The effect of pumice powder on the self-compactability of pumice aggregate lightweight concrete

Construction and Building Materials 103 (2016) 36–46

Contents lists available at ScienceDirect

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

The effect of pumice powder on the self-compactability of pumice aggregate lightweight concrete Murat Kurt a, Muhammed Said Gül b, Rüstem Gül b, Abdulkadir Cüneyt Aydin b,⇑, Türkay Kotan c a

Atatürk University, Faculty of Architecture & Design, Department of Architecture, 25240 Erzurum, Turkey Atatürk University, Faculty of Engineering, Department of Civil Engineering, 25240 Erzurum, Turkey c Erzurum Technical University, Faculty of Engineering & Architecture, Department of Civil Engineering, 25070 Erzurum, Turkey b

h i g h l i g h t s  The self-compacting pumice aggregate concrete satisfies the strength requirement of semi-structural lightweight concrete.  The flowing ability and thermal conductivity of self-compacting pumice aggregate concrete is expressive.  Increasing pumice ratio increased the tendency to segregation of fresh concrete.  The grading and moisture content of pumice aggregate are the main parameters of self-compacting concrete mix.

a r t i c l e

i n f o

Article history: Received 29 July 2015 Received in revised form 4 November 2015 Accepted 20 November 2015

Keywords: Lightweight aggregate concrete Self-compacting concrete Pumice Pumice powder

a b s t r a c t This paper presents the results of an experimental study about the effects of pumice powder, different water/(cement + mineral additive) ratios and pumice aggregates on some physical and mechanical properties of self-compacting lightweight aggregate concrete. In this study, pumice was used as lightweight aggregates. Several properties of self-compacting pumice aggregate lightweight concretes such as unit weight, flow diameter, T50 time, flow diameter after an hour, V-funnel time and L-box tests, 7, 28, 90 and 180-day compressive strength, 28-day splitting tensile strength, dry unit weight, water absorption, thermal conductivity and ultrasonic pulse velocity tests were investigated. For this purpose, 24 series of concrete samples were prepared in two groups. In the first group, pumice aggregate at the rate of 100% was used for the production of self-compacting lightweight aggregate concrete with constant w/(c + m) ratios as 0.35, 0.40 and 0.45 by weight. Furthermore as the second group, pumice aggregate was used as a replacement of natural aggregate, at the levels of 0%, 20%, 40%, 60%, 80%, and 100% by volume with fly ash and blast furnace slag mineral additives at the constant rate of 40%. The flow diameters, T50 times, paste volumes, 28-day compressive strengths, dry unit weights, thermal conductivities and ultrasonic pulse velocity of self-compacting lightweight aggregate concrete were obtained in the range of 560–800 mm, 2–11 s, 435–558 l/m3, 10.5–65.0 MPa, 840–2278 kg/m3, 0.347–1.694 W/mK and 2611–4770 m/s respectively, which satisfies not only the strength requirement of semi-structural lightweight concrete but also the flowing ability requirements and thermal conductivity requirements of self-compacting lightweight aggregate concrete. Ó 2015 Published by Elsevier Ltd.

1. Introduction Concrete is a multiphase exceedingly complex heterogeneous material and one of the principal materials for structures. However, the heterogeneous structure of concrete results in some undesirable effects. Heterogeneity and other properties of concrete are mostly concerned with the hydration. Hydration, the chemical

⇑ Corresponding author. E-mail address: [email protected] (A.C. Aydin). http://dx.doi.org/10.1016/j.conbuildmat.2015.11.043 0950-0618/Ó 2015 Published by Elsevier Ltd.

reaction between water and ingredients of cement, is one of the most important properties of its strength gain process. This property of hydration caused volume change of hydrated cement, varying hydration rate through the concrete and time dependency of strength gain. One of the main effects of strength gain is the improved mechanical properties of concrete. The mechanical properties of cement based material is needed by designers for stiffness and deflections evaluation, and is a fundamental property required for the proper modelling of its constitutive behaviour and for its proper use in various structural applications. For this reason,

37

M. Kurt et al. / Construction and Building Materials 103 (2016) 36–46

Nomenclature SCC SCLC LWAC UPV SGF SG w c m CS FA BFS NP80

self-compacting concrete self-compacting lightweight aggregate concrete lightweight aggregate concrete ultrasonic pulse velocity specific gravity factor specific gravity water cement mineral admixture control sample fly ash blast furnace slag %80 pumice as aggregate (the number varies according to percentage of pumice aggregate)

determination of mechanical properties of concrete has become very important in terms of design. However, due to economic considerations, there is strong demand on natural resource usage. Moreover, when structure weights are considered, not only natural light weight aggregates but also artificial light materials such as gas concrete are used. Incorporation of natural/artificial resources in concrete brings environmental, economic and/or technological benefits [1–13]. Self-compacting concrete (SCC) is considered as a concrete which can be placed and compacted under its self-weight with little or no vibration effort and which is at the same time cohesive enough to be handled without segregation or bleeding [14]. SCC was originally developed at the University of Tokyo, Japan in 1986 by Prof. Okamura and his team to improve the quality of construction and to overcome the problems of defective workmanship [15]. It is used to facilitate and ensure the proper filling and good structural performance of the restricted areas and heavily reinforced structural members. SCC can also provide a better working environment by eliminating the vibration noise [16]. Self-compacting lightweight aggregate concrete (SCLC) is a kind of high-performance concrete developed from self-compacting concrete (SCC). SCLC combines the favourable properties of lightweight aggregate concrete (LWAC) and SCC, needs no external vibration, and can spread into place, fill the formwork and encapsulate reinforcement without any bleeding or segregation [17,18]. On the other hand, the use of chemical admixtures is always necessary when producing SCC in order to increase the workability and reduce the segregation. The content of coarse aggregate and water to binder ratio in SCC are lower than those of normal concrete. Therefore, SCC contains large amounts of fine particles such as, blast-furnace slag, fly ash and lime powder in order to avoid gravity segregation of larger particles in the fresh mix. The wide variety of the lightweight aggregate source result in distinguishing behaviour among the SCLCs. Thus, properties of SCLCs have to be examined individually [19–21]. Pumice is a natural material of volcanic origin produced by the release of gases during the solidification of lava, and it has been used as the aggregate in the production of lightweight concrete in many countries around the world. So far, the use of pumice was dependent on the availability and limited to the countries where it is locally available or easily imported. Approximately, 7.4 billion m3 (40%) of the total 18 billion m3 of pumice reserve is located in Turkey [22]. Therefore, the use of pumice as aggregate or mineral additive in the production of self-compacting concrete may be a good approach for the production of lightweight, easy workable, economic and environmentalist concrete.

