Investigation of additive incorporation on rheological, microstructural and mechanical properties of 3D printable alkali-activated materials

Abstract This study investigates the addition of Poly-vinyl Alcohol (PVA) fibres and attapulgite nanoclay to alkali-activated materials (AAMs) with the aim of enhancing the mechanical performance and optimizing the printability and buildability of AAMs. The fresh properties of six mix formulations, including flowability, slump values, rheology, shape retention, and extrusion window, were evaluated. The best performing mixes, that exhibited optimal fresh properties, were 3D printed, and their mechanical performance, microstructure, and buildability were investigated. The addition of 1 wt.-% attapulgite nanoclay (i.e. A-1) showed the desirable fresh properties required for 3D printing, as well as providing sufficient mechanical reinforcement to the samples. The 3D printed A-1 samples showed an improved flexural and compressive strength by 43% and 20%, respectively, compared to both the casted and printed control mixes. Moreover, microstructure analysis, including SEM, Rapidair measurement, and micro-CT, provided evidence of the compatibility by showing the lowest pores anisotropy and mixture homogeneity, between attapulgite and AAMs.

3 2000 nm. This resulted in a high aspect ratio making nano-clays highly reactive [21]. In line with the aforementioned results and author's previous works [14,22], the fresh property investigation proved to be an effective and useful tool for identifying the suitable mix in terms of both mechanical property and printability. Therefore, in this study, PVA fibre with the dosages of 0.25 wt.-% and 0.5 wt.-%, attapulgite nano-clay with the dosages of 0.5 wt.-% and 1 wt.-%, and a combined mix composed of 0.25 wt.-% of PVA and 1 wt.-% of attapulgite have been added to the alkaliactivated mixture. All the mixtures were characterized via fresh property tests, including flowability, rheology, mini slump, extrudable window, shape retention, and buildability. The chosen mixtures were then printed and investigated in the hardened state to evaluate the effect of additives on microstructure (i.e. SEM, RapidAir measurement, and µ-CT) and mechanical properties (i.e. flexural strength, and compressive strength). Finally, the correlation between fresh properties and mechanical performance was examined.

Materials and experimental testing methods
Outlined Figure 1 is the comprehensive framework of all the mix design and experimental procedures employed in this study. According to the author's previous study [14], the most effective alkali-activated cementitious composite for 3D printing was composed of (i) fly ash (Cemex, UK) which meets normal fineness (N) Category of British standard BS EN 450-1:2012. The bulk density of fly ash is ranged between 0.8 -1.0 g/cm 3 with the maximum permitted residue of 40% on a 45 µm sieve in dry state, (ii) ground-granulated blast furnace slag (Hanson Heidelberg Cement, UK) which complies with EN15167-1 with the bulk density of approximately 2.4 -3 g/cm³ (20 °C). D50 and D90 percentiles of 17.3 µm and 49.27 µm, respectively, were registered for this product. More details on the material characterizations, including the material size distribution and compositions, can be found in previous researches [23,24]  Industries Ltd, UK) were incorporated as a reinforcement and thixotropic modifier additives, respectively. As indicated in Figure 2b, attapulgite nano-clay particles have an angular shape with sharp edges. Based on the data provided by the supplier, the material has a bulk density of 0.769 g/cm³ and maximum permitted residue of 45% on a 44µm sieve in a dry state.   Table 2, a total of six alkali-activated mixes were produced. The FA, GGBS, and SF content were set to 60, 25, and 15 wt.-%, respectively, by the overall weight of the binder. Moreover, the activator solution dosage (i.e. NaOH+Na 2 SiO 3 ) used was 40 wt.-% of the total weight of the binder. Two particle size ranges of 0 to 0.5 mm and 0.5 to 1 mm were obtained after oven-drying and sieving the fine river sand. The sand to binder ratio used in this study is 0.55. The alkali-activated cementitious composites were prepared in three separate stages. In the first stage, the precursor ingredients, i.e. binders (FA+GGBS+SF) and graded sand, were dry-mixed using a planetary mixer (Kenwood, Germany) for 2 min at 250 rpm. The additives (i.e. PVA fibres and attapulgite nano-clay) were slowly added to the dry-mixture and mixed for another 2 min at 250 rpm. The additives volume fractions have been chosen based on the previous studies conducted by   [25] and   [11]. The alkali activators i.e., sodium hydroxide and sodium silicate solutions with the fixed mass ratio of 2:1 were mixed at 700 rpm for 5 min, and then gradually added to dry-mixture and stirred for 2.5 min at 250 rpm, followed by mixing at 450 rpm until a homogeneous alkali-activated mixture was achieved.

