Aggregate Gradation Variation on the Properties of Asphalt Mixtures

Abstract

National highway projects present a crucial role in economic growth, as they have a great influence on the national income. Therefore, the decision makers plan to construct these projects at a rapid rate. To achieve the just aforesaid, the utilization of asphalt of adequate quality and gradation is essential. The key problem which lies in recent decades is that many types of asphalt mixtures are rejected and reconstructed in the site due to the gradation variation of aggregates in the asphalt mixture which waste raw materials, cost and time. Thus, this research seeks to assess the possibility of accepting asphalt mixes with aggregates gradation variation (within the range from +4% above the upper specification limit to −2% below the lower specification limit). A wearing surface mix with gradation 3B was prepared according to the Egyptian code. The gradation variation was presented as the aggregate gradation is out of the specification limits during Hot Mix Asphalt (HMA) production. The aggregate gradations lie above and below the upper and lower specification limits, respectively, by ±2%, ±4% and ±6%. The design gradation of the control mix was included as a reference case. The different mix properties were measured using the Marshall Mix design method. Then, the performance of HMA mixes was evaluated under the effects of high temperature and water cycles through applying wheel loading tracking and Indirect Tensile Strength (ITS) tests. The results show that the 3B mixes with a gradation within a range of +4% to −2% of the upper and lower specification limits recorded the lowest rutting depth and the highest water damage resistance in hot regions compared to ordinary asphalt mixes. In summary, the new aggregate gradation limits will provide a reference for the design of asphalt mixture in hot climate regions.

1. Introduction

National highways are the basic axes of economic development and growth and bring important social benefits. They are of vital importance in order to make a nation grow and develop. In addition, providing access to employment, social, health and education services makes a road network crucial to fighting against poverty [1]. Thus, the government has been motivated to finish these projects rapidly to reflect their performance. However, the rapid completion of such projects has a negative effect on the final product due to the aggregate gradation variation that occurs at execution and production processes. On the other hand, the construction process of asphalt pavement typically includes mixing, transportation, paving and compaction stages [2]. There are many details in each process that are difficult to accurately control, which results in the presence of uncertain factors in the construction processes [3,4]. Indeed, it is a well-documented fact that inconsistency in asphalt design and construction parameters has consistently caused premature performance failures [5]. Therefore, exploring the negative implications associated with variability in asphalt mix design and construction parameters may be worthwhile [6].

Asphalt mixtures used in modern pavement construction are mainly comprised of air void, asphalt binder, coarse aggregates, fine aggregates and filler [7,8,9]. Asphalt binder cements aggregates, and filler to form a multiphase composite with air voids [10]. To ensure a sound pavement performance, the skeletal structure should consist of well-interlocked aggregates [11]. As an essential parameter for asphalt mixture, the gradation is closely related to the quality and performance of the pavement [12]. Gradation variation in the asphalt mixture may result in many distresses of pavement [13].

Many types of asphalt mixtures are rejected and reconstructed in the site due to the gradation variation of aggregates in the asphalt mixture, which waste raw materials, cost and time [14]. Thus, this study seeks to evaluate the possibility of accepting mixes with aggregates gradation variation (within the range from +4% above the upper specification limit to −2% below the lower specification limit).

The pavement design and construction literature shows that many studies on the effect of the gradation variation on HMA properties, or rather the effect of variance of gradation that may occur during production on HMA properties have been undertaken in the past [15,16,17,18,19,20].

