Laboratory Investigation of Open Graded Friction Courses with Different Binder Grades and Aggregate Gradations

Abstract

This study investigates the effectiveness of the volumetric mix design procedure used by the New Mexico Department of Transportation based only on the tensile strength ratio (TSR) test for designing open graded friction course (OGFC) materials. It also evaluates the effects of different binders (or binder grades) and aggregate gradations on the performance of OGFC materials. Nine OGFC mixes were produced using three different binders (PG 64-28+, PG 70-28+, and PG 76-22+) and three different aggregate gradations (two of 9.5 mm nominal maximum aggregate size [NMAS] and one of 12.5 mm NMAS). Next, a thorough laboratory investigation was conducted to evaluate their various performances. This study found that the traditional volumetric mix design based solely on the TSR test cannot ensure adequate durability performance. The PG 64-28+ OGFCs, for example, met all the volumetric and TSR requirements during mix design but performed poorly in all performance tests. This suggests that additional performance testing for OGFCs is required to ensure adequate longevity. Cantabro abrasion and semi-circular bending tests are found to be useful in characterizing the durability and moisture conditioning, and the fracture performances of an OGFC, respectively. Furthermore, a new test is proposed for evaluating the durability performance of an OGFC in wet conditions using the Micro-Deval device. According to this study, OGFC's durability and fracture performance improves as the grade of binder used increases. The OGFCs prepared with aggregates of smaller NMAS outperformed the OGFCs prepared with larger NMAS aggregate in both durability and fracture tests despite having similar asphalt contents.

Keywords

infrastructure materials asphalt materials selection and mix design open graded friction course

Open graded friction course (OGFC) is a special type of asphalt concrete (AC) surface layer that is placed on top of dense graded layer in a flexible pavement ( 1 ). It is a porous material whose primary objectives are to drain out the rainwater quickly and to improve frictional characteristics. To achieve these special features, the OGFC material consists of a relatively higher percentage of air void, usually between 15% to 22%, whereas the typical air void for the traditional dense mix is between 5% and 7% ( 2 ). Furthermore, the OGFC material consists of mostly coarse size aggregates with low fraction of fines to ensure a higher content of connected air voids ( 3 ). Past studies showed that the use of OGFC layer in a pavement can improve wet weather skid resistance, reduce the potential for hydroplaning effect, reduce water splash and spray, reduce noise, and reduce night-time wet pavement glare. In summary, a pavement with OGFC as its surface layer can provide better ride quality compared with a pavement with dense graded material as its surface layer. Despite these benefits, the use of OGFC has diminished over the years because of its poor service life issues ( 4 , 5 ). This shorter durability is a major problem for the OGFC material. Because of its open graded aggregate skeleton and high air void content, the aggregate-to-aggregate bond strength of an OGFC is significantly low compared with a dense graded material. Loss of materials at the surface of the pavement can easily be caused by repeated abrasion from the traffic wheels and this can be aggravated by the presence of moisture ( 6 ). This is because the OGFC layers are permeable, water can easily penetrate them and hamper the bond between aggregate and asphalt binder. Moreover, the high air content in OGFC leads to accelerated aging because of higher accessibility of air, resulting in brittle OGFC that helps initiate cracking within the thin OGFC layer. For these reasons, OGFC has a shorter expected service life than the traditional dense mixes. For example, the expected service life of an OGFC ranges between six and 12 years, whereas it is about 12 to 18 years for a dense graded mix ( 7 ). Therefore, several US states have discontinued the use of OGFC in their pavements.

The New Mexico Department of Transportation (NMDOT) uses OGFC material/layer for surfacing its high-volume traffic roads; however, it is found that the OGFC layers perform poorly in most cases. More importantly, the application of OGFC is suspected of playing a significant role in the statewide top-down cracking problem in New Mexico’s roads. After easy crack initiation within the thin OGFC layer, cracks propagate into the lower dense graded layer. It should be noted that NMDOT uses the traditional volumetric method, which is based on only air void, tensile strength ratio (TSR), and permeability requirements, for designing the OGFC layers without considering their performance. However, it is known that the volumetric mix design method cannot provide adequate performance of dense graded ACs in many cases. Therefore, there is a national trend for state DOTs to move toward performance-based mix design for AC, where an optimized AC mix is chosen based not only on its volumetric requirements but also on its performance. The question then arises whether OGFC is also required to have some performance criteria in its mix design procedure to represent its durability performance. Furthermore, NMDOT restricts its engineers to using only one binder grade, PG 70-28+ (polymer modified), all over the state, without regard to climate and/or traffic condition(s). NMDOT also has a restriction for selecting aggregate gradation for the OGFC mix. This gradation consists of 9.5 mm nominal maximum aggregate size (NMAS) aggregates. However, there is great variation in climatic conditions (temperature and rainfall) within the state ( 8 ). Like climate, the traffic pattern also differs significantly from location to location ( 9 , 10 ). Past studies have also shown that aggregate size has a substantial influence on the performance of an AC mixture ( 11 , 12 ). The use of the same materials over the whole state may not be a wise decision because it may fail to provide the required performance or be uneconomical. From the literature, it is found that different state DOTs use various different binder grades and aggregate gradations ( 13 , 14 ). Moreover, the effects of binder grade and aggregate gradation on OGFC's performance are not well understood.

The primary objectives of this research are to examine whether NMDOT’s current OGFC volumetric mix design method is sufficient to ensure long-term durability and to add some performance tests to the current procedure to provide adequate durability performance. In addition, this study examines whether NMDOT can remove the restriction on binder grade and aggregate gradation for its use of OGFC.

