CHAPTER 4
RESULTS AND DISCUSSION
4.1 Introduction
This chapter discusses the engineering properties and performance of porous asphalt based on the laboratory test results. The engineering properties include specimen volumetric properties such as porosity, density, voids in mineral aggregate and voids filled with bitumen. In addition, the mix stressstrain behaviour and other performance characteristics will be analysed based on mechanical testing results of the bituminous mixes. The mechanical tests can be either fundamental, simulative or empirical in nature. The results obtained in this investigation are analysed according to binder types and varying percentage of DAMA used. The test results obtained are also compared with other works in order to assess the general performance of the mixtures tested.
4.2 Volumetric Properties
4.2.1 Porosity
In porous asphalt, binder drainoff takes place once the critical binder content is exceeded. In porous asphalt and unlike in dense mix, the binder will never be able to completely fill up the voids. The binder in an open mixture is required to partially fill the voids so that a high porosity mix is obtained but the magnitude of porosity should not sacrifice its strength and other related properties. Hence, the porosity in a porous mix must be carefully selected.
In this study, the average porosity reduces linearly as the binder content increases. This is evident from the porositybitumen content relationships for the porous asphalt mix investigated shown in Figure 4.1 while Table 4.1 shows the linear regression parameters linking porosity and bitumen content.
4.2.1.1 Effect of Bitumen Modifier
The results also indicate that the porosity of the base bitumen and SBS mixes exceed 21% and 20%, respectively while subsequent increment of DAMA quantities causes a general drop in porosity. The porosity of SBS mix is higher than that of the DAMA mix but lower than the porosity of conventional mix as shown in Figure 4.1.
4.2.1.2 Effect of DAMA Quantities
From Table 4.1, the relationship between average porosity and bitumen content is linear with a correlation coefficient of more than 90% regardless of bitumen type used and the percentage of DAMA. The slopes of all lines are not significantly different. However, specimens with higher percentage of DAMA exhibits lower porosity. The results also indicate that the addition of DAMA at 0.5% interval causes a general drop in porosity. The porosities of the 0.5% to 1.5% DAMA mix grouping show a marked difference with mixes made with 2.0% DAMA. The VMA test results also exhibit similar trend.
Figure 4.1 Relationship between Porosity versus Binder Content
Table 4.1 Linear Equations and Regression Values For Porosity
Mix with Binder Type

Linear Equation

RSquare

70P

Y = 1.3x + 27.962

0.99

SBS

Y = 1.320x + 27.427

0.99

70P + 0.5%DAMA

Y = 1.317x + 26.668

0.99

70P + 1.0%DAMA

Y = 1.302x + 26.018

0.99

70P + 1.5%DAMA

Y = 1.305x + 25.480

0.99

70P + 2.0%DAMA

Y = 1.297x + 22.625

0.99

4.2.2 Mix Density
In dense mixes, the bitumen content can be optimized from the densitybitumen content relationship. The results of this investigation show otherwise. As reflected in Figure 4.2, the average mix density increases proportionally with bitumen content while Table 4.2 shows the parameters of the linear regression linking porosity and bitumen content. It is well known that porous mixes are characterized by its high voids contents. Any amount of bitumen added will fill up the voids and consequently results in higher mix densities. However, unlike dense mixes, as the bitumen content increases, the mix density also increases progressively since the binder will not be able to fill up all voids in the porous mix. Within the narrow range of bitumen content investigated, the average mix density and bitumen content are found to be linearly correlated with RSquare values exceeding 0.89.
4.2.2.1 Effect of Bitumen Modifier
The effect of using bitumen modifier is the increase in mix density. From Figure 4.2, mixes made with SBS exhibit a lower average mix density compared to mixes with DAMA but higher than 70P base bitumen mixes. Mixes made with SBS bitumen exhibit a higher average mix density compared to mixes prepared using 70P bitumen. The density of SBS mixes lies between the 0% DAMA and 0.5% DAMA mixes. The rate of increase in density with respect to binder content of all modified mixes is higher than the conventional mix. Among the modifier types, the SBS mix registers the lowest rate of increase.
4.2.2.2 Effect of DAMA Quantities
The results also showed that the mixes become denser as the DAMA content increases. For instance, at 2.0% DAMA, the mix density at 6% binder content is 1.979g/cm^{3} while at 0% and 1% DAMA, the corresponding density is 1.955 and 1.976g/cm^{3}, respectively. Hence, a 1% increase in DAMA content can cause an increase of mix density by 0.15%. At all binder contents investigated, the 0% DAMA content mixes record the lowest mix density. From the slope of the graph shown in Table 4.2, the rate of increase of mix density with respect to binder content also increases as the DAMA content increases. This rate of increase for mixes containing DAMA is also higher than the SBS mixes.
Figure 4.2 Relationship between Density versus Binder Content
Table 4.2 Linear Equations and Regression Values For Density
Mix with Binder Type

