Three-dimensional printing of concrete (3DPC) has a potential for the rapid industrialization of the housing sector, with benefits of reduced construction time due to no formwork requirement, ease of construction of complex geometries, potential high construction quality and reduced waste. Required materials adaption for 3DPC is within reach, as concrete materials technology has reached the point where performance-based specification is possible by specialists. This paper aims to present an overview of the current status of 3DPC for construction, including existing printing methods and material properties required for robustness of 3DPC structures or structural elements.
This paper has presented an overview of three categories of 3DPC systems, namely, gantry, robotic and crane systems. Material compositions as well as fresh and hardened properties of mixes currently used for 3DPC have been elaborated.
This paper presents an overview of the state of the art of 3DPC systems and materials. Research needs, including reinforcement in the form of bars or fibres in the 3D printable cement-based materials, are also addressed.
The critical analysis of the 3D concrete printing system and materials described in this review paper is original.
Paul, S., van Zijl, G., Tan, M. and Gibson, I. (2018), "A review of 3D concrete printing systems and materials properties: current status and future research prospects", Rapid Prototyping Journal, Vol. 24 No. 4, pp. 784-798. https://doi.org/10.1108/RPJ-09-2016-0154Download as .RIS
Emerald Publishing Limited
Copyright © 2018, Emerald Publishing Limited
The success of rapid industrialization in various parts of the world may be explained by the automation processes which led to a faster and cheaper way of production. However, the concrete construction sector has not been automated in the same manner and to the same extent as other industrial sectors. In the past two decades, the traditional way of mixing and casting concrete on site has to a significant degree been replaced by pre-cast or pre-fabrication construction in several developed and newly industrialized countries. However, the construction sector can benefit significantly from further automation towards reducing labour and construction time, improved quality and reduced environmental impact. Three-dimensional printing (3DP) was first introduced in 1987 as a means of rapid prototyping (Hull, 1986; Chua and Leong, 2014). Today, there are various 3DP techniques, but the basic principle remains that of additive manufacturing (AM), meaning that it adds material layer by layer, in contrast to the most traditional manufacturing methods that subtract material. In recent years, the AM technique has been used to build artificial organs, for instance, for knee/hip/other joint replacements, limbs (arms, legs, feet, hands, etc.) and enabled higher levels of customization in toys, jewellery and consumable products (Gardiner, 2011). It is postulated that AM and 3D printing will be at the centre of a shift in design and fabrication due to their unique capabilities for fabricating geometrically complex and highly customized products using a variety of materials. This dramatic change in the way objects are designed and fabricated presents significant opportunities for the building and construction (B&C) industry (Gardiner, 2011).
Typically, AM processes use computer-aided design (CAD) models as blue prints to fabricate complex 3D designs in a layer-by-layer manner. Early applications of 3DP were for modelling and prototype purposes, but due to developments in material science, it has since been used in many industrial applications such as the automobile, aerospace, military and medical sectors (Bak, 2003; Gibson et al., 2015). More recently, development of 3D printable construction material has evolved this technology to B&C applications. A number of 3D-printed construction projects have been executed to demonstrate the successful application of this form of automated construction (Khoshnevis, 2004; Khoshnevis et al., 2006). However, 3DP application in B&C is in its infancy, and to significantly grow its application, challenges of the scale of the objects for B&C applications; strict control through standards and governing bodies; and the level of conservatism in B&C industries world-wide are to be addressed.
Initial efforts to realize automated construction were made by Japanese companies in 1980s, which were mostly towards two categories of work (Taylor et al., 2003). In the first category, single task robots replaced simple labour activities on construction sites. Robots are used for concrete floor finishing, spray painting, tile inspection and materials handling. The second category consists of fully automated systems that can construct high-rise steel buildings or steel reinforced concrete (RC) buildings using prefabricated components as shown in Figure 1. An example of this approach is Big-Canopy, which was the world’s first automated construction system for building a precisely defined concrete structure. It has four independent masts supporting an overhead crane that delivers components at the control of a joystick. All tasks are scheduled and controlled by a centralized information control system (Khoshnevis et al., 2006; Wakisaka et al., 2000).
Concrete is used worldwide as one of the major construction materials, both in situ and prefabricated. This is due to several benefits it has as a construction material, including low production cost, mouldability into various shapes, high thermal resistance and relatively high durability. Typically, RC construction can be divided into three components: concrete, formwork and reinforcement. Formwork may account for 35-54 per cent of the total construction cost and consume 50-75 per cent of the total construction time (Jha, 2012). An example of estimated cost distribution in a typical construction is shown in Figure 2 (Jha, 2012). However, as a new construction method, the data on time and cost for construction by 3D printing of concrete (3DPC) are limited. The construction industry produces a significant amount of construction and demolition waste (C&D waste). According to Lawson et al. (2001), the construction industry in the UK produces 53.5 Mt of C&D waste annually, of which 51 per cent goes to landfill, 40 per cent for land reclamation and 9 per cent is crushed for re-use. It is anticipated that adoption of 3DPC technology would be able to significantly reduce construction waste, as the amount of concrete mixed and placed is carefully controlled in such automated system. As such, 3DPC structures could potentially significantly save on construction cost, improve productivity and could, above all, significantly limit the environmental impact by using less material and producing less waste.
WinSun built their first 3DPC structures shown in Figure 3 (WinSun, 2018). Over a period of about a decade, they made significant efforts to address matters of shape retention, open time, pumpability and extrudability of concrete material (Le et al., 2012a). These matters remain important research and development areas for 3DPC, considering that various classes of concrete materials are used in the B&C industry, and that strict material quality control is required, which may be challenging especially in on-site construction as opposed to prefabrication in controlled environments in factories.
This paper presents an overview of the state of the art of 3DPC systems and materials. Research needs, including reinforcement in the form of bars or fibres in the 3D printable concrete materials, are also addressed.
2. Current 3D printing of concrete systems
The main advantage of a 3DPC layered approach is that it can manufacture complex, non-standard geometries and details rapidly (Chua and Leong, 2014; Gardiner, 2011; Gibson et al., 2015). The 3DPC processes have developed in three major categories named by their inventors as contour crafting (CC), D-shape and concrete printing. Descriptions of these processes can be found in Khoshnevis (2004), Le et al. (2012a) and Dini (2015). Typically, these printing processes follow two approaches, namely, pumping and extrusion, and methods based on selective binding. In both approaches, an STL (standard triangulation language) file is collected from a 3D CAD model and then divided into multiple 2D layers in a so called “slicing” process. Cartesian coordinates from the 2D layers along with printing parameters such as speed of the printing head, extrusion rate and binder deposition rate are then delivered to a 3D printer in machine-readable language. The desired structure is then built layer by layer using the defined coordinates and the given printing rates (Wolfs, 2015; Nerella et al., 2016). This section describes 3D concrete printers, nozzles, pumps and control systems. To achieve better 3D-printed objects, there must be a good combination of these elements in the system.