Materials and Mix Design

This study used two 9.5 mm and one 12.5 mm NMAS aggregate gradations, as shown in Figure 1. Between two 9.5 mm NMAS aggregates, one (denoted as NMAS-9.5-mm-L) has a lower fraction of fines than the other (denoted as NMAS-9.5-mm-H). The 12.5 mm NMAS aggregate (denoted as NMAS-12.5-mm) has the lowest fraction of fines among the used aggregate gradations. To make this study consistent, all aggregates were collected from the same source. In addition to three aggregate gradations, binders of three different performance grades (PG) were also used. The binder grades are PG 64-28+, PG 70-28+, and PG 76-22+, respectively. It should be noted that NMDOT specifies using either polymer or rubber modified asphalt binder in OGFC materials. Past research has shown that the polymer modified binders perform better than the neat (or unmodified) binders against various damages ( 15 – 17 ). Therefore, this study selected three styrene-butadiene-styrene (SBS) polymer modified binders from the same binder supplier. After collecting the binder, Fourier transform infrared (FTIR) spectroscopy was performed to confirm the polymer modification status ( 18 ). Furthermore, all binders passed the elastic recovery requirement (minimum 65%) as per NMDOT specification.

Figure 1.

Aggregate gradations used in this study.

The NMDOT specification ( 19 ) uses the ASTM D7064 procedure to design OGFC, which is a variant of the National Center of Asphalt Technology (NCAT) OGFC mix design procedure. However, NMDOT does not require any standard for durability (e.g., Cantabro abrasion test). The design requirements according to ASTM D7064 are shown in Table 1. In this study, no durability performance requirement was considered during the mix design. To make the study simple, PG 70-28+ was used for the mix design purpose, and the determined binder contents were used for other binders. Hydrated lime was added as an anti-stripping additive at a rate of 1.0% by weight of aggregate. Figure 2 describes the mix design steps used in this study. At first, aggregates were sieved and then recombined according to their required proportions (Figure 2a). Next, dry-rodded density, γ_w and bulk specific gravity of coarse aggregate, G_CA were determined, as shown in Figures 2b and c , respectively. These values were applied to calculate the voids in coarse aggregate, VCA_DRC using Equation 1.

VC A_{DRC} = \frac{G_{CA} γ_{w} - γ_{s}}{G_{CA} γ_{w}} \times 100

(1)

Table 1.

New Mexico Department of Transportation Mix Design Requirements for Open Graded Friction Course (OGFC) (According to ASTM D7064)

Test	Specification
Air void	≥ 18%
Drain-down loss	≤ 0.30%
Voids in coarse aggregate of compacted mix (VCA_MIX)	< VCA_DRC (voids in coarse aggregate)
Permeability	≥ 100 meters per day
Tensile strength ratio	≥ 80%

Figure 2.

Open graded friction course (OGFC) mix design procedure used in this study: (a) sieve analysis, (b) density of coarse aggregate (γ_s) test, (c) bulk specific gravity of coarse aggregate (G_CA) test, (d) bucket mixing, (e) theoretical maximum specific gravity of loose mix (G_mm) test, (f) drain-down test, (g) bulk specific gravity of loose mix (G_mb) test, (h) tensile strength ratio (TSR) test, and (i) permeability test.

where G_CA is the bulk specific gravity of the coarse aggregate, γ_w is the density of water, and γ_s is the dry-rodded density of coarse aggregate. After that, a bucket mixer (Figure 2d) was used to prepare aggregate-asphalt binder blends with binder contents ranging from 6.0% to 7.5% (by weight of total mixture) at 0.5% increments. The prepared loose mix was used to determine the theoretical maximum specific gravity, G_mm, and drain-down loss. The core-lock method (Figure 2e) was used to determine the G_mm of the loose mix. The drain-down test, as shown in Figure 2f, was performed at two different temperatures: mixing temperature and 10° C above the mixing temperature. This was done to determine the potential loss while storing and transporting plant-produced mix to the field. Next, the loose mix was compacted using the Superpave gyratory compactor using 50 gyrations to determine the bulk specific gravity, G_mb of compacted OGFC specimen, as depicted in Figure 2g. Finally, air voids, V_a, and voids in the coarse aggregate of the compacted mixes, VCA_MIX, were calculated from G_mb using Equations 2 and 3, respectively.

V_{a} = 100 \times (1 - \frac{G_{mb}}{G_{mm}})

(2)

VC A_{MIX} = 100 - (\frac{G_{mb}}{G_{CA}} \times P_{CA})

(3)

where P_CA is the fraction of coarse aggregate, G_mb is the bulk specific gravity of compacted mix, and, as above, G_mm is the theoretical maximum specific gravity. Three different criteria were used to determine the breakpoint sieve to calculate the P_CA in a mix:

Criterion 1: The breakpoint sieve is the no. 4 (4.75 mm) sieve for all mixtures ( 3 ).

Criterion 2: The breakpoint sieve is the sieve below which the slope of the gradation curve begins to flatten out ( 20 ).

Criterion 3: The breakpoint sieve is the smallest sieve in which a minimum of 10% of the aggregate is retained ( 20 ).

This study found that the first criterion does not work for the used gradations and the last two criteria give similar results. The optimum binder content was determined as the binder content at which the air void, drain-down loss, and VCA_MIX values met the recommended criteria. Lastly, TSR (Figure 2h) and permeability (Figure 2i) tests were performed to determine whether these requirements were also met (results are presented in following section).

