Recent development and perspective of lightweight aggregates based self-compacting concrete

Abstract The utilization of natural and artificial lightweight aggregates in lightweight self-compacting concrete (LWSCC) is gaining popularity in research field. Extensive research has been carried out in the past decade all over the world to utilize lightweight aggregates (LWA) in self-compacting concrete (SCC). LWSCC, which uses renewable aggregates, has great potential to become an alternative material to conventional concrete. The paper is aimed to review the more recent research of physical properties of lightweight aggregates used in developing mix design of lightweight self-compacting concrete. In design, the mix proportion of LWSCC is a crucial factor to achieve the desired fresh and hardened concrete properties. The methods to develop LWSCC mix design with anticipated fresh and hardened concrete are reviewed. Research shows that the mix design LWSCC is preferably proportioned by aggregates packing concept. In addition, discussion on the fresh and hardened concrete properties is made and summarized in this paper. Studies indicate that there is a promising future for the use of lightweight aggregates in SCC as it shows satisfactory filling ability, passing ability, segregation resistance and compressive strength. Research gaps recommendations are then identified through this review to further discover lightweight self-compacting concrete in several aspects, particularly in term of sustainability.


Concrete sustainability problem
Concrete is a very common construction material which has been widely used throughout the world due to its versatility, availability and economy (Rodriguez de Sensale et al., 2015).
According to Samson et al. (2016), concrete is considered as the most heavily consumed construction materials in the world due to its low cost. The more recent statistics shows that there is more than 26.8 billion tonnes of normal concrete being produced globally per year (Senaratne et al., 2016). This huge production has caused the construction sector to face the issue of gradual exhaustion of natural resources as well as the difficulty in accessing them. In the aspect of the environmental impact of the concrete production, study shows that it can be reduced through the use of alternative materials (Mehta, 2001). The more common practice is the partial replacement of cement and aggregates alternatives.
Also for the reason of the high demand of concrete in construction industry, a large amount of normal weight aggregates (NWA) is consumed which has resulted in gradual depletion of natural gravel and crushed rock. The situation warrants the urgency to intensify the research and development of more sustainable construction materials. As such, great opportunity exists to incorporate construction and demolition wastes into concrete mix as aggregates in order to improve its resource productivity (Mehta, 2001). Research has been carried out for utilizing recycled aggregate from demolition waste (Duan & Poon, 2014;Etxeberria et al., 2007;Xiao et al., 2005). The recycle process involves stock piling, crushing, presizing, sorting, screening and contaminant elimination. However, processing of recycled aggregates requires large amount of energy and cause higher carbon dioxide emission. Alternatively, other materials such as lightweight aggregates (LWA), either arising naturally or being generated as by-product from industrial processing, can be used to replace NWA in the concrete production. This leads to the production of lightweight concrete (LWC). LWC is commonly produced by replacing the normal weight aggregates with LWA. Extensive research has been carried out to utilize the waste generated as alternative construction materials in concrete due to growing of sustainability consciousness. (Alengaram et al., 2013;Aslam et al., 2016)

Lightweight self-compacting concrete
With the advancement of concrete technology, several attempts have been made in developing new high performance materials that possess the benefits and characteristics of SCC and LWC in the past decades. An innovative concrete, lightweight self-compacting concrete (LWSCC), which possesses the properties of both LWC and SCC has been developed. LWSCC is produced by the replacement of NWA with LWA in SCC. According to ACI 213 (2014), the density of structural lightweight concrete must falls within the range of 1120 kg/m 3 to 1920 kg/m 3 .
Aggregates contribute to the most of the weight of concrete and commonly constitute about 60% by volume of SCC (Topçu & Uygunoğlu, 2010). As such, due to the porous structure of LWA, it is able to reduce the density as well as the thermal conductivity of concrete. The use of LWSCC brings about several benefits such as reduced self-weight, shorter construction period, lower construction cost and elimination of noise emitted from vibration machines as well as better heat and sound insulation due to the voids in LWA (Grabois et al., 2016;Papanicolaou & Kaffetzakis, 2010;Vakhshouri & Nejadi, 2016). Since the present construction industry is experiencing the shortage of skilled workers as well as the difficulty in hiring new generation of skilled workers (Kim et al., 2010), LWSCC which is less labour intensive, can be a timely solution to these shortcomings. In addition, LWSCC, which is very suitable for manufacturing precast units, can be used to promote mechanisation or even automation processes in construction industry. The assembly of precast building components units on site has made the construction methods more straightforward.

Application of lightweight self-compacting concrete
LWSCC has been employed as alternative construction materials in structural construction such as cable stayed bridge construction since 1992 in Japan (Ohno et al., 1993). Dymond (2007) had designed and constructed a 20m pre-stressed beam by using LWSCC while Lahkega and Stenah (2011) studied the possibility of utilizing LWSCC in full scale wall. Also, Shi and Yang (2005) had utilized LWSCC in the application of thin precast C-shaped wall. Hubertova and Hela (2007) made use of LWSCC in the construction of stadium walkway structural elements.
Lately, the use of LWSCC has become popular in construction and research field.

