Effect of Aggregate Size and Compaction on the Strength and Hydraulic Properties of Pervious Concrete

Abstract

Pervious concrete is one of the emerging sustainable materials that has recently gained the attention of many researchers. The importance of pervious concrete mainly depends on its application and on a modern integrated approach in which it is employed to reduce the effects of flooding. The main goal of this experimental analysis is to study the significance of aggregate size and the degree of compaction on the mechanical and hydraulic properties of pervious concrete. Eleven concrete mixture proportions were investigated by controlling the constituents with different aggregate fractions. The important variables considered were the aggregate sizes, viz., 0/4 mm, 4/8 mm, and 8/16 mm, with four different degrees of compaction. The porosity of the concrete structure was obtained by the partial filling of the voids in the aggregates with cement paste. The ingredients of the pervious concrete were also varied to study their significance and to evaluate the predominant factor that controls the mechanical and hydraulic properties based on the test results. Tests were conducted to determine properties such as compacting factor, compressive strength, splitting tensile strength, abrasion resistance, porosity, and hydraulic conductivity. The study revealed that the degree of compaction was one of the critical factors governing the strength and hydraulic properties of the pervious concrete; the maximum strength and minimum hydraulic conductivity were achieved with a higher degree of compaction. The test results imply that the cement content is the predominant factor determining the fresh and tensile properties of the pervious concrete, rather than the size of the aggregates used. In addition, the results also illustrated that the highly compacted pervious concrete samples made with 4/8 mm aggregates exhibited improved abrasion resistance and strength properties, but slightly reduced hydraulic conductivity, despite the designed porosity.

1. Introduction

Adequate drainage on horizontal surfaces is one of the major issues that most countries have recently faced. High-volume surface run-off in urban areas is a common problem because of the wide use of non-pervious surfaces, leading to high flood risks. Proper surface run-off is essential to alleviate flooding and also for the rejuvenation of groundwater. Pervious concrete is a special material broadly used to address these issues, and it possesses many other environmental benefits. It was first employed in the United Kingdom in the year 1852 for building construction [1]. The distinct feature of pervious concrete as compared with conventional concrete is the interconnected pore structure. The interconnected pores permit the stormwater to percolate through the pavement and reach the sub-base/base layer [2]. Some other important advantages of pervious concrete are its high sound absorption, heat island effect reduction, water purification, reduced temperature, and solar reflectance when used for pavements [3,4,5]. The essential properties of pervious concrete are its porosity that ranges (from 11% to 35%) and hydraulic conductivity (0.2 cm/s–3 cm/s) [6,7]. Due to the highly porous structure of pervious concrete, the compressive strength and flexural strength range from 2.8 MPa to 28 MPa [8] and from 1.5 MPa to 3.2 MPa [9], respectively, values much lower than those of conventional concrete. This low strength has limited the utilisation of pervious concrete in roadways that receive excessive traffic and has led to considerable research into ways of enhancing the mechanical properties of pervious concrete [10,11,12,13]. Hence, improving its mechanical properties will lead to the wider applications of pervious concrete.

The properties of pervious concrete are highly influenced by many factors, which include water–cement ratio, degree of compaction, gradation of aggregates, and volume of binder. Yu et al. [14] studied the impact of aggregate size on compressive strength and reported that compressive strength was enhanced with an aggregate size of up to 7 mm, beyond which there was no effective improvement in strength. They also reported that the strength improved as the cement paste thickness around the aggregates increased up to 1.15 mm, after which it was comparatively constant. The effect of the water–cement ratio on the compressive strength of planting concrete, i.e., concrete with pores in which plants can germinate, was examined by Chen et al. [15]. They noted that the water–cement ratio affects the pores in the concrete to a great extent; as the water–cement ratio increases, the pores decrease. Similar behaviour was observed by Sahdeo et al. [16], who noticed that a water–cement ratio beyond 0.35 resulted in the cement paste descending to the ground surface of the concrete.

