Topology Optimization of 3D Printed Concrete Structural Walls Through Artificial Intelligence Techniques

Abstract

Concrete is one of the most widely used building materials worldwide. Various types of concrete, such as self-compacting concrete, high-strength concrete, and chemical-resistant concrete, are used globally to improve the performance of structures. Digital technology applications like 3D printers offer new possibilities for the production of concrete structures. There are many advantages to numerically modeling the 3D concrete printing process. Simulating the printing process allows for predicting the structural behavior of concrete during production, and calculating the potential damage that may occur at this stage. This is important for reducing material waste and preventing wear on the printing machine. With the increasing demand for sustainable product development, scientists have begun to focus on producing hollow structures with high strength and minimum weight. Topology optimization procedures will contribute to the more economical and robust design of structures, providing new contributions to the engineering, architecture, and construction fields. Many researchers have developed physics-based models to improve the properties of hollow elements printed through 3D printers. However, these formulations require a deep understanding of the production process, which can be very laborious. As an alternative to this method, predictions can be made using only the data obtained. In this context, various artificial intelligence methods can be used. In the 3D printing process of concrete structures, the design process should be done in the office environment, while the printing process should be done in the construction site. This simultaneous occurrence of the design and printing processes makes it impossible to fully automate the 3D concrete printing process. Additionally, there is no formulation in the literature for the capacity calculation of concrete walls printed through 3D printers and subjected to combined axial and lateral loads (such as design effects occurring during wind loads or seismic demands). To achieve fully automatic 3D concrete printing, in this study, the optimization of wall topology in concrete building construction printed by 3D printers has been realized with artificial intelligence techniques, and automated design tools have been developed using these artificial intelligence models. The dataset required for the artificial intelligence model was obtained through numerical modeling. With this study, the most economical and lightest section topology with the desired strength was determined through the artificial intelligence model without resorting to trial and error, saving material and energy. Additionally, by comparing two different artificial intelligence models, the model with the highest accuracy and the lowest number of parameters was found, reducing the calculation cost. In addition, the pre-trained version of the proposed artificial intelligence-based design tool has been uploaded to permanent storage, and the relevant link has been added in the conclusion section of this study.