Cornell Aims to 3D Print Sediment-Based Underwater Arches in DARPA Challenge

Cornell University researchers are racing to 3D print concrete structures directly on the seafloor, testing a sediment-first approach that could cut the need to ship tons of cement and change how subsea repairs and construction are done. The effort is part of a DARPA one-year challenge that awarded Cornell a milestone-driven $1.4 million grant in May 2025 and calls for teams to deposit printable material several meters underwater at a March demonstration.

Assistant professor Sriramya Nair leads an interdisciplinary team from Cornell’s David A. Duffield College of Engineering combining materials science, robotics, and additive manufacturing. Cornell began work in 2024 after DARPA issued a call late that year. Cornell is one of six teams competing; the project culminates when each team must 3D print a concrete arch underwater during the March event.

The core constraint is unusual: DARPA requires that most of the printable material be derived from seafloor sediment rather than traditional cement. That mandate aims to reduce transport logistics and enable in-situ fabrication for ports, pipelines, and offshore energy assets. Achieving a printable slurry from local sediment creates difficult tradeoffs in rheology and kinetics. Water-driven washout, where cement particles fail to bind during deposition, is a principal failure mode. Adding stabilizers too early thickens the slurry to the point it cannot be pumped. Disturbing fine sediment produces turbidity that can drop visibility to near zero, making hands-on inspection impractical.

Cornell’s practical countermeasures mix print hardware with a two-stage materials approach. The team has run frequent test prints in a large water-filled tank at Cornell’s Bovay Civil Infrastructure Laboratory Complex using an approximately 6,000-pound industrial robot for large-scale concrete printing. The printable feed is a flowable, high-sediment low-cement base that can be pumped through hose and nozzle. At the nozzle, additives or “unique mix-ins” trigger super-fast solidification on contact with water so layers adhere and resist washout. That approach accepts a narrow window between pumpable flowability and rapid setting at the deposition point.

Sensorization and robotic control are central to the workflow. Engineers mounted a control box with sensors on the robot arm to check layer deposition and print quality, enabling real-time adjustments when visibility and manual inspection fail. The Bovay lab’s controlled tank lets researchers inspect geometry, strength, and texture that would be impossible in murky field conditions. Cornell demonstrated progress toward the DARPA high-sediment targets during a September demonstration to agency officials.

Nair frames the work bluntly: “Nobody is doing this right now. Nobody takes seafloor sediment and prints with it.” She said the team took on DARPA’s call to learn the hurdles: “When the call for proposals came out, we said, ‘Hey, let’s just do this and see, so that we will at least understand what the challenges are,’” and she added that the team’s mixture allowed underwater printing after adjustments for continuous water exposure.

For the 3D printing community this is a practical pivot from above-water extrusion. The March demonstration will show whether sensor-driven control and nozzle-triggered chemistry can tame washout and turbidity in a live challenge. Expect the next wave of detail to be technical performance metrics, sensor logs, and concrete mix characterizations after the arch prints.