bitume rubber asphalt.pdf

Proceedings of the 8th Conference on Asphalt Pavements for Southern Africa (CAPSA'04) 12 – 16 September 2004 ISBN Number: 1-920-01718-6 Sun City, South Africa Proceedings produced by: Document Transformation Technologies cc

BITUMEN RUBBER ASPHALT: YEAR 2003 DESIGN AND CONSTRUCTION PROCEDURES IN SOUTH AFRICA

C.J. Potgieter

Potgieter Hattingh and Raspi Inc. PO Box 23083, Randburg Waterfront, 2167 South Africa. E-mail: [email protected]

ABSTRACT

This paper describes the latest design procedures and construction practices used for the construction of bitumen rubber asphalt in rehabilitation projects in South Africa. Bitumen rubber products are normally at a disadvantage as they are placed where the conventional asphalts have already failed or where the existing base courses are in severe distress, and on top of this must now provide additional structural capacity for the next 10 to 15 years under severe traffic conditions.

This paper describes the following aspects:

• Basic constituent materials used in the manufacture of B-R asphalt. The South African materials specification is furnished and discussed.

• The SA asphalt mix design process including the type selection; the volumetric design process as well as the B-R asphalt end point specification.

• Further developments (rutting, fatigue, energy methods, moisture sensitivity, Dynamic Creep).

• B-R structural equivalency.

• Construction of B-R asphalt.

• Potential problems.

• Summary and conclusions.

This paper conclude that the somewhat higher initial construction cost of B-R asphalt are negligible in view of the savings in traffic accommodation, bypasses and the extended life rendered by the B-R asphalt when assessed over the full life cycle. In South Africa B-R asphalt is considered the Rolls Royce of asphalts for rehabilitation purposes.

1. INTRODUCTION

The outstanding success of Bitumen Rubber (B-R) asphalt over the past 20 years in South Africa has revived interest in this product recently. This success became prevalent in 1998 with the report on the 12-year service life of bitumen rubber asphalt on the Buccleuch Interchange [Potgieter et al, 1998]. This asphalt eventually provided a 16-year service under extreme traffic conditions before being overlayed.

This paper focus on the practical experiences gained in recent years during the design and construction of the new generation B-R asphalt with special reference to the behavior of the B-R binder over time. This paper concludes that B-R asphalt is the superior product for heavy trafficked asphalt roads when calculated over its full life cycle.

mailto:[email protected]

8th CONFERENCE ON ASPHALT PAVEMENTS FOR SOUTHERN AFRICA

Paper 015

2. BASIC CONSTITUENT MATERIALS

As with any asphalt the long-term performance of the B-R asphalt is dependent on the quality of the materials, the binder content and the grading. Attention to detail for each of these components is critical.

2.1 Bitumen Rubber Binder

The B-R asphalt is manufactured using non-homogeneous bitumen rubber binder. The binder is manufactured from blending penetration grade bitumen (60/70 or 80/100), plus extender oil (2%) plus rubber crumbs (18 – 24%) in a patented high-speed blender (3 000 r.p.m.). The more reliable manufacturers in South Africa nowadays prefer to standardize on a 20% rubber content and also to pre-select the type of tyres for the buffing process. The bitumen rubber blend is then circulated in a holding tank and heated to temperatures (190 – 210oC) to facilitate the chemical digestion process.

The bituminous binder used in the production of the bitumen-rubber must be a B12 or B8 road-grade bitumen (60/70 or 80/100 penetration-grade bitumen respectively) or a blend of these grades to provide a product with a particular viscosity and other prescribed properties.

The rubber is obtained by processing and recycling pneumatic tyres. The rubber crumbs is produced by a mechanical comminuting process. Crumbs produced by cryogenic-mechanical techniques are not permitted in South Africa. It must be pulverized, free of fabric, steel cords and other contaminants. A maximum of 4% by mass of fine particle size calcium carbonate, or talc, may be added to the rubber crumbs to prevent the rubber particles from sticking together. The crumbs must be free flowing, dry and comply with the requirements in table 1.