2.1 3D concrete printers
At present, three categories of 3D concrete printers, namely, gantry (WinSun, 2018; Le et al., 2012a; Dini, 2015; Wolfs, 2015; Nerella et al., 2016; TU Eindhoven, 2018; BetAbram, 2018), robotic (Concrete Plans, 2018) and crane (Apis Cor, 2018) systems shown in Figure 4, are successfully used in universities and industry. Note that the gantry is also one type of crane, but its height is typically fixed as shown in Figure 4(a). However, the crane system shown in Figure 4(c) is adjustable in the vertical direction. The advantage of using gantry and crane printers is that they are easily scalable in size. On the other hand, robots typically have a fixed dimension and are difficult to scale up. However, the speed and degrees of freedom of a six-axis robot allow it to perform many tasks that may not be possible with a four-axis gantry printer. Moreover, if the design of printed objects does not require any complexity, a gantry printer may be preferred over a robotic printer, as the cost of robot is higher and the pay load on a robotic arm is generally lower than on a gantry. The size of gantry printers currently in use varies from small laboratory versions up to 40 × 10 × 6.6 m (length × width × height) to print large-scale building components (WinSun, 2018). Figure 4(a) shows a four-axis gantry system adopted by the Technical University of Eindhoven in The Netherlands, with an adjustable printing envelope of about 11 × 6 × 4 m (TU Eindhoven, 2018). With this technique (gantry printer), it is reported (WinSun, 2018) that WinSun has built multiple houses, a five-story apartment block and an 1100 m2 mansion. The printing method is similar to CC, i.e. an inner and outer bead is printed in layered fashion, as well as a cellular inner structure [Figure 4(c)]. In some cases, reinforcement is placed manually between layers during printing. The concrete material mixture WinSun used contains glass fibre, steel, cement, hardening agents and recycled construction waste materials (WinSun, 2018). Note that the multiple houses built by WinSun were not fully 3D printed. The structural elements were printed in segments in the factory and transported to the site, where they were assembled. Suitable concrete materials for 3DPC are discussed in more detail in Section 3.
Another company that has adopted the CC technique is CyBe additive industries, founded in The Netherlands. They print a mortar that reaches a bearable strength within 5 min. CyBe has experimented by attaching a print head to an imprecise six-axis robot-arm [Figure 4(b)], allowing for high diversity in printer speed and strategy (Concrete Plans, 2018). The additional rotational axis of a robot over gantry printers gives the designer more freedom to design complex shape. This robotic printer has a 6-m diameter reach, which enables printing of relatively large elements within a short period. A Russian company Apis Cor (2018) has developed a crane type printer [Figure 4(c)] that enables a large printing plan area of up to 58 m2, with virtually no height limitation.
The printer head speed plays an important role in the dimension of filament layers in 3DCP. For a fixed nozzle orifice and material flow, too slow movement of the printer head may deposit more materials in a region and increase the bead dimension, i.e. the bead dimension may be higher than the nozzle orifice. Similarly, too fast movement may cause an insufficient amount of material deposition and reduce the bead dimension. The selected printing speed is a function of the size and geometrical complexity of the element to be printed, linked to the pump speed and quality of the extruded concrete material. It also has to consider the so-called open time, i.e. how long the fresh concrete material can be formed before it sets, as well as the interlayer placement time gap which influences interfacial bond. Nerella et al. (2016) used a printing speed of about 75 mm/s for materials with 90 min open time. By injecting accelerating chemical additive into the nozzle, they significantly reduced the initial setting time of the concrete to 3 min, considering that usual concrete setting time may be several hours. In their case, the time gap between two layers was 30 s. In a study by Perrot et al. (2016), the time gap between the printing layers was in a range of 11 to 60 s, which allowed them to print a 3D column at a rate ranging from 1.1 to 6 m/h. The influence of the interlayer time gap on interlayer mechanical properties is described in more detail in Section 3.3. Extrusion rates of analogous processes in concrete construction range between 15 and 125 mm/s for typical hollow core slab extrusions and road pavement curb extrusion. It appears that quality of surface textures may limit printing speed to about 200 mm/s, based on extrusion of 15 × 70 mm plates (Visser, 2007). However, this may vary for different concrete materials used for 3DPC.
2.2 Nozzles for concrete extrusion
The end part of the printer head is the nozzle that forms the desired shape and size of the concrete layer (Bos et al., 2016). To achieve the desired shape and good buildability on bedding layers, it is necessary to have an appropriate nozzle for 3DPC. The orientation of the nozzle should remain tangent to the tool path (Bos et al., 2016) to prevent twisting and displacement of the fresh layer. Displacement of layers may create eccentricity, which may lead to instability and collapse. Various nozzle shapes have been used as orifice of the extruder, such as circular, ellipse, square and rectangular shapes. For better surface finishing, side trowels can also be used in the nozzle orifice as shown in Figure 5(a). In general, a circular nozzle gives freedom of printing without having to adjust the printing angle for shape continuity at vertices or changes in angle of the printed element. However, the smaller contact surface between layers or beads may influence buildability or stability of layers in quick printing construction. A study by Kwon (Kwon, 2002) concluded that the surface finish created with a square orifice is better compared to an ellipse orifice. Also, the ease of manufacture is reported to be greater with a square orifice than an elliptic one (Anell, 2015). The size of the orifice also depends on the size and shape of the objects. A range in size of 9 × 6 mm to 38 × 15 mm of rectangular orifices is currently being used for 3DPC (Nerella et al., 2016; Le et al., 2012b), but this should not be considered to be limiting. In case of circular nozzles, diameters of 4-22 mm are in use (Lim et al., 2011). The existing nozzles are designed in such a way that their orifice size and printing head speed deliver printing volumes in the range 0.04-0.09 L/s (Nerella et al., 2016; Malaeb et al., 2015). However, printing volume rates of analogous extrusion processes of concrete are up to 11 L/s (hollow core slabs), and a speed of roughly 2 L/s is considered to be a realistic upper target rate for 3DPC. Depending on the shape of printed objects, multiple nozzle orifices can also be used for printing as shown in Figure 5(b) (Contour crafting, 2018). This may also give solutions to printing various materials, for example foamed concrete as insulation between concrete outer layers.