PP1

%100 pumice aggregate and pumice powder as mineral admixture, w/(c + m) = 0.35 PP2 %100 pumice aggregate and pumice powder as mineral admixture, w/(c + m) = 0.40 PP3 %100 pumice aggregate and pumice powder as mineral admixture, w/(c + m) = 0.45 EFNARC The European Federation of Specialist Construction Chemicals and Concrete Systems T50 time to flow a diameter of 50 cm T400 the time for SCLC to reach 400 mm from three steel bars V-funnel flow time Tv Dmax maximum grain size STS splitting tensile strength

There has been an increase in the usage self-compacting concrete (SCC) in recent years and a number of papers have been published on this topic [19]. However, there is very little documentation on selfcompacting lightweight aggregate concrete, which has superior advantages as using natural materials, lightness and easy workability. Thus, a study was performed under the light of the literature information given above. For this purpose, experimental studies were carried out in two base groups. In the first group, concrete specimens with three different water/(cement + mineral additive) portions, were prepared by using volcanic originated pumice aggregate at 100%. In the second group, concrete specimens with constant water/(cement + mineral additive) portions and complemented by blast furnace slag instead of cement were produced by replacing pumice five different ratios instead of the normal aggregate. Then, some physical and mechanical properties such as workability, unit weight, compressive and splitting tensile strength, thermal conductivity and ultrasonic pulse velocity (UPV) of selfcompacting concrete were investigated. 2. Materials and experimental study 2.1. Materials In this study, CEM I 42.5 N type cement, fly ash and blast furnace slag were used. The fly ash that was supplied from Nallıhan Factory of Park Thermal Electric & Trading Co. is accepted as class F fly ash, because CaO content is less than 10%. Table 1 Chemical properties of Portland cement, fly ash, blast-furnace slag and pumice aggregate. Components

Portland cement (%)

Fly ash (%)

Blast furnace slag (%)

Pumice aggregate (%)

SiO2 Al2O3 Fe2O3 CaO MgO SO3 Na2O K2O CI Free CaO Na2O + K2O Insoluble matter Ignition loss Undetermined*

20.79 5.17 3.43 60.29 3.03 3.12 0.41 0.66 0.0251 0.34 – 2.47

47.50 15.95 16.60 6.60 4.65 3.94 1.74 1.96 – 0.56 3.70 –

38.54 14.90 1.50 33.50 8.20 0.62 0.22 1.50 – – – –

69.78 11.16 2.11 2.47 0.60 0.06 4.33 2.87 0.0496 – – –

2.79 0.32

– –

1.00 –

4.66 –

* Undetermined: acid insoluble residue, a non-cementing material which is present in Portland cement.

38

M. Kurt et al. / Construction and Building Materials 103 (2016) 36–46

Table 2 Physical and mechanical properties of cement and pozzolans. Properties

Portland cement

Blast furnace slag

Fly ash

Specific gravity (g/cm3) Specific surface (cm2/g) Retained on the sieve of 0.09 mm Initial set (hour–minute) Final set (hour–minute) Volume expansion (Le Chatelier, mm) Compressive strength 2th day (MPa) 7th day 28th day

3.13 3751 1 2 h–38 m 3 h–15 m 2

2.93 4300 – – – –

2.05 3500 – – – –

23.6 37.9 48

– – –

– – –

Blast-furnace slag was supplied from Karabük Factory of Karçimsa Cement Industry and Trade Inc. The pumice aggregate, supplied from volcanic slug furnace from Demirdöven region of Pasinler/Erzurum was used as lightweight aggregate. Chemical properties of cement, fly ash, blast-furnace slag and pumice aggregate are given in Table 1; physical and mechanical properties of them are given in Table 2. For the natural aggregates, natural sand from Aras River and crushed gravel aggregates supplied at Yag˘an region of Erzurum were used as normal aggregates. The grading curves of pumice and normal aggregates used in experiments were presented in Fig. 1. Specific gravity factor (SGF) incorporates compensation for absorption of free water by the pumice aggregates, but it is used in exactly the same way to calculate the volume relationship. Specific gravity factor of the lightweight aggregate for 10 min were used in the calculation of effective volume for 1 m3 concrete; the specific gravity and moisture percentages of normal aggregate were also used. In addition, specific gravity factor, specific gravity (SG) and water absorption values of normal and lightweight aggregate for different grain sizes were given in Table 3. Besides, the ratio of fine aggregate was determined as 50% and 60% in the experiments to ensure the condition of self-compacting of fresh concrete, respectively.

Table 3 The specific gravity factor, specific gravity and water absorption values of aggregates (According to TS EN 1097). Grain Size

0/2 2/4 4/8 8/16

Lightweight Aggregate

Normal aggregate

SGF

SG

Water absorption (%)

1.68 1.18 0.98 0.92

2.41 2.60 2.59 2.61

6.48 1.69 2.92 2.24

As a result of this adjustment, aggregate ratios’ grain class were determined as 30% and 42% for 0–2 mm, 20% and 18% for 2–4 mm, 35% and 20% for 4–8 mm, 15% and 20% for 8–16 mm in the first and second group of experiments, respectively. The fineness modules were 3.86 for Group I and 3.50 for Group II of mixes. Furthermore, a third generation modified polycarboxylate-based hyper-plasticizer was used in the concrete mix to provide viscosity and an air entraining admixture was used to reduce the risk of segregation and to increase the cohesion in some cases. 2.2. Experimental study This study was carried out in two groups. In the first group of experiments, pumice at 100% was used as the aggregate in concrete. SCC samples were produced at three different w/(c + m) ratios (35%, 40%, 45%) and three different pumice powder ratios (20%, 30%, 40%). In the second group of experiments, pumice at five different ratios (20%, 40%, 60%, 80%, 100%) was used instead of normal aggregate for every grain grade. The w/(c + m) ratio was constant as 0.30, fly ash and blast furnace slag at 40% of cement weight was replaced instead of cement. Hyper plasticizer and air entraining agents were used in SCC mixes to increase workability and to decrease segregation and bleeding. Thus, fresh and hardened concrete properties such as workability, unit weight and compressive strength, splitting tensile strength, UPV and thermal conductivity of produced normal, structural semi-

Fig. 1. Grading curves of pumice and normal aggregates used in the first and second group of experiments.