A considerable amount of research has been performed to improve the mechanical properties of pervious concrete by incorporating fibres, admixtures, waste cloth strips, etc. Juradin et al. [17] stated that pervious concrete is highly susceptible to steel reinforcement corrosion and researched whether adding waste cloth strips could increase the mechanical properties. A study in which macro synthetic fibres were added to pervious concrete was carried out by Kervern et al. [18], who noted that the inclusion of macro synthetic fibres is beneficial for the application of pervious concrete in high-traffic highways due to its improved toughness and fracture resistance. It was also pointed out that incorporating microfibres at higher dosages results in the degradation of the properties of pervious concrete due to inadequate bonding between the cement paste and the aggregate. Investigations on the utilisation of recycled aggregates also provided some valuable outcomes in terms of enhancing the properties of pervious concrete. Guo et al. [19] investigated the effect of recycled aggregates in pervious concrete. They found that concrete with 25% recycled aggregates exhibited greater compressive strength than the other mixture proportions considered in the study. Studies have also indicated that the employment of 100% recycled aggregates reduce the performance of pervious concrete [20,21]. Zhang et al. [22] analysed the effect of admixtures on the behaviour of pervious concrete and concluded that mineral admixtures effectively improved the strength rather than the pore characteristics. They also found that the anti-stripping property was improved by including mineral admixtures and worsened by adding a superplasticiser.

All of these studies on pervious concrete have mainly highlighted the influence of various factors such as aggregate gradation, water–cement ratio, fibres, and admixtures on the mechanical and hydraulic properties of pervious concrete. A few studies were also carried out on cylindrical specimens to evaluate the effect of compaction on the properties of pervious concrete, and these concluded that the compaction methods highly influenced the strength properties [23,24,25]. However, the impact of the aggregate size with different compaction levels on pervious concrete has not yet been investigated in detail. Hence, it is desirable to know the factors, such as compaction and aggregate size, which influence the mechanical and hydraulic properties of the pervious concrete. In addition, the predominant factor that controls the different properties of pervious concrete needs to be identified to improve its performance in various applications. This research is significant because it highlights the important factor which controls the various properties with different compaction levels and aggregate sizes. This study also focuses on achieving the optimal mixture proportion for pervious concrete that will improve its mechanical properties while still maintaining its high hydraulic conductivity. In this study, cubic and cylindrical samples were compacted using a compaction device developed in the laboratory. The outcomes of this research can facilitate the choosing of the right aggregate size and ingredients with proper compaction based on the requirements of the application of the pervious concrete. Four compaction levels and three aggregate sizes were adopted to investigate their influence on the strength and hydraulic properties of pervious concrete, and the test results were reported and analysed.

2. Experimental design

2.1. Materials

The main raw materials required for pervious concrete are cement, coarse aggregate, and water. Hence, pervious concrete is also termed a no fines concrete. Cement of the type CEM II/B-M (S-LL) with a grade of 42.5 N, pursuant to HRN EN 197-1:2012 [26] and obtained from CEMEX Hrvatska d.d., Croatia, was used in this research. The chemical and physical properties of the cement used are shown in

Table 1. Chemical composition of CEM II/B-M (S-LL) 42.5 N.

CaO	SiO2	Al2O3	SO3	Fe2O3	MgO	Insoluble Residue	Loss on Ignition
59.17%	20.22%	5.35%	3.00%	2.35%	2.50%	0.25%	5.00%

and

Table 2. Physical properties of CEM II/B-M (S-LL) 42.5 N.

Initial Setting Time (Minutes)	Final Setting Time (Minutes)	Soundness (mm)	Specific Gravity (g/cm3)	Standard Consistency (%)	Residue on 45 µm Sieve (%)
190	260	0	3.10	25	11

. Crushed limestone, procured from Alas-Seget quarry, Split, complying with HRN EN 12620:2013 [27], was used as the aggregate. Three different aggregate fraction sizes of 0/4 mm, 4/8 mm, and 8/16 mm were used to study their effect on pervious concrete. The properties of the aggregates are listed in

Table 3. Properties of aggregates.

Properties	0/4 mm	4/8 mm	8/16 mm
Saturated surface dry density	2620 kg/m3	2630 kg/m3	2620 kg/m3
Water absorption	1.4%	1.3%	1.1%
Crushing value	-	24.3%	24.3%
Shape index	-	10	15
Flakiness index	-	13	12

. Potable water obtained from the laboratory was used for the manufacturing and curing of the pervious concrete. Isoflow, a polycarboxylate ether-based superplasticiser (SP), and Isosphere, an air-entraining agent (AEA) manufactured by CEMEX, were also used in this investigation.