If used, extender oil must be a resinous, high flash point aromatic hydrocarbon conforming with the requirements in table 2.

The chemical composition of the base bitumen and the nature of the rubber dictate the need to introduce an extender oil. The bitumen and extender oil, when combined must form a material that is chemically compatible with the rubber.

Table 1. Rubber crumbs [COLTO, 1988].

Property Requirements Test method Sieve analysis, as a percentage by mass passing through the

1,18 mm sieve, % 0,60 mm sieve, % 0,075 mm sieve, %

100 50 - 70

0 - 5

Sabita [1988] BR6T

Natural rubber hydro-carbon content (minimum), % 30

BS903 Parts [British Standards]

B11 & B12

Resilience (raw) (minimum), % 40 Sabita [1988] BR7T

Loss in resilience (maximum), % 60 Sabita [1988] BR7T

Relative density, g/cm3 1,10 - 1,25 Sabita [1988] BR9T

Table 2. Extender oil [COLTO, 1988]. Property Requirements Test method Viscosity at 40oC (minimum) 2 500 ASTMD 88 [ASTM] Flash point, COC, oC 200 ASTMD 92 [ASTM] Compositional analysis Asphaltness, % by mass (maximum) Aromatics, % by mass (minimum)

0,1

55,0

ASTMD 2007 [ASTM]


Paper 015

At least two weeks prior to commencement of the blending operation, the contractor must state in writing a method statement that includes details of the proposed plant, the percentage of rubber, the blending / reaction temperature, reaction time and percentage of extender oil he intends to use, substantiated by graphs of time against ring and ball softening point and flow for the full reaction period, taking cognisance of the actual equipment and method to be used for the mass production of bitumen-binder. These are referred to as behaviour curves. The contractor are not allowed to deviate from the stated percentage of rubber crumbs by more than 1 per cent and from the stated reaction temperature by more than 10oC without producing new behaviour curves. Reaction time is defined as half the time required for adding the rubber to the bitumen plus the time required for acceptable reaction between the bitumen and rubber as determined from the behaviour curves in agreement with the engineer. No bitumen-rubber blend may be used until it has complied with the specified criteria.

The bitumen-rubber blend, containing where necessary extender oils, must satisfy the requirements in table 3. The reconstitution of the bitumen-rubber blend is discouraged in South Africa.

Table 3. B-R blend [COLTO, 1988].

Property Requirements Percentage of rubber by mass of the total blend, % 18 - 24 Percentage extender oil by mass of the total blend (maximum), % 4 Blending / reaction temperature, oC 180 - 210 Reaction time, hours 1 - 4

A continuous record is kept on the site by the contractor of the percentage rubber in the mix and the reaction temperatures and times. The bitumen-rubber binder must be sampled within five minutes prior to the mixing of the asphalt and must comply with the requirements in table 4.

Table 4. Modified B-R binder [COLTO, 1988].

Property Requirements Test method Compression recovery, % after 5 minutes after 1 hour after 4 days

80 -100 70 - 95 25 - 55

Sabita [1988]

BR3T

Ring-and-ball softening point (minimum and range), oC 55 - 62 ASTM D36 [ASTM]

Resilience, % 15 - 35 Sabita [1988]

BR2T

Flow, mm 15 - 55 Sabita [1988]

BR4T

Viscosity at 190oC, dPas 20 - 50 Sabita [1988]

BR5T The bitumen-rubber mixture after reaching the desired consistency must not be held at temperatures in excess of 160oC (with no extender oil) or 190oC (with 4% extender oil) for more than 4 hours.

The patented blender is shown diagrammatically in figure 1.


Paper 015

Figure 1. Blending of bitumen rubber.

2.2 Aggregate and Fillers [Colto, 1988]

With the increase in traffic and loading on our provincial and national roads the quality of the aggregate has become of great importance in the manufacturing of a high quality asphalt.

2.2.1 Aggregate [Colto, 1988] • The aggregate crushing value (ACV) of the coarse aggregate, must not exceed 25. The

minimum dry 10% FACT values of the –13,2mm +9,5mm fraction must be at least 210 kN. The wet / dry ratio must not be less than 75%.

