penetration grade bitumens have a rate of
application equal to 100 per cent, then the binder application rate can be increased by
about 120 per cent for polymer-modified binders (slightly more for SBS modified binders),
and by about 160-180 per cent for bitumen rubber. This is largely because of the increased
viscosity of the modified binders, resulting in reduced plastic strains and bleeding at high
temperatures and high traffic. It has also been observed that the aggregate orientation in
modified seals allows for more binder to be applied (Semmelink, 1997). Polymer-modified
binders are, however, highly susceptible to flushing at high application rates, due to the
punching of aggregates into a soft base. Bitumen rubber has been found to be the least
susceptible of all with respect to susceptibility to high temperature and high traffic stresses.

The structural strength of the seal is required to prevent cracking and to retain the
aggregates. There is a lower limit to the amount of binder which is needed to structurally
hold the aggregate under heavy traffic. Modified binders have improved adhesive
properties and thus can withstand higher traffic stresses without loss of stone. Furthermore,
the structural strength of modified binders is more rapidly achieved and maintained for
longer periods. The elastic recovery of modified binders is enhanced, thus allowing for
recovery from very high deformations.

Bitumen seals, when placed on unstabilized material and when mechanically damaged,
results in rapid deterioration of the base into potholes. Modified binders are more tolerant
of marginal quality materials than unmodified binders. The following material limits
(AASHTO classification) are recommended for warm binders:
C penetration grade bitumens can be effective on A-1-a(o) to A-2-4(o) materials
(approximately equivalent to G5 materials and above in the South African materials
code), provided that the base CBR is greater than 50;
C polymer-modified binders have the same requirements as above but are not

Muthen KM et al -12- CAPSA’99

 Design recommendations for highly flexible surfacings

recommended for materials below A-2-4(o) as punching of the aggregate will be a

problem, and
C bitumen rubber can be used on A-1-a(o) up to A-2-6(o) materials (approximately
equivalent to G7 materials and above in the South African materials code) but could
also be used on lower quality materials if the seal and the base are properly
designed.
Generally, because of the relatively high volumes applied, bitumen rubber binders result in
the most effective seals for the prevention of water ingress and, on account of their ability
to maintain their flexible properties, they can accommodate some settlement of the
pavement system and overcome some possible detrimental effects of passive and active
cracks resulting from overstressed bases. The superior fatigue characteristics of bitumen
rubber, especially in the retardation of reflective cracking, has been proven both in
laboratory simulation tests as well as field trials (Coetser, 1994 and Rust et. al., 1992).
Intuitively bitumen rubber seals have the characteristics required of a superior highly flexible
surfacing.

4. DESIGN IMPLICATIONS / RECOMMENDATIONS

The general aspects of seal design contained in Draft TRH3 (1997), as well as the
guidelines provided in SABITA Manual 10 (1992) provide a sound basis for the design of
surface seals. However, for highly flexible light pavement structures, there are additional
implications on the surfacing.

When existing structures are upgraded, there are certain advantages and disadvantages
which should be addressed. The advantages of upgrading existing roads to paved
standards are:
C existing structures/roads/streets have been subjected to traffic compaction and the
lower layers in the structure have probably settled to a fair state of "equilibrium";
C there may be layers of imported material which are probably up to subbase
standard;
C where no gravel has been imported, the existing subgrade material is probably
sound enough for use as a subbase or base layer;
C the stormwater system is probably in place and only minor improvements may be
required, and
C for design purposes, the DCP and a minimum of materials testing would be
adequate for assessment.

Possible disadvantage of upgrading existing roads are :

C lack of maintenance and erosion could result in major earthworks and/or importation
of gravel to re-establish the shape of the road;
C drainage structures are probably silted up or damaged;
C the improved service conditions will attract increased traffic which will accelerate the
degradation of the road. Limited enforcement of overloading regulations in rural
areas may aggravate this problem;
C sealed pavements require regular specialized maintenance programmes to ensure
adequate performance. Rural road authorities may not have the budget, manpower

Muthen KM et al -13- CAPSA’99

 Design recommendations for highly flexible surfacings

skill or equipment to carry out the required maintenance, and

C the geometry and pavement standards may be poor and may not live up to the
travelling public’s expectations of a surfaced road (Visser et al, 1983).