Table 2 presents the mix design summaries. Based on the NMDOT mix design method, the optimum binder contents for the NMAS-9.5-mm-L OGFC and NMAS-9.5-mm-H OGFC are the same and which is 6.5%. However, the optimum binder content for the NMAS-12.5-mm OGFC is 6.4%. A lack of fines made it more prone to drain-down loss.

Table 2.

Mix Design Summaries

Asphalt concrete (%)	G_mm	G_mb	V_a (%)	VCA_MIX	VCA_DRC	Drain-down loss (%)
NMAS-9.5 mm-L
6.0	2.380	1.918	19.4	30.9	42.3	0.13
6.5	2.372	1.942	18.1	30.0	42.3	0.17
7.0	2.363	1.962	17.0	29.3	42.3	0.59
7.5	2.341	1.941	15.9	29.1	42.3	0.72
NMAS-9.5 mm-H
6.0	2.401	1.928	19.7	33.0	45.2	0.08
6.5	2.378	1.942	18.3	32.5	45.2	0.13
7.0	2.354	1.953	17.0	32.2	45.2	0.35
7.5	2.315	1.934	16.5	32.8	45.2	0.54
NMAS-12.5 mm
6.0	2.361	1.901	19.5	30.8	44.6	0.17
6.5	2.344	1.917	18.2	30.2	44.6	0.31
7.0	2.325	1.922	17.3	30.0	44.6	0.59
7.5	2.308	1.934	16.2	29.6	44.6	0.73

Note: G_mm = theoretical maximum specific gravity of compacted mix; G_mb = bulk specific gravity of compacted mix; V_a = air voids; VCA_MIX = voids in coarse aggregate of compacted mix; VCA_DRC = voids in coarse aggregate.

Experimental Design

After determining the optimum binder contents, nine different OGFC mixtures were produced from three binders and three gradations. Next, TSR and permeability tests were performed on each mix to be sure that they met the NMDOT design requirements. Their durability performance was evaluated using the widely used Cantabro abrasion (CA) test. It should be noted that the CA test is done by using the Los Angeles (LA) abrasion drum and it determines the durability performance of the OGFC mixture without the presence of the steel charges. During this test, a gyratory compacted cylindrical specimen of OGFC is lifted to a certain height and then released to free fall. As a result, instead of causing damage from the combination of abrasion, impact, and grinding, it causes loss only from impact loading. Therefore, the loading condition in the CA test is not properly representative of field loading conditions. Furthermore, the CA test is performed only on dry specimens (without presence of water) in both cases. In a practical scenario, it is very common for a wet OGFC layer to be subjected to repeated traffic loadings during/after rainfall. Even after the rainfall has ceased, water can be clogged inside the voids. Under wheel load this entrapped water can create tremendous pressure on the void walls and cause damage to the pavement materials. However, currently there is no test available that measures the durability performance of OGFC material in such wet conditions or in the presence of water. Therefore, this study proposes a new durability test using a Micro-Deval (MD) device to determine the durability performance of OGFC material in wet conditions. The primary reason for choosing the MD test is that it can be performed in the presence of both water and steel charges, which facilitates capturing durability performance against mainly abrasion load in wet conditions. In addition to durability performance, the cracking performance of each OGFC mix was determined using the semi-circular bending (SCB) test.

Results and Discussion

TSR Test

The TSR test was used to determine the moisture damage susceptibility of the AC mixture. It was done by performing indirect tensile (IDT) strength tests on cylindrical AC specimens in both unconditioned and moisture conditioned states. The unconditioned state refers to undamaged samples, whereas the moisture conditioned state refers to samples damaged by moisture after a moisture conditioning or freeze-thaw cycle. The moisture conditioning was done according to AASHTO T 283. The IDT test was conducted by loading a cylindrical specimen across its vertical diametral plane at a deformation rate of 50 mm/min (2 in./min) and 25°C test temperature. The peak load at failure was recorded and used to calculate the IDT strength of the specimen using Equation 4.

IDT = \frac{2 \times P}{π Dt}

(4)

where IDT is the indirect tensile strength, P is the maximum compressive strength, D is the diameter of the sample, t is the thickness of the sample. Finally, the TSR value can be calculated from Equation 5. In this study, three replicate specimens were used to determine the IDT strengths at unconditioned and moisture conditioned states.

T S R = \frac{I D T_{a g e d}}{I D T_{u n a g e d}} \times 100 (%)

(5)

The IDT strengths of tested OGFCs are presented in Figure 3. Figure 3a shows that IDT strength in unconditioned state increases as the binder grade increases, regardless of aggregate gradation. It is known that binder stiffness increases as the temperature grade of the binder increases. Thus, the PG 76-22+ OGFCs have the highest IDT strengths, followed by the PG 70-28+ OGFCs regardless of aggregate gradation. For the same reason, the PG 64-28+ OGFCs have the lowest IDT strengths among the tested OGFCs. While comparing IDT results among different aggregate gradations, it can be seen that IDT strengths are greater for the NMAS-9.5-mm OGFCs than the NMAS-12.5-mm OGFCs. The relatively fine aggregates in the NMAS-9.5-mm OGFCs are closely packed together, and as a result, they have improved load carrying capacities than the NMAS-12.5-mm OGFCs. Like unconditioned states, a similar trend of IDT strengths is also observed for moisture conditioned states, as shown in Figure 3b. However, as a result of moisture conditioning, every OGFC has a lower IDT strength than its unconditioned state. Figure 3c compares the TSR values among the nine test OGFCs. It shows that the TSR values of all the tested OGFCs are around 80%. This indicates that all OGFCs meet the TSR requirement. It should be noted that NMDOT only uses the TSR value as a performance measure for OGFC materials. Using only the TSR value, it is difficult to distinguish the performance of different OGFCs. This indicates that a suitable performance test/parameter is required to properly characterize the durability performance of an OGFC.