39

M. Kurt et al. / Construction and Building Materials 103 (2016) 36–46

The L-box test is used to evaluate the fluidity of SCLC and the ability for SCLC to pass through steel bars [17]. The L-box consists of a ‘‘chimney” section and a ‘‘channel” section as shown in Fig. 4. With the L-box, the height of concrete in the chimney, h1, the height of concrete in the channel section, h2, and the time for SCLC to reach 400 mm from three steel bars, T400, can be measured. According to EFNARC [23], when the ratio of h2 to h1 is larger than 0.8, SCC has good passing ability. The measured values of h2/h1 are shown in Tables 6 and 7.

3. Experimental results 3.1. Fresh concrete results

Fig. 2. Slump flow test.

lightweight, structural lightweight and lightweight SCC were investigated. The concrete mixture proportions are given in Tables 4 and 5 for Group I and II concrete samples, respectively.

2.3. Workability tests The self-compacting ability of SCC may be defined by three parameters: filling ability, resistance to segregation and passing ability [23]. Haist et al. [24] proposed three mix proportions for SCLC and assessed their self-compacting properties by the slump flow, J-ring, V-funnel, and L-box tests. It has been found that, compared to SCC, there is no significant difference in the mix proportion design except for the aggregate used [17]. To determine the fluidity and workability properties of SCC, V-funnel tests were performed to gather information about flowing ability and viscosity with flow diameter and time of fresh concrete (see Fig. 2). Besides, the L-box tests were performed to determine the passing ability from narrow sections of fresh concrete. These fresh concrete tests were conducted according to the standards of EFNARC [23], prepared by the European Working Group on Self-Compacting Concrete. Furthermore, to produce proper selfcompacting mixes, several preliminary trials on self-compacting pumice concrete were carried out. For each mixture, the flow diameter, time to flow a diameter of 50 cm (T50 time), flow diameter after one hour, V-funnel flow time, V-funnel times delayed 5 min, L-box ratio, air temperature and the unit weight of fresh concrete were measured. The details of the fresh concrete tests for SCC were given elsewhere [17,23]. The V-funnel flow test is performed to evaluate the fluidity of SCLC and the ability for SCLC to change its path and to pass through a constricted area. For this test, the V-funnel apparatus is shown in Fig. 3, the total time for SCLC to flow through the V-funnel was measured. According to EFNARC [11,21], class 1 SCC Tv is smaller than 8 s and class 2 SCC Tv is 9–25 s [17]. The measured values of Tv and T50 times are presented in Tables 6 and 7.

According to EFNARC [23] standard, the flow diameter, time to flow a diameter of 50 cm (T50 time), flow diameter after an hour, V-funnel flow time, V-funnel times delayed 5 min, L-box ratio, air temperature and the unit weight of fresh concrete were measured for each mixture. The results of fresh concrete experiments obtained for the samples in Groups I and II are given in Tables 6 and 7 as follows. The flow diameters of concrete containing pumice powder and not containing pumice powder were measured as 560–640 mm and 600–650 mm respectively. For all mixtures, as w/(c + m) rate increases, flow diameter also increases because of shear stress and viscosity of concrete decreased. Pumice powder replacing instead of cement has caused to the increase in the workability and flow diameter. The flow diameter of concrete samples in Group II varies between 645-800 mm as seen in Table 7. For this group, the relationships between flow diameter and unit weight of concrete are shown in Fig. 5. The mixtures have satisfied 650–800 mm value of flow diameter, proposed by EFNARC [23]. The results show that as density of SCLC increases, its workability increases, too. This is a prospective result since the spread and placement properties of SCC are provided by its weight. Due to increasing weight of the mix, the spreading capability will be enhanced at the fresh stage. The times to flow a diameter of 50 cm were measured as 5–11 s and 2–9 s in, the first and second group of experiments, respectively. According to the results of Dowson [25], the time to flow a diameter of 50 cm is not more than 3 s. EFNARC [23] also suggested that the time to flow a diameter of 50 cm is 2–5 s. The time to flow a diameter of 50 cm is related to the flow rate and plastic

Table 4 Concrete mixture proportions of the Group I samples and control samples (CS). Components

CS1

CS2

CS3

PP1

PP2

PP3

PP4

PP5

PP6

PP7

PP8

PP9

Cement (kg/m3) Water (kg/m3) Pumice Powder (kg/m3) Super plasticizer (kg/m3) Air entraining (kg/m3) Pumice (kg/m3) 0/2 2/4 4/8 8/16 Theor. unit weightb (kg/m3) Powder quantity (kg/m3) Paste volume (l/m3)

550 193 – 11 1 262 123 178 72 1390

550 220 – 11 1 248 116 169 68 1384

550 248 – 11 1 235 110 160 64 1378

440 193 98 11 1 269 120 174 70 1376

440 220 98 11 1 255 114 165 66 1370

440 248 98 11 1 240 107 156 63 1364

385 193 153 11 1 266 119 172 69 1369

385 220 153 11 1 252 112 163 66 1363

385 248 153 11 1 273 106 154 62 1356

330 193 208 11 1 263 117 170 69 1362

330 220 208 11 1 248 111 161 65 1355

330 248 208 11 1 234 104 152 61 1349

550 480

550 507

550 535

550 491

550 519

550 546

550 497

550 525

550 552

550 503

550 530

550 558

0.35 1.10a 0

0.40 1.25 0

0.45 1.41 0

0.35 1.03 20

0.40 1.17 20

0.45 1.32 20

0.35 1.00 30

0.40 1.14 30

0.45 1.28 30

0.35 0.97 40

0.40 1.11 40

0.45 1.24 40

1.54 0.50

1.54 0.50

1.54 0.50

1.59 0.50

1.59 0.50

1.59 0.50

1.59 0.50

1.59 0.50

1.59 0.50

1.59 0.50

1.59 0.50

1.59 0.50

Design parameters w/(c + m) (weight) w/(c + m) (volume) Mineral admixture (cement %) Sand/fine ag. (weight) Fine ag./aggregate (volume) a b

w/(c + m) (volume) = 193/(550/3.13) = 1.10. Theoretical unit weight is the air free unit weight.