Printing setup
Similar to our previous works [14,22], a gantry 3D printer equipped with a custom-designed auger extruder was utilized with a circular-shaped nozzle of size 20 mm, motion speed of 20 mm/s, and nozzle standoff of 10 mm. All the printable mix formulations were printed in a rectangular-shape and then cut to the 160 x 40 x 40 mm prisms (three samples for each composition) to keep it consistent with the casted samples.

Fresh properties assessment
Several test methods were used to obtain parameters which are interlinked and give evidence which explains the rheological properties of AAMs: (1) Where, FP is flow percentage, d Average is average spread-diameter of the paste in two perpendicular direction spread, and d 0 bottom cone diameter.
 The extrusion efficiency has been associated with an 'extrusion window', where a specific volume of material shows required workability to be extruded smoothly, without causing any discontinuity [26]. The extrusion efficiency test was conducted by printing a straight-lined filament of size 200mm in length. A filament was printed every 5 minutes rest-intervals from 0 minutes until the mix was no longer extrudable.
 The mini-slump test was also carried out to determine the workability or fluidity of the fresh alkali-activated composites as it is quantitatively related to the yield stress of fresh mixtures [27,28] Where, is shear stress, is yield stress, K is consistency, is shear rate, and n is flow index.

Mechanical property assessment
The flexural and compressive strength tests were conducted on the casted and printed prismatic samples with dimensions of 160 × 40 × 40mm 3 . The mechanical property tests were carried out after seven days of curing in which for the first 24 hours, the samples were placed in the oven at 60°C [29] and the remaining six days air-cured. These tests conformed to BS EN 196-1:2016 and were completed using an Instron 5960 Series Universal Testing System.

Microstructure analysis
A small cross-section of the printed samples was cut using broken portions of samples in flexural strength test to investigate the microstructure using SEM (Supra 35VP, Carl Zeiss, Germany) coupled with Energy dispersive spectroscopy (EDS) analyser (EDAX, U.S.A). Prior to the SEM analysis of each composition, a minimum number of ten specimens with the approximate size of 10 mm 3 , were gold-coated by means of Edwards S150B sputter coater to provide electrical conductivity on samples surface.

RapidAir measurements and micro-computed tomography
In order to assess the microstructural characteristics of 3D printed specimens, two image-based approaches were applied: (i) 2D standardized technique based on a linear traverse method and (ii) X-ray micro-computed tomography (micro-CT) method enabling to obtain a 3D volumetric images of concretes. Our previous study showed that both techniques can be successfully applied to evaluate the microstructural properties of cementitious composites with satisfactory correlation between obtained results [30]. Specimens dedicated for this measurement were rectangular specimens with an approximate area of 96 cm², obtained from vertical cross sections of specimens (Figure 3). Prior to testing, cured concrete samples were cut into 1 cm-thick slices and one face of the sample was thoroughly polished. The surfaces of the samples were then painted with a black marker and zinc (II) oxide paste to increase the contrast between the air voids and the matrices. Two specimens were tested twice in each case following the BS EN 480-11 standard. The surface in question had an area of 120×80 mm 2 . The RapidAir device enables to perform the air void analysis with a resolution of 1 to 2.5 µm. Two measurements of each specimen were performed and mean value was taken.
To assess the concrete characteristics along the height of specimen, three representative areas were defined: upper part, center part and lower part (Figure 3). X-ray micro-computed tomography (micro-CT) evaluations were performed on 20×20×20 mm 3 specimens cut out from the upper, middle, and lower part of each extraction zone of printed element. The used micro-CT device is SKYSCAN1172, ex-vivo at Seoul National University in South Korea. The anisotropy of the pore shape was confirmed using a probabilistic method, which is used to evaluate relative phase clustering size. Here, a probabilistic function, lineal-path function (L 2 ) [32,33], was used to investigate directional pore characteristics of the specimens. This function can be utilized to investigate and continuity of pores (L 2 ). Detailed descriptions of the used function can be found in [33,34], and general formulations are presented as follows (see Eq.4): Where r is the distance between the two points, while and are the angle between the test line and the z-axis, and the angle between the projections of the test line on the xy-plane, respectively. denotes the pore volume fraction. Using the L 2 function, the relative length along each direction can be compared. However, it is limited to be used for quantitative comparison because the feature of the L 2 function is a type of distribution. To overcome the limitation, the each graph was integrated, and the computed value was used as the index to compare the degree of anisotropy of pores in each specimen.

