Rheology Control and 3D Concrete Printing with Fly ash-based Aqueous Nano-silica Enhanced Alkali-activated Binders

E�cient alkali-activated binder pastes of slag and �y ash-based aqueous nano-silica (FABANS) that provide enhanced strength are used for developing extrusion-based 3D printing applications. The binder pastes of slag made with FABANS are not suitable for printing due to low yield stress and insu�cient thixotropy. Rheology control for enhancing the yield stress and thixotropic buildup is evaluated using bentonite clay and carboxymethyl cellulose (CMC). There is a synergistic enhancement in yield stress and thixotropic buildup provided by the combined use of bentonite and CMC that provides improved printability and buildability. Very rapid increase in yield stress with excess clay content in the presence of CMC, however, produces choking of �ow and printability loss. The proportion of CMC and clay that provides the required thixotropic buildup for buildability is established. Printability and buildability of concrete mixture made with binder paste of FABANS with the proportion of rheology modi�ers is demonstrated.


Introduction
Concrete 3D printing has been gaining much importance in the past few years due to the advantages of automated free-form construction without the use formwork.The process is swift, cost effective and requires relatively less labor.One of the challenges with the implementation is the production printable material.A material suitable for 3D printing would be expected to hold certain properties such as printability and buildability.The fundamental requirements of the material to satisfy the requirements of printability vary with the type of pump and printer, the delivery system, size and shape of nozzle, standoff distance etc. Due to these constraints, the basic material which is proportioned for strength is often not ideal for 3D Concrete Printing (3DCP).The rheology of the mix needs to be modi ed so as to possess a low viscosity, a high yield stress and adequate thixotropy [1][2][3].Since the possibility of a material manufactured for strength to possess these properties is very low, it often needs the addition of rheology modi ers.The rheology modi ers that have been signi cantly reported were Kaolinite clay, Attapulgite clay, nano clay, bentonite clay, lime powder, plasticizers, Viscosity Modifying Agents (VMA) and methyl cellulose.Adding the required rheology modi ers alone or in combinations in adequate quantities can alter the rheology as per the requirements of printable material.
Research on 3DCP has mostly been with the use of cement-based binders.In cement-based binders, y ash and slag have also been used as rheology modi ers.3DCP has also been made possible using alkaliactivated binders (AAB).External silica is often added in the form of nano silica (NS) or sodium silicate (SS) to enhance the strength of AABs.In addition, different forms of silica extracted from other sources such as rice husk ash (RHA), sugarcane straw ash (SCSA), lead bearing sludge (LBS), Olivine mineral, waste glass (WG), bagasse ash (BA) have also been used in producing AABs [5][6][7][8][9].Rheology of alkaliactivated mixtures is complex and is in uenced by the composition of the activating solution.The silica content and the molar ratio of silica to sodium (molar ratio) in the activating solution in uence the rheological response of the AAB.The mix with a lower silica content exhibit a yield-type behavior under applied strain rate.At lower silica modulus and higher silica dosage, the mix is found to exhibit a Maxwell-type of behavior [4].Maxwell ow under a constant shear rate is characterized by a viscous dominated response and results in a continuous deformation at a constant applied stress [5,6].Mixes with Maxwell-ow type behavior, did not have buildability in the printed structure.The shape of the bottom layers in the printed structures continuously deform under the weight of the additional layers.
Printing with AABs require rheology control for improving the yield stress and thixotropy to achieve buildability.Mixtures that have a Maxwell ow behavior can only be printed with the use of additives that help produce a yield type response.In the previous studies, rheological control was achieved with the use of nano clays.Mostly Montmorillonite [7,8] and Attapulgite clays [9,10] have been used.While nano clays provide rheology control, their effectiveness depends on the dispersion.Additionally, nano-clays signi cantly enhance the cost of the mixtures.
Previous studies have shown that in y ash-slag AABs, the property development is determined by the slag reaction while the y ash provides reactive silica.The primary reaction product in the y ash-slag AAB produced by slag reaction is a calcium silicate hydrate gel with aluminum substitution (CASH).The y ash contributes additional silica for enrichment of CASH.Direct silica extraction from y ash has also been attempted and typically involve processes of acid dissolution, electrolysis and centrifuging which however involve high energy [11].In a recent study, the y ash based nano silica (FABANS) enriched activator has been extracted in a room temperature process.Slag activated with FABANS activator produced enhanced strength gain and a higher ultimate strength.FABANS allows effective utilization of silica in the y ash in promoting the CASH formation.The nano silica in the FABANS activator, however, has an in uence on the rheological properties of the AAB binder paste.The focus of this study is to develop printable alkali activated slag mixture made with FABANS activator.The rheological behavior of the FABANS enhanced activator is evaluated and rheology modi cations required for extrusion-based 3D printing are assessed.The requirement of yield stress and thixotropy for 3DCP is provided with the use of additives like clay and CMC.The use of Bentonite is experimented with in this paper since it is cost effective.The e ciency of using Bentonite clays in enhancing the yield stress and thixotropic buildup suitable for 3DCP is evaluated.The reduction in the requirement of clay by utilizing synergy with CMC is explored.