2.2. General Philosophy and Mixture Proportioning of Pervious Concrete

The mix proportioning for the pervious concrete was carried out as per the requirements of ACI 522R-10 [8]. The mix proportions were developed using the trial and error method and then investigated as they related to the anticipated characteristics of the concrete. The design principle adopted for the present study is given in Figure 1. Initially, the aggregates in saturated surface dry condition were compacted in three layers, according to ASTM C29 [28], in a cylindrical metal container, as represented in Figure 1a. The top surface of the container was levelled, and water was poured into the container until it reached the top level to measure the voids percentage in the aggregates using known volume and mass. The voids percentage obtained from the 4/8 mm aggregates was found to be 41.66%. Similarly, the voids percentage was measured by replacing the 4/8 mm aggregates with 8/16 mm aggregates at different levels, and the obtained results are shown in Figure 2. It can be observed that the maximum voids percentage was achieved without the blending of aggregates. The cement paste volume was calculated in order to partially fill these voids and achieve the targeted porosity of the concrete (Figure 1b). The same principle was followed with different gradations of aggregates for the different mixtures considered in this study.

Table 4. Mix proportions of pervious concrete.

Mix ID	Porosity (%)	Cement (kg/m3)	Aggregate (kg/m3)			Water (kg/m3)	W/C Ratio	SP (kg/m3)	AEA (kg/m3)
			0/4 mm	4/8 mm	8/16 mm
PC1	20.0	339.26	-	1530.66	-	108.56	0.32	-	-
PC2	20.0	339.26	-	-	1524.84	108.56	0.32	-	-
PC3	16.1	321.47	146.42	1464.41	-	122.16	0.38	-	-
PC4	17.5	339.26	-	796.95	796.95	108.56	0.32	-	-
PC5	15.0	417.07	-	1530.66	-	133.46	0.32	-	-
PC6	15.0	417.07	-	-	1524.84	133.46	0.32	-	-
PC7	20.8	359.00	-	1530.66	-	90.46	0.25	3.05	-
PC8	20.8	359.00	-	-	1524.84	90.46	0.25	3.05	-
PC9	13.7	355.97	146.42	1464.41	-	135.27	0.38	-	-
PC10	13.5	400.57	-	796.95	796.95	128.18	0.32	-	-
PC11	13.5	400.57	-	796.95	796.95	128.18	0.32	-	0.30

summarises the various mix proportions and the ingredients employed in this work.

2.3. Mixing, Casting, and Compacting Procedures

The concrete mixture was prepared in a tilting drum concrete mixer. Initially, the aggregates in saturated surface dry condition were mixed with one-third of the required water and mixed thoroughly for 2–3 min to ensure that the surface of the aggregates was entirely coated with water. The cement was then added to the aggregates, followed by the leftover water with chemical admixtures, as per the requirements, and mixed for another 3–4 min to achieve a uniform mixture. The fresh concrete was poured into the respective moulds in three layers and compacted.

In this research, four compaction methods were performed. The first compaction method was in accordance with HRN EN 12390-2:2019 [29], using a standard tamping rod. The other compaction methods were performed with a compaction device developed at the laboratory for cubic and cylindrical specimens. It consisted of two parts, namely the shell and the rammer, as shown in Figure 3. The device was made of steel and designed in such a way that the bottom of the shell fit perfectly over the respective moulds. The rammer plate was designed to be smaller than the shell, with an offset of 5 mm on all sides to ensure free fall over the fresh concrete. The different compaction levels were achieved by dropping the rammer repeatedly from a definite height of 475 mm. By dropping the rammer 5, 15, and 25 times, producing an equivalent energy of 99.3 KJ/m³, 298.1 KJ/m³, and 496.9 KJ/m³, respectively, 3 levels of compactions were obtained. The maximum number of drops was limited to 25 since the aggregates were crushed after that due to the higher impact energy. The specific energy can be calculated using the given Equation (1). The weight of the rammer was fixed at 4.8 kg for cubes and 2.23 kg for cylinders to produce the same compaction energy for the respective number of drops.