Type of lightweight aggregates
Lightweight aggregates can generally be categorized into natural and artificial types. The common natural LWA are pumice, diatomite, volcanic cinders, scoria and tuff (ACI-213, 2003;Neville, 2008). As for the artificial LWA, it can be further categorized into industrial wastes and processed natural materials (Aslam et al., 2016). Sintered slate, sintered pulverized fuel ash, expanded or foamed blast furnace slag and colliery wastes are more common industrial wastes used as LWAs. In addition, there are also processed natural materials such as shale, expanded clay, slate, vermiculite and perlite which can be used as LWA in manufacturing concrete (Mahmud, 2010). Numerous researches have been concentrated on utilizing artificial LWA in developing LWSCC.

Problems in lightweight self-compacting concrete
There are several common issues in developing mix design of LWSCC. As LWA is porous materials and generally irregular in shape, its workability is poor and compressive strength is relatively low when compared to gravels. As such, a large amount of cement paste is required for LWSCC to achieve desired workability and targeted compressive strength. Due to the porous structure of LWA, it has high water absorption capacity which tends to absorb the water during batching, resulting in poor workability. The high water absorption of LWA makes it difficult to estimate the required water volume for batching. The common practice to overcome this issue is to allow LWA to achieve saturated surface dry (SSD) condition before batching (Domagała, 2015). However, care must be taken since different type of LWA has different water absorption rate. Excessive water can increase the risk of bleeding and segregation (Illidge, 2010;Juradin et al., 2012). Moreover, the densities of lightweight aggregates are generally lower than those of the mortar matrix and natural aggregates in concrete (Topçu & Uygunoğlu, 2010). Therefore, the difference in density between LWA and normal weight sand can alter the fresh properties of LWSCC mixture. The resulting poor self-compaction and segregation of aggregates can severely affect the durability and structural performance of concrete in hardened state (Juradin et al., 2012;Kwasny et al., 2012). Thus, the use of LWA in SCC is still regarded as new development in concrete technology and further investigation and study are required. In addition, no code of practice or guideline has been published for developing mix design of LWSCC.

Objective
As LWSCC brings about advantages in many aspects, research to understand the complicated nature of LWSCC is gaining popularity. Therefore, the main objective of this paper is to review the lightweight aggregates (LWA) that have been used in developing lightweight selfcompacting concrete. Identification of the physical properties as well as comparisons of LWA are conducted. In addition, the effect of using LWA in SCC mixture on fresh and hardened concrete properties will be discussed. The methodology to develop LWSCC mix design is reviewed too. In summary, the LWSCC properties and mix design can be improved significantly upon the review of the currently available literature.

Lightweight aggregates
Extensive research has been carried out by many researchers in utilizing lightweight aggregates (LWA) in SCC. Hwang and Hung (2005) utilized reservoir fine sediment as coarse aggregates in SCC while Bogas et al. (2012) and Hubertová and Hela (2013)  Tayfun Uygunoğlu & Topçu, 2009). Also, Shi and Wu (2005) and Lo et al. (2007) have utilized expanded shale as LWA for SCC. Moreover, Kanadasan and Razak (2014) used agriculture waste, palm oil clinker, as aggregates in SCC. The physical properties of the selected lightweight aggregates including pumice, expanded shale and expanded clay will be discussed in the following part of the paper. The fresh and hardened state properties of LWSCC are highly depends on the physical properties of LWA used. In this connection, specific gravity, size distribution, shape thickness and texture, bulk density and water absorption characteristic of lightweight aggregates will be elaborated.

Specific gravity
Specific gravity is defined as the ratio of the material mass to the mass of an equal volume of water at the temperature of 23°C. Based on the research done by several researchers, all the three types of lightweight aggregate (LWA) have different values of specific gravity which are not more than specific gravity of normal weight aggregates of 2.4-2.9. The specific gravity values of all these three types of aggregate falls within the range of 0.42-2.25 as shown in Table   1.
The specific gravity for pumice aggregates is within the range of 0.  Gopi et al. (2015) found the lowest specific gravity of 0.42 of expended clay aggregates while Shanker (2016) found the highest of 1.75. This inconsistency of specific gravity may be due to the situation whereby the aggregates are supplied from different sources as well as the different ways they are processed in the industry. By comparing the LWA and NWA (shown in Table 1), the specific gravity of LWA is 10% to 80% less than that of NWA. Aggregate specific gravity is important in the calculation of weight-to-volume relationships and to compute various volumerelated quantities such as voids presented in aggregate, and that the voids that must be filled by cementitious materials. It affects the resulting workability and final density of designed LWSCC.