Source materials
Alkali activated binders were prepared using y ash and slag as the binder materials.Class F or Siliceous, low-calcium y ash found in the Indian sub-continent procured from local thermal power station con rming to IS 3812 − 2003 [12] was used.Ground granulated blast furnace slag (referred to as slag hereafter) procured from JSW suppliers conforming to IS 12089 − 1987 [13] was used as the other source material.Bentonite clay and CMC were used as rheology modi ers in the alkali activated binder mixtures.Previous studies reported the use of Kaolinite clay [5,6], Attapulgite clay [9] and Nano clays [14].It has been observed that Attapulgite clay and Nano clays are highly effective in enhancing the yield stress and thixotropy.However, these materials are very sensitive to dispersion and are expensive.Kaolinite was found to be effective, but a signi cantly large quantity of replacement for the required enhancement [5].
The use of clay has also been reported to reduce the phase separation within the mixture under pressure [6, 15,16].In the present study, Bentonite clay, is used to provide a low-cost alternative to nano clays.
Further, with a higher content of Montmorillonite, Bentonite clay is expected to provide enhanced yield stress and thixotropy for rheology control at a smaller dosage compared to Kaolinite.A commercially available CMC acquired from Quali-tech chemicals was used in the present study.Laboratory grade NaOH was used as the activator.The oxide compositions of the source materials and Bentonite clay as measured using X-ray uorescence spectroscopy are presented in Tables 1 and 2, respectively.The particle size distributions (PSD) of Fly ash, Slag and Bentonite clay (referred to as clay hereafter) were measured using Iso-propyl alcohol dispersant in a Microtac S3500 Particle Size Analyzer are reported in Fig. 1.The PSD shows that all the solid materials have a comparable size distribution and thus, the addition of clay is not as a ller material.

Formulation of alkali activated mixtures and 3D Printing
The Control alkali-activated y ash-slag mixture was made by using a binder consisting of y ash and slag in a mass proportion of 1:1.The activation was provided using 3M NaOH solution that was mixed with the binder material in the mass proportion of 1.0: 0.4.AAFS mixtures were also prepared using FABANS enhanced activating solutions.An aqueous solution was prepared by pre-leaching all the y ash in 8M NaOH for 24 hours.The pre-leaching of y ash results in the dissolution of the reactive precursors from the y ash into the alkaline solution [17].Previous studies established that the dissolution of lowcalcium y ash results in the release of nano silica and alumina from the y ash [18].Slag was combined with the aqueous solution that included the NaOH, y ash leachate and its residue.Additional water was added to reduce the molarity of NaOH to 3M in the nal paste.The AAFS mixture made with FABANS is referred to as FAB and it has the same binder composition as the Control mixture except the availability of nano silica in the aqueous solution results in enhanced e ciency with a strength enhancement [19].
The 28-day strengths of the cast Control and FAB mixtures were 26 MPa and 36 MPa, respectively.
The rheology of the FAB mixture was evaluated, and the rheology modi cation provided by the additives were assessed.A total of nine trials were performed varying the quantities and the proportions of rheology modi ers.The FAB mixture without any additional rheology modi er is labelled as M1.To understand the in uence of each rheology modi er on the mix, Bentonite clay was varied in proportions of 0%, 1.25%, 1.8% and 2.4% by mass of total binder (Fly ash + Slag).Similarly, CMC was varied in proportions of 0%, 0.25%, and 0.375% by weight of total binder.The mix proportions of all the nine trials are listed in Table 3.

