A Comparative Study Of Roller And Hammer Compacted Asphalt Concrete

In this work, Dense and Gap graded Asphalt Concrete samples were compacted in the laboratory using two modes, the first one was the traditional Marshall method of hammer compaction; Cylindrical specimens were constructed using three different Asphalt percentages; While the second mode was the TRRL roller wheel compaction, slab samples of (30×30×7) cm were constructed using the same Asphalt percentages, core specimens were obtained from the slabs. All of the Asphalt Concrete specimens were subjected to Marshall properties determination, indirect tensile stress test, Hveem stability and cohesion tests. A comparative analysis of testing results was conducted. It was concluded that Dense graded Asphalt Concrete shows superior quality when compared to Gap graded one when roller wheel compaction was adopted. Gap graded Asphalt Concrete shows higher quality when Marshall Hammer compaction was adopted

The demand is increasing for high quality and more durable paving mixture for modern traffic.Cost for producing maximum density or well graded aggregate is also increasing because of rapid depletion of natural deposits that meets the specifications of continuous grading, and because of increased cost of labor, transportation and processing.It is therefore desirable to examine the suitability of other types of aggregates grading as compared to well graded maximum density mix, so that more efficient use can be made of the available sources of aggregates.On the other hand, it was believed that conventional sample preparation such as that of Marshall method which uses small compacted cylindrical specimens could be replaced by another method better adapted to the new requirements that must be met by Asphalt Concrete.It was felt that the roller wheel compactor developed by TRRL is more representative of conditions at the job site.The roller wheel compactor could give more homogeneous samples in which the element arrangement pattern come close to those obtained on the job site.

Previous Experience
Gradation of aggregate to a curve of maximum density, developed by (Fuller and Thompson-1909) and later modified and confirmed by a number of other investigators (Lees-1973) is generally accepted as the most desirable grading for the production of good and economical Portland Cement and Asphalt Concrete.These grading are also referred to as well graded or continuously grading.On the other hand, an aggregate is said to be gap graded or (skip or discontinuously graded) when certain particle sizes in the grading of aggregate are missing.The absence of such sizes can be achieved by deliberately omitting them to obtain certain desired properties of the mixture.A large amount of literature, especially theoretical, can be found on the packing of aggregate particles and maximum density or minimum porosity grading including the classical work on concrete proportioning by (Fuller and Thomption-1909).There is also abundant published information on gap graded concrete as compared to the corresponding continuously graded concrete (Lees-1973;Sarsam-2002); However, reported data on gap graded Asphalt Concrete mixtures are few and scattered (Brien-1972;Lees-1974;Marias-1979;Sarsam-1987).The effect of gradation (Dense and Gap) on physical properties of asphalt concrete was studied by (Sarsam-1987;Sarsam-1997) using usual hammer compaction, he concluded that Gap graded mixes shows superior Hveem stability, cohesion, specific gravity, compatibility and strain resistance at various Asphalt content when compared to dense mixes.The effect of gradation on the behavior of rubber modified asphalt  3 shows its properties.

Asphalt Cement
Asphalt Cement of grade (40-50) was obtained from Gayara oil refinery stock.Table 4 shows its various properties.

Testing Program 4.1 Marshall samples construction
The required amount of aggregate of different sizes to meet the dense or gap gradation were heated to 160˚C and combined, the required amount of Asphalt Cement heated to 150˚C was added and mixed with the aggregates for two minutes using a mechanical mixer, then the mix was poured into the preheated moulds and subjected to Marshall hammer compaction using 75 blows for each side as per the standard procedure.Three different Asphalt percentages (4, 5, 6) % were adopted for each gradation type and triplicate samples were constructed for each Asphalt percentage.A total of 36 Marshall Specimen was prepared.

