STRENGTH AND DURABILITY OF COMPOSITE CONCRETES WITH MUNICIPAL WASTES

5 The influence of different types of polyethylene (PE) substitutions as partial aggregate 6 replacement of micro-steel fiber reinforced self-consolidating concrete (SCC) incorporating 7 incinerator fly ash was investigated. The study focuses on the workability and hardened 8 properties including mechanical, permeability properties, sulfate resistance and 9 microstructure. Regardless of the polyethylene type, PE substitutions slightly decreased the 10 compressive and flexural strength of SSC initially, however, the difference was compensated 11 at later ages. SEM analysis of the interfacial transition zone showed that there was chemical 12 interaction between PE and the matrix. Although PE substitutions increased the permeable 13 porosity and sorptivity, it significantly improved the sulfate resistance of SCC. The influence of 14 PE shape and size on workability and strength was found to be more important than its type. 15 When considering the disposal of PE wastes and saving embodied energy, consuming recycled 16 PE as partial aggregate replacement was more advantageous over virgin PE aggregate 17 replaced concrete. 18


INTRODUCTION
In the last decades, sustainable development in the construction industry has been gaining 1 increasing attention. Sustainable development can combine economic growth and 2 environmental protection by conserving natural resources and saving embodied energy. 3 Recycling of waste materials has been accepted as one of the most beneficial option to achieve 4 sustainable development for construction industry 1 . Depending on the availability and price, 5 several industrial wastes can be used as parts of the binder, i.e. cement and the filler (natural 6 aggregate). For example, industrial by-products such as fly ash, municipal fly ash, ground 7 granulated blast furnace slag and silica fumes have been used in construction industry as 8 cement replacement or supplementary cementitious materials 1 . Recycled concrete 2 which is 9 produced from demolishing concrete structures and recycled polymers 3 obtained from waste 10 polymers are the most common wastes used as natural aggregate substitutes in the building 11 industry 4,5 . 12 13 Construction industry has been using recycled polymeric wastes as aggregate and fiber, 14 because of its economic and ecological advantages 3,6-8 . Different types of polymeric wastes 15 such as polypropylene (PP) 9,10 , polyethylene (PE) [11][12][13] , and polyethylene terephthalate (PET) 14-16 18 have been used as filler in concrete. However, most of these studies were conducted for 17 conventional concrete and polymeric wastes were utilized as fine aggregate replacement. 18 Therefore, little information is presently known regarding the use of polymers as coarse 19 aggregate in the formulation of new concretes, especially self-consolidating concrete (SCC). 20 21 Qatar has been one of the largest producer and consumer of polymers in the Gulf region 19,20 . 22 Effective disposal of polymeric wastes are constrained by its non-biodegradable nature and 23 emission of dangerous gases when combusted. Therefore landfilling and incineration are not 24 mixtures in order to compare the fiber reinforced property of the polyethylene fiber 1 incorporated mixtures, the results of the comparison has been presented in another 2 publication of the authors 52 . 3 4 As plastic waste material, virgin high density polyethylene (v-HDPE) samples collected from 5 Qatar Chemical Company (QCHEM) and virgin low density polyethylene (v-LDPE) samples 6 collected from Qatar Petrochemical Company (QAPCO) (Fig. 3) were used in the mixtures in 7 order to compare their effect on the workability, durability and mechanical properties of SCC. 8 Both v-HDPE and v-LDPE was in the form of spherical granules with an average diameter of 9 3.0 mm (0.118 in.). Recycled polyethylene granules were collected from one of the plastic 10 recycling companies in Qatar. In this plant, firstly HDPE and LDPE municipal plastic wastes 11 were separated and then they were processed. Separated HDPE and LDPE wastes were 12 washed and crushed into scraps. Prior to extrusion, these crushed scraps were dried and then 13 fed into the extruder and extracted as plastic wires. Finally, these plastic wires were cut into 14 cylindrical granules by cutter (Fig. 3). The average diameter and length were 4.0 mm (0.157 15 in.) and 3.5 mm (0.138 in.) for r-HDPE, and 3.0 mm (0.118 in.) of both for r-LDPE, respectively. 16 The properties of polyethylene used in this research are given in Table 2, while the shape and  17 size of polyethylene aggregates along with their SEM images are presented in Fig. 3. 18 19 Within the scope of experimental program, five concrete mixtures have been prepared and 20 summarized in Table 3. The control mixture included OPC, MSWI FA, silica fume as 21 cementitious materials and steel fiber. In the remaining mixtures, 10% by weight of coarse 22 aggregate was replaced with virgin and recycled polyethylene granules. The partial PE 23 substitution in this research kept 10% by weight as higher PE substitution rates were reported 24 to decrease the mechanical strength of manufactured concrete drastically 11-13 while lower 1 rates may not be feasible in terms of economy. For all mixtures, the amount of OPC, MSWI FA, 2 silica fume, steel fiber and superplasticizer (SP) content were kept constant to reduce the 3 number of cases to be studied in this study. As seen in Table 3, the only variable was the type 4 of polyethylene aggregate substituted in the mixes 2-5. 5 6 SCC mixtures were prepared using an electrically driven concrete mixer. The preparation 7 procedure was the same for all mixtures: firstly all aggregates (sand, coarse and PE), cement, 8 MSWI FA, silica fume and steel fiber were mixed in a dry state. Then three quarters of mixing 9

