1. Introduction
Lightweight aggregates applied to concretes, according to the definition included in PN-EN 206+A1:2016-12 standard, have the density of <2000 kg/m
3 in an oven-dry condition, or loose bulk density of <1200 kg/m
3 [
1]. They may be divided into natural lightweight aggregates which include volcanic tuff, scoria, or perlite, and artificial (processed) lightweight aggregates that include expanded clay, expanded perlite, expanded shale, lightweight aggregate from processed fly ash, or vermiculite [
2].
Perlite is a natural volcanic rock, and approximately 97% of it is extracted in six countries (Turkey, Greece, Japan, Italy, USA, and Hungary) [
3]. Perlite is an amorphous rock with a high content of silica in its composition [
4]. When natural perlite is heated at 900–1200 °C, a lightweight porous material is created, so-called expanded perlite (EP) [
3]. During the heating process, the base material increases its volume as much as 20 times, which leads to the production of a very lightweight porous material with outstanding thermal insulation properties and a low thermal conductivity value. Moreover, EP shows good sound-proofing properties, good fire resistance, and a high resistance to chemicals [
3].
Porosity of lightweight aggregate is associated with its density, which is the main parameter that determines its suitability for use in structural concretes [
5]. Expanded perlite is a very porous material. According to authors [
6], the total porosity of perlite may reach up to 26 vol.%, which explains the low density of this material. Porosity of the material is also connected with its water absorption, which in case of perlite is significant and may amount to as much as 35% [
7]. Due to the significant water absorption of lightweight aggregate, designing and production of lightweight concrete with those aggregate is complicated [
8,
9].
Expanded perlite may be used in building materials as an additive to cement (in a form of powder) or as a lightweight aggregate in thermal insulating materials. Due to its high silica and aluminate content, EP may also be the main component of geopolymer materials [
3,
10].
Generally, it can be stated that the application of EP to the concrete mix instead of natural aggregates leads to the reduction of the hardened concrete strength as the perlite content in the concrete mix increases [
11,
12,
13,
14]. The above is explained by the introduction of the porous aggregate with a low strength to the concrete mix. Reduction of the concrete strength was also found in case of ground EP applied as the additive to cement in the concrete mix [
15,
16,
17]. It was explained by a significant impact of ground expanded perlite on the loss of concrete mix fluidity. Therefore, despite using superplasticizers, at low water to binder ratios (<0.35) the hardened concrete showed increased porosity and therefore lower density and strength [
18]. As per test results, application of the ground EP as an additive to cement may also increase the strength of the hardened concretes [
19,
20]. This may be explained by the fact that EP ground to the particle size characteristic of cement is able to to activate the pozzolanic properties of expanded perlite. Consequently, the content of C-S-H phase formed in the reaction of portlandite with EP increases, and thus, the porosity of the interfacial transition zone (ITZ) between aggregate and cement matrix in hardened concretes is reduced [
21].
Many authors conducted research on lightweight ultra hight performance concerts (LUHPC) [
22,
23,
24]. It was demonstrated that it is possible to achieve LUHPC with compressive strength ranging from 35 to 70 MPa, with a density ranging from 1440 to 1840 kg/m
3, with the addition of pozzolan materials such as silica fume, fly ash, metakaolin, volcanic ashes, calcined clays, and shales [
22]. During the determination of the usefulness of the lightweight concrete as a structural material, it is very important to define its flexural and tensile strength. While testing the LUHPC, the authors of the paper [
22] proved that lightweight concrete with lower density demonstrated higher flexural strength than concrete with the same compressive strength but higher density. The authors of the paper [
25] demonstrated the possibility of increasing flexural strength by incorporating steel microfibers into concrete mix, achieving concrete with a density of approximately 1700 kg/m
3 [
25]. Moreover, they stated that the introduction of fibers in amounts up to 0.75 vol.% does not cause significant workability loss in the concrete mix. With the aim of achieving lightweight concrete with lower density, containing lightweight aggregates while also demonstrating good mechanical properties, the authors of papers [
26,
27] introduced basalt fibers in lightweight concretes. However, the addition of those fibers to the concrete mix decreased its workability significantly.
It is necessary to point out that numerical methods, applied to estimate the behavior of structure elements under loads, are very helpful to analyze the tensile strength of concrete or to predict areas and intensity of crack formations [
28]. However, the application of these numerical methods to modeling the lightweight concrete, composed as they are with materials of very different elastic modulus, is more complicated compared to normal concrete or high performance concrete [
29].
