MODELING THE FRESH PROPERTIES OF 3D PRINTABLE CONCRETE
Advisor: Professor David Lange
Abstract
Additive manufacturing of concrete materials has recently gained attention in the field of civil engineering. The construction industry has shown substantial interest in exploring the use of 3D concrete printing as a supplementary to conventional technology considering the reduced labor, formwork, and carbon emissions. A major impediment to the application of 3D printable concrete lies in designing the fresh material properties. Ideally, a fresh concrete needs to be fluid-like for pumping and extruding but solid-like during placement to support the printed structure. In practice, it is fairly challenging to actualize these needs. A thorough understanding of the concrete flow behavior in 3D printing process is necessary to optimize the fresh properties for 3D concrete printing. Despite extensive experimental methods have been developed for characterizing the concrete flow behavior over the past decades, they are insufficient to resolve the complexity encountered during the 3D printing construction especially in full scale. Leveraging the advances in computer techniques, this study aims to model the fresh properties of 3D printable concrete based on numerical simulation and machine learning.
To fulfill the goal, this research can be divided into three parts: 1) understand the rheological behavior of cement-based materials; 2) build a numerical model to simulate the flow behavior of cement-based materials and determine the intrinsic and external control parameters that make concrete; and 3) develop a machine learning model to predict fresh properties and 3D printability of concrete from material design.