Figure 3.

Tensile strength ratio (TSR) test results (M = mean, SD = standard deviation): (a) unconditioned state, and (b) moisture conditioned state, (c) comparison of TSR values among samples.

Permeability Test

The permeability test was used to determine the water conductivity of the compacted AC sample. The water conductivity represents the rate of flow of water through the specimen. In this study, a falling head permeability test apparatus was used to determine the rate of flow of water through each OGFC specimen. In this test, water in a graduated cylinder was allowed to flow through a saturated asphalt sample, and the interval of time taken to reach a known change in the head was recorded. The coefficient of permeability of the asphalt sample was then determined using Darcy’s Law, as shown in Equation 6.

k = \frac{aL}{At} \ln (\frac{h_{1}}{h_{2}})

(6)

where k is the coefficient of permeability, a is the cross-sectional area of the tube, A is the cross-sectional of the sample, L is the thickness of the sample, t is elapsed time, h₁ and h₂ are the water heights during the test.

The permeability test was done on three replicate specimens and the test results are presented in Figure 4. It shows that binder grade has no effect on the permeability performance. Permeability performance depends on the aggregate gradation of OGFC. It can be seen that the NMAS-12.5-mm OGFCs have the highest permeability values regardless of binder type. Though the NMAS-9.5-mm OGFCs have the same air voids as the NMAS-12.5-mm OGFCs, the presence of more fine aggregates in the the NMAS-9.5-mm OGFCs clogged their interconnected pores and prevented the water from flowing. As a result, permeability performance is lower for the NMAS-9.5-mm OGFCs. With a little more fines, the NMAS-9.5-mm-H OGFCs have further lower permeability than the NMAS-9.5-mm-L OGFCs. The NMDOT’s permeability performance requirement is a minimum 100 m/day and all OGFCs passed the requirement.

Figure 4.

Permeability test results (M = mean, SD = standard deviation).

CA Test

The CA test is one of the most widely used tests to evaluate the durability of OGFC mixtures ( 2 ). The test is done by damaging AC samples using the LA abrasion drum by applying 300 revolutions at 30–33 rpm in the absence of steel spheres at room temperature (generally at 25°C). The mass change from abrasion is reported as the percentage of mass loss using Equation 7.

% l o s s = \frac{M_{1} - M_{2}}{M_{1}} \times 100 (%)

(7)

where M₁ and M₂ are the total mass of three specimens before and after the test, respectively. The NCAT recommended values for the CA test are 20% and 30% for unaged and oven-aged specimens, respectively. However, this study used moisture conditioning instead of oven aging procedure to better understand the moisture damage effect.

The CA test was performed, as shown in Figure 5, on three replicate specimens at each condition state for each OGFC mix. The specimen height and diameter are 115 mm and 150 mm, respectively. The moisture conditioning was done according to the AASHTO T283 procedure. The three replicate specimens of each OGFC were tested simultaneously. Thus, the percent loss was calculated using the total mass of the three specimens. The CA losses of all mixes are presented in Figure 6. Figures 6a and b show that the CA loss decreases as the binder grade increases regardless of aggregate gradation for both unconditioned and moisture conditioned states, respectively. This is probably a result of the modification of binders, binders with higher grades have better damage tolerances ( 21 , 22 ). Thus, the PG 76-22+ OGFCs have the lowest CA mass losses, followed by the PG 70-28+ OGFCs, regardless of aggregate gradation. For the same reason, the PG 64-28+ OGFCs have the highest CA losses among the tested OGFCs. While comparing CA results among different aggregate gradations, it can be seen that CA losses are higher for the NMAS-12.5-mm OGFCs than the NMAS-9.5-mm OGFCs. The lack of fine aggregates in the NMAS-12.5-mm OGFCs is the main reason for their poor performance. It is known that if fine particles are missing from the aggregate matrix, then the asphalt binder is only able to bind the coarse particles at their relatively few contact points. Thus, this kind of mixture has relatively lower stability than a mixture with more fines. While comparing CA performance between unconditioned and moisture conditioned specimens, every moisture conditioned OGFC sample demonstrated more loss than the same mix in unconditioned state.

Figure 5.

Cantabro abrasion (CA) test: (a) Los Angeles (LA) abrasion tester, (b) undamaged specimen, (c) abraded specimen after the test, and (d) comparison of abraded status between unconditioned and moisture conditioned specimens.

Figure 6.

Cantabro abrasion (CA) test results (M = mean, SD = standard deviation): (a) unconditioned state and (b) moisture conditioned state.