• The flakiness index for B-R surfacing asphalt must not exceed 25% for the 19,0mm and 13,2mm fraction and 30% for the 9,5mm and 6,7mm aggregate. In addition, at least 95% of all particles must have at least three fractured faces.

• The polished stone value of aggregate, when determined in accordance with SABS method 848, must not be less than 50.

• When tested in accordance with TMH1[1986] method C5, the immersion index of a mixture of the binder and aggregate proposed for use must not be less than 75%. The aggregate used for the test mixture must have a grading within the actual limits for the mix concerned.

• When tested in accordance with TMH1[1986] methods B14 and B15, the water absorption of the coarse aggregate must not exceed 1% by mass, and that of the fine aggregate must not exceed 1,5% by mass, unless otherwise permitted.

• The total fine aggregate used in all asphalt mixes must have a sand equivalent of at least 50, when tested in accordance with TMH1[1986] method B19, and the natural sand where it is permitted to be mixed with the aggregate must have a sand equivalent of at least 30.


Paper 015

• The contractor must, by conducting the necessary tests, satisfy himself that he will be able to produce a mixture meeting the design requirements specified hereinafter, using the aggregate he proposes to supply, within the grading limits specified.

• The grading of the combined aggregate including any filler added in an approved working mix must be within the limits stated in table 5 for the various mixes. The approved grading is designated the target grading. The mean grading of each lot of the working mix (minimum of 6 test per lot) determined from samples obtained in a stratified random sampling procedure, must conform to the approved target grading within the tolerances specified.

2.2.2 Fillers [Colto, 1988] If the grading of the combined aggregates for asphalt surfacing mixes shows a deficiency in fines, an approved filler may be used to improve the grading. Filler may consist of active filler as defined hereinafter or of inert materials such as rock dust having the required grading necessary to improve the grading of the combined aggregates up to a maximum of 2%.

The engineer may order the use of any active filler to improve the adhesion properties of the aggregate. Active filler consist of milled blast furnace slag, hydrated lime, ordinary Portland cement, Portland blast furnace cement, fly-ash, or a mixture of any of the above materials. Active filler must have at least 70% by mass passing the 0,075mm sieve and a bulk density in toluene falling between 0,5 and 0,9 g / m ℓ. The voids in dry compacted filler must be between 0,3% and 0,5%, when tested in accordance with British Standard 812.

Table 5. Grading for B-R asphalts in South Africa [COLTO, 1988].

Percentage passing by mass

Continuously graded Semi-open Open 13,2 mm Minimal

Sieve size (mm)

13,2mm max

19,0mm max

19,0mm max Type 1 Type 2 Type 3

GAP and SEMI GAP

19,0 100 100 100 100 100 100 13,2 100 84 - 96 70 - 100 90 - 100 70 - 100 100 9,5 80 - 100 70 - 84 50 - 82 30 - 50 50 - 80 50 - 70 4,75 50 - 70 45 - 63 16 - 38 10 - 20 15 - 30 20 - 30 2,36 32 - 50 29 - 47 8 - 22 8 - 14 10 - 22 5 - 15 1,18 - 19 - 33 4 - 15 - - - 0,6 13 - 25 13 - 25 3 - 10 - 6 - 13 3 - 8 0,3 8 - 18 10 - 18 3 - 8 - - - 0,15 - 6 - 13 2 - 6 - - -

0,075 4 - 8 4 - 10 1 - 4 2 - 6 3 - 6 2 - 5

NO

T P

RE

FER

ED

A

T P

RE

SE

NT

Nominal properties by

mass

Aggregate 91,0% 91,0% 90,5% 93,5% 93,5% 93,5% Modified Binder 7,0% 7,0% 8,5% 5,5% 5,5% 5,5%

Active Filler 2,0% 2,0% 1,0% 1,0% 1,0% 1,0%

3. ASPHALT MIX DESIGN

3.1 B-R Asphalt Type Selection [Interim Guidelines, 2001]

The characterization of a mix type depends primarily on the spatial composition of the mix (e.g. nominal size, gradation, aggregate, filler and binder characteristics of the mineral components). The selection of a mix type can be optimised by considering the relative demand for each of the different design objectives (i.e. stability, durability, rutting, fatigue, etc.) which in turn is influenced by external factors such as expected traffic, pavement and climatic situation, environmental influence (noise, stormwater, etc) and other special design considerations.