Zhang et al. [15] evaluated the aggregate gradation in asphalt mixtures on the characterize load carrying capacity and rutting resistance. Zhang recorded that the aggregates retaining on sieve sizes of 2.36 and 4.75 mm provide more than 50% contribution to resist load and rutting, and the aggregates retaining on sieve sizes of 1.18, 0.6 and 0.3 mm provide more than 50% contribution to strength the structure. The Influence of Aggregate Gradation on Clogging Characteristics of Porous Asphalt Mixtures was evaluated by martin et al. [16], results revealed that aggregate gradation was found to be strongly correlated to the macrotexture depth of the porous pavement and the permeability of the mixes both before and after clogging. Abo-Qudais et al. [17] study the Effect of aggregate properties on asphalt mixtures stripping and creep behavior and results documented that unconditioned HMA specimens prepared using basalt aggregate resist creep better than those prepared using limestone. However, after conditioning, mixes prepared using basalt were less resistant to creep strain than those prepared using limestone aggregate. Percent absorbed asphalt was found to be directly related to stripping resistant. Also, mixes prepared using aggregate following American Society for Testing and Materials (ASTM) upper limit of dense aggregate gradation presented the highest resistance to stripping. Moghaddam et al. [18] Study the effects of using different types of additives and aggregate gradation on fatigue and rutting resistance of Asphalt Concrete (AC) mixtures and it is concluded that that fatigue and rutting resistance of AC mixture could be enhanced considerably by utilization of different aggregate gradation and types of additives such as fibers. Bazi [19] reported that variability of gradation has a considerable effect on the performance of pavement. To enhance resistance to asphalt pavement rutting, Lv et al. [20] was of the opinion that adjusting the percentage of particles passing the 4.75 mm sieve should be considered. Amir Golalipour et al. [21] examined the effect of aggregate gradation variation on rutting characteristics of asphalt mixture and concluded that the aggregate gradation played a considerable role in resisting the permanent deformation of pavement. Yu et al. [22] investigated the influence of the aggregate gradation during construction on the asphalt pavement performance and reported that the rutting resistance increased first and then decreased with regard to the changing of the gradation from fine to coarse. In research prepared by Moustafa et al. [23], performed a laboratory study to evaluate the performance of asphalt mixtures using a composite mixture of diatomite powder and lignin fiber. The results showed that the addition of diatomite powder and lignin fiber are greatly enhanced the overall performance of asphalt mixes, and the compound mixture was more effective for improving the overall asphalt performance than either lignin fiber or diatomite powder separately. Elliott et al. [24] reported that the mixtures with lower limit gradations almost had the greatest effect on HMA properties. Moreover, they found that asphalt mixtures gradations deflected to lower limit gradation showed the best performance. Awan et al. [25] applied Multi Expression Programming to predict the output parameters Marshall Flow and Marshall Stability of Asphalt Pavements, it is demonstrated that the novelette models have produced results that are consistent with the experimental data and function equally well for unknown data as well as The models developed have successfully incorporated input parameters and have the capability to predict the trends of MS and MF for flexible pavements, as revealed from the parametric study. Rafiq et al. [26] conducted a comparative comparison among Hot Mix Asphalt (HMA) and Reclaimed Asphalt Pavement (RAP) through using Life Cycle Cost (LCC) and the results denoted that the total LCC measurement, a total of 14% cost reduction was reported using RAP as compared to HMA. Moreover, the two materials (HMA and RAP) are manufactured in different types of manufacturing plants. Thus, in analyzing the cost difference between the two chosen manufacturing plants for virgin materials and RAP, a total of 57% cost reduction was observed for a RAP manufacturing plant. Besides this, no cost difference was observed in the rest of the phases, such as manpower, materials transportation, and construction activities, as the same procedures and types of machinery are used.

Accordingly, this study is concerned with the aggregate gradation variation that occurred to HMA. Limited studies have been conducted to determine how far to deviate beyond the upper and lower specification limits of aggregate gradations to achieve an acceptable performance level for HMA mixtures. Therefore, this study investigated the performance of HMA mixtures in the case of the aggregate gradation variation beyond the upper and lower specification limits.

The main objectives of this study are:

• To establish deviations from the upper and lower specification limits of the aggregate gradation curve that are not inimical to the performance of HMA mixtures. Thereafter, the properties of HMA mixtures prepared thereof would also be evaluated.
• Relative to the control mix, the performance of the best HMA mix that lies above the upper specification limit and the best mix beneath the lower specification limit would be compared via wheel tracking and indirect tensile strength tests.
• Determination of the HMA mixture with an optimized skeletal structure capable of providing enhanced resistance to high temperature induced rutting distress and low temperature engendered thermal cracking.

