Influencia De Los Agregados De Escoria De Hormigón Y Acero Reciclados En Las Propiedades De Mezcla Asfáltica En Caliente.pdf

Construction and Building Materials 185 (2018) 684–696

Contents lists available at ScienceDirect

Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat

Influence of recycled concrete and steel slag aggregates on warm-mix asphalt properties F.C.G. Martinho a, L.G. Picado-Santos b, S.D. Capitão c,b,⇑ a

FM Consult & Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1049-001 Lisboa, Portugal CESUR, CERIS, Instituto Superior Técnico, Universidade de Lisboa, Av. Rovisco Pais, 1049-001 Lisboa, Portugal c Instituto Politécnico de Coimbra, Instituto Superior de Engenharia de Coimbra, Rua Pedro Nunes, 3030-199 Coimbra, Portugal b

h i g h l i g h t s
WMA blends with a chemical additive showed low resistance to permanent deformation.
The WMA blend with 60% of RCA and organic wax was the best performant material.
Adding 30% EAFS or 60% RCA did not notably influence mechanical performance of WMA.
Fatigue resistance and stiffness modulus of WMA with RCA or EAFS were satisfactory.
Incorporating EASF or RCA in WMA was not a problem vis-a-vis water sensitivity.

a r t i c l e

i n f o

Article history: Received 15 March 2018 Received in revised form 4 June 2018 Accepted 8 July 2018

Keywords: Additives Electric arc furnace slag Performance properties Recycled concrete aggregate Resistance to fatigue Resistance to permanent deformation Stiffness modulus Warm-mix asphalt

a b s t r a c t This paper focuses on the comparison of the mechanical performance of three warm-mix asphalt (WMA) blends with recycled concrete aggregate (RCA) or electric arc furnace slag (EAFS) as substitutes of part of the aggregate. A conventional hot mix asphalt (HMA) and a WMA without by-products were used as references. The evaluation was carried out in a laboratory by means of testing specimens taken from experimental pavement sections built in real production, laying and compaction circumstances. Performance testing included wheel-tracking tests, four-point bending tests, and indirect tensile strength to assess water sensitivity. An organic wax and a chemical surfactant were applied to lower handling temperatures of the WMA under study. Apart from the HMA and the WMA used as references, the study evaluated the influence of introducing 60% of RCA or 30% of EAFS into the WMA blends as substitutes of the aggregate. The obtained results for the WMA with by-products tested allowed to conclude that the introduction of EAFS or RCA into the WMA blends increases Marshall stability and may increase or decrease resistance to rutting. Findings also showed that stiffness modulus is somewhat reduced and fatigue resistance does not change significantly. Additionally, water sensitivity is slightly reduced. Comparing these results with the performance observed elsewhere for WMA without by-products revealed that the addition of RCA or EAFS is satisfactory. The construction of experimental sections used conventional batch plant, paver and compactors without any noticeable technical problems. Ó 2018 Elsevier Ltd. All rights reserved.

1. Introduction Warm-mix asphalt (WMA) are asphalt mixtures generally handled at lower temperatures than conventional asphalt concrete. Their constituents are typically mixed together at temperatures varying from 100 to 140 °C [1,2]. This also allows lower compaction temperatures and longer haulage distances and additional ⇑ Corresponding author at: Department of Civil Engineering, Instituto Polite´cnico de Coimbra, Instituto Superior de Engenharia de Coimbra, Rua Pedro Nunes, 3030- 199 Coimbra, Portugal. E-mail address: [email protected] (S.D. Capitão). https://doi.org/10.1016/j.conbuildmat.2018.07.041 0950-0618/Ó 2018 Elsevier Ltd. All rights reserved.

time to carry out construction activities [3]. The authors have summarised the production technology and properties of WMA in a previous publication [3], in which they presented a detailed description of additive based WMA and foamed bitumen techniques. The idea underlying the project partially described in this paper is contributing to improve knowledge on more sustainable asphalt mixtures, potentially suitable to substitute hot mix asphalt (HMA) in many paving situations. The chosen way to achieve this goal was studying the reuse of recycled concrete and steel slag as substitutes of part of natural aggregates to produce WMA, which were manufactured by adding additives to the blend. The use of those

F.C.G. Martinho et al. / Construction and Building Materials 185 (2018) 684–696

by-products is aligned with the goals of eliminating waste going to landfill, contributing to the valorisation of by-products and, thus, increasing circular economy. In addition, decreasing handling temperatures is also recognized as an important way of reducing emissions associated with paving construction and maintenance [3]. Although WMA show suitable volumetric properties and mechanical performance, there is some weaknesses referred to in the literature that can be found elsewhere [3,4]. Hence, this introduction presents below a brief summary of the issues associated to the reuse of recycled concrete aggregates (RCA) and steel slag aggregates (SSA) in asphalt mixtures found in the literature. Crushing concrete elements produce materials known as RCA. Unlike natural aggregates produced in quarries, RCA have cement mortar around the surface of the coarse natural aggregates. This highly porous mortar is responsible for the high values of porosity and water absorption of RCA [5,6]. Therefore, to assure enough binder to involve the aggregate particles, the use of RCA generally requires higher bitumen content in asphalt concrete (AC) because part of the binder is absorbed by the aggregate [5,6]. Because fine particles of RCA have much more voids than coarse aggregates, porosity of blends increases if fines are added to the blend [6]. In addition, RCA is formed by angular and rough textured coarse and fines particles, providing high Marshall stability and small flow values [5,6]. This trend was not observed in a number of studies, apparently because some of them used coarse RCA while others also added fine fractions [7]. In addition, replacing part of the aggregates by RCA generally degrades the resistance to moisture [5–9]. Likewise, RCA generally lead to AC with lower stiffness [5–9] mostly when coarse RCA is used, and leads to higher stiffness mainly if RCA is included as the fine fraction [7,10]. When the blends are submitted to ageing by heating them in an oven before compaction, the level of binder absorption increases, which contributes to increase stiffness [10]. The resistance to permanent deformation is usually reported to increase when RCA are incorporated into the blends [5–8,11]. Nevertheless, several studies, referred to in [7], revealed the opposite trend for several reasons: more RCA requires higher binder content; coarse and fine fractions of RCA tend to decrease resistance to permanent deformation compared to those with coarse fraction only, amongst others. The fatigue resistance of HMA is adversely affected by the introduction of RCA, decreasing fatigue resistance as the amount of RCA increases [12]. On the contrary, several studies, reported by Pasandín & Pérez [7], reveal that RCA tend to improve resistance of HMA to fatigue. Steel slag is a by-product of the steel industry, produced from impurities separated from molten steel, which solidifies after cooling (generally by adding water) [13,14]. There are three types of slags: basic oxygen furnace (BOF), electric arc furnace (EAF) and blast furnace (BF) [14,15]. EAF slag (EAFS) was used in the present study. This type of slag is obtained from processes of melting recycled scrap to produce different types of steel. A detailed review on the use of EAFS in asphalt mixes can be found in [15]. EAFS is formed by a several chemical components, such as iron oxides, lime, silica, magnesia and alumina and other minor components [15]. EAFS are used as aggregates in asphalt mixtures because their physical and mechanical properties are usually suitable for that purpose. In general, the amount of fine particles is low and the coarser ones have good shape, rough texture and high angularity. Although EAFS have relatively high density (3200–3800 kg/m3), they exhibit elevated porosity and, therefore, higher water and binder absorption than typical natural aggregates. Nevertheless, expansion potential of asphalt mixes with EAFS in water is negligible [13]. Also, leaching of EAFS when used in asphalt mixes is very low [16], showing practically no health risk [15]. EAFS have