Size Distribution of LWA
Lightweight aggregates (LWA) generally occur in different particle shape and size. Sieve analysis or gradation test is a common method for determining the particle size distribution.
The particle size distribution of LWA is crucial in engineering application as it can be used to verify the compliance of design requirement, production control and specifications. Typical particle size distribution curves of pumice, expanded shale and expanded clay are shown in Figure 1 (Lotfy et al., 2016;Topçu & Uygunoğlu, 2010). It is noted that pumice aggregates possess better particle distribution curve than expanded shale and expanded clay aggregates.
The use of well graded aggregates in SCC will minimize the voids which leads to optimum workability and strength. As such, selection of appropriate size distribution of aggregates is important in designing LWSCC mix design.

Shape thickness and texture
According to Tviksta (2000), the performance of SCC is very sensitive to the characteristics of aggregates. These characteristics include shape, texture, maximum size, grading and morphology. The shape and size of coarse aggregates have significant influence on the particle packing and aggregate interlocking within the matrix. They are factors in determining the amount of paste volume to cover all particles. LWAs commonly exist in angular and flaky shape. Khaleel et al. (2011) had studied the effect of maximum aggregate size on flowability of SCC.
The authors found that the flowability of SCC decreased with the increase of coarse aggregate size. The authors also recommended the use of coarse aggregates with maximum 10mm size as it can produce higher strength SCC than that produced by using coarse aggregates of maximum 20mm size. From the review of LWA of several researchers as summarized in Table 1, most of the coarse LWA maximum size used in LWSCC is either 12.5mm or 16mm. This is to promote a good interlocking effect between them to enhance the packing characteristics and flowability of SCC which will guarantee the strength of concrete (Kanadasan & Razak, 2014).

Bulk density
Bulk density of aggregates measures the volume of their solid aggregate particles as well as the voids between them that they occupy in the concrete. The bulk density is used in the volume method of concrete mix proportioning. Many researchers did not provide the compacted bulk density of the LWA aggregates used. As shown in Table 1, the loose bulk density of LWA from different sources generally shows variation. The bulk density of expanded clay, expanded shale and pumice aggregates is in the range of 300-1280 kg/m 3 , 750-1500 kg/m 3 and 330-1010 kg/m 3 respectively. Ahmad et al. (2007) stated that aggregates with density within the range 700-1400 kg/m 3 are preferable for structural application. By comparison, the bulk densities of all these three LWA are 10-80% lesser than normal weight aggregates. The lightweight characteristic of LWA is generally due to its porous characteristics.
a. b. c.

Water absorption
LWA are generally porous materials which tend to absorb water. LWA will absorb and hold more moisture than normal weight aggregates. As a result, pre-wetting of LWA is required before batching and this practice has been used in manufacturing lightweight concrete (LWC).
Depending on the cellular structure of LWA, it may also take longer time to achieve saturated surface dry (SSD) condition (Peters, 1999). The 24-hour water absorption of these three aggregates is in the range of 5-80%. By comparing these three LWA, pumice is found to have the highest water absorption capacity. LWSCC is sensitive to the water content of LWA as it can alter the resulting workability and compressive strength of concrete. The water/binder ratio of concrete can also be affected by the water absorption of LWA (Liu et al., 2011). The water absorption capacity of LWA must be specified in order to maintain the consistency of LWSCC.
According to Shafigh et al. (2012), concrete with porous aggregates is less sensitive to poor curing as the strength may vary only 6-11%. This is due to the fact that the water present in aggregate pores is capable of providing internal curing. The sensitivity can be reduced when lower water/binder ratio is used. The water present in aggregates is able to reduce plastic shrinkage due to unfavourable drying condition and provide internal curing which allows for more complete hydration of cement (Pierce, 2007).

Remark
The mix proportion of LWSCC and its corresponding performance in terms of both fresh and hardened state are greatly dependent on the physical properties of LWA incorporated.
Concerning the characteristics of LWA such as specific gravity, size gradation, shape, texture, and water absorption capacity, they can significantly alter the amount of material used in mix design. Specific gravity of LWA used can affect the resulting concrete density. From the review above, it is noted that the specific gravity of LWA of less than 2.0 is used to produce lightweight concrete in order to produce concrete of density below 1920kg/m 3 . Aggregate size, gradation and texture can greatly influence the amount of cement paste used to lubricate aggregates in order to achieve self-compacting ability as well as to fill in the voids between aggregates. Since LWA is generally present as angular and flaky shape, most of the researchers have limited the maximum coarse aggregates size up to 12.5 or 16mm. This can reduce the surface-to-volume ratio in order to minimize the cement used to achieve better workability and hence lower cost.
Moreover, the water absorption of LWA can greatly affect both fresh and hardened properties.
High water absorption LWA can cause workability loss when it is used as dry condition during batching. Saturated LWA can greatly alter the water to cementitious material ratio used which will result in poor compressive strength of concrete. In the light of considerable influence of water absorption of LWA, LWA must be pre-wetted and allowed to achieve saturated surface dry (SDD) condition in order to prevent either water loss or high water content before batching.

