In general, bituminous hot mixtures consist of two main components: bitumen and aggregates. The function of the bitumen, which typically represents 4–7% of the pavement, is to act as a binder in between the aggregate skeleton, giving the asphalt sufficient internal cohesion. It is, therefore, of vital importance that the bitumen has a strong adhesion to the aggregate surface. Adhesion between the aggregate and the bitumen is important with respect to the long-term performance of bituminous hot mixtures. If good adhesion between aggregate and bitumen cannot be provided, durability and resistance to shear stresses-plastic deformation-stripping decrease, whereas susceptibility to low-temperature cracks and potholes due to moisture increases [1, 2] Previous studies indicate that adhesion between aggregate and bitumen is affected by numerous factors, such as the content and chemical properties of the bitumen, the type of aggregate, the surface properties of the aggregate, and the pH of the water [3–10]. An important factor affecting adhesion is dust on the aggregate surface [10, 11]. Dusty aggregates may generally be referred to as aggregates coated with materials smaller than 75 μm . Chadbourn et al.  state that the generation of dust can be caused by aggregate degradation due to aggregate handling during loading, hauling, mixing, and construction, and degradation can cause more fines, significantly changing the gradation of bituminous hot mixtures. Kiggundu and Roberts  report that dust on the aggregate surface can enhance the formation of blisters and pits. These forms of film defects may lead to the rupturing of the film and cause moisture damage. Tarrer and Wagh  state that dust on the aggregate surface weakens asphalt-aggregate bonding by preventing intimate contact between the aggregate and bitumen, and also causes stripping.
This study was designed to investigate the effects of different dust contents on the aggregate surface of bituminous hot mixtures. For this purpose, a Marshall design was performed using two different aggregates with three different dust contents. The test results show that the properties of bituminous hot mixture produced from dusty aggregates deteriorated, no matter the amount of dust, and demonstrate that deterioration varies by type of aggregate.
2 Materials and methods
Two different limestone aggregate specimens (A, B) were used. Dust was artificially generated on the aggregate surfaces to create two different dust contents. Aggregate specimens were grouped according to dust content. A1, B1; A2, B2; and A3, B3 specimens had no, medium, and high dust content, respectively. Specimens with no dust content were considered washed aggregate (A1, B1).
A 50/70 penetration bitumen, acquired from Aliaga refineries, was used. The rheological properties of the bitumen are given in Table 1.
Rheological properties of the bitumen.
|Penetration at 25°C||63||ASTM D 5|
|Specific gravity||1.060||ASTM D 70|
|Softening point (°C)||49||ASTM D 36|
|Loss on heating (%)||2||ASTM D 6|
|Flash point (°C)||296||ASTM D 92|
|Ductility (5 cm/dk)||>100 cm||ASTM D 113|
|Viscosity at 135°C||0.420 Pa s||ASTM D 4402|
|Viscosity at 165°C||0.114 Pa s||ASTM D 4402|
The testing procedure consisted of four different stages. In the first stage, aggregate specimens were cycled for artificial dust generation on the aggregate particle surfaces. The second stage was aggregate tests, the third stage was stripping and adhesion tests, and the fourth stage consisted of bituminous hot mix tests.
The first stage consisted of obtaining different dust contents on the aggregate surface. Dust on bituminous hot mix aggregates is one of the most common factors that cause raveling, stripping, and segregation. Although the aggregates are washed when they are produced at the source, a significant amount of dust may be generated during transportation and handling, particularly with crushed limestone. To investigate the effect of dust on the aggregate-binder bond, it was decided to prepare bituminous hot mix aggregate test specimens using aggregates with three different dust contents (zero, medium, and high).
Instead of artificially adding dust to the aggregate specimen, it was decided to allow the aggregate to generate its own dust under conditions somewhat similar to field conditions (i.e., handling). This was done in the laboratory using cylindrical drum test equipment. This cylindrical drum was placed on a milling device and rotated for a specified duration. Preliminary tests showed that too much dust is generated when steel balls are used, which also do not accurately mimic field dust generation conditions. Therefore, the milling drum equipment was used without steel balls. The duration of the drum rotation affects the amount of dust generated.
After preliminary trials, the dust contents were evaluated, and 30 and 40 min of drum rotation were selected to represent the conditions of medium and high dust content. Washed aggregate specimens (A1 and A2) were accepted as not having any dust content. Before putting them in the milling drum equipment, all aggregates were washed then dried in a drying oven for 24 h. The aggregates were then stored in bags until the bituminous hot mixture specimens were prepared.