685

generally good mechanical resistance and they reveal suitable adhesion with bitumen [15]. The asphalt mixtures that incorporate EAFS as aggregate have usually bulk densities 15 to 20% higher than similar asphalt mixes without EAFS. Because of manifest angularity of the EAFS particles, the porosity of asphalt concrete made with EAFS is also higher [15]. Generally speaking, using EAFS to substitute part of the natural aggregate tend to improve mechanical behaviour of the resulting asphalt mixes. This tendency is referred to in the literature for Marshall stability, indirect tensile strength, stiffness, fatigue and resistance to permanent deformation. This improved behaviour is general attributed, on the one hand, to a better interlock between aggregate particles as well as the roughness of steel slag assuring a better adhesion binder-aggregate [13] and, on the other hand, to a higher resistance to heavy loads and shear stress [15]. Most of the authors state that the best results were obtained for a partial substitution of natural aggregates by EAFS (around 30%). However, there is a number of cases reported in which the level of substitution was greater [15,17] and the mechanical behaviour remained improved [15]. In what concerns water sensitivity, some divergent results were found in the literature. Passeto and Baldo [18,19], for instance, mentioned in [15], consistently observed that asphalt mixtures with coarse and fine EAFS aggregates revealed higher resistance to water damage in indirect tensile strength tests. They attribute this behaviour to either a thicker bitumen film involving the aggregates in mixtures with EAFS due to a higher binder content or a good slag-binder adhesion that protect the particles against water damage. On the contrary, some studies [20] reveal that water sensitivity was higher in mixtures with coarse slag aggregates as compared to similar mixtures with limestone aggregates. The observed performance seems to be related with a worse affinity of the slag with binders than that of limestone aggregate. The information collected in the literature shows some opportunities and challenges related with the use of RCA and EAFS to produce and lay WMA incorporating these by products in substitution of part of natural aggregates. It’s necessary, on the one hand, to evaluate the expected volumetric and mechanical properties of that type of asphalt mixtures and, on the other hand, to verify the technological viability of these bituminous products in real production conditions. This paper summarises and examines the results found in a study involving the evaluation of specimens produced in laboratory and sawed from experimental sections constructed in real construction circumstances with conventional equipment. Chemical and organic additives in the form of pellets were used to allow lowering production and compaction temperatures. The main objective of the study was to evaluate the effect of using warmmix asphalt technologies, together with by-products aggregates (EAFS and RCA), on mechanical properties of asphalt concrete.

2. Materials 2.1. Aggregates, RCA, EAFS, bitumen and additives Three fractions of limestone crushed rock, RCA and EAFS were used as aggregates to produce the warm-asphalt mixtures produced and evaluated throughout the study. In the early stage of the project, two blends of limestone aggregates were used to produce the reference asphalt mixtures, one HMA with 35/50 paving grade bitumen and three WMA with different additives to reduce handling temperatures. These compositions were characterized in the Marshall test and in terms of volumetric properties. The same was carried out to the six additional WMA blends in which

686

F.C.G. Martinho et al. / Construction and Building Materials 185 (2018) 684–696

summarizes the aggregates’ characteristics measured according to the Portuguese specifications. The EAFS and RCA aggregates were examined in terms of their potential to release dangerous compounds, according to EN 12457-4 [21] and EN 1744-3 [22]. The obtained analysis revealed that both used by-products easily satisfy all the requirements indicated in the European regulations for leaching limits [23]. Conventional 35/50 paving grade bitumen, with a penetration, at 25 °C, of 45  0.1 mm, and a ring & ball softening point of 56 °C, was the binder used for all the blends evaluated throughout the project. One organic wax (SasobitÒ) and one chemical additive (RedisetTM) were used to lower the handling temperatures of the WMA. The additives were added to the blends in the form of solid pellets.

part of the natural aggregate was replaced by RCA, EAFS or both. These blends were mixed and compacted in the laboratory to choose, based on the properties achieved, those that would be applied in the trial sections to be built in the later stage of the project. The blends selected were mixed in asphalt plants, laid and compacted in trial sections of pavement in real production conditions. Furthermore, taking into account the laboratory results, the three chosen blends suffered some adjustments prior to their use in the field. The grading Portuguese envelope for dense graded mixtures (AC 20 base) was the reference to derive the aggregate blends. Fig. 1 shows the blends’ grading curves applied in the trial sections as well as the grading curves of RCA and EAFS, and two images to show the appearance of the two by-products. Table 1

100

80

EAFS 0/16 mm

89.0

RCA 0/20 mm

70 60

60.0

57.02

50 40 32.0

32.01

30

22.0

20

Particles size (mm)

100 90

TWc(EAFS)3550

80

TWc(RCA)3550 = TWw(RCA)3550

70

Grading envelope

100

Particles size (mm)

10

100 0.01

10

1

0.1

0.01

9.0

1.2

7.98

1.76

0

1

10

% of material passing

100.0

RCA (Recycled Concrete Aggregate)

87.24

0.1

% of material passing

99.84

EAFS (Electrical Arc Furnace Slag)

90

60 50 40 30 20 10 100

10

1

0.1

0.01

0

Particles size (mm)

Fig. 1. Grading curves of the aggregate blends, EAFS and RCA, and general view of EAFS and RCA.

Table 1 Properties of aggregates. Properties

Sand Equivalent (%) Methylene Blue (g/kg) Flakiness Index (%) Los Angeles (%)1 Density after drying (Mg/m3) Water absorption (%)1 Floating material (cm3/kg) Concrete, conc. products and mortars Unbound aggr., natural stone and aggr. treated with hydraulic binders Bituminous materials Masonry elements of clay materials, calcium silicates and cellular concrete; Plastics, rubbers, metals, non-floating wood and stucco

Symbol

SE MBF FI LA

qa WA FL Rc Ru Ra Rb X

Standard

EN 933-8 EN 933-9 EN 933-3 EN 1097-2 EN 1097-6 EN 1097-6 EN 933-11 EN 933-11 EN 933-11 EN 933-11 EN 933-11 EN 933-11

RCA

EAFS

10/20 (mm)

Limestone Aggregates 4/10 (mm)

0/4 (mm)

0/20 (mm)

0/16 (mm)

– – 12.8 27 2.6 1.74 – – – – – –

– – 13.6 – 2.6 1.09 – – – – – –

62 2.3 – – 2.5 0.99 – – – – – –

30 3.7 6.0 43 2.3 6.1 0.0 84.0 9.0 0.7 5.0 0.6

78 0.2 2.0 25 3.6 1.7 0.0 0.0 99.0 0.0 0.0 0.9

1 Note: Requirements of LA and WA24 commonly used in Portugal for natural aggregates to be applied in asphalt mixtures: 1. LA40 and WA24  2 (for base layers); 2. LA20 and WA24  2 (for binder layers).

F.C.G. Martinho et al. / Construction and Building Materials 185 (2018) 684–696

687

Table 2 Composition of the studied blends.