Paper 015

The spatial composition of the mix (i.e. stone- or sand-skeleton) and hence the type of gradation are perhaps the most important choices to be made as far as mix type selection is concerned. The aggregate packing characteristics to a large extent determine the binder content and volumetric properties of the final mix. These elements in turn determine the relative resistance of the mix to deformation, deterioration caused by the environment, etc. table 6 lists the types of gradation covered by these guidelines and also shows a relative rating of the most important performance properties associated with each type.

The selection of a nominal aggregate size is limited by the asphalt layer thickness. Current specifications limit the maximum aggregate size to not more than half of the thickness of the compacted asphalt layer. Designers should however, consider decreasing this limit (i.e. increasing the ratio of layer thickness to maximum aggregate size) whenever conditions are anticipated in which compactibility or segregation may pose problems during construction.

Increasing the nominal aggregate size generally increases the stability of the asphalt but reduces the workability of the mix. Segregation also becomes more problematic when larger aggregate sizes are used.

Table 6. Mix types and typical performance properties.

Performance rating* for

Type of gradation and binder Typical applications

Ease

of d

esig

n

Rut

ting

resi

stan

ce

Dur

abili

ty /

Fatig

ue

resi

stan

ce

Skid

resi

stan

ce

Impe

rmea

bilit

y to

w

ater

Noi

se re

duct

ion

Ease

of c

onst

ruct

ion

Continuous with bitumen rubber

flexible surfacing/overlay 2 3 4 3 4 3 2

SMA with bitumen rubber rut-resistant surfacing 3 5 4 4 2 3 4

Open graded with bitumen-rubber functional layer 3 4 3 4 1** 5 4

Semi-open with bitumen rubber

flexible surfacing/overlay 3 4 5 3 3 4 3

* 1 = Poor, 5 = Good ** Impermeable support layer or membrane required

3.2 Volumetric Design and Performance Testing [Interim Guidelines, 2001]

After the design objectives have been determined, the mix type has been selected and the various components have been evaluated, the actual design process can begin. The basic design procedure consists of the following steps:

• Sample preparation (including sourcing of suitable materials and the proper sampling thereof), compaction and volumetric calculations. This involves compacting specimens at different binder contents and understanding both the compaction characteristics as well as the volumetric and engineering characteristics of the mix as determined in the laboratory and as expected immediately after construction and during its lifetime in the field.

• Engineering properties. Where there is uncertainty and risk, some engineering properties should be determined to increase the level of confidence to that which is required for the application, and to verify the properties expected from the volumetric calculations.

• Field trials. During construction some field trials should be carried out to assess whether the field mixing and compaction processes can produce a layer with the required properties.


Paper 015

The basic steps required for volumetric design and performance testing are shown in figure 2. Although the overall process as illustrated in figure 2 is common to all mix types, specific criteria and procedures apply to individual mix types.

Figure 2. Design process for spatial design and binder content selection.

Before the selection of a target gradation and the calculation of volumetric design parameters are started, designers should be aware of the intended spatial composition of the planned mix. In particular, designers should be aware of the packing characteristics of the planned mix type and how this influences the volumetric design parameters. The type of skeleton structure that is aimed for in the design should be kept in mind and the evaluation and selection of the gradation should ensure that the appropriate packing mechanism is attained. Two opposing packing mechanisms govern the packing of aggregates:

• Substitution, in which the space occupied by the fine aggregate fraction is replaced by an increase in the concentration of the coarse aggregate fractions. This mechanism applies to sand skeleton mixes.

• Filling, in which the spaces between coarse aggregates are filled by an increase in the concentration of fine aggregate. This mechanism applies to stone skeleton mixes.