The second stage consisted of aggregate tests. These were gradation analysis, unit volume weight, Los Angeles abrasion, aggregate impact, freezing-thawing, and resistance loss after freezing-thawing tests.
Adhesion and stripping tests were conducted in the third stage of the study. To determine the effects of different dust contents on adhesion, the Vialit adhesion test was conducted. Similarly, to determine the stripping resistance of aggregate specimens, the Nicholson stripping test was performed.
The fourth stage consisted of bituminous hot mixture tests. Marshall stability-flow tests were conducted to determine the optimum bitumen ratios. Marshall stability-bitumen content %, Marshall flow-bitumen content %, void %-bitumen content %, bulk specific gravity-bitumen content %, void filled with bitumen-bitumen content %, and void in mineral aggregates (VMA) %-bitumen content % charts were plotted. A mechanical immersion test  was performed for each of the aggregate specimens, then afterwards compared with the calculated loss of stability values of the specimens. An indirect tensile test  was also performed on the bituminous hot mixture specimens produced with optimum bitumen content to compare the fatigue life of the specimens. A flowchart of the testing procedure is shown in Figure 1.
3 Test results
3.1 The results of artificial dust generation on the aggregate surface
The amount of dust retained on the aggregate surfaces was determined and plotted in Figure 2 as “percent dust retained”. The trend line of these 40-min points turned out to be linear. Dust generation on the aggregate surface obtained from drum rotation for 30 and 40 min was enough to obtain aggregates specimens with medium (A2, B2) and high (A3, B3) dust content.
In this study, aggregates with artificial dust generated on their surfaces were used. The most important reason for this was to create two different dust contents on the same specimen. If natural dusty aggregate was used, we would not have been able to create specimens with two different dust content for each aggregate specimen. With artificial dust generation on aggregate surfaces, we are able to see clear differences for all specimens of mixtures.
3.2 Aggregate test results
The selected design gradations of aggregate specimens used to produce bituminous hot mixture specimens for wearing course layer are given in Table 2. Specific gravity, water absorption, and unit volume weight test results are shown in Tables 3 and 4. As shown in the results of the water absorption test, aggregate B was more porous than aggregate A. This means the surface of the B specimen kept more dust than the A specimen.
Selected design gradation of the aggregate specimens.
|Sieve no.||Sieve (mm)||Passing (A) %||Passing (B) %||Lower–upper limits|
Test results of specific gravity and water absorption.
|Samples||Apparent specificgravity (g/cm3)||Bulk specific gravity (g/cm3)||Water absorption (%)|
|A (>No. 4)||2.84||2.84||0.34|
|A (No. 4–No. 200)||2.81||2.75||0.68|
|A (mineral filler)||2.87||–||–|
|B (>No. 4)||2.72||2.72||0.80|
|B (No. 4–No. 200)||2.69||2.66||1.40|
|B (mineral filler)||2.65||–||–|
Test results of unit volume weight.
|Properties/samples||A (>No. 4)||A (||B (>No. 4)||B (|
|Loose unit volume weight (g/cm3)||1.55||1.62||1.62||1.46|
|Dense unit volume weight (g/cm3)||1.62||1.69||1.69||1.51|
Various wearing tests were performed on the aggregate specimens. The wearing loss of both of the specimens was within specification limits. However, the wearing loss of the B specimen was higher than that of the A specimen. Similarly, freezing-thawing and loss of resistance after freezing and thawing were higher in the B specimen than in the A specimen (Table 5).
Test results of the aggregate specimens.
|Sample A||Sample B|
|Los Angeles abrasion value (%)||23.7||29.1||<35||ASTM C131-89|
|Aggregate impact value (%)||10.2||11.4||<18||ASTM D3744|
|Loss of freezing-thawing (%)||1.3||1.9||<12||TS EN 1097-2|
|Loss of resistance after freezing-thawing test (%)||35.2||38.9||–||TS EN 1097-2|
3.3 Vialit adhesion and Nicholson stripping test results
To determine the stripping resistance of the aggregate specimens, a Nicholson stripping test was performed. The test results showed that the A specimen was more sensitive to stripping under water than the B specimen. Inadequate resistance to stripping in aggregates may adversely affect the physical properties of the mixture. Indeed, the A specimen’s indirect tensile strength was lower than that of the B specimen. The Nicholson stripping test results are shown in Figure 3.
To determine the adhesion resistance of the aggregate specimens and the bitumen, a Vialit adhesion test was performed. The test results showed that the number of falling aggregates increased parallel to the dust content of the aggregate surfaces (Figure 4). Because of the higher porosity of the B specimens, the number of falling aggregate particles was less than in the A specimens.