40

M. Kurt et al. / Construction and Building Materials 103 (2016) 36–46

Table 5 Concrete mixture proportions of the Group II samples. Components

Cement (kg/m3) Water (kg/m3) Fly Ash (kg/m3) Blast-Furnace Slag (kg/m3) Super plasticizer (kg/m3) Pumice (kg/m3) 0/2* 2/4* 4/8* 8/16* Normal (kg/m3) 0/2* 2/4* 4/8* 8/16* Theoretical unit weight Powder quantity (kg/m3) Paste volume (l/m3) Design parameters w/(c + m) (weight) w/(c + m) (volume) Mineral admixture (cement %) Sand/fine ag. (weight) Fine ag./aggregate (volume) *

FA

BFS

NP100

NP80

NP60

NP40

NP20

NP0

NP100

NP80

NP60

NP40

NP20

NP0

375 188 231 – 13 392 112 104 97 0 0 0 0 1512 625 471

375 191 225 – 13 314 90 83 78 119 49 56 56 1650 625 471

375 195 225 – 13 235 67 62 58 238 98 111 112 1789 625 471

375 198 222 – 13 157 45 41 39 356 147 167 168 1927 625 471

375 201 219 – 13 78 22 21 19 475 196 222 224 2066 625 471

375 204 216 – 13 0 0 0 0 594 245 278 280 2204 625 471

375 188 – 230 13 418 120 111 104 0 0 0 0 1558 625 435

375 191 – 227 13 335 96 89 83 127 52 59 60 1706 625 435

375 195 – 224 13 251 72 66 62 254 104 119 119 1854 625 435

375 198 – 220 13 167 48 44 42 381 157 178 179 2002 625 435

375 202 – 217 13 84 24 22 21 507 209 237 239 2150 625 435

375 206 – 214 13 0 0 0 0 634 261 296 299 2298 625 435

0.30 0.81 0.40

0.83

0.78

0.78

0.78

0.78

0.30 0.91 0.40

0.91

0.91

0.91

0.91

0.91

2.51 0.40

2.10 0.40

1.86 0.40

1.70 0.40

1.59 0.40

1.50 0.40

2.51 0.40

2.10 0.40

1.86 0.40

1.70 0.40

1.59 0.40

1.50 0.40

0/2–2/4–4/8–8/16 are the grading of aggregates.

Fig. 3. V-funnel test.

viscosity of concrete. Shear stress and viscosity of fresh concrete decreased as the amount of water increased for all mixtures. Fly ash and blast-furnace slag replacing instead of cement increased the flow rate of the concrete. On the condition that the w/(c + m) rate and the type of mineral additive were constant, when the pumice rate was increased in the aggregate, the time to flow a diameter of 50 cm extended (Fig. 6). The mineral additives slow the strength gain because of their low pozzolanic activity when the amount of mineral additive increased in the mixture. As a result, loss of workability decreased

generally as mineral additive and amount of water increased. The V-funnel flow times were measured as 17–24 s in the first group of experiments. Besides, in this group, increasing pumice powder rate decreased the passing ability of SCC, and flow time was extended because of the segregation as w/(c + m) rate exceeded the optimum value. The V-funnel flow times were measured as 8–26 s in the second group of experiments. The increasing lightweight aggregate rate, i.e. decrease of unit weight, extended the V-funnel flow time. For the second group of experiments, the relationship between unit weight of fresh concrete and V-funnel flow time was given in Fig. 7. Khurana and Topçu [26] has suggested boundary values as following; 8–12 s for Dmax = 15 mm, for 11–15 s for Dmax = 20 mm, for the flow times through V-funnel with 5  5 cm span of SCC with different maximum grain size (Dmax). The EFNARC [23] Committee has also suggested that the V-funnel flow time becomes 6–12 s. In this test, the exit time of concrete through orifice is measured. Extension of this time in SCLC is ordinary, since increasing weight of the mix enhanced the spreading capability at the fresh stage. Therefore, the V-funnel flow times on the amount of recommended values can be considered suitable for V-funnel flow time. The difference of V-funnel time delayed 5 min was 8–13 s Group I. The EFNARC [23] Committee indicated that if there is a difference of more than 3 s according to first flow time, static segregation occurs. It was seen that the flow time decreased in w/(c + m) ratio 0.35–0.40 mixtures, but flow time increased in w/(c + m) ratio 0.45 mixtures as the amount of water increased. Therefore, the static segregation risk increases, since the viscosity of fresh concrete decreases, as the amount of water increases on the optimum value in fresh concrete. As a result of this, V-funnel time, delayed 5 min, extends. Besides, the static stability of fresh concrete has increased and V-funnel time, delayed 5 min, has been shortened since the viscosity of fresh concrete has increased without further increasing the shear stress of the fresh concrete as the amount of fly ash has increased. The V-funnel time, delayed 5 min, increased when the lightweight aggregate ratio increased as w/(c + m) was constant. The blocking occurred in the mixtures produced with the pumice aggregate of 100% and lack of air entraining admixture.

41

M. Kurt et al. / Construction and Building Materials 103 (2016) 36–46 Table 6 The results of fresh concrete experiments for Group I mixture. Mixtures Flow diameter T50 flow time Flow diameter after 1 hour V-funnel flow time V-funnel time delayed 5 min. L-box ratio Unit weight of fresh concrete Ambient temperature

mm s mm s s h2/h1 kg/m3 °C

CS1

CS2

CS3

PP1

PP2

PP3

PP4

PP5

PP6

PP7

PP8

PP9

600 7 550 14 17 0.81 1185 18

640 6 580 11 15 0.88 1175 18

650 5 600 9 14 0.88 1160 19

590 8 540 20 25 0.77 1175 24

620 6 570 18 22 0.77 1163 22

640 6 600 17 20 0.81 1154 18

570 11 530 22 27 0.77 1175 25

600 7 570 19 23 0.77 1164 25

630 6 600 17 21 0.81 1158 23

560 10 500 24 30 0.73 1176 24

590 8 550 21 26 0.77 1166 24

610 8 570 19 23 0.81 1158 23

Table 7 The results of fresh concrete experiments for Group II mixture. Mixtures

FA

Flow diameter T50 flow time V-funnel flow time V-funnel time delayed 5 min L-box ratio Unit weight of fresh concrete Ambient temperature

mm s s s h2/h1 kg/m3 °C

BFS

NP100

NP80

NP60

NP40

NP20

NP0

NP100

NP80

NP60

NP40

NP20

NP0

650 8 21 Blocked 0.77 1357 16

670 7 14 20 0.81 1500 18

730 4 12 16 0.88 1712 17

750 4 11 14 0.93 1858 16

760 3 10 11 0.93 2074 17

800 2 8 9 0.93 2303 18

645 9 26 Blocked 0.77 1364 17

660 6 19 Blocked 0.77 1480 17

710 4 13 19 0.81 1719 16

730 3 12 16 0.88 1834 17

750 3 10 13 0.88 2087 16

770 3 9 11 0.93 2351 17

L-box (h2/h1) ratio and unit weight of fresh concrete in Group II was shown in Fig. 8. Therefore, the L-box (h2/h1) ratio is equal to 1 (one) in a very fluid material. The report of the EFNARC [23] Committee indicated that if this ratio is smaller than 0.8, there is a risk of aggregate being blocked. However, Bernabeu and Laborde [27] reported that the mixtures of L-box ratio 0.65 (flow diameter of 60 cm) easily filled the formwork, according to results of their experiments. The localized aggregate source/type etc. in the work of Bernabeu and Laborde [27] showed the main difference with the EFNARC [23]; which is a standardization process, including not only general criteria for aggregate blocking, but also the other selfcompactability affecting parameters. Fig. 4. L-box test.