2. Materials and Methods

2.1. Materials

Coarse and Fine aggregates, with particles either retained or passing, respectively, on the 4.75 mm (No. 4) sieve were used. Similarly, aggregate particles finer than the 0.075 mm (No. 200) sieve served as fillers. All the aggregates were uniformly graded, conforming to the accepted gradation limits specified by the Egyptian Code [27]. Two types of coarse dolomite, grade (I) and grade (II) with the physical properties shown in

Table 1. Properties of coarse aggregate.

Test No.	Test	Designation No.	Values		Specification Limits
			Type 1 (Grade I)	Type 1 (Grade II)
1	Bulk specific gravity (gm/cm3)	AASHTO T-85	2.52	2.5	N/A
	Saturated surface dry specific gravity (gm/cm3)		2.59	2.57
	Apparent specific gravity (gm/cm3)		2.7	2.6
2	absorption %	AASHTO T-85	2.56	2.67	≤5
3	Los angles abrasion test %	AASHTO T-96	22.2	24.5	≤40
4	Stripping Test %	AASHTO T-182	>95	>95	≥95

were also used as components in the asphalt. Fine siliceous sand with a bulk specific gravity of 2.65 g/cm³ and limestone dust with a bulk specific gravity 2.85 g/cm³ (filler) were used as asphalt concrete ingredients. The two types of coarse aggregate were obtained from the “ATAKA” quarry, Suez, Eygt, where the natural siliceous sand obtained from “ELREHAB” quarry, Cairo, Eygt. The binder used is Suez asphalt cement with 60–70 penetration grades, and physical properties shown in

Table 2. Physical properties of asphalt binder.

Test No.	Test	Designation No.	Values	Specification Limits
1	Penetration test (0.1 mm)	T-49	62	60–70
2	Softening Point (°C)	T-53	52	45–55
3	Flash point (°C)	T-48	+250	+250
4	Kinematics viscosity (cSt)	T-201	395	+320
5	Ductility (cm)	T-51	+100	≥ 95

.

2.2. Mix Gradation

The asphaltic wearing surface mix (mix 3B) tested in this study consists of 30% grade I coarse aggregate, 20% grade II coarse aggregate, 15% natural sand, 30% crushed sand and 5% limestone dust as a mineral filler. The design gradation of the control asphalt concrete mix (M0) was presented in

Table 3. Different gradations of asphalt concrete mixes.

Sieve Size	3B Asphalt Concrete Mixes
	M0	M1	M2	M3	M4	M5	M6	Specification Limits
1″	100	100	100	100	98	96	94	100–100
3/4″	100	100	100	100	98	96	94	100–100
1/2″	85	100	100	100	74	72	69	75–100
3/8″	73	100	100	100	59	58	55	60–85
No. 4	48	66	77	88	34	33	32	35–55
No. 8	29	42	49	56	20	19	18	20–35
No. 30	18	26	31	35	10	9	9	10–22
No. 50	11	19	22	26	6	5	5	6–16
No. 100	8	14	17	19	4	3	3	4–12
No. 200	5	10	11	13	2	2	2	2–8

. The design gradation of the control asphalt concrete mix (M0) is shown in Table 3, and the design gradation lies within the limits of Egyptian standard specifications for wearing surface (mix 3B) [27]. To highlight the different phases obtainable during asphalt production, aggregate gradation of mixtures was also extended (1) above and (2) below the upper and lower specification limits of the Egyptian standard specifications for wearing surface (mix 3B) [27]. For the three mixes with gradations lying above the upper specification limit, the deviations were +2% (M1), +4% (M2) and +6% (M3). Conversely, the gradations of the three mixes below the lower specification limit by −2%, −4% and −6% were denoted as M4, M5, and M6, respectively.

3. Experimental Work

The test program involves four stages, as displayed in Figure 1.

• In the first stage, after selection materials, five types of aggregate characterization tests will be applied; Los Angeles abrasion, water absorption, specific gravity, stripping value and selection of design gradation to attain the condition of stone-on-stone contact were undertaken. As well as the chosen asphalt binder will be evaluated through applying penetration, softening point, flash point, viscosity and ductility tests.
• In the second stage, the control asphalt mix (M0) and the other five asphalt mixes (from M2 to M6) were designed and prepared according to the Egyptian code [16].
• In the third stage, the effect of the aggregate gradation variation on HMA properties was examined based on the Marshall Mix Design Method.
• In the final stage, wheel loading tracking and indirect tensile strength tests were conducted on the M0 and samples obtained from the best mixes lying above and below the upper specification limit, and the lower specification limit, respectively.