Rheological measurements
The AR-G2™ strain-controlled rheometer tted with a cup and vane shear measurement system was used to perform the rheological measurements on the FAB mixtures with the different additives in different proportions.The FAB pastes were prepared in a paddle mixer.The mixing protocol was performed as mentioned in [20] at 300 rpm for 2.5 minutes and placed in the cup of the measurement system within 60 s.A pre-shear of 20 s − 1 was applied for 5 minutes followed by equilibration of 2 minutes before starting the measurement.The entire system's temperature was kept constant at 25 0 C by means of a Peltier cooling system.
Rheological measurements were performed for determining yield stress, viscosity, and thixotropic buildup.The vane shear measurement system was used to reduce the in uence of wall-slip, which is commonly observed in water-based particulate suspensions [21,22].To measure the yield stress and to understand the basic ow behavior, a constant strain rate test was performed with a prescribed angular velocity of 0.1 rad/s.The peak resistance to increasing angular strain was used to determine the yield stress of the mixture [23].The viscosity was determined by tting the Bingham model to the decreasing strain response measured in hysteresis test.The hysteresis test was performed by linearly ramping up the shear rate from 0 to 40/s and then linearly decreasing the strain rate from 40 to 0/s at a ramp rate of 0.33/s 2 .Thixotropic buildup was determined from the increase in yield stress with time.The yield stress of the mixture was measured 10 minutes, 20 minutes, and 30 minutes after performing pre-shear and equilibration.The increase in yield stress with time provides a measure of the structure build-up within the mixture.

Printing requirements
The printing performance of the FAB mixtures was evaluated using an extrusion-based ram-type extruder.
Binder pastes have been previously pre-quali ed using the ram-type, extrusion-based printer, which has also been utilized to assess the printability and buildability of the paste [6,24].The print performance provides insights into the behavior of paste under induced shear ow under con nement like in concrete during pumping and extrusion [25].The printability of paste is affected by the segregation of water under pressure.The thixotropic build-up in the paste during shearing regulates the buildability of the concrete mixture [16].The piston-type printer can assess both qualities and is used to evaluate the relative in uence of the additives on enhancing speci c aspects of printability and buildability.
The ram in the extruder was 76 mm in diameter and the paste was pressurized in a barrel of 300 mm length.The barrel was tted with a 10 mm nozzle.The printer setup using the ram-type extruder consisted of microprocessor-controlled stepper motors for horizontal planar axes movement control.The vertical movement control was provided by an independent stepper motor and the depended on the planar contour and the print speed of the extruder.The G-code regulating the print speed, extrusion rate and the coordinated movements in the planar and perpendicular axes was transferred directly to the microprocessor controller.
Printability was classi ed as the ability of the mix to be extrudable as well as have a constant width over the entire lament without any signi cant cracking.To determine this, a lament of 10 cm length and 0.1 cm width was printed and checked for width control along with cracking.The mixtures that showed a deviation in width of more than 0.02 cm or showed an interrupted pulsated ow (cracking) were considered not printable.Buildability is de ned as the ability of the mix to sustain its weight and allow for build-up of layers.This property was determined by measuring the height of the specimen and comparing it with the actual height.If the measured specimen differs by more than 10%, then the mixture was not considered to be buildable.