3.4 Bituminous hot mix test results
3.4.1 Marshall test results
Both of the aggregate specimens with three different dust contents were used, and Marshall design with six different wearing course layer bituminous hot mixtures were carried out (A1, A2, A3, B1, B2, B3). Bituminous hot mixture specimens were prepared according to the curve within the Turkish General Directory of Highways (TCK)  wearing course specification limit, and gradation curves and tests were scheduled on the basis of 0.5% increments of bitumen content. Three bituminous hot mixture specimens were prepared for each different bitumen content. Marshall stability-flow tests were conducted; the weights of specimens in air and water and the surface dry-saturated weights of the specimens were measured; and stability-% bitumen content, bulk specific gravity-% bitumen content, void filled with bitumen-% bitumen content, and void %-bitumen content curves were plotted. Optimum bitumen contents were calculated from the curves for each of the design groups. While the amounts of optimum bitumen were determined as 4.35, 3.93, and 4.18 for specimens A1, A2, and A3, respectively, the amounts of optimum bitumen were determined as 4.98, 4.70, and 4.93 for the specimens B1, B2, and B3, respectively. The relation between % bitumen and Marshall stability is shown in Figure 5.
Stability is one of the important features of bituminous hot mixtures. It determines the strength of bituminous hot mixtures against pressure, shear, and horizontal stresses.
As can be seen in Figure 6, with the increase in the amount of dust on the aggregate surface, the values of maximum stability decreased for both A and B design series. However, the decrease in the stability of A3 and B3 was clearer than in the others. Bitumen-aggregate adhesion was weakened, as the dust content of A3 and B3 was more than the other specimens.
Density is another important feature for bituminous hot mixtures. An increase in density leads to better impermeability, durability, and resistance to aging, stripping, and raveling [17, 18]. According to the test results, the maximum stability value is inversely proportional to the dust content on the aggregate surface. However, a decrease in the values of bulk specific gravity was not observed. The researchers proposed that some of the dust on the aggregate surface exhibited a mineral filler effect in the mix and caused an increase in mix density. The maximum density values of A1, A2, and A3 were obtained as 2.572, 2.583, and 2.568 g/cm3, respectively, whereas the maximum density values of B1, B2, and B3 were obtained as 2.425, 2.441, and 2.445 g/cm3, respectively. In the A and B series of specimens, the relation between bitumen and maximum bulk specific gravity is shown in Figure 6.
The feature of “void filled with bitumen” controls the plasticity, durability, and friction coefficient of the mixes, and also provides a final bitumen film around the aggregate particle [17, 19]. Because of its effect on resistance to stripping, raveling, and atmospheric effects (rain, snow, freezing-thawing), the value of the void filled with bitumen is important. The percentage of void filled with bitumen versus the optimum bitumen ratio was 76%, 73%, and 76% for A1, A2, and A3, respectively, and was 75%, 72%, and 74% for B1, B2, and B3, respectively. With the increase of dust content on the aggregate surface, the percent values of void filled with bitumen has decreased. The decrease in the B-series mixtures was more stable. Therefore, if the dust content on aggregate surface increases, the durability of bituminous hot mixtures decreases. According to the Turkish State Highway Specification, the value of the void filled with bitumen should be between 65% and 75% . The relation between the void filled with bitumen and bitumen in the A and B specimens is given in Figure 7.
Percent void is another important parameter for bituminous hot mixtures. The lower and upper limits have been described in the specification for bituminous hot mixtures of wearing courses . If the void value is higher than the specification’s upper limit, the stability of the bituminous hot mixtures can decrease or premature failures can be seen on the pavement. However, to avoid flushing in hot climates, hot bituminous mixtures should have some amount of void volume . The percentage void versus the optimum bitumen content was 3.0%, 2.5%, and 2.8% for A1, A2 and A3, respectively, and was 3.4%, 3.8%, and 3.3% for B1, B2, and B3, respectively. The test results showed that the percentage void of A2 and A3 are lower than the specification limit. The percentage void in other specimens was within the specification limits. The researchers believe that the dust content of the A2 and A3 specimens behaved like a filler, such that the void decreased in these specimens. In the B series of specimens, the dust content on the aggregates caused a decrease of adhesion and compaction, and caused an increase in the void. The relation between percentage void and percentage bitumen in the A and B series of specimens is shown in Figure 8.
