Receiving a $1.3 million Advanced Research Projects Agency (ARPA-E) award, the partners will focus on high-performance parts for turbomachinery applications. Their target is to reduce the design and validation timeline by up to 65% and demonstrate a defect-free, high-performance multi-functional design in the hope of contributing to an increased adoption of additive manufacturing in the power industry.
The three organisations intend to optimise structural characteristics and thermal and fluid properties using AI-generated surrogate models and integrating multi-physics design optimisation techniques. In the design of new components for complex power products, whether it be jet engines or gas turbines, dozens of experts are typically responsible for the structural, thermal and fluid properties. When those new components are set to be additively manufactured, engineers must consider how material composition will respond to heat and stresses and how the design of the part impacts things like airflow and aerodynamics.
Moving through these evaluations, adjusting the design as and when required and validating the final part can take between two and five years. As it stands, this timeframe means traditional manufacturing technologies are typically the preferred method of production.
“One of the keys to enabling the widespread use and benefits of 3D printing is the reduction of the time it takes to create and validate defect-free 3D component designs,” explained Brent Brunell, leader of GE Research’s Additive efforts. “Using multi-physics enabled tools and AI, we think we can beat the timeline for some traditional manufacturing processes by automating the entire process.”
“The combination of model-based and data-driven AI to accelerate generative design is a key innovation that will dramatically reduce the time to synthesise and fabricate quality parts,” commented Saigopal Nelaturi, Manager of Computation for Automation in Systems Engineering area in the System Sciences Lab at PARC. “Surrogate models (built using machine learning) that encapsulate complex couplings between process physics and part quality will help guide the optimisation models in feasible regions of very high dimensional design spaces. This combination of AI techniques enables automatic multi-functional part synthesis to meet real-world application demands, for which AM can provide truly novel solutions.”
Throughout the project ORNL’s Summit supercomputer will be used to create the AI-based surrogate models with ‘unprecedented precision’ and the organisation’s High Flux Isotope Reactor will be deployed to analyse additively manufactured components and generate data necessary for training and evaluating AI-based models.
“This is the type of project that leverages the unique capabilities at ORNL – experimental and computational facilities – as well as expertise in computational science and additive manufacturing,” said John Turner, Computational Engineering Program Director at ORNL.
At the project’s end, the partners hope to have an additively manufactured multi-functional design which demonstrates the ability to withstand high temperatures and stresses with an improved performance when compared against conventional casting.