BASIC FRESH-STATE PROPERTIES OF EXTRUSION-BASED 3D PRINTED CONCRETE

Trodimenzionalno štampanje (3D štampanje) ili aditivna proizvodnja u opštem smislu predstavlja sukcesivno dodavanje slojeva materijala kako bi se formirao željeni model iz 3D digitalnog geometrijskog modela. Važno je napomenuti da se, iako postoje različite vrste aditivnih tehnologija (deponovanje istopljenog filamenta, vezivna 3D štampa, stereolitografija, selektivno lasersko sinterovanje itd.), vrste materijala (materijali u čvrstom, tečnom, gasovitom stanju, praškasti materijali, laminati) i primene (za izradu prototipova ili proizvodnju), izraz 3D štampanje koristi kao opšti termin za sve aditivne tehnološke procese. Drugi termini koji su takođe u upotrebi jesu digitalna fabrikacija, tehnologija brze izrade prototipa ili CAD dizajn [27]. Proces 3D štampanja sastoji se od kreiranja 3D modela u computer-aided-design (CAD) formatu i njegovom eksportovanju u stereolitografski (STL) format,


INTRODUCTION
Three -dimensional printing (3D printing) or additive manufacturing is in general defined as successive assembling of the material layers in order to make objects or structures from a 3D data model. It is important to notice that, although there are several types of additive technologies (material extrusion, binder jetting, vat photo polymerization, powder bed fusion, etc.), materials used (i.e. materials in solid, liquid, gas state, powders or sheets)and application (for making of prototypes or for production), the term 3D printing is usually used as an umbrella term for all additive manufacturing processes. Other terms used as a synonym for additive manufacturing are digital fabrication, rapid prototyping or computer-aided design [27].
According to the American Society for Testing and Materials, there are seven categories of additive manufacturing technologies: binder jetting, directed energy deposition, material extrusion, material jetting, powder bed fusion, sheet lamination and vat photo polymerization [7].
The two most researched and developed additive manufacturing technologies used for 3D printing of concrete are material extrusion and binder jetting [27,42]. Material extrusion is based on selective dispensing of the material through a nozzle, making a multi-layer objects, with no need for formwork use [27]. Given the nature process, reinforcement is not used, although there have been attempts to incorporate it in extrusion-based 3D printing [1,2,35]. Under the extrusion-based concrete printing, two technologies have been eventually developed: Contour Crafting and concrete printing [27,42]. Contour Crafting (Figure 1b) is based on extrusion of a mortar or cementitious materials in layers, against a trowel which forms the smooth surface of the printed object [12,17]. Concrete printing is extrusion process as well, but with smaller deposition resolution and retained 3-dimensional freedom ( Figure 1c) [16,17]. The extrusion-based processes are used for "in-situ" construction, while powder based 3D concrete printing(i.e. binder jetting)is suitable for making a precast structure elements. This printing technology is based on selective deposition of a liquid binder into a powder bed, creating the 3D object at the targeted areas where the powder is bound [42]. Powder-based 3D concrete printing is developed on the basis of methods for polymers and metals additive manufacturing, adjusted to concretepowder bed can be made of aggregate in which a fluid cement paste is jetted, or the powder bed can be a cement-based or geopolymer-based binder in which the water is jetted [27]. This additive manufacturing technology applied in construction is also known as "D-shape" process ( Figure 1a) [4,27].

Materials used for extrusion based process
The compositions of concrete for extrusion-based 3D printing and conventional cement concrete are significantly different. Concrete for 3D printing has larger amount of Portland cement and consequently, smaller amount of aggregate. In addition, to achieve printability of concrete, the coarse aggregate is omitted from this composite. The most commonly found nominal maximum aggregate size in literature is 1-2mm [13,20,24,25,33,37,40,41], but nominal maximum sizes up to 4,75 mm is reported in experimental research as well [1,3,5,36]. However, there has been a recent study on the impact of coarse aggregate (maximum size 10mm) on basic 3D concrete fresh-state properties [32]. Coarse aggregate has been used for constructing a house with specific huge 3D extrusion-based printer [35].
improve the performance of printing mixtures. The challenges of 3D concrete mix design that need to be addressed to are establishing relation between the rheological and technological properties with printability of 3D printed concrete and, early age drying shrinkage reduction, due to the absence of formwork and large quantities of cement -more than 500 kg/m 3 which is why polypropylene fibres are commonly used in mixture design [11,19,29,33,34,37,38]. According to some studies, Portland cement is, for now, the most reliable binder that can ensure achieving the required 3D concrete properties [11,29]. However, large amount of cement used has a negative impact on environment, enlarges consumption of natural raw materials and increases cost of 3D concrete printing [8]. In order to resolve these problems, the previous experience in using recycled lightweight aggregates and supplementary cementitious materials in traditionally casted concrete mixtures [18,30,31] can be adjusted and applied for 3D printed concretes. The use of lightweight coarse aggregate [32] and alternative binders are investigated, in order to meet the needs of an eco and environmental friendly concrete and required properties for 3D concrete printing (e.g. supplementary cementitious materials and geopolymer binder) [23,25].