Rheology of Control mix
The constant strain rate response obtained from the M1 FAB mixture is shown in Fig. 2 (a).The response obtained from the AAF paste made by directly blending the y ash and slag with the activating solution, referred to as 'Control', is also shown in the gure for reference.The constant strain rate response of the Control mixture exhibits a yield-type response.There is initially an increase in the shear resistance to applied strain, which reaches a maximum and then with increasing strain the resistance decreases to a constant value.The peak stress value from the stress -time response is the dynamic yield stress.The resistance to the increasing strain measured from the particulate suspension is derived from the percolated network of particles within the uid medium.The yield stress is the peak resistance of the internal structure with a long-range order that is disrupted inducing ow with smaller agglomerates of particles.The yield stress also indicates the stress at which the material starts to ow [22,26,27].
The MI FAB mixture prepared using the FABANS enhanced activating solution had a lower yield stress than the control paste.The presence of dissolved silica from the alkaline leaching of y ash in uences the rheological behaviors of the AAF paste.The silica from the leaching is known to be present in a nano form [19].The FABANS is a form of nano silica which is extracted from y ash in a room temperature process.For a comparative evaluation, mixtures were also prepared with a commercially available nano silica at 1.0 and 3.3% by mass of the binder.The nano silica (NS) used for comparison was a clear solution of colloidal nano silica particles with an average particle size of 5-7 nm that were dispersed in water at 25% concentration.The constant strain rate response of the Control mixture with 1 and 3.3% dissolved nano silica by mass of the binder are plotted in the gure.There is initially a decrease in the yield stress with the NS content in the activating solution.The decrease in the yield stress with NS addition in the activating solution has been reported previously and is due to the in uence of the colloidal nano-silica on reducing the inter-particle a nity indicated by the zeta potential [4,20].The constant strain rate response obtained from the M1 FAB mixture containing FABANS clearly lies between the 1 and 3.3% NS cases as shown in Fig. 2 (a).
The storage modulus measured as a function of time after mixing is plotted in Fig. 2 (b).The increase in the storage modulus is indicative of the internal structure buildup that leads to an increase in the rigidity.
The control mixture exhibits a very rapid buildup of elastic stiffness with time.This indicates a very rapid internal structuration, which occurs on the order or 10 minutes that enhances the storage modulus with time.The increase in storage modulus is signi cantly slower in the FABANS mix.The in uence of silica in the activating solution, therefore, signi cantly reduces the thixotropic buildup within the mixture.The in uence of the FABANS on rheology is identical to an activating solution containing dissolved nano silica.Dissolved silica is known to produce a retardation of the early kinetics of reaction in y ash slag blend.