3.2. Hardened concrete results FA

BFS

Linear (FA)

Linear (BFS)

Spread Diameter (mm)

850 800

y = 0.16x + 444.40 R² = 0.95

750 700

y = 0.13x + 475.14 R² = 0.95

650 600 1200

1400

1600

1800

2000

2200

2400

Unit weight of fresh concrete (kg/m3) Fig. 5. Effect of unit weight to flow diameter.

The L-box (h2/h1) ratios were measured as 0.73–0.88 and 0.77– 0.93 in the first and second group of experiments, respectively. It was seen that the L-box (h2/h1) ratio increased as the amount of mineral additives and w/(c + m) ratio increased for all mixtures. Besides, it increased when the amount of normal aggregates increased in all mixtures, i.e. the passing ability increased as unit weight of fresh concrete increased. The relationship between

The hardened concrete properties were examined separately for the first and second group of experiments as in fresh concrete test results. The test results of dry unit weight, compressive strength for 7, 28, 90 and 180 days, splitting tensile strength for 28 days, thermal conductivity and ultrasonic pulse velocity (Fig. 9) were given in Tables 8 and 9 for the first and second groups of specimens, respectively. The ultrasonic pulse velocities were measured by a pulse meter with an associated transducer pair. The transducer pair had a nominal frequency of 54 kHz. The principle of ultrasonic pulse velocities measurement involves sending a wave pulse into concrete and measuring the travel time for the pulse to propagate through the samples. The pulse is generated by a transmitter and received by a receiver. In the experimental studies, the transmitter and receiver were placed at the top and bottom surfaces of a cylindrical specimen, respectively. For each mixture, three samples of 100  200 mm cylinders (totally 432 specimens) were prepared and cured in lime-saturated water at 20 ± 3 °C until the testing time. At the testing age, the samples were tested for compressive strength, and splitting-tensile strength in accordance with ASTM C-192 and ASTM C-496, respectively. In the first group, the dry unit weights of concrete samples were specified between 840–1011 kg/m3 and 1014–1037 kg/m3 for those produced with lightweight aggregate of 100% and control samples not containing mineral additive. Although there is not a very big difference between them, replacing of pumice powder

42

M. Kurt et al. / Construction and Building Materials 103 (2016) 36–46

for BFS

Time to spread diameter of 50 cm (sec)

Time to spread diameter of 50 cm (sec)

for FA 10

y = 2.13e0.01x R² = 0.95

8 6 4 2 0

10

y = 2.42e0.01x R² = 0.84

8 6 4 2 0

0

20

40

60

80

100

0

20

Light weight aggregate (%)

40

60

80

100

Light weight aggregate (%)

Fig. 6. The relationship between lightweight aggregate rate with time to spread a diameter of 50 cm for fly ash (on the left) and blast-furnace slag (on the right) mineral additives.

for FA

for BFS 30 y = 1,955,167.00x-1.60 R² = 0.93

20

V-funnel flow time (sec)

V-funnel flow time (sec)

25

15

10

5 1200

1400

1600

1800

2000

2200

2400

Unit weight of fresh concrete (kg/m 3)

y = 21,673,042.02x-1.91 R² = 0.94

25 20 15 10 5 1200

1400

1600

1800

2000

2200

2400

Unit weight of fresh concrete (kg/m 3)

Fig. 7. Effect of unit weight to V-funnel flow time for fly ash (on the left) and blast-furnace slag (on the right) mineral additives.

FA

BFS

Linear (FA)

Linear (BFS)

1.00 y = 0.00x + 0.55 R² = 0.82

L-box (h2/h1)

0.95 0.90 0.85

y = 0.00x + 0.53 R² = 0.92

0.80 0.75 0.70 1200

1400

1600

1800

2000

2200

2400

Unit weight of fresh concrete (kg/m3) Fig. 8. The relationship between L-box (h2/h1) ratio and unit weight of fresh concrete in Group II.

instead of cement in mixture has reduced unit weight. Besides, the unit weights of concretes reduced as w/(c + m) ratio and amount of mineral additive increased in mixture. The reason of this is that the increase of the spaces in concrete structure with the increase in the w/(c + m) ratio and mineral additives replaced instead of cement have lower specific gravity than cement. The dry unit weights of mixtures in the second group were also found between 1187 and 2278 kg/m3. In this group, the unit weight of concrete significantly decreased with the increasing lightweight aggregate ratio. Demirbog˘a [28] indicated that depending on the production method of lightweight concretes and type of aggregate, the unit weight of lightweight concretes might vary between 1360 and 1840 kg/m3 for structural lightweight concretes and 320–1120 kg/m3 for heat insulating concretes.

Fig. 9. Schematic diagram of pulse velocity measurement.