The use of surface seals on light structures can be justified on the following grounds:
C seals are more flexible than hotmix asphalt and they can accommodate
approximately 2 to 3 times the deflection of an asphalt surfacing (Bergh, 1997);
C seals are cheaper to construct than asphalt surfacings because of the reduced
aggregate and binder costs, and
C well-constructed seals are impermeable, and thus allow for the use of marginal
local materials in place of the expensive option of importing materials.

Surface dressings may be the least expensive form of surfacing but could lead to high
maintenance costs if not applied correctly (construction quality). Regular maintenance of
a seal has a major effect on improving the performance and life of a seal (SABITA Manual
10, 1992). Hence, the maintenance environment must be considered in the selection of an
appropriate surfacing. It is important to ensure that the road authority is capable, both
technically and financially, of carrying out the required maintenance for the particular
surfacing option. This is a possible advantage in the use of highly flexible surfacings as it
affords the opportunity for the training of specialised maintenance teams (sustained job
creation) in these areas.

Since it is virtually impossible to prevent some ingress of water into the pavement, any
sealed pavement should be complemented with a good drainage system to prevent water
from collecting in the pavement layers. In previous investigations it has been observed that
the provision of good drainage always resulted in improvement in the surface life (Emery
et al, 1991).

Although it has been agreed that the South African climate is moderate, certain inland
regions experience overnight temperatures well below freezing point during cold seasons.
The brittle surfacing coupled with the possibility of frost heave or a single heavy vehicle will
result in cracking (Bergh, 1997). It is thus important to select a binder according to the
minimum expected temperatures.

In order to enhance flexibility and durability, it is recommended that the binder/bitumen

application rates be maximized. This may result in bleeding but this is not critical for lightly
trafficked roads. Modified binders are suitable because of the increased application rates
and reduced susceptibility to bleeding. The use of larger aggregates as well as of graded
aggregates which can accommodate high binder contents is also recommended.
Aggregate precoating is also recommended in order to ensure proper adhesion of the
aggregates in the seal.

The selection of surfacing should be based on the pavement deflection measured by

deflection tests (FWD or Benkelman beam), traffic volume, and climate. Generally, a
greater deflection under a standard load requires a thicker seal with more binder. Table
4.1, which is an adaptation of Medelsky and Lovecek (1987) , based on experience with

Muthen KM et al -14- CAPSA’99

 Design recommendations for highly flexible surfacings

forestry roads, is a rough guide to surfacing thickness required for different traffic and
deflection conditions.

Table 4.1: Surface recommendation based on traffic and deflection for a service
life of 10 years

Benkelman Beam Deflection (mm)(Dry

E80's/day Season)
0.50-1.80 1.81-2.10 2.11-2.40 2.41-2.70
7 SS SS S1 S2
14 SS S1 S2 *
24 SS S1 S2 *
adapted from Medelsky and Lovecek(1987), based on experience with forestry roads.
SS=sand seal; S1=6.7 single seal; S2=13.2/6.7 double seal; * base strengthening required.

Although the use of marginal materials for the pavement layers is advocated, it is stressed
that the standard specifications for seal aggregates should be adhered to strictly. The size,
shape, texture and grading of the aggregate influences the amount of binder that is
required for the seal. The thermal and solar radiation characteristics and the
chemical/mineral structure of aggregates affects the rate of binder hardening by possibly
catalysing or retarding the oxidation process. Excessive fines/dust in the cover aggregate
will be absorbed by the binder and thus reduce flexibility. Bergh (1997) is of the opinion that
the role of aggregate absorption on binder hardening is underestimated and that it would
be wise to consider this in the seal design.

The upper portion of the base is especially significant to the performance of the seal. This
part of the base should preferably be strengthened so as to resist horizontal shear
stresses, punching in of aggregates and to improve the resistance of the material to
damage by water and traffic. Adequate penetration of the prime (low viscosity prime is
recommended) is required to enhance the adhesion of the surfacing to the base, reduce
oxidation from the base and to reduce absorption of binder from the seal. In order to fulfil
the roles of priming, stabilization, and strengthening, treatment with diluted emulsion is
recommended( i.e. ETB) as follows:
C Class (II) roads: 40 mm of ETB
C Class (III) roads: 100 mm of ETB
This may be achieved by ripping the top 40 mm or 100 mm and treating it with diluted
emulsion before final compaction. This procedure can only be effectively used on materials
with Pl’s of less than approximately 8. Materials with PI’s greater than 8 can be modified
by stabilizing the top 50 mm or 100 mm of base with lime. A less expensive alternative to
stabilisation is armouring of the base with aggregate chips rolled into the surface of the
base layer.