Linear Equation

RSquare

70P

Y = 0.0025x + 1.9406

0.96

SBS

Y = 0.0027x + 1.9523

0.97

70P + 0.5%DAMA

Y = 0.0029x + 1.9584

0.98

70P + 1.0%DAMA

Y = 0.0031x + 1.9579

0.98

70P + 1.5%DAMA

Y = 0.0031x + 1.9589

0.98

70P + 2.0%DAMA

Y = 0.0038x + 1.9566

0.98

4.2.3 Voids in Mineral Aggregate
VMA is important from the viewpoint of durability. Mixture with high VMA will promote binder hardening that could lead to pavement cracking and distress. On the other hand, mixes with deficient VMA will be prone to suffer from bleeding. This explains the general short service life of porous asphalt as compared to dense asphalts. The VMAbitumen content relationship for the porous asphalt investigated is shown in Figure 4.3. Within the short range of binder content investigated, it is obvious that the VMA increases proportionally with bitumen content.
4.2.3.1 Effect of Bitumen Modifier
The VMA is primarily a function of the aggregate gradation and of the shape and surface texture of the aggregate. When the VMA is too high, no amount of bitumen added will overfill the voids. The direct effect of adding modifier is a reduction in VMA. From Figure 4.3, VMA of SBS mixes lies between that of the 0% DAMA and 0.5% DAMA mixes.
4.2.3.2 Effect of DAMA Quantities
The control or 0% DAMA mix exhibits the highest VMA followed by SBS and other DAMA mixes in increasing DAMA contents. Figure 4.3 shows that the VMA for 2.0% DAMA and 1.5% DAMA mixes at 6% bitumen content are 26% and 29%, respectively. The 2.0% DAMA mix gives the similar VMA value to Hamzah (1995) investigated is in the region of 30%. From Table 4.3, the different percentage of DAMA mixes show RSquare values exceeding 0.99 which also shows other parameters of the linear equation.
Figure 4.3 Relationship between VMA versus Binder Content
Table 4.3 Linear Equations and Regression Values For VMA
Mix with Binder Type

Linear Equation

RSquare

70P

Y = 0.6167x + 27.949

0.95

SBS

Y = 0.642x + 27.351

0.99

70P + 0.5%DAMA

Y = 0.636x + 26.576

0.99

70P + 1.0%DAMA

Y = 0.638x + 25.99

0.99

70P + 1.5%DAMA

Y = 0.649x + 25.389

0.99

70P + 2.0%DAMA

Y = 0.647x + 22.6

0.99

4.2.4 Voids Filled with Bitumen
Voids Filled with Bitumen (VFB) is the percentage of the volume of the VMA that is filled with bitumen. The VFB is inversely related to the air voids. As the percentage of air voids approaches zero, the VFB approaches 100. Dense mixes are initially constructed to VFB values ranging between 70% to 85% (NAPA, 1991). In service, the mix undergoes traffic overcompaction and hence the VFB increases as the mix continues to densify under traffic. When the VFB exceed approximately 80% to 85%, the dense mix typically becomes unstable and rutting is likely to occur.
According to JKR specification (JKR, 1988), the VFB should range between 75% to 85%. This can be achieved by incorporating binder contents between 3.0%  5.0%. However, the local specifications do not specify limiting values for porous asphalt. From the results of the investigation shown in Figure 4.4, the VFB increases proportionally with bitumen content. Table 4.4 shows the parameters of the linear equation.
4.2.4.1 Effect of Bitumen Modifier
The general effect of modifier on mix properties is to increase their VFB. For instance from Figure 4.4, at 5.7% bitumen content, the VFB of the SBS mix and 70P bitumen mix are approximately 36% and 34.8%, respectively. Therefore, addition of SBS modified bitumen increases the VFB by 3.3%. However, the SBS mixes exhibit lower VFB compared to any of the DAMA mixes.
4.2.4.2 Effect of DAMA Quantities
Figure 4.4 illustrates the linear increase in VFB with bitumen content for all DAMA mixes. Generally, the VFB increases as the DAMA quantity increases. From the slopes of the linear equation given in Table 4.4, mix 2.0% DAMA is the most sensitive to variations in binder content.
Figure 4.4 Relationship between VFB versus Binder Content
Table 4.4 Linear Equations and Regression Values For VFB
Mix with Binder Type