After performing the respective compaction method, the samples were levelled and covered with a thin plastic sheet to maintain the moisture level for 24 h. After 24 h, the samples were removed from the mould and then immersed in water at a controlled temperature of 20 ± 2 °C until the day of testing.

$$\mathrm{Specific} \mathrm{Energy}=\frac{M\times h\times n\times N }{V_m}$$

(1)

where $M$ is the weight of the rammer (kg), $h$ is the height of the fall (m), $\mathrm{n}$ is the number of concrete layers, $N$ is the number of blows per layer, and $V_m$ is the volume of the mould (m³).

2.4. Test Methods

2.4.1. Tests on Fresh Concrete

ACI 522R-10 [8] characterises pervious concrete as near-zero slump concrete, which means that the conventional slump cone test [30] cannot define the scale of workability properties of pervious concrete. This shortcoming has resulted in some new approaches for obtaining effective details on stiff concrete [31]. One useful method is to observe a the visual appearance of the pervious concrete. Tennis et al. [32] recommended a method to estimate the stiffness of pervious concrete by performing a ball test. They stated that a ball made from fresh pervious concrete should not crumble into pieces if a proper water/cement (W/C) ratio has been used. It has also been noted that a high W/C ratio will leave cement slurry stuck on the hands. The visual observation methods have some serious shortcomings since these tests do not provide specific values, are based on personal predictions, and are subjective. Hence, in this research work, additional efforts were made to study the consistency of pervious concrete by performing inverted slump [33] and compacting factor tests [34], as shown in Figure 4.

2.4.2. Tests of Mechanical Properties and Abrasion Resistance

Compressive strength (CS) and split tensile strength (SS) tests were performed to examine the mechanical characteristics of all the mixtures considered in this study. The compressive strength test was carried out on the 150 mm cubic samples using a digital compressive testing machine of 3000 kN capacity, in accordance with HRN EN 12390-4:2019 [35]. The split tensile strength test was performed on the 100 mm diameter × 200 mm height cylindrical specimens by applying the load transversely on the sample, in accordance with HRN EN 12390-6:2010 [36]. The abrasion resistance (AR) of the hardened pervious concrete was examined using a Cantabro test performed in a Los Angeles abrasion machine, as shown in Figure 5. This test was performed on the cylindrical specimens with a height and diameter of 100 mm to evaluate the weight loss due to abrasion. The test was carried out in a Los Angeles abrasion machine without adding steel ball charges, and the observations were made after 100 revolutions, 200 revolutions, and 300 revolutions [37].

2.4.3. Tests on Porosity and Hydraulic Conductivity

Porosity (P) and hydraulic conductivity are the two essential characteristics of pervious concrete. In this research work, the porosity and hydraulic conductivity of the 100 mm diameter × 200 mm height cylindrical specimens were tested. The samples were immersed in water for 24 h and then surface-dried before the test was performed. The porosity of the pervious concrete sample was measured based on the recommendations of HRN EN 12390-7:2019 [38] using the method of immersing the samples in water, as shown in Figure 6a, and measuring the dimensions. The effective porosity of the concrete is measured using Equation (2).

$$\mathrm{Porosity}, P=\left(1-\left[\frac{W_2-W_1}{{\rho }_w V}\right]\right) \times 100\%$$

(2)

where $W_1$ is the weight of the sample immersed in water (g), $W_2$ is the saturated surface dry weight of the sample (g), ${\rho }_w$ is the density of the water (1 $\frac{\mathrm{g}}{\mathrm{c}{\mathrm{m}}^3})$, and $V$ is the volume of the sample (cm³).

Hydraulic conductivity is defined as the ability of water to flow through the pore structure of the pervious concrete. It is highly dependent on the pore structure of the material and on the fluid characteristics. This work evaluates hydraulic conductivity using the falling head method, in accordance with ASTM D5084 Method C [39]. The test setup assembled at the laboratory for testing the hydraulic conductivity is given in Figure 6b. The sample was placed in position and completely sealed using duct tape, as shown in the figure. The samples were kept in a fully saturated condition before testing, and care was taken to ensure that no water leaked through the sides of the specimen. The hydraulic conductivity (K) according to the falling head method was calculated using Equation (3) [40].

$$K=\frac{\mathrm{L} a}{\mathrm{A} t}\ln \frac{{\mathrm{H}}_{\mathrm{i}}}{{\mathrm{H}}_{\mathrm{f}}}$$

(3)

where $\mathrm{L}$ is the length of the concrete sample (mm), $a$ is the cross-sectional area of the tube above the sample (mm²), $\mathrm{A}$ is the cross-sectional area of the sample (mm²), ${\mathrm{H}}_{\mathrm{i}}$ is the initial height of the water level (mm), ${\mathrm{H}}_{\mathrm{f}}$ is the final height of the water level (mm), and $t$ is the duration of the water head from ${\mathrm{H}}_{\mathrm{i}}$ to ${\mathrm{H}}_{\mathrm{f}}$ (s).