MD test

The MD test is proposed in this study to determine the durability performance of an OGFC in wet conditions. Figure 7 demonstrates the procedure of the MD test that was followed in this study. Figures 7a and b show the MD test device and drum/container with steel charges. In this test, cylindrical AC samples were prepared using a gyratory compactor. The height and diameter of the samples are 115 mm and 150 mm, respectively. The sizes of the drum/container of the MD test are smaller than the LA abrasion test. Therefore, each sample was trimmed with a 100 mm core (Figure 7c) to insert it into the test container. The wet samples were kept on a rack for at least 24 h to naturally dry out the moisture, as illustrated in Figure 7d. After drying, the weight of each sample was measured (Figure 7e). The MD apparatus is capable of performing two tests at a time. Therefore, two specimens were soaked in water at room temperature for 2 h, as depicted in Figure 7f. For the test, each wet specimen was inserted into one test container separately. Then, each container was filled with 2 L of water (at room temperature) and 5 kg steel charges, as shown in Figure 7g. Next, each container was sealed with a cap and placed inside the MD device. The test was started by selecting the duration of the testing and pressing the start button. In this study, the duration of testing was chosen as 4 h, and the speed of the revolution was 100 rpm. After the completion of the test, the broken specimens were taken from the test container, washed thoroughly, and left for 24 h for natural drying. When the damaged samples were completely dry, the final weight was measured (Figure 7h). Finally, the percent mass loss was calculated using Equation 8, where M_i and M_f are the mass of each specimen before and after the test, respectively.

% l o s s = \frac{M_{i} - M_{f}}{M_{i}} \times 100 (%)

(8)

In this study, three replicate specimens were investigated individually for each OGFC. Figure 8 compares the damaged specimens prepared with different binders after completion of the test. In this case, all specimens were prepared with the NMAS-9.5-mm-L aggregate. Figure 8a shows three replicate damaged specimens prepared with the PG 64-28+ binder. It is clearly evident that all three specimens experienced significant damage during the test. It can also be seen that all three specimens are in similar damaged conditions. Because each specimen was tested individually, this observation confirms the repeatability of the MD test. Figures 8b and c show the damaged specimens prepared with the PG 70-28+ and PG 76-22+ binders, respectively. In these two cases, all specimens experienced relatively less damage than the specimens prepared with the PG 64-28+ binders. The primary reason is that the PG 64-28+ binder is relatively less stiff and weaker than the other two binders. Thus, it experienced more damage in the MD test. Figure 8d compares the damaged specimens prepared with three different binder grades. It can be seen that the damage ranking is based on their binder grades. This confirms that the MD test can successfully rank the OGFC based on their performances.

Figure 7.

Micro-Deval (MD) test procedure used in this study: (a) MD tester, (b) container and steel charges, (c) coring the gyratory compacted specimen, (d) specimens in drying rack, (e) weight of undamaged specimen, (f) specimens soaking in water, (g) specimen, steel charges, and water in container, and (h) weight of damaged specimen.

Figure 8.

Abraded specimens after Micro-Deval (MD) test prepared with the NMAS-9.5-mm-L aggregate and different binder grades: (a) PG 64-28+, (b) PG 70-28+, (c) PG 76-22+, and (d) comparison of effects of binder grade.

Figure 9 presents the MD test results. Clearly, the PG 64-28+ OGFCs have the highest losses compared with OGFCs made with the other two binders, regardless of the aggregate gradation. During the MD test, heat is generated by the friction between steel charges, and between steel charges and the specimen. This heat increases the temperature of the internal water which makes the PG 64-28+ binder soften and thus it accumulates damage more easily. Among different aggregate gradations, the NMAS-12.5-mm OGFCs experienced more damage than the NMAS-9.5-mm OGFCs regardless of the binder grade. It can also be seen that the standard deviations of all cases are very small. Thus, it can be concluded that the MD test is repeatable.

Figure 9.

Micro-Deval (MD) test results.

Unlike the CA test results, the poor performance of the PG 64-28+ OGFCs compared with the PG 70-28+ and PG 76-22+ OGFCs are clearly identifiable from the MD test. To account for this poor performance, this study proposes a new criterion, a maximum 25% loss in the MD test, to ensure adequate durability performance of an OGFC in wet conditions. A one-tailed t-test was performed to find out which OGFC has MD loss greater than 25% at 95% confidence level (p-value = 0.05). Table 3 shows the t-test results to see whether any of OGFC mixes failed to meet the requirement. It can be seen that all three OGFCs made with the PG 64-28+ binder failed to meet the requirement.

Table 3.

t-Test Results

Parameter	NMAS-9.5-mm-L(%)			NMAS-9.5-mm-H(%)			NMAS-12.5-mm(%)
Parameter	PG 64-28+	PG 70-28+	PG 76-28+	PG 64-28+	PG 70-28+	PG 76-28+	PG 64-28+	PG 70-28+	PG 76-28+
number of samples	3	3	3	3	3	3	3	3	3
degrees of freedom	2	2	2	2	2	2	2	2	2
Mean	38.88	15.29	13.24	29.17	13.54	10.86	48.40	21.04	17.77
Std error	3.409	0.519	0.270	2.824	0.430	0.535	2.676	0.769	0.233
t-statistic	4.07	−18.73	−43.62	3.06	−26.65	−26.42	8.74	−5.15	−31.00
Design value	25	25	25	25	25	25	25	25	25
p-value	0.028	0.999	1.000	0.048	0.999	0.999	0.006	0.982	0.999
Significant	Yes	No	No	Yes	No	No	Yes	No	No

SCB Test

The SCB test, as shown in Figure 10, was used to determine the fracture performance of OGFC. In this test, a pre-notched semi-circular specimen was loaded monotonically until fracture failure under a constant deformation rate of 0.5 mm/min and at a room temperature of 25°C. During the test, the load and deformation were continuously recorded and used to calculate two parameters: strain energy to failure, U and J-integral. The U is defined as the work required to initiate a crack. The J-integral is known as the critical fracture energy release rate, and it can be determined using Equation 9. The J-integral represents the mixture’s resistance to fracture propagation. When the J-integral value of a mix is higher, it is more resistant to fracture or crack propagation.