Mixture proportions
water mixed with the superplasticizer (SP) was added in the mixer and the mixture was mixed 10 for 2 min period. The remaining water was added gradually into the mixture to provide 11 uniformity in the mixture and mixed for a period of 2 more min. After completing the mixing 12 procedure, fresh concrete tests including V-funnel, slump flow time and diameter and setting 13 time tests were performed on the mixtures. From each concrete mixture, twelve Ø100x200 14 mm (Ø4x8 in.) cylinder specimens were cast for determination of compressive strength and 15 permeability properties including water absorption, sorptivity and rapid chloride permeability 16 tests, and six 160x40x40 mm (6.30x1.57x1.57 in.) beam specimens were cast for the 17 determination of flexural strength, and nine 280x25x25 mm (11x1x1 in.) bar specimens were 18 cast for sulfate exposure determinations in accordance with the related ASTM standards. Note 19 that all specimens were cast in one layer without compaction as all mixtures were accepted 20 as SCC. After 24 h, the specimens were demoulded and stored in water tank till the age of 21 testing. 22 Tests on fresh concrete Tests on hardened concrete 12 Tests performed on hardened concrete can be grouped into two as tests to evaluate 13 mechanical and durability properties of SCC. 14 Mechanical properties 15 In order to determine mechanical properties, the compressive strength of SCC specimens was 16 determined at 7, 28 and 90 days in accordance with ASTM C39 57 the flexural strength of SCC 17 specimens was determined at 7 and 28 days in accordance with ASTM C293 58 . For mechanical 18 tests, three specimens from each mixture was tested and average of these were calculated. 19

Durability properties 20
Permeability properties 21 To determine the permeability properties, 28 days age of Ø100x200 mm (Ø4x8 in.) cylinder 22 specimens were cut into Ø100x50 mm (Ø4x2 in.) disc specimens and the absorption, sorptivity 23 and rapid chloride permeability tests (RCPT) were performed on mentioned specimens. Water 24 absorption which is defined as the amount of water absorbed under specified conditions was 25 determined in accordance with ASTM C642 59 . As water absorption can only take place in pores 26 which were emptied during drying and filled with water during the immersion period, water 1 absorption indicates the degree of permeable porosity of a material. 2 The sorptivity was determined in accordance with ASTM C1585 60 . In this test, the rate of 3 absorption of water by unsaturated SCC specimens were measured by the increase in the mass 4 of a disc specimen at given intervals of time (1,5,10,20,30,60,180,240, 300 and 360 min) 5 when only one surface of the specimen was exposed to water, with the depth of water 6 between 3 to 5 mm (0.12 to 0.20 in.). 7 8 The RCPT test was performed to determine the chloride permeability of 28 days SCC 9 specimens in accordance with ASTM C1202 61 . At the end of this test, the total charge passed, 10 in coulombs, is determined. Higher coulombs value indicates lower resistance to chloride ion 11 penetration, while lower coulombs value indicates higher resistance. 12

