Investigation of concrete mixtures for additive construction

Purpose This paper aims to qualify traditional concrete mixtures for large-scale material extrusion in an automated, additive manufacturing process or additive construction. Design/methodology/approach A robust and viable automated additive construction process must be developed that has the capability to construct full-scale, habitable structures using materials that are readily available near the location of the construction site. Accordingly, the applicability of conventional concrete mixtures for large-scale material extrusion in an additive construction process was investigated. A qualitative test was proposed in which concrete mixtures were forced through a modified clay extruder and evaluated on performance and potential to be suitable for nozzle extrusion typical of additive construction, or 3D printing with concrete. The concrete mixtures were further subjected to the standard drop table test for flow, and the results for the two tests were compared. Finally, the concrete mixtures were tested for setting time, compressive strength and flexural strength as final indicators for usefulness in large-scale construction. Findings Conventional concrete mixtures, typically with a high percentage of coarse aggregate, were found to be unsuitable for additive construction application due to clogging in the extruder. However, reducing the amount of coarse aggregate provided concrete mixtures that were promising for additive construction while still using materials that are generally available worldwide. Originality/value Much of the work performed in additive manufacturing processes on a construction scale using concrete focuses on unconventional concrete mixtures using synthetic aggregates or no coarse aggregate at all. This paper shows that a concrete mixture using conventional materials can be suitable for material extrusion in additive construction. The use of conventional materials will reduce costs and allow for additive construction to be used worldwide.