The literature contains very few studies dedicated to the implementation of reactive powder concrete (RPC) in lightweight concretes technology. Sadrekarimi [
30] obtained the lightweight RPC with density ca. 1900 kg/m
3 and the height compressive strength (ca. 280 MPa), by simply increasing the content of silica fumes in the RPC composition, without the use of lightweight aggregate, while applying the increased temperature (240 °C) during 24 h of concrete curing. The author also showed that the concretes cured in temperatures of 240 °C are characterized by lower density compared to the concretes cured in temperatures of 90 °C [
30].
Gökçe et al. [
31] demonstrated that it was possible to obtain lightweight RPC with density of 1840 kg/m
3 and the compressive strength of ca. 69 MPa when quartz was replaced with a lightweight volcanic aggregate (2400 kg/m
3). Achieving higher compressive strength was possible with the application of lightweight aggregates, with the parallel application of steel micro fibers and pressure curing. However, such a treatment increased the RPC density to about 2400 kg/m
3 with an increase of the compressive strength up to 176 MPa and a decrease of water absorption from 6.5% (for concrete cured in ambient conditions) down to 1.7% at the same time.
Studies [
32,
33] applied lightweight aggregates such as Pollytag and expanded clay [
32] or expanded polystyrene beads-EPS [
32,
33]. The results showed that it was possible to obtain lightweight RPC without the application of thermal or pressure treatment, with the compressive strength in the range of ca. 17 MPa to ca. 83 MPa after 28 days, depending on the type and quantity of lightweight aggregate used. It was found that the addition of expanded clay to RPC had no impact on the increase of water absorption of this lightweight RPC, compared to the reference RPC without lightweight aggregate [
32]. However, water absorption of RPC with Pollytag and expanded polystyrene beads-EPS is higher by about 1.5% compared to RPC without lightweight aggregates [
32].
The research into lightweight reactive powder concrete with perlite was conducted by Al-Jumaily [
34]. The author showed that with the increase in perlite content in the RPC mix, in quantities from 2.5% to 10% by cement mass, the compressive strength and elastic modulus decrease. After 28 days, the compressive strength, with 2.5% of perlite in the RPC, equals 82.5 MPa, and for higher content of perlite (10%) the compressive strength equals 53.9 MPa, whereas the elastic modulus is 43.1 GPa and 31.0 GPa, respectively. Contrary to the findings of the authors of this paper, the author of [
34] demonstrated that increasing the content of perlite in the RPC mix up to 10% causes the water absorption rate to increase by approximately 2%.
In this study, the influence of expanded perlite in quantities of 30%, 45%, and 60% by volume to the RPC mix, on the lightweight RPC properties was tested. Particular attention was paid to changes in density, compressive strength, and water absorption with the increase in quantity of EP in RPC. Microstructure analyses were also conducted, as well as investigation into the influence of pozzolanic activity of EP on the compressive strength of lightweight RPC.
2. Materials and Methods
For testing, the RPC mix was prepared with the blast furnace slag cement-CEM III/A 42.5 N, (specific surface area of cement acc. to Blaine was 360 m
2/kg), with GGBFS content in amounts of ca. 60% by mass. Finely ground quartz sand, with maximum grain size < 800 μm, quartz powder, and waste silica, with the content of SiO
2 > 90% by mass, were used. Theater to binder ratio equaled 0.2 and was obtained by the use of the polycarboxylate superplasticizer in the amount of 3.0% of the cement mass. The research of the chemical and physical properties (such as chemical composition and particle size distribution) of the RPC mix composition are presented in this paper [
32]. The optimalization of the RPC mix components, in order to increase the particles packing, was carried out based on a Funk and Dinger curve [
35], as well as on the experience of Zdeb and Śliwiński [
36].
RPC mixes with perlite added in the amounts of 30%, 45%, and 60% by volume of the concrete mix were prepared to be tested. The expanded perlite of 0/4 mm fraction was used (
Figure 1). The particle size distribution of the lightweight aggregate used is given in
Table 1.
Expanded perlite (0/4 fraction) primarily contains grain sizes >2 mm (52%). The aggregate used contained ca. 4% of powder with the particle size <0.063 mm.
Increasing the quantity of expanded perlite (EP) in the RPC mix caused significant deterioration of its fluidity and required an increase in the amount of a superplasticizer from 3.0% to 5.0% by mass, at 60% of the EP content. The composition of tested mixes is presented in
Table 2. The procedure of mixing the ingredients was maintained for all mixes as follows: to begin with, the dry ingredients were mixed (cement, waste silica, quartz powder, and quartz sand). Then water was added, followed by the superplasticizer. When the homogeneous and fluid RPC mix was obtained, the lightweight aggregate was added.