3.1. Marshall Test

The resistance of asphalt mixtures to plastic flow was determined in the current test by measuring the stability (kg) and flow values (mm) for each mix by using the Marshall apparatus, Marshall machine model TO-550-1 which obtained from USA. Marshall Test was performed using three 10 cm × 6 cm cylindrical specimens from each mix. After 24 h of specimen compaction, they were immersed in a water bath at 60 °C for 30 min before the Marshall test commenced. The test criterion used was the AASHTO T-245, 75-blow Marshall Compaction [28], highlighted in Figure 2 and Figure 3. For the stability component of the Marshall test, all specimens were weighed in air and submerged in water, with the maximum load (kg) designated as the stability value. Note that a correction factor was implemented for specimens with thickness differing from 6 cm. On the other hand, the deformation undergone by specimens during loading to the peak value was measured using the flow meter and reported as the flow value.

3.2. Wheel Loading Tracking Test

This test investigates the capability of a pavement to withstand rutting, and was performed according to the Egyptian Code specifications. Figure 4 shows the wheel tracking test machine used, which manufactured in Egypt according to BS-EN 12697-22 specifications. One slab measuring 440 mm × 330 mm × 50 mm according to the LTG 2015 [29,30] was prepared for each mixture, and tested at 60 °C under a wheel load of 53.5 kg. The track depth was recorded at regular intervals up to 45 min using a spring-less dial gauge. A description of the test setup is shown in Figure 5.

3.3. Indirect Tensile Strength

Tensile characteristics of bitumen mixtures were determined according to AASHTO T-283 test method, by loading the Marshall specimen along with its diametric plan with a constant rate, producing uniform stress [31]. In this test, two sets of specimens from each mixture (the control mix, the best mix that lies above the upper specification limit and the best mix that lies below the lower specification limit) were fabricated and evaluated. While one set of specimens was conditioned by soaking in water at 60 °C for 24 h, the other set was left unconditioned. The ratio of the average indirect tensile strength of the conditioned samples to the average indirect tensile strength of the unconditioned specimens was recorded as the Tensile Strength Ratio (TSR).

4. Results

4.1. Marshall Test Results

Marshall Test was conducted to determine the OAC for each mix, as reported in

Table 4. Effect of aggregate gradation variation on the investigated mixes.

	Mixes	M0	M1	M2	M3	M4	M5	M6	Specification Limits *
Properties
% OAC		4.5	4.3	4.0	3.5	5.0	5.2	5.5	3%–6%
Stability (Kg)		1203	1182	1175	985	1306	1318	1334	900 kg (min)
Flow (mm)		3.2	3.6	4.0	5.2	2.8	1.9	1.6	2–4 mm
Stiffness (kg/mm)		376	328	294	189	466	694	834	300–500 kg/mm
Bulk specific gravity (Gmb) (gm/cm3)		2.345	2.358	2.361	2.350	2.305	2.263	2.245	--
% Air voids in total mix (Va)		3.54	4.35	4.22	6.0	4.44	5.92	7.66	3%–5%
% Air voids in mineral aggregate (VMA)		16.44	15.8	15.43	15.39	18.3	19.96	20.84	--
% Air voids filled with asphalt (VFA)		87.5	72.5	72.7	61.4	75.7	65.3	63.2	--

* Egyptian code of practice (ECP 104) limits [27].

. Then evaluate the properties of the control mix and other mixes at different gradations. Different gradations showed different views of the aggregate gradation variation during asphalt production. The mixes of M0 through M6 for 3B mix showed this difference of gradations as shown in Figure 6. Where M0 presented the mix applied at the design gradation curve (control mix). M1, M2 and M3 presented the mixes applied at +2%, +4% and +6%, respectively, above the upper specification limit. M4, M5 and M6 presented the mixes applied at −2%, −4% and −6%, respectively, below the lower specification limit. The previous mixes of M0 through M6 differed at the OAC. The Marshall properties (stability, flow, bulk specific gravity, air voids, voids in mineral aggregate and voids filled with asphalt) were measured for all the mixes studied as shown in Table 4. Then the data of the results would be collected and analyzed.