Buildability test
An ad hoc testing method was employed by printing 5500 g of alkali-activated fresh composite in a circular-shape printing path with a diameter of 150 mm. After the printing process, the quality of samples in terms of maximum height of printed objects, vertical distortion, and layer deformation was investigated visually. High-quality pictures of the printed samples were used to perform a visual-investigation of each mix's buildability and outer-surface finish.
ImageJ software was also employed to measure accurate reduction in bottom layers height and maximum printing height of each sample.

Effect of additives incorporation on fresh properties
As reported by several authors, e.g. [7,[35][36][37], the cement-based composites have the colloidal interaction and are considered as fully thixotropic, which is attributed to the presence of Van der Waals colloidal forces in the materials in the fresh state. This can lead to generating initial yield stress. However, because of the high-viscosity behaviour of potassium silicate chemical reagents within the alkali-activated cementitious composites, the hydrodynamic effect is presented in these materials [7]. Previously, attapulgite nano-clay [13] and PVA fiber [38] were incorporated to improve the thixotropic behaviour over the hydrodynamic effect in the alkali-activated mixtures. The attapulgite particle edges with the opposite electrical charge induce the electrical attraction between the particles to generate high yield stress at the resting stage. Incorporation of attapulgite with the fine particle size can form Van der Waals bonding within mixture by filling the gaps between binder components [25]. On the other hand, the electrostatic J o u r n a l P r e -p r o o f bonding can be formed due to intermolecular interactions between the PVA's hydroxyl group (-OH) and cementitious binders' surfaces, improving the thixotropic behaviour in the fresh state [39].
Following the methodology specified in BS EN 1015-3:1999, an initial flowability (i.e. 5 minutes) of 32% was recorded for the control sample (CS). As can be seen in Figure 5, the flowability of all PVA fibre-reinforced mixtures (i.e. P-0.25 and P-0.5) were reduced compared to the CS. The highest reduction of 26.5% was obtained by the 0.5% PVA sample.
The presence of oxygen-containing functional groups (i.e. hydroxyl group -OH) on the outer surface of PVA fibers absorbs free water of the fresh mixture. This results in increased friction between the fibres and mixture's solid components making it stiffer compared to control sample [38,40]. As reported by Zhu et al. (2018), fiber type and fiber dosage are the most critical parameters affecting the flowability. They also reported that the PVA fiber incorporation accelerates the reaction kinetics of fresh mixtures at early ages, which was evidenced by increasing the reaction heat release rate during the activation process [41]. Moreover, the long PVA fibers (i.e. 12 mm) with the flexible nature can easily twist and ball together, forming an agglomerated aggregate within the mixture during the mixing procedure, which leads to a decrease in the flowability [42].
The inclusion of attapulgite nano-clay reduces the flowability compared to the CS (see Figure 5). However, the attapulgite incorporation shows a more considerable influence on decreasing the overall flowability than PVA fibre incorporation when compared to the control sample. The results revealed that the initial flowability was reduced by The yield shear stress of each fresh mixture with and without additives was determined. It is commonly known that the yield stress is related to the slump values (i.e. slump height and relative slump value) of an alkali-activated composite [28]. In general, higher yield stress and lower mini-slump values are found to be the required conditions for a printable mix formulation [10]. Therefore, the mini-slump test was also carried out for the analysis of the flow behaviour of the fresh mixtures to check workability and shape stability of the mix. The results show that the yield shear stress of alkali-activated mixture increased from 1216 Pa for the control sample to 1361 Pa and 1522 Pa for P-0.25 and P-0.5, respectively. As observed in Figure 6a, b and Table 3 The author's previous work and several other researches have confirmed the direct relation between extrusion window and the other fresh properties of alkali-activated composites [14,20,43]. Moreover, the results obtained from the extrusion window test are known as the most pertinent and relevant parameter among other fresh property factors as it exactly imitate the 3D printing process of fresh AAMs [44]. This is due to the fact that all the other properties of a mix are insignificant if the mix cannot be extruded [45]. Therefore, the impact of additive   The shape retention of all the samples were measured, and the results are shown in Figure 7. Among all samples, the mixture with 0.5 wt.-% of attapulgite (i.e. A-0.5) was excluded since it was not printable, as previously confirmed by the considerably low fresh property efficiency, including flowability, yield stress, slump values, and extrusion window, that was out of the employed operating extrusion system's limit. For the printable mixtures, the results of the shape J o u r n a l P r e -p r o o f Journal Pre-proof retention were compared in terms of first layer's height (H 1 ) and the height difference between the first and the sixth printed layer (H dif ).
The results indicate that the H 1 and H dif values for the sample with 1 wt.-% attapulgite particles (i.e. A-1) was improved and reached 6.3 mm and 5.9 mm, compared to the values obtained for CS (i.e. 5 mm and 6.3 mm). The increased dosage of PVA fibres also improved the shape retention of the layers. The H 1 value of P-0.25 and P-0.5 was increased by 20% and 90%, respectively, which was higher than that of CS. The H dif , on the other hand, was decreased