A discussion on the main differences between them and the expected influence on WMA’s mechanical characteristics can be found elsewhere [3]. 2.2. Compositions of blends The information concerning the constituents and compositions of all the studied blends are summarized in Table 2. Although the data presented refers to all the blends studied, the objective of this paper is highlighting the results derived from the blends applied in real production conditions, whose compositions were adjusted based on a previous evaluation of Marshall specimens produced from analogous blends in the laboratory. The rates of additives considered to produce the mixtures followed the manufacturers’ recommendations. With the aim of making easier the identification of the blends, the following key is used throughout the text: L – laboratory blends; T – trial sections of pavement; W – warm-mix asphalt; H – hot-mix asphalt; c – chemical additive; s – surfactant additive; w – wax additive; RCA – recycled concrete aggregates; EAFS – electric arc furnace slag aggregates; D – upper size sieve of the blend (D = 14; 16; 20 mm); bitumen penetration @ 25 °C in 0.1 mm – (10/20 = 1020; 35/50 = 3550; 50/70 = 5070; 160/220 = 160220). A direct match was considered between some of the laboratory mixtures and the blends produced in an asphalt plant, such as LH3530 and TH3550, LWc3550 and TWc3550, LWc(EAFS)3550 and TWc(EAFS)3550, LWc(RCA)3550 and TWc(RCA)3550, LWw (RCA)3550 and TWw(RCA)3550 (Table 2). The remaining laboratory blends LWw3550, LWw(EAFS)3550, LWw(RCA + EAFS)3550 and LWc(RCA + EAFS)3550 were not considered for testing in real production conditions as the observed preliminary characteristics were not considered promising. 2.3. Production processes and handling temperatures Throughout the early stage of the testing program, the production and compaction processes of the blends were established in the laboratory. This was done by moulding cylindrical specimens with (101.6 ± 0.1) mm in diameter. Mixing was carried out according to EN 12697-35 [24] and compaction was accomplished according to EN 12697-30 [25]. The laboratory blends were evaluated in what concerns some volumetric properties as well as in the Marshall test. In the second phase of the project, the blends with better results for those properties were mixed in a batch plant, laid

with a common road paver and compacted with conventional compactors. For further evaluation of mechanical properties of TWc3550, TWc(EAFS)3550, TWc(RCA)3550 and TWw(RCA)3550 in the laboratory, slabs were sawed from the trial sections to produce the needed specimens for mechanical testing. Following the guidelines from the additive manufacturers and found in the literature [3], the WMA blends were mixed to a target temperature of 120 °C whereas the HMA blends were prepared at 165 °C. Compaction in the laboratory was accomplished between 100 and 110 °C for the WMA mixtures and between 120 and 150 °C for the HMA. In what concerns the experimental sections, all the trial sections were compacted in less than 45 min after laying, allowing compaction above the minimum temperature of 80 °C required for the WMA applied.

2.4. Volumetric properties, Marshall stability and flow of mixtures The cylindrical specimens were characterized in terms of Marshall stability and flow as well as the most common volumetric properties: porosity, voids in mineral aggregates (VMA) and voids filled with binder (VFB). These volumetric properties were measured on the laboratory specimens and on the samples collected from the experimental sections, whereas stability and flow were determined for laboratory specimens only (see Table 2: LH3550 to LWc(RCA + EAFS)3550) in the first stage of the project. Fig. 2 shows the average results measured for the aforementioned properties. For ease of comparison, the results of matching blends produced in laboratory and cut from the trial sections pavements are adjacent to each other in the graphs. Fig. 2 shows that the WMA blends LWw3550 and LWc3550 returned low stability values, below the minimum limit currently used in Portugal. Although in terms of stability/flow ratio, all the blends satisfied the minimum of 2 kN/mm typically required for AC 20 base in Portugal, LWw(RCA)3550, LWc(RCA)3550, LWw (RCA + EAFS)3550 and LWc(RCA + EAFS)3550 revealed higher flow values than the Portuguese requirements. Typically, AC 20 has porosity values from 3 to 6%. In this study, blends LWw(RCA) 3550, LWc(RCA)3550, LWw(RCA + EAFS)3550 and LWc(RCA + EAFS)3550 revealed low porosity. As shown in Table 2, no similar materials to LWw3550, LWw (EAFS)3550, LWw(RCA + EAFS)3550 and LWc(RCA + EAFS)3550 were applied in the experimental sections. Based on the results and some practical reasons, such as the blends’ workability

688

F.C.G. Martinho et al. / Construction and Building Materials 185 (2018) 684–696

Porosity (%)

VMA (%)

9% 8% 7% 6% 5% 4% 3% 2% 1% 0%

LH3550 TH3550 LWw3550 LWc3550 TWc3550 LWw(RCA+EAFS)3550 LWc(RCA+EAFS)3550 LWc(EAFS)3550 TWc(EAFS)3550 LWw(EAFS)3550 LWc(RCA)3550 TWc(RCA)3550 LWw(RCA)3550 TWw(RCA)3550

LH3550 TH3550 LWw3550 LWc3550 TWc3550 LWw(RCA+EAFS)3550 LWc(RCA+EAFS)3550 LWc(EAFS)3550 TWc(EAFS)3550 LWw(EAFS)3550 LWc(RCA)3550 TWc(RCA)3550 LWw(RCA)3550 TWw(RCA)3550

18% 17% 16% 15% 14% 13% 12% 11% 10%

Stability (kN) 6 5 4

MAX: Portuguese specificaons limit

3 2 1

LWc(RCA+EAFS)3550

LWw(RCA+EAFS)3550

LWc(EAFS)3550

LWw(EAFS)3550

LWc(RCA)3550

LWw(RCA)3550

LWc3550

LWw3550

LWc(RCA+EAFS)3550

LWw(RCA+EAFS)3550

LWc(EAFS)3550

LWw(EAFS)3550

LWc(RCA)3550

0

LWw(RCA)3550

LWc3550

LWw3550

LH3550

Min.

Flow (mm) MAX: Portuguese specificaons

LH3550

16 14 12 10 8 6 4 2 0

Fig. 2. Average values of Marshall stability and flow, and volumetric properties of the studied blends.

observed and the search for better results in the performance evaluation stage, LWw(RCA)3550, LWc(RCA)3550 and LWc(EAFS)3550 were the compositions selected to the later stage of the project. The volumetric properties were somewhat different in the specimens cut from the trial sections and in the specimens produced in the laboratory. The differences observed can be mainly attributed to compaction method (impact compaction) standardised for the Marshall method, which does not reproduce the field compaction method [26]. 3. Mechanical properties methods and results 3.1. Specimens for testing In the field, the studied asphalt mixtures were mixed in a batch plant, transported in lorries, laid with a paver in the trial sections and compacted with tyre and metallic rollers. Slabs with 300  400  60 mm3 were cut from the trial sections’ pavements to carry out wheel-tracking tests. Some of those slabs were cut again to make prismatic beams with 400x55x55 mm3 to submit to four-point bending tests. Samples of loose mixtures, collected during the construction of pavements trial sections, were used to mould and compact cylindrical specimens with 101.6 ± 0.1 mm in diameter, aiming at submitting them to indirect tensile tests.

3.2. Methods Wheel-tracking (WT) tests, performed according to EN 1269722 [27], were carried out to evaluate permanent deformation performance of the blends. The wheel of WT applies a stress of about 700 kPa and passes over the specimen up to 10,000 times. When a rut depth of 20 mm is attained the tests finish, even if the specified number of cycles is not achieved. The evaluation of permanent deformation resistance of the studied blends was carried out at 50 and 60 °C. Four-point bending tests (EN 12697-26 [28]) were carried out to assess the stiffness modulus and phase angle of the blends. The followed standard considers prismatic beams to perform repetitive four-point bending tests, under controlled displacement conditions (strain level of 50 lm/m) at 20 °C. The applied strain varied with time according to a sinusoidal wave. Each specimen was submitted to loading frequencies of 30, 20, 10, 5, 3 and 1 Hz. Fig. 4 presents the measured stiffness moduli and phase angles for the strain level of 50 lm/m. Evaluation of fatigue resistance of the blends was also performed in four-point bending tests following the procedure described in EN 12697-24 [29], testing eighteen specimens (with a cross section of 55  55 mm2), six per strain level. The tests were carried out at 20 °C in displacement control conditions, applying repetitive sinusoidal loads along time, with a frequency of 10 Hz