Sulfate resistance 13
Sulfate resistance tests were performed on nine 280x25x25 mm (11x1x1 in.) bar specimens 14 per mix. To determine the sulfate resistance, the bars were immersed into sulfate solution in 15 accordance with ASTM C1012 62 and the length changes were measured at 1 week, 2 weeks, 3 16 weeks, 4 weeks, 8 weeks, 13 weeks, 15 weeks, 4 months and 6 months. 17

18
Fresh concrete properties 19 The  (19.69 in.) for the Japan Society for Civil Engineers (JSCE) 64 . Moreover, since all the 1 mixtures in this study filled the molds by their own weight without any vibration, and neither 2 segregation nor considerable bleeding was visually observed in any of the mixtures during 3 mixing, testing and casting, they were accepted as SCC. As seen from Table 4 blocked by large and uniform size of PE aggregates and hence increased the V-funnel time 20 compared to the reference mixture (Mix1). While preparing the mixtures, the water content 21 was adjusted to keep the same workability conditions in each SCC mixture. As seen in Table 3, 22 the ratio of water-to-cementitious material (w/cm) was 0.49 for all mixtures, except Mix3. The 23 w/cm ratio of Mix3 was 0.46 for similar workability characteristics. This could be explained by 1 the smooth surface texture and spherical shape of v-HDPE 68 (Fig. 3). 2 3 The type of PE aggregate had no significant influence on workability of SCC, so it can be 4 concluded that addition of virgin/recycled LDPE and HDPE has no significant level of negative 5 effect on the fresh properties of SCC when they were used as partial aggregate replacement 6 from workability points of view. 7 Hardened concrete properties 8 Mechanical properties 9 The compressive strength and flexural strength test results were given in Table 5. As it was 10 expected, PE aggregate substitution reduced the compressive strength, except Mix3, 11 especially at earlier ages (7 and 28 days), but the difference was compensated at 90 days. surface of v-HDPE seen from SEM image (Fig. 3) was the reason for a decrease in its w/cm 5 ratio while keeping the same workability measures as the remaining mixtures. 6 7 Moreover, the results of flexural strength tests at 7 and 28 days were presented in Table 5. All 8 SCC mixtures gained strength with age ( Fig. 4-b). Like in compressive strength, reduction in Durability properties 5 Permeability properties 6 The water absorption, sorptivity and chloride ion permeability tests performed at 28 days 7 were reported for all SCC mixtures ( Table 5). Water absorption values for all SCC mixtures 8 were in the range of 5.7-8.5% and indicated low water absorption characteristic (less than 9 10%) which is in agreement with other SCC studies containing mineral admixtures 71,72 . As seen 10 in Table 5, the water absorption capacity, namely permeable pore volume and the sorptivity 11 values of all SCC mixtures with PE were higher than the reference mixture, except Mix3. 12 Increase in total porosity with polymer addition was also reported by several studies 3,17,65,73,74 . 13 The poor chemical bonding between PE and the cement matrix may lead to formation of micro 14 cavities in the interfacial transition zone which in turn is responsible for increase in porosity. 15 Excess gas trapped in the blend 75 due to uniform and large size of PE aggregates and 16 hydrophobicity of PE could have also contributed to the porosity increase in mixtures 17 containing PE. Contradictory to the above explanations, Mix3 showed lower permeable pore 18 volume than the reference. The reason for having lower permeable pore volume in Mix3 was 19 probably related with its lowest w/cm ratio as 0.46 among the all mixtures. The significant 20 influence of w/cm ratio in SCC mixtures were also reported in the literature 76 . The reverse 21 relation between strength and total permeable pore volume was clearly seen in Fig. 5, which 22 also supports our previous statement. As well, for plastics with non/low-absorption 23 characteristics, high plastic contents are resulted in more free water which surrounds and 24 accumulates around the plastic aggregates and increases the voids and pores hence increases 25 the water absorption 15,67,77 . In Mix3, w/cm ratio was lowest and this may be the reason of its 1 lowest water absorption property among others. 2 As far as the sorptivity index was concerned, the reference mixture had the lowest sorptivity 3 with a sorptivity index of 131x10 -4 mm/min 1/2 (5.16 x10 -4 in./min 1/2 ). As seen from Table 5,  4 incorporation of PE, increased the sorptivity of the SCC mixtures (Mix2-5) about three times 5 higher than the reference. This indicated that incorporation of PE as aggregate replacement 6 has significant effect on the water sorptivity of SCC. The mechanism of increasing sorptivity 7 can be explained by larger, permeable and connected pores in the PE substituted mixtures. 8 However, further research should be conducted to establish the microstructure and the 9 porosity structure of SCC mixtures incorporating PE substitutes. 10 11 The total charge passed from each SCC mixtures during RCPT test was presented in Table 5. 12 Total charge passed was below 1000 Coulombs for all SCC mixtures, therefore all SCC mixtures 13 were rated as "very low" according to limits suggested by ASTM C1012 61 . The main factors 14 determining the resistance of SCC to chloride ion ingress are reported as binder type, binder 15 content and admixtures 78-84 . In the current study, as seen in Table 3 Considering all of these we can conclude that, regardless of the type of PE, there was no 20 significant influence of partial PE aggregate substitution on the chloride permeability of SCC. 21 Sulfate resistance 22 The results of the length measurement of sulfate exposed bars at specified periods were 23 illustrated in Fig. 6 conditions. Further research is needed to establish the influence of external sulfate exposure 6 on the microstructure and pore networking of PE incorporated SCC. 7 8 The interfacial transition zone (ITZ) of the manufactured SCC mixtures and the cement paste 9