Roller wheel slab samples construction
The required amount of aggregate of different sizes to prepare a slab specimen of (30×30×7) cm size was weighted, heated to 160˚C and combined.Asphalt Cement was also heated to 150˚C, and then the predetermined amount of Asphalt was added to the aggregate into the preheated mixing bowel.Mechanical mixing was conducted for two minutes, then the mix was poured into the preheated slab mould of the TRRL roller wheel tracking machine, leveled with a spatula, then it was subjected to 10 passes of the steel roller wheel for each of the three stages of compaction using different compaction effort for each stage by changing the applied normal load.A primary compaction by roller was applied using 10 passes of the machine shoe with a normal load of 10kg/cm width, followed by 10 passes using a normal load of 20kg/cm width.Such compaction may represent the primary and heavy compaction applied by steel and pneumatic tire rollers in the field.The final compaction was demonstrated by the application of 10 passes of the roller using 45kg/cm width normal load representing the finishing compaction by steel rollers in the field (Sarsam-2002).Samples were kept overnight in the mould for cooling, then withdrawn from the mould for further testing.A total of 12 slabs were compacted as above using both dense and gap gradations with three Asphalt percentages of (4, 5, 6) %.Drilled core samples of 10cm diameter were obtained from the slabs.A total of six cores were obtained from each slab.

Testing of samples
All of the hammer compacted samples and the core samples were subjected to bulk density determination, and divided into three groups, then tested for Marshall stability and flow, Hveem cohesion and stability, and indirect tensile strength test as per the Asphalt institute(MS-2) and (ASTMC496-64T).

Marshall properties
The first group of hammer compacted and core samples were tested for Marshall Properties at 60˚C.As demonstrated in figure 1 Figure 3 shows the changes in voids percentages at different asphalt content.Such changes is further supported by figure 4 which illustrates the changes in the voids filled with bitumen.Gap gradation shows higher voids content and lower voids filling with asphalt.Gap gradation has higher bulk specific gravity than Dense mixes when hammer compaction was employed, while Dense mixes have the highest specific gravity when roller compaction was adopted as demonstrated in figure 5, such behavior agrees well with other research findings (Marias-1974;Takallow-1986;Sarsam-1987;Sarsam-1997;Sarsam-2000), when Gap gradation showed hardly any traffic compaction, and were at time of laying were close to laboratory density, while Dense gradation densified progressively under the traffic for a period of three years after which, it reached an asymptotic value close to the laboratory density as indicated by (Marias-1974) .Hammer compaction slightly increases the bulk specific gravity by 0.95%.Such behavior is further supported by the voids and the voids filled with bitumen relationships.

5.2Hveem cohesion
The second group of samples was subjected to Hveem stability and cohesion test at 60˚C. Figure 6 shows that for roller compacted samples, Dense gradation have better cohesion than Gap gradation at low and high Asphalt content, while both gradations have almost the same cohesion at 5% Asphalt content.This may be attributed to the good packing of aggregates of Dense gradation which is effective at both low and high Asphalt percentages, the particles interlock seems to be much better than that of Gap gradation.
When hammer compaction was employed, Gap gradation shows higher cohesion at medium Asphalt content and lower cohesion at 4 and 6% Asphalt content.This situation may indicate that Gap gradation resist the sliding of aggregate particles over each other at medium Asphalt content when compacted with impact loading such as that of hammer compaction, and the sufficient Asphalt will increase the particles contact to each other.Hammer compaction increases Hveem cohesion by 70%.

Hveem stability
Gap gradation has higher Hveem Stability than Dense gradation and the Hammer compaction shows higher stability values than roller compaction at all of the tested Asphalt percentages as illustrated in figure 7. It was felt that the impact load has a great effect on changing the particles orientation and giving high Hveem stability values which reflects the high rutting resistance of Gap gradation.This agrees well with the findings of (Marias-1972;and Sarsam-1987).Hammer compaction increases Hveem stability by 25%.

Indirect tensile strength
The third group of samples was subjected to indirect tensile strength test at 25˚C. Figure 8 illustrate that for roller compaction, Dense gradation shoes higher tensile strength than Gap gradation for the range of Asphalt Cement used, such behavior may be related to good cohesion.On the other hand, same behavior was noticed when hammer compaction was adopted.
Hammer compaction increases tensile strength values by 68%; This may be attributed to the change in particle interlock pattern when using impact loading.

Table 1 : Physical properties of aggregates Table 2: Grain size distribution of the design mixes
*Dept.0fCivilEng.College-M0SUL Univ.536