Concrete 3D printing is revolutionizing the building industry by cutting costs and construction time.
However, most current systems use an extrusion-based method, laying down concrete layer by layer directly under the printer’s nozzle.
This approach limits what builders can design and makes it difficult to print around steel reinforcement bars—known as rebars—without collisions, reducing both strength and flexibility in finished structures.
Researchers at Carnegie Mellon University’s Computational Engineering and Robotics Laboratory (CERLAB) have now developed a new simulation tool that could change that.
Led by mechanical engineering professor Kenji Shimada, the team has created a model that accurately predicts how concrete behaves when sprayed, rather than extruded.
This innovation supports a process known as spray-based 3D printing, which builds up layers of concrete by spraying a modified shotcrete mixture onto a surface.
“Spray-based concrete 3D printing is a new process with complex physical behaviors,” said Shimada. “In this method, the mixture can be sprayed freely, even around rebar, making it possible to construct stronger and more flexible designs.”
This ability is particularly valuable in earthquake-prone regions such as Japan and California, where reinforced concrete structures are essential. However, to make the technology practical, engineers must be able to predict exactly how the sprayed concrete will behave—how it will spread, drip, and harden into its final shape.
To address this challenge, Shimada’s team developed a simulator that models how shotcrete mixtures behave in real-world conditions. The software accounts for multiple factors, including viscosity, rebound of particles, solidification time, and even how the material spreads after impact.
Contractors can use the simulator to test different printing strategies on a computer before committing to real construction, helping them decide whether spray-based 3D printing is suitable for a particular design.
The team tested their simulator in Tokyo, Japan, where Shimizu Corporation already uses spray 3D printing robots.
In trials, the model proved impressively accurate—predicting the height of sprayed concrete with 90.75% accuracy and the shape of concrete printed over rebar with more than 97% accuracy in some measurements.
According to Soji Yamakawa, the study’s lead author, the simulator achieves these results much faster than traditional physics models. “By simplifying extremely complex physics into efficient algorithms, we were able to achieve highly realistic results in a fraction of the time,” he said.
The researchers plan to refine their model further by adding new factors such as humidity, plastering effects, and surface smoothness. Ph.D. student Kyshalee Vazquez-Santiago said the team sees enormous potential: “Spray 3D printing opens up an entirely new way to build—faster, stronger, and smarter.”
Their research, published in IEEE Robotics and Automation Letters, marks an important step toward safer and more sustainable construction powered by robotics.
