Concrete pumping, an on-site technology thanks to which the material can be poured where it would otherwise be unfeasible, hastens construction, thereby raising output and lowering costs. In additive manufacturing or 3D printing it is the method most usually deployed to carry the freshly mixed material to the printer nozzle. Alternative systems involving intermediate deposition to ensure a continuous flow of material have also been tested, however. Although less common than pumping, these procedures (Figure 2), have proved able to suitably maintain a constant supply of material while avoiding possible pumpability issues.

In additive manufacturing, the material is generally pumped through a flexible pipe to ensure nozzle positioning and mobility (Figure 3). The pressure used must be sufficient for the material to deform and travel through the pipe. Depending on the size of the aggregates, this material can be considered a mortar or micro-concrete.. Any change in the composition may affect fresh mix performance. Characteristics such as the water/cement ratio, aggregate particle size distribution or the presence of superplasticizers or other chemical admixtures modify material rheology and must therefore be determined a priori.

That determines a need to estimate the pressure required to pump a given mix of fresh material a given distance or height at a given known and controlled speed. Such estimates lay the grounds for optimising the composition of the cement-based material and defining the pumping pressure required to feed the printer, for an uninterrupted and constant flow rate is indispensable in additive manufacturing. The extensive research conducted on that issue in recent decades has given rise to a number of prediction and test models as well as numerical simulations of in-pipe particle flow (11), using different types of rheometers and tribometers to explore concrete pumpability (12, 13).

Nonetheless, in actual on-site practice pumping is still estimated with simple and empirical methods such as the slump test to determine freshly mixed concrete or mortar consistency. Based on the results of that test, expressed in centimetres (14), the material is classified under one of four consistency categories: true slump, zero slump, shear slump or collapse. There are values for a concrete to be pumpable, for 3D printed materials it is necessary to test in each case and it is not possible to determine a common value for all materials and pumping systems.

For mortar, characterised by a smaller aggregate size, fresh state consistency can be determined with two tests. The first involves placement on a spread or flow table to establish consistency in terms of the mean diameter of the resulting spread (15). The second is based on the penetration of a plunger in the sample (16). Traditionally a mortar has been said to be pumpable if it has a flowable consistency. Neither of these tests has proven suitable for accurately estimating pumping pressure, however.

Estimation can be found with slump tests and numerical simulation. ACI standard 304.2R-96-2000 (17) on placing concrete with pumping methods, like many of the diagrams used by concrete pump manufacturers, accepts the simplifying assumption that head loss and flow rate are linearly correlated. The empirical formula used is Equation [1]. Simplified formula for calculating pumping pressure, defines the pressure (p)-flow (q) p/q, m³/h relationship as linear during pumping. The coefficient that relates the two is equal to a constant that depends on circuit geometry (determined as length, L (m) and diameter, D (m)) and mix properties. That constant, b (coefficient on the tables published in the standard), is expressed as the mean slump.

New procedures have been developed in recent years that deliver a greater quantity of more accurate information than those traditionally used to find the pumping pressure required to ensure a suitable supply of material from the mixer to the printer nozzle. These procedures attempt to be more sensitive to water/cement ratio, aggregate shape and admixtures than the aforementioned viscometers and concrete dispersion tables. Some of these methods are the sliper tube rheometer (12) or the study of pumping pressures with a tribometer (13), which measures the friction between the fresh material and the pipe through which it flows.

In extrusion, the second stage of AM consisting in controlled layer-by-layer deposition of the material pumped through the nozzle to attain the desired geometry (Figure 4), rheology is the determinant parameter.

Nonetheless concrete rheology, understood as its fluid and fresh state resistance to deformation, i.e., its workability, is not the sole factor involved in extrusion. Other properties such as homogeneity, suitable setting time and high thixotropy also affect the outcome.

As the material must remain sufficiently workable to be extruded without clogging, workability in concrete refers to how readily it can be laid on site after mixing. The American Society for Testing and Materials (ASTM) defines it as the property that determines the ease with which a freshly mixed concrete can be mixed, placed, consolidated and finished to a homogeneous condition (18).

At this time a wide variety of tests is applicable to determine the ‘open’ time (19), or the time during which a product is workable before hardening. Specific tests are in even place for different types of concrete. In self-compacting concrete, for instance, the most suitable procedure depends on the self-compacting method (20-23). European standard EN 196-3 (24) specifies the methods for determining ordinary cement setting times, although many other cement-based materials also apply the tests described, in light of their benchmark status. For bespoke materials expressly designed for additive manufacturing tests must be adapted and studied for calibration.