In light of the previous successful use of bitumen rubber seals the following low risk
surfacing for highly flexible pavements is recommended:
C single seal using bitumen rubber with an application rate of about 2.5 litre/m2; and
C 13.2 mm aggregate.

Muthen KM et al -15- CAPSA’99

 Design recommendations for highly flexible surfacings

This seal can be applied on any material ( A-2-6(o) or A-6(o) i.e. approximately equivalent
to G8 materials and above in the South African materials code) which can be compacted
to a smooth finish and shape.

4.1 Special Considerations for Sand Seals for Class III Roads
Sand seals, when placed on ETB, provide a very economical surfacing solution for lightly
trafficked roads provided that they are correctly designed and constructed. The life of a
sand seal is highly dependent on the smoothness of the final surface obtained during
completion of the compaction of the base. It has also been observed that sand seals are
more durable when applied in two layers. The quality of finish of the base is more readily
improved by special treatment with diluted emulsion for the final compaction of the base
(as explained above). This approach will eliminate the need for application of the prime and
one of the traditional double application of sand seals. Cationic emulsions are preferable
for the binder in sand seals. The sand should be screened to eliminate all material larger
than 6.75 mm.

5. ECONOMIC ANALYSIS OF HIGHLY FLEXIBLE SURFACING PRODUCTS

The analysis in this section is based on the principles of discounted life cycle cost.
Essentially this type of analysis calculates the current value (nett present value) of the cost
involved in the construction, maintenance and operation of a pavement surfacing over a
certain analysis period (typically 10 years for surface treatments). The analysis assumes
that none of the surfacings has any value at the end of its service life. The life remaining at
the end of the analysis period is accounted for by assigning a salvage value. Only surfacing
products identified during the industry survey were evaluated. The costs determined in the
economic analysis were compared with that of the no-change option of a gravel road. This
analysis included both economic and financial costs.
C Economic analysis: includes indirect costs, such as accident costs, travel time and
vehicle operating costs.
C Financial analysis: includes only the direct costs to the road authority, such as
construction and maintenance costs.

5.1 Concepts and Assumptions associated with the Economic Analysis

Discount Rate
This is the real rate of return that can be expected for an investment (ie the nominal interest
rate minus the inflation rate). The discount rate was assumed to be 10 per cent in this
analysis. The discount rate is used to adjust all future monetary values to the present
year(1997) equivalent.

Maintenance Costs
The routine maintenance costs (surface cleaning and patching) and the cyclic maintenance
(resealing) cost are taken to be the same for all alternatives. The routine maintenance
costs as well as the road user costs were obtained from the SABITA SURF+
software(1993), based on inflated values from the software’s internal database. The reseal
maintenance cost assumes that all alternatives are resealed with a single seal, the cost of
which was obtained from Draft TRH4 (1996) and inflated by 10 per cent. The SURF+

Muthen KM et al -16- CAPSA’99

 Design recommendations for highly flexible surfacings

software does not take into account the lower maintenance requirements of modified
binders. Also, the maintenance costs of gravel roads seem to be slightly high, but not
unreasonably so. (Bergh, 1997).

Surfacing Life
The maintenance free life of the surfacings was taken to be the average expected life given
in Draft TRH3 (1997), as well as in SABITA Manual 10 (1992). The range of life cycle costs
based on the expected range of service life was calculated.

Construction Cost
The construction cost of the surfacing was calculated using a bill of quantities approach,
using assumptions and data from Emery et al (1991). Since the cost can vary significantly
depending on the location and size of job, as well as on the condition of the existing
pavement, average assumptions were made in this regard. The costs of the binders were
obtained from the industry survey.

Analysis Period and Salvage Value

The analysis period was selected as 10 years in keeping with Emery et al (1991).The value
of the remaining life of the surfacing was taken into account in the analysis as a salvage
value, using the following equation:

Life remaining
Salvage value ' Construction cost x
Surfacing life

Accidents Costs, Road User Costs and Vehicle operating costs.

The cost of these items are not incurred by the road authority but by the road user and,
ultimately, by the economy of a region. These costs usually form the basis of socio-political
decision for surfacing roads. The SABITA SURF+(1993) software was used to estimate
these cost items.

5.2 Results of the Economic Analysis

(Note that all costs exclude taxes.)