According to Table 4, the aggregate gradation variation above the upper specification limit had a negative effect on the mix stability. The stability value for the control mix (M0) was 1203 Kg. The following two mixes (M1 and M2), in which the gradation variation was increased to +2% and +4% above the upper specification limit respectively, the mix stability almost was not affected as it decreased by 1.7% and 2.3%. For the following mix (M3), as the gradation was increased to +6% above the upper specification limit, the mix stability decreased by 18%, reaching its lowest value of 985 kg but remained higher than the minimum stability value (900 kg).

Concerning values of flow for 3B mixes that lie above the upper specification limit, the flow value for M0 was 3.2 mm. For the following two mixes (M1 and M2), as the gradation variation was increased above the upper specification limit by 2% and 4% respectively, the flow increased by 12.5% and 20% but still comply with the specification (2 mm ≤ Flow ≤ 4 mm). Mix (M3) lies beyond the specification requirement for flow, as its flow was (5.2 mm).

Values of mix stability for 3B mixes that lie below the lower specification limit are displayed in Table 4. The mix stability for the following three mixes (M4, M5, and M6) increased by 8.5%, 9.6%, and 10% respectively as the gradation variation was below the lower specification limit by −2%, −4% and −6%, achieving stability values of 1306 kg at −2%, 1318 kg at −4%, and 1334 kg at −6% below the lower specification limit.

Based on the results in Table 4 show values of flow for 3B mixes that lie below the lower specification limit. The flow value for M0 was 3.2 mm. For the following mix (M4), in which the gradation was decreased to −2% below the lower specification limit, the flow decreased, reaching 2.8 mm, but remained higher than the minimum flow value (2 mm). This value decreased the flow by 12.5% compared with the control mix (M0). The following two mixes (M5 and M6) gave flow values beyond the specification value for flow (> 4 mm) according to the Egyptian code.

4.2. Performance Evaluation Tests

As M2 mix was the highest diffracted mix above the lower specification limit for 3B mixes gradation that comply with the specification requirements and M4 was the highest diffracted mix below the lower specification limit for 3B mixes gradation that comply with the specification requirements, M0, M2, and M4 were chosen to conduct to the performance evaluation tests (wheel load tracking and indirect tensile test).

4.2.1. Wheel Loading Tracking Test Results

Figure 7 presents the rutting depth test results of the three mixes; M0 (Control), M2 (the best mix that lies at +4% above the upper specification limit) and M4 (the best mix that lies at −2% below the lower specification limit). Figure 7 shows a rutting value of 3.92 mm for M0 mix, 4.85 mm for M2 mix, and 4.34 mm for M4 mix. Relative to the M0, the increase in rutting depth was 18% for the M2 and 9.5% for the M4. This implies that compared to the mix M2, the mix M4 had a superior rutting resistance. It is attributed to the increase in stability by 11.1% and the reduction in flow by 30% for M4 when compared with M2.

Accordingly, it could be concluded that the 3B mix at −2% below the lower gradation limit showed high resistance to rutting phenomena compared with the 3B mix at +4% above the upper gradation limit.

4.2.2. Indirect Tensile Strength Test

The TSR results of the three mixes are shown in Figure 8. The results shown in Figure 8 indicate that while the TSR for the mix M0 was 83.87%, a value of 81.62% was recorded for the mix M2. Compared to the mix M0, this represents about 2.7% reduction in the TSR. Figure 8 also shows that at a TSR value of 82.84%, the TSR of the mix M4 was slightly better than that of the mix M2. The main inference from these results is that comparing mix M2 and mix M4, the latter possesses a better resistance to moisture induced damage.

Accordingly, it could be concluded that the 3B mix at −2% below the lower gradation limit showed good resistance to moisture damage phenomena compared with the 3B mix at +4% above the upper gradation limit.