J - integral = - \frac{1}{b} (\frac{dU}{da})

(9)

where J-integral is the critical strain energy release rate (J/m²); b is the sample thickness (m); a is the notch depth (m); U is the strain energy to failure (J), and dU/da represents the change of strain energy with notch depth (J/m). In this study, the semi-circular specimens were prepared from cylindrical specimens compacted by the Superpave gyratory compactor. Three replicate specimens were tested at three different notch depth levels (15 mm, 25 mm, and 32 mm). The average thickness of the tested specimens is 60 mm.

Figure 10.

Semi-circular bending (SCB) test: (a) test specimen and (b) test setup.

Figure 11 demonstrates the SCB test results for the OGFC prepared with the NMAS-9.5-mm-L aggregate and the PG 70-28+ binder. Figures 11a to c show the load–displacement curves for different notch depths. It can be seen that at each notch depth the load–displacement curves for the three replicates are close to each other. This confirms the repeatability of test results. As expected, the peak failure load decreases with an increase in notch depth. It was also found that the failure displacement corresponding to the peak load decreases with notch depth. Using these load–displacement plots, respective strain energies up to failure were calculated. Figure 11d shows the area of the load–displacement curve, which is defined as strain energy up to failure, U in this study.

Figure 11.

Semi-circular bending (SCB) test results for open graded friction course (OGFC) prepared with the NMAS-9.5-mm-L aggregate and the PG 70-28+ binder at different notch depths and strain energy up to failure point: (a) 15 mm, (b) 25 mm, (c) 32 mm, and (d) area of the load–displacement curve.

Table 4 summarizes the strain energies up to failure point of all OGFC samples determined from the SCB test. It can be seen that the strain energies of the PG 64-28+ OGFCs are lower than the OGFCs prepared with the other two binders, regardless of notch depth and aggregate gradation. This is because the PG 64-28+ binder is significantly weaker than the other two binders. Similarly, strain energies of the PG 76-22+ OGFCs are higher than other OGFCs at 15 mm and 25 mm notch depths regardless of aggregate gradation. At 32 mm notch depth, the strain energies of the PG 76-22+ and PG 70-28+ OGFCs are comparable. From the table, it can also be observed that the NMAS-9.5-mm OGFCs have higher strain energies than the NMAS-12.5-mm OGFCs. The lack of fines in the NMAS-12.5-mm OGFCs is detrimental to their load carrying mechanisms and makes them more vulnerable to fracture. The NMAS-9.5-mm-H OGFCs perform slightly better than the NMAS-9.5-mm-L OGFCs. This is because the former have more fines than the latter, which makes them more stable and improves their loading carrying capacities.

Table 4.

Semi-Circular Bending (SCB) Test Results

Replicate no.	Notch depth (mm)	NMAS-9.5-mm-L (J)			NMAS-9.5-mm-H (J)			NMAS-12.5-mm (J)
Replicate no.	Notch depth (mm)	PG 64-28+	PG 70-28+	PG 76-22+	PG 64-28+	PG 70-28+	PG 76-22+	PG 64-28+	PG 70-28+	PG 76-22+
1	15	0.272	0.655	0.670	0.313	0.735	0.811	0.130	0.488	0.529
2	15	0.241	0.694	0.660	0.357	0.858	0.942	0.159	0.461	0.503
3	15	0.215	0.593	0.697	0.319	0.919	1.055	0.149	0.392	0.507
1	25	0.171	0.375	0.416	0.209	0.427	0.525	0.096	0.294	0.304
2	25	0.166	0.337	0.364	0.176	0.433	0.470	0.084	0.313	0.329
3	25	0.181	0.355	0.381	0.195	0.525	0.458	0.093	0.267	0.296
1	32	0.054	0.258	0.209	0.112	0.297	0.308	0.064	0.208	0.173
2	32	0.063	0.209	0.206	0.134	0.282	0.317	0.068	0.198	0.170
3	32	0.079	0.251	0.181	0.138	0.316	0.276	0.087	0.154	0.194

After determining strain energies, the relationship between strain energies and notch depths (Figure 12) was evaluated to compute the dU/da value for each OGFC. It is known that strain energy decreases as the notch depth increases, and a steep slope represents a better fracture resistance than a mild slope and vice versa. Figure 12a compares the relationships between strain energies and notch depths among three NMAS-9.5-mm-L OGFCs prepared with three different binders. It can be observed that the slopes of OGFCs prepared with different binders are different. The PG 76-22+ OGFC has the steepest slope among all the tested OGFCs. On the other hand, the PG 64-28+ OGFC has the mildest slope. The slope of PG 70-28+ OGFC is lower than PG 76-22+ OGFC. Thus, it can be said that fracture resistance increases as the binder grade increases. Similar observations can be seen for the NMAS-9.5-mm-H and NMAS-12.5-mm OGFCs, as shown in Figures 12b and c , respectively.

Figure 12.

Notch depth versus fracture energy for open graded friction course (OGFC) samples prepared with different binders: (a) NMAS-9.5-mm-L, (b) NMAS-9.5-mm-H, and (c) NMAS-12.5-mm.