Table 5.1 shows the expected financial life cycle costs for the various surfacing options.
These values are based on the expected life of the particular surfacing. It should be
remembered, however, that surfacing life varies according to operating conditions. Figure
5.1 shows the likely range of life cycle costs based on the range of surfacing lives.

Muthen KM et al -17- CAPSA’99

 Design recommendations for highly flexible surfacings

Table 5.1: Expected Financial life cycle costs in 1997.

Seal Type Aggregate Binder Expected Life cycle
(m3/m2) (l/m²) Life cost (R/m²)
13.2/6.7 Double Seal using Vialastic 65/5 0.0158 2.70 9 21.69
13.2/6.7 Double Seal using Vialastic 70/5 0.0158 2.50 9 21.47
13.2/grit Double Seal using Vialastic 65/5 0.0145 2.20 8 21.20
13.2/grit Double Seal using Vialastic 70/5 0.0145 2.00 8 20.96
13.2 Single Seal using Viagrip 3 (RMB3) 0.0095 1.50 6 20.53
13.2 Single Seal using Elastomod 3 0.0095 1.60 6 21.10
Graded Seal using Rubberized Gravel 0.0100 1.80 8 18.95
Binder
Sand Seal on 40 ETB using Vialastic 65/5 0.0060 1.60 7 20.82
Sand Seal on 40 ETB using Viagrip 3 0.0060 1.20 7 20.27
13.2 Single Seal using Bitumen Rubber 0.0095 2.5 15 18.62

As can be seen from Table 5.2, the financial and economic life cycle costs of a gravel road
as well as the economic life cycle costs of the surfacings increase with traffic volumes. The
Nett Present Value of a surfacing (i.e. the benefit of the surfacing compared with a gravel
road) can be calculated by subtracting the life cycle cost of the surfacing from the life cycle
cost of the gravel road (Table 5.2).

Table 5.2: Life cycle costs in 1997 of an existing gravel road over 10 years.
(using SABITA SURF+ Software,1993).
Traffic Financial Cost Economic Cost
(ADT in evu) (R/m²) (R/m²)
50 15.00 59.40
150 45.01 145.85
300 90.02 275.53
500 150.03 448.43
750 225.04 664.55

In the case of traffic volumes of 50 evu/day, it was found that, financially, all surfacing
options are more costly than the gravel. On further analysis (Figure 5.2) it was found that
the financial break-even traffic (cost of surfacing equals the cost of the maintaining the
gravel road i.e. NPV = 0) was approximately 75 evu/day, while the economic break-even
traffic was slightly less than 50 evu/day. This will obviously vary, depending on the
construction cost of the surfacings, but did not vary significantly for the surfacing options
considered in this study.

Muthen KM et al -18- CAPSA’99

 Design recommendations for highly flexible surfacings

26

Lower bound for seal life (years)

24 3
6

Sand Seal on 40 ETB using

6 3
5
5

Sand Seal on 40 ETB using

Lifecycle Cost (R/m²)

Vialastic 65/5
13.2/6.7 Double Seal using

4
13.2/6.7 Double Seal using

13.2/grit Double Seal using

13.2/grit Double Seal using

4
Vialastic 65/5

Viagrip 3
22
Vialastic70/5

Vialastic 65/5

13.2 Single Seal using Elastomod3

Vialastic 70/5 7

13.2 Single Seal using Bitumen

13.2 Single Seal using Viagrip 3
6 7 10

20

Graded Seal using Rubberised

Rubber
14
14 12
12

Gravel Binder
15

18
Upper bound for seal life (years) 14
14
14

16
Figure 5.1: Range of life cycle costs for various surfacing options.

A significant outcome of the economic analysis is the competitiveness of the costs of sand
seals on emulsion treated bases (ETB) relative to those of the other options (see Table
5.1), largely due to the high life expectancy of the sand seal on this structure. The bitumen
rubber single seal is also economically attractive as it has the lowest life cycle cost and
least variability in its range of possible life cycle costs, as shown in Figure 5.1.

Muthen KM et al -19- CAPSA’99

 Design recommendations for highly flexible surfacings

250

200
Nett Present Value

150

100

50

-50
0 100 200 300 400 500 600 700 800
Traffic (evu/day)

Financial Nett Present Value Economic Nett Present Value

Figure 5.2: Variation of financial and economic NPV with traffic volume (for 13.2/6.7
double seal using Vialastic 65/5)

The selection of an appropriate surfacing should be based on the lowest life cycle cost and
not on the lowest construction cost. Ideally, the road authority should have sufficient funds
available at the time of construction to cover the financial life cycle cost of the selected
surfacing.