These two packing mechanisms serve different purposes and have different advantages as far as stability, durability and compactibility are concerned. The selection of a target gradation and analysis of volumetric parameters should be relevant for the particular type of packing mechanism that is aimed for in the design.


Paper 015

3.3 End Point Specification

Table 7 lists the project specification requirements for the process and acceptance control during the site construction period.

Table 7. B-R asphalt end product specification.

Graded Asphalt Requirements

Continuously Semi-Open Open Void in mix (table 5) 2 - 6 3 - 7 Voids in mineral aggregate (%) 17 min -

Indirect tensile strength (MPa) 550 min 600 min

Dynamic Creep (MPa) at 40oC 15 min 15 min

*Marshall stability (kN) 8 - 15 6,5 -12,5 *Marshall flow (mm) 2 - 5 2 - 5 Filler/bitumen Ratio 0,4 - 1,0 0,4 - 1,0 Film thickness 15 min 15 min

Density (% max) 97% - design voids but not less than 93%

97% - design voids but not less than 93%

SHRP Gyratory Nin = 9, Ndes = 128 Nmax = 208

Nin = 9, Ndes = 128 Nmax = 208

NO

T R

EA

DY

FO

R P

UBL

ICA

TIO

N Y

ET

*The stability and flow in the RSA are considered not to be all that applicable to B-R asphalt. It is rather investigated using the X-Y recorder in conjunction with the Marshall press and ITT test. This enables the elastic modulus, plastic modulus, energy stored, hysteresis, toughness of the B-R asphalt, etc, (refer section 4) to be determined.

4. FURTHER DEVELOPMENTS

Standard laboratory tests in use in the RSA for quality control are those listed for rubber crumbs (table 1), extender oil (table 2), B-R modified binder blend (table 3 and 4), gradings for B-R asphalt (table 5), B-R asphalt end-product specifications such as ITT, Cantabro, Schelenberg drainage test, permeability, etc (refer table 7).

Tests discussed hereafter for permanent deformation, fatigue and energy methods are not obligatory for contractors as yet, since they are still in the development phases. However, the RSA Consulting Engineers already use these in the design phase.

4.1 Permanent Deformation

B-R asphalts are almost always used in very highly trafficked area where permanent deformation or rutting is rife. With the expected increases in traffic volumes, vehicle loading and tyre pressure, there is a definite need to adapt hot mix design methods to ensure that B-R asphalt mixes are sufficiently stable to accommodate these increases. Of the available tests, wheel-tracking tests appear to have the strongest correlation with rutting in the field. The test comprises the compaction of a slab while a solid rubber wheel (400mm diameter, 100mm wide) is used for rutting evaluation. Transportek RSA performs this test at 60°C using a 600kg wheel load and the permanent deformation at various wheel pass repetitions.

Typical graphs are included in figure 3. The initial results indicate that a continuously graded asphalt perform the same with B-R binder as with conventional binder. However, the semi-open grading B-R asphalt out performed all.


Paper 015

Figure 3. Typical wheel tracking results.

4.2 Fatigue

In the RSA the four point Bending Beam Test at 5°C using a sinusoidal load with a frequency of 10Hz is used to assess fatigue characteristics. Two modes of loading can be applied i.e. constant strain or constant stress. Failure is defined as the point at which the stiffness of the beam is reduced to 50% of its initial stiffness. Typical results are shown in figure 4. It is clear that B-R asphalt (all grades) outperform conventional asphalt by more than 100 times using fatigue as the criteria.

Figure 4. Typical wheel tracking results.

4.3 Energy Methods

At present we have no measure as to when a B-R asphalt will perform average or excellently. Also, B-R asphalt cost more than conventional asphalt and research to justify this extra expenditure in terms of B-R performance is needed.

Typical laboratory results obtained thus far appears in figure 5 below. Note that Stability-Flow and ITT graphs are used in practice as simplified stress-strain diagrams. Figure 6 demonstrate how much more energy can be stored in the area under the curve for a B-R asphalt.


Paper 015

Figure 5(a). Typical stress-strain diagram for

compression or tension. Figure 5(b). Typical stress-strain

diagram showing elastic hysteresis.