Using dU/da values and thicknesses of the specimen, the J-integral value for each mix was determined from Equation 9. The computed J-integral values are listed in Table 5. It is known that a higher J-integral value represents that the mix is more resistant to fracture or crack propagation. It can be seen that the PG 64-28+ OGFCs have the lowest J-integral values regardless of aggregate gradation. Similarly, the PG 76-22+ OGFCs have the highest J-integral values followed by the PG 70-28+ OGFCs. Among different aggregate gradations, the NMAS-9.5-mm-H OGFCs have the highest J-integral values than the other two gradations, while the NMAS-12.5-mm OGFCs have the lowest values regardless of binder type.

Table 5.

J-Integral Values

Binder type	J-integral (J/m²)
Binder type	NMAS-9.5-mm-L	NMAS-9.5-mm-H	NMAS-12.5-mm
PG 64-28+	170	200	74
PG 70-28+	405	535	255
PG 76-22+	460	620	328

Conclusion

Based on this study, the following conclusions can be made:

The IDT test results reveal that the IDT strength increases as binder grade increases in both unconditioned and moisture conditioned states. This is because binder stiffness increases as binder grade increases. It should be noted that the TSR value is not affected by the binder grade. For example, the PG 64-28+ OGFCs have lower IDT strengths than the PG 70-28+ OGFCs regardless of the aggregate gradation. However, their TSR values are close to each other. In this study, all binders passed the TSR requirement. Like the binder grade, the aggregate gradation also affects the IDT performance of an OGFC. The NMAS-9.5-mm OGFCs have higher IDT strengths than the NMAS-12.5-mm OGFCs regardless of binder type. This is because the NMAS-12.5-mm OGFCs have fewer fine aggregates than the NMAS-9.5-mm OGFCs. The lack of fine aggregates in the NMAS-12.5-mm OGFCs deteriorates the inter-aggregate load transferring mechanism and reduces their overall load carrying capacities.

It is observed from the permeability test that the binder grade has no influence on the permeability performance of an OGFC. Permeability performance is directly related to the aggregate gradation. The NMAS-12.5-mm OGFCs exhibited higher permeability performance than the NMAS-9.5-mm OGFCs. Though all of them have the same air voids, relatively more fine aggregates in the NMAS-9.5-mm OGFCs clogged the interconnecting pores and reduced their permeability performance.

From the CA and MD tests, it is found that the PG 64-28+ OGFCs experienced more damage than the other two OGFCs. On the other hand, the PG 76-22+ OGFCs experienced less damage followed by the PG 70-28+ OGFCs. Among different aggerate gradations, the NMAS-9.5-mm OGFCs performed better in both CA and MD tests than the NMAS-12.5-mm OGFCs regardless of binder grade.

The SCB test results reveal that the PG 64-28+ OGFCs have the lowest fracture energies and J-integrals compared with the other two OGFCs. The PG 76-22+ OGFCs slightly outperformed the PG 70-28+ OGFCs in the SCB test. Among different aggregate gradations, the NMAS-9.5-mm-H OGFCs have the highest J-integral values followed by the NMAS-9.5-mm-L OGFCs regardless of binder type. NMAS-12.5-mm OGFCs have the lowest values in all cases.

The NMDOT’s current volumetric mix design based on only the TSR test cannot provide any concrete information about the durability performance of an OGFC mixture. For example, all three binders investigated in this study met the volumetrics and TSR requirements during mix design. However, the PG 64-28+ OGFCs performed worst in all performance tests (CA, MD, and SCB).

Based on the test results, this study recommends using the PG 76-22+ binder as an alternative to the current PG 70-28+ binder for better performance on high traffic roads or in hotter regions.

The CA test can be useful to characterize both abrasion and moisture conditioning performance of an OGFC mixture. Similarly, the SCB test can be used to investigate fracture performance.

The MD test proposed in this study can be a good addition to other performance tests for evaluating the durability performance of an OGFC in wet conditions.

The findings presented in this study are based on the performance of nine OGFC samples prepared with three binders and three aggregate gradations. The effects of different material sources, different amounts of fines, and different polymers on the performance of OGFC were not investigated here. Thus, further investigation is needed to explore these effects.

Footnotes

Acknowledgements

The authors would like to express their sincere gratitude and appreciations to the Project Technical Panel Members, Project Advocate, Project Sponsor, and Project Manager at the New Mexico Department of Transportation.

Author Contributions

The authors confirm the contribution to the paper as follows: study conception and design: M. A. Hasan, R. A. Tarefder; laboratory testing: M. A. Hasan, Z. H. Khan; analysis and interpretation of test results: M. A. Hasan, R. A. Tarefder; manuscript preparation: M. A. Hasan, R. A. Tarefder. All authors reviewed the results and approved the final version of the manuscript.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was funded by the New Mexico Department of Transportation (NMDOT).

ORCID iDs

Md Amanul Hasan

Rafiqul A. Tarefder

References

Huber

Performance Survey on Open-Graded Friction Course Mixes. Transportation Research Board, Washington, D. C., 2000.

Alvarez

A. E.

Epps-Martin

Estakhri

Izzo

Evaluation of Durability Tests for Permeable Friction Course Mixtures. International Journal of Pavement Engineering, Vol. 11, No. 1, 2010, pp. 49–60.

Kandhal

P. S

. Design, Construction, and Maintenance of Open-Graded Asphalt Friction Courses. Information Series 115. National Asphalt Pavement Association, Washington, D. C., 2002.