Effect of additives incorporation on mechanical properties
The mechanical performance of alkali-activated composites has been investigated to identify the ideal printable mixes to be used as feasible feedstock for the eventual upscaling process.  Before the mechanical property tests, the samples' density, both printed and conventionally casted, were measured (shown in Figure 8a). For samples with PVA fibres, the density of conventionally casted samples increased from 2.19 g/cm 3 for P-0.25 to 2.24 g/cm 3 for P-0.5, which are higher than the value registered for casted CS (i.e. 2.14 g/cm 3 ). On the other hand, the density of the printed samples also gradually increased from 2.33 g/cm 3 for CS to 2.37 g/cm 3 and 2.42 g/cm 3 for P-0.25 and P-0.5, respectively. The gradual densification by increasing the fibre content, both for printed and casted samples, is attributed to the adequate compatibility between the fibres and matrices. Further to this, a slightly higher density value was obtained for printed samples when comparing to the corresponding casted samples. This provides a confirmation that adequate compaction and integration of each extruded layer occurs during the extrusion process. Hence, the 3D printing process has the potential of minimising the formation of voids, which could be detrimental to the mechanical properties of the samples.
The incorporation of Attapulgite (i.e. A-1) caused a slight increase of 4% and 3% in density of casted and printed samples, respectively, compared to that of CS. The results are aligned with the results obtained from other studies [46], which provides further evidence that the addition of nano-clay leads to an increase in the density of alkaliactivated mix. This suggests that the enhanced density recorded for the A-1 mix was attributed to the effective dispersion of these particles resulting in a less porous mix. For the hybrid sample (i.e. P-0.25+A-1) unexpected results were obtained, where the density of the casted sample was slightly higher (about 2.5%) than the printed one. This has to be associated with the high yield stress of the aforementioned mix, among the printable mixes, in the fresh state (see section 3.1), which led to inconsistent printing. This phenomenon generates trapped air babbles within the samples that reduces the density of printed sample compared to casted one.
The mechanical property results (Figure 8b and 16%, respectively, to that of the printed control sample. However, the incorporation of more PVA fibres (i.e. P-0.5) was shown to be less effective as the flexural and compressive strength increased by 18% and 1% compared to the printed control sample.
The enhanced flexural and compressive strength is attributed to three major mechanisms. First, the improved integration of the fibres within the alkali-activated matrices is reinforced by the extrusion process [47]. Figure 9 indicates that the fibre was well embedded in the matrices as large amounts of activation products are observed on the fibre surface [48]. This is further supported by the use of EDS, which identified similar elements of the matrix presented on the fibre surface. As reported by Xu et al. (2017), PVA fibres react with alkali-activated composites, a transitional layer composed of activation products are produced at the interface between fibre and matrices, which make strong interfacial bonding between the matrices and fibres. Hence, PVA fibres can be more effectively distributed and connected within the cementitious paste [49]. The second mechanisms are the crack-bridging and crack blocking phenomena induced by the incorporation of fibres. As shown in Figure 10, after the micro-cracks initiation, the micro-cracks tend to propagate within the cementitious paste to form macro-cracks. However, PVA fibres that have crossed the crack and are anchored in the matrices on both sides, effectively 'sewing' the two sides together and preventing the crack from widening [50]. This results in increased flexural strength of the fiberreinforced samples compared to both printed and casted CS. The third mechanism is the fibre alignment generated during the extrusion process within the alkali-activated paste [12,51]. In this hypothesis, during the extrusion process, the nozzle enables fibre alignment along the printing path, which can lead to an increase in flexural strength of the printed structures in a perpendicular load direction [52]. Several studies have demonstrated the feasibility of developing a 3D-printable fibre-reinforced cementitious composite and reported the positive impact of fibre alignment on the samples' mechanical performance [53][54][55]. attributed to the induced thixotropic behaviour of a mix through the addition of nano-clay generated by the rapture of the inter-particle links within the matrix under the shear-force of the extrusion process. After the printing process, due to the reversible behaviour of the thixotropic effect, the inter-particle links are rebuilt more robustly than before, resulting in the increased mechanical performance of the alkali-activated composite [16,25]. Another possible explanation, as demonstrated in Figure 11b, is that nano-clay incorporation accelerates the alkali-reaction process, leading to a higher amount of activation products and lower unreacted fly ash particles [46,56,57]. Moreover, the effective dispersion of nano-clay particles within the cementitious matrices leads to an improved porosity structure (see Table 5), thus, producing a more compact matrix structure compared to that of CS (see Figure 11a). Assaedi et al. (2016), also reported that the addition of up to 2% Cloisite 30B nano-clay particles to a geopolymer mix induced a remarkable increase in both flexural and compressive strength [46]. For P-0.25+A-1 sample, the flexural strength was increased by 29% however the compressive strength slightly decreased (i.e. 2%) compared to that of printed CS. As can be seen in Figure 8b and c, both flexural and compressive strength of P-0.25+A-1 are relatively lower than the values registered for individual A-1 and P-0.25. This is attributed to their respective poor fresh properties as previously discussed. The microstructure analysis further proved that the adhesion between the fibers and matrices decreased due to the lower workability of the mixture in a fresh state. As it is evident in Figure 11c and d, some unreacted fly ash and nano-clay particles are present between the fiber and matrices interface, which increases the stress concentration, hence decreasing mechanical performance.