3. Results and Discussion

3.1. Properties of Fresh Pervious Concrete

During the visual inspection, all of the concrete mixtures appeared to have a metallic shine, and the aggregates were completely coated with the cement paste. This appearance confirmed that the cement content was adequate for the quality binding of the aggregates. All of the concrete mixtures considered in this work could be formed into a ball using one’s hands, as per the recommendations of Tennis et al. [32]. This visual observation ensured that the concrete possessed the quality and consistency required for a pervious concrete. The inverted slump test and compacting factor test results are presented in

Table 5. Fresh concrete properties.

Mix ID	Slump (mm)	Compacting Factor	Fresh Concrete Density (kg/m3)
			Tamping Rod	5 Drops	15 Drops	25 Drops
PC1	180	0.814	1848	1731	1867	1897
PC2	180	0.804	1792	1692	1808	1846
PC3	160	0.923	1884	1851	1961	1982
PC4	160	0.785	1846	1740	1881	1942
PC5	140	0.797	1962	1837	1968	2011
PC6	145	0.777	1859	1773	1884	1909
PC7	180	0.842	1880	1788	1888	1931
PC8	175	0.849	1769	1729	1819	1870
PC9	155	0.883	1895	1810	1902	1972
PC10	155	0.784	1877	1785	1885	1960
PC11	160	0.805	1897	1808	1920	1989

. The tests were carried out with the standard compaction method using the tamping rod. A light shake was given during the test if the concrete became stuck in the cone. It was also observed that a too-stiff concrete mix blocked the cone, resulting in a fresh concrete consistency with which it was difficult to work. Additionally, a high W/C ratio resulted in the flow of cement paste through the bottom of the inverted cone. From the quantitative inverted slump test results, it can be observed that the workability was lower for the mixtures with higher cement content and a lower void percentage. This behaviour shows that the fine particles influence the workability of the pervious concrete, rather than the aggregate size [41]. The compacting factor test results ranged between 0.7 and 0.98, which shows that this test can be satisfactorily performed to measure the workability of pervious concrete [34]. This range confirms the suitability of the concrete for road pavements and normal reinforced concrete with manual compaction [42]. It can be seen from Table 5 that the W/C ratio and binder content plays a vital role in the compacting factor. The values show that the mixture with a high W/C ratio and a low binder content exhibits high compacting factor values compared with the other mixtures considered in this study. The added level of SP in the mix resulted in more or less the same workability with a lesser W/C ratio. The incorporation of AEA slightly improved the workability, which could be visually observed by holding the materials together.

The fresh concrete densities of all the mixtures prepared using the different compaction methods are also tabulated in Table 5. The density of the typical fresh concrete examined in this work ranged from 1692 kg/m³ to 2011 kg/m³. It is clear from the table that the compaction level significantly affects the fresh density of the pervious concrete. The samples compacted with higher specific energy resulted in higher density values. The mixtures compacted with five drops resulted in low-density concrete since the energy produced during this compaction level allowed the ingredients to settle rather than become compact [43].

3.2. Properties of Hardened Pervious Concrete

3.2.1. Hardened Concrete Density

The hardened density of the concrete was obtained from the ratio of mass to the corresponding volume of the samples.

Table 6. Hardened concrete density.

Mix ID	Hardened Concrete Density (kg/m3)
	Tamping Rod	5 Drops	15 Drops	25 Drops
PC1	1898	1762	1897	1928
PC2	1813	1711	1829	1864
PC3	1947	1891	2003	2024
PC4	1902	1766	1910	1975
PC5	2030	1872	2017	2050
PC6	1921	1791	1922	1930
PC7	1970	1835	1958	1984
PC8	1862	1753	1843	1895
PC9	1972	1849	1970	2011
PC10	1973	1850	1975	2020
PC11	1990	1865	1988	2032

shows the hardened density of the samples measured after 28 days of curing. The results from Table 6 confirm that the density of the hardened concrete depended on the energy of compaction, irrespective of the other ingredients in the mixture. The highest density was achieved with 25 drops of the rammer, corresponding to 496.9 KJ/m³. The results also suggest that the density of the concrete produced with 15 drops was more or less equal to that of concrete produced with standard compaction using the tamping rod. This trend shows that the compaction energy produced with the standard tamping rod is similar to the energy produced with 15 drops of the rammer, as performed in this research. The test results show that the hardened concrete density typically ranges from 1711 kg/m³ to 2050 kg/m³.