689

F.C.G. Martinho et al. / Construction and Building Materials 185 (2018) 684–696

and three strain levels (150, 250 and 350 lm/m). As usual, fatigue failure of specimens was considered the moment in which a 50% loss of the initial stiffness modulus was observed. The evaluation of water sensitivity was carried out based on ITSR – Indirect tensile strength ratio – following the procedures indicated in EN 12697-12 [30] and EN 12697-23 [31]. This evaluation was performed on six cylindrical specimens with (100 ± 3) mm in diameter. They were moulded in the laboratory and compacted by impact in the Marshall compactor. Three of the specimens were firstly conditioned in water at (40 ± 1) °C for 68 to 72 h and then in air at (15 ± 1) °C for a time greater than or equal to 2 h. Finally, indirect tensile strength (ITS) was evaluated in diametrical load compression tests. ITSR is the ratio between ITS obtained for the conditioned specimens (ITSw) and ITS measured to the dry specimens (ITSd) not submitted to the conditioning procedure. 3.3. Evaluation of rutting resistance Fig. 3 shows the results of slope in air, WTSAIR (wheel-tracking slope), and RDAIR (rut depth, in mm), obtained for the two parameters considered to evaluate resistance to permanent deformation, according to EN 12697-22 [27], determined from the time – deformation curves (small device in air). The error bars show the range of tests’ results obtained on two specimens. The results show that permanent deformation resistance of TWc(RCA)3550 (blend with a chemical additive and RCA) is very sensitive to temperature. Although it has a reasonable performance at 50 °C, its behaviour significantly worsened at 60 °C as the tests did not achieved 10,000 loading cycles. Fig. 2 had already shown that LWc(RCA)3550, the matching preliminary blend, revealed a tendency to deform at high temperature in the Marshall compression test. Fig. 3 also shows that the blends produced with chemical additive revealed more temperature sensitivity than the blends produced with organic wax. The low stability obtained in the Marshall test for LWc3550 (with chemical additive and limestone natural aggregate), shown in Fig. 2, could foresee the weak resistance observed for TWc3550, the equivalent reference blend tested in the trial sections.

TWc(EAFS)3550 (WMA with chemical additive and EAFS) had an intermediate performance amongst the blends tested even though it revealed low S/f ratio (2 kN/mm) in the Marshall test. However, the rut depth developed is quite high, suggesting a weak resistance to permanent deformation. The reference HMA (TH3550) and TWw(RCA)3550 (WMA with organic additive and RCA) were the best performant blends in terms of resistance to permanent deformation. TWw(RCA) 3550 and TWw(RCA)3550 (chemical additive and RCA) have the same types and contents of bitumen and aggregates but not the same additive to reduce handling temperature. This seems to be the most important factor in what concerns permanent deformation resistance at 60 °C in this case. These results show that despite the Marshall stability and flow of LWc(RCA) 3550 and LWw(RCA)3550 were almost the same, the corresponding blends applied in the test site, respectively, TWc(RCA) 3550 and TWw(RCA)3550, revealed very different permanent deformation resistance. Looking at the WT results it can be stated that the effect of including EAFS as part of the aggregate in TWc(EAFS)3550 is inconsistent in terms of permanent resistance performance. Despite the level of rut depth reached at 60 °C is slightly lower for TWc(EAFS)3550 than for TWc3550 (the reference WMA), the first one revealed weaker resistance at 50 °C. Also, TWw (RCA)3550 had an increase in rutting performance when compared with TWc3550. These results reveal that adding organic wax and RCA as part of the aggregate contributed to increase permanent deformation performance of the WMA, allowing a performing level similar to HMA. Despite there are no requirements established for the WT test results in Portugal, the high levels of rut depth reached by TWc (EAFS)3550 and TWc(RCA)3550 at 50 °C seem to be incompatible with a suitable permanent deformation resistance for the climatic conditions in Portugal, where the pavements’ binder layers can achieve up to 50 °C in the hottest regions in summer. 3.4. Stiffness Fig. 4 presents WMA with EAFS and RCA in green. The reference WMA, TWc3550, with chemical additives, had stiffness moduli

WTSAIR (mm/103 cycles) 4.0

RDAIR (mm) 25 50°C

50°C

3.5

20

60°C

3.0

20

3.10

60°C

17.09 14.22

2.5

13.1

15

2.0 10

1.5

6.62

0.91 0.73

1.0

0.71

0.21

0.25

0.34

0.5

0.23

0.14

3.97

5

6.9

5.13 3.59

4.09

0.25

c(E A

TW

TW

c3 55 0 F TW S )3 c(R 550 C T W A )3 55 w( 0 RC A) 35 50 TH 35 50 c(E A

TW

TW

c3 55 0 FS )3 TW c(R 550 C TW A)3 55 w( 0 RC A) 35 50 TH 35 50

0

0.0

Note: For TWw(RCA)3550, WTSAIR was determined for a number of loading cycles higher than 2000. Fig. 3. Wheel-tracking test results: WTSAIR and RDAIR (50 & 60 °C). Note: For TWw(RCA)3550, WTSAIR was determined for a number of loading cycles higher than 2000.

690

F.C.G. Martinho et al. / Construction and Building Materials 185 (2018) 684–696

100 TWc35 50

TWc(EAFS)3550

TWc35 50

TWc(EAFS)3550

TWw(RCA)3550

TWw(RCA)3550

TWw(RCA)3550

TWw(RCA)3550

Phase angle (degree)

Stiffness modulus (MPa)

10000

TH3550

1000

TH3550

10 1

5

25

1

5

Frequency (Hz)

25

Frequency (Hz)

Fig. 4. Stiffness moduli and phase angles for the studied mixtures (at 50 lm/m).

log E ¼ a  log f þ b  V m þ c  log e þ d

ð1Þ

8000

Sffness modulus (MPa)

higher than the other blends. TWW(RCA)3550 is the one that revealed a deformability behaviour closest to TWc3550, with a stiffness at 10 Hz around 4700 MPa and a phase angle of 19.4°. In general, HMA have higher stiffness modulus and lower phase angles than similar WMA. However, in this study TH3550 had greater porosity than usually (see Fig. 2). Because of this, the stiffness values for this blend were moderated in comparison with typical values for conventional HMA and lower than those of the reference WMA blend (TWc3550). When EAFS was introduced into the blend (TWc(EAFS)3550), stiffness modulus has reduced close to 29% and the phase angle slightly increased (from 23.2 to 24.5) in comparison to TWc3550. In what concerns stiffness the results were good to TWw(RCA) 3550 (with chemical additive and RCA) in comparison with a typical hot-mix asphalt, even if the values were the lowest of the study (reducing circa 31% as compared to the reference blend TWc3550), with a modulus around 4000 MPa at 10 Hz. The results for TWw(RCA)3550 showed that when RCA was used as part of the aggregate blend the stiffness moduli are close to those measured to TWc3550 (the WMA without by-products). However, the values at 10 Hz or higher were slightly lower for TWw(RCA)3550 (e.g. stiffness of 4722 and 5273 MPa for TWw (RCA)3550 and TWc3550, respectively), although phase angles have been around 16% lower for TWw(RCA)3550 (the coefficient of variation of phase angle is 6.4%). This indicates that including RCA into the WMA blends the viscous component seems to decrease as compared to the blends without by-products and with EAFS. A multivariate regression analysis was also carried out with SPSS (Statistical Package for the Social Sciences) to complement the evaluation in terms of stiffness modulus. The used statistical tools allowed the validation of regression analysis assumptions. The best-fit models (higher adjusted R2), after several types of models tried [26], were based on the independent variables determined in the laboratory. Stiffness modulus (E) was considered as the dependent variable whereas loading frequency (f in Hz), porosity (Vm in %) and strain level (e in lm/m) were used as independent variables as they were statistically significant and led to the best fit models [expression (1)]. The models presented in Fig. 5 only consider the range of porosity observed throughout the experimental framework. The derived regression model is useful as a complement of Fig. 4 because it allows us to observe the response of the studied warm-mix asphalt in terms of stiffness if porosity and loading conditions (strain and frequency) vary in the range evaluated in this study.

