Johns Hopkins researchers have used AI to refine titanium alloy manufacturing, accelerating production while preserving or enhancing strength, a breakthrough with potential benefits for aerospace, defense, and medical industries.
Published in the journal Additive Manufacturing, the study focuses on Ti-6Al-4V, a widely used titanium alloy valued for its strength and low weight.
By utilising AI-driven models, the researchers mapped out previously unexplored manufacturing conditions for laser powder bed fusion, a metal 3D-printing method.
This approach challenges long-standing assumptions about process limitations and expands the possibilities for producing high-quality titanium components with customizable properties, the university said in a news release.
Morgan Trexler, program manager for Science of Extreme and Multifunctional Materials at APL, emphasised the importance of accelerating manufacturing to meet evolving operational demands.
“At APL, we are advancing research in laser-based additive manufacturing to rapidly develop mission-ready materials, ensuring that production keeps pace with evolving operational challenges,” Trexler said.
The AI-driven method identified new processing regions that allow for faster printing while preserving or even improving material properties, such as strength and ductility. Senior materials scientist Brendan Croom explained that traditional manufacturing approaches often dismissed certain processing parameters due to quality concerns.
“For years, we assumed that certain processing parameters were ‘off-limits’ because they would result in poor-quality end products,” Croom said.
“But by using AI to explore the full range of possibilities, we discovered new processing regions that allow for faster printing while maintaining — or even improving — material strength and ductility, the ability to stretch or deform without breaking.”
The findings also align with broader efforts to advance additive manufacturing in aerospace and defense. Somnath Ghosh, a researcher at the Whiting School of Engineering, is working on AI-driven simulations to predict the performance of additively manufactured materials in extreme environments.
Ghosh co-leads the NASA Space Technology Research Institute, which aims to accelerate material qualification and certification, reducing the time required to develop materials for space applications.
APL has spent years refining additive manufacturing techniques, focusing on defect control and material performance.
The team previously developed a rapid material optimization framework, patented in 2020, which laid the foundation for this AI-driven study.
By using machine learning, the researchers were able to explore a vast range of processing parameters efficiently, revealing new high-density processing regimes that had been previously dismissed due to concerns about material instability.
Steve Storck, chief scientist for manufacturing technologies at APL, emphasized the significance of these findings.
“We’re not just making incremental improvements,” Storck said. “We’re finding entirely new ways to process these materials, unlocking capabilities that weren’t previously considered.”
Traditionally, identifying the best manufacturing settings required extensive trial-and-error testing. Instead, the team used Bayesian optimisation, a machine learning technique that predicts promising experimental conditions based on prior data.
This allowed them to rapidly home in on optimal processing settings, reducing the need for slow and costly physical testing.
Beyond titanium, the research has broader implications for manufacturing. The AI-driven approach is being applied to other metals and additive manufacturing techniques.
Future research will focus on real-time monitoring of the manufacturing process, potentially enabling automated adjustments to ensure high-quality production without extensive post-processing.
Storck envisions a future where advanced metal additive manufacturing operates with the efficiency of consumer 3D printing.
