Experimental Method to Identify Viable Mixture Proportions and Printing Parameters for Additive Manufacturing Using Calcined Clay-Based Cementitious Inks

Proposed in this paper is a systematic methodology to identify cementitious paste mix design and printing parameters suitable for yielding dimensionally accurate 3D-printed specimens. Using a Box–Behnken surface response experimental design approach, the first work phase generated mathematical models to predict ink flow based on the paste water–binder ratio and dosage of superplasticising and viscosity-modifying admixtures. In the second work phase, favourable mixes were investigated further, again using a Box–Behnken surface response experimental design and corresponding mathematical models to predict printing accuracy in relation to printing parameters including ranges of print speed, extrusion multipliers and layer height. Confirmed by parallel preliminary print trials, favourable ink mix designs were efficiently identified for inks comprising either Portland cement only, or ternary blends of calcined clay, silica fume and Portland cement. For the Portland cement binder, the favourable mix design comprised a water–binder ratio of 0.26 and superplasticising and viscosity-modifying admixture dosages of 1.12% and 1.10% by mass of the binder respectively. Corresponding values for the calcined clay/silica fume/Portland cement binder ink were 0.27, 1.1% and 1.1% respectively. For both binder types, corresponding favourable values of the above listed print parameters were identified as 5 mm/s, 1.1% and 1.0 mm respectively. In the final work phase, the outputs from phases one and two were validated via print trials of hollow cube, beam, hexagonal and circular elements enabling measurements of the dimensional accuracy and buildability. With deviations in element height and width ranging from only ±0.3 to 2.6% from corresponding CAD designs, the appropriateness of the methodology was established.