Rheology control with additives
The enhancement in the yield and thixotropic buildup of the FAB mixture was evaluated with the use of additives, bentonite clay and CMC.The constant strain rate test response of the FAB mixture with variation in Bentonite clay and CMC are shown in Fig. 3 (a) and (b).The mixtures selected in Fig. 3 (a) show the in uence of clay addition, while Fig. 3 (b) shows the in uence of CMC.The FAB mixtures M2 and M3 contain clay at 1.8% and CMC at 0.25% by mass of the binder, respectively.The measured yield stress from all the mixtures is listed in Table 3.The addition of clay and CMC increase the yield stress of the FAB mixture signi cantly.The shear resistance of the M1 mixture is insigni cant compared to the other mixtures containing CMC and clay.The combined use of clay and CMC appears to enhance the yield stress.Among the mixes M3, M4, M5 and M6, that contain 0.25% CMC while the clay content varied from 0 to 2.4%, it can be observed that the yield stress increases with clay dosage.The M4 mixture, which contains lower clay dosage than the M2 mixture has a signi cantly higher peak resistance.
The mixtures M2, M5, and M7 that contain clay at 1.8% dosage while the CMC content varies from 0 to 0.375%.also exhibit an increase in the peak resistance with CMC content.There is however a distinction in the evolution of stress resistance with increasing strain with increasing CMC content at a xed clay content.The shear stress in M7 increases with strain and the resistance evolves continuously to reach a constant value asymptotically.The continuously evolving shear resistance under an applied constant strain rate, that gradually attains a constant value is typical of Maxwell ow behavior.A sharp peak is absent in the constant strain rate response.The peak load in the Maxwell ow response is identi ed with an apparent yield stress.The shear resistance in a Maxwell ow response attains a constant value after the apparent yield stress.The continuously evolving resistance is in uenced by the viscoelastic deformation of the uid which is enhanced by the addition of CMC.The continuous deformation of the uid between the particles of the suspension results in a continuous deformation under applied stress.The addition of CMC therefore induces a viscous response from the activating solution that transforms the basic shear resistance derived from the suspension of y ash and slag particles from a yield-type to Maxwell-ow response.The viscosity measured from the FAB mixture with the different CMC and clay additions is listed in Table 3.The viscosity increases with increases in Clay and CMC contents but the increase in viscosity is more signi cant in case of CMC when compared to clay.The viscosity scales sensitively with the CMC content.
The mix with only CMC and no Clay (M3) showed a very high increase in viscosity when compared to Control mix (M1) when compared to the mixture with only Clay and no CMC (M2).There is also a synergy in the viscosity increase obtained with the combination of CMC and clay.In the presence of CMC there is a large scaleup in the viscosity with addition of clay.The CMC has been known to provide a thickening effect of clay suspension [16], which is evident in the observed synergistic increase in viscosity.
The in uence of clay can be assessed in terms of increasing yield stress without in uencing the viscosity (Table 3).The CMC increases the viscosity signi cantly while there is also an increase in the yield stress.The combined in uence of CMC and clay is to enhance the peak ow resistance and the viscosity.The viscosity enhancement with the combined use of CMC and clay results in transforming the response of mixture from a yield-type to a Maxwell ow behavior when the proportion of CMC is increased relative to clay.A general observation pertains to the shear resistance as a proportion of clay and CMC.In mixtures where the proportion of clay was increased relative to CMC, al low clay contents (M4), a viscous type response is obtained.M4 shows an approach that approaches a Maxwell-type behavior which is caused due to the viscous dominance of the activating solution due to a larger proportion of CMC relative to clay.On increasing the proportion of clay in M5 a clear yield type response is obtained.Further increasing the clay content produces a signi cant increase in viscosity and the constant strain rate response in M6 indicates a transition to Maxwell ow.Increasing the proportion of CMC relative to clay as in M7 produces a Maxwell ow.The proportion of CMC and clay therefore is important in controlling the viscous response while ensuring a yield stress improvement.
The increase in yield stress with time brought about by the variations in bentonite clay and CMC is shown in Fig. 4. The increase of the yield stress measured in the M1 mixture provides a reference to evaluate the enhancement provided by the other additives.The time frame of printing has been marked in the gure which indicates the variation in yield stress during the print.Adding clay or CMC by itself does not result in a large increase in yield stress with time.The increase in the internal resistance to ow with time seems to scale with the clay content in the mixture in the presence of CMC.There is a clear enhancement in the thixotropy provided by combinations of CMC and clay when compared with the use of only clay or CMC.
In the mixtures with a xed CMC content, there is an increase in the yield stress buildup with increasing clay content.Comparing M4, M5, and M6, the most rapid increase occurred in M6.In FAB M6 mixture there is also an abnormal increase in yield stress after 20 minutes.Similar trends are observed on increasing the proportion of CMC relative to clay.

Printing
of all the FAB mixtures in extrusion-based 3D printing are presented in Table 3.The baseline mixture, M1, was by itself not printable.It owed under gravity and had a very high spread.The introduction of bentonite and CMC resulted in speci c changes in the print behavior of the M1 mixture.The images of the fresh prints are shown in Fig. 5.With clay addition, Mix M2 exhibited a large spread but had a better shape retention when compared to M1.The bottom layers of M3 exhibited a high spread with continued deformation under the weight of additional layers.The inclusion of CMC in the absence of bentonite had no signi cant in uence on the specimen though it performed better than M1.M4 through M7 contained both bentonite and CMC.There was an improvement in buildability exhibited with the inclusion of both clay and CMC.With increasing clay content, there was improved shape retention and buildability.The layer de nitions improved and the difference with target height decreased from M4, M5 and M6.Mixture M6, which had very large clay content, had the best shape retention during printing the bottom layers but was not extrudable after a few layers.The ow choked at the nozzle after the ninth layer.Pressure of extrusion increased signi cantly, and the ow seized.
In mixtures where the clay content was kept xed while CMC was increased, Mixture M8 with a large CMC content, exhibited good shape retention in the beginning but the layers heaved after the deposition of a few layers.M5 was the best mix with optimal dosages of Bentonite clay and CMC.To understand the in uence of excess clay and excess CMC in the absence of the other.M8 with excess CMC and M9 with excess clay were printed.M8 had bad buildability and M9 had poor shape retention.
Absence of clay in the mixtures resulted in poor shape retention with the addition of layers, which indicates inadequate thixotropic buildup within the mixture.Even mixes with lower dosages of Clay showed signi cant improvement in buildability, and this property got enhanced with addition of CMC.The addition of CMC to clay by about 0.5% (M-5) had a better in uence than increasing the clay content from 1.8% (M-2) to 5% (M-10).A similar study was performed by increasing the CMC content from 0.125% (M-3) to 0.5% (M-9).However, the mix with clay and CMC had better performance than the other two.This shows that a synergy needs to be formed between Bentonite and CMC in improving the yield stress and thixotropic yield stress buildup with time.