30 Hz

7000 6000

10 Hz

TWc(EAFS)3550

5000

30 Hz

4000

1 Hz 10 Hz

3000

TWw(RCA)3550

2000

1 Hz 1000 0 4.0 4.5 5.0 5.5 6.0 6.5 7.0 7.5 8.0

Porosity (%) Fig. 5. Results of stiffness modulus for TWc(EAFS)3550 and TWw(RCA)3550 derived from Eq. (1).

The coefficients a, b, c and d as well as the adjusted R2 are presented in Table 3. Fig. 5. shows the variation of stiffness modulus calculated from Eq. (1) as a function of porosity in the range of values observed on the specimens submitted to testing. Although porosity being higher in TWw(RCA)3550 than in TWc(EAFS)3550, the results show that using organic wax as additive and RCA as substitute of part of the aggregate increased stiffness moduli of the WMA.

3.5. Resistance to fatigue Fig. 6 presents the fatigue laws derived from the tests carried out. Although the results in Fig. 6 reveal that fatigue behaviour of the blends is somehow similar, they also show that slope of the lines is slightly higher for the WMA blends (TWc3550 to TWw (RCA)3550) than for the HMA (TH3550). Also, TWc3550, the WMA without by-products, has a line less sloping than the blends with by-products. Amongst these, TWw(RCA)3550, with organic wax and RCA, has the less sloping line. This finding can be analysed in more detail by calculating e4, e5 and e6, which represent the strain values to induce specimens’ failure after 10.000, 100.000 and 1 million loading cycles, respectively (Fig. 7). Fig. 7 confirms different rankings for the tested blends depending on the number of loading cycles applied. TH3550, for example, is the worst performant blend at 10.000 cycles and is the best one

691

F.C.G. Martinho et al. / Construction and Building Materials 185 (2018) 684–696 Table 3 Models of stiffness modulus derived from multivariate regression analysis. Blends

By-product

Additives

TWc(EAFS)3550 TWw(RCA)3550

EAFS RCA

Chemical Organic Wax

Adjusted R2 (%)

Coefficients a

b

c

d

0.306 0.219

0.048 0.065

0.058 0.051

3.581 4.042

94.8 79.6

μ

μ μ

ω

μ

Fig. 6. Fatigue laws at 20 °C derived from four-point bending tests by regression analysis.

Failure strain (

/m)

ε4

3.6. Water sensitivity TWc3550 TWc(EAFS)3550 TWc(RCA)3550

366

387 TWw(RCA)3550

350 353 342

TH3550

ε5 220 224 208 211 217

ε6 133 129 124 126 138

10 000

100 000

1 000 000

Number of loading cycles

Since the heating temperature of constituents in WMA blends is relatively low, some humidity can remain in the blend. This is referred to in the literature [3] as potentially harmful to water resistance of WMA. Fig. 8 shows the results for the studied blends. All the results revealed good performance in what concerns water sensitivity. In general, the specifications require ITSR values higher that 80%. TH3550 (HMA) reached a value of 93.1% and TWc3550, the reference WMA, achieved 94.2% for ITSR. When EAFS or RCA have been incorporated into the blend as substitutes of the natural aggregate, both the ITS and ITSR decrease as compared to the value measured to TWc3550 but the observed performance was still good. However, the incorporation of RCA resulted in lower resistance to water damage in TWc(RCA)3550 and TWw(RCA)3550

Fig. 7. Fatigue performance in terms of strain needed to specimen’s failure. 4 000

100 ITS d (kPa)

ITSR (%) 95

94.2

3 500

93.1

91.6

90 84.7

85 82.8

2 500

80

2 297 2 139

ITSR (%)

3 000

ITS (kPa)

at 1 million cycles. Taking into account that evaluation of fatigue resistance aims to anticipate which mixtures will behave better and worse after being submitted to a high number of loading cycles, e6 is typically the preferred parameter to evaluate fatigue resistance of asphalt concrete. The e6 values show that TWc3550, the blend without byproducts, is the best performant WMA blend (only 3% less than HMA, TH3550). The introduction of EAFS or RCA slightly decreased resistance to fatigue as compared to TWc3550: 3% in TWc(EAFS)3550, 7% in TWw(RCA)3550 and 5% in TWw(RCA) 3550. Moreover, the results also reveal that the WMA blends, including the ones with EAFS or RCA, resist better to higher strain levels that the HMA (TH3550) whereas for low strain levels the opposite occurs.

ITS w (kPa)

75 2 000

1 870

1 831 1 725

70

1 549

1 463

1 500

1 340

1 290

65 1 093

1 000

60 TWc35 50

TWc(EAFS)3550 TWc(RCA)3 550 TWw(RCA)3550

TH3550

Fig. 8. Results of ITSR, ITSw and ITSd for the blends assessed in the study.

692

F.C.G. Martinho et al. / Construction and Building Materials 185 (2018) 684–696

than the use of EAFS. Therefore, the results showed that for the additives and production temperatures used for the studied blends, no problems were observed regarding susceptibility to moisture of the tested WMA blends, when EAFS or RCA was added as part of the aggregate. 3.7. Comparative performance of the studied mixtures Fig. 9 compares the relative performance of the tested blends in what concerns resistance to permanent deformation (RD and WTS), stiffness, fatigue resistance and water sensitivity. The best performant blend represented in the graph for a parameter achieves 100% in that parameter. The remaining mixtures have a lower percentage, which represents the relative performance of each blend as compared to the best performant one. As Fig. 9 shows, TH3550 and TWw(RCA)3550 (WMA with wax and RCA) are the blends with the closest performance in most of the assessed parameters. Moreover, it also reveals that for the WMA blends with a chemical additive (TWc3550, TWc(EAFS) 3550 & TWw(RCA)3550) revealed similar behaviour, with a remarkable low resistance to permanent deformation. Consequently, the observed performance was apparently commanded by the type of additive used instead of the effect of introducing EAFS or RCA as substitutes of part of the aggregates. In what concerns resistance to fatigue, stiffness and water sensitivity of the blends, the use of EAFS or RCA did not significantly influence the observed performance. Fig. 9 also shows that although TWw (RCA)3550 has incorporated RCA, its general behaviour was not comparable to the remaining WMA blends. Generally speaking, TWw(RCA)3550 revealed good properties, comparable to TH3550 in all the parameters considered. As far as permanent deformation is concerned, TWw(RCA)3550 revealed even better resistance than

RD

3.8. Relative performance as compared with other WMA blends

100% 80%

Results overview

ITSR TH3550 TWc3550 TWc(EAFS)3550 TWc(RCA)3550 TWw(RCA)3550

fague

60% 40% 20% 0%

the observed for TH3550 (HMA). This behaviour seems to derive more from the organic wax used as additive than from the incorporation of RCA. Because engineers should have a procedure to decide on the blends to use or to reject, complementarily to the visual analysis based on Fig. 9, a process to rank the blends was also considered. Briefly, it consists in using the results of mechanical performance measured in the laboratory, to derive a single score. Six parameters were considered to rank the blends: RDAIR, WTSAIR, E, phase angle (/),e6 and ITSR. Taking into account that AC 20 is typically applied in base or binder pavement layers, some of those parameters are more important than the others. Therefore, the parameters were weighted to consider different contributions for an overall performance indicator. Although this procedure is somehow subjective, it expresses the authors’ perspective regarding the suitable properties for a blend to be applied in base and binder layers: RDAIR 15%; WTSAIR 15%; E 20%; / 5%; e 6 40%; ITSR 5%. The first step involves a transformation of each parameter to allow expressing it in a scale from 0 to 100%, dividing each individual value by the maximum (or minimum) of them. After this, these partial scores are summed to determine a global score. Values closer to 100% represent a better global performance than lower scores. Fig. 10 shows the scores calculated to the blends considering the aforesaid procedure. Taking all the evaluated parameters and the corresponding weights into consideration, TWw(RCA)3550 was the best blend, even better than TH3550 (HMA). These results also show that the combination in the same blend of a chemical additive and RCA, has slightly reduced the general performance of TWc(EAFS)3550 and TWc(RCA)3550 in comparison to TWc3550. This small score decrease is mainly associated with resistance to fatigue and stiffness.