Mix design of LWSCC
The mix proportions of LWSCC are crucial in its application as the selected proportions can affect the required properties in fresh and hardened states. Similar to SCC, LWSCC must attain the desired fresh properties such as filling ability, passing ability and segregation resistance so as to fulfil the self-compacting requirement. Filling ability, which is also known as flow ability, is the capability of concrete to flow and fill the formwork completely under its own weight.
Meanwhile, passing ability refers to the capability to flow past the confined spaces between steel reinforcement congested area without segregating and clogging within the space of formworks. Segregation resistance is the capability to stay homogeneous during the process of transporting, placing and after placing without tendency to bleed and separation of aggregates from mortar. Similar to any other type of concrete, strength, volume stability and durability of the hardened LWSCC are important in structural applications (Sethy et al., 2016). The performance of LWSCC is greatly influenced by the constituent of raw materials, the dosage of chemical and mineral admixtures, types of aggregate used, packing density, water to cement ratio (W/C) and design procedures.
At the present moment, standardized method for obtaining mix design of SCC does not exist.
Many researchers have developed and proposed several design methods for SCC based on scientific theories and empirical expressions. In the context of SCC, the design methods can be classified into five categories based on their design principles, which are empirical design method, compressive strength method, close aggregate packing method, statistical factorial method and rheology of paste model (Shi et al., 2015). However, there is limited mix design method has been developed for LWSCC. The majority of the available LWSCC mix design methods in literatures are mainly based on close aggregate packing method. Many researchers prefer to develop the mix design of LWSCC by trial and error method as most of the proposed methods are not suitable to be used once the requirement of application is changed. This is commonly done by varying the binder content, binder/water ratio, admixture dosage, fine and coarse aggregate ratio. The review of LWSCC mix design method will be presented in the following section.

Shi and Wu Method
The combination of the least void volume for binary aggregate mixture, excess paste theory and ACI 211 has been adopted by Shi and Wu (2005) in proportioning the mix design of LWSCC.
The relationship between void volume or density of combined aggregates and coarse to fine aggregates volume ratio is determined by using particle packing concept in accordance with ASTM C29/ C29M. The least void volume of combined aggregates was found to be 0.5 in their study. However, the authors recommended to use coarse to fine aggregates ratio of 0.6 as it does not increase much void but decrease the density significantly. Excess paste theory is then used to determine the minimum quantity of paste required to fill in the void among the aggregates and also to allow SCC to flow with minimum frictions between aggregates as well as to balance the mixture by the quantity of water retained by the aggregates as illustrated in Figure 3. The required volume of excess paste is highly dependent on the characteristics of LWA, such as gradation, shape and surface texture, which can be determined through laboratory tests. The cement content and water to cement ratio are then determined from ACI 211 based on the designed compressive strength. The cement content is fixed from the chosen value while excess paste is produced from powders including fly ash and glass powder. The workability is then adjusted by varying the SP dosage. The authors successfully design LWSCC with satisfactory flowability and segregation resistance by using the proposed method. However, the proposed method requires intensive laboratory work to obtain the necessary information to proportion mix design.

Hwang and Hung method
For DMDA, Hwang and Hung (2005) developed this method based on ACI 318 and the fact that high physical density can produce optimum physical properties. In DMDA method, the mixture proportion algorithm is classified into aggregate and paste phase. Aggregate phase comprises lightweight aggregate, normal weight fine aggregate and fly ash while cement, slag, water and superplasticizer constitute paste phase. Finer particles fill the voids of the coarse aggregates to minimize the porosity in order to form the major skeleton of aggregates phase as shown in Figure 4. This in turn increases the density of solid materials and reduces the content of cement paste as illustrated in Figure 5. Paste phase is mainly used for lubricating aggregates in order to achieve concrete workability. This method is suitable for mix proportion design aimed to reduce water and cement content by using the physical packing density of aggregate which results in lower permeability of LWSCC. Though, this method does not take into account the optimum weight of concrete as long as the optimum properties are obtained. This may result in high density concrete. The authors recommended to use high water to binder (w/b) ratio of more than 0.42 to prevent autogenous shrinkage of the cement paste due to cement hydration and pozzolanic reaction. In fact, it is not necessary to use high w/b ratio when LWA are pre-soaked and achieved saturated surface dry condition (SSD) before casting. The water from internal pores is able to prevent the autogenous shrinkage. Moreover, in this method, the aggregates packing density can be enhanced by adding fly ash which fill the voids in LWA.
Fly ash should not be considered as the part of aggregate phase as fly ash is supplementary cementitious materials.   Kaffetzakis and Papanicolaou (2012) proposed another LWSCC mix design method based on optimum packing point (OPP) concept and workability criteria. This method involves the investigation of paste, mortar and concrete phase of material. Cement paste and mortar are assessed through wet packing method, which is used to determine the packing density of cement paste and mortar. This concept involves the determination of total voids and air voids as well as the solid concentration factor of a given water to cementitious materials volumetric ratio ( / ). High / ratio is used as trial initially. The ratio is then decreased until solid concentration factor is about to decrease. Void ratio versus / curve will be plotted based on the trials as shown in Figure 6 . Optimum packing and void ratio can be determined from the curve. The derived mortars from OPP concept must be assessed for self-compactness through slump-flow and V-funnel test. This method assumes that the least void volume of mixture corresponds to the optimum flowability in both paste and mortar. For concrete phase, the aggregate packing index is first determined from aggregate apparent and particle density.