3.2.2. Compressive Strength

The load transfer through the pervious concrete is highly dependent on the pore characteristics. The cubic samples were tested after 7 days and 28 days of curing, and the results are presented in

Table 7. Compressive strength test results.

Mix ID	7 Days Compressive Strength (MPa)				28 Days Compressive Strength (MPa)
	Tamping Rod	5 Drops	15 Drops	25 Drops	Tamping Rod	5 Drops	15 Drops	25 Drops
PC1	12.30	8.07	12.73	13.44	13.03	9.44	13.56	14.53
PC2	8.95	6.26	10.03	9.88	9.70	7.68	10.58	11.73
PC3	15.14	11.60	15.23	15.31	17.02	13.30	17.88	18.93
PC4	10.60	7.82	11.62	13.06	12.14	8.40	13.14	15.87
PC5	17.55	11.99	19.05	19.69	21.37	14.39	22.16	24.40
PC6	13.48	9.80	11.72	12.35	16.32	10.09	14.13	13.94
PC7	14.63	14.35	18.44	20.47	18.05	15.14	20.41	22.70
PC8	9.66	8.30	12.92	14.72	11.73	8.94	13.43	15.34
PC9	12.02	10.35	12.62	12.74	14.52	11.60	14.84	15.98
PC10	14.38	11.57	15.14	15.82	15.71	12.69	16.02	17.87
PC11	15.14	12.08	16.03	16.85	17.57	14.76	17.88	18.50

. All the values in the table are the mean of three samples tested under the same conditions. It can be clearly observed from the table that the specimens compacted with 25 drops exhibited the highest compressive strength, regardless of the mix proportions considered in this investigation. This confirms that compaction energy greatly affects the compressive strength of pervious concrete. It may also be noted from the results that the increase in compressive strength after 28 days compared with that after 7 days varies from 5% to 25%, unlike normal concrete which can range from 30% to 35% [44]. This large variation may be because the presence of pores in all the concrete samples with different mix proportions was not uniform. The highest compressive strength was achieved with PC5 due to its high cement content and low percentage of pores in the concrete. The concrete samples with larger aggregates produced lower compressive strength than those with smaller aggregates, even with the same cement and pore content. This may be due to the crushing behaviour of the larger aggregates surrounded by pores during testing. The results obtained from the standard compaction method with a tamping rod were close to those obtained with 15 drops of the rammer, as was the trend for concrete density. Hence, further study of the method of compaction using 15 drops of the rammer was skipped in this research. The SP resulted in enhanced strength results. This could be attributed to the fact that the mixture with the SP contained a lower W/C ratio than the other samples considered in this work. The relation between the compressive strength and the hardened concrete density is shown in Figure 7. The figure presents a good correlation between the density and compressive strength for all the aggregate sizes and compaction methods considered in this study.

3.2.3. Splitting Tensile Strength

The results of the splitting tensile strength test are tabulated in

Table 8. Splitting tensile strength test results.

Mix ID	7 Days Splitting Tensile Strength (MPa)			28 Days Splitting Tensile Strength (MPa)
	Tamping Rod	5 Drops	25 Drops	Tamping Rod	5 Drops	25 Drops
PC1	1.45	1.25	2.00	1.70	1.68	2.40
PC2	1.60	1.10	2.00	1.63	1.95	2.00
PC3	1.75	1.65	2.80	2.05	1.85	3.00
PC4	1.85	1.60	2.80	2.03	1.88	3.10
PC5	1.85	1.95	2.90	2.35	2.13	3.13
PC6	2.05	2.40	2.70	2.85	2.53	3.03
PC7	1.25	1.43	2.40	1.40	1.60	2.85
PC8	1.80	1.75	2.60	1.98	2.00	2.88
PC9	2.00	1.93	2.95	2.53	2.51	3.16
PC10	2.45	2.43	3.00	2.53	2.44	3.25
PC11	2.58	2.48	3.06	2.70	2.68	3.28