Nelson

J. P.

Meta-Analysis of Airport Noise and Hedonic Property Values. Journal of Transport Economics and Policy, Vol. 38, No. 1, 2004, pp. 1–27.

Cooley

Brumfield

Mallick

Mogawer

Partl

Poulikakos

Hicks

Construction and Maintenance Practices for Permeable Friction Courses. NCHRP Report 640. Transportation Research Board, Washington, D.C., 2009.

Shu

Huang

Song

Onyango

M. A.

Optimize Application of Open Graded Friction Course (OGFC). Final Report. Tennessee Department of Transportation, Nashville, 2017.

Yildirim

Dossey

Fults

Tahmoressi

Trevino

Winter Maintenance Issues Associated with New Generation of Open-Graded Friction Courses. Report FHWA/TX-08-0-4834-2. Center for Transportation Research, Austin, TX, 2007.

Hasan

M. A.

Tarefder

R. A.

Development of Temperature Zone Map for Mechanistic Empirical (ME) Pavement Design. International Journal of Pavement Research and Technology, Vol. 11, No. 1, 2018, pp. 99–111.

Hasan

M. A.

Islam

M. R.

Tarefder

R. A.

Clustering Vehicle Class Distribution and Axle Load Spectra for Mechanistic-Empirical Predicting Pavement Performance. Journal of Transportation Engineering, Vol. 142, No. 11, 2016, p. 05016006. https://doi.org/10.1061/(ASCE)TE.1943-5436.0000876.

10.

Hasan

M. A.

Tarefder

R. A.

Linking Traffic Volume and Weight Data for Mechanistic-Empirical Pavement Design. Journal of Transportation Engineering, Part B: Pavements, Vol. 146, No. 2, 2020, p. 04020015.

11.

Hasan

M. M.

Ahmad

Hasan

M. A.

Faisal

H. M.

Tarefder

R. A.

Laboratory Performance Evaluation of Fine and Coarse-Graded Asphalt Concrete Mix. Journal of Materials in Civil Engineering, Vol. 31, No. 11, 2019, p. 04019259.

12.

Hasan

M. A.

Khan

Tarefder

R. A.

Effects of Aggregate Gradations on the Performance of Open Graded Friction Course (OGFC) Mixtures. In International Conference on Transportation and Development ( Bhat

C. R.

, ed.), American Society of Civil Engineers, Reston, VA, 2021, pp. 274–280.

13.

Jackson

N. C.

Vargas

Pucinelli

Quieter Pavement Survey. Report No. WA-RD 688.1. Washington State Transportation Commission, Department of Transportation, Olympia, 2008.

14.

Watson

Johnson

Jared

Georgia DOT’s Progress in Open-Graded Friction Course Development. Transportation Research Record: Journal of the Transportation Research Board, 1998. 1616: 30–33.

15.

Hasan

M. A.

Ahmad

Mannan

U. A.

Tarefder

R. A.

Laboratory Investigation and Modeling of High Temperature Performance of Polymer Modified Binders (PMBs) with Different Concentration of Styrene-Butadiene-Styrene (SBS) Polymer. In Airfield and Highway Pavements 2019: Testing and Characterization of Pavement Materials ( Al-Qadi

Ozer

Loizos

Murrell

, eds.), American Society of Civil Engineers, Reston, VA, 2019, pp. 248–254.

16.

Hasan

M. A.

Tarefder

R. A.

Laboratory Investigation of Low-Temperature Performance of the SBS Modified Mixtures. In Airfield and Highway Pavements 2019: Testing and Characterization of Pavement Materials ( Al-Qadi

Ozer

Loizos

Murrell

, eds.), American Society of Civil Engineers, Reston, VA, 2019, pp. 109–113.

17.

Hasan

M. A.

Tarefder

R. A.

New Mechanistic Procedure to Predict the Critical Cracking Temperature of Asphalt Concrete from Bending Beam Rheometer and Indirect Tension Test Data. Transportation Research Record: Journal of the Transportation Research Board, 2021. 2675: 438–453.

18.

Hasan

M. A.

Mannan

U. A.

Tarefder

R. A.

Determination of Polymer Content in SBS Modified Asphalt Binder Using FTIR Analysis. Proc., Ninth International Conference on Construction in the 21st Century (CITC 9), Dubai, UAE, March 5–7, 2017.

19.

New Mexico Department of Transportation (NMDOT). Standard Specifications for Highway and Bridge Construction. NMDOT, Santa Fe, NM, 2019.

20.

Watson

D. E.

Masad

Moore

K. A.

Williams

Cooley

L. A.

Verification of Voids in Coarse Aggregate Testing: Determining Stone-on-Stone Contact of Hot-Mix Asphalt Mixtures. Transportation Research Record: Journal of the Transportation Research Board, 2004. 1891: 182–190.

21.

Hasan

M. A.

Hasan

M. M.

Bairgi

B. K.

Mannan

U. A.

Tarefder

R. A.

Utilizing Simplified Viscoelastic Continuum Damage Model to Characterize the Fatigue Behavior of Styrene-Butadiene-Styrene (SBS) Modified Binders. Construction and Building Materials, Vol. 200, 2019, pp.159–169.

22.

Hasan

M. A.

Tarefder

R. A.

Investigating the Effects of SBS Concentration on Overall Performances of Polymer Modified Binders and Mixes. Presented at 100th Annual Meeting of the Transportation Research Board, Washington, D.C., 2021.