The SCLC produced with pumice 100% is lighter than normal concrete when the unit weight of normal concrete is considered to be around 2300 kg/m3. Since, the unit weights of all other mixtures except for mixtures containing normal aggregate 100% and 80% is lower than 1840 kg/m3, these are incorporated into the class of structural lightweight concretes. The compressive strengths of concrete in the first group for 7, 28, 90 and 180 ages were found between 7.1–9.3, 10.5–13.2, 11.1–14.5 and 11.7–15.3 MPa, respectively as seen in Table 8. The changes of compressive strength depending on the age of concrete are given in Fig. 10. The compressive strengths for 7, 28, 90 and 180 days in the second group of experiments were given in Table 8. The relationship

43

M. Kurt et al. / Construction and Building Materials 103 (2016) 36–46 Table 8 The hardened concrete properties of the samples. MA %

w/(c + m)

MIX

7 day

28 day

90 day

180 day

0

0.35 0.40 0.45

CS1 CS2 CS3

11.8 10.3 8.8

13.9 12.7 10.6

14.4 13.1 11.9

20

0.35 0.40 0.45 0.35 0.40 0.45 0.35 0.40 0.45

PP1 PP2 PP3 PP4 PP5 PP6 PP7 PP8 PP9

9.3 8.2 7.2 8.7 7.8 7.1 8.1 7.2 7.1

13.2 11.7 10.7 12.5 11.4 10.5 12.3 11.5 10.5

40

0.30

FANP100 FANP80 FANP60 FANP40 FANP20 FANP00

14.0 15.6 18.7 21.6 25.3 32.4

40

0.30

BFSNP100 BFSNP80 BFSNP60 BFSNP40 BFSNP20 BFSNP00

14.5 17.1 20.7 24.1 28.4 52.3

30

40

STS (MPa)

Dry unit weight (kg/m3)

Water absorption (%)

Thermal conductivity (W/mK)

UPV (m/s)

15.5 14.3 12.3

1.9 1.9 1.7

1037 1028 1014

12.62 14.38 16.03

0.613 0.598 0.545

2867 2815 2685

14.5 12.9 11.1 13.8 12.6 11.5 13.5 12.6 11.6

15.3 13.6 11.7 14.5 13.2 12.1 14.3 13.3 12.2

1.9 1.8 1.7 1.8 1.8 1.7 1.8 1.8 1.7

1011 962 915 988 935 884 980 915 840

17.0 21.7 25.9 17.8 22.5 26.1 18.1 23.3 29.1

0.565 0.473 0.384 0.532 0.434 0.362 0.496 0.403 0.347

2884 2781 2685 2872 2767 2659 2838 2719 2611

19.9 22.2 26.7 30.9 36.1 53.3

21.4 23.9 28.7 33.2 38.8 57.2

22.3 24.9 29.9 34.6 40.4 60.0

1.8 2.2 2.2 3.1 3.4 4.2

1187 1392 1622 1780 1976 2156

16.8 12.8 10.7 8.6 7.2 5.7

0.606 0.629 0.845 1.078 1.238 1.508

3022 3256 3549 3743 3994 4340

21.3 25.1 26.8 32.6 37.8 65.0

23.2 27.0 28.9 35.0 40.7 71.2

23.7 27.5 29.5 35.7 41.5 72.6

2.1 2.8 3.0 3.3 4.0 7.3

1266 1411 1630 1791 2079 2278

19.0 15.9 14.3 14.6 12.8 8.4

0.642 0.667 0.938 1.132 1.581 1.694

3152 3363 3578 3813 4124 4770

Compressive strength (MPa)

Table 9 Water absorption and dry unit weight of the mixtures. Mixture number

w/(c + m)

Dry weight (g)

Saturated surface dry weight (gr)

Water absorption by weight (%)

Water absorption by volume (%)

Dry unit weight (kg/m3)

20 20 20 30 30 30 40 40 40

0.35 0.40 0.45 0.35 0.40 0.45 0.35 0.40 0.45

1608 1571 1533 1594 1558 1533 1579 1547 1527

1838 1859 1865 1834 1881 1869 1822 1823 1876

14.3 18.4 21.6 15.1 20.7 22.0 15.4 17.8 22.9

14.6 18.4 21.1 15.3 20.6 21.4 15.5 17.5 22.3

1024 1000 976 1015 992 976 1005 985 972

FANP00 FANP20 FANP40 FANP60 FANP80 FANP100

0 20 40 60 80 100

0.30 0.30 0.30 0.30 0.30 0.30

3387 3104 2796 2548 2187 1865

3578 3328 3036 2820 2466 2177

5.7 7.2 8.6 10.7 12.8 16.8

12.2 14.3 15.3 17.3 17.8 19.9

2156 1976 1780 1622 1392 1187

BFSNP00 BFSNP20 BFSNP40 BFSNP60 BFSNP80 BFSNP100

0 20 40 60 80 100

0.30 0.30 0.30 0.30 0.30 0.30

3578 3266 2813 2561 2217 1989

3710 3467 3043 2785 2467 2287

3.7 6.2 8.2 8.8 11.3 15.0

8.4 12.8 14.6 14.3 15.9 19.0

2278 2079 1791 1630 1411 1266

PP1 PP2 PP3 PP4 PP5 PP6 PP7 PP8 PP9

Mineral additive (%)

between lightweight aggregate ratio and time-dependent compressive strength was also shown in Fig. 11. Generally, the compressive strengths of mixtures including pumice powder for all curing periods stayed under control as samples not including mineral additive. This is because, mineral additive replaced instead of cement gained strength more slowly than cement. Besides, the strength development was provided in all w/(c + m) ratios because of the increase effect on the workability of pumice powder. The decrease of strength and unit weight with an increase of the amount of lightweight aggregate in mixture is explicit. In the second group of experiments, when the ratio of lightweight aggregate was 20% in concrete mixture unit weight, 28-day compressive

strength decreased approximately 8% and 32% for FA, 9% and 42% for BFS, respectively. The water absorption values of concrete samples in Group I and control samples were 17.0–29.1% and 12.6– 16.0% by weight, respectively. Those were also specified as 5.7– 16.8% and 8.4–19.0% for FA and BFS in the second group of mixtures, respectively (Table 9). The water absorption ratios of control samples not containing mineral additive were less for the same w/ (c + m) ratio. Furthermore, the increase rate of mineral additive in the mixture also increased the water absorption ratios. The water absorption ratios of the control samples not containing mineral additive had been less than the samples with lightweight aggregate at the same rate and pumice powder. Furthermore, the

44

M. Kurt et al. / Construction and Building Materials 103 (2016) 36–46

Compressive strength (MPa)

18 7 days

16

28 days

14

90 days

12

180 days

10 8 6 4 2

0 Mixture number

PP1

PP2

PP3

PP4

PP5

PP6

PP7

PP8

PP9

w/(c+m)

0.35

0.40

0.45

0.35

0.40

0.45

0.35

0.40

0.45

20

20

20

30

30

30

40

40

40

Mineral add. (%)

Fig. 10. The time-dependent compressive strength of the samples in Group I.

