Analysis of shrinkage and creep behaviors in polymer-coated lightweight concretes

1 Introduction

Pumice, which is an industrial raw material, is a porous, natural, light type of rock formed by volcanic activities. Pumice aggregates that are used in the production of lightweight concretes reduce the dead load on the structure, provide partial thermal and acoustic insulation, and increase fire resistance when compared with the concretes produced with limestone aggregates. However, the fact that the aggregates have a high water absorption capacity owing to their porous structure is a disadvantage in terms of lightweight concrete production [1]. The half-open porous structure of pumice aggregates mostly causes the increase of compressive strength by improving the adherence between the binder and the aggregate. However, because of the half-open surface structure of pumice, the polymer concrete (PC) needs more binding material than a PC with a normal aggregate [2]. Some researches have stated that the different mechanical features found in polyester-based PCs are related to the type of thermosetting resin or mineral aggregates, and to their concentrations in composite [3]. Therefore, aggregates were coated with polymers.

Polymers form a chemical bond, a chain, as a result of a chemical reaction (polymerization) of many organic molecules (monomers) [4]. The most important properties of polymers include their highly variable characteristics and easy adhesion to various materials. Such composite materials are known as polymer concretes and boast several advantages such as higher strength and a shorter curing process [5]. They can be used in concrete production owing to their properties such as applicability for all concrete classes, high resistance against the abrasive property of acids and bases, resistance to atmospheric conditions, higher abrasion resistance when compared with granite stones, and damping of vibrations owing to their resin content [6]. However, the fact that they also contribute to the creep and shrinkage behavior of concretes is important.

Creep deformation refers to the temporal deformation of the concrete under a constant load of concrete. Shrinkage deformation refers to the temporal deformation of concrete even in the absence of a load. The “unit deformation-time” curves obtained in both cases are quite similar. Significant levels of permanent deformation occur both in creep and shrinkage [7]. The gel structure of cement paste and non-free water loss in paste pores has a significant impact on creep. It can cause cracks with deformations reaching two to three times of the first deformation and visual deformities, and can cause the collapse of the building elements under high stress [8].

In another study on the mechanical properties of lightweight concrete modified with polymer [by adding styrene-butadiene rubber (SBR) as a polymer into two types of Brazilian lightweight aggregates], it was found that the 7-day compressive strength values differed between 39.7 and 51.9 MPa, and, as a conclusion, that thin prefabricated component material could be produced by an SBR modification of Brazilian lightweight aggregates [9]. Chen and Liu [10] used SBR latex as a polymeric admixture in expanded polystyrene (EPS) concrete, by wet-dry curing. They found that, first, wet curing had more advantages at the early curing periods for the development of the cement matrix, and the following process, dry curing with SBR additions, improved the adhesion between the cement matrix and EPS particles. They also indicated that compared with normal EPS concretes, the compressive strength of polymer-modified EPS concretes increased gradually even after 28 days.

In a study on the estimation of temporal deformations of concrete, Akperov and Akyüz [11] analyzed the creep, foundation creep, and shrinkage deformations of normal- and low-resistance concretes with and without additives. Furthermore, they examined the compatibility of the results they obtained from shrinkage and creep estimation models to the experimental shrinkage and creep values. They reported that excessive water amount and thinner fine aggregate particles increased the shrinkage deformation.

Nilsen and Aitcin [12] analyzed the shrinkage values of lightweight concretes made of expanded shale. They reported that after a 28-day curing, the shrinkage values of 56-day concretes were between 34 and 230 microns (1×10^-6), while the values of normal concretes were 203 microns (1×10^-6) in the same process. This was attributed to presence of water in the aggregate. Furthermore, in a study on fiber-reinforced lightweight concretes containing fly ash, Kayali et al. [13] found that the shrinkage values of 400-day samples were between 900 and 1100 microns (1×10^-6).

The aim of this study is to examine compressive strength, and creep and shrinkage properties of 500-dose lightweight concretes made of pumice aggregates coated with three different polymers. The 3-, 7-, and 28-day compressive strengths of the lightweight samples; 840-h and 12-month creep; and shrinkage measurements were determined.