Another parameter that must be established is mix temperature prior to extrusion. Like any other chemical reaction, cement setting is affected by temperature. Early-age properties of 3D-printable mortars consequently depend heavily on the initial stages of (heat-generating) hydration kinetics. Hence the importance of assessing the extent to which the properties measured under standard laboratory conditions (20 °C) may be affected by variations in temperature in real out-of-laboratory printing.

Lastly, the material to be used must be thixotropic or exhibit accelerated early-age setting to maintain its shape after extrusion. Further to the definition of thixotropy, that entails studying the variation in material consistency when exposed to external actions such as extrusion and deposition.

Based on Mewis (25), Sonebi & Yahia (26) define thixotropy as a reversible, isothermal, time-dependent decrease in the apparent viscosity when a material is subjected to increased shear rate. There is no specific test to measure mortar or concrete thixotropy. Tests for self-compacting concretes are in place, as well as a test described in European standard EN 13062:2004 (27), to determine thixotropy in products for protecting reinforcing steel, which could be likened to the type of material at issue here.

That test consists in applying the freshly mixed material on two steel plates, one set horizontally and the other vertically. Thixotropy is the quotient of the mean thickness at the top of the vertical plate divided by the mean of the thicknesses on the horizontal plate.

In this final stage of 3D printing the conditioning factor is the buildability of the fresh mix used. Here the most important factors are the mechanical properties of the fresh mix, which must be highly thixotropic (or contain accelerating admixtures) to accommodate the deposition of one layer over another with acceptable and controlled deformation.

In this final stage of 3D printing, the conditioning factor is the buildability of the fresh mix used. Compared to traditional concrete, there are two mechanical problems that condition the buildability.

The first one is the evolution of the mechanical properties of the concrete in the early ages, from the plastic and deforming state to the hardened state, the layer-by-layer construction method requires a rapid development of the compressive strength to withstand the tension of the previous layers. Here the most important factors are the mechanical properties of the fresh mix, which must be highly thixotropic (or contain accelerating additives) to accommodate the deposition of one layer on top of another with acceptable and controlled deformation.

The second problem is the strength of the bond between layers, this also occurs in other types of concretes such as self-compacting concretes with so-called “cold joints’’ or “distinct layer castings’’, which are the interfaces where there is a reduction in mechanical properties.

The strength of this interlayer bond can be affected by a number of factors. Firstly, the parameters of the printing process itself, including the interval time, printhead speed and the height of the printing nozzle Panda et al. (28, 29) found that low values of the printhead speed and the separation distance of the printing nozzle from the printing nozzle led to an increase in the interface bond strength. On the other hand, Wolfs et al (30) reported that there was no clear relationship between nozzle height nozzle height and interface bond strength.

Other properties have effects on this adhesion, Roussel and Cussign (31) observed that the thixotropic property of cementitious materials had a negative effect on the interface bond strength, which is in contradiction with other requirements.

Two other properties of particular significance for characterising the mix and assessing its fresh state variations in volume are fresh mix compactability (32) and entrained air content (33, 34).

The mixing and pumping system that feeds material to the printer is common to most 3D printing setups used in international projects.

In all those cases the pre-prepared materials, delivered in sacks, were mixed with water on site for subsequent uninterrupted pumping from the mixer to the extrusion nozzle (Figure 5).

The pre-prepared material contained the admixtures needed to ensure the necessary workability, pumpability, resistance to premature ageing and compliance with all other requisites.

Another approach, consists in preparing several litres of very fluid material apt for a given printing time, designed to flow under its own weight via a screw-based system to a mixing chamber placed immediately prior to the nozzle (Figure 6).

There viscosity modifier agents (VMAs) or accelerating admixtures are added to the mix for extrusion to the geometry defined. Due to such late-stage inclusion of the admixtures, the material acquires a rheology that ensures workability as well as early strength development.

When tested, this process met the aims established, with the material exhibiting ready pumpability, ample open times and high early-age mechanical strength.

Fibres are sometimes added to cementitious materials designed for 3D printing to improve mechanical properties, mainly at flexural strengths where these materials have their limitations. When fibres are included in the cementitious material it is necessary to study the effect of different content percentages on workability and mechanical properties (35, 36). That is, the tensile strength should improve while maintaining an acceptable workability for 3D printing (37). There are also studies aimed at investigating the effect of fibre orientation in 3D printed concrete, determining its possible effects on the mechanical properties of 3D printed specimens and comparing them with those of conventionally moulded specimens. Some results revealed that smaller nozzle size and higher fibre volume fraction significantly improved fibre alignment parallel to the printing direction (38).