2.3 Pumping and control system
Pumping of concrete materials is an essential part of the 3DPC system, given the importance of its fast delivery without segregation of particles. The pump transports the materials from the mixing unit to the nozzle, where it is subsequently extruded. The pump has to be able to transport the specified concrete mix, which may be a challenge due to large particle size aggregate and a range in water to cement (w/c) ratio. High viscosity gaining in fresh concrete is typically required for shape retention once printed, which demands relatively high pumping pressures in the range 1 to 4 MPa (10 to 40 Bar). Concrete material mix adjustments are required to avoid segregation under these pumping pressures, and formation of a thin smearing layer for good pumpability (Secrieru et al., 2016). Segregation during high-pressure pumping may lead to loss of the smearing layer and cause blockage of material in the pipe. For better printing, it is desirable to have a balance between the feeder system, nozzle and material properties. Depending on the size and shape of the printed objects, pumping speed may need to be adjusted. At angle changes in the printing path, the deposition of materials should be controlled to avoid higher deposition of material. The printer head (nozzle) can typically change direction reasonably quickly; nevertheless, geometrical perfection will require that the pump speed is reduced at especially sharp curves or discontinuities. Therefore, it is necessary to have a proper control system which will allow adjusting the printer speed, nozzle rotation and pumping of materials from one control unit. Alternatively, lower viscosity concrete can be used for improved pumpability, and chemical additives are added in the nozzle to accelerate setting once printed, as done by, for instance, Nerella et al. (2016).
3. 3D printable materials
Because the printing process requires a continuous, high degree of control of the material during printing, high performance building materials are preferred (Lim et al., 2011). Also, as no supporting formwork is used for 3DPC, traditional concrete cannot be directly used. For ensuring little or no deformation in the bedding layers, either low to zero slump concrete is required as shown in Figure 6. Or low viscosity concrete can be used for ease of pumping, requiring intervention by adding a chemical accelerator at the nozzle for quick setting once printed. Production of low slump concrete needs special care of granulometric properties of the fines. In this respect, the particle shape has a large impact on both compaction behaviour and green strength (Hüsken and Brouwers, 2012). Thixotropic behaviour is desired, whereby material of high static viscosity can undergo microstructural changes to become less viscous by de-flocculation when agitated or otherwise stressed, but rebuilds or re-flocculates to become highly viscous once it has been extruded and comes to rest (Mitsoulis, 2007; Barnes, 1997; Roussel, 2006; Khayat and Assaad, 2006; Khayat and Assaad, 2008). A typical response of thixotropic material subjected to rheological testing is shown in Figure 7. It can be seen that when material is at rest, it needs higher torque (static torque) to flow, but over time, the torque required for a particular angular speed in a rheological test is reduced, known as dynamic torque. In this regard, lessons may be learnt from experience with self-consolidating concretes (Khayat and Assaad, 2006; Khayat and Assaad, 2008). For which mix adjustments in terms of the aggregate grading, water to binder (w/b) ratio and chemical additives are successfully used to adjust the flocculation process, or microstructural thinning (de-flocculation) and rebuilding (re-flocculation).
3.1 Material composition
The material compositions for 3DPC by several authors are summarized in Table I. Most authors used several trial mixes before choosing the best mix composition. A significant amount of fine particles were used in all mixes and no reports of the use of coarse aggregates for 3DPC was found in the literature. Typically, in most cases the maximum sand particle size was limited to less than 2 mm (WinSun, 2018; Nerella et al., 2016; Perrot et al., 2016; Visser, 2007; Anell, 2015; Le et al., 2012b; Malaeb et al., 2015). However, coarse aggregate with particle size up to 10 mm have been used in 3DPC (Rushing et al., 2017). To control the open time and setting period of the materials, superplasticizer (SP), accelerator and retarder were also used in the mixes. Based on the current research reports, water to binder ratios in the range 0.23-0.41 and binder/sand ratios in the range 0.63-0.73 were used, which produced best printability and mechanical properties of the printed material once hardened.
3.2 Rheological properties of 3D printable materials
The fresh properties of shear stress, viscosity, open time and green strength of concrete material are very important, as they relate to pumpability, extrudability, buildability and interlayer adhesion. Open time of concrete relates to the change of flowability with time, which allows the material to be printed without detrimental effects on the quality of printing or hardened properties. Usually, a Vicat apparatus is used to measure the concrete open time, by typically establishing the initial setting time and (final) setting time. Open time is typically defined as the time between completion of mixing and initial setting time. On the other hand, shear stress and viscosity have direct influence on the pumpability and extrudability, usually measured with a rheometer. Shear stress of fresh concrete can also be measured indirectly from the slump flow test (Pierre et al., 2013; Roussel and Coussot, 2005). The green strength of 3DPC is related to its fracture on the surface when load is applied on top of the bed layer. In 3DPC, after printing the layer, the material must achieve sufficient strength to carry the load of subsequent layers. To measure the green strength the plate deposition method as shown in Figure 8 can be followed. In this case, a plate with a known weight can be placed on top of cylindrical sample at a certain time gap until failure can be seen on the surface (Perrot et al., 2016). The deformation of the bed layer can also be recorded in this test method.
3.2.1 Pumpability of concrete
Pumpability refers to the material’s mobility and stability under pressure while maintaining its initial properties (Jolin et al., 2009). At the pump, a relatively soft material that will be easy to pump is required; at the nozzle, a stiff material is wanted so it does not sag or lose shape. Concrete is a heterogeneous material and has particles of various sizes, shapes and densities. Therefore, to ensure good pumpability, carefully designed and optimized mixture design is required (Mechtcherine et al., 2014). Two methods of improving pumpability are including sufficient paste content so that there is enough grout to form a thin smearing boundary layer on particles, and suitable grout consistency and structure between aggregate grains to hinder forced or pressurized bleeding during high pressure pumping (Spiratos et al., 2003). It is necessary to obtain suitable consistency, combined with resistance against squeezing out of water from the concrete due to the high pressure in the pipe, so-called pressurized bleeding (Wallevik, 2002). It has been proposed to measure the pumpability of concrete with a sliding pipe rheometer as described in Mechtcherine et al. (2014).
3.2.2 Extrudability of concrete
In typical concrete extrusion, for instance, for manufacturing of pre-cast hollow-core slab elements or pavement curbs, a highly viscous, plastic-like mixture is forced through a die/nozzle, which is a rigid opening with a desired cross section. In this production process, the materials are formed under high shear and high compressive forces (Visser, 2007; Shao and Shah, 1997; Akkaya et al., 2000). In 3DPC, this is changed by the requirement of pumping the fresh material over appropriate distances depending on the printer (gantry, robot or crane) size and working volume (length, width and height). As described in the previous section, a thixotropic concrete material is desired to be pumpable. Through this agitation-reduced viscosity caused by the pumping pressure, the fresh concrete reaching the nozzle is less viscous than in traditional extrusion. Shape retention requires that the extrudate be sufficiently stiff to retain its geometrical form, by fast rebuilding through re-flocculation. At the nozzle, increased pressure may arise due to the nozzle size being smaller than the pipe cross-section. Here, segregation must also be avoided (Le et al., 2012a). To achieve the appropriate thixotropic and tribological (pumpability) characteristics, material ingredients and proportions should be chosen carefully and be properly controlled. Because of poor matrix composition, segregation may lead to material blocking in the pipe or at the nozzle.