J o u r n a l P r e -p r o o f
Journal Pre-proof  Table 4 and Figure 12 present the details of the Rapid Air void measurements obtained from vertical cross sections of printed specimens except P-0.5 that is eliminated due to poor mechanical performance compared to CS. Evaluation of the chord length frequency (Figure 12) confirms that specimens mostly consists fine voids. However, a significant difference between chord length frequency curves can be distinguished between the control sample (CS) and modified specimens. CS specimen exhibit significantly broader void distribution with higher contribution of coarser pores, while modified specimens possess higher quantity of fine voids. The contribution of so-called micro air voids (<300 microns) to the total air content of specimens CS, P-0.25, A-1 and P-0.25+A-1 is 53 %, 80 %, 83 % and 70 %, respectively. The total air content of specimens is relatively comparable, however, the lowest air content was reported for specimen 0.25-PVA which is 11 % lower when compared to CS specimen. The refinement of the void structures in specimens is also confirmed by other parameters such as average chord length and spacing factor, which confirms the decrement of the average void size as well as overall refinement of the void structure. This could be responsible for the observed strength improvements in modified specimens (Figure 8).  In each additive case, three different positions were selected at the top, middle, and bottom part of the specimens, and a total of 12 cases were investigated (Figure 3). Since the main target of the micro-CT measurement is the pore characteristics, the pore structure of each specimen was visualized as shown in Figure 13. In this figure, the gray part denotes the pores inside the specimens, and it can be confirmed that the pores within the specimens show directional dependency, although the degree of anisotropy varies according to the positions and additive types.