The tests of the physical properties of EP (the loose bulk density, the water absorption, and the resistance to crushing), were performed in accordance with PN-EN 13055:2016-07 standard—“Lightweight aggregates”.
The consistence of the lightweight RPC mixes was tested by means of the flow table method, in accordance with PN-EN 1015-3:2000—“Methods of test for mortar for masonry—Part 3: Determination of consistence of fresh mor-tar (by flow table)”, by measuring the concrete flow diameter after 10, 30, and 60 min.
The tests of concrete density were conducted on the samples sized 150 × 150 × 150 mm3 according to PN-EN 12390-7:2011—“Testing hardened concrete—Part 7: The density of hardened concrete”.
The tests of water absorption were carried out for 28 days after the samples’ preparation. The sample size was 100 × 100 × 100 mm3. The samples were saturated with water and then dried to constant mass at the temperature of 50 °C. The test procedure is described in Polish standard PN-B 06250:1988 “Ordinary concrete”.
To test strength, the prism samples were prepared sized 40 × 40 × 160 mm3. After mix preparation, and casting them in steel molds, the samples were demolded after 24 h and cured in water at a temperature of 20 ± 2 °C until the strength test (i.e., for 365 days). The procedure of flexural and compressive strength testing is described in PN-EN 196-1:2016-07—“Methods of testing cement—Part 1: The determination of strength”.
The freeze-thaw resistance of lightweight RPC in the presence of de-icing salt was tested by means of Slab test (PKN-CEN/TS 12390-9:2017-07—“Testing hardened concrete—Part 9: Freeze-thaw resistance with de-icing salts. Scaling”), the classification of the concrete under its freeze-thaw resistance was based on a Swedish standard SS 13 72 44 “Concrete testing; Hardened concrete; Scaling at freezing—2005”.
The microstructure of lightweight RPC was carried out with the use of the NOVA NANO SEM 200 scanning microscope (FEI Europe B.V., Eindhoven, The Netherlands) with the possibility of obtainment the EDS spectra in selected points.
The test of the phase composition of EP and hardened lightweight RPC were conducted by means of powder X-ray diffraction within the angle range of 5° to 60° 2θ. The Philips X’PertSystem diffractometer (Amsterdam, The Netherlands) and CuKα radiation was used. The JCPDS-ICDD database was used to define the crystalline minerals based on known patterns [
37].
4. Discussion
It is well known that expanded perlite may be used, in powder form, as a mineral additive to cement, where the advantages of its pozzolanic properties may be utilized [
21]. However, EP which is not ground is used as a lightweight aggregate, and it was in this form utilized to RPC in this paper.
The results of consistency tests of RPC mix with EP showed that increasing the content of aggregate in an RPC mix causes significant workability loss (
Table 4). After 10 min, the flow diameter of RPC containing 60 vol.% of EP is almost two times shorter than the flow diameter of RPC without lightweight aggregate. The slight decrease in flow diameter over a period of 1 h suggests that the workability of the mix remains at approximately the same level. Similar phenomena are observed in RPC mixes with a lower content of EP (30 and 45 vol.%). The amount of EP in an RPC mix not only has an impact on the significant workability loss, but also the fact that expanded perlite is characterized by a low resistance to crushing (0.66 MPa) and, during mixing, all ingredients of an RPC mix are ground. The elements of ground EP were observed on microscopic images (SEM) of hardened lightweight RPC (
Figure 11).
A longer period of mixing time of an RPC mix which shows high viscosity [
32], which is required to achieve a homogeneous and fluid concrete mix, causes the grinding down of the expanded perlite. Therefore, the increase of EP in RPC above 30 vol.%, does not cause a significant decrease in density, and compressive strength remains at approximately the same level (
Figure 7,
Figure 8 and
Figure 17). Such an occurrence is not observed in RPC with lightweight aggregates, which represent higher resistance to crushing (Pollytag and expanded clay aggregate), as well as with expanded polystyrene beads (EPS) [
32]. In the case of the abovementioned lightweight aggerates, the increase in aggregates in RPC from 30 vol % to 60 vol.% leads to a decrease in the density of hardened lightweight RPC from ca. 6% to 15%. This observed phenomenon of the grounding down of expanded perlite in RPC mix allows us to state that utilization of higher amounts of EP (45 and 60 vol.%) in RPC is not recommended, bearing in mind the obtainment of lightweight concrete.
It is notable that conducted analysis by the authors of this paper of the research presented in a previous paper [
34], concerning the density of lightweight RPC including perlite with the increase in its participation in RPC mix, allows us to state that, also in this case, a significant decrease in lightweight RPC density with increase in perlite in the mix, in range from 2.5 to 10 mas.%, is not observed. However, the author of the paper [
34] does not comment on this phenomenon.