Microstructural properties
in the ITZ were investigated by SEM-EDX and selected SEM images are presented in Fig.7

-I and 10
Fig. 7-II, respectively. As seen from the figure, the bonding between natural aggregate (N.A) 11 and the cement matrix ( Fig. 7(a-I)) was stronger than the bonding between PE aggregate and 12 the cement matrix. The voids between PE aggregate and the cement matrix can be clearly 13 seen in Fig. 7(b,d,e-I). The composition of calcium silicate hydrate (C-S-H) and calcium 14 hydroxide (CH) could not be determined by SEM-EDX because of intermixing with other 15 phases (i.e. the elemental ratio of Ca, Si, Al were not found to differentiate the C-S-H and CH). 16 Therefore, the morphology was used to distinguish C-S-H and CH. The ITZ between natural 17 aggregate and the cement matrix was characterised by the presence of large and dense C-S-H 18 and CH crystals (Fig. 7(a-II)). While smaller CH crystals intermixed with C-S-H, unhydrated 19 cement and MSWI FA was observed in ITZ of PE incorparated mixtures. (Fig. 7(b-e,II)). Among 20 the SEM of PE incorparated SCC mixtures, C-S-H gel was only clearly seen in Mix3 (Fig. 7(c-II)). 21 This dense and compact structure of C-S-H gel observed in Mix3 agreed with its low water 22 absorption capacity and its higher strength, and hence confirmed our previous statemet and 23 highlighted the influence of w/cm ratio on the strength of SCC. Moreover, some hydation 24 products were observed on the surface of the PE aggregates (Fig.7-I). This indicated that there 1 is a chemical interaction between PE aggregate and the cement matrix.