3.2.3 Buildability of concrete
Buildability, which combines properties of printed layers being self-supportive and shape retention, refers to the ability of the bedding layers to hold the subsequent layers on top of them without collapsing or deforming. Layer imperfection and settlement (Kazemian et al., 2017) may lead to instability once successive layers are added. In traditional concrete construction, this is not an issue as the concrete is placed and supported in formwork. As 3DPC is formwork-free, there is a need for it to be self-supporting.
Apart from adjusting the concrete material properties, buildability of printed structures can be achieved in several ways. One simple way to improve the structural buildability is by changing the nozzle type. For instance, in case of a circular nozzle orifice, the contact area between the two beads is less compared to a rectangular or square orifice as shown in Figure 9(a).
Another way to improve buildability is by increasing the number of adjacent filament layers as shown in Figure 9(b). Note that the term filament in 3DPC refers to a printed layer of concrete which looks like a slender thread-like object. This method may not be suitable for thin 3DPC objects. For the same material composition and nozzle orifice, Le et al. (2012a) found significant improvement in the buildability by increasing the number of adjacent filament layers from one to six as shown in Figure. 9(b). Similarly, a supporting filament as shown in Figure 9(c) creating a cellular-type structure can also be used to improve the buildability. Note that these adjacent filament layers can be printed using both single and multiple nozzle heads. For printing two adjacent layers using a single nozzle head, the printing loop has to be closed, i.e. start and end of printing should be the same point. However, for multiple nozzle head, these two layers can be printed simultaneously, i.e. the printing loop can be open, and in this case, the start and end points should not be the same.
From the above discussion, it is postulated that the rheology of printable concrete is crucial, as several factors are related to it. It is useful to model printable material to predict its behaviour accurately for 3DPC. In this regard, Perrot et al. (2016) developed a time-dependent yield stress (τ0 (t)) model and failure time (tf) model of layer by layer 3DPC as shown in equations (1) to (3). Equation (4) is the yield stress calculation, and equation (5) gives the deformability from the slump flow test of concrete (Roussel and Coussot, 2005). Equation (6) describes the variation of viscosity with time (Petit et al., 2007).
Yield stress development with time can be expressed as follows:
Critical failure time, i.e. the minimum time between printing of layers to allow the bedding layer to develop enough strength, is given by:
The yield stress (τc) of concrete can be estimated from slump flow test results as follows:
For testing deformability, a small slump cone of diameter d0 = 100 mm is used. In the slump tests, no external vibration is applied to consolidate the fresh concrete. The maximum diameter of the spread d1 and the diameter perpendicular to it d2 are then measured. The deformability (Γ) is then calculated as follows (Kim et al., 2003):
Typically, a quick reduction in deformability is required after printing for shape retention. Alternatively, a more direct measure is the plastic viscosity of the fresh concrete. The variation of plastic viscosity with time can be expressed as follows:
It is also worth mentioning that the rheological behaviour of printable concrete is closely influenced by concreting temperature and the elapsed time. Therefore, the variation of the plastic viscosity and the yield stress with elapsed time and temperature must be accurately quantified to forecast the variation of workability of cement-based materials (Petit et al., 2007), and to design matching printing speeds for appropriate 3DPC.
3.3. Mechanical properties of 3D printable materials
To date, limited information is available on mechanical properties of 3DPC. However, research results of extruded concrete material indicate reduced porosity, leading to increased strength and stiffness (Peled and Shah, 2003). This section discusses reported mechanical properties of 3DPC.
3.3.1 Compressive and flexural strength of 3D printing of concrete
Although the w/c ratio is dominant in determination of concrete paste strength, it is well-documented that air entrainment reduces strength significantly. Strength is also influenced by curing, test direction relative to interlayer joints and, in cases of fibre reinforcement, the fibre orientation. These factors are considered in the interpretation of compressive and flexural test results reported by various researchers on 3DPC specimens and control specimens prepared by standard casting and curing, as summarized in Table II.
By high-pressure extrusion, air content is reduced and strength increased as shown in Figure 11. The flexural responses of cast and extruded beams manufactured from steel fibre RC (SFRC) are shown in Figure 11 (De Koker, 2004). The SFRC reported here contained 2 per cent (by volume) of 13-mm long, straight steel fibres of diameter 0.16 mm. De Koker (2004) also showed by X-ray computer tomography imaging that extrusion aligns fibres dominantly parallel to the extrusion direction, compared with random 3D orientation in cast specimens. Together with reduced air entrainment, the fibre alignment significantly increases the flexural resistance of the extruded specimens compared with specimens manufactured from the same SFRC, but by traditional casting and vibration.
Nerella et al. (2016) found increased compressive and flexural strength in 3DPC specimens compared with cast specimens. They extracted specimens by sawing and coring from 3DPC elements in the different orientations shown in Figure 12(a) and (b). However, Le et al. (2012b) found lower compressive strengths for specimens collected from the 3DPC specimens than the cast specimens, as shown in Table II. Various mechanisms could have led to these contradicting results, including reduced air content, segregation by printing of non-optimized material and curing conditions. Reduced air content could have led to the strengthening in the specimens of Nerella et al. (2016). Le et al. (2012b) moist cured printed specimens under damp hessian, while the cast specimens were cured at 20°C after stripping at the age of 1 day, until the test age of 28 days, which could explain lower strength development in printed specimens. However, flaws in the specimens due to segregation could also have caused lower strength of printed specimens.
The direction of loading of the printed specimen has influence on the strength properties, as shown in Table II and Figures 13(a) and (b) (Feng et al., 2015). By testing parallel to the layer deposition [see testing direction I in Figure 12(a)], Feng et al. (2015) found higher compressive strength than when testing perpendicular to the layer depositions, as shown in Figure 12(a) and Figure 13. This phenomenon is difficult to explain. It is noteworthy that they heat cured their specimens and tested at the young age of 3 h. The failure mechanism of printed cubes under compression test was also investigated by the Feng et al. (2015). Specimen loaded in X, Y and Z directions as shown in Figure 13 had almost similar failure patterns. In all cases, the specimens exhibited diagonal failure with two sets of triangular cracks intersecting near the centre to form an hourglass shape on the two opposite sides (Feng et al., 2015).
The influence of the position of the joints is clearer in the results of Nerella et al. (2016) and Feng et al. (2015). In both groups’ results, compressive and flexural strengths were consistently the lowest in testing direction III [Figure 12(b)], which is when loaded parallel to joints. In this orientation, compressive and flexural splitting may occur along weak joints in 3DPC specimens under compressive and flexural loading, respectively.