. The trend of the compaction level test results is similar to that of the compressive strength test results. In the splitting tensile strength test, it was observed that most of the cracks developed between the binder and the aggregates, irrespective of the aggregate size. Here, the effect of the aggregate crushing that occurred during the compressive strength test is not highly significant. The mix proportions with higher binder and lower void content resulted in a better performance. The relation between compressive strength and splitting tensile strength is not as uniform as in conventional concrete. Hence, the equation for obtaining the splitting tensile strength of normal concrete cannot be applied to pervious concrete [12]. This difficulty is due to the non-uniformity of the pore structures in pervious concrete. The splitting tensile strength increases as the hardened density of the samples increases. The addition of AEA to the mixture considerably promotes strength by enhancing the fluidity of the cement paste filling between the aggregates [45].

3.2.4. Abrasion Resistance

Abrasion resistance was investigated for all the mixtures in order to evaluate the durability characteristics of the pervious concrete. The percentage of weight loss during the Cantabro test for every 100 revolutions is tabulated in

Table 9. Cantabro test results.

Mix ID	Weight Loss (%)
	Tamping Rod			5 Drops			25 Drops
	100 Rev.	200 Rev.	300 Rev.	100 Rev.	200 Rev.	300 Rev.	100 Rev.	200 Rev.	300 Rev.
PC1	22.6	34.1	46.7	15.5	42.5	65.4	11.7	19.2	25.5
PC2	35.4	89.0	100.0	62.9	100.0	100.0	12.7	46.9	80.2
PC3	13.0	21.4	29.5	10.0	18.2	25.1	7.0	12.3	18.0
PC4	16.7	31.9	47.1	17.2	46.7	73.5	9.6	15.7	22.0
PC5	9.2	16.3	23.1	9.8	17.2	25.9	6.2	11.1	15.4
PC6	13.0	29.7	91.1	14.7	42.8	85.3	8.7	18.7	36.5
PC7	20.1	41.0	51.9	12.4	22.1	31.5	7.9	13.1	17.8
PC8	16.1	45.6	82.5	34.5	72.8	92.8	10.6	17.3	34.1
PC9	12.3	18.1	25.7	8.8	15.2	21.6	5.7	10.5	15.0
PC10	15.2	28.7	42.3	15.3	43.1	71.4	8.2	13.1	19.5
PC11	14.7	26.8	38.5	13.4	41.6	68.2	7.1	11.6	18.0

. From the table, it can be understood that the weight loss of the samples was indirectly proportional to the compaction energy applied to the samples during the casting process. It can also be observed that the aggregate size impacts weight loss significantly, as is depicted in Figure 8. Typical images of the tested samples are given in Figure 9. Figure 8 shows that the weight loss was high for the samples with larger aggregate sizes [46]. This may be because the samples with larger aggregates are more susceptible to the impact force generated during the revolutions than to abrasion [37]. The weight loss of the samples with the same aggregate size was further influenced by the binder content, which can be noted from PC2 and PC6. The mixture with a small aggregate size and high binder content (PC5) exhibited the highest abrasion resistance. This mode of behaviour is similar to that seen in the compressive strength tests, in which the aggregate size was the prominent factor affecting the test results.

3.3. Porosity and Hydraulic Conductivity

The porosity measurement test results are provided in

Table 10. Porosity and hydraulic conductivity test results.

Mix ID	Designed Porosity (%)	Actual Porosity (%)			Hydraulic Conductivity (mm/s)
		Tamping Rod	5 Drops	25 Drops	Tamping Rod	5 Drops	25 Drops
PC1	20.0	25.7	27.2	18.7	18.0	26.7	9.7
PC2	20.0	27.9	27.3	21.0	22.8	35.2	13.9
PC3	16.1	20.6	23.0	12.9	12.7	18.9	3.5
PC4	17.5	22.6	25.1	15.3	16.5	23.8	9.1
PC5	15.0	18.5	23.1	11.2	10.0	16.3	6.0
PC6	15.0	19.1	21.8	14.4	11.2	17.9	9.9
PC7	20.8	27.5	29.2	18.7	23.3	29.3	12.0
PC8	20.8	26.8	29.6	18.3	26.0	34.5	13.8
PC9	13.7	14.9	19.6	12.3	11.3	16.6	5.6
PC10	13.5	17.0	19.3	12.1	13.3	19.5	6.6
PC11	13.5	17.0	19.3	12.1	11.9	17.5	5.9