Kaffetzekis and Papanicolaou
LWSCC is then proportioned by modifying the mortar to aggregates void volumetric ratio based on the equation derived by Jacobsen and Arntsen (2008). The workability must be assessed using SCC fresh concrete test. The authors argue that maximizing packing density should be solely used to determine the mix proportion of LWSCC, which contradicts with the method proposed by Hwang and Hung (2005).

Kanadasan and Razak method
Kanadasan and Razak (2014) modified the particle packing method of SCC which was originally proposed by Choi et al. (2006) to allow for the substitution of palm oil clinker (POC) aggregates in SCC. The substitution can be made on either fine or coarse aggregates at the level of 0% to 100%. It is based on the concept of minimizing the void of concrete by using appropriate size and gradation of aggregate with the use of minimum volume of paste as shown in Figure 8. The authors introduced an additional correction lubrication factor (LCF) to particle packing factor (PP) to allow for the characteristics of LWA aggregates when aggregates substitution is made in LWSCC mix design. The authors highlighted that the voids produced by flaky and porous structure of POC aggregates could be filled and lubricated by the binder paste. The proposed method fixed the fine aggregates ratio at 0.5 and 0.6 to allow wider range of ratios for SCC. The authors studied the cement content varied from 380 to 420kg/m 3 and recommended that 420kg/m 3 could produce the optimum performance SCC. However, the authors also mentioned that trial has to be carried out to ensure the required performance. The authors also demonstrated experimentally that the proposed method is able to produce LWSCC when 100% substitution of LWA is incorporated. PP theory is able to produce LWSCC mix design with minimum void volumes relative to the coarse aggregate, water to binder ratio, maximum cementitious materials density as well as the optimum fresh concrete properties. This theory provides good understanding of the consumption of aggregate and paste volume for a given unit volume of concrete. The proposed method is also applicable for a variety combination of other aggregates. However, the PP factor and CLF have to be determined in laboratory if other types of aggregates and their combinations are used. Besides, the actual performances of the designed mix must be checked in laboratory.

Remark
Although Mazaheripour et al. (2011) recommended to apply high performance concrete mix design method for LWSCC to avoid segregation and maintain the strength, the method cannot produce optimum LWSCC mix proportion in terms of fresh and hardened properties. The resulting density is not within the upper limit of lightweight concrete in accordance to ASTM.
It is clear from the research reviewed above that most of the proposed methodologies for proportioning LWSCC mix design are based on close aggregate packing principle. Aggregates packing principle is used to determine the least void among the aggregates in order to minimize the void produced by LWA as well as to determine the optimum coarse to fine aggregates ratio in order to produce the lowest density LWSCC. From the literatures above, it is noticed that the coarse to fine aggregates ratio used is generally in the range of 0.5 to 0.6 and the ratio of 0.6 is recommended by most of the researchers as it is the most cost efficient. The paste is then applied to fill the voids to become LWSCC which can be determined through either excess paste theory or rheological study of cement paste or mortar. However, intensive laboratory work is required in obtaining the necessary information. Most of the proposed methodologies is not able to proportion the LWSCC mix design based on required performance such as specified workability and compressive strength criteria. Furthermore, durability requirements such as shrinkage, creep, physical durability and chemical durability are not considered in proposed mix design methodology by researchers. As such, further statistical data analysis is required in order to simplify and produce performance based LWSCC mix design methodology.