The value of Young’s modulus (E-mod) is also given in Table II for the specimens tested by Feng et al. As in the case of compressive strength, different E-mod was also found for different directions of loading relative to the printing direction.
The printing time and quality may be contributing factors that may also have significant influence on the results. An example is illustrated in Figure 14 with a 3DPC wall made from a single batch of concrete, and with a size of 500 × 300 mm and 50 mm thickness. At the beginning, the printing quality is considered to be good, but with time, the hydration process accelerates and matrix becomes harder (changing initial to final setting time). Also, some moisture may be lost from the mix, leading to reduced printing quality. As a result, concrete layers and interfaces of poorer quality might have been created in the upper layers. For the same reason, mechanical properties of the specimens obtained from Samples 1, 2 and 3 as indicated in Figure 14 will also be different, and because of this, there may be variation in the results, which could explain the scatter reported by Le et al. (2012b).
From the above discussion, it is postulated that in 3DPC, the printing direction and printing time have significant effect on the overall load bearing capacity of the printed objects. This implies the significance of considering the anisotropic properties of 3D printed objects, as well as proper printing speed when designing the structures.
3.3.2 Interlayer bond strength of 3D printing of concrete
The interface between concrete layers with different ages must assure sufficiently high shear and tensile strength as a main requirement to resist actions on the structure. However, the shear and tensile stress transfer mechanisms between two concrete layers is complex. They involve the combination of different interactions and depend on several parameters that influence the transmission process, such as the amount of reinforcement (if any) crossing the interface, the compressive resistance of concrete, the roughness of the interface, the presence of cracking or the stress caused by normal forces across the interface (Kang et al., 2015).
The interlayer bond strength of 3DPC was investigated by Le et al. (2012b) with a varying time gap between printing subsequent layers, and the results are shown in Figure 15. To determine the bond strength, direct tensile testing was performed on the cylindrical specimens cored from 3DPC elements. The curing process of the specimens was the same as their compressive and flexural specimens discussed in Section 3.3.1. The printing time gap between the layers was varied from 0 min (virtually no time gap) to 7 days, i.e. cold joints, and the results were compared with the cast specimens without interfacial joints. It is clear from Figure 15 that the printing time gap has a significant influence on the bond strength. However, the main mechanisms of the bond strength were neither studied nor discussed by the authors. It is not yet clear whether chemical or mechanical interfacial bonding mechanisms dominated the bond strength variation with the interlayer printing time gap in their results. For shorter interlayer gaps, hydration or chemical reaction of binders in the fresh concrete layers may lead to formation of a crystalline structure which strengthens the bond between the two layers. With increased interlayer time, it is postulated that the chemical bond is reduced as the lower layer sets, hardens and cures. In addition, differential inter-layer drying shrinkage may occur and significantly influence the eventual interlayer bond strength. Therefore, for better bond strength of layer concrete, the printing time should be shorter.
3.3.3 Drying shrinkage of 3D printing of concrete
In current 3DPC, fine aggregates of maximum particle size up to 2 mm are used (Nerella et al., 2016; Le et al., 2012b). The finer particles have higher water demand due to larger surface to mass ratio, so this may lead to higher drying shrinkage. Also, in the mixes of Table I, the aggregate content is significantly lower, at about 50 per cent of the concrete volume, while the combination of fine and coarse aggregate in normal concrete may amount to about 70 per cent of the volume. The higher paste content may also lead to larger drying shrinkage in these concretes. Le et al. (2012b) investigated the drying shrinkage of similar 3DPC specimens in different exposure conditions for 6 months, namely, in water, moist under hessian that was kept wet for the first 70 days where after they were allowed to dry, and in a controlled climate of 60 per cent relative humidity and temperature of about 20°C. The average shrinkage of five specimens subjected to each curing regime after 6 months was 0.177, (immersed in water), 0.58 (moist) and 0.855 mm/m (60 per cent RH, 20°C). This is an acceptable shrinkage level. Nevertheless, in different materials, shrinkage will also be different. Note that in this particular mix, roughly 0.1 per cent by volume polypropylene fibre was included. Such an amount is typically included to control plastic shrinkage cracking. Nevertheless, it is clear from Le et al. (2012b) that the drying shrinkage is dominated by the exposure conditions. The influence of the manufacturing method, i.e. 3DPC or standard casting, on shrinkage has not yet been reported in the literature.
4. Reinforcement in 3D printing of concrete
For better structural performance and structural reliability, concrete structures generally need addition of reinforcement. Previous studies on extrusion of SHCC with steel bars are illustrated in Figure 16 (Van Zijl et al., 2016). For reinforced SHCC (R/SHCC), two steel bars were extruded with the SHCC as shown in Figure 16. In this case, steel bars were entered horizontally through specially provided openings at the back of the extruder. Once entered, the steel bars were drawn automatically by the extrudate, which was forced by a vertical piston extruder around the 90° bend.
However, it is a challenging task to install reinforcement in the automated layer-by-layer 3DPC process. The placement of vertical reinforcement as well as the connection between the reinforcement is not simple with the current printing technology. Therefore, to strengthen the 3DPC structure, alternative ways of reinforcement, such as composite fibre mixed into concrete, carbon fibre or fibre reinforced polymer (FRP), might be introduced, but these require thorough investigation and innovative adjustments to 3DPC technology. Preliminary study has shown that the structural capacity and ductility can be significantly improved for 3DPC columns and beams when strengthened with FRP (Kim et al., 2003). Much benefit in structural performance would be gained when a hybrid printing system is developed that can print concrete along with such reinforcement in different arrangements. Manual placement of such reinforcement is possible and has been done as shown in Figure 17 (Shanghai-based WinSun 3D Prints 6-Story Apartment Building and an Incredible Home, 2018; WASP, 2018; Architect Plans, 2018; Exquisite 400 m2 villa 3D printed on-site in Beijing in just 45 days, 2018) on intermediate layers [Figure 17(a)], or between the filament layers [Figures 17(b) and (c)], or directly by extrusion of concrete through the side of the manually pre-tied reinforcement cage [Figure 17(d)]. However, elimination of such intervention would fully automate the printing process, with benefits of geometrical accuracy, reduced manufacturing time, reduced labour cost and waste.
Apart from the approaches discussed here, the researchers in Singapore Centre for 3D Printing (SC3DP) are currently working on two other ways of installing reinforcement in 3DCP as illustrated in Figure 18. In the first method, researchers are designing a new nozzle system where steel wire can be extruded with the material as shown in Figure 18(a). In this method, the flexibility of the wire reinforcement must suit the process, especially when the nozzle needs to turn at a geometrical curve or to climb from one layer to the next. In the second method, a hybrid printing system entails automated injection of steel fibres into an existing 3DPC layer by a robot before the next layer is deposited, as illustrated in Figure 18(b). These approaches are at a preliminary stage and much research is required before adopting them successfully in 3DCP.