Linear Equation

RSquare

70P

Y = 5.52x + 3.2227

0.99

SBS

Y = 5.7147x + 3.2461

0.99

70P + 0.5%DAMA

Y = 5.813x + 3.408

0.99

70P + 1.0%DAMA

Y = 5.876x + 3.731

0.99

70P + 1.5%DAMA

Y = 5.987x + 3.832

0.99

70P + 2.0%DAMA

Y = 6.511x + 4.926

0.99

4.3 Coefficient of Permeability
The coefficient of permeabilitybitumen content relationship for the porous asphalt studied is shown in Figure 4.5 which illustrates a general drop in permeability values as the bitumen content increases. It could also be seen that the effect of bitumen content on permeability can be significant.
4.3.1 Effect of Bitumen Modifier
From Figure 4.5, the net effect of using modifier is a reduction in permeability coefficient. All mixes tested that incorporates DAMA are less permeable than the SBS mixes. However, regardless of modifier type, the coefficient of permeability of all mixes reduces as the binder content increases. For instance, for SBS mixes, the permeability reduces from 0.285 to 0.136cm/s as the binder content increases from 4.4% to 5.9%. The extra binder added replaced the air voids causing blockage of some capillaries hence making it more difficult for the passage of water.
4.3.2 Effect of DAMA Quantities
From Figure 4.5, mix without DAMA is the most permeable compared to all other mixes but the general trend remains that the permeability reduces as bitumen content increases. It could be seen that bitumen content has a significant effect on permeability. For instance, for 1% DAMA mixes, the permeability drop from 0.179 to 0.129cm/sec as the binder content increases from 4.5% to 5.5% . This represents a permeability coefficient drop in the order of 27.9%. However, all mixes meet the Korean specifications for permeability (Darin Tech, 2001) which should not be less than 0.01 cm/s. The lowest 0.09 cm/s coefficient of permeability recorded is the 2.0% DAMA mix prepared at 6% binder content.
Figure 4.5 Relationship between Coefficient of Permeability versus
Binder Content
4.4 Marshall Stability Test Results
For dense mixes, it is possible to optimize the binder content from the Marshall stability results when the stability increases with increment in bitumen content up to a peak value beyond which further addition of bitumen will cause a general drop in stability. For porous mixes, as shown in Figure 4.6, this form of definite relationship is inconsistent. The variation in stability with binder content is small and indicates that the stability is not sensitive to variations in binder content. This is one of the reasons why the Marshall stability method could not be used to design porous mixes.
4.4.1 Effect of Bitumen Modifier
The effect of bitumen modifier on stability is shown in Figure 4.6. Clearly, the trend is quite inconsistent for majority of mixes. However, using modified bitumen has the positive effect of increasing stability. Generally, with the exception of 2.0% DAMA mix, the SBS mixes exhibit the highest stability compared to DAMA mixes. Nevertheless, all mixes indicated stability values more than the required 5 kN stipulated in the Korean specification (Darin Tech, 2001).
4.4.2 Effect of DAMA Quantities
The stability test results of all DAMA mixes can be compared from Figure 4.6. Higher DAMA percentages have the net effect of increasing mix stability. The average stability of conventional porous mix is 6.3 kN. However, with the addition of DAMA, the stability value can increase by as much as up to 1.4 times compared to mixes without DAMA. Nevertheless, the stability remains insensitive to binder content variations.
Figure 4.6 Stability Test Results
Table 4.5 The Marshall Properties and Marshall Quotient Test Results
Types of Binder

Binder Content (%)

Marshall Properties

Marshall Quotient (kN/mm)

Stability (kN)

Flow (mm)