In order to correctly characterise a ultra-high performance fiber-reinforced concrete, it is necessary to determine the influence of its dosage, fibre content and fibre type. And finally the choice of test methods (39).

In this procedure mortars (material containing no coarse aggregate) or concretes (material containing coarse aggregate) are air-blown onto a substrate on site at high pressure from a hose. The purpose is to cover the soil itself where otherwise impossible due to sloping terrain or other factors, ultimately to enhance strength and durability and optimise weatherproofing thanks to the scant porosity of the cover.

This technique is commonly used to stabilise embankments, clad tunnels and build swimming pools.

Gunned concretes and mortars must meet a number of requirements that concur with the characteristics called for in 3D printing. As readily pumpable materials bearing setting stabilisers, they have flexible open times, high early-age strength and good substrate bondability, all features shared with additive manufacturing.

As those features of the materials used in sprayed concrete are shared with 3D printing materials, subject to the imperative prior study, these materials are candidates for adaptation to use in the additive manufacturing.

The mortars used as protective indoor and outdoor surface coverings and finishes are called rendering or float coats. Depending on the characteristics of the material they can be prepared for single or multilayer application.

The properties of these mortars that have prompted their use in a few pioneering 3D printing projects include: early-age strength (although the values are not comparable to those of other mortars), impermeability, substrate bonding, flexibility and quality of finish.

As in the case of gunned mortars, some of the characteristics of the rendering materials have induced their use in early tests of 3D printing projects (Figure 7). Those first tests gave very suitable values for some characteristics such as interlayer or substrate bonding, however others such as early-age strength, while acceptable were not very high. The result was that their use was discarded later in the project and ad-hoc materials were developed.

Repair mortars, polymer-modified cement-based mortars and epoxy-based materials are developed to specific requirements to remedy damaged or deteriorated concrete and mortar.

Repair mortars are characterised primarily by excellent workability and finish, good substrate and interlayer bonding and excellent bonding to materials such as steel.

Their setting times can be controlled and they exhibit high early-age strength, essentially no drying shrinkage and high thixotropy.

With such features these mortars are worth exploring for their 3D printability.

In some projects, tests have been run to determine setting times as specified in standards for ordinary cement (24). Such tests are not applicable to the materials used in other projects, however, because the standard method is not apt for measuring very short setting times, as when the mortar or concrete bears accelerating admixtures.

Another standard method deploys a manual device call the Vicat apparatus (Figure 8), comprising a probe or needle that penetrates a standardised mould containing the test material. The probe is lowered at pre-set intervals and the depth of penetration recorded.

Figure 9 shows the setting times measured with a Vicat apparatus, and in table 1 we can see the beginning of setting of the mortar. It can be seen the comparison between mortars formulated for other applications and mortars developed ad-hoc for a project. The figure shows that the onset of setting is earlier in the mortars developed for 3D printing, while commercial premixed mortars show much longer setting times.

The slump test, used to determine the consistency of fresh masonry mortars (15) may also be applied to potential printing materials.

This test delivers good results for mortars where consistency depends on the water/cement ratio or other parameters, but is unsuitable when a thixotropy-high, freshly mixed material holds its shape (see photographs reproduced in Figure 10).

Both compressive and flexural strength (Figure 11) are among the parameters most routinely tested in most projects. The usual procedure is to run the test on a 4x4x16 cm prismatic specimen, further to the standard for mortars (42).

Figure 12 gives the results for a number of materials tested, according to which strength varied substantially from 15 MPa for 28 d rendering mortars to over 90 MPa for an ultra-high performance concretes (UHPC). Further to those findings, whilst both materials are printable, their performance in other respects is not comparable.

These tests can be conducted after several days or at very early ages, from 30 min to 120 min, and also accommodate comparison among materials such as the ultra-high performance concretes (UHPC) shown in Figure 13.

After extrusion-mediated layered deposition, data must be collected (Figure 14) to ensure that any settling of the material under its own weight lies within the pre-established tolerance interval.

Printing stability must also be verified on both horizontal and vertical substrates as the layers are applied. As shown in Figure 15 on horizontal substrates the aim is to ensure constant verticality and stability or at least to determine the number of layers that can be deposited before the operation must be momentarily suspended. Similarly, on vertical surfaces the aim is to establish the number of layers ensuring good substrate and interlayer bonding. Those three parameters - shape, stability and tolerances - define a material’s buildability.