The flow value provides information about the features of plasticity and flexibility of bituminous mixtures under traffic loads, and also measures the interface friction of compacted mixtures [17, 18, 20]. The values of flow were 2.92, 2.96, and 3.14 mm for A1, A2, and A3, respectively, and were 3.16, 3.40, and 3.48 mm for B1, B2, and B3, respectively. According to the Turkish State Highway Specification, the value of flow should be between 2 and 4 mm . Thus, according to the results of the flow test, with the increase in dust content on the aggregate surface, the susceptibility to plastic deformation of bituminous hot mixture specimens also increased. The relation between % bitumen and % Marshall flow is given in Figure 9.
VMA is an important parameter with respect to durability and the locking of aggregates to each other [17, 19]. The % VMA versus percentage optimum bitumen of A and B series of specimens was 12.8%, 11.6%, 12.4% and 13.9%, 14.0%, 13.7%, respectively. In general, with an increase in dust content on the aggregate surface, VMA also decreased, as dust on the aggregates absorbs the available bitumen in the mix, decreasing the percent value of VMA. Chadbourn et al.  also report that with increasing dust content, % VMA decreased in both of their specimens. The relation between % VMA and % bitumen is shown in Figure 10.
3.4.2 Mechanical Marshall immersion test results
Another disadvantage of dust on the aggregate surfaces is stripping due to water action. To determine how dust on the aggregate surface affects stripping in the bituminous hot mixtures, a Marshall immersion test was performed. The loss of stability value of the A and B series of aggregate specimens increased parallel to the dust content on the aggregate surface. The mechanical Marshall immersion test results are shown in Figure 11. The test results show that if the dust content on the aggregate surfaces is increased, the durability and strength of the bituminous hot mixtures against environmental effects will decrease.
3.4.3 Indirect tensile test results
To determine the strength of the bituminous hot mixtures against plastic deformation in different temperature conditions (5°C, 25°C, 40°C), an indirect tensile test was performed. The indirect tensile test determined how indirect tensile strength varied with different dust contents of the aggregate for the six different bituminous hot mixture specimens (A1, A2, A3, B1, B2, B3). The variation of indirect tensile strength of the A and B specimens is shown in Figures 12 and 13, respectively.
The test results show that the B specimens have a fine adhesion strength and stripping resistance, and also have better indirect tensile strength than the A specimens. As the dust content on the aggregate surface increased, the indirect tensile strength of bituminous hot mixes decreased in both of the A and B specimens, at 5°C and 25°C. Because the bitumen may reach its softening point at 40°C, the indirect tensile strength of the specimens decreased at this temperature; however, different dust contents did not affect the indirect tensile strength of the specimens at 40°C.
4 Conclusions and recommendations
On the basis of the results derived from the laboratory, the following conclusions can be drawn:
- According to the Marshall stability test results, the variation of the maximum stability of the A and B series of specimens was A1>A2>A3 and B1>B2>B3. Owing to their high dust content, the decrease in the value of maximum stability in the A3 and B3 specimens was clearer than in the others.
- The Marshall immersion test results show that with an increasing dust content, both A and B specimens had decreased resistance to moisture effects. The main reason for this decrease could be a further weakening of aggregate adhesion with increasing dust content on the aggregate surfaces.
- With an increase of the A specimen’s dust content, the percent void of the specimens decreased. The most important reason for this is that dust affected the A specimens like a filler, while it affected the B specimens as dust. Indeed, the loss of Marshall stability in the B specimens was higher than in the A specimens, even as an increase of dust content in both of the specimens resulted in an increase in the densities of the mixtures. Increasing dust content on the aggregate surfaces led to a decrease of VMAs.
- The flow test results show that if the dust content on the aggregate surface is higher, the susceptibility trend to plastic deformation of the bituminous hot mix pavement was slightly higher. While this trend is clear in all of the A series of specimens, only B3 displays a clear effect within the B series of specimens.
- The test results show that dust on the aggregate surface has a negative effect on the properties of hot bituminous mixtures without taking into consideration the differences between aggregates. The dust contents of the aggregate surface were ≥0.8% and ≥1.1% in the A and B specimens, respectively, while the bituminous hot mixture properties also markedly deteriorated.
- According to the Marshall immersion test results, confirming the results of Kandhal et al. , when bituminous hot mixtures produced from dusty aggregates were exposed to moisture, early deteriorations, such as stripping, occurred. This result is due to the ability to keep dust on the surface, which was higher in B specimens than in the A specimens; also, the loss of stability in the B specimens was higher than in the A specimens.
- Factors such as type of aggregate, porosity, wearing, and surface properties also affect dust generation on the aggregate surfaces. Thus, according to the artificial dust generation test results, more dust collected on the surface of the B specimens, which had higher wear values and porosity.