6. CONCLUSION

In South Africa, the successful use of surface seals has been widespread, even on relatively
heavily trafficked roads. This has fuelled the demand for the use of seals as a cost-effective
upgrading option for unpaved roads, an application which requires a highly flexible
surfacing. The use of light pavement technology and highly flexible surfacings offers cost-
effective solutions to the challenge of providing economical pavement infrastructure in
developing regions in South Africa. Although all thin surface dressings (seals) may be
regarded as flexible by comparison with asphalt, highly flexible surfacings refer to those
seals which offer performance characteristics superior to those of traditional surfacings
under conditions of abnormal pavement deflections. Surfacings on highly flexible light
pavements can be expected to have reduced service lives and high risks of failure because
of the aggravating service conditions of high deflections and possible marginal base
quality. Stringent construction quality control of the surfacing is thus required in order to
avoid premature failure of the entire pavement.

The design of highly flexible surfacings using the conventional seal design would be
inadequate, since the functions of highly flexible surfacings differ from those of conventional
seals. While the main functions of conventional seals is to provide a skid-resistant surfacing

Muthen KM et al -20- CAPSA’99

 Design recommendations for highly flexible surfacings

and prevent the ingress of water, the waterproofing role of highly flexible surfacings is
usually critical in preventing pavement failure, while skid-resistance may be of secondary
consequence on low volume roads. Conventional methods of seal design concentrate on
optimising the skid resistance of the surfacing. In contrast, the prime objective of a design
procedure for highly flexible surfacings would be to optimise the flexibility and fatigue
resistance of the surfacing. Furthermore, conventional seal design methods do not
adequately take the enhanced properties of modified binders and the characteristics of the
base into account. A design procedure for highly flexible surfacings needs to combine the
existing seal design methods with procedures for taking the specific functional
requirements and deterioration mechanisms of highly flexible surfacings into account. It was
found that the following variables had to be considered in the selection of a highly flexible
surfacing:
C the type and quality of the base;
C expected deflections;
C climatic conditions, and
C the engineering properties of the binder.

Essentially, a design procedure for highly flexible surfacings should aim to:
C optimise flexibility and fatigue resistance;
C limit brittle fracture;
C limit the binder degradation with age, and
C reduce the temperature susceptibility of the surfacing.

Climatic conditions such as extremes of temperature and heavy rainfall play a dominant
role in the performance of light pavement structures surfaced with thin surface dressings.
The performance and lives of these pavements can be enhanced and prolonged by the use
of modified bituminous seals and timeous surfacing maintenance programs. Any means
of delaying the degradation of the binder in highly flexible surfacings, such as:
C regular application of fog-sprays (emulsified binder or solvents);
C increased binder film thicknesses;
C the use of non-absorptive aggregates, and
C the use of high penetration prime treatments,
will have a positive influence on the performance of the surfacing.

It was evident from the results of the industry survey that the binder requirements of highly
flexible surfacings had been identified by practitioners and that their needs had been
fulfilled to some extent by the development of specialised modified bituminous binders,
which include:
C SBR, SBS and EVA polymer-modified bituminous binders, and
C Bitumen rubber modified bituminous binders.

However the lack of information on the field performance of the identified products, except
for bitumen rubber, meant that the manufacturers’ claims on the performance of the
specific products could not be confirmed. Generally, it can be accepted that polymer
modified binders offer longer seal lives than un-modified bitumen binders, as has been
proven in certain field applications.

Muthen KM et al -21- CAPSA’99

 Design recommendations for highly flexible surfacings

Bitumen rubber seals have proved to be ideal for application as highly flexible surfacings,
both functionally and economically. The average life cycle cost of a bitumen rubber single
seal was calculated to be approximately R19.50 per square metre. Certain types of
polymer-modified binders were found to have economic advantages over bitumen rubber,
such as semi-priming cutback binder used in a graded seal (average life cycle cost of
R18.40 per square metre), but lacked the field performance track record of bitumen rubber.
The economic analysis tended to be conservative (i.e. higher costs) because of the lack
of information on the performance and maintenance requirements of modified seals. Highly
flexible surfacings were found to be economically viable for application on gravel roads with
traffic volumes as low as 50 equivalent vehicle units per day, by comparison with the cost
of gravel road maintenance.