7 days

28 days

90 days

180 days

for BFS

Compressive strength (MPa)

Compressive strength (MPa)

for FA 80 70 60 50 40 30 20 10 0 0

20

40

60

80

7 days

28 days

90 days

180 days

80 70 60 50 40 30 20 10 0

100

0

20

40

60

80

100

Light weight aggregate (pumice) ratio (%)

Light weight aggregate (pumice) ratio (%)

0.65

Coefficient of thermal conducvity (W/mK)

Coefficient of thermal conducvity (W/mK)

Fig. 11. The relationship between lightweight aggregate ratio and compressive strength of Group II for fly ash (on the left) and blast-furnace slag (on the right) mineral additives.

y = 2E-09x2.8469 R² = 0.9407

0.55 0.45 0.35 0.25 820

860

900

940

980

1020

Dry unit weight (kg/m3)

2.00 1.80 1.60 1.40 1.20 1.00 0.80 0.60 0.40 1000

FA Exp. (FA) Exp

BFS Exp. (BFS) Exp

y = 0.16e0.00x R² = 0.98

y = 0.17e0.00x R² = 0.97 1250

1500

1750

2000

2250

2500

Dry unit weight (kg/m3)

Fig. 12. The relationship between thermal conductivity coefficient and dry unit weight for Group I (on the left) and II (on the right).

increase in the rate of water absorption was an expected result since the increase in the rate of lightweight aggregate in mixture caused the increase of space volume in concrete. The coefficients of thermal conductivity of SCLC in the control and first group samples were found as 0.545–0.613 W/mK and 0.347–0.565 W/mK, respectively. The decrease of thermal conductivity was an expected result since the volume of space in concrete increased as the w/(c + m) ratio increased. In addition, the thermal conductivity of the mixtures decreased as rate of pumice powder with lower unit weight than cement increased. The control samples not containing mineral additive had the higher thermal conductivity value for same w/(c + m) ratio. It was understood from all these results that thermal conductivity of SCLC to the unit weight as in conventional concrete. This relationship was shown for the first group mixtures in Fig. 12.

The coefficients of thermal conductivity of SCLC in the second group were also found as 0.606–1.508 W/mK for FA and 0.642– 1.694 W/mK for BFS. The factors affecting the coefficient of thermal conductivity in the first group of experiments also generated similar effects for the second group of experiments. On the other hand, the increase in rate of lightweight aggregate decreased the coefficient of thermal conductivity as expected. Therefore, the thermal conductivity decreased 18%, 44%, 60% and 7%, 45%, 62% for FA and BFS in second group when the lightweight aggregate ratios in the mixture were 20%, 60%, and 100%, respectively. The changes in the coefficients of thermal conductivity with the rate of lightweight aggregate and dry unit weight for the second group were shown in Fig. 12. As a result, the path length of the ultrasonic pulse was the length of the specimen, which was measured using a vernier with

45

M. Kurt et al. / Construction and Building Materials 103 (2016) 36–46

Ultrasonic pulse velocity (UPV) (m/s)

Ultrasonic pulse velocity (UPV) (m/s)

BFS Log. (BFS) 3000 2900 2800 2700 y = 96.84x + 1634.22 R² = 0.93

2600 2500 10

11

12

13

14

Compressive strength for 28 days (MPa)

FA Log. (FA)

y = 1464.86ln(x) -1293.07 R² = 0.99

5000 4500 4000 3500

y = 1337.54ln(x) -890.65 R² = 0.98

3000 2500 10

20

30

40

50

60

70

Compressive strength for 28 days (MPa)

Fig. 13. The relationship between ultrasonic pulse velocity and 28-day compressive strength for Group I (on the left) and II (on the right).

a minimum reading of 0.01 mm. The ultrasonic pulse velocities of SCLC in the first and second group of experiments were found between 2611–2884 m/s and 3022–4770 m/s, respectively. The variables affecting the compressive strength also affected ultrasonic pulse velocity. Therefore, the ultrasonic pulse velocities also increased linearly with the increase in the compressive strength. The ultrasonic pulse velocities of the samples in the second group also reduced significantly with the increase in the amount of lightweight aggregate in mixture as in compressive strength. The relationship between ultrasonic pulse velocities and 28-day compressive strength of samples in the first and second groups were given in Fig. 13. 4. Conclusions Some results and recommendations that can be inferred from all of this experimental study are summarized below. (1) In all mixtures, the increase in the amount of pumice powder did not improve the self-compatibility properties. As a result of visual inspection of pumice, the shear stress of concrete may have decreased because of more porous and roughness surface structure of pumice. (2) Because of the pumice and natural aggregate usage together in SCLC, the unit weight, thermal conductivity and ultrasonic pulse velocity values of the concrete samples decreased while the compressive strengths and water absorption ratios increased with replacing fly ash instead of cement in the second group. However, this reason is not related to the fly ash but the pumice ratio. (3) It was seen that the workability was also easier as density of SCLC increased. This was an expected result since the compacting and spreading properties of SCC were provided by its own weight. (4) The increase of lightweight aggregate rate (i.e. decrease of density) increased the V-funnel flow time of SCC. There are two reasons for this. First, the flow time extends as unit weight decreases since the flow reason of concrete is its own weight exceeding the threshold stress. Therefore, the V-funnel flow time which was a little above the recommended value may be acceptable. The second reason is that the increase in the amount of lightweight aggregate in mixture also increased the tendency to segregation of fresh concrete. (5) The flow test is not sufficient alone to determine the selfcompatibility. While the spread and V-funnel tests are used in the control of flowability properties, L-box or U-box and V-funnel time delayed 5 min flowability should be used to specify the passing ability and resistance to segregation, respectively.

(6) When the humidity control of aggregate was neglected and was not considered in concrete mix design, the significant fluctuations occurred in the data of self-compactibility tests. Therefore, especially grading and moisture content of pumice aggregate should be checked frequently and a quality entry plan with common-period must be created for this. Besides, the pumice aggregate should be stocked in closed storage areas and to be made provision for its properties not to change up to production stages. (7) The water absorption rate of lightweight aggregates is quite high. The continuance to adsorb the water of lightweight aggregate has affected the homogeneity of experiments during mixing and fresh concrete tests. Hence, the saturated lightweight aggregates should be used in future studies. (8) In the case of blended cement use in mixture, the type of mineral additives used in cement mixture should be determined and considered in the mixture account. If the type of mineral additives is not considered, the self compactability and strength evaluation process of the mixture may become a phenomenon. (9) It was determined that the percentage of entrained air is very important to avoid the segregation problems. Despite that the smaller amount of air is entrained, the production of mixture not to generate the segregation problems will enable the production of concrete with higher hardened concrete properties. (10) In the future studies, the adequacy of adherence with reinforcement, shrinkage conditions, and durability properties of SCLC should be investigated. Besides, the lightweight aggregate and normal aggregate combinations may lead to maximum compacting, and less segregation may be investigated to leave the minimum air void in the design of SCLC.

