Building information modelling (BIM) enhanced 3D concrete printing

• Main Author: Liu, Qinyan
• Other Authors: Qian Shunzhi
• Format: Final Year Project
• Language: English
• Published: Nanyang Technological University 2024
• Subjects: Engineering; 3D concrete printing; Point cloud data; Printing path automation
• Online Access: https://hdl.handle.net/10356/176963
• Institution: Nanyang Technological University

3D concrete printing (3DCP) is the process of using additive manufacturing techniques to fabricate concrete structures layer by layer, based on a digital 3D model. As compared to conventional construction methods, 3DCP offers advantages. It increased design flexibility, allowing architects to create intricate structures which might be challenging to achieve using the conventional construction method. Moreover, 3DCP minimises material wastage by precisely depositing concrete only where it is needed according to the designed printing path. Additionally, the automated nature of 3DCP can increase the overall construction time, accelerate project timelines, yet reduce overall costs. However, current process of 3DCP is still labour-intensive, making it unable to be fully automated. 3D point cloud data obtained from methods like photogrammetry, 3D scanning, and laser scanning, serves as a comprehensive digital representation of physical objects. These points capture the spatial coordinates of the scanned object accurately. This efficient and precise process has been widely adopted in the construction industry, enhancing project visualisation and documentation. Many researchers have tried to integrate 3D point cloud data into traditional 3D printing processes, but few have attempted to adopt 3D point cloud data into 3DCP. Thus, this study aims to analyse the efficiency of integrating 3D point cloud data into 3DCP. The point cloud data was first simplified to reduce the number of points while maintaining its shape using Douglas-Peucker algorithm. Afterwards, two different methods, DBSCAN and a hybrid code developed, were used to rearrange, and convert the point cloud data into machine-readable code that can be used for 3DCP directly. The efficiency of these two methods were analysed. Result obtained using the new hybrid code is able to generate a continuous printing path that is smooth and with no outliers. Hence, the aim of this study is achieved.