70P

4.7

6.5

3.52

1.9

5.2

6.7

5.02

1.3

5.7

6.2

3.74

1.7

SBS

4.4

8.4

6.09

1.4

4.9

9.0

4.58

2.0

5.4

9.1

3.89

2.3

5.9

8.9

6.00

1.5

0.5% DAMA

4.5

7.4

3.91

1.9

5.0

6.9

2.94

2.4

5.5

7.2

5.09

1.4

6.0

7.0

3.82

1.8

1.0% DAMA

4.5

8.2

4.93

1.7

5.0

7.1

3.27

2.2

5.5

8.0

4.61

1.7

6.0

7.8

4.19

1.9

1.5% DAMA

4.5

8.5

3.61

2.4

5.0

8.3

4.05

2.0

5.5

8.4

3.64

2.3

6.0

8.6

3.57

2.4

2.0% DAMA

4.5

9.1

6.00

1.5

5.0

8.7

3.36

2.6

5.5

9.4

6.59

1.4

6.0

9.0

5.90

1.5

4.5 Cantabrian Test Results
The resistance of compacted porous mixes specimen to abrasion loss is analyzed by means of the Cantabrian test. This combined abrasion and impact test is carried out in the Los Angeles Abrasion machine in accordance to the procedure outlined in ASTM Method C131 (ASTM, 1999). The abrasion loss is expressed in terms of the percentage mass loss compared to the original mass as discussed in Chapter 3.
Figure 4.7 shows the relationship between abrasion loss and bitumen content obtained from the Cantabrian test carried out on all mixes investigated. The amount of abrasion loss indicates the interaggregate particle cohesion loss in the porous mixes tested. This cohesion is provided by the adhesive forces present in the bitumen film that coats every aggregate particle. The lower the abrasion loss, the less prone are the mixes to disintegration. The results also indicate that the resistance to disintegration is highly sensitive to variations in bitumen content, particularly in the lower range of bitumen content.
4.5.1 Effect of Bitumen Modifier
Resistance to abrasion usually improves with an increase in bitumen content. However, this resistance is also related to the rheological properties of the bitumen. Porous mixes containing unmodified bitumen generally exhibits less resistance to abrasion than mixes containing modified bitumen. Figure 4.7 shows that the SBS mix is the most resistant to abrasion compared to conventional and DAMA mixes. At 4.0% binder content, the abrasion loss of SBS mix is 20% while the equivalent value for base bitumen mix is 37.8%. At 5.0% bitumen content for all types of mixes, the abrasion loss is below 10% and the only exception is the 70P bitumen mix which experiences an 11.3% abrasion loss.
4.5.2 Effect of DAMA Quantities
From the general trend shown in Figure 4.7, DAMA mixes become more resistant to abrasion loss as the DAMA quantity increases. Thus, mixes with 0.5% DAMA exhibit the highest abrasion loss while the 2.0% DAMA mixes perform the best. From the resistance to disintegration perspective, addition of DAMA in higher quantities can increase resistance to abrasion and the subsequent mix durability. At 4.0% bitumen content, the abrasion loss of the 70P bitumen and 0.5% DAMA mixes are respectively 37.8% and 30.8%. This means addition of a mere 0.5% DAMA to porous mixes via the dry process can reduce the percentage of abrasion loss up to 7%.
Figure 4.7 Cantabrian Test Results
4.6 Binder Drainage Test Results