LWSCC workability criteria
As previously stated, LWSCC must be assessed for filling ability, passing ability and segregation resistance and they are used to measure the workability of LWSCC. There are several methods for assessing each of these properties. Several publications such as EFNARC (2002) and ACI-237 (2007) provide the guidelines to carry out workability test for SCC. The methods to assess the filling ability are slump flow, T 500 , Kajama box, v-funnel, o-funnel and orimet. Assessing the filling ability is the most fundamental test for any type of SCC as it can be used to assess the consistency of SCC to meet the guideline requirements. The test for assessing passing ability are L-box, U-box, J-ring and Kajama box. These tests adopt the concepts of allowing SCC to pass through a pre-set spacing. This spacing is the smallest gap whereby SCC can flow continuously to fill the formwork. Also, segregation resistance can be assessed through penetration, sieve segregation, settlement column and visual segregation. SCC is mostly prone to segregation during and after placing. Segregation is a crucial problem in the casting of vertically tall structural element as it can lead to the uneven distribution of aggregates and mortar in LWSCC. The workability performance requirements of EFNARC (2002) for SCC are shown in Table 2. According to EFNARC (2002), these criteria are developed based on the current knowledge and research. SCC with fresh properties outside these criteria may be acceptable if it is able to perform properly under the required conditions. Future developments will likely produce different requirements for these criteria. For example, these criteria may be relaxed if the formwork design is very simple or the spacing between the reinforcement is large. packing density and less void between the aggregates particle, allowing the excess paste in LWSCC to achieve better flowability and segregation resistance. The excess paste required for improving workability highly depends on the gradation, shape and surface texture of aggregates.
They agreed the research outcome of Shi and Wu (2005). In short, the workability of LWSCC is highly dependent on the aggregates packing density and void volume. were able to be used for casting the "U"-shape thin wall panel. The aggregates and fibers were found to be homogenously distributed along the panel length. The findings in their study provided a new understanding that LWSCC is able to fill the narrow formwork even with the flow time outside the SCC workability requirement as stated in Table 2.
On the other hand, Mohammadi et al. (2015) examined the effect of silica fume with 0% to 15% of binder replacement on the properties of LWSCC workability with expanded clay and perlite as aggregates. The flowability and segregation resistance of LWSCC were found to be improved with the replacement as well as the increased dosage of silica fume. They also concluded that LWSCC with expanded clay as aggregates achieved better workability compared to LWSCC with perlite as aggregates. Corinaldesi and Moriconi (2015) studied the effect of the addition of synthetic fibers in LWSCC with expanded clay as aggregates and recycled concrete aggregate as partial replacement. It was noticed that the incorporation of fibers is able to improve the filling ability while it had negative effect on the passing ability. Silica fume was also studied. They observed that addition of small amount of silica fume can result in higher viscosity. Poor flow ability and passing ability were observed but the segregation resistance was improved. Similar observation was obtained with addition of silica fume in LWSCC with synthetic fibers. However, the findings of Corinaldesi and Moriconi (2015) had contradicted with the findings obtained by Mohammadi et al. (2015).
A comprehensive study of LWSCC was done by Floyd et al. (2015) on the effect of cementitious material and aggregate type on the workability of LWSCC. Two types of aggregates, which are expanded clay and expanded shale, were studied by them. They found that better visual stability of LWSCC was achieved by increasing the cement content. For common finding similar with other researchers, the increase in superplasticizer dosage could result in improved filling and flowing ability. With the constant amount of SP dosage and w/b ratio, the increase of volumetric sand to total aggregate ratio was found to be able to produce better fresh properties with optimum ratio of 0.51. Also, no significant improvement in fresh properties was noted by incorporation of silica fume with 5% and 10% in LWSCC with lower cement content. For LWSCC with high cement content, the fresh properties tend to be improved with only 5% or 10% incorporation. Poorer fresh properties were achieved by LWSCC with Type I cement compared to Type III cement. The fresh properties of LWSCC with Type III cement can be improved by partially replacing binder with fly ash as shown in their study . Floyd et al. (2015) stated that LWSCC with expanded shale exhibited better fresh properties compared to expanded clay with the same amount of other mixture content which agreed with the findings of Lotfy et al. (2015a). Also, the authors changed the coarse aggregates distribution in their study by limiting the maximum aggregate size to 12.5mm. This resulted in better fresh properties. In short, the fresh properties of LWSCC are highly dependent on binder content, SP dosage, type of aggregates used and volumetric sand to total aggregate ratio. Kurt et al. (2015) investigated the effect of fly ash, different water to binder ratio and replacement of pumice aggregates with natural aggregates on LWSCC. The filling ability was found to be improved with the increasing of water to binder ratio as well as fly ash replacement.
Due to the low pozzolanic activities of fly ash, its increase could retard the bonding of water to mixture and hence the loss of workability. However, segregation was observed in their research when water to binder ratio exceeded the optimum value. Also, the spreading capability of slump flow was found to be increased with the density increase of LWSCC as the spread and placement properties of LWSCC were highly dependent on its own weight. With the increase of pumice aggregates in LWSCC, the time required to spread 50cm diameter also increased as well as the V-funnel flow time. This can be explained by the loss of weight with the replacement of LWA in LWSCC resulted in self-weight to be less than threshold stress. Since the self-weight was below the threshold stress, the authors implied that it could increase the tendency of static segregation. Bozkurta and Taşkin (2017) studied the effect of the use of barite, fly ash and pumice as powder on the LWSCC fresh properties. The authors observed that LWSCC with barite powders are the best among three types of powder in improving the fresh properties in terms of flowability and filling ability. However, the authors reported that the use of barite as power content in LWSCC could cause bleeding due to its poor adhesiveness and viscosity resistance. As such, the ratio of low adhesive powder content is crucial in developing LWSCC to prevent bleeding. Ardalan et al. (2017) investigated the effect of fly ash, pumice and slag as binder partial replacement in LWSCC on retention workability after 50minutes. The authors stated that pumice blend require more superplasticizer dosage to achieve target slump flow among the three types of supplementary cementitious materials. Conversely, fly ash blend requires lesser dosage of SP in order to achieve target slump flow. It was explained that the spherical geometry of fly ash particles is able to reduce the fraction resistance of cement particles and enhancement of the mixture fluidity. Among the three types of blend mixture, fly ash blended LWSCC showed significant slump flow loss after 50minutes while pumice blended LWSCC showed the best retention capacity.