Figure 5. B-R Asphalt energy curves.

Figure 6. Energy stored (area under graph): B-R vs conventional.

For the above research work a load cells and X-Y plotters is coupled to the Marshall press as well as the ITT equipment. The resultant elastic modulus, plastic modulus, area under the stress-strain curve, energy stored, hysteresis, toughness, etc are all being investigated at present. Hopefully we will be able to identify those characteristics which make a particular B-R asphalt a poor or excellent performer so that we can adjust payments accordingly, i.e. should we find that the B-R binder is largely enhanced by more rubber crumbs and more natural rubber, then we need a method to quantity this property and to adjust measurements and payments accordingly.

4.4 Moisture Sensitivity (Modified Lottman Test)

The modified Lottman test for measurement of moisture sensitivity basically relies on ITS measurements taken before and after conditioning by freeze-thaw cycles. A tensile strength ratio (TSR) is then determined. A similar freeze-thaw test is recommended by [Lourens and Jordan, 1989] to assess the potential for stripping in B-R asphalt.


Paper 015

For routine mix design purposes, a minimum TSR of 0,7 is usually specified for mixes in high rainfall areas and high traffic applications, a minimum TSR of 0,8 is recommended. All good B-R asphalts tested have TSR values above 0,8.

4.5 Dynamic Creep Test for Evaluating Rutting Potential

The dynamic creep test specimen is subjected to repeated dynamic loads and the accumulated permanent deformation is monitored as a function of the number of load repetitions. The dynamic creep modules are then calculated as the applied strain vs the permanent strain.

In recent years, research work has raised some doubts concerning the ability of the dynamic creep test to properly and consistently evaluate the rutting property and consistently evaluate the rutting potential of different mix types. For this reason the dynamic creep test must be used in conjunction with other tests such as the wheel-tracking test.

Typical dynamic creep values for bitumen rubber asphalt range between 20 MPa and 30 MPa for both continuously grade and semi-open graded B-R asphalt mixes.

4.6 B-R Structural Equivalence [Sabita, 1998]

The B-R asphalt survey revealed that the B-R asphalt, used in conjunction with SAMI’s in numerous cases, provided superior structural capacity. It also contributed towards dampening reflective cracks. The savings on asphalt thickness when compared to conventional asphalts are considered significant. The somewhat higher initial construction costs of the B-R asphalt is more than offset by the extended service life. In fact, if the lesser thickness of the bitumen rubber asphalt is constructed, the initial construction costs will be less than for a conventional dense graded asphalt overlay.

Table 8. Structural equivalents.

Conventional dense graded asphalt (mm)

Equivalent bitumen rubber gap-graded asphalt

(mm)

Equivalent bitumen rubber gap-graded asphalt on SAMI

(mm) 37 30 - 60 30 - 75 45 30 90 45 30

105 60 45 120 65 45 135 45 BR + 45 (AC) 60 150 45 BR + 60 (AC) 60

Table 9. Crack reflection retardation equivalencies.

Conventional dense graded asphalt (mm)

Equivalent bitumen rubber gap-graded asphalt

(mm)

Equivalent bitumen rubber gap-graded asphalt on SAMI

(mm) 45 30 - 60 30 - 75 45 - 90 45 -

105 60 30

5. FURTHER DEVELOPMENTS

5.1 Manufacturing of B-R Asphalt

The B-R asphalt is manufactured in standard asphalt plants. A high-speed blender and digestion tanks is added to the asphalt plant to facilitate the blending of the B-R binder as shown in figure 7.


Paper 015

Figure 7. B-R asphalt process.

A batch of between 10 tons (minimum) and 25 tons (maximum) of B-R binder is blended into an insulated holding tank equipped with heating flues and auger. The blending process takes approximately 90 minutes. The bitumen-rubber blend is heated to 190 – 205oC and allowed to digest for approximately 60 minutes after which it is ready for use in the manufacturing of the asphalt.

The B-R binder may not be stored for longer than four hours after completion of the reaction process. Should this time be exceeded the material could be re-constituted but which practice is seriously discouraged. The properties of the re-constituted blend must comply in full with the specified requirements of a fresh mix.