WTS

sffness

Fig. 9. Relative performance of the tested blends in terms of five evaluated parameters.

An expected delivery of this study was to assess the influence of substituting in WMA blends part of the natural aggregate by RCA or EAFS. Apart from the results provided by this study both for conventional WMA and WMA with by-products as aggregates, some additional results are available in the literature, particularly for WMA manufactured with natural aggregates. Notice that the compositions of the blends studied elsewhere can differ in some degree from the mixtures under study. Therefore, these sources of information should be considered only to compare the general tendencies observed for conventional WMA in terms of the resulting properties, when using RCA and EAFS. Fig. 11 presents a summary of the comparisons carried out.

Fig. 10. Global score of the blends assessed in the study.

693

F.C.G. Martinho et al. / Construction and Building Materials 185 (2018) 684–696

a) Resistance to permanent deformaon WTSAIR (@50oC) 8: 14Ww5070 H

M25 T20Ww(RCA)3550

0.8 0.9

9: 20WwI

6: 14Wc5070F

(mm/103 cycles)

0.2 0.6 1.0 1.4 1.8 2.2 2.6 3.0 3.4 3.8

T20Wc(RCA)3550 M23

0.5 0.6

G 7: 14Ww5070

WTSAIR (@60oC)

T20Wc(EAFS)3550 M22

M22 T20Wc(EAFS)3550 0.0 0.2 0.3

M23 T20Wc(RCA)3550

A 14Wc5070 1:

5: 14Ww5070E

B 14Ww5070 2:

4: 20Ww1020D

C 14Ww160220 3:

RDAIR (@50oC)

M25 T20Ww(RCA)3550

4: 20Ww1020 D

T20Wc(RCA)3550 M23 11: Ww5070K

3: 14Ww160220C

RDAIR (@60oC)

T20Wc(EAFS)3550 M22

M22 T20Wc(EAFS)3550 2 4 6 8 10 12 14 16 18 20

(mm)

0 2 4 6 8 10 12 14 16 18 20

M23 T20Wc(RCA)3550

M25 T20Ww(RCA)3550

10: 20WwJ 2: 14Ww5070B

M25 T20Ww(RCA)3550

A 14Wc5070 1: I 9: 20Ww

b) Resistance to fague and sffness ε6 (20oC)

T20Wc(EAFS)3550 M22 17: 16Ww5070 Q 8000

T20Wc(EAFS)3550 M22

(μm/m)

180 160 140 120 100 80 60 40 20 0

18: 16Ww5070 R

D 4: 20Ww1020

T20Wc(RCA)3550 M23

16: 16Ws5070 P 15: 16Ws5070 O

T20Ww(RCA)3550 M25

7000 6000 5000 4000 3000 2000 1000

T20Ww(RCA)3550 M25 1: 14Wc5070 A

N 14: 16Ws5070

3: 14Ww160220C

E5: 14Ww5070

M 13: 16Ww5070

A 14Wc5070 1:

F6: 14Wc5070

12: 16Ws5070L B 2: 14Ww5070

Sffness (MPa) M23 T20Wc(RCA)3550

G 7: 14Ww5070 H 8: 14Ww5070

c) Indirect tensile strength and water sensivity ITSdry (kPa) 13: 16Ww5070 M

12: 16Ws5070L

T20Wc(EAFS)3550 M22 2500 2250 2000 1750 1500 1250 1000 750 500 250 0

T20Wc(RCA)3550 M23

ITSR (%) T20Wc(EAFS)3550 M22 18: 16Ww5070R 100 T20Wc(RCA)3550 M23 90 16: 16Ws5070P T20Ww(RCA)3550 80 M25 15: 16Ws5070O

T20Ww(RCA)3550 M25

14: 16Ws5070N M 13: 16Ww5070

D 4: 20Ww1020

1: A 14Wc5070 2: B 14Ww5070 3: C 14Ww160220

A 14Wc5070 1:

12: 16Ws5070L 3: 14Ww160220C

70 60 50 40 30 20 10 0

B 14Ww5070 2:

8: 14Ww5070H 7: 14Ww5070

D 20Ww1020 4: E 14Ww5070 5: F6: 14Wc5070 G

Fig. 11. Comparison of performance parameters observed in the study with results collected from the literature. 1: AC 14 surf 50/70: chemical [32], 2: AC 14 surf 50/70: organic wax [33], 3: AC 14 surf 160/220: organic wax [33], 4: AC 20 base 10/20: organic wax [33], 5: AC14 surf 50/70 mix@145 °C: organic wax [34], 6: AC14 surf 50/70 mix@130 °C: chemical [34], 7: AC14 surf 50/70 mix@145 °C: organic wax [33], 8: AC14 surf 50/70 mix@125 °C: organic wax [33], 9: AC20, [email protected] °C: organic wax [34], 10: AC20, mix@125 °C: organic wax [35], 11: 50/70: organic wax [36], 12: AC16 surf 50/70: surfactant [37], 13: AC16 surf 50/70: organic wax [37], 14: AC16 surf 50/70: surfactant A [38], 15: AC16 surf 50/70: surfactant B [38], 16: AC16 surf 50/70: surfactant C [38], 17: AC16 surf 50/70: organic wax [38], 18: AC16 surf 50/70: organic wax [39].

694

F.C.G. Martinho et al. / Construction and Building Materials 185 (2018) 684–696

In terms of rutting resistance (Fig. 11a), TWw(RCA)3550 performed very well either at 50 or 60 °C in comparison with a substantial number of other WMA without by-products. This allow us to anticipate good performance for a combination of wax & RCA. The performance obtained for WTSAIR at 50 °C, for instance, is quite close to the obtained for a WMA produced with a hard bitumen 10/20 (blend 4), which is known to perform very well against permanent deformation. In what concerns TWc(EAFS)3550 (chemical & EAFS) and TWc (RCA)3550 (chemical & RCA) the permanent deformation performance was quite irregular. At 50 °C TWc(RCA)3550 almost achieved a resistance similar to the best performant blends produced with 50/70 paving grade bitumen but at 60 °C its performance was weak. Although for TWc(EAFS)3550 the results were inconsistent when analysing the two measured rutting performance parameters, the comparison with the other blends considered reveals that combining a chemical additive and EAFS as substitute of aggregates has a tendency to originate weak resistance to rutting. Looking at resistance to fatigue (Fig. 11b) all the studied blends performed quite well in comparison with the results taken from the literature, even better than the observed for the blends 1, 2 and 18, all of them produced with a bitumen softer than the studied blends and, therefore, with increased expectations in terms of fatigue resistance. For the combination of additives (wax or chemical) and by-products (RCA or EAFS) used as substitutes of part of the aggregate, resistance to fatigue seems not to be a concern. In terms of stiffness moduli at 20 °C and 10 Hz, TWc(EAFS)3550, TWc(RCA)3550 and TWw(RCA)3550 revealed lower values than practically all the blends used for comparison. Nevertheless, the values of stiffness modulus measured for the studied blends are approximately in the range 4000–5000 MPa. According with the authors experience as consultants, this is satisfactory for typical pavement asphalt layers. Therefore, stiffness level of these WMA produced with RCA and EAFS is not a problem. Although TWc(EAFS)3550 and TWc(RCA)3550 have achieved values of ITSdry generally lower than those observed for the other WMA without by-products considered for assessment, their sensitivity to water was good, with ITSR values above 80%. Also, TWw (RCA)3550 reached good performance in terms of ITSdry and ITSR as compared to all the WMA mixtures used for comparison. Based on this analysis, it can be concluded that the incorporation of EAFS or RCA in WMA is not a problem vis-a-vis resistance to water of WMA. As a corollary of the comparisons performed above, a procedure close to that proposed in [40] was followed to summarise the trends referred to in the literature (already mentioned at the

beginning of this paper) as well as the general tendencies observed in this study (Table 4). They are shown together to highlight the effects on the properties of WMA blends derived from substituting part of the aggregates by RCA or EAFS.