5. Discussion

This section aims to introduce the analysis of the main results of physical properties of six types of asphalt mixes, as well as the pavement performance under high temperature and water cycles effects.

5.1. Analysis of Marshall Test Results

Based on the results of the Marshall test mentioned in Table 4 and Figure 6, it can be concluded that the aggregate gradation variation over the upper specification limit causes an increased in the mixture stability and decreased the flow. On the contrast, the variation of aggregate gradation lower than specification limit result in decreased the stability and increased the flow of mixtures.

The variation in aggregate gradation above the upper specification limit leads to a damaging influence on the mixture stability and it may be due to the excessive asphalt in the mix due to the 30% reduction in the voids filled with asphalt (VFA) as a result to the fineness of the stone matrix. On contract, the flow of mixtures which prepared above the upper specification limit was increased above the upper specification limit. The increase in flow is attributed to the excessive asphalt in the mix as it is the same reason of stability reduction.

The aggregate gradation variation under the lower specification limit leads to a significant effect on the stability of mixes. This increase in stability is attributed to the adequate asphalt content in the mixes due to the increase in the voids in stone matrix (VMA) so the voids filled with asphalt (VFA) included the entire content of the asphalt. The flow values of mixes that lie under the lower specification limit was decreased, and it is may be because of the absence of excessive asphalt in the mixes and the reduction in fines in the stone matrix.

5.2. Analysis of Pavement Performance

5.2.1. Wheel Loading Tracking Test

The outcomes of wheel loading tracking test revealed that the permanent deformation of M2 and M4 mix is slightly increased comparing with Control asphalt mixture.

The main reason for increasing the rutting depth for M2 and M4 mix is that the total air voids among the skeleton of aggregate is increased which lead to increase the bitumen absorbed by the mixes, thus the rutting depth of M2 and M4 mixes is slightly increased.

5.2.2. Indirect Tensile Test

The indirect tensile strength results from the soaking group in water and unconditional group, and the tensile strength ratio (TSR) are displayed in Figure 8. Results observed that the moisture damage resistance of M2 and M4 was increased as opposed to Control asphalt Mixture.

It may be due to the fineness of its stone matrix, that led to an increase in the cohesion between the stone matrix and the low asphalt content that was affected harmfully by the elevated temperature of the water path. The tensile strength of asphalt mixes was increased, accordingly; as well as, the increase of adhesion force among asphalt mastic and aggregate led to the improvement of the anti-shear strength and rutting of the asphalt mixes.

Generally, based on this discussion, it could be concluded that the 3B mix at −2% below the lower gradation limit presented a good performance against moisture induced damage compared with the 3B mix at +4% above the upper gradation limit.

6. Conclusions

This paper presented a novel aggregate gradation variation with range (2%, 4%, and 6%) from the upper and lower specification limits of aggregate. The HMA mixtures were prepared and performed in laboratory according to the Egyptian Code specifications, afterward Marshall, wheel loading tracking and Indirect Tensile Strength (ITS) tests were applied in order to evaluate the HMA properties, high temperature performance, and water stability of different asphalt mixes, respectively. Based on the research outcomes, the following conclusions can be drawn.

Both of M2 (aggregate gradation within +4% the upper specification limit) and M4 (aggregate gradation within −2% the lower specification limit) significantly enhances the asphalt mixes performance. M2 mix has clearly enhanced the water stability performance, but the enhancement of rutting resistance is limited. M4 mix has a great effect on improving both of high temperature performance and water damage resistance. The rutting resistance of the mix M4 was superior to that of the mix M2. The resistance of mix M4 to moisture damage was higher than that of the mix M2. In summary, compared to the mix M2, the capacity of the mix M4 to enhance the service life and ride quality of pavements is determined to be higher.

7. Recommendation and Future Works

The following main recommendations can be outlined:

• For cold climates, the study should be performed and evaluated.
• Study the influence of OAC variation with the aggregate gradation variation on the performance of asphalt mixes.
• The economic considerations should be studied extensively regarding the possibility of accepting down to −2% variation in 3B mix gradation below the lower specification limits.