Remark
The studies presented thus far provide evidence that the workability of LWSCC is highly dependent on the aggregates packing density and void volume. In general, similar to normal SCC, the performance of LWSCC workability with respect to filling ability, passing ability and segregation resistance is greatly influenced by water to binder ratio, superplasticizer dosage and total binder content. The inclusion of different types of supplementary materials has different effects on LWSCC workability. When silica fume is used, and with increasing replacement level, the segregation resistance of LWSCC is found to be improved while it has negative effect on filling and passing ability. The inclusion of fly ash as binary or ternary blend can not only improve all the three fresh properties but also reduce the amount of SP required. In addition, the incorporation of fibers such as steel and synthetic fibers is able to improve the filling ability but it causes negative effect on passing ability.

Compressive strength
The most important required property of any innovative material is its compressive strength.
The compressive strength of concrete has great influence on its structural performance. As increased with the decreasing of w/b ratio. The 28-day compressive strength also increased with the increase of total binder content. The amount of superplasticizer dosage was found to have no effect on the LWSCC strength. These findings conformed to the basic knowledge of concrete property. Grabois et al. (2016) observed that their LWSCC mix design were able to achieve 70% of the 28-day strength in a day. Their mix design is suitable for high early strength applications. Also, the incorporation of steel fibers in LWSCC could result in lower compressive strength. For failure mode, they noticed that the rupture was occurred through the LWA and yet the interfacial transition zone was still intact. The authors explained that the mortar was stronger than LWA in lightweight concrete which was in conformity with the findings of Lotfy et al. (2015a). The use of expanded clay aggregates could result in better paste-porous LWA bonding. Mohammadi et al. (2015) studied the effect of silica fume on LWSCC containing perlite and expanded clay as LWA. They observed that the LWSCC containing expanded clay as LWA achieved higher compressive strength compared to perlite as LWA. However, the compressive strength differences decreased when the increase of silica fume replacement. The replacement of silica fume in LWSCC would increase LWSCC compressive strength. Nevertheless, Mohammadi et al. (2015) only studied the silica fume replacement up to 20% of total binder.
The result is yet to be known if the silica fume replacement is more than 20%. The optimum replacement percentage is also not known. Kurt et al. (2015) conducted a series of experimental test to investigate the effect of fly ash, different water to binder ratio and replacement of pumice aggregates with natural aggregates on LWSCC. With the increasing percentage of pumice aggregates replacement, the compressive strength of LWSCC decreased significantly. This concurred with the findings of Floyd et al. (2015) and Grabois et al. (2016) that the LWA are generally weaker than mortar even though the LWA used by both authors are different. Also, Kurt et al. (2015) found the compressive strength decreased with higher water content which is generally true. LWSCC with fly ash replacement gain strength at the slower rate than that those without fly ash replacement at the early stage (e.g. 7 days). Nevertheless, they achieved almost similar strength at later age (e.g. 90 days). The authors attributed the findings to low pozzolanic activity of fly ash at the early stage when its content increased. The replacement of fly ash in LWSCC could significantly improve the fresh concrete properties but require longer time to gain strength.
A comprehensive study was done by Floyd et al. (2015) to investigate the effect of cementitious The authors reported that the use of BTS in LWSCC could result in weak bond between the binder paste and the aggregates, thereby creating a weak interfacial transition zone and hence reduction in compressive strength. Perlite based LWSCC showed most significant strength loss when the LWA content was increased. The authors explained the excess pore water in the perlite was released due to crushing during mixing.

Flexural strength
Flexural strength is one of the parameters measuring the tensile strength of concrete. No significant improvement on the flexural strength of LWSCC with the addition of synthetic fibers was noticed in the works of Corinaldesi and Moriconi (2015). In the research done by

Tensile strength
Concrete is generally weak under tension action. The tensile strength of concrete is commonly used to estimate the load that will cause the development of cracking in the member under flexural loading. Once the concrete cracks, the concrete behaviour will be affected (Malárics & Müller, 2010 The study conducted by Grabois et al. (2016), the tensile strength of LWSCC was determined under direct tensile loading. Tensile strength of LWSCC was found to be improved for about 30% with the addition of steel fibers. They stated that addition of small amount of steel fibers in LWSCC could improve the tensile strength up to the first crack under direct tensile loading.
Nevertheless, more study concerning the tensile strength of LWSCC is essential for it to fully replace conventional concrete in any structure.