Discussion
After hardening, the printed shapes were cut along the cross-sections for evaluating the layer structure and buildability of the mixtures.In mixes that exhibited instability, the printing was stopped one layer before the onset of instability and allowed to harden.The cross-sections in the hardened state are shown in Fig. 6.M2 and M3 did not have printability due to lack of shape retention within the layer structure.Under the load of the additional layers, there was a large spread in the bottommost layers and a collapse in the structure which is classi ed as a plastic collapse [14].Improvement in shape stability within the layer structure can be seen with M4 compared to M2 and M3.M4 exhibited an instability in the form of continued accumulated lateral displacement of a middle layer with added layers, which has been classi ed as buckling instability [9,28].M4 also exhibits a spread in the base.The base layers continued to spread under the weight of additional layers.Mixtures M5 and M6 exhibited printability and buildability.The layer deformations were limited, and shape was retained.While Mixture M6 exhibited superior shape retention and buildability, the ow was choked after a few layers.Mixture M7 exhibited an instability associated with localized deformation within the printed structure.With additional layers, there is continuous deformation resulting in increasing localization that produced the instability.The print performance can be related to the rheology of the paste mixtures.The printability, and buildability can be related to the fundamental rheological behavior of yield and Maxwell ow responses and thixotropy.The shape retention improved with an increase in the yield stress.Mixtures M5 and M6 with high yield stress and thixotropic buildup from rheological measurements exhibited good shape retention of bottom layers under the weight of increasing number of layers.However, the very large increase in yield stress in M6, within the duration of printing, brought about by excess clay resulted in a choked ow.Very large yield stress has been shown to result in ow discontinuities and a choking of the ow [29].The importance of the ow behavior determined in constant strain rate rheological measurement is illustrated considering M7, which had a high resistance to shear-induced ow.However, the shear stress under increasing applied strain of M7 was a Maxwell ow response.The Maxwell-ow type response has previously been shown to result in a continuous deformation produced by viscous dominated response of the uid [23].There is a continuous distortion of shape in Mixture M7 though it had a very high resistance to ow, indicated by the apparent yield stress.The localization of the spread is produced by the continuous deformation under applied stress exhibited by the Maxwell behavior of the mix.The mixtures with Maxwell behavior have been shown to exhibit continuous deformation within the layers of the printed structure [16].Mixtures with yield stress response, however exhibited shape stability with printing.
The dissolved silica in the activating solution reduces the zeta potential and increases the dispersion of the particles in the medium [4].The presence of silica produces a continuous relaxation of the uid under applied stress.This produces a reduced internal structure developing within the uid medium and a decrease in the resistance to the disruption of the structure under shear stress.The early reactivity in the y ash slag blend is due to slag reaction [4].The presence of silica in the activating solution also has a retarding effect on the dissolution of slag [30,31].The addition of clay and CMC are required to overcome the in uence of dissolved silica and provide the required rheology control.Addition of CMC induced a viscous dominance in the mixture.The use of Bentonite clay provided a yield stress enhancement, which in combination with CMC resulted in synergistic increases in yield stress and thixotropic buildup.
The buildability on the time scale of observation is provided by the structuration within the medium due to both reversible and irreversible processes.On the time scale of relevance to the printing, the kinetics of the reaction is typically not a factor contributing to the buildup.Flocculation and aggregation of particles is enhanced by the clay that contributes to buildability.The presence of CMC has been shown to increase the occulation of clay [19].Thixotropy as a function of yield stress with time is an important measure of buildability which clearly shows that very high thixotropy with a rapid increase in yield stress can lead to blockage at the nozzle and, therefore, is not suitable for printing.This is demonstrated in M6 though it has the best shape retention in the printed layer structure.
The evaluation of rheological responses of the mixtures indicated certain trends produced with the variations in the proportions of the clay and CMC in the mixture.In the presence of CMC there was a synergistic increase in the yield stress, viscosity and thixotropy with clay content.Good shape retention and buildability was obtained on increasing the proportion of clay relative to CMC.However, increasing the clay in proportion to CMC also resulted in a very rapid buildup of yield stress with time (as seen in M6) that resulted in choked ow.A larger proportion of CMC produces a viscous dominated Maxwell ow type response, which did not have shape retention under the load of additional layers due to the continuous deformation under applied stress.The proportion of CMC and bentonite is therefore important in controlling the rheological parameters within the range suitable for printing.The plot between the bentonite content and CMC from all the FAB mixtures is shown in Fig. 8. Considering shape retention and buildability, the proportion of bentonite clay to CMC in M5 produced the best performance for printability.
The bentonite clay at 1.8% by mass of the binder provided adequate rheology control when used in conjunction with 0.25% CMC.
The printability and buildability of concrete mixtures made using FAB paste was demonstrated using a gantry-type printer.A concrete mixture was prepared using the FAB paste containing clay and CMC.The aggregate content in the concrete was proportioned at a mass proportion relative to the binder of 1.5:1.
The aggregate The ratio of clay and CMC was kept constant in the same proportion as M5 FAB mixture.
The material was pumped into the hopper using a piston-type of pump at ow rate of 10 kg/min.The print speed was xed at 30 mm/s.The nozzle diameter was 30 mm, and the standoff distance was 20 mm which is the height or thickness of each layer.The width of the layer was 50 mm.The printed shape shown in Fig. 9 exhibits good shape retention and buildability with a total height of 200 mm (10 layers).
The packing of slag and y ash did not change between the mixes.The rheology of the solution was altered by the addition of clay and CMC.Generally, the addition of clay and CMC resulted in an increase in the yield stress, viscosity, in addition to enhancing the thixotropy.As a general observation, the uidity of the particulate suspension is controlled by the yield stress and the viscosity of the uid medium between the particles.The uidity of the suspension is inversely proportional to the viscosity of the paste.Similarly, the shape retention is proportional to the yield stress of the uid medium.For a given packing, a certain uidity is required from the paste to ensure adequate yield stress.