The B-R binder is pumped from the reaction / holding tank into a header tank. From here it is metered on a volumetric basis through a calibrated variable speed pump into the mixing drum. The output of the pump is set to provide the required volume of binder as indicated by the rate of asphalt mixing (dry aggregate). The B-R asphalt is then mixed in a conventional manner. Once the quality of the asphalt mix is acceptable it is conveyed to a mixed material storage silo from where it is loaded onto trucks.

B-R asphalt must only be transported in mechanically sound and fully roadworthy vehicles, with special attention being given to oil and diesel leaks. The fresh B-R asphalt is very susceptible to such fluids and will show fatty patches shortly after construction should such leaks occur.

5.2 Placing and Compaction

The B-R asphalt is placed and compacted in a conventional way. The paving and tamping screeds of the paver must be set very accurately as the B-R asphalt is very sticky and streaks easily.

The B-R asphalt is normally compacted using three-point rollers and tandem high frequency vibratory rollers. Pneumatic rollers will only be used when continuously graded B-R asphalt is compacted.

The B-R asphalt normally arrives on site at a temperature of approximately 170 – 190oC. After paving the mix is still extremely hot and very susceptible to roller pick-up. The asphalt that is picked up by the roller wheels and allowed to fall back onto the asphalt mat will cause fatty spots on the final surface if rolled in. It is therefore recommended that the rollers stay some distance behind the paver to allow the mat to cool down. Washing-up liquid soaps should also be added to the water used to wet the wheels of the rollers to prevent pick-up.


Paper 015

6. POTENTIAL PROBLEMS

6.1 Binder Quality

A complication in testing B-R binder is that the chemical reaction changes whilst the product is still above 120oC. Hence, if parameters such as softening point, viscosity, resilience, compression recovery, etc. is specified, then the time at which it must be tested must also be specified. It is practice in South Africa to specify the time of testing as 5 minutes before asphalt manufacture. The B-R binder is only used once it is tested and conforms to the standards of tables 1 to 4. A typical example of time influences is furnished in figure 8.

2.295 2.298

2.303

2.3092.309

2.299

2.2902.286

2.283

2.275

2.285

2.295

2.305

2.315

0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0

Time (Hours)

Bul

k R

elat

ive

Den

sity

Figure 8. Bitumen rubber asphalt density vs. time.

6.2 Binder Content

The binder content can be determined by one of the following three methods:

• Extraction method

• Nuclear binder content gauge

• Incineration oven

The extraction method is still the pre-scribed method. This method is however, time-consuming, messy and can be a health risk due to the use of toluene during the extraction. The nett extracted bitumen content is multiplied with a sample factor that is determined by extracting a lab mixed control sample. The determination of the sample factor is very important and it is recommended that the control sample be extracted directly after being made up. No curing of the control sample in an oven is recommended as this lead to poor repeatability in the determination of the sample factor.

The nuclear binder content gauge (AC2) was found to be a very accurate measuring instrument. The instrument is also calibrated using laboratory made up samples. The advantage of this instrument is its ease of use and short testing cycle.

The incineration oven (Troxler oven) burns of the binder at a temperature of approximately 500oC and automatically calculates the mass loss to determine the binder content. This method is also a quick and easy method for binder content determination.

6.3 Bitumen Rubber Binder

The bitumen-rubber binder is highly viscous and special pumps are required to pump it. These pumps also have a limited life span and should be replaced timeously.


Paper 015

From previous studies [Lourens and Jordaan, 1989] it was also found that the high viscosity of the bitumen rubber binder when mixed with the aggregate to make asphalt is the most important factor which governs the adhesion of the binder to the aggregate to prevent stripping or internal pumping of the asphalt.

From laboratory studies on particular bitumen rubber blends it was found that there is a critical time period in which the viscosity reaches its peak before it decreases. Mixing should not be done during this time period, as the ‘wetting power’ of the blend is then at it’s lowest.

