Control of mortar rheology for 3D concrete printing

Additive manufacturing is an emerging technology that is being used in construction to remove the need for moulds and formwork. The 3D concrete printing technique developed at Loughborough University involves the extrusion of cement-based mortar, thus having broad similarities with the fused deposition modelling type of additive manufacturing, but on a larger scale. The firstgeneration gantry beam was successful in printing flat and curved layers but was relatively slow. In order to maintain sufficient open time for printing, a retarding admixture was added to the mortar mix along with a superplasticizer to ensure that the mix was the correct workability. More recently a second-generation printer has been developed which utilises a robotic arm instead of a gantry. The speed of the process has increased considerably but the process still has challenges in terms of control. The rheological properties of the mortar affect its printability, in particular the number of layers that can be placed at a time, and the quality and strength of the final printed object. The rheology of cementitious systems, and mortar especially, are not well described by existing theory, and this work aims to add to the understanding of mortar rheology. This research investigates the effect of the mortar rheology on the printing process, which can be broken up into 3 areas: comparison of three techniques for measuring the rheology, measurement of the deformation during printing, and synthesis and characterisation of a superplasticizer. Due to the nature of this work it therefore embodies both engineering and chemistry research, in an effort to explore the structure property relationship between the mortar, its rheology and the polycarboxylate based superplasticizer. The three techniques for measuring the fluid properties of the printing mortar were a building materials rheometer, a modified version of the geotechnical shear vane and the slump flow test (BS EN 12350-8:2010). The yield stress and plastic viscosity were calculated from the rheometer data, along with the viscosity index, a measure of the Newtonian nature of the mortar. The plastic viscosity decreased as superplasticizer concentration increased, but a poor trend was seen for yield stress, although the yield stress for mixes containing the lowest superplasticizer concentration was significantly higher than for the other mixes. It was found that above a superplasticizer concentration of 1.25% the viscosity index was close to zero, suggesting that a yield stress does not exist. The shear vane test was used as it is a relatively cheap, easy to carry out test but only gives a single value; the vane shear strength, commonly correlated to the yield stress. Comparisons showed though that the vane shear strength was related to both yield stress and plastic viscosity. Videos were taken of the slump flow test to provide a more dynamic measurement; as again the slump flow is usually a single point test. Curve fitting was then undertaken to monitor the growth of the slump flow area with time, and it was found that the slump flow values were linked to both plastic viscosity and yield stress. Mortars with three different superplasticizer concentrations were then printed to observe the effect of rheology on the printability and in particular the height and width of extrusions and the number of layers that could be built on top of each other. The sizes of the voids in the printed samples were also determined by image analysis. It was found that the mortar mixes containing the highest levels of superplasticizer deformed more from the expected size. The mix containing 0.75% superplasticizer had the least deformation, but had a high proportion of large voids in the hardened samples. This mix had a significantly larger yield stress that the others, suggesting that the yield stress is critical for void sizes. When deciding on the appropriate rheological properties for use in printing there will be a balancing act between expected shape and strength of hardened object. A known polymer superplasticizer was then synthesised and the rheological properties of mortar containing the superplasticizer measured. As the polymer contained none of the additives associated with commercial superplasticizers the direct effect of the polymer could be analysed. The polymer structure was investigated by gel permeation chromatography, showing that specific synthesis conditions were required. The polymers were then added to mortar and the rheological properties assessed before deformation samples were produced. The polymer had to be added to the mortar mix at a higher concentration than the commercial superplasticizer used for the rest of the project. The reasons for this were investigated further through the use of tuneable resistive pulse sensing (TRPS) to measure the zeta potential of silica fume and superplasticizers.