2 Materials and methods

2.2 Polymer coating of aggregates

Polymer coating of aggregates was applied on 4–8 mm and 8–16 mm aggregate particle groups. Then, the aggregates were dried at 105±5°C for 24 h. The coating process was performed by spraying polymer on the aggregates using the conventional paint spray gun method. The polymers that could not be sprayed from a paint spray gun (size range, 0–4 mm) because of their thickness were thinned by using a cellulosic thinner. On the basis of preliminary work, the amount of the cellulosic thinner added to the polymers was determined to be 25% for Sonomeric1 (SNMC) [16], 20% for KB Pur 214 (KBP) [17], and 10% for Polipol3455 (PLP) [18]. Three layers of coating were performed to obtain sufficient coating thickness on the surface of the aggregate (Figure 1). Marble powder was spread on the sticking aggregates during polymer coating. The coated aggregates were dried at 23±2°C temperature for an average period of 96 h [19].

Figure 1

Aggregates coated with polymers.

The physical properties (specific weight, loose bulk density, and water absorption rates) of uncoated and polymer-coated aggregates were tested, and the results are shown in Table 3.

Table 3

Results of physical testing of uncoated and coated aggregates.

Mixtures	Specific weight (g/cm3)			Loose bulk density (kg/m3)			Water absorption rates (%)
	0–4 (mm)	4–8 (mm)	8–16 (mm)	0–4 (mm)	4–8 (mm)	8–16 (mm)	0–4 (mm)	4–8 (mm)	8–16 (mm)
Control	1.57	1.03	0.98	225	272	221	48.2	43.1	34.0
SNMC	1.57	1.25	1.20	225	282	230	48.2	10.2	7.6
KBP	1.57	1.37	1.28	225	315	290	48.2	2.1	4.8
PLP	1.57	1.51	1.45	225	330	245	48.2	8.5	8.1

3 Results and discussion

3.1 Compressive strength

A compressive strength test was conducted in accordance with the TS EN 12390-3, “Determining compressive strength in hardened concrete samples,” standard [20]. A total of 36 cubic specimens 100 mm in size were produced, and then these specimens were broken at the end of the 3-, 7-, and 28-day curing periods for each experiment series. A uniaxial concrete compressive test device with an adjustable loading speed (2 kgf/cm²/s), a digital control unit, and a capacity of 3000 kN were used in the test. The compressive strength test results are presented in Figure 2.

Figure 2

Test results of compressive strength (%).

The comparison of 3-, 7-, and 28-day compressive strength values of concrete samples with those of control values revealed that the compressive strength values of SNMC and PLP samples decreased by 11% and 40% in 3-day samples, 19% and 41% in 7-day samples, and 13% and 52% in 28-day samples. In contrast, the compressive strength values of KBP samples were found to increase by 10% in 3-day samples, 11% in 7-day samples, and 4% in 28-day samples.

3.2 Creep and shrinkage

The creep experiment was conducted in accordance with TS EN 3454 [21], while the shrinkage experiment was conducted in accordance with TS ISO 1920-8 [22]. The 500-dose 100×300 mm cylindrical concrete samples were prepared for the tests. The samples were removed from the molds 24 h after casting and were kept in standard water cure for 28 days. The samples were kept at 23±2°C temperature and 50±4% relative humidity until the end of creep and shrinkage test periods.

The length variations in creep and shrinkage experiments were measured using special pins (thickness, 3 mm; diameter, 10 mm) that are resistant to corrosion, with a hollow center for placing the tips of the measurement device. The pins were placed around the samples with equal distances on the basis of three different measurement directions. A 1997 model portable comparator produced by M.H. Mayes & Son Windsor Lim. Company, which can measure the unit length change in samples with a minimum sensitivity of 10^-6, was used in the creep and shrinkage experiments (Figure 3). A load of up to 40% of the compressive strength values of the samples was taken as a basis for determining creep behaviors (Table 5).

Figure 3

Length measuring device.

Table 5

Stresses applied to the test specimens (MPa).