6.4 Construction

As mentioned previously the bitumen rubber asphalt is extremely sticky. Handwork should therefore be limited to the minimum. All excess material should also be removed from the road surface as soon as possible as these materials can be carried over onto newly completed surfacing. The newly completed surface is also very sensitive to oil, hydraulic oil and diesel spillages creating fatty patches.

The sticky nature of the asphalt also makes it prone to segregation and accumulation of fines on auger blades and paver vibrators. Care should always be taken to prevent segregation and all equipment should be thoroughly cleaned at the end of each shift.

A well worked out rolling pattern should be established during the construction of the trial sections to prevent roller pick-up under the wheels and to prevent over or under compaction.

6.5 Quality Control and Testing

The non-homogeneous character of the B-R binder possesses certain challenges during the monitoring of the work. The best way to overcome these challenges is to have a very well worked out testing programme in place and to do time studies on site to determine to the sensitivity of the mix to time and temperature.

7. SUMMARY AND CONCLUSIONS

• Only basic constituent materials such as bitumen, rubber crumbs, extender oils and aggregate of high quality which conforms to the latest specifications should be used to manufacture a bitumen rubber binder.

• Stricter quality control measures must be in place and only bitumen rubber binder conforming to all the requirements must be used in the manufacturing of B-R asphalt. Specific indicators such as plate flow, softening point, viscosity, compression recovery, resilience etc. will confirm that the chemical digestion process followed the correct manufacturing route and that the reaction was carried to completion. If these indicators fall outside the specified limits it would indicate that the chemical process followed an undesirable route and that the binder should not be used.

• The Marshall method is still the bases for the design of the B-R asphalt, but is supplemented by other criteria such as rutting, fatigue, modified Lottman, dynamic creep, resilient modules etc.

• Binder content and grading determination should be done just before asphalt batching to obtain the most accurate results. This is due to the continued digestion characteristic of the B-R binder.

• The asphalt mix design process including the type selection, the volumetric design process as well as the B-R asphalt end point specification is furnished in this paper.


Paper 015

• Bitumen rubber asphalt is an excellent product with a proven track record of over 17-years in South Africa. With the necessary attention to detail the expected life span can be as much as 20-years [Potgieter et al, 1998]. The somewhat higher initial construction cost of B-R asphalt are negligible in view of the savings in traffic accommodation, by-passes and the extended life rendered by the bitumen rubber asphalt overlay when assessed over the full life cycle.

8. ACKNOWLEDGEMENTS

• The authors wish to acknowledge the input by Ronnie Renshaw (Pr. Eng.) who contributed largely to the importation of Arm-R-Shield bitumen rubber technology to the RSA.

• All the State Road Authorities and personnel (National Roads Agency, Provinces and Metropoles) who pioneered this fine product into the RSA.

• All the consulting engineers and their technicians who diligently worked on these projects.

9. REFERENCES

SABITA - South African Bitumen and Tar Association, 1988. Test methods for Bitumen Rubber, Manual 3.

British Standards, No 903.

American Society for Testing and Materials. COLTO - Committee of Land Transport Officials-South Africa, 1988. Standard Specifications for Road and Bridge Works for State Road Authorities. South African Bureau of Standards. Technical Methods for Highways – South Africa (TMH1), 1986. Standard Test Methods for Road Construction Materials. Interim Guidelines for the Design of Hot-Mix Asphalt in South Africa, June 2001.

Lourens T. and Jordaan E., The Prevention of Stripping of Bitumen Rubber Asphalt, 1989. Proceedings of the 5th Conference on Asphalt Pavements for Southern Africa, Swaziland.

SABITA - South African Bitumen and Tar Association, 1988. Modified Binders ’98 – Seminar. Potgieter C.J., Pike L.L. and Balmaceda P., 1998. The Bitumen Rubber Asphalt of Buccleuch Interchange – Report on the 1986 Construction and 12-years service life, Proceedings of the South African Transport Conference.

Potgieter C.J., Sadler D.E. and De Villiers E.M., 1998. Bitumen Rubber Asphalt: Report on the long term performance in South Africa, Proceedings of the 9th International Conference on Asphalt Pavement, Copenhagen.

bitume rubber asphalt.pdf

Documents