3D printing in reactor construction: ORNL shortens material testing for nuclear power components

Nuclear engineering is under pressure to bring new reactor concepts to deployment readiness more quickly. Researchers at Oak Ridge National Laboratory (ORNL) are specifically relying on additive manufacturing to produce complex components such as pump housings, valves, or compact heat exchangers using 3D printing and to reduce delivery times from years to months or weeks.

In the Nuclear Fuel Materials Group, Caleb Massey is investigating how fuel and structural materials age under intense neutron irradiation and in corrosive media.

Although research reactors reproduce many conditions, they often differ from the planned operating parameters of advanced reactors, which delays test campaigns.

Approving 3D-printed materials is particularly demanding. For conventionally manufactured alloys, it can usually be assumed that a single specimen is representative of the entire semi-finished product. Additively built components, on the other hand, exhibit variations in properties within a single part, depending on geometry, local cooling rates, and process parameters. ORNL is therefore developing test series that capture this variance and integrate it into existing standards as well as testing concepts for non-destructive examinations and non-standard specimen geometries.

Digital tools are intended to reduce testing efforts. Numerical models are calibrated so that they can predict deformation behavior even when material properties are direction-dependent. In parallel, AI-based methods are used to select from thousands of potential high-temperature tests those that maximize the information gain.

Supported by programs of the U.S. Department of Energy, ORNL is linking this materials research with demonstration reactors to validate the data obtained in real systems. In the long term, a qualified 3D-printing supply chain is intended to ease bottlenecks in reactor operation and facilitate the deployment of compact reactors for applications such as dedicated power supply for computing and AI data centers.