4. Conclusions The use of by-products in WMA blends was the main focus of this study aiming at verifying, on the one hand, the feasibility of these techniques in real production conditions and, in the other hand, to evaluate the influence on the WMA properties resulting from the incorporation of RCA or EAFS as substitutes of part of the aggregates. With the exception of handling temperatures of the mixtures, lower than the commonly applied for HMA, the construction procedures used in the field to lay and compact the WMA blends with 60% of RCA or 30% of EAFS did not require any modification of the traditional construction techniques. These percentages of aggregate substitution were derived from preliminary testing carried out in laboratory. The specimens collected from the field allowed the development of a mechanical characterisation program in the laboratory. The trends mentioned in the literature for HMA incorporating the same types of by-products considered in this study were not completely confirmed (Table 4). Although the blends produced with RCA, for instance, had the same binder content that the other blends, the resulting porosity was not consistently higher. Only TWw(RCA)3550 (wax and RCA) revealed porosity considerably higher but the same did not happen for the similar laboratory blend (LWw(RCA)3550). This shows that the porous cement mortar of the RCA did not consistently absorb the blends’ binder. The results produced this study as well as the information gathered from the literature led to the following conclusions:
The general performance of WMA with 60% of RCA was very good (even better than HMA used as reference) when the additive was an organic wax, but rutting resistance was weak when a chemical additive was used instead of wax;
Because all the WMA blends with chemical additive tested had rather weak permanent deformation resistance, apparently the effect of adding a chemical additive overlapped the effect of incorporating RCA into the blend;
The comparison of results of this study with the obtained for other blends evaluated elsewhere reveals that combining chemical additives and EAFS as substitute of aggregates has a tendency to originate mixtures with weak resistance to rutting;

Table 4 Tendency of change in properties of bituminous mixtures (adapted from [40]). Parameter

Marshall stability

HMA with RCA1 HMA with EAFS1 WMA technology (at HMA temperature)1 WMA technology (with temperature reduction)1 WMA with RCA (in this study: with temperature reduction)2 WMA with EAFS (in this study: with temperature reduction)2 usually increases. may increase. usually unchanged. may decrease. usually decreases. 1 Comparing with conventional HMA. 2 Comparing with a similar WMA blend without by-products as aggregates.

Porosity

Rutting resistance

Stiffness

Fatigue resistance

Water sensitivity

F.C.G. Martinho et al. / Construction and Building Materials 185 (2018) 684–696


Although the use of by-products as substitutes of part of the aggregate decreased the stiffness moduli achieved in comparison with the blend without by-products, typical values, at 20 °C, were about 4000–5000 MPa, which is satisfactory for a conventional base/binder layer;
Fatigue life of WMA with RCA or EAFS was very satisfactory and close to the performance observed for the blends (HMA and WMA) used as references; also, comparing the results obtained with other determined elsewhere on conventional WMA led to the conclusion that resistance to fatigue seems not to be a concern for WMA with RCA or EAFS;
Sensitivity to water slightly decreased for the WMA blends with by-products as compared with the conventional WMA and HMA and the performance achieved by ITSR was over 80%, which is the typical lower level requirement for asphalt concrete; so, water sensitivity was not a problem when EAFS or RCA are added to the blends to produce WMA;
Although conventional equipment was used to build the experimental sections of pavement, no specific problems were observed in the field for the WMA blends with RCA or EAFS in any of the necessary operations: mixing, transport, laying and compaction. Concluding, it can be stated that based on the authors’ experience throughout this study WMA produced with a high percentage of RCA or EAFS is practicable but some combinations of them with additives should be evaluated with caution. In fact, in situations where good performance to permanent deformation is required the combination in a WMA of RCA and a chemical additive should be avoided. Since the convenient evaluation is carried out, solutions based on WMA with RCA or EAFS are suitable alternatives to conventional HMA. It must also be emphasized that the use of RCA or EAFS as part of the aggregates fulfil the European goals of increasing circular economy, reducing the waste going to landfill and increasing valorisation of by-products in different value chains. Conflicts of interest The authors declare no conflicts of interest. Acknowledgements The authors want to thanks to the technical personnel working at the Laboratory of Highways and Transportation (LVTC) of the CERIS, Instituto Superior Tecnico, Universidade de Lisboa. This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. References [1] F. Olard, C. Noan, Low Energy Asphalts. Routes Roads 336/337, PIARC (World Road Association), 2008, pp. 131–145. [2] EAPA, The Use of Warm Mix Asphalt – EAPA Position Paper, European Asphalt Pavement Association, Brussels, 2010. [3] S. Capitão, L. Picado-Santos, F. Martinho, Pavement engineering materials: review on the use of warm-mix asphalt, Constr. Build. Mater. 36 (2012) 1016– 1024, https://doi.org/10.1016/j.conbuildmat.2012.06.038. [4] F. Martinho, L. Picado-Santos, S. Capitão, Mechanical properties of warm-mix asphalt concrete containing different additives and recycled asphalt as constituents applied in real production conditions, Constr. Build. Mater. 131 (2017) 78–89, https://doi.org/10.1016/j.conbuildmat.2016.11.051. [5] J. Zhu, S. Wu, J. Zhong, D. Wang, Investigation of asphalt mixture containing demolition waste obtained from earthquake-damaged buildings, Constr. Build. Mater. 29 (2012) 466–475, https://doi.org/10.1016/j.conbuildmat.2011.09. 023. [6] S. Paranavithana, A. Mohajerani, Effects of recycled concrete aggregates on properties of asphalt concrete, Resour. Conserv. Recycl. 48 (2006) 1–12, https://doi.org/10.1016/j.resconrec.2005.12.009.