5. Construction industry automation through 3D printing of concrete
3DPC is an emerging high-tech development in construction technology (Bos et al., 2016; Kwon, 2002; Tay et al., 2016; Lim et al., 2012; Tay et al., 2017). The major advantages of this technique are potential high speed construction, formwork-free, minimal labour and, importantly, increased freedom to design complex geometries and shapes. This technique allows for mass customization, as it does not require every (structural) element of a building to be identical for matters of speed or cost. Besides, as the printer only prints where it is desired, with the 3D printing philosophy of additive construction as opposed to subtractive construction, materials waste can be eliminated or reduced significantly. Also, 3DP creates new shapes in an efficient manner which may solve some difficulties in the traditional construction especially for the complex type of structures. It is too early to say whether 3DPC can fully replace current cast-in situ and pre-fabricated concrete construction methods. At this moment, it is necessary to find the complexity in the current construction methods and justify whether the 3DP can be used to remove that complexity. Concrete printing can improve the current fabrication methodology by automation, which reduces the reliance on labour and can increase precision. However, considering the challenges, there is a need for intelligent system design as well as advanced material development that can lead to environment friendly and affordable construction. Finally, it is also important to have standard test methods for this type of layer by layer construction, due to particular construction process potentially causing anisotropy, i.e. properties are different in different directions. A standard data set of different benchmarking properties of 3DPC as discussed in Section 3 would be required for this new research area.
Automation already exists in the construction industry in terms of prefabrication. 3DPC presents an alternative to such prefabrication but may extend automated prefabrication from factories to the construction site. In some construction areas, especially in underground construction, removing formwork can be troublesome and costly. While innovative solutions are developed for automated reinforcement in 3DPC, a potential short-term application lies in 3DPC permanent formwork. 3DPC may present an efficient automated production process for typical thin-walled formwork, for instance for columns, into which reinforcement cages are placed and interior concrete is subsequently cast in situ to form a composite structural element.
Finally, each of the 3DPC techniques described in Section 2 uses different materials and fabrication processes and thus each has different potential. Currently, there are no integrated systems such as hardware and software available for 3DCP which can be bought from the market and used directly for printing. Different types of software are being used to create 3D modelling of objects, slicing them for layer–by-layer printing and then generating G-code for the printer to read and perform the job. Although there are computer numeric control (CNC) tools available in the market which are widely used in the automation industry, they are mainly for subtracting metal as opposed to depositing of materials as in 3DCP. Therefore, there is a need for integrated systems of hardware and software for 3DCP.
6. Standard for 3D printing of concrete
Due to required changes to concrete materials and reinforcement in 3DPC, existing design and execution standards in B&C are not applicable, which fact may be a significant deterrent. Standards are typically developed over several years, once sufficient, reliable data and sound understanding have been developed. Currently, ASTM committee F42 and ISO/TC 261 members are working on developing a standard for AM. However, there is no committee and standard for 3DPC, nor standardized tests. The authors believe there is a need for a testing guideline to develop a pool of standardized test results for comparison and modelling of mechanical properties of 3DPC structural elements. Guidelines should include the printed object size for testing, minimum number of layers/interfaces for objectively characterizing mechanical properties, procedures for compressive, flexural, tensile and bond strength tests. Guidelines for steel bar reinforcement in 3DPC, including bond between the reinforcement and the cement-based composite, are also essential for reliable design and construction. In its simplest form, a separate chapter in existing standards might be required, analogous to the treatment of precast structural concrete in Eurocode 2 (BS EN 1992-1-1:2004, 2018).
7. Concluding remarks
This paper presents the state of the art of 3DPC, motivated by its potential to revolutionize B&C industries. Through gantry, robotic and crane systems, the technology has been applied successfully in demonstration construction projects. Options for improved shape retention and buildability on sub-layers include the development of thixotropic concrete materials, or alternatively the use of low viscosity concrete that can be easily pumped, and then treated chemically at the nozzle for accelerated setting. However, the pool of data on suitable materials, structural performance, safety and economy is limited. The following research needs are identified:
To enable appropriate engineering design and construction, it is imperative that mechanical properties of 3DPC elements are characterized by standardized material tests. Tests for compressive and indirect tensile strength of 3DPC elements, based on existing test methods for concrete, have been performed by individual laboratories, as described in this contribution. Interlayer bond is integral to structural integrity and must be tested for appropriate design calculations.
Reinforcement is required to overcome quasi-brittle failure of concrete materials. Manual placement of steel reinforcement bars or assemblies of bars as inter-layer or inter-filament reinforcement as reported in this paper, presents a solution. Placement of reinforcement may be automated as part of the 3DPC process, or through other automated devices. In both manual and automated processes, the bond of reinforcement with the concrete remains to be studied carefully to provide design guidelines for appropriate anchorage and curtailment of reinforcement in 3DP concrete structures and structural elements.
For acceptance of 3DPC as a construction technology by B&C authorities, standards for materials, specification, manufacturing, testing and structural design are required. These standards and their appropriate application in design and construction must ensure that appropriate levels of reliability are achieved.
Materials compositions of 3D printable concrete
|Materials compositions (kg/m3)|
|Nerella et al. (2016)||430||170||180||1240||180||10||–|
|Le et al. (2012b)||579||165||83||1241||232||16.5||1.2 (PP)|
|Anell (2015)||659||87||83||1140||228||11.6||1.2 (PP)|
|Perrot et al. (2016)||Binder content expressed as a weight:
Cement 50%, limestone filler 25%, kaolin 25%, w/c = 0.41,SP/cement = 0.3%
|Malaeb et al. (2015)||Cement 125 g, sand 80 g, filler 160 g, water:cement = 0.39, SP = 0.5-1 mL|
SP: superplasticizer; PP: polypropylene (12/0.18 mm length/diameter)
Comparison of mechanical properties of cast and printed objects
|Printed specimens||Cast specimens|
|Authors||Specimens retrieved from||Type of test||Testing direction||3 days||21 days||28 days||21 days||28 days|
|Nerella et al. (2016)||1000 × 300 × 35-mm wall||Compressive strength (MPa) (size: 35-mm cube)||I||45.9||83.5||–||73.4||–|
|Flexural strength (MPa) (size: 160 × 35 × 35 mm)||I||4.8||5.8||5.1|
|Le et al. (2012b)||350 × 350 × 120-mm slab||Compressive strength (MPa)
(size: 100-mm cube)
|500 × 350 × 120-mm slab||I||–||–||102|
|Curvy bench||Compressive strength (MPa) (size: 63-mm cube)||I||–||–||74|
|500 × 350 × 120-mm slab||Flexural strength (MPa) (size: 400 × 100 × 100 mm)||I||–||–||16.5||–||11|
|Curvy bench||Flexural strength (MPa) (size: 220 × 63 × 50 mm)||I||–||–||12|
|Feng et al. (2015)||70.7-mm cube||Compressive strength (MPa)||Testing time 3 h after casting|
|50-mm cube||Compressive strength (MPa)||I||16.8 (7.1)*|
*The values in the brackets are E-modulus and unit in GPa
Akkaya, Y., Peled, A. and Shah, S.P. (2000), “Parameters related to fibre length and processing in cement composites”, Materials and Structures, Vol. 33 No. 8, pp. 515-524.