. In this research, all the mixture proportions were derived from a design principle based on the voids percentage in the aggregates. Hence, the designed porosity was compared with the actual porosity obtained from the test results. The results show that the designed porosity was always lower than the actual porosity of the concrete samples compacted with the tamping rod. The role of aggregate size was insignificant in the actual porosity compared with the designed ones. This discrepancy could be due to the effect of the dense binder paste filling around the aggregates, which resulted in an increased volume of aggregates [47]. On the other hand, the samples compacted with 5 drops exhibited much higher porosity since the ingredients were not compacted adequately with the applied energy [43]. In comparison, the highly compacted samples obtained with 25 drops showed lower porosity values than those intended.

The hydraulic conductivity values of all the samples tested using the falling head method are presented in Table 10. The difference between the initial $({\mathrm{H}}_{\mathrm{i}})$ and final $({\mathrm{H}}_{\mathrm{f}})$ water levels were fixed at 30 cm for all the samples, and the final $({\mathrm{H}}_{\mathrm{f}})$ level was 6 cm above the sample. The typical hydraulic conductivity values ranged from 3.5 mm/s to 35.2 mm/s for the samples considered in this investigation, and this satisfies the criteria for a pervious concrete [7]. The results clearly show that the compaction level is one of the critical factors affecting the hydraulic properties of the concrete samples. The compaction process expels the entrapped air from the concrete and distributes the aggregate particles to reduce the hydraulic conductivity of the hardened concrete. The hydraulic conductivity of the pervious concrete improves as the aggregate size increases from 4/8 mm to 8/16 mm, but not significantly. This may be due to the larger pore size formed by larger aggregates when compared with smaller aggregates [48]. It can also be observed from the results that the designed porosity values influenced the obtained results. The specimens with higher porosity values exhibited higher hydraulic conductivity, and this shows that the porosity is directly proportional to the hydraulic conductivity of the pervious concrete [21].

4. Conclusions

This experimental investigation mainly focused on the effects of aggregate size and compaction method on the mechanical and hydraulic properties of pervious concrete. From the obtained test results, we have arrived at the following conclusions for limestone aggregates:

• The workability of the pervious concrete was significantly affected by the binder content rather than the aggregate size. The results obtained from the compacting factor test are within the range of 0.70–0.98, which shows that this test method is suitable for testing fresh pervious concrete and can be used for pavement and other reinforced concrete works.
• The applied compaction energy influenced the density of the concrete; the density increased by around 9% as the applied specific energy increased from 99.3 KJ/m³ to 496.9 KJ/m³.
• The compressive strength was notably affected by the aggregate size; the compressive strength of the concrete samples with 8/16 mm aggregates exhibited 20–40% lower values than the samples with 4/8 mm aggregates.
• The test results exhibited a good correlation between the strength and the density of the concrete samples. The strength increased as the density of the pervious concrete sample increased.
• The cement and void contents played a more significant role in determining the splitting tensile strength than the size of the aggregates. The improvement in strength ranged from 25% to 75% as the cement content increased by 35–78 kg/m³.
• The concrete with 8/16 mm aggregates possessed lower abrasion resistance (90–220%) compared with the samples produced with smaller aggregates. Hence, the effect of aggregate size on the abrasion resistance of the pervious concrete is more prominent.
• The measured porosity of the concrete samples obtained using standard compaction was always less than the designed porosity, and this variation ranged from 10% to 40%.
• Even with high compaction, the samples exhibited hydraulic conductivity within the accepted range (3.5–35.2 mm/s). Hence, a designed porosity of 13.5% and above is recommended to achieve sufficient hydraulic conductivity for pervious concrete.
• The experimental investigation concluded that the utilisation of smaller aggregates (4/8 mm) in pervious concrete significantly improved the strength properties and simultaneously slightly reduces the hydraulic conductivity. The reduction in hydraulic conductivity of the samples with 4/8 mm aggregates was less significant than that observed in the samples produced with larger aggregates (8/16 mm) and was highly influenced by the porosity of the pervious concrete.