History and application
The first officially patented attempt to automatize concrete casting was made by Thomas Edison in 1917. The process consisted of single-pouring a Portland cement mixture into the single-piece mould made of cast iron, through a pump and a hose system, from the top to the bottom of the mould. The mould is supposed to be in shape of entire structure (e.g. a house or a building) after assembling, making an integral construction after hardening of the cement mixture. However, due to inability to overcome the complexity of concrete mixture properties and high cost of described mould, this patent was implemented only a few times [2,6,39,47].
Significant increase in automation of in-situ construction of concrete structures started in 1990s. In the first part of the decade, this reflected in Japanese automized assembling of prefabricated elements using specialized robots [39]. Research on possibilities of the additive manufacturing application in construction of concrete structures begun with powder bed fusion technology published by Joseph Pegna (Department of Mechanical Engineering, Aeronautical Engineering and Mechanics, Rensselaer Polytechnic Institute, Troy, New York, USA) in 1997. Extrusion-based 3D concrete printing using Contour Crafting process was introduced in 1998 by Berokh Khoshnevis, Professor of Engineering at the University of Southern California. In 2004, the 1:1 scale printed wall was shown, and this 3D printing process further developed to large scale on-site construction technology [2,12,39].
Since then, 3D concrete printing has been researched more extensively, especially since 2012. The freeform construction is of great interest for architects, giving them more design freedom. The use of formwork and moulds for traditional concrete casting limits the creative expression of architects, since the complex geometry formwork is very expensive and irrational to make. Furthermore, it is estimated that formwork costs are 35%-65% of total concrete construction costs, while formwork has limited or no possibility for re-use [2]. Making concrete structures without formwork would decrease the construction cost as well as construction waste and increase the low annual growth in productivity common for construction industry [2,27,35].
Although the research in the field of 3D concrete printing is at its early stage and still improving and the standardized methods for 3D printed concrete mixtures design and testing still do not exist, there have been impressive examples world-wide of potential successful application of this technology. Some of them are illustrated in the following figures.

BASIC FRESH-STATE PROPERTIES OF EXTRUSION-BASED 3D PRINTED CONCRETE
Fresh-state properties of 3D printed concrete are complex, overlapping and highly time-dependent. They can be divided into rheological (yield stress, plastic viscosity and thixotropy), technological properties (pumpability and flowability) and printability properties (extrudability, print quality and buildability). The printability of the 3D concrete is the most important property for 3D printing process. It is not strictly defined, but it refers to fresh-state properties needed to successfully conduct the printing process, from extrusion to the end of the process [39].
Since there are no standardized methods and procedure for defining and evaluating these complex 3D concrete properties, the terminology found in the body of se sprovelo štampanje, od ekstrudiranja do kraja procesa [39].
Pumpability (or deliverability) is the ability of concrete mixture to be transported from mixer, through a hose, to a nozzle. It is related to plastic viscosity and yield stress of the concrete mixture (i.e. rheology properties) and power of the pump as well as its technology must be chosen in accordance with them [27]. The challenges of pumping 3D concrete mixtures reflect in the timedependent fresh-state properties, bleeding and segregation. The sufficient amount of cement paste is necessary so it could form a coat around aggregate particles and reduce the friction between them and reduce the shear stress. This will lead to increase of the 3D mixture workability in terms of pumpability/in terms of transporting through the pumping system [37]. In the study on printability region for 3D printed concrete [37], authors introduce the pumpability index to quantify the pumpability of the printing mixture. For each trial mixture, pumpability index is calculated as ratio of the mixture flow rate and water flow rate (ml/s) for constant pumping speed, in a selected time interval, to obtain the results. Higher pumpability index indicates easier mixture pumping [37].
Flowability is the ease with which concrete flows under given conditions, and is usually tested by the slump flow test [19].The impact of an optimum aggregate content on flowability is investigated by Zhang et al. [43].
Extrudability can be compromised by water drainage and phase separationsimilar as with pumping flow, although extrusion is a process with slower flow velocities [39]. In addition, pumpability and extrudability are similar due to shearing of the mixture, although during extrusion, the mixture is sheared in the nozzle, under different condition than shearing in the pipes during pumping [34]. It is examined visually, usually on the layer extruded in predefined time period. There has been no recommendations for more reliable extrudability testing method [15,33]. According to some authors, extrudability can be assessed through the print quality properties [33], which are explained in the following paragraph.