Anell, L.H. (2015), “Concrete 3d printer”, MSc thesis, Civil engineering, Lund University.
Apis Cor (2018). Apis Cor printer, available at: https://3dprint.com/100229/apis-cor-3d-printer-buildings/ (accessed 27 April 2016).
Architect Plans (2018), “Architect plans to 3d print a 2-story home in Minnesota using a homemade cement printer”, available at: https://3dprint.com/2471/3d-printed-home-in-minnesota (accessed 3 May 2016).
Bak, D. (2003), “Rapid prototyping or rapid production? 3D printing processes move industry towards the latter”, Rapid Prototyping Journal, Vol. 23 No. 4, pp. 340-345.
Barnes, H.A. (1997), “Thixotropy – a review”, Journal for Non-Newtonian Fluid Mechanics, Vol. 70 Nos 1/2, pp. 1-33.
BetAbram (2018), BetAbram 3D printer, available at: http://betabram.com/ (accessed 28 April 2016).
Bos, F., Wolfs, R., Ahmed, Z. and Salet, T. (2016), “Additive manufacturing of concrete in construction: potentials and challenges of 3D concrete printing”, Virtual and Physical Prototyping, Vol. 11 No. 3, doi: 10.1080/17452759.2016.1209867.
BS EN 1992-1-1:2004 (2018), “Eurocode 2: design of concrete structures – Part 1-1: general rules for buildings”, British Standards Institute.
Chua, C. and Leong, K. (2014), 3D Printing and Additive Manufacturing: Principles and Applications, 4th ed, World Scientific.
Concrete Plans (2018), “Concrete plans: CyBe’s Berry Hendriks describes plans to 3D print with mortar”, available at: www.3Dprint.com (accessed 7 January 2015).
Contour crafting (2018), “Contour crafting: automated construction”, available at: http://contourcrafting.com/ (accessed 30 June 2017).
De Koker, D. (2004) “Manufacturing processes for engineered cement-based composite material products”, MSc thesis, Stellenbosch University.
Dini, E. (2015), “D-Shape – steriolithography 3-D printing technology”, Monolite- UK-Ltd, available: www.d-shape.com/cose.htm (accessed 15 June 2016).
Exquisite 400 m2 villa 3D printed on-site in Beijing in just 45 days (2018), Exquisite 400 m2 villa 3D printed on-site in Beijing in just 45 days, available at: www.3ders.org/articles/20160614-exquisite-400-m2-villa-3d-printed-on-site-in-beijing-in-just-45-days (accessed 14 June 2016).
Feng, P., Meng, X., Chen, J.F. and Ye, L. (2015), “Mechanical properties of structures 3D printed with cementitious powders”, Construction and Building Materials, Vol. 93, pp. 486-497.
Gardiner, J.B. (2011), Exploring the Emerging Design Territory of Construction 3D Printing – Project Led Architectural Research, PhD thesis, RMIT University.
Gibson, I., Rosen, D.W. and Stucker, B. (2015), Additive Manufacturing Technologies: 3D Printing, Rapid Prototyping, and Direct Digital Manufacturing, Springer-Verlag, New York, NY, doi: 10.1007/978-1-4939-2113-3-1.
Hull, C.W. (1986), Apparatus for Production of Three-Dimensional Objects by Stereolithography, US Patent 4575330.
Hüsken, G. and Brouwers, H.J.H. (2012), “On the early-age behaviour of zero-slump concrete”, Cement and Concrete Research, Vol. 42 No. 3, pp. 501-510.
Jha, K.N. (2012), Formwork for Concrete Structures, Tata McGraw Hill Education Private Limited, New Delhi, ISBN (13): 978-1-25-900733-0.
Jolin, M., Burns, D., Bissonnette, B., Gagnon, F. and Bolduc, L.S. (2009), “Understanding the pumability of concrete”, in Proceedings for the Conference Shotcrete for Underground Support (XI), Davos.
Kang, J.Y., Park, J.S., Jung, W.T. and Keum, M.S. (2015), “Connection between concrete layers with different strengths”, Engineering, Vol. 7 No. 7, pp. 365-372.
Kazemian, A., Yuan, X., Cochran, E. and Khoshnevis, B. (2017), “Cementitious materials for construction-scale 3D printing: laboratory testing of fresh printing mixture”, Construction and Building Materials, Vol. 145, pp. 639-647.
Khayat, K.H. and Assaad, J.J. (2006), “Effect of w/cm and high-range water-reducing admixture on formwork pressure and thixotropy of self-consolodating concrete”, ACI Materials Journal, Vol. 103 No. 3, pp. 186-193.
Khayat, K.H. and Assaad, J.J. (2008), “Thixotropy-enhancing agent to reduce formwork pressure exerted by self-consolidating concrete”, ACI Materials Journal, Vol. 105 No. 1, pp. 88-96.
Khoshnevis, B. (2004), “Automated construction by contour crafting – related robotics and information technologies”, Automation in Construction, Vol. 13 No. 1, pp. 5-19.
Khoshnevis, B., Hwang, D., Yao, K.T. and Yeh, Z. (2006), “Mega-scale fabrication by contour crafting”, International Journal of Industrial and Systems Engineering, Vol. 1 No. 3, pp. 301-320.
Kim, Y.Y., Kong, H.J. and Li, V.C. (2003), “Design of engineered cementitious composite suitable for wet-mixture shotcreting”, ACI Materials Journal, Vol. 100 No. 6, pp. 511-518.
Kwon, H. (2002), “Experimentation and analysis of contour crafting (CC) process using uncured ceramic materials (Electronic)”, PhD thesis, University of southern California, USA.
Lawson, N., Douglas, I., Garvin, S., McGrath, C., Manning, D. and Vetterlein, J. (2001), “Recycling construction and demolition wastes – a UK perspective”, Environmental Management and Health, Vol. 12 No. 2, pp. 146-157.
Le, T.T., Austin, S.A., Lim, S. and Buswell, R.A. (2012a), “Mix design and fresh properties for high-performance printing concrete”, Materials and Structures, Vol. 45 No. 8, pp. 1221-1232.