695

[7] A. Pasandín, I. Pérez, Overview of bituminous mixtures made with recycled concrete aggregates, Constr. Build. Mater. 74 (2015) 151–161, https://doi.org/ 10.1016/j.conbuildmat.2014.10.035. [8] J. Mills-Beale, Z. You, The mechanical properties of asphalt mixtures with recycled concrete aggregates, Constr. Build. Mater. 24 (2010) 230–235, https:// doi.org/10.1016/j.conbuildmat.2009.08.046. [9] A. Zulkati, Y. Wong, D. Sun, Mechanistic performance of asphalt-concrete mixture incorporating coarse recycled concrete aggregate, J. Mater. Civ. Eng. 25 (9) (2013) 1299–1306. [10] A. Pasandín, I. Pérez, Effect of ageing time on properties of hot-mix asphalt containing recycled concrete aggregates, Constr. Build. Mater. 52 (2014) 284– 293, https://doi.org/10.1016/j.conbuildmat.2013.11.050. [11] W. Gul, M. Guler, Rutting susceptibility of asphalt concrete with recycled concrete aggregate using revised Marshall procedure, Constr. Build. Mater. 55 (2014) 341–349, https://doi.org/10.1016/j.conbuildmat.2014.01.043. [12] S. Wu, B. Muhunthan, H. Wen, Investigation of effectiveness of prediction of fatigue life for hot mix asphalt blended with recycled concrete aggregate using monotonic fracture testing, Constr. Build. Mater. 131 (2017) 50–56, https:// doi.org/10.1016/j.conbuildmat.2016.11.045. [13] A. Kavussi, M. Qazizadeh, Fatigue characterization of asphalt mixes containing electric arc furnace (EAF) steel slag subjected to long term aging, Constr. Build. Mater. 72 (2014) 158–166, https://doi.org/10.1016/j.conbuildmat.2014.08. 052. [14] J. Xie, W. Shaopeng, J. Lin, J. Cai, Z. Chen, W. Wei, Recycling of basic oxygen furnace slag in asphalt mixture: material characterization and moisture damage investigation, Constr. Build. Mater. 36 (2012) 467–474. [15] M. Skaf, J. Manso, Á. Aragón, J. Fuente-Alonso, EAF slag in asphalt mixes: a brief review of its possible re-use, Resour. Conserv. Recycl. 120 (2017) 176–185, https://doi.org/10.1016/j.resconrec.2016.12.009. [16] J. Xie, W. Shaopeng, L. Zhang, Y. Xiao, W. Ding, Evaluation the deleterious potential and heating characteristics of basic oxygen furnace slag based on laboratory and in-place investigation during large-scale reutilization, J. Cleaner Prod. 133 (2016) 78–87, https://doi.org/10.1016/j.jclepro.2016.05. 106. [17] M. Zhao, W. Shaopeng, C. Zongwu, L. Chao, Function investigation of stone mastic asphalt (SMA) mixture partly containing basic oxygen furnace (BOF) slag, J. Appl. Biomater. Funct. Mater. 14 (Suppl. 1) (2016) S68–S72, https://doi. org/10.5301/jabfm.5000307. [18] M. Pasetto, N. Baldo, Experimental evaluation of high performance base course and road base asphalt concrete with electric arc furnace steel slags, J. Hazard. Mater. 181 (2010) 938–948, https://doi.org/10.1016/j.jhazmat.2010.05.104. [19] M. Pasetto, Baldo, Laboratory investigation on foamed bitumen bound mixtures made with steel slag, foundry sand, bottom ash and reclaimed asphalt pavement, Road Mater Pavement Des. 13 (2012) 691–712, https://doi. org/10.1080/14680629.2012.742629. [20] M. Ameri, S. Hesami, H. Goli, Laboratory evaluation of warm mix asphalt mixtures containing electric arc furnace (EAF) steel slag, Constr. Build. Mater. 49 (2013) 611–617, https://doi.org/10.1016/j.conbuildmat.2013.08.034. [21] CEN. EN 12457-4, Characterization of Waste – Leaching – Compliance Test for Leaching of Granular Waste Materials and Sludges – Part 4 – One Stage Batch Test at a Liquid to Solid Ratio of 10 l/kg for Materials with Particle Size Below 10 mm (without or with size reduction), European Committee for Standardization, Brussels, 2002. [22] CEN. EN 1744-3, Tests for Chemical Properties of Aggregates – Part 3 – Preparation of Eluates by Leaching of Aggregates, European Committee for Standardization, Brussels, 2002. [23] The Council of The European Union, Council Decision of 19 December 2002 establishing criteria and procedures for the acceptance of waste at landfills pursuant to Article 16 of and Annex II to Directive 1999/31/EC, Official Journal of the European Communities (2003). 16.1.2003. [24] CEN. EN 12697-35, Bituminous Mixtures – Test Methods for Hot Mix Asphalt – Part 35 – Laboratory Mixing, European Committee for Standardization, Brussels, 2016. [25] CEN. EN 12697-30, Bituminous Mixtures – Test Methods for Hot Mix Asphalt – Part 30 – Specimen Preparation, Impact Compactor, European Committee for Standardization, Brussels, 2012. [26] F. Martinho, Misturas betuminosas temperadas com incorporação de subprodutos industriais: caracterização e validação tecnológica (Warm mix asphalt with incorporation of industrial by-products: characterisation and technological validation), Universidade de Lisboa (in Portuguese), Instituto Superior Tecnico, 2014. http://files.isec.pt/DOCUMENTOS/SERVICOS/BIBLIO/ Teses/Tese_Dout_Fernando-Martinho.pdf. [27] CEN. EN 12697-22, Bituminous Mixtures – Test Methods for Hot Mix Asphalt – Part 22 – Wheel Tracking, European Committee for Standardization, Brussels, 2003/2007. [28] CEN. EN 12697-26, Bituminous Mixtures – Test Methods for Hot Mix Asphalt – Part 26 – Stiffness, European Committee for Standardization, Brussels, 2012. [29] CEN. EN 12697-24, Bituminous Mixtures – Test Methods for Hot Mix Asphalt – Part 24 – Resistance to Fatigue, European Committee for Standardization, Brussels, 2012. [30] CEN. EN 12697-12, Bituminous Mixtures – Test Methods for Hot Mix Asphalt – Part 12 – Determination of the Water Sensitivity of Bituminous Specimens, European Committee for Standardization, Brussels, 2008. [31] CEN. EN 12697-23, Bituminous Mixtures – Test Methods for Hot Mix Asphalt – Part 23 – Determination of the Indirect Tensile Strength of Bituminous Specimens, European Committee for Standardization, Brussels, 2017.

696

F.C.G. Martinho et al. / Construction and Building Materials 185 (2018) 684–696

[32] J. Oliveira, H. Silva, L. Abreu, J. Gonzalez-Leon, The role of a surfactant based additive on the production of recycled warm mix asphalts – less is more, Constr. Build. Mater. 35 (2012) 693–700, https://doi.org/10.1016/ j.conbuildmat.2012.04.141. [33] H. Silva, J. Oliveira, J. Peralta, J. Zoorob, Optimization of warm mix asphalts using different blends of binders and synthetic paraffin wax contents, Constr. Build. Mater. 24 (2010) 1621–1631, https://doi.org/10.1016/ j.conbuildmat.2010.02.030. [34] H. Silva, J. Oliveira, C. Ferreira, P. Pereira, Assessment of the performance of warm mix asphalts in road pavements, Int. J. Pavement Res. Technol. 3 (3) (2010) 119–127. [35] Bolzan P, Balige M, Mezclas asfalticas en caliente elaboradas y compactadas a menores temperaturas, in: XXXVI Reunión del Asfalto Bicentanário de los Andes. Buenos Aires, Argentina, 2010. [36] E. Toraldo, C. Brovelli, E. Mariani, Laboratory investigation into the effects of working temperatures on wax-based warm mix asphalt, Constr. Build.

[37]

[38]

[39]

[40]

Mater. 44 (2013) 774–781, https://doi.org/10.1016/j.conbuildmat.2013.03. 085. E. Sanchez-Alonso, A. Vega-Zamanillo, D. Castro-Fresno, M. DelRio-Prat, Evaluation of compactability and mechanical properties of bituminous mixes with warm additives, Constr. Build. Mater. 25 (5) (2013) 2304–2311, https:// doi.org/10.1016/j.conbuildmat.2010.11.024. S. Redondo, J. León, A. González, Sánchez E, Estudio del ángulo de contacto de betunes aditivados y su efecto sobre distintas propiedades mecánicas de las mezclas asfálticas, Revista Carreteras (2011) 178. Oliveira J, Silva H, Fonseca P, Kim Y, Hwang S, Pyun J, Lee H, Laboratory and field study of a WMA mixture produced with a new temperature reduction additive, in: Proceedings of the 2nd International Warm-mix Conference, St. Louis, USA, 2012. M. Zaumanis, L. Poulikakos, M. Partl, Performance-based design of asphalt mixtures and review of key parameters, Mater. Des. 141 (2018) 185–201.