Modulus of elasticity
The modulus of elasticity (E -value) is defined as the ratio between normal stress to strain below the proportional limit of a material. It is used to measure instantaneous elastic deformation which represents the stiffness of materials. According to Neville (2008), the E-value of concrete decreased with the use of LWA. The stiffness of LWA is generally very weak which is proven by a few researchers (Floyd et al., 2015;Grabois et al., 2016;Lotfy et al., 2015a

Remark
In contrary to normal SCC, the compressive strength of LWSCC is mainly governed by the homogeneity of the batched concrete. The uniformity and homogeneity of LWSCC are governed by the mixing time and procedure. As highlighted by Li et al. (2017), mixing time should not be longer than 3 minutes in order to avoid segregation. Longer mixing time can cause LWA to segregate and float at the top part of specimen. Consequently, the hardened specimen has unbalanced aggregates distribution with more aggregates at top part and more cement mortar at the bottom part which can result in poor compressive strength. Well distribution of aggregates throughout the matrix of concrete can maximize its compressive strength. It can be said that the strength variability of LWSCC can be related to its aggregates distribution and hence is the function of segregation resistance.
Since the mortar of LWSCC is normally stronger than LWA, the compressive strength of LWSCC is also dependent on the strength and proportion of LWA. The compressive strength of LWSCC is sensitive to changes in mix component properties and their proportions such as water to binder ratio, binder content and the incorporation of supplementary cementitious materials. These factors must be considered properly in mix design in order to achieve anticipated workability in fresh state and compressive strength in hardened state. The optimum implementation of supplementary materials such as fly ash, slag and silica fume can improve compressive strength. In addition, the incorporation of fibres such as steel, synthetic and macro fibers will increase compressive strength of LWSCC.

Prospective and Future challenges
The recent and present research works provide framework for further investigation and study for utilization of lightweight aggregates in self-compacting concrete. Future research should concentrate on the investigations of the followings.
1. The current methods for developing mix design of lightweight self-compacting concrete are complicated and require the validation through trial laboratory work. Further research is required to develop easy and simple guidelines for developing mix design of lightweight self-compacting concrete. One recommendation is to carry out statistical analysis of the relationship between mix design and performance in terms of fresh and hardened properties.
2. Most of the current research done on LWSCC is restricted to a few types of lightweight aggregates only. Furthermore, there is limited research on the use of other types of lightweight aggregates such as sintered slate, sintered pulverized fuel ash, oil palm shell, colliery waste, etc in LWSCC as they have been used as LWA in lightweight concrete (LWC). Effort must be made to identify more variety of suitable aggregates.
3. There is limited research on the long term durability behaviour of LWSCC such as shrinkage, creep, corrosion and bond strength. Moreover, these properties are not considered in the mix design methodology. Further research is required in this area.
4. More study is recommended to understand the tensile strength, flexural strength, elastic modulus, shear characteristic, and pre-stressing application of LWSCC.
5. The use of LWSCC technology at the present moment is restricted to research only. It will be a big challenge to further develop and refine this technology to be adopted and widely used in construction industry.

Conclusions
The application of lightweight aggregates in lightweight self-compacting concrete is reviewed based on recent literatures and the outcome is reported in this paper. The physical properties of LWA, mix design methodology, fresh and hardened properties of LWSCC were discussed.
From the literature review, the following conclusion can be made.
1. Different LWA exhibit different specific gravity, size gradation, shape characteristic, bulk density and water absorbability which lead to different performance of LWSCC. LWA of specific gravity less than 2.0 is commonly utilized to manufacture LWSCC. The maximum LWA sizes are limited to the range of 12.5 to 16mm. The shape of LWA commonly varies from spherical to flaky. Different LWA exhibit different water absorption which varies from 5%-80%. In this regards, saturated surface dry condition has to be achieved to reduce water sensitivity.
2. The workability of LWSCC depends on the aggregates packing density and void volume.
Water to binder ratio, superplasticizer dosage and total binder content have great bearing on the performance of LWSCC workability. The inclusion of different types of supplementary materials has different effects on LWSCC workability. The use of silica fume as well as with its increasing replacement level, improve the segregation resistance of LWSCC but have negative effect on filling and passing ability. The inclusion of fly ash as binary or ternary blend will not only improve all the three fresh properties but also reduce the amount of SP required.
3. The compressive strength of LWSCC is highly dependent on the strength of LWA as they are weaker than the mortar. Factors such as water to binder ratio, binder content and the incorporation of supplementary cementitious materials will affect the compressive strength of LWSCC significantly and they must be considered properly in mix design. Optimum inclusion of supplementary materials such as fly ash, slag and silica fume will improve the compressive strength. Also, fibres such as steel, synthetic and macro fibers will increase the compressive strength of LWSCC 4. This review enhances the understanding of LWSCC mix design methodology with close aggregate packing method being most commonly practiced. Close aggregate packing method establishes the relationship between paste and aggregates. Some researchers have introduced statistical analysis to simplify and improve the design procedures. Close aggregate packing principle provides clear insight into the understanding of consumption of aggregate and paste volume for a given unit volume of concrete.