Print quality in the literature refers to three printing mixture requirements. The first one is surface quality of the deposited layers, which has to be free of defects, i.e. discontinuities. The tearing of the layers (Figure 9) is often reported in different studies and it appears due to the excessive stiffness and low cohesion of the printing mixturepoor workability. Next requirement for the print quality is squared edges, which means that the edges of the deposited layer must be visible. The term squared edges is proposed by the researchers that conducted studies using squared printer nozzles, but it refers to the consistency of the deposited layer's shape with the shape of the nozzle, in general/regardless of the nozzle type [11]. . Surface qualitytearing of the printed layer due to low cohesion and excessive stiffness of the printing mixture [10] Treći zahtev je dimenziona usaglašenost i konzistentnost dimenzija. Da bi ovaj kriterijum bio ispunjen, potrebno je definisati prihvatljiv raspon dimenzija sloja. Ukoliko su dimenzije odštampanog sloja u okviru predviđenog raspona, dimenziona usaglašenost je postignuta. Dimenziona usaglašenost se odnosi na sposobnost mešavine da bude odštampana u slojevima s prihvatljivim rasponom ciljanih dimenzija. Konzistentnost dimenzija odnosi se na variranje širine odštampanog sloja. U studiji o svojstvima mešavine za štampanje u svežem stanju, Kazemian i dr. objavili su da variranje prethodno definisane širine sloja do 10% daje prihvatljiv kvalitet štampanja. Ova variranja odnose se samo na sloj u svežem stanju, odmah nakon deponovanja i ne uzimaju u obzir promene u dimenzijama usled skupljanja betona [10,11]. Dimenziona usaglašenost ispitivana u Kazemian i dr. eksperimentu prikazana je na slici 10, a rezultati merenja konzistentnosti dimenzija na slici 11. Svojstva kvaliteta štampanja ispituju se vizuelno i preporučeno je da se podešavaju kroz više proba s nekoliko ponovljenih mešavina, jer ne postoje predložene metode za ispitivanje i ocenu ovih svojstava [7].
Kao kvantitativni parametar za stabilnost oblika, Panda i dr. koriste faktor stabilnosti oblika, izračunat kao odnos poprečnog preseka ekstrudiranog materijala i poprečnog preseka dizne [26]. Prema eksperimentalnim istraživanjima, napon na granici tečenja raste linearno tokom perioda mirovanja, uzrokujući linearan porast stabilnosti oblika s vremenom. Stabilnost oblika utiče na očvrsla svojstva štampanog betona, preko povezanosti s vremenskim razmakom štampanja između sukcesivnih slojeva (engl. printing time gap). Kraći vremenski razmak uzrokovaće jaču mehaničku vezu dva odštampana sloja nakon očvršćavanja i obrnuto. Ipak, štampanje u kratkom vremenskom razmaku izvodljivo je samo za mešavine s visokom stabilnosti oblika, jer odštampani sloj mora imati sposobnost da izdrži težinu narednih slojeva odmah nakon ekstrudiranja. Dalje, vremenski razmak štampanja će zavisiti od rastojanja koje prelazi dizna tokom štampanja sloja (tj. dužine sloja) i brzine štampanja [11]. Ovo naglašava složenost i zavisnost više svojstava štampanih betona u svežem i očvrslom stanju od karakteristika štampača. rheological properties of the highest importance for withstanding the gravity-induced compressive stresses in single and multiple layers [34]. Even if this stress does not exceed the yield stress, significant deformation can occur, due to the cumulative stress in layers, and it can impact the geometry of the printed layers. Additionally, slender vertical printed models are exposed to the buckling, after a certain number of printed layers/model height is achieved, which is why printing concrete must develop high elastic modulus relatively early. It is proven that, with the height increase of the printed object, yield stress needed to resist compressive stress increases linearly, while the elastic modulus needed to resist the buckling increases with the power of 3 of the printed object height. This means that below certain object height (number of printed layers) compressive stress failure will be critical for the buildability. Above this height, the buckling will have most impact on the buildability of the printing concrete [39]. Although all researchers agree that the further research and experimental data is needed for better understanding the mechanisms of forming the shape stability, the use of nanoclay in mixtures is reported to increase the build-up or "structuration at rest" of the printing mixtures [11,22,23,46]. Figure 12 shows the buildability analysis in the experimental study by Panda et al. Besides the number of deposited layers before plastic deformation (i.e. collapsing), elastic deformation of layers is examined on the control mixture and clay modified mixture, which showed higher buildability [22]. Figure 13 shows the consequences of the relations between buildability and pumpability, flowability and shape retention, presented by Tay et al. [37].