References [1] A.C. Aydin, A. Oz, R. Polat, H. Mindivan, Effects of the different atmospheric steam curing processes on the properties of self-compacting-concrete containing microsilica, SADHANA Acad. Proc. Eng. Sci. 40 (4) (June 2015) 1361–1372. _ Türkmen, A.C. Aydın, Thermo-mechanıcal propertıes of [2] R. Gül, E. Okuyucu, I. fiber reınforced raw perlıte concrete, Mater. Lett. 61 (29) (2007) 5145–5149. [3] A.C. Aydın, A. Arslan, R. Gül, Mesoscale simulation of cement based materials’ time dependent behavior, Comput. Mater. Sci. 41 (1) (2007) 20–26. [4] A.C. Aydın, Self compactibility of high volume hybrid fiber reinforced concrete, Constr. Build. Mater. 21 (2007) 1146–1151. [5] A.C. Aydın, R. Gül, Influence of volcanic originated natural materials as additives on the setting time and some mechanical properties of concrete, Constr. Build. Mater. 21 (2007) 1274–1278. [6] A.C. Aydın, A. Tortum, M. Yavuz, Prediction of concrete elastic modulus using adaptive neuro-fuzzy inference system, Civ. Eng. Environ. Sys. 23 (2006) 295– 309. [7] O.A. Düzgün, R. Gül, A.C. Aydın, Effect of steel fibers on the mechanical properties of natural lightweight aggregate concrete, Mater. Lett. 59 (2005) 3357–3363.

46

M. Kurt et al. / Construction and Building Materials 103 (2016) 36–46

[8] A. Tortum, C. Celik, A.C. Aydın, Determination of the optimum conditions for tire rubber in asphalt concrete, Build. Environ. 40 (2005) 1492–1504. [9] E. Og˘uz, A.C. Aydın, Prediction of adsorption velocity of phosphate removal from wastewater with gas concrete, including Ph, temperature, and agitation speed, Int. J. Environ. Pollut. 19 (2003) 603–614. [10] U.C. Hasar, O. Simsek, A.C. Aydın, Application of varying-frequency amplitudeonly technique for electrical characterization of hardened cement-based materials, Microwave Opt. Technol. Lett. 52 (4) (2010) 801–805. [11] U.C. Hasar, G. Akkaya, M. Aktan, C. Gozu, A.C. Aydın, Water-to-cement ratio prediction using ANNs from non-destructive and contactless microwave measurements, Prog. Electromag. Res. 94 (2009) 311–325. [12] A.C. Aydın, M.B. Karakoç, O.A. Düzgün, M.S. Bayraktutan, Effect of low quality aggregates on the mechanical properties of lightweight concrete, Sci. Res. Essays 5 (10) (2010) 1133–1140. [13] A.C. Aydın, O.A. Düzgün, A. Tortum, Determination of the optimum conditions for steel fibers on the mechanical properties of natural lightweight aggregate concrete, Pollack Perodika 3 (1) (2008) 101–112. [14] O. Boukendakdji, S. Kenai, E.H. Kadri, F. Rouis, Effect of slag on the rheology of fresh self-compacted concrete, Constr. Build. Mater. 23 (2009) 2593–2598. [15] B. Sukumar, K. Nagamani, Raghavan R. Srinivasa, Evaluation of strength at early ages of self-compacting concrete with high volume fly ash, Constr. Build. Mater. 22 (2008) 1394–1401. [16] R. Siddique, Properties of self-compacting concrete containing class F fly ash, Mater. Des. 32 (2011) 1501–1507. [17] Z. Wua, Y. Zhang, J. Zheng, Y. Ding, An experimental study on the workability of self-compacting lightweight concrete, Constr. Build. Mater. 23 (2009) 2087– 2092. [18] M. Kurt, A.C. Aydin, M.S. Gül, R. Gül, T. Kotan, The effect of fly ash to selfcompactibility of pumice aggregate lightweight concrete, SADHANA Acad. Proc. Eng. Sci. 40 (4) (2015) 1343–1359.

[19] J.M. Khatib, Performance of self-compacting concrete containing fly ash, Constr. Build. Mater. 22 (2008) 1963–1971. [20] E. Güneyisi, M. Gesog˘lu, E. Booya ve, K. Mermerdasß, Strength and permeability properties of self-compacting concrete with cold bonded fly ash lightweight aggregate, Constr. Build. Mater. 74 (2015) 17–24. [21] E. Güneyisi, M. Gesog˘lu, E. Booyave, Fresh properties of self-compacting cold bonded fly ash lightweight aggregate concrete with different mineral admixtures, Mater. Struct. 45 (2012) 1849–1859. [22] T. Kotan, R. Gül, Effect of atmospheric pressure steam curing to mechanical properties of lightweight concrete produced with Erzurum-Pasinler pumice, Mach. Technol. Mater. Int. Vir. J. 4–5 (2010) 66–69. [23] EFNARC, Specifications and Guidelines for Self-Compacting Concrete, EFNARC, Association House, 99 West Street, Farnham, UK, 2002. www.efnarc.org, ISBN 0 953973344, p. 32. [24] M. Haist, V. Mechtcherine, H. Beitzel, H.S. Müller, O. Wallevik, I. Nielsson, Retrofitting of building structures using pumpable self-compacting lightweight concrete, International RILEM Symposium on Self-Compacting Concrete, RILEM Publications SARL, 2003. pp. 776–785. [25] A.J. Dowson, The application, self-compacting concrete (SCC) in precast products, in: BIBM 17th International Congress of the Precast Concrete Industry, 2002, Istanbul (Turkey), 2002, pp. 425–472. [26] R. Khurana, O. Topçu, Role of Superplasticizers in the development of selfcompacting concrete, in: Second International Symposium on Cement and Concrete Technology in the 2000s, vol. I, Istanbul, Turkey, 2000, pp. 473–482. [27] M. Bernabeu, Laborde, Brite Euram Project: Rational Production and Improved Working Environment Through Using Self Compacting Concrete- FINAL REPORT: TASK 8.3 Production System for Civil Engineering, GTM Construction, 2000. _ Örüng, R. Gül, Effects of expanded perlite aggregate and [28] R. Demirbog˘a, I. mineral admixtures on the compressive strength of low-density concretes, Cem. Concr. Res. 31 (2001) 1627–1632.