Effect of additives incorporation on 2D and 3D structure
Porosities of specimens measured on different height of specimen exhibits variation and in generally higher quantities of voids can be reported in lower and middle sections of specimens (Table 5). However, on average the total porosities from 3 tested heights are comparable to that of total air content value determined with RapidAir ( Table 4). The slight discrepancy between the results is attributed to the different measuring ranges of both techniques.
For qualitative and quantitative investigation, the lineal-path function (L 2 ) was adopted in this study. Figure 14 shows the results of the L 2 function for the cases in Figure 13. As can be seen in the figures, all cases show anisotropic characteristics in their pore structures. This indicates that the pores within the specimens are anisotropic, particularly J o u r n a l P r e -p r o o f in the relatively large pores. In addition to the pore anisotropy, the porosity in each specimen at different locations can be found from the function at r/D is equal to zero.
For a quantitative investigation, the L 2 functions were integrated over the r/D, and the area in each function was computed. Table 5 shows the quantitative results of the L 2 function and the porosity in each case. In this table, the integrated value denotes the degree of pore size along the direction, and the directional difference between the values in each specimen represents the degree of pore anisotropy (anisotropy ratio). The higher anisotropy ratio indicates that the specimen at that location contains more anisotropic pores, which can affect the stability of printed samples. Specimens CS and the P-0.25+A-1 show the highest anisotropic indexes, which indicate instability of the pores (Figure 14). Among the studied samples, the A-1 specimen shows the most stable and small anisotropy ratio that guarantees the quality of the printed material. These observations are in line with the results from the effect of additive incorporation on fresh properties and buildability investigations. The A-1 sample has satisfactory shape retention characteristics as well as the best buildability among the tested specimens, which will be discussed in the next section. It can be concluded that A-1 is the most homogenous mixture over the height between all the investigated samples. The lowest pores anisotropy as well as overall refinement of the microstructure in A-1 can also be correlated with the best mechanical performance.
J o u r n a l P r e -p r o o f

Buildability test
The buildability performance of each printable mix, selected in previous section (i.e. section 3.3), is displayed in Moreover, due to the compressive load induced by the deposition of top layers, the reduction in bottom layers height was clearly observed for all mixtures. The maximum printing height of each object after printing was also measured.
The results indicated that by printing the constant quantity of fresh materials (i.e. 5500 g), the incorporation of additives led to an increase in the maximum height of printed cylinders from 117 mm for the control sample (see Figure 15a) to 125 mm (see Figure 15b), 120 mm (see Figure 15c) and 138 mm (see Figure 15d) for A-1, P-0.25, and P-0.25+A-1 samples, respectively. As it can be seen in Table 3

Conclusions
The main objective of this study was to investigate the impact of two additives (i.e. PVA fibres and attapulgite nanoclay) on the fresh properties of alkali-activated materials in terms of flowability, rheology, mini slump, extrusion window, and shape retention. Further, the printed samples' mechanical properties (i.e. density, flexural and compressive strength) were evaluated and compared to the corresponding casted samples. Finally, the reinforcing mechanisms of additives were elucidated by the correlations between the fresh properties, buildability, and mechanical properties as well as the 2D and 3D microstructure analysis (i.e. SEM, RapidAir measurement, and µ-CT scan) of the specimens. The following conclusive statement can be drawn from this study:  The incorporation of both attapulgite and PVA fibres in specific dosages can lead to improvements in the fresh properties of alkali-activated mix. Both additives at specific dosages induce an adequate decrease in J o u r n a l P r e -p r o o f flow ability, slump height along with a sufficient increase in shape retention, yield shear stress, and extrusion window.
 Amongst the investigate AAM mixes, P-0.25, A-1, and P-0.25+A-1 meet the basic buildability requirements needed for 3D printing. The inclusion of additives slightly increased the maximum height of printed objects compared to the control sample.
 P-0.25 and A-1 were found to be the best performing sample in terms of 3D printing and mechanical performance. For P-0.25, the flexural and compressive strength increased by 24% and 10 %, respectively, compared to the printed control sample. Samples with 1% attapulgite (i.e. A-1) also experienced increased flexural and compressive strength by 43% and 20%, respectively, compared to the printed control sample.
This increase in mechanical performance is most likely attributed to good dispersion of nano-clay particles and fibers crack-bridging mechanisms throughout the composite.
 The microstructure analysis clearly showed the effectiveness of attapulgite and PVA fibres incorporation in filling the voids and crack bridging mechanisms.
 The study found that both RapidAir and micro-CT techniques can be successfully used as complimentary measuring approaches towards understanding the microstructural characteristics of the 3D printed elements.
Noticeable pores anisotropy has been over the height in selected specimens, however, appropriate mixture design helps to overcome this obstacle.
 The incorporation of 1 wt.-% attapulgite clay exhibits the most satisfactory microstructural characteristics with the lowest pores anisotropy as well as mixture homogeneity. This results in the best mechanical performance and shape retention followed by buildability.
The promising breakthroughs of this study on the positive impacts of fibre and nano-clay additives in 3D printing of alkali-activated materials can lead to potential applications for large-scale construction. Future studies need focus on a more comprehensive mix design formulations to yield conclusive outcomes with statistical analysis and verification.
Moreover, the reaction mechanisms and functionalities of the additives in determination of various properties of 3D printed AAMs should be elaborated.