- Dust content on the aggregate surface leads to a weakening of adhesion, such that the characteristics of the hot bituminous mixture gradually deteriorated. Because dust contents on the aggregate surface can vary according to the aggregate type, the progression of deteriorations in bituminous hot mixtures can also vary.
The authors wish to thank the General Director Assistant of ISFALT, Dr. İbrahim Sönmez; the Director of the Asphalt Department of Afyonkarahisar Municipality, Güven Kayhan; Municipality Asphalt Laboratory assistant Ahmet İba; and MSc student/lecturer Ayfer Elmacı for their assistance.
Mansell T. Raveling in Hot Mix Asphalt Pavements. Available from http://www.graniterock.com/technical_notes/raveling_in_hot_mix_asphalt_pavements.html, 2010. Accessed on 24 October, 2010.
AkzoNobel. Adhesion Promoters. Technical Bulletin, 2007, p. 3.
Tarrer AR, Wagh VP. The Effect of the Physical and Chemical Characteristics of the Aggregate on Bonding. Strategic Highway Research Program, National Research Council, Report SHRP-A/ UIR-91-507: Washington, DC, 1992.
Curtis CW, Ensley K, Epps J. Fundamental Properties of Asphalt-Aggregate Interactions Including Adhesion and Absorption. Strategic Highway Research Program: Washington DC, 1993, pp. 3–6.
Yoon HH, Tarrer AR. Transp. Res. Rec. 1988, 1171, 37–43.
Senadheera S, Wm Tock R, Hossain MS, Yazgan B, Das S. A Testing and Evaluation Protocol to Assess Seal Coat Binder-Aggregate Compatibility. Texas Tech University, Center for Multidisciplinary Research in Transportation, Project Summary Report 0-4362-S, 2006.
Karaşahin M, Saltan M, Gürer C, Aktaş B, Uz VE, Çetin S, Taciroğlu M. Proceedings of the 5th National Asphalt Symposium, Ankara, Turkey, 2009, pp. 207–215 (in Turkish).
Lu Y, Wang L. Int. J. Pavement Eng. 2010, 11, 393–401.
Gürer C. Determination of Parameters Affecting Seal Coat Performance and Development a Performance Model. PhD Thesis, Suleyman Demirel University: Isparta, Turkey. 2010, 4, pp. 115–121 (in Turkish).
Doğan H. An Investigation on Effect of Aggregates Dusty Degree on Bituminous Hot Mixes. MSc Thesis, Afyon Kocatepe University: Afyonkarahisar, Turkey, 2011, 2, pp. 26–37 (in Turkish).
Little DN, Jones DR. Chemical and Mechanical Processes of Moisture Damage in Hot-Mix Asphalt Pavements. Moisture Sensitivity of Asphalt Pavements-National Seminar, Transportation Research Board: San Diego, CA, Chap 2, pp. 37–70.
Chadbourn BA, Stok EL, Newcomb Jr DE, Crow BL, Spindle S. The Effect of Voids in Mineral Aggregate (VMA) on Hot-Mix Asphalt Pavements. Minnesota Department of Transportation, Final Report, 1999, 6, pp. 71–72.
Kiggundu BM, Roberts FL. Stripping in HMA Mixtures: State-of-the-Art and Critical Review of Test Methods. National Center for Asphalt Technology NCAT Report 88-02: Auburn, AL, 1988, Chap 1, p 5.
AASHTO. Standard Method of Test for Effect of Water on Cohesion of Compacted Bituminous Mixtures. AASHTO, T 165-2: Washington, DC, 2002.
ASTM. Standard Test Method for Indirect Tensile (IDT) Strength of Bituminous Mixtures. ASTM D6931-07: West Conshohocken, PA, 2007.
Turkish State Highway Specifications. Republic of Turkey, Ministry of Transport, General Directory of Highways-TCK: Ankara, Turkey, 2006 (in Turkish).
Akbulut H, Gürer C. Build. Environ. 2007, 42, 1921–1930.
Akbulut H, Gürer C, Çetin S. Inst. Civil Eng. Transp. 2011, 164-TR2, 111–123.
Umar F, Ağar E. Road Pavement. İTÜ: İstanbul, Turkey, 1991.
ACMA. Bituminous Mixes and Flexible Pavements. Product Group of the British Aggregate Construction Materials Industries (BACMI): London, 1992.
Kandhal PS, Lynn CY, Parker F. Test for Plastic Fines in Aggregates Related to Stripping in Asphalt Paving Mixtures. Report 98-3, National Center for Asphalt Technology, 1998.