Conclusions
In this study, the relationship between y ash-slag blends and alkali activation in terms of composition, rheological behavior, and 3D printing performance is assessed.The behavior of blends activated using a new source of silica (FAB) was evaluated.The in uence of additives like Bentonite and CMC on the shape retention and buildability of the mixes were assessed.The conclusions that can be drawn from this study are: Fly ash based aqueous nano silica (FABANS) is a viable source of nano silica that can be implemented to improve the strength of the alkali-activated binders.The rheology of an alkali-activated binder made with FABANS is not suitable for printing.
Bentonite clay produces improvements in yield stress and viscosity.There is a synergistic relationship between clay and CMC which helps in better performance of the binder in terms of printability and buildability.
Printing is favored only by yield-type behavior and the Maxwell-type behavior results in an uncontrolled deformation along with shape instability which makes it unfavorable for extrusion-based 3D printing.

Figure 2 .
Figure 2. Comparison of M1 FAB mixtures with Control, 1% NS and 3.3% NS in terms of (a) Constant Strain Response and (b) Storage modulus

Figure 3 .
Figure 3. Constant Strain Response of mixtures with variation in (a) Bentonite clay and (b) CMC.

Figure 4 .
Figure 4. Variation of yield stress with time of mixtures with variation in (a) CMC and (b) Bentonite clay.

Figure 5 Print
Figure 5

Table 3
Rheology and print behavior of 3D printed FAB mixtures with different additives * Percentage mass of binder consisting of y ash and slag.