As a quantitative parameter for shape stability, Panda et al. use shape retention factor, calculated as ratio between cross section area of extrudate and cross section area of the nozzle [26]. According to the experimental research, yield stress increases linearly during the dormant period, causing linear increase of shape stability with time. The shape stability impacts the hardened properties of printed concrete through the linkage with printing time gap between successive layers. The shorter printing time gap will result in higher mechanical bond between layers and vice versa. However, printing with short time gaps can be feasible only if the mixture has high shape stability, meaning that the printed layer must be capable of withstanding the weight of the following layer right after extrusion. Furthermore, printing time gap will depend on nozzle travelling distance during layer printing (i.e. the layer length) and printing speed [11]. This emphasizes the complexity and dependence of multiple fresh and hardened printing concrete properties and printer properties. Slika 12. Laboratorijsko ispitivanje "buildability" svojstva odštampane mešavine, opisano preko broja deponovanih slojeva pre obrušavanja [22] Figure 12. Laboratory testing of buildability of the printing mixtures describe through number of deposited layers before the structure collapsed [22] Imajući na umu navedeno, većina naučnika predlaže kombinovano ispitivanje stabilnosti oblika. Kazemian i dr. procenjivali su stabilnost oblika preko sleganja slojeva i ispitivanja cilindrom, za vremenski razmak štampanja slojeva od 0 i 19 minuta. Sleganje slojeva ispitano je na dva odštampana sloja, jedan preko drugog, i merenjem sleganja analizom fotografija u program za obradu fotografija. Ispitivanje za vremenski razmak od 19 minuta pokazalo je da nema deformacija u donjem sloju za mešavinu s dodatkom silikatne prašine, polipropilenskih vlakana i nano-gline, dok je prosečna vrednost deformacija za pet merenja na mešavinama s Portland cementom bila 1,5 mm [11]. Ovo ispitivanje prikazano je na slici 14, a uzorak od pet sukcesivnih slojeva prikazan je na slici 15.
Additionally, in this experimental study, shape stability was measured by cylinder stability test, while the similar was proposed by Perrot et al. [10,29]. While Kazemian et al. proposed imposing a constant load of 5.5kg and measuring the change in specimen height, Perrot et al. increased the load in 1.5N increments, in order to obtain the maximum stress before the plastic deformation [10,29]. For example, for 17 seconds printing time gap, the 4 specimens average failure time was 656 seconds after load imposing, at 4,76 kPa, average. For 60 second time gap, no plastic deformation was detected [29]. The specimens tested by Perrot et al. for different printing time gaps are shown at Figure 16 and Figure 17.

CONCLUSIONS
This paper presents the basic fresh-state properties of extrusion-based 3D printed concrete -pumpability, flowability (technological properties), extrudability, print quality and buildability (printability properties). All analyzed properties are highly time-dependent and printer-dependent as well. Their definitions are often overlapping and their complexity is challenging for finding general solutions and procedures for designing the printable mixtures with satisfying performance.
Although the field of 3D concrete printing is recently extensively researched, further experimental results are needed in order to propose reliable models for linking technological and rheological properties of printing concrete with printability properties. There is still a great need for theoretical knowledge to express, quantitatively and qualitatively, the desirable printing mixture properties. Specific fresh -state properties of 3D printing concretes required establishing the new terminology compared to the traditionally casted concretes. For example, one of the basic properties of the fresh concreteworkabilitycannot be defined and described in the same way for 3D concrete as for traditional concrete [11]. Workability is considered as the ability of concrete to "be properly compacted and also transported, placed and finished sufficiently easily without segregation", or, more strictly, as "the amount of useful internal work or energy required to overcome the internal friction between concrete and the formwork or reinforcement" [21]. For example, the important property of 3D concrete is buildability, i.e. its correlation with structuration rate, which is time dependent. The consistency and setting time of the printing mixture must be in certain interval to meet the needs of extrusion, but on the other hand, printed layers must obtain green strengths almost immediately after extruding, in order to bare the load of the following layers. Therefore, the usual definition of workability should be adjusted, since this term in 3D printing process combines a whole set of inter-dependable factors. It is recommended to evaluate the workability of 3D printing mixtures by investigating fresh-state concrete properties relevant mora biti prilagođena, jer ovaj termin u procesu 3D štampanja obuhvata širok spektar međuzavisnih faktora. Preporučeno je da se obradljivost 3D mešavina za štampanje ocenjuje ispitivanjem svojstava betona u svežem stanju koja su relevantna za 3D štampane betone -kvalitet štampe, stabilnost oblika i svojstvo printability window [11].
Since there are no standardized methods for testing and evaluation of 3D printed concrete properties, most of the research is still based on trial-and-error experimental approach. That is why every research result in the field of 3D concrete printing contributes to further development and improvement of this technology.