The fact that ground EP shows pozzolanic properties [
21] prompted the authors of this paper to conduct the research concerning the pozzolanic properties of this material in RPC mix. The compressive strength tests, which were canaried out on RPC containing 30 vol.% of EP with addition of 1.0% by cement mass of hydrated lime, did not show an evident strength increase in the material (the compressive strength increased just a few tenths of an MPa). The above emphasizes the lack of significant impact the ground EP pozzolanic properties have on the lightweight RPC compressive strength. However, phase composition tests of hardened lightweight RPC containing 30 vol.% of EP, with the addition of 1.0% hydrated lime and cured for 28 days (
Figure 17), showed the disappearance of characteristic reflections for portlandite and muscovite, the identified phase in EP (
Figure 2), proving that the pozzolanic reaction between EP and calcium ions occurs.
Yao et al. [
52] demonstrated that muscovite is a mineral able to react with calcium hydroxide, forming the C-S-H gel with potassium cations built in. These authors also showed that grinding the perlite had a significant impact on the pozzolanic activity of component minerals, including the pozzolanic activity of muscovite. Taking the above statements and the results of our own study into consideration, a slight increase in the compressive strength of the lightweight RPC with 30% EP and the addition of 1% hydrated lime, as compared to the compressive strength of the concrete without this additive, may be associated with the pozzolanic reaction of the silicate ions originating from the perlite phases with the calcium ions (Ca
2+) additionally introduced to the RPC mix.
The pozzolanic properties of the expanded perlite are revealed to a greater extent when the ground perlite is added to the cement [
21]. Its addition clearly reduces the quantity of portlandite in hydrated cement pastes. However, the pozzolanic activity of expanded perlite is slightly lower than the activity of fly ash (
Table 6). Authors of the paper [
53] used expanded perlite with other pozzolanic additives, such as fly ash, and demonstrated that in these configurations the perlite showed almost inert pozzolanic properties. The compressive strength of the mortars where 10% of fly ash was replaced with perlite, after 28 days, reduced by half, as compared to mortars with fly ash.
RPC with EP in quantity from 30% to 60% by volume are characterized by very low water absorption, ranging from 3.3% to 3.5%, which is not significantly different than the water absorption of RPC without a lightweight aggregate (3.3%). The above was explained by the very compact RPC microstructure, without visible larger pores (
Figure 9 and
Figure 11). The C-S-H phase precisely adheres to quartz and EP grains, also penetrating lightweight aggregate’s pores (
Figure 10 and
Figure 12). The consequence of such a compacted lightweight RPC microstructure is presented in the paper as an excellent resistance to freezing and thawing cycles in the presence of de-icing salt. RPC with EP, according to classification presented in Swedish standard SS 13 72 44, may be considered as concrete with very good freeze-thaw resistance.
5. Conclusions
Introduction of the expanded perlite-fraction 0/4 mm-to the RPC mix in quantities from 30% to 60% by volume causes a reduction in mix fluidity and homogeneity, which increases as a larger content of aggregate is applied in the mix. Maintaining the mix fluidity while using the larger amount of EP requires a larger quantity of a superplasticizer (to 5% by mass) and a longer time to mix ingredients.
EP’s low resistance to crushing and required longer mixing time of the mix components causes grounding down the expanded perlite, which therefore has no influence on the decrease in the lightweight RPC density.
The addition of EP to RPC mix in the quantity of 30 vol.%, causes a decrease in the concrete density from 2200 kg/m3 to 1900 kg/m3. However, a further increase in EP content in RPC mix, from 30 to 60 vol.%, results in a lack of decrease in density and compressive strength of those lightweight concretes, which should have been expected with a significant increase in EP in the RPC. This is the result of the weak aggregate getting finer due to its very low resistance to crushing. The above test results indicate that application of larger amounts of EP (45 and 60 vol.%) is not recommended in order to obtain lightweight RPC.
The water absorption of lightweight RPC with EP (30, 45, and 60 vol.%) is in the range of 3.3% to 3.5%, and is comparable to the water absorption of RPC without lightweight aggregate (3.3%). The low water absorption of lightweight concrete was explained by very compacted microstructures of hardened lightweight RPC.
Expanded perlite in the form of aggregate shows pozzolanic properties. Its pozzolanic activity, however, is much lower than the activity of silica fume and quartz powder, the ingredients of the RPC mix. It has been demonstrated that EP in the RPC mix reacts with calcium hydroxide, but the pozzolanic activity it shows has a marginal impact on the increase of the lightweight RPC strength.