The use of ETB technology combined with highly flexible surfacings also seems to be cost-
effective and possibly to have the additional socio-economic advantages of providing
labour-intensive employment opportunities during the construction phase and sustained job
creation resulting from the specialised maintenance requirements of highly flexible
surfacings.

7. ACKNOWLEDGEMENTS

This paper is based on research conducted at CSIR Transportek which was funded by the
South African Bitumen Association (SABITA). The author is grateful to both organisations
for permission to publish this paper.

8. REFERENCES

Bergh, A. O., Personal Discussion , Consultant, CSIR Transportek, Pretoria, 1997.

Coetser, K., Guidelines for the selection of materials for use in surface treatments
to retard reflective cracking, CSIR Transportek, Pretoria, 1994.

Emery, S., Van Huyssteen, S., Van Zyl, G., Appropriate standards for effective
bituminous surfacings: Final report, CSIR Transportek, Pretoria, 1991.

Dickinson, E. J., The performance of thin bituminous pavement surfacings in

Australia, 11th Australian Road Research Board Conference, Melbourne, 1982.

Dorman, G. M. and Lamb, D. R., The properties of bitumen in relation to the

performance of road surfacings, 6th Australian Road Research Board Conference,
Canberra, 1972.

Muthen KM et al -22- CAPSA’99

 Design recommendations for highly flexible surfacings

Guimmarra, G. J. (editor), Sealed local roads manual: guidelines to good practice for
the construction, maintenance and rehabilitation of pavements, Australian Road
Research Board, Victoria, 1995.

Jooste, F. J., Evaluation of the HDM performance prediction model for use in the
Gauteng Province, CSIR Transportek, Pretoria, 1997.

Mclennan, A. K., Towards a strategy for the use of marginal and naturally occurring
materials in pavements, 24th Australian Road Research Board Regional Symposium,
Bundaberg, Queensland, 1986.

Medelsky, V and Lovecek, Z, Question VI - Roads in developing regions,

Czechoslovakia, XVIIIth World Road Congress, Brussels, 1987.

Oliver, J. H. W., An interim model for predicting bitumen hardening in Australian

sprayed seals, 12th Australian Road Research Board Conference, Hobart, Tasmania,
1984.

Paterson, W. D. O., Road deterioration and maintenance effects, models for

planning and management, John Hopkins University Press for The World Bank,
Baltimore, 1987.

Paige-Green, P., Recommendations on the use of marginal base course materials

in low volume roads in South Africa, CSIR Transportek, Pretoria, 1994.

Paige-Green, P., Personal discussion , Senior Researcher, CSIR Transportek, Pretoria,

1997.

Rust, F. C., Coetser, K. and Verhaeghe, B. M. J. A., The evaluation of six modified
binders for retardation of crack reflection through laboratory studies and field
work, 7th International Conference on Asphalt Pavements, Nottingham, 1992.

SABITA Manual 7 (SURF+), Economic warrants for surfacing roads, South African
Bitumen and Tar Association, Cape Town, 1993.

SABITA Manual 10,Appropriate standards for bituminous surfacings for low volume
roads, South African Bitumen and Tar Association , Cape Town, 1992.

Muthen KM et al -23- CAPSA’99

 Design recommendations for highly flexible surfacings

Semmelink, C. J., Personal discussion , Senior Researcher, CSIR Transportek, Pretoria,

1997.

Technical Recommendation for Highways 3 (Draft TRH3), Surface seals for rural and
urban roads, CSRA, Pretoria, 1997.

Technical Recommendation for Highways 4 (Draft TRH4), Structural design of

interurban and rural road pavements, CSRA, Pretoria, 1996.

Technical Recommendation for Highways 14 (TRH14),Guidelines for road construction

materials, CSRA, Pretoria, 1987.

Visser, A. T., Maree, J. H., Marais, G. P., Implications of light bituminous surface
treatments on gravel roads, 3rd International Conference on Low Volume Roads,
Transportation Research Board, Washington D.C., 1983.

Visser, A. T., van Niekerk, T., The implementation of appropriate technology in the
design of light pavement structures, 4th International Conference on Low Volume
Roads, Transportation Research Board, Washington D.C., 1987.

Muthen KM et al -24- CAPSA’99