Le, T.T., Austin, S.A., Lim, A., Buswell, R.A., Law, R., Gibb, A.G.F. and Thorpe, TT. (2012b), “Hardened properties of high-performance printing concrete”, Cement and Concrete Research, Vol. 42 No. 3, pp. 558-566.
Lim, S., Buswell, R.A., Le, T., Austin, S.A., Gibb, A.G.F. and Thorpe, T. (2012), “Developments in construction-scale additive manufacturing processes”, Automation in Construction, Vol. 21, pp. 262-268.
Lim, S., Buswell, R., Le, T., Wackrow, R., Austin, S.A., Gibb, A.G.F. and Thorpe, T. (2011), “Development of a viable concrete printing process”, in proceeding for 28th International Symposium on Automation and Robotics in Construction (ISARC2011), Seoul, pp. 665-670.
Malaeb, Z., Hachem, H., Tourbah, A., Maalouf, T., Zarwi, N.I. and Hamzeh, F. (2015), “3D concrete printing: machine and mix design”, International Journal of Civil Engineering and Technology, Vol. 6 No. 6, pp. 14-22.
Mechtcherine, V., Nerella, V.N. and Kasten, K. (2014), “Testing pumpability of concrete using sliding pipe rheometer”, Construction and Building Materials, Vol. 53, pp. 312-323.
Mitsoulis, E. (2007), “Flows of viscoplastic materials: models and computations”, Rheology Reviews, pp. 135-178, ISBN: 0-9547414-6-3.
Nerella, V.N., Krause, M., Näther, M. and Mechtcherine, V. (2016), “Studying printability of fresh concrete for formwork free concrete on-site 3D Printing technology (CONPrint3D)” in Proceeding for the 25th Conference on Rheology of Building Materials, Regensburg.
Peled, A. and Shah, S.P. (2003), “Processing effects in cementitious composites: extrusion and casting”, Journal of Materials in Civil Engineering, Vol. 15 No. 2, pp. 192-199.
Perrot, A., Rangeard, D. and Pierre, A. (2016), “Structural built-up of cement-based materials used for 3Dprinting extrusion techniques”, Materials and Structures, Vol. 49 No. 4, pp. 1213-1220.
Petit, J.Y., Wirquin, E., Vanhove, Y. and Khayat, K. (2007), “Yield stress and viscosity equations for mortars and self-consolidating concrete”, Cement and Concrete Research, Vol. 37 No. 5, pp. 655-670.
Pierre, A., Lanos, C. and Estelle, P. (2013), “Extension of spread-slump formulae for yield stress evaluation”, Applied Rheology, Vol. 23 No. 6, pp. 638-649.
Roussel, N. (2006), “A thixotropic model for fresh fluid concretes: theory, validation and applications”, Cement and Concrete Research, Vol. 36 No. 10, pp. 1797-1806.
Roussel, N. and Coussot, P. (2005), “Fifty cent rheometer for yield stress measurements: from slump to spreading flow”, Journal of Rheology, Vol. 49 No. 3, pp. 705-718.
Rushing, R.S., Al-Chaar, G.K., Eick, B.A., Burroughs, J.F., Shannon, J., Barna, L.A. and Case, M.P. (2017), “Investigation of concrete mixtures for additive construction”, Rapid Prototyping Journal, Vol. 23 No. 1, pp. 74-80.
Secrieru, E., Mechtcherine, V., Schröfl, C. and Borin, D. (2016), “Rheological characterisation and prediction of pumpability of strain-hardening cement-based composites (SHCC) with and without addition of superabsorbent polymers (SAP) at various temperatures”, Construction and Building Materials, Vol. 112, pp. 581-594.
Shanghai-based WinSun 3D Prints 6-Story Apartment Building and an Incredible Home (2018), Shanghai-based WinSun 3D Prints 6-Story Apartment Building and an Incredible Home, available at: https://3dprint.com/38144/3d-printed-apartment-building/ (accessed 3May 2016).
Shao, Y. and Shah, S.P. (1997), “Mechanical properties of PVA fibre reinforced cement composites fabricated by extrusion processing”, ACI Materials Journal, Vol. 94 No. 6, pp. 555-564.
Spiratos, N., Pagé, M., Mailvaganam, N., Malhotra, V.M. and Jolicoeur, C. (2003), “Superplasticizers for concrete: fundamentals, technology, and practice”, Supplementary Cementing Materials for Sustainable Development, Incorporated, Ottawa, pp. 322, ISBN 0-9731507-1-8.
Taylor, M., Wamuziri, S. and Smith, I. (2003), “Automated construction in Japan”, Proceedings of the Institution of Civil - Civil Engineering, Vol. 156 No. 1, pp. 34-41.
Tay, Y.W.D., Panda, B., Paul, S.C., Tan, M.J. and Leong, K.F. (2017), “3D printing trends in building and construction industry: a review”, Virtual and Physical Prototyping, Vol. 12 No. 3, pp. 261-276.
Tay, Y.W., Panda, B., Paul, S.C., Tan, M.J., Qian, S., Leong, K.F. and Chua, C.K. (2016), “Processing and properties of construction materials for 3D printing”, Materials Science Forum, Vol. 861, pp. 177-181.
TU Eindhoven (2018), “TU Eindhoven starts using king size 3D concrete printer”, available at: www.tue.nl/en/university/news-and-press/news/22-10-2015-tu-eindhoven-starts-using-kingsize-3d-concrete-printer/ (accessed 29 April 2016).
Van Zijl, G.P.A.G., Paul, S.C. and Tan, M.J. (2016), “Properties of 3D printable concrete”, Proceeding for 2nd International Conference on Progress in Additive Manufacturing (Pro-AM), pp. 421-426.
Visser, C.R. (2007) “Mechanical and structural characterization of extrusion moulded SHCC”, MSc thesis, Stellenbosch University.
Wakisaka, T., Furuya, N., Inoue, Y. and Shiokawa, T. (2000), “Automated construction system for high rise reinforced concrete buildings”, Automation in Construction, Vol. 9 No. 3, pp. 229-250.
Wallevik, O. (2002) “Pumping of concrete”, Rheology course, IBRI Concrete Industry, Rheology of cement suspensions, Practical applications, pp. 623-627.
WASP (2018), “WASP develops new 3D printing method for modular reinforced concrete beams”, available at: www.3ders.org/articles/20150811-wasp-develops-new-3d-printing-method-for-modular-reinforced-concrete-beams.html (accessed 3 May 2016).
WinSun (2018), “WinSun China builds world’s first 3D printed villa and tallest 3D printed apartment building”, available at: www.3ders.org (accessed 14 May 2018).
Wolfs, R.J.M. (2015), “3D printing of concrete structures”, MSc thesis, Eindhoven University of Technology.