The binder drainage test results, expressed in terms of retained binder versus mixed binder are shown in Figure 4.8. In the absence of drainage, the graph will be linear when the retained binder equals the mixed binder. Each graph also exhibits a distinct peak beyond which further addition of bitumen will cause a drop in retained binder when the binder drainage becomes more serious.
4.6.1 Effect of Bitumen Modifier
Figure 4.8 illustrates the significant role played by the SBS and DAMA modifiers as antidraining agents. Use of SBS mixes and additives help to mitigate the problem of binder drainage compared to base bitumen 70P mix. Figure 4.8 also shows that the retained binder of SBS mixes lie between the 1.0% DAMA and 1.5% DAMA mixes. It can be concluded that the cohesion of SBS mixes is probably equal to the range of 1.0% up to 1.5% DAMA mixes. At 6.5% mixed binder, the retained binders of the SBS mixes are 12% higher than that of conventional mixes.
4.6.2 Effect of DAMA Quantities
Figure 4.8 illustrates the increment in retained binder as the DAMA content increases. Hence, the problem of binder runoff can be mitigated by an increment in DAMA content. When the DAMA content increase from 0.5 to 2.0%, the maximum retained binder content that takes place at 6.1% increases by 6.5%. From Section 6.4, addition of DAMA also has the effect of increasing the target binder content.
Figure 4.8 Binder Drainage Test Results
4.7 Resilient Modulus Test Results
The resilient modulus is simply the modulus of elasticity when the asphalt sample is loaded within its elastic range where the deformation is fully recoverable. It is defined as the ratio of the applied stress to the recoverable strain when a dynamic load is applied. The results of the resilient modulus test for mixes prepared at varying binder contents are shown in Figure 4.9. For all mixes, modified and unmodified bitumen exhibits similar pattern. Table 4.6 shows the mean resilient modulus of five pulse with the various binder types and binder contents. Each value is the average reading of two samples.
4.7.1 Effect of Bitumen Modifier
Figure 4.9 exhibits a distinct peak beyond which further addition of bitumen will cause a drop in modulus except 70P bitumen. The resilient modulus for SBS mixes is between that of the 1.0 and 1.5% DAMA mixes. At 4.9% binder content, the SBS and 1.5% DAMA mixes give the same value of modulus, which is 5160 MPa. Nevertheless, the resilient modulus of the SBS mixes is twice that of the conventional mixes.
4.7.2 Effect of DAMA Quantities
A comparison between graphs shown in Figure 4.9 indicates that in general, addition of DAMA causes an increase in the resilient modulus. The results also indicate that the increment of bitumen content causes an increase in the resilient modulus up to a binder content of 5.0% to 5.5% beyond which the resilient modulus starts to decrease. The mix without DAMA does not exhibit this pattern. However, addition of DAMA in higher quantities can increase the resilient modulus value compared to conventional mix. Within the limited DAMA content tested, the resilient modulus increases as the DAMA content increases.
Figure 4.9 Resilient Modulus Test Results
Table 4.6 Mean Resilient Modulus For Types of Binder
Types of Binder

Binder Content (%)

Mean Resilient Modulus at Pulse Number (MPa)

Mean Resilient Modulus per Sample (MPa)

1

2

3

4

5

70P

4.7

2110

2237

2246

2140

2217

2190

5.2

2118

2256

2430

2360

2421

2317

5.7

2460

2268

2478

2412

2242

2372

SBS

4.4

4212

4109

4256

4115

4218

4182

4.9

5210

5120

5170

5220

5080

5160

5.4

5087

4990

4960

5012

5001

5010

5.9

4760

4790

4880

4814

4856

4820

70P + 0.5%DAMA

4.5

3760

4287

4294

2399

4221

3792

5.0

4963

4745

4710

4558

4469

4689

5.5

4577

4620

4322

4339

4270

4426

6.0

3187

3073

3072

2933

3050

3063

70P + 1.0%DAMA

4.5

4179

4036

4332

4195

4114

4171

5.0

5387

5183

5028

4969

5065

5126

5.5

4459

4483

4499

4558

4395

4479

6.0

3920

3861

3723

3749

3630

3777

70P + 1.5%DAMA

4.5

4831

4617

4757

4776

4776

4751

5.0

5170

5272

5160

5220

5263

5217

5.5

5279

5147

4937

5127

5355

5169

6.0

4571

4530

4612

4594

4577

4577

70P + 2.0%DAMA

4.5

4524

4762

4798

4834

4909

4766

5.0

5344

5365

5487

5487

5198

5367

5.5

6079

6104

6281

5843

6128

6087

6.0

5868

5778

5669

5823

5464

5720

4.11 Comparison of Results
4.11.1 Comparison of Porosity Test Results
According to the DoT (1993), the porosity of porous mix incorporating binder contents between 3.4% to 5.0% ranges between 17% to 26%. In this investigation, the porosity ranges between 14.8% to 21.9% at binder contents ranging from 4.4 to 6.0%. The relationship between porosity and binder content is linear for all mixes. This linear relationship is in agreement with the results published by Ruiz et al. (1990) and Bucchi and Arcangeli (1990). Hamzah (1995) obtained the porosity ranges between 16.9% to 17.7% respectively for mixes prepared using two slightly different gradations. These results have shown the linear relationship between porosity and binder content. This is because the porosity value is very much influenced by the binder content.
4.11.2 Comparison of Marshall Stability Test Results
In this study, the Marshall stability values range between 6.2 kN to 9 kN for 70P bitumen and 2.0% DAMA mixes, respectively. The addition of DAMA with 0.5% interval from 0.5% to 2.0% DAMA reveals that the stability value can increase proportionally. The increasing of bitumen content does not show any constant trend in the stability result. This is because, a unique correlation between stability and bitumen content does not exist.
This finding is consistent with the results obtained by other researchers (Smith et al. 1974, Hamzah 1995 and Peter 2003). In another study carried out by Walsh et al. (2004), the dense mix with 5% bitumen content incorporating modified and unmodified bitumen were 16.1 kN and 12.8 kN, respectively. In Korea, the stability result for porous mixes with and without DAMA were 10.5 kN and 4.0 kN as reported by Darin Tech (2001). These results indicated that the modified binder can be improved the stability value of the mixes. However, when compared to dense mix, the value of Marshall stability of porous asphalt is significantly lower.
4.11.3 Comparison of Resilient Modulus Test Results
In this study, the resilient modulus mixes incorporating 4.4% modified and 4.7% unmodified binder were 4182 MPa and 2190 MPa, respectively. According to Walsh et al. (2004), the resilient test result for mixes incorporating 5.4% modified binder and 5.6% unmodified binder were 5500 MPa and 1680 MPa, respectively. The mixes using modified binder achieved 69% more stability than mixes using unmodified binder and the result shows the same trend for other bitumen contents. Tayfur (2004) also obtained the same pattern where the resilient modulus for modified mixes was higher than unmodified mixes. The difference between the effect of modified and unmodified bitumen was quite evident with SBS and conventional mixes with a difference 4300 MPa and 1000 MPa, respectively. From the above deliberation, it is clear that the use of modified bitumen increases the mix resilient modulus.
In another investigation, the resilient modulus of the specimen prepared with modified and unmodified bitumen were 2443.5 MPa and 1677.5 MPa respectively (Clifford et al., 1996). Darin Tech (2001) found that the resilient values for mixes with and without DAMA were obviously different. The 1.0% DAMA mixes gave the value of 2935 MPa whereas the mixes of 0% DAMA gave the lowest value at 889 MPa.
4.11.4 Comparison of Dynamic Creep Test Results
The highest value for dynamic creep stiffness in this investigation is 30.9 MPa achieved with SBS mixes at 5.9% bitumen content while the lowest value is 8.3 MPa achieved with 70P mixes at 4.7% bitumen content. Although the lowest bitumen content for SBS mixes is 4.4%, the stiffness value is still higher than 70P mixes with 5.7% bitumen content. This indicates that the SBS mixes exhibit higher stiffness modulus value when compared with unmodified mixes. The strain value of mixes with SBS at 4.4% and 5.0% bitumen contents are 0.0051 mm and 0.0041 mm, respectively. In one investigation by (Brennan et al. 1996), the average strain value for dynamic creep stiffness test result conducted in a similar manner on SBS modified porous mixes was 0.0042. The strain value obtained is comparable with the tests results obtained in this study.

Summary
The engineering properties and performance of porous asphalt investigated are evaluated by the appropriate laboratory tests. The analysis of results obtained from these tests were carried out according to binder types and percentage of DAMA incorporated in the porous mixes. Furthermore, comparison between results from this study and other works are carried out for greater visualization of the performance of the mixes produced. The results indicate an improvement on the engineering properties and performance with the addition of DAMA. For instance, the resilient modulus value of sample prepared at 4.5% bitumen content increases by 20% as the DAMA content increases from 0.5% to 2.0%. The dynamic creep stiffness also increases by 35% as the DAMA content increases from 0.5% to 2.0%. From the study of varying DAMA concentrations, mixes with 2% DAMA contents exhibits the most favourable stability, resilient modulus, creep and rut depth. However, mixes incorporating 2% DAMA exhibits porosity less than 20% and low permeability values which are very important parameters in open mixes. On this basis, the mix with 1% DAMA is recommended for use in field construction of porous pavements.