Mixtures	Maximum strength	Creep strength	No. used in the spring
	MPa	(% 40) MPa
Control	16.17	6.50	1
SNMC	14.83	5.95	2
KBP	17.33	6.95	2
PLP	8.00	3.20	2

A loading framework setup was prepared in the experiments. The setup applies the load on the sample uniformly distributed and keeps the load unchanged in case of deformation of the sample. Three steel plates, each with a thickness of 10 mm and placed on four Ø24-mm bolts, and bolts were used in the specially produced setup. The fixed bottom plate in the experimental setup is important for the upper and lower free movement of the other plates. Leaf springs that are commonly used in train tampons were used between the bottom two plates in the framework (Figure 4).

Figure 4

Loading framework setup.

The creep behaviors of the concretes made of aggregates that were and were not coated with polymer were examined for 840 h and 12 months. The 840-h and 12-month creep change graphs are presented in Figures 5 and 6.

Figure 5

Change of the creep function (840 h).

Figure 6

Change of the creep function (12 months).

The ductility (elastic deformations) of the samples coated with SNMC and KBP were found to increase, while the ductility of PLP-coated samples was found to decrease when compared with the control samples. Analysis of creep functions (elastic deformations) of the samples in 12 months revealed that creep measures showed fluctuations rather than following a linear course. The highest creep value (18%) occurred in the SNMC and KBP samples. The PLP samples gave the lowest creep function values, with a decrease of 26%. This means that the KBP and SNMC samples experienced less deformation than the control samples, and they showed a parallel course.

Figure 7 presents the 840-h changes of shrinkage deformations. Figure 8 shows the 12-month shrinkage deformation graphs. Analysis of shrinkage deformations of samples revealed that length deformations showed fluctuations in polymer-coated samples, rather than following a linear course. Shrinkage deformations in concrete samples decreased by 41% in SNMC and BP samples and by 28% in PLP samples when compared with the control samples. This means that the highest shrinkage deformation was found in the control samples, while the lowest shrinkage deformation was found in the SNMC and KBP samples. This finding is consistent with the studies of Akperov and Akyüz [11] and Oymael [8].

Figure 7

Change of shrinkage (840 h).

Figure 8

Change of shrinkage (12 months).

4 Conclusions

In this study, the physical properties (specific weight, loose bulk density, and water absorption rates) of uncoated and polymer-coated pumice aggregates were examined. Then, the compressive strength, creep, and shrinkage changes of concretes due to these aggregates were investigated. The results are given below in order:

• The specific weights for coated and uncoated lightweight aggregates varied between 0.98 and 1.64 g/cm³. It was determined that the specific weight values of the aggregates with a mesh size of 4–8 mm and 8–16 mm increased with polymer coating compared with the control. The loose bulk density of aggregates changed in the range of 221–330 kg/m³. The average water absorption rates of uncoated lightweight aggregates (30–40%) was compared with those of the polymer-coated aggregates; their values were decreased (2–10%).

• The compressive strength values (at 3, 7, and 28 days) of concrete samples were compared with those of the control samples; while the compressive strength values of KBP samples increased up to 4–11%, the compressive strength values of SNMC and PLP samples decreased.

• Analysis of creep deformation showed that the ductility of SNMC and KBP samples increased by 57–100% at the end of 840 h and by 1–18% at the end of 12 months when compared with the control sample. In contrast, the ductility of PLP samples decreased by 3% at the end of 840 h and by 26% at the end of 12 months.

• Analysis of the shrinkage deformations of samples revealed that the 840-h shrinkage deformation values of the SNMC-, KBP-, and PLP-coated samples were lower than those of the control sample. However, the 12-month shrinkage deformations of the PLP-coated samples were observed to be higher than those of the control sample.

In conclusion, the difference between the deformation values of temporal creep under load and autogenous shrinkage values in the air yields real creep. Accordingly, it was found that the creep and shrinkage properties of concretes produced with SNMC- and KBP-coated pumice aggregates were positively affected. Furthermore, it was concluded that the compressive strength of lightweight concretes produced with coated samples further increased. Especially, KBP-coated lightweight concretes were found to give positive values in terms of compressive, shrinkage, and creep values when compared with the control sample.

Lightweight concretes that were produced here fall into a special class of concretes. Polymer-coated lightweight concretes that are recommended to reduce dead load and increase earthquake resistance of a building can be produced. The increase in ductility of a concrete, in addition to its compressive strength